Aspartame, Review of Safety
Aspartame, Review of Safety
W. Wayne Stargel
Research and Development, The NutraSweet Company, Mt. Prospect, Illinois
C. Phil Comer
Graystone Associates, Inc., Macon, Georgia
Dale A. Mayhew
Regulatory Affairs, The NutraSweet Company, Mt. Prospect, Illinois
1 To whom correspondence should be addressed at Medical and Scientific Affairs, The NutraSweet Company, 699 Wheeling Road, Mt.
S1 0273-2300/02 $35.00
C 2002 Elsevier Science (USA)
Donald L. Schomer (Neurochemistry, Seizures and EEGs, Behavior, Cognitive Function, and Mood)
Department of Neurology, Division of Neurophysiology and Epilepsy, Beth Israel Deaconess Medical Center,
Harvard Medical School, Boston, Massachusetts
Bennett A. Shaywitz (Behavior, Cognitive Function, Mood in Children, Seizures, and EEGs)
Departments of Pediatrics, Neurology, and Child Study, Yale University School of Medicine, New Haven, Connecticut
DEDICATION
The authors dedicate this supplement to the memories of Lewis D. Stegink, Ph.D., and L. J. Filer, Jr., M.D.,
Ph.D., from the University of Iowa. Their early research on aspartame metabolism in humans formed the basis for
much of the future research on aspartame that is discussed in this supplement. Their objectivity and long-standing
dedication to science as well as their medical and scientific expertise are greatly missed.
PREFACE
More than 30 years have elapsed since the foundations of todays aspartame safety database were laid. Since that
time the portfolio of studies assessing the safety of aspartame has continued to grow. A search of the scientific
literature on the U.S. National Library of Medicines MEDLINE reveals almost 700 citations for aspartame
with a number of these relevant to aspartame safety. The extensive body of research undertaken on aspartame
clearly and overwhelmingly demonstrates its safety for its intended use. The aspartame safety data have been
evaluated and found satisfactory by regulatory scientists in all major regulatory agencies and expert committees,
including the U.S. Food and Drug Administration (FDA), the EU Scientific Committee for Food (SCF), and the
Joint FAO/WHO Expert Committee on Food Additives (JECFA). Further, aspartame has been approved for human
consumption by regulatory agencies in more than 100 countries and received wide consumer acceptance with
consumption by hundreds of millions of people over the past 20 years, representing billions of man-years of safe
exposure.
Thus, one would ask, Why is there a need for a comprehensive review of aspartame safety data at this point in
time? Questions about its safety continue to be raised by a few individuals. Safety issues long ago resolved to the
satisfaction of regulatory agencies and expert committees are today resurrected by some as if new. Early on these
issues were based on scientific hypothesis, misinterpretation of data, and/or anecdotal reports of adverse health
effects. Recently spurious and imaginative, hypothetical questions that are lacking even anecdotal support or a
logical scientific rationale have compounded these issues. Given the extensive safety database and many years
of safe human exposure, the continuing debate about aspartame is most unusual.
Those who have followed the aspartame story are familiar with the issues that continue to be raised.
Aspartame is a simple molecule, which is hydrolyzed entirely to its constituent amino acids, aspartate and
phenylalanine, and methanol which are then absorbed. The constituents of aspartame are also derived in much
larger amounts from common foods. Are theoretical concerns for these components of any relevance to the small
amounts that are produced following aspartame consumption under intended conditions of use?
By what mechanism could the constituents of aspartame be handled differently in the body than are the
chemically identical constituents from common foods? Since aspartame is broken down upon ingestion, what is
the relevance of in vitro and parenteral studies that administer intact aspartame?
Following the marketing of aspartame and the rapid and widespread consumer acceptance of aspartame-
containing products, there were anecdotal reports of adverse health effects. Are these actually due to aspar-
tame consumption? Why does the incidence wax and wane precisely when the media focus on these anecdotal
reports?
Critics have argued that aspartame induces brain tumors in rats. This conclusion was based originally on
post hoc analyses of data combined from independent dose groups in animal carcinogenicity studies, but the
conclusions reached by the critics were rejected by regulatory agencies. Thirty years later, the same critics again
argued that aspartame induces brain tumors, on this occasion based on misinterpretation of epidemiology data;
again this argument was considered in detail by regulatory bodies around the world and dismissed. How is the
late-comer to distinguish credible concern from historical grievance?
To some extent, the current unwarranted concern about aspartame is fueled by the efforts of a small number of
anti-aspartame activists who often seek to inflame as much as to inform. These individuals invest considerable
personal time and energy in organizing and alerting others to what they believe to be the dangers of aspartame.
Such individuals may be perceived by the public and the non-specialist media as representing an independent
source of unbiased information, despite the fact that the conclusions they reach are not supported by the rigorous
safety evaluations undertaken by respected regulatory agencies and expert committees or by publications in the
peer-reviewed scientific literature.
One chain e-mail widely disseminated by lay anti-aspartame activists throughout the world purports to link
aspartame with a long list of maladies including headache, joint pain, depression, memory loss, tinnitus, vertigo,
blindness, carbohydrate cravings, Alzheimers disease, birth defects, brain cancer, diabetes, Gulf War syndrome,
lupus erythematosus, multiple sclerosis, and seizure disorders. Unfortunately, the scientific literacy of the public
is sometimes insufficient to differentiate Internet fiction from Internet fact. Well-meaning citizens repeatedly
have brought these findingsperpetrated only on the Internetto the attention of governments and regula-
tory agencies. Governments have to be responsive to such consumer concerns, despite the fact that continuing
reevaluations are not warranted by the scientific evidence.
Misinformation about aspartame abounds elsewhere on the Internet, from websites such as
www.aspartamekills.com to frenetic news group alerts flashed by anti-aspartame activists around the world.
These inaccuracies have spilled over into the popular press, have affected governments and regulators, have been
S3
S4 BUTCHKO ET AL.
published in scientific textbooks, have caused concern among practicing physicians, have required responses from
organizations of health professionals, and have even affected research funding by charities. An Australian daily
newspaper recently reported on aspartame detox centers in America as a serious news item. Todays consumers
frequently consult the Internet regarding a variety of health and nutritional issues; unfortunately, many users
regard online information as authoritative and valid without questioning its source. Inappropriate advice on the
Internet could pose a serious health danger to individuals needing medical treatment, because some activists
recommend extended periods of aspartame abstinence in lieu of seeking medical attention for what could be
symptoms of a serious and treatable illness. Whatever the goal of these individuals, it is not public health.
Considerable academic, regulatory, and industrial resources have gone into addressing issues raised regarding
aspartame prior to its approval and during the many reaffirmations of its safety. As the present review will attest,
considerable additional resources have been invested by many after the approval of aspartame. Readers should
be aware that regulatory agencies have stringent premarket requirements for safety assessment; aspartame met
and exceeded those requirements long ago and on numerous occasions since. However, the case of aspartame is
unusual in that conclusions reached by regulatory agencies and authorities about the different issues are not
accepted by some individuals, and the same allegations of adverse health consequences continue to be raised,
despite their resolution to the satisfaction of competent authorities.
The present review considers all new published studies, puts recent reports into the context of previous experi-
ence, and discusses key aspects such as study design and methods, route of administration, and reproducibility.
Anecdotal reports that aspartame has produced adverse health effects can only be considered by a rigorous evalu-
ation of all the evidence. For example, in the case of epilepsy, given the prevalence of the condition and the millions
of people who consume aspartame, it is inevitable that some individuals will experience their first seizure soon
after consuming aspartame; such a temporal relationship is not evidence of causation, but it will appear convinc-
ing to the individual concerned. Such anecdotal experiences could be an indication of a causeeffect relationship,
and therefore all those interested in ensuring public health, including the present reviewers, must consider the
totality of the evidence that either supports or refutes such a relationship.
There is a new generation of scientists, regulators, and consumers who are factually unfamiliar with thirty
years of aspartame research. The size and complexity of published literature documenting aspartames safety
make issues and the resolution of questions difficult for those who have not followed the particulars from the
beginning. Consequently, many today are ill-equipped to draw their own informed conclusions regarding the
safety of aspartame. Thus, this comprehensive review of aspartames safety is both timely and warranted.
In this supplement, a group of esteemed scientists has accomplished the feat of compiling a comprehensive
review of the extensive scientific literature published worldwide on aspartame safety. The reviewers have at-
tempted to place more recent research into the continuum of earlier investigations. They have organized their
reviews to address legitimate issues as well as theoretical concerns of critics.
The extensive list of contributing authors includes several scientists employed by the main producer of aspar-
tame, and therefore activists may dismiss the findings here as biased. However, considering the credentials and
bibliographies of all the contributing authors, and the detailed analysis of all relevant published data, it is clear
that the reader can have confidence in using this review as an authoritative evaluation of the scientific literature.
It is the intention of all the authors to present the scientific facts clearly and accurately so that the informed
reader can decide whether there is a basis for continuing the debate on aspartame safety. Articles relevant to
aspartame safety have been identified through ongoing searches of electronic databases such as Medline, Chem-
ical Abstracts, Science Citation Index, Biosis, Food Science and Technology Abstracts, Agricola, and Toxline. The
articles referenced in this review represent those identified from the 1970s into 2002.
The conclusion of this up-to-date review is in agreement with that of international regulatory agencies and
expert committees: the weight of scientific evidence is clear that aspartame is safe for its intended use. The one
exception is patients with the rare genetic disease, phenylketonuria, which is diagnosed at birth in the United
States, Canada, Europe, and some other countries. These individuals are treated with low phenylalanine diets
for a period of years, and while on this diet, they are required to monitor their phenylalanine intake from all
dietary sources, including aspartame.
After 30 plus years of rigorous scientific research, it is time to put questions of aspartame safety to rest. It
is difficult to identify any dietary constituent that has been more thoroughly evaluated than aspartame. The
breadth and depth of scientific data available on aspartame and reviewed here are unlikely to exist for any other
food additive. The continuing debate over such a nonissue only serves to divert attention and the allocation of
resources from more important health issues that need to be addressed.
Ian C. Munro
Andrew G. Renwick
Introduction to Aspartame: Review of Safety
Further, the components of aspartame are derived
Over 20 years have elapsed since aspartame was
approved by regulatory agencies as a sweetener and
from these common foods in much larger amounts. For
flavor enhancer. The safety of aspartame and its example, a glass of no-fat milk provides about 6 times
metabolic constituents was established through exten- more phenylalanine and 13 times more aspartic acid
sive toxicology studies in laboratory animals, using and a glass of tomato juice provides about 6 times more
much greater doses than people could possibly con- methanol than an equivalent volume of beverage sweet-
sume. Its safety was further confirmed through studies ened 100% with aspartame. Thus, much of the scientific
in several human subpopulations, including healthy research, both before and after regulatory approval, has
infants, children, adolescents, and adults; obese in- focused on the safety of these components.
dividuals; diabetics; lactating women; and individ- Prior to marketing, the safety of aspartame was
uals heterozygous (PKUH) for the genetic disease firmly established through numerous studies using in-
phenylketonuria (PKU) who have a decreased abil- ternationally accepted models for safety assessment.
ity to metabolize the essential amino acid, phenylala- Extensive toxicologic and pharmacologic research was
nine. Several scientific issues continued to be raised done in laboratory animals using much greater doses
after approval, largely as a concern for theoretical of aspartame than people could possibly consume. In
toxicity from its metabolic componentsthe amino addition, the safety of aspartame and its metabolic con-
acids, aspartate and phenylalanine, and methanol stituents was further confirmed in several human sub-
even though dietary exposure to these components is populations: healthy infants, children, adolescents, and
much greater than from aspartame. Nonetheless, addi- adults; obese individuals; diabetics; lactating women;
tional research, including evaluations of possible asso- and individuals heterozygous (PKUH) for the genetic
ciations between aspartame and headaches, seizures,
disease phenylketonuria (PKU) who have a decreased
behavior, cognition, and mood as well as allergic-type
ability to metabolize the essential amino acid pheny-
reactions and use by potentially sensitive subpopu-
lalanine. The results of the animal and human studies
lations, has continued after approval. These findings
are reviewed here. The safety testing of aspartame has
provided convincing evidence that aspartame consump-
gone well beyond that required to evaluate the safety tion is safe, including by pregnant women and children.
of a food additive. When all the research on aspartame, As a result, regulatory agencies in over 100 countries
including evaluations in both the premarketing and approved aspartame for its intended use as a sweetener.
postmarketing periods, is examined as a whole, it is In the postmarketing period, consumption studies
clear that aspartame is safe, and there are no unre- demonstrated that actual aspartame intake at the 90th
solved questions regarding its safety under conditions percentile was only 510% of the acceptable daily in-
of intended use. C 2002 Elsevier Science (USA) take, thus providing small amounts of aspartate, pheny-
lalanine, and methanol to the diet. Nonetheless, several
scientific issues were raised and further evaluated in
Since regulatory approval about 20 years ago, the the postmarketing period, largely as a result of asser-
high-intensity sweetener aspartame (L--aspartyl-L- tions of potential toxicity of its three metabolic compo-
phenylalanine methyl ester) is consumed in more than nents when given in extremely high doses. These issues
6000 products by hundreds of millions of people around centered on potential excitotoxicity of aspartic acid, po-
the world who want to enjoy the sweet taste of sugar tential effects of phenylalanine on brain function, and
without all the calories. The introduction of aspartame toxicity of methanols formate metabolite. The results
changed forever the quality of life of diabetic individ- of the research demonstrated that it is not possible for
uals by allowing them to enjoy good-tasting foods and a human to ingest enough aspartame to raise plasma
beverages and still comply with dietary requirements. concentrations of its metabolic constituents to those as-
In addition, aspartame-containing products have been sociated with adverse effects.
shown to be useful as part of a comprehensive multidis- In response to anecdotal reports of adverse health
ciplinary program to promote weight loss and control of effects from aspartame from some consumers, a post-
body weight in obese individuals. marketing surveillance program was implemented to
Aspartame is unique among high-intensity sweeten- document and evaluate these reports. Although such re-
ers because it is metabolized by digestive esterases and ports cannot establish a cause-and-effect relationship,
peptidases to three common dietary componentsthe the information was used to guide additional research
amino acids, aspartic acid and phenylalanine, and a efforts, including evaluations of possible associations
small amount of methanol. These components are used between aspartame and headaches, seizures, behavior,
by the body in the same ways as when they are also cognition, and mood, as well as allergic-type reactions.
derived from foods, such as meat, milk, fruits, and veg- In addition, studies were done to evaluate safety of
etables. aspartame use by potentially sensitive subpopulations
S5
S6 BUTCHKO ET AL.
(i.e., individuals heterozygous for the rare genetic dis- aspartame, NutraSweet, aspartic acid, phenylalanine,
ease phenylketonuria and individuals with Parkinsons methanol, and diketopiperazine. Research by scientists
disease, dizziness, depression, liver disease, and renal who have been involved with aspartame is also fre-
disease). Further evaluations of aspartame and the en- quently monitored. For the most part, the referenced lit-
docrine system and usefulness in weight control were erature represents articles published in peer-reviewed
completed. In addition, an allegation that aspartame journals; however, a few abstracts and book chapters
may be associated with brain tumors in humans was were included when the information had not been pub-
raised by a long time aspartame critic almost 20 years lished in peer-reviewed journals. The articles refer-
after he first raised the same issue in animals. This alle- enced in this review represent those identified from the
gation was evaluated by scientists and regulators with 1970s into 2002.
the conclusion that the claims were not valid. The testing of aspartame has been far beyond the
The extensive literature included in this review rep- standard safety testing required to evaluate the safety
resents over 30 years of research on aspartame safety. of a food additive. When all the research on aspartame,
Articles relevant to aspartame safety were identi- including evaluations in both the premarketing and
fied through ongoing searches of electronic databases postmarketing periods, is examined as a whole, it is
such as Medline, Chemical Abstracts, Science Cita- clear that aspartame is safe, and there are no unre-
tion Index, Biosis, Food Science and Technology Ab- solved questions regarding its safety under conditions
stracts, Agricola, and Toxline. Search terms included of intended use.
Preclinical Safety Evaluation of Aspartame
Introduction established a no-observed-effect level of 2000 mg/kg/day
for the aspartame preclinical studies (FDA, 1974) based
The original data used to support aspartame safety upon small changes in body weight in 4000 mg/kg body
consisted of a comprehensive battery of studies in ani- wt/day dose groups that other regulatory bodies ac-
mals on acute, subchronic, and long-term toxicity, car- cepted as being secondary to decreases in food con-
cinogenicity, genetic toxicity, and reproductive toxicity sumption. The FDA (1974) set an ADI of 20 mg/kg body
and teratogenicity. In addition, there was a study to wt but subsequently raised it to 50 mg/kg body wt based
evaluate postnatal developmental effects in infant pri- on additional data from human studies (FDA, 1984).
mates as well as studies of effects on the central ner- In addition to aspartame, the safety of its conver-
vous system, gastrointestinal tract, endocrine system, sion product, aspartylphenylalanine diketopiperazine
and reproductive system (Molinary, 1984; Kotsonis and (DKP), has been assessed in safety studies. DKP may
Hjelle, 1996). The definitive preclinical studies done form over time under conditions of extremes of pH and
with aspartame are listed in Table 1. These studies temperature, particularly in liquid matrices. The defini-
demonstrated that aspartame is not toxic, carcinogenic, tive preclinical safety studies done with DKP are listed
mutagenic, or teratogenic and has no effects on repro- in Table 2. JECFA established a no-observed-effect level
duction. Based on the results of these studies, a no- of 750 mg/kg body wt/day for DKP from preclinical
observed-effect level of at least 4000 mg/kg body wt was safety studies noting apparent findings in higher dose
established by the Joint FAO/WHO Expert Committee groups of one long-term rat study (JECFA, 1980). The
on Food Additives (1980), the Scientific Committee for U.S. FDA considered these observations incidental in
Food (SCF) (1985), and the Health Protection Branch older rats and established a no-observed-effect level of
of Health and Welfare Canada (1979). As a result, an 3000 mg/kg body wt/day for DKP; FDA also evaluated
acceptable daily intake (ADI) of 40 mg/kg body wt was additional data that became available after JECFAs de-
set by those agencies based on the no-observed-effect liberations (FDA, 1983). Thus, JECFA (1980) and the
level of 4000 mg/kg in the animal studies. The U.S. FDA SCF (1985) set the ADI for DKP at 7.5 mg/kg body wt;
TABLE 1
Summary of Definitive Preclinical Toxicology Studies with Aspartamea
Maximum dose
Study type Specific studies Species Outcome (mg/kg body wt)
S7
S8 BUTCHKO ET AL.
TABLE 2
Summary of Definitive Preclinical Toxicology Studies with Aspartylphenylalanine Diketopiperazine (DKP)a
Maximum dose
Study type Specific studies Species Outcome (mg/kg body wt)
U.S. FDA (1983) set the ADI for DKP at 30 mg/kg body aspartame and brain tumors; however, in a post hoc
wt. Actual exposure to DKP in the United States at the analysis, data from independent dose groups were com-
mean and 90th percentile daily intakes have been esti- bined to imply a doseresponse relationship. Specifi-
mated to be less than 0.25 and 0.56 mg/kg body wt/day, cally, data from the two different lowest dose groups
respectively (Kotsonis and Hjelle, 1996), well below the were combined and contrasted with the combined data
ADIs set by FDA, SCF, and JECFA. from the two different highest dose groups to allege an
It is not unusual for regulatory bodies to establish dif- apparent dose response. Clearly, manipulations of data
ferent no-observed-effect levels for the same preclinical and dose groups in this manner are scientifically and
safety data based on the different policies and practices statistically meaningless according to internationally
of individual agencies. Consequently, regulatory bod- accepted standards. Further, the assertion was made
ies may apply different ADIs to compounds. For exam- that another long-term bioassay in rats was invalid be-
ple, in the absence of toxicity, it is the normal practice cause the incidence of brain tumors in controls exceeded
of JECFA to recognize body weight changes due to re- the evaluators expectations. In actuality, no carcino-
duced food consumption as not adverse when establish- genicity study with aspartame or DKP (Tables 1 and
ing a no-observed-effect level; the FDA generally con- 2) reveals a doseresponse relationship between as-
siders these changes in body weight when establishing partame and brain tumors. Regardless of these long-
a no-observed-effect level. For example, in the case of lived claims, scientists in regulatory bodies and ex-
sucralose, JECFA considered reduced body weight gain pert committees around the world, including the U.S.
at 1500 mg/kg as being due to reduced palatability of FDA (1981), the Canadian Health Protection Branch
diet and established an ADI of 015 mg/kg body wt/day (Health and Welfare Canada, 1981), the UK Committee
based on a no-observed-effect level at this dose (JECFA, on Toxicity (MAFF, 1982), the EU Scientific Committee
1991). In contrast, FDA considered body weight changes for Food (1985) and the Joint FAO/WHO Expert Com-
and established a no-observed-effect level of 500 mg/kg mittee on Food Additives (1980), which have evaluated
body wt in rats and, consequently, an ADI for sucralose the preclinical carcinogenicity data according to stan-
of 5 mg/kg body wt/day (FDA, 1998). dard and internationally accepted criteria, have con-
Notwithstanding the findings of national and inter- cluded that aspartame is not a carcinogen. For a com-
national regulatory bodies that aspartame is not car- plete discussion of the brain tumor issue see Evaluation
cinogenic in animals, critics continue to cite a flawed of Aspartame and Brain Tumors below.
and discredited analysis by Olney from the 1970s alleg- The large body of preclinical safety data that has
ing that aspartame is associated with brain tumors in been reviewed by regulatory bodies worldwide and es-
rats. Specifically, data from a long-term carcinogenicity tablishes the safety of aspartame has been summa-
study showed no doseresponse relationship between rized elsewhere (Molinary, 1984; Ishii, 1984; Kotsonis
ASPARTAME: REVIEW OF SAFETY S9
and Hjelle, 1996). Several studies in additional mod- mosomal aberration count in bone marrow cells. As-
els have since become available which further expand partame alone appeared to have antimutagenic prop-
the toxicology database for aspartame. These studies erties when injected 5 days before exposure to the mu-
discussed below include evaluations of genetic toxicity, tagen, whereas joint administration of aspartame did
teratogenicity, and developmental milestones and be- not change clastogenic activity of the compounds. More
havioral development. recently, Mukhopadhyay et al. (2000) evaluated the ef-
fect of blends of aspartame (3.5, 35, 350 mg/kg body wt)
and acesulfame-k (1.5, 15, 150 mg/kg body wt) by gav-
Additional Preclinical Safety Evaluations
age in mice. There was no genotoxic effect as assessed
with Aspartame
by the presence of chromosomal aberrations in femoral
Genetic toxicology. Tests done before regulatory ap- bone marrow.
proval demonstrated that aspartame was not genotoxic
in the dominant lethal mutation assay in rats, the host- Teratogenicity. Bantle et al. (1990) evaluated five
mediated assay in rats and mice, the in vivo cytogenet- compounds, including aspartame, in the frog embryo
ics assay in rats, and the Ames test (Kotsonis and Hjelle, teratogenesis assay (FETAX), a screening test devel-
1996). Several additional studies have since been re- oped as an alternative to mammalian teratogenesis.
ported which further confirm the results of the original FETAX reportedly has a high degree of developmen-
studies that aspartame is not genotoxic. tal relevance to mammals. For each compound, 816
Karikas et al. (1998) measured direct molecular inter- concentrations were used in duplicate, and testing was
action of aspartame and its metabolites with DNA as an done in a blinded manner except for one colored com-
indicator of possible carcinogenic potential in an in vitro pound. The embryos were exposed to the compounds for
model. They observed measurable, dose-related molec- 96 h. Malformations and embryo growth were evalu-
ular interaction with DNA by aspartame, as well as its ated. Based on the results of the study, the authors con-
amino acid components aspartic acid and phenylala- cluded that aspartame does not pose a teratogenic haz-
nine. This was further verified in another article pub- ard. These results are consistent with those from the
lished by the same group (Schulpis et al., 1998). What is animal teratogenicity studies done preapproval with as-
being measured in this study is actually the interaction partame (Kotsonis and Hjelle, 1996).
with DNA by amino groups characteristic of amino acids
and is not an effect specific to aspartame. Further, as- Developmental milestones and behavioral develop-
partame is metabolized into its component amino acids ment. Brunner et al. (1979) used pregnant rats to eval-
and methanol which are then absorbed; i.e., aspartame uate three doses of aspartame in the diet (2, 4, and 6%,
itself is not absorbed (Ranney and Oppermann, 1979; equivalent to doses in pups of up to 3000, 6000, and
Burgert et al., 1991). Thus, aspartame is not avail- 9000 mg/kg body wt/day) compared to a negative con-
able to interact with DNA in vivo. In another study, trol as well as a positive control (hydroxyurea). In addi-
Jeffrey and Williams (2000) tested the DNA-damaging tion there was a group fed phenylalanine at 3% (up to
activity of several sweeteners in a rat hepatocyte/ 5000 mg/kg body wt/day) of the diet. Thus, pups were
DNA repair assay. None of the sweeteners, including exposed from conception through weaning to 90 days of
aspartame, showed an increase in NNG (net nuclear postnatal life. At aspartame doses to pups of 9000 mg/kg
grain) counts; thus aspartame had no effect on DNA. body wt/day or phenylalanine doses up to 5000 mg/kg
In an in vitro study, Kasamaki and Urasawa (1993) body wt/day in the diet, deficits were seen in prewean-
evaluated the ability of various food chemicals to induce ling rats in time to eye opening, surface righting, lo-
cell aging in human diploid fibroblast cell cultures. Us- comotion (decreased), swimming, startle response, and
ing known genotoxic materials (e.g., aflatoxin B1 ), the open-field activity. Dramatically fewer effects were seen
cumulative cell population doubling (CPD) was reduced at the 6000 mg/kg body wt/day dose, and no effects were
by 812 CPDs, whereas nongenotoxic compounds, such seen at the 3000 mg/kg body wt/day dose. Effects seen
as aspartame, slightly shortened life span (26 CPDs) at the highest dose of aspartame were essentially the
compared to untreated cells. However, the relevance of same as those observed with the phenylalanine dose. As
this research to human exposure is questionable, as as- reported in this study, pups received much higher doses
partame is not absorbed into the systemic circulation. than dams during the latter stages of lactation when
Several studies have been done to evaluate potential pups began to eat solid food and consume the dams
clastogenic activity via chromosomal aberration evalu- aspartame-containing diets. For example, the young
ations. Durnev et al. (1995) gave daily aspartame doses adult dams consuming 2, 4, and 6% aspartame in diet
of 40 and 400 mg/kg body wt by gavage over 5 days would be expected to be dosed with approximately 2000,
to mice. There was no clastogenic activity after aspar- 4000, and 6000 mg/kg body wt/day at these concentra-
tame. Kulakova et al. (1999) evaluated the effect of tions, respectively. Thus, the extremely high doses ad-
aspartame (0.4 to 40 mg/kg body wt) on the cytogenic ministered to pups during this study were sufficiently
effects of dioxydin and cyclophosphan in mice with chro- large to result in potential nutritional deficits in these
S10 BUTCHKO ET AL.
young, growing animals, which may account for any ob- averaged 835 mg/kg body wt/day. After weaning, pups
served effects. These extremely large doses in pups are consumed an average of 32, 154, 836, and 3566 mg/kg
not relevant to human exposure to aspartame. body wt/day aspartame, and those on phenylalanine av-
In a study comparing the effect of aspartame to eraged 1795 mg/kg body wt/day. No effects of aspartame
saline-treated and untreated controls on development were observed on reflex development or morphologi-
in mice, Mahalik and Gautieri (1984) administered cal development. In addition, contrary to the results
doses by gavage to gravid mice on days 1518 of ges- of Yirmiya et al. (1989), there were no effects of aspar-
tation. Two doses of aspartame (1000 and 4000 mg/kg tame compared to water on spatial memory as assessed
body wt) were evaluated. In the neonatal mice, there in the radial arm maze. Evaluations included the num-
were no differences between either of the aspartame ber of arms chosen before one was reentered, overall
groups and historical controls in three of four parame- time, and time in the arms. In addition, no differences
ters tested, including negative geotaxis or surface and were found between groups when pups were tested in
air righting. However, using the rope method, achieve- the additional learning challenge of the milk maze.
ment of the visual placing response was significantly In a testing paradigm in guinea pigs, Dow-Edwards
delayed in both aspartame groups compared to his- et al. (1989) evaluated the effect of prenatal expo-
torical control values and in the high-dose aspartame sure (from conception to parturition) to aspartame
group compared to the low-dose group. Concurrent con- (500 mg/kg body wt/day) in sesame oil compared to
trols were not used. McAnulty et al. (1989) tested the sesame oil alone and an untreated control group on
same strain of mice with maternal dosages of 500, 1000, odor-aversion testing. There were no significant effects
2000, and 4000 mg/kg body wt aspartame by gavage on on maternal weight gain, litter size, or birth weight of
days 1518 of gestation. In contrast to the findings of pups. At day 15 of age, pups were injected with lithium
Mahalik and Gautieri, McAnulty et al. found no effect chloride (LiCl) or saline and exposed to vanilla odor for
on the visual placing response when using concurrent 30 min to create an aversion to vanilla odor. Twenty-
controls. Specifically, there were no effects of any dose of four hours later, pups were permitted to choose between
aspartame on maternal body weight, food consumption, vanilla and lemon odors. The control and untreated
length of gestation, reproduction indices, or litter size. groups showed the expected conditioned aversion to
In the pups, birth weight, negative geotaxis, and surface vanilla; however, pups in the aspartame group did
and midair righting reflexes were not affected. Neither not. The authors concluded that aspartame exposure
was there any delay in development of visual placing, throughout gestation disrupts odor-associative learn-
regardless of the method used for assessment (grid or ing in 15-day-old guinea pigs. However, the validity of
rope) or the manner in which data were analyzed af- this model was questioned by a report from McAnulty
ter any dose of aspartame compared to concurrent con- et al. (1992), who did a feasibility study using sesame
trols. In addition, there were no differences between oil and untreated groups in anticipation of investi-
groups in eye opening, reflex pupil closure, and ophthal- gating the effects described by Dow-Edwards et al.
mologic examinations. Thus, the results of Mahalik and (1989). They could not replicate either the conditioned
Gautieri were not reproducible. odor aversion or the reported natural preference for
Yirmiya et al. (1989) evaluated the effect of aspar- vanilla over lemon odors in guinea pigs. As the group
tame given in drinking water to dams starting 30 days mean scores for vanilla preference were in the range
prior to conception until rat pups were 30 days old. of 50 15% after saline or LiCl injections in either of
The average daily dose of aspartame in water was the groups, there was no preference; pups sometimes
1188 mg/kg body wt. Compared to rats in the plain showed preference for vanilla or lemon, but generally
water group, there were no differences in morpholog- pups tended to remain in the first chamber entered re-
ical development and reflexes in pups exposed to aspar- gardless of the odor cue present. However, conditioned
tame. At 30 days of age, there was a difference in perfor- avoidance was readily demonstrated in an avoidance
mance in the radial arm maze in the aspartame-treated paradigm using electric shock, thus establishing that
group. The authors observed that the number of entries learning ability can be measured in 15-day-old guinea
prior to repeat were greater for aspartame-exposed rats pigs. These results indicate that conditioned odor aver-
than controls. They suggested that aspartame may ei- sion in guinea pigs is not reproducible, and, thus, the
ther facilitate the development of components involved methodology could not be used to investigate an odor
in this procedure or reduce factors that may interfere aversion effect, such as that attributed to aspartame
with this task. Holder (1989) reported another study to by Dow-Edwards and co-workers.
evaluate aspartame and development, including spa-
tial memory in a radial arm maze. Adult female rats
Conclusion
were exposed to aspartame or phenylalanine in drink-
ing water for 12 days prior to mating until the pups Since regulatory approvals of aspartame in the early
were 38 days old. Doses of aspartame were 14, 68, 347, 1980s, there have been a few additional animal safety
and 1614 mg/kg body wt/day, and phenylalanine dosing studies completed to evaluate further the safety of
ASPARTAME: REVIEW OF SAFETY S11
aspartame. In no cases have subsequent studies al- performance on odor-aversion testing following prenatal aspartame
tered original determinations of safety established in exposure in the guinea pig. Neurotoxicol. Teratol. 11, 413416.
the complete battery of preclinical studies evaluated by Durnev, A. D., Oreshchenko, A. V., Kulakova, A. V., Beresten, N. F.,
and Seredenin, S. B. (1995). Clastogenic activity of dietary sugar
regulatory agencies as part of the approval process. One substitutes. Vopr. Med. Khim. 41, 3133.
study suggested an effect of aspartame on DNA. How- Food and Drug Administration (FDA) (1974). Food additives permit-
ever, other studies by the same investigators demon- ted in food for human consumption: Aspartame. Fed. Regist. 39,
strate the effect to be the result of an interaction with 2731727320.
DNA by the amino group of amino acids in general and Food and Drug Administration (FDA) (1981). Aspartame: Commis-
not an interaction specific to aspartame. Other inves- sioners final decision. Fed. Regist. 46, 3828538308. July 24, 1981.
tigators using another model assessing DNA damage Food and Drug Administration (FDA) (1983). Food additives permit-
ted for direct addition to food for human consumption: Aspartame.
found no effect of aspartame. One in vitro study sug-
Fed. Regist. 48, 3137631382.
gested an effect of a number of food ingredients, includ- Food and Drug Administration (FDA) (1984). Food additives permit-
ing aspartame, on cell aging. Again, the lack of speci- ted for direct addition to food for human consumption: aspartame.
ficity of the response and the lack of applicability of Fed. Regist. 49, 66726677.
these data to the in vivo situation eliminate the rel- Food and Drug Administration (FDA) (1998). Food additives permit-
evance of these data to the safety of aspartames use ted for direct addition to food for human consumption: Sucralose.
in food. Several studies confirmed the results of earlier Fed. Regist. 63, 1641716433.
studies demonstrating the lack of genotoxicity of aspar- Health and Welfare Canada (1979). Proposal on Aspartame, Informa-
tion Letter No. 564. Health Protection Branch, Ottawa.
tame. Confirming the results of the preapproval safety
Health and Welfare Canada (1981). Aspartame. Information Letter
studies, aspartame was also shown not to pose a risk of No. 602. Health Protection Branch, Ottawa. July 31, 1981.
teratogenicity in an in vitro screening model. Holder, M. D. (1989). Effects of perinatal exposure to aspartame on
In a study assessing developmental milestones and rat pups. Neurotoxicol. Teratol. 11, 16.
behavioral development after extremely large doses of Ishii, H. (1984). Chronic feeding studies with aspartame and its dike-
aspartame, dose-related effects were seen but only at topiperazine. In Aspartame: Physiology and Biochemistry (L. D.
doses sufficiently large as to cause nutritional deficits. Stegink and L. J. Filer, Jr., Eds.), pp. 307319. Marcel Dekker, New
York.
However, the same effects were also seen at an equiv-
Joint FAO/WHO Expert Committee on Food Additives (JECFA)
alent dose of phenylalanine demonstrating the effects
(1980). Aspartame: Toxicological Evaluation of Certain Food Addi-
were not specific to aspartame. Two studies appeared tives, WHO Food Additive Series No. 15, pp. 1886. JECFA, Geneva.
to indicate either an effect of aspartame on develop- Joint FAO/WHO Expert Committee on Food Additives (JECFA)
ment of the visual placing response or a disruption of (1991). Trichlorogalactosucrose: Toxicological Evaluation of Cer-
odor-associative learning. However, reported effects of tain Food Additives and Contaminants, WHO Food Additive Series
aspartame on visual placing were not reproducible in an No. 28, pp. 219228. JECFA, Geneva.
enhanced study using a concurrent control group by an- Jeffrey, A. M., and Williams, G. M. (2000). Lack of DNA-damaging
activity of five non-nutritive sweeteners in the rat hepatocyte/DNA
other laboratory, further the model of odor-associative repair assay. Food Chem. Toxicol. 38, 335338.
learning also was not reproducible by others. In addi- Karikas, G. A., Schulpis, K. H., Reclos, G., and Kokotos, G. (1998).
tion, reported effects of aspartame on performance in a Measurement of molecular interaction of aspartame and its
radial arm maze were not reproducible. metabolites with DNA. Clin. Biochem. 31, 405407.
In conclusion, an extensive body of preclinical safety Kasamaki, A., and Urasawa, S. (1993). The effect of food chemicals
data demonstrates the safety of aspartame and its hy- on cell aging of human diploid cells in in vitro culture. J. Toxicol.
Sci. 18, 143153.
drolytic cyclization product, DKP. This body of work
Kotsonis, F. N., and Hjelle, J, J. (1996). The safety assessment of
demonstrates that aspartame and DKP do not produce
aspartame: Scientific and regulatory considerations. In The Clin-
adverse effects, even at doses several orders of magni- ical Evaluation of a Food Additive: Assessment of Aspartame
tude greater than human consumption. (C. Tschanz, H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.),
pp. 2341. CRC Press, Boca Raton, FL.
Kulakova, A. V., Belogolovskaya, E. G., Oreshchenko, A. V., Durnev,
REFERENCES A. D., and Seredenin, S. B. (1999). The antimutagenic activity of
aspartame. Eksp. Klin. Farmakol. 62, 4850.
Bantle, J. A., Fort, D. J., Rayburn, J. R., DeYoung, D. J., and Bush, Mahalik, M. P., and Gautieri, R. F. (1984). Reflex responsiveness of
S. J. (1990). Further validation of FETAX: Evaluation of the devel- CF-1 mouse neonates following maternal aspartame exposure. Res.
opmental toxicity of five known mammalian teratogens and non- Commun. Psychol. Psychiatry Behav. 9, 385403.
teratogens. Drug Chem. Toxicol. 13, 267282. McAnulty, P. A., Collier, M. J., Comer, C. P., and Allen, J. L. (1992).
Brunner, R. L., Vorhees, C. V., Kinney, L., and Butcher, R. E. (1979). The Odour Aversion Technique Claimed to Demonstrate Effects of
Aspartame: assessment of development psychotoxicity in a new ar- Aspartame in Guinea Pigs is Non-reproducible, Presented at the
tificial sweetener. Neurobehav. Toxicol. 1, 7986. 32nd annual meeting of the Teratology Society, Boca Raton, FL,
Burgert, S. L., Andersen, D. W., Stegink, L. D., Takeuchi, H., June 1992.
and Schedl, H. P. (1991). Metabolism of aspartame and its L- McAnulty, P. A., Collier, M. J., Enticott, J. M., Tesh, J. M., Mayhew,
phenylalanine methyl ester decomposition product by the porcine D. A., Comer, C. P., Hjelle, J. J., and Kotsonis, F. N. (1989). Absence of
gut. Metabolism 40, 612618. developmental effects of CF-1 mice exposed to aspartame in utero.
Dow-Edwards, D. L., Scribani, L. A., and Riley, E. P. (1989). Impaired Fundam. Appl. Toxicol. 13, 296302.
S12 BUTCHKO ET AL.
Ministry of Agriculture, Fisheries, and Food (MAFF) (1982). Food of the aspartyl moiety of aspartame in experimental animals and
Additives and Contaminants Committee Report on the Review of man. J. Environ. Pathol. Toxicol. 2, 979985.
Sweeteners in Food. Committee on Toxicity of Chemicals in Food, Schulpis, K. H., Karikas, G. A., and Kokotos, G. (1998). Direct molec-
Consumer Products and the Environment. FAC/REP/34:HMSO ular interaction of phenylalanine with DNA measured by reversed-
1982. United Kingdom. phase HPLC. Clin. Chem. 44, 178179.
Molinary, S. V. (1984). Preclinical studies of aspartame in nonprimate Scientific Committee for Food (SCF) (1985). Food-Science and Tech-
animals. In Aspartame: Physiology and Biochemistry (L. D. Stegink niques, Reports of the Scientific Committee for Food, sixteenth se-
and L. J. Filer, Jr., Eds.), pp. 289306. Marcel Dekker, New York. ries. Commission of the European Communities, Luxembourg.
Mukhopadhyay, M., Mukherjee, A., and Chakrabarti, J. (2000). Yirmiya, R., Levin, E. D., Chapman, C. D., and Garcia, J. (1989).
In vivo cytogenetic studies on blends of aspartame and acesulfame- Morphological and behavioral effects of perinatal exposure to as-
K. Food Chem. Toxicol. 38, 7577. partame (Nutrasweet) on rat pups. Bull. Psychon. Soc. 27, 153
Ranney, R. E., and Oppermann, J. A. (1979). A review of metabolism 156.
Intake of Aspartame vs the Acceptable Daily Intake
Introduction the FDA has concluded that the 99th percentile is un-
duly conservative and unreliable (FDA, 1986) because
Prior to regulatory approval, extensive toxicologic
the very small number of consumers in the 99th per-
and pharmacologic research was done in laboratory an-
centile may have large and variable intakes from day to
imals using much greater doses of aspartame than peo-
day, which may markedly skew the data. Thus, the FDA
ple could possibly consume. From the results of the
currently uses projections at the 90th percentile as
toxicology studies, a no-observed-effect level of at least
the benchmark of high-level consumers. In the United
4000 mg/kg body wt was established for aspartame by
Kingdom, the more conservative 97.5th percentile is
the Scientific Committee for Food (SCF) (1985), the
used by regulators as representative of intake by high-
Canadian Health Protection Branch (Health and Wel-
level consumers (MAFF, 1990).
fare Canada, 1979), and the Joint FAO/WHO Expert
Committee on Food Additives (JECFA, 1980). Based on
Aspartame Intake in the United States
these assessments, an acceptable daily intake (ADI) of
40 mg/kg body wt was established for aspartame. When Aspartame was approved by FDA for use in dry appli-
aspartame was first approved in the United States, the cations (e.g., tabletop, gelatins) in 1981 (FDA) followed
FDA established an ADI of 20 mg/kg body wt for aspar- by approval for use in carbonated soft drinks in 1983
tame (FDA, 1974). (FDA). By 1984 about 30% of the U.S. population was
In addition to the preapproval animal studies, the estimated to consume aspartame, and MRCA Informa-
safety of aspartame and its metabolic constituents tion Services (Northbrook, IL) began monitoring as-
was assessed in humans in several subgroups: healthy partame consumption in the U.S. population (Abrams,
infants, children, adolescents, and adults; obese indi- 1986, 1992; Butchko and Kotsonis, 1989, 1991, 1994,
viduals; diabetics; lactating women; and individuals 1996; Butchko et al., 1994, 2001; Butchko and Stargel,
heterozygous (PKUH) for the genetic disease phenylke- 2001, 2002). Intake was monitored from 1984 to 1992
tonuria (PKU) who have a decreased ability to me- via detailed menu census surveys from over 2000 house-
tabolize the essential amino acid, phenylalanine. The holds a year. During the 14-day survey, all foods eaten
results of the human studies, along with the animal re- both at home and away from home were recorded.
search, provided convincing evidence that aspartame Data were recorded by age group: 023 months, 2
was safe for general use, including by pregnant women 5 years, 612 years, 1317 years, and 18 years and over,
and children. The FDA responded to these additional as well as all age groups together. Intake by children
data by increasing the ADI for aspartame to 50 mg/kg was of special interest because of their smaller body
body wt in 1983 (FDA, 1984). The ADIs for aspartame in weights as they may consume more of an additive on a
the European Union and United States are the sweet- milligram per kilogram basis than adults. In addition,
ness equivalent of a 60 kg person consuming approxi- intakes by special subpopulations, such as diabetics
mately 500600 g of sugar daily over a lifetime, obvi- and people on weight-reduction programs, who might
ously an amount well above consumption patterns for be enthusiastic users of aspartame with potentially
sugar. higher intakes, and women of childbearing potential
As part of the safety evaluation for a food additive, and pregnant women, were also monitored.
regulators evaluate projected use levels relative to the Because aspartame is so intensely sweet, only small
ADI. If projected intake levels approach or exceed the amounts are needed to sweeten foods (Table 1). Thus,
ADI, restrictions may be imposed, such as limiting ap- it would be expected that average daily intake of as-
provals for some categories of use to decrease potential partame would be low. From the MRCA survey, the
exposure in the general population. Before approval,
projected average intake levels of aspartame in the
United States ranged from 8.3 mg/kg body wt/day if all TABLE 1
sucrose in an average-sized persons diet was replaced Approximate Aspartame Content of Some Common
by aspartame to 25 mg/kg body wt/day if all dietary car- Foods and Beveragesa
bohydrate could be replaced by aspartame. Based on
dietary records from about 12,000 individuals, it was Aspartame
Food Serving size content (mg)
estimated that, if all possible foods were replaced with
aspartame-containing foods, the 99th percentile daily Beverage 12 fluid oz. (355 ml) 180
consumption of aspartame would be 34 mg/kg body wt Yogurt 8 fluid oz. (240 ml) 125
(FDA, 1981). Hot chocolate 6 fluid oz. (180 ml) 50
At the time of aspartame approval, FDA considered Tabletop sweetener 1 packet 35
Pudding dessert 4 fluid oz. (120 ml) 25
the 99th percentile estimated intake as representa-
a Source:
tive of high-level consumers. However, since that time, Butchko and Stargel (2001).
S13
S14 BUTCHKO ET AL.
TABLE 2
Aspartame Intake (mg/kg body wt/day) in the United
States (90th Percentile, Eaters Only, 14-Day Average)
from 1984 to 1992a
Individuals
on a Women of
Dates General 25 reducing childbearing
of survey population years diet age Diabetics
1.15 mg/kg body wt/day, less than 3% of the ADI in body wt/day, respectively, for a 60-kg person. In 1994,
Europe (40 mg/kg body wt) (Virtanen et al., 1988). the 97.5th percentile of aspartame consumption in di-
abetics, who would likely be frequent consumers of as-
France. Chambolle et al. (1994) reported that from partame, was found to be 10.1 mg/kg body wt/day, only
1991 to 1992, aspartame intake was 0.6 and 1.0 mg/ about 25% of the ADI (MAFF, 1995).
kg body wt/day at the 90th and 95th percentiles, re-
spectively. However, this study was limited because Aspartame Intake in Other Countries
data for some categories were missing, and there were
no data for food consumed outside the home. More Australia. Mean consumption levels of aspartame
recently, aspartame intake was evaluated in insulin- were 6 and 7% of the ADI (40 mg/kg body wt) for all
dependent diabetic children ranging in age from 2 to respondents to a 7-day survey and total consumers, re-
20 years (Garnier-Sagne et al., 2001) using a 5-day di- spectively. Although the 90th percentile consumption
ary questionnaire. Intakes by aspartame consumers at was 23% of the ADI, the small sample rendered a precise
the mean, 97.5th percentile, and maximum were 2.4, estimate of 90th percentile intake difficult (National
7.8, and 15.6 mg/kg body wt/day, respectively. These re- Food Authority Australia, 1995).
sults, however, are very conservative as all sugar-free
products were assumed to contain only one sweetener at Brazil. Aspartame intake at the median by the users
its maximum authorized level whereas many products of intense sweeteners was 2.9% of the ADI (40 mg/
actually contain blends of several sweeteners. kg body wt); intakes at the median by diabetics and in-
dividuals on weight control regimens were 1.02 mg/kg
Germany. In a survey done in 19881989, consump- body wt/day (2.6% of the ADI) and 1.28 mg/kg body
tion of the sweeteners, aspartame, cyclamate, and sac- wt/day (3.2% of the ADI), respectively (Toledo and Ioshi,
charin was evaluated in Germany. The 90th percentile 1995).
average daily intake for aspartame was 2.75 mg/kg body
wt/day (Bar and Biermann, 1992). Canada. The general population of aspartame
eaters in Canada consumed 5.5 mg/kg body wt/day dur-
Italy. Italian teenagers who were known to be users ing cold weather months and 5.9 mg/kg body wt/day
of diet products had estimated average aspartame in- during warm weather months (7-day average, 90th per-
takes of only 0.03 mg/kg body wt/day; the maximum as- centile) in 1987. Intake by children and special pop-
partame intake was 0.39 mg/kg body wt/day (Leclercq ulations, who might consume higher amounts of as-
et al., 1999). partame, varied from 5.5 to 11.4 mg/kg body wt/day
(eaters only, 90th percentile, 7-day average) (Heybach
The Netherlands. Based on food frequency question- and Ross, 1989).
naires, mean aspartame intake was estimated to be
2.4 mg/kg body wt/day. Intake at the 95th percentile
Conclusion
intake was 7.5 mg/kg body wt/day. Using food intake
records, mean and 95th percentile intakes were 1.9 Dietary surveys with over 5000 individuals per year
and 7.5 mg/kg body wt/day, respectively (Hulshof and were done from 19841992 to monitor intake levels of
Bouman, 1995). aspartame in the United States. Average daily aspar-
tame intake at the 90th percentile (eaters only) in the
Norway. The average estimated intake of aspartame general population ranged from about 2 to 3 mg/kg body
varied from 0.9 to 3.4 mg/kg body wt/day among males wt over the course of the surveys. Consumption by 2- to
and females and various age groups (Bergsten, 1993). 5-year-old children in these surveys ranged from about
United Kingdom. In a survey from 1988 in the 2.5 to 5 mg/kg body wt/day. Aspartame intake was also
United Kingdom, aspartame consumption at the 90th monitored in several other countries, including seven
percentile was 4% of the ADI (40 mg/kg body wt) or European countries. Although survey methodologies
about 1.6 mg/kg body wt/day. Children and diabetics may have differed among these studies, aspartame in-
ingested only 7 and 6%, respectively, of the ADI at the take is remarkably consistent across studies and is well
90th percentile (Hinson and Nicol, 1992). From another below the ADI.
survey done by the UK Ministry of Agriculture, Fish-
eries, and Food (MAFF, 1990), median and maximum REFERENCES
aspartame intakes in 2- to 5-year-old children were
1.0 and 1.60 mg/kg body wt/day, respectively. In 35- to Abrams, I. J. (1986). Using the menu census survey to estimate di-
64-year-old adults, mean and maximum aspartame in- etary intake: Postmarket surveillance of aspartame. In Interna-
tional Aspartame Workshop Proceedings, November 1721, 1986,
takes were 0.25 and 6.20 mg/kg body wt/day, respec- Marbella, Spain. International Life Sciences InstituteNutrition
tively. In the general population, median, maximum, Foundation, Washington, DC.
and 97.5th percentile intakes were 16, 372, and 109 mg, Abrams, I. J. (1992). Using the menu census survey to estimate di-
respectively, which are equal to 0.3, 6.2, and 1.8 mg/kg etary intake: Postmarket surveillance of aspartame. In Food Safety
S16 BUTCHKO ET AL.
Assessment (J. W. Finley, S. F. Robinson, and D. J. Armstrong, Eds.), Garnier-Sagne, I., Leblanc, J. C., and Verger, P. (2001). Calculation of
pp. 201213. Am. Chem. Soc., Washington, DC. the intake of three intense sweeteners in young insulin-dependent
Bar, A., and Biermann, C. (1992). Intake of intense sweeteners in diabetics. Food Chem. Toxicol. 39, 745749.
Germany. Z. Ernahrungswiss. 31, 2539. Health and Welfare Canada (1979). Proposal on Aspartame, Informa-
Bergsten, C. (1993). Intake of Acesulfame-k, Aspartame, Cyclamate, tion Letter No. 564. Health Protection Branch, Ottawa.
and Saccharin in Norway, SNT Report 3. Norwegian Food Control Heybach, J. P., and Allen, S. S. (1988). Resources for inferential esti-
Authority, Oslo, Norway. mates of aspartame intake in the United States. In Dietary Pheny-
Butchko, H. (1990). Response to Calculation of aspartame intake lalanine and Brain Function (R. J. Wurtman and E. Ritter-Walker,
in children by Thomas-Doberson (letter). J. Am. Diet. Assoc. 90, Eds.), pp. 365372. Birkhauser, Boston.
360361. Heybach, J. P., and Smith, J. L. (1988). Intake of aspartame in 1950
Butchko, H. H., and Kotsonis, F. N. (1989). Aspartame: Review of year old women from the USDA continuing survey of food intakes
recent research. Comments Toxicol. 3, 253278. by individuals (CSFII 85). FASEB J. 2, A1615.
Butchko, H. H., and Kotsonis, F. N. (1991). Acceptable daily intake Heybach, J., and Ross, C. (1989). Aspartame consumption in a repre-
vs actual intake: The aspartame example. J. Am. Coll. Nutr. 10, sentative sample of Canadians. J. Can. Diet. Assoc. 50, 166170.
258266. Hinson, A. L., and Nicol, W. M. (1992). Monitoring sweetener con-
Butchko, H. H., and Kotsonis, F. N. (1994). Postmarketing surveil- sumption in Great Britain. Food Addit. Contam. 9, 669680.
lance in the food industry: the aspartame case study. In Nutritional Hulshof, K. F. A. M., and Bouman, M. (1995). Use of Various Types
Toxicology (F. N. Kotsonis, M. Mackey, and J. Hjelle, Eds.), pp. 235 of Sweeteners in Different Population Groups: 1992 Dutch National
250. Raven Press, New York. Food Consumption Survey. TNO Nutrition and Food Research In-
Butchko, H. H., and Kotsonis, F. N. (1996). Acceptable daily in- stitute, Netherlands.
take and estimation of consumption. In The Clinical Evaluation Joint FAO/WHO Expert Committee on Food Additives (JECFA)
of A Food Additive: Assessment of Aspartame (C. Tschanz, H. H. (1980). Aspartame: Toxicological Evaluation of Certain Food Ad-
Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 4353. CRC ditives, WHO Food Additive Series No. 15, pp. 1886. JECFA,
Press, Boca Raton, FL. Geneva.
Butchko, H. H., Tschanz, C., and Kotsonis, F. N. (1994). Postmarket- Leclercq, C., Berardi, D., Sorbillo, M. R., and Lambe, J. (1999). Intake
ing surveillance of food additives. Regul. Toxicol. Pharmacol. 20, of saccharin, aspartame, acesulfame K and cyclamate in Italian
105118. teenagers: Present levels and projections. Food Addit. Contam. 16,
Butchko, H. H., Stargel, W. W., Comer, C. P., Mayhew, D. A., and 99109.
Andress, S. (2001). Aspartame. In Alternative Sweeteners (L. Ministry of Agriculture, Fisheries, and Food (MAFF) (1990). Intakes
Nabors, Ed.), 3rd ed., pp. 4161. Marcel Dekker, New York. of intense and bulk sweeteners in the UK 19871988, Food Surveil-
Butchko, H. H., and Stargel, W. W. (2001). Aspartame: Scientific eval- lance Paper No. 29. Ministry of Agriculture, Fisheries and Food,
uation in the postmarketing period. Regul. Toxicol. Pharmacol. 34, London, UK.
221233. Ministry of Agriculture, Fisheries, and Food (MAFF) (1995). Survey
Butchko, H. H., and Stargel, W. W. (2002). Postmarketing surveillance of the Intake of Sweeteners by Diabetics, Food Surveillance Info.
in the food industry: The aspartame case study. In Nutritional Toxi- Sheet. Ministry of Agriculture, Fisheries and Food, Food Safety
cology (F. N. Kotsonis and M. Mackey, Eds.), second ed., pp. 287315. Directorate, London, UK.
Taylor & Francis, London/New York. National Food Authority Australia (1995). Survey of Intense Sweet-
Chambolle, M., Collerie de Borely, A., Renault, C., and Volatier, J. ener Consumption in Australia: Final Report. Canberra, Australia.
(1994). Intake of Intense Sweeteners in France (19891992). CRE- Scientific Committee for Food (1985). Food-Science and Techniques,
DOC, Departement Prospective de la Consommation, Paris. Reports of the Scientific Committee for Food, sixteenth series. Com-
Endres, J. (1990). Response to Calculation of aspartame intake in mission of the European Communities, Luxembourg.
children by Thomas-Doberson (letter). J. Am. Diet. Assoc. 90, 360. Stegink, L. D. (1984). Aspartame metabolism in humans: Acute dos-
Food and Drug Administration (FDA) (1974). Food additives permit- ing studies. In Aspartame: Physiology and Biochemistry. (L. D.
ted in food for human consumption: Aspartame. Fed. Regist. 39, Stegink and L. J. Filer, Jr., Eds.), pp. 509553. Marcel Dekker, New
2731727320. York.
Food and Drug Administration (FDA) (1981). Aspartame: Commis- Stenzel, T. E. (1990). Response to Calculation of aspartame intake in
sioners final decision. Fed. Regist. 46, 3828538308. children by Thomas-Doberson (letter). J. Am. Diet. Assoc. 90, 360.
Food and Drug Administration (FDA) (1983). Food additives permit- Thomas-Doberson, D. (1989). Calculation of aspartame intake in chil-
ted for direct addition to food for human consumption: Aspartame. dren. J. Am. Diet. Assoc. 89, 831833.
Fed. Regist. 48, 3137631382. Thomas-Doberson, D. (1990). Calculation of aspartame intake in chil-
Food and Drug Administration (FDA) (1984). Food additives permit- dren (letter). J. Am. Diet. Assoc. 90, 362, 364.
ted for direct addition to food for human consumption: Aspartame. Toledo, M. C. F., and Ioshi, S. H. (1995). Potential intake of intense
Fed. Regist. 49, 66726682. sweeteners in Brazil. Food Addit. Contam. 12, 799808.
Food and Drug Administration (FDA) (1986). Food additives permit- Virtanen, S. M., Rasanen, L., Paganus, A., Varo, P., and Akerblom,
ted for human consumption: Aspartame. Fed. Regist. 51, 42999 H. K. (1988). Intake of sugars and artificial sweeteners by adoles-
43000. cent diabetics. Nutr. Rep. Int. 38, 12111218.
Metabolism of Aspartame
Introduction effects in sensitive species (e.g., optic nerve toxic-
ity), the methanol derived from the methyl ester group
Aspartame (L--aspartyl-L-phenylalanine methyl es-
of aspartame was studied (see Safety of Methanol from
ter) is metabolized in the GI tract into three
Aspartame and the Diet below). Thus, an extensive
componentsthe amino acids, aspartic acid and pheny-
research program was undertaken before regulatory
lalanine, and methanol (Fig. 1). Aspartame may be
approvals to evaluate the plasma concentrations of
hydrolyzed into its components in the intestinal lu-
aspartate, phenylalanine, and methanol after various
men followed by absorption of these components into
doses of aspartame under different conditions. For the
the systemic circulation. At times, methanol is hy-
purpose of scientific completeness, additional studies
drolyzed in the intestinal lumen with transport of
were completed even after regulatory approvals were
the aspartylphenylalanine dipeptide into mucosal cells,
granted.
where it is metabolized to aspartate and phenylala-
The metabolism of aspartame and the pharmaco-
nine, which are then absorbed into the systemic circu-
kinetics of its components have been studied in healthy
lation. Aspartame may also be absorbed by intestinal
adults, healthy infants, lactating females, adults
mucosal cells where it is hydrolyzed to its components,
heterozygous for phenylketonuria (PKUH), adults
which are then absorbed into the systemic circulation
homozygous for phenylketonuria (PKU), and individu-
(Stegink 1984, 1987). Upon absorption, these con-
als thought to be sensitive to monosodium glutamate
stituents are metabolized, utilized, and/or excreted by
(MSG) (Chinese restaurant syndrome). Studies in-
the body in the same pathways as when they are derived
clude acute bolus doses, repeated doses, administration
from any other dietary source.
in capsules to evaluate the appropriateness of capsules
Common foods, such as milk, meat, fruits, and veg-
as a dosage form in blinded clinical studies, and as-
etables, actually provide far greater amounts of as-
partame ingested with meals, as well as long-term dos-
partic acid, phenylalanine, and methanol than does
ing studies. These safety studies have been extensively
aspartame in the diet. Initially, there were concerns
reviewed (Filer et al., 1984; Filer and Stegink, 1988,
that the plasma concentrations of aspartames con-
1989; Stegink, 1984, 1987; Stegink and Filer, 1996a,b).
stituents might be elevated above those known to be
In addition, effects of aspartame on the ratio of pheny-
well tolerated. There were concerns expressed that as-
lalanine (Phe) to the other large neutral amino acids
partame may result in plasma concentrations of as-
(LNAAs) that compete for transport across the blood
partate that may be associated with neurotoxicity (i.e.,
brain barrier (Phe/LNAA) were evaluated (see Phenyl-
neuronal necrosis) (see Aspartate and Excitotoxicity
alanine and Neurochemistry below).
below). It was also hypothesized that aspartame may
markedly increase plasma phenylalanine concentra-
Studies of Aspartame Metabolism
tions and thereby affect brain function, which was
a concern considering the adverse effects of elevated Extensive metabolism studies using radiolabeled as-
plasma phenylalanine concentrations associated with partame were done in several animal species, includ-
the rare genetic disease phenylketonuria (PKU) (see ing nonhuman primates, and humans to evaluate the
Phenylalanine and Neurochemistry below). In addition, metabolic fate of aspartame (Oppermann et al., 1973;
as ingestion of large amounts of methanol, with conse- Ranney et al., 1976; Ranney and Oppermann, 1979;
quent formation of formate, are associated with toxic Karim and Burns, 1996). The results of these studies
demonstrated that aspartame is metabolized into its
components in the GI tract; these are then absorbed
and utilized by the body via the same metabolic path-
ways as when these same constituents are derived from
common food.
O
Under conditions of extremes in pH, increased tem-
HO O
NH CH3
perature, and long storage times, aspartame, like other
aspartyl dipeptides, may cyclize to form the dike-
O NH2 O
topiperazine cycloaspartylphenylalanine (DKP) and
small amounts of -aspartame. The safety of these con-
L-aspartyl-L-phenylalanine methyl ester version products of aspartame has been established
(Kotsonis and Hjelle, 1996). DKPs are common com-
Aspartate Phenylalanine Methanol
ponents in food, e.g., protein-rich foods such as co-
FIG. 1. Aspartame is metabolized by esterases and peptidases
coa, cheese, protein, and casein hydrolysates, and in
in the gastrointestinal tract into its three constituentsthe amino roasted malts used in brewing. Exposure to DKP from
acids, aspartate and phenylalanine, and methanol. aspartame in products is limited because of product
S17
S18 BUTCHKO ET AL.
acceptability issues due to loss of sweetness. Studies In studies of the effect of aspartame on the activ-
in animals and humans indicate that DKP is poorly ity of various liver xenobiotic-metabolizing enzymes,
absorbed, is not biotransformed by mammalian en- Tutelyan et al. (1989, 1990) administered aspartame
zymes, and is rapidly eliminated in the urine. The - (40 and 4000 mg/kg body wt) for 90 days to rats (these
aspartyl dipeptides are formed from the rearrangement doses are about 131333 times 90th percentile aver-
of -aspartyl dipeptides and may also be synthesized age daily aspartame consumption in the United States)
by kidney enzymes. At least 14 -aspartyl dipeptides and concluded that aspartame, even in enormous doses,
and 6 -aspartyl tripeptides have been identified in hu- did not substantially affect hepatic microsomal enzyme
man urine. -Aspartylphenylalanine (-AP), the free function.
acid of -aspartame, has also been isolated and identi-
fied in plasma and urine samples obtained from normal Studies in Healthy Adults and Infants
healthy subjects who had not consumed aspartame (see Single bolus dosing studies in healthy subjects.
Kotsonis and Hjelle, 1996). Studies using doses of aspartame ranging from those
Burgert et al. (1991) evaluated the intestinal relevant to current consumption levels (4 and 10 mg/kg
metabolism of aspartame at doses of 735 mg/kg body body wt), to doses representative of premarketing pro-
wt, phenylalanine methyl ester (PME) (450 mg/kg body jections of high-level intake (34 mg/kg body wt) and the
wt), and phenylalanine (413 mg/kg body wt) admin- ADI in the United States (50 mg/kg body wt), to abu-
istered to the proximal jejunum in pigs. The pig was sive doses of 100, 150, and 200 mg/kg body wt were
chosen as an appropriate model for humans because of undertaken (Table 1). The 200 mg/kg body wt dose ap-
many structural and functional similarities with hu- proximated the amount of aspartame provided in about
mans. Neither aspartame, nor phenylalanine methyl 25 liters of 100% aspartame-sweetened beverage con-
ester, nor aspartylphenylalanine (-AP) was detected sumed at one time by a 70-kg adult. Studies were also
in portal blood, confirming the results of earlier stud- done in healthy 1-year-old infants given 34, 50, and
ies that aspartame is not absorbed intact. This dose of 100 mg/kg body wt of aspartame.
aspartame is approximately 245 times 90th percentile In healthy adults, single bolus doses of aspartame
average daily consumption of aspartame by the general administered at 4, 10, 34, and 50 mg/kg body wt
population in the United States (Butchko and Kotsonis, (Stegink et al., 1977, 1979a,b, 1987c; Wolf-Novak et al.,
1991, 1996). 1990) did not significantly affect plasma aspartate con-
Stegink et al. (1995) evaluated -aspartylphen- centrations. Mean peak plasma aspartate concentra-
ylalanine hydroxylase activity and -AP concentrations tions in healthy adults administered abuse doses (100,
in preschool and school-age children (Wolraich et al., 150, and 200 mg/kg body wt) of aspartame increased
1994) who had received aspartame, sucrose, and sac- slightly after aspartame dosing (Stegink et al., 1980,
charin for 3 weeks on each treatment. No -AP was de- 1981a). However, all plasma aspartate concentrations
tected, and plasma phenylalanine and aspartate con- (maximum approximately 1.0 mol/dl) were well within
centrations were within the normal range. Although normal plasma aspartate concentrations observed post-
one subject had -aspartylphenylalanine hydroxylase prandially in healthy human infants and adults
activity well below normal, there were no behavioral or (Stegink and Filer, 1996a). The results of the stud-
cognitive effects observed after aspartame. ies in infants given doses of aspartame of 34, 50, and
Lipton et al. (1991) evaluated intestinal absorption
of aspartames metabolic components and conversion
products in steady state perfusion studies of rat je-
junum. The results demonstrated that -AP and pheny- TABLE 1
lalanine were rapidly taken up by a carrier-assisted sys- Mean (SD) Fasting Plasma Phenylalanine Con-
tem in the GI mucosal cells, but that -AP and DKP centrations, Mean Peak (SD) Plasma Phenylalanine
may be absorbed through passive diffusion or excreted Concentrations, and Time to Mean Peak after Various
in the GI tract. Hooper et al. (1994) used microvillar Bolus Doses of Aspartame in Healthy Adultsa
membrane preparations from human duodenum, je- Dosage Fasting Mean peak Time to mean peak
junum, and ileum and from pig duodenum and kidney to (mg/kg body wt) (mol/dl) (mol/dl) (min)
evaluate the metabolism of aspartame, -AP, - aspar-
tame, and DKP by GI tract aminopeptidases. Expect- 0 5.50 0.54 5.61 1.42 30
edly, this study demonstrated that aspartame and - 4 5.48 0.85 5.67 0.48 30
10 5.08 0.82 6.73 0.75 30
AP were rapidly metabolized by aminopeptidases, with 34 5.66 1.21 11.1 2.49 30
aminopeptidase A being the major activity in the mi- 50 4.61 1.72 16.2 4.86 45
crovillar membranes involved with the metabolism of 100 5.40 1.05 20.2 6.77 45
aspartame. -AP and DKP were essentially resistant 150 6.72 1.93 35.0 9.43 90
200 5.26 0.67 48.7 15.5 90
to hydrolysis in human and pig microvillar membrane
preparations. a Adapted from Stegink et al. (1987c).
ASPARTAME: REVIEW OF SAFETY S19
A number of studies examined the effect of various partame and Seizures and Electroencephalograms
doses of aspartame on Phe/LNAA. Expectedly, a high- (EEGs); Evaluation of Aspartame and Headache; and
protein meal provides phenylalanine in conjunction Evaluation of Aspartame and Behavior, Cognitive
with other LNAAs and increased the plasma pheny- Function, and Mood sections below).
lalanine concentration but did not raise the Phe/LNAA
above baseline (0.101 0.010). However, ingestion of a Solution versus capsule administration in healthy
single aspartame dose of 34 mg/kg body wt increased subjects. In order to conduct randomized, double-
the Phe/LNAA from a baseline value of 0.102 0.014 blind, placebo-controlled studies to evaluate various
to a mean peak value of 0.230 0.039 in healthy adults issues raised regarding aspartame (e.g., headaches, sei-
(Stegink and Filer, 1996a; Stegink et al., 1979b). Plasma zures, allergic-type reactions) aspartame and placebo
Phe/LNAA values were lower when approximately the must be administered in a capsule form in order to
same dose of aspartame was administered in divided mask aspartames sweetness to maintain the integrity
doses to healthy adults. For example, ingestion of three of the blinded condition. To evaluate whether there
successive 10 mg/kg body wt doses resulted in a mean were pharmacokinetic differences in amino acid profiles
peak Phe/LNAA of 0.161 0.021 (Stegink et al., 1988). after aspartame administered in solution compared to
When 34 mg/kg body wt of aspartame was added to capsules, Stegink et al. (1987b) administered 3000 mg
a high-protein meal, the Phe/LNAA increased from of aspartame (50 mg/kg body wt for a 60-kg adult) in so-
0.1000.017 at baseline to a mean peak of 0.1300.032 lution on one occasion and in capsules on another occa-
(Stegink and Filer, 1996a; Stegink et al., 1982). Thus, sion. Using this very large dose of aspartame, the mean
ingestion of protein with aspartame, as would normally peak plasma phenylalanine concentrations were higher
occur throughout the day, would reduce the change in and occurred earlier when aspartame was given in so-
the plasma Phe/LNAA related to aspartame. lution, and the 4-h area under the plasma phenylala-
Using a larger dose of aspartame (600 mg every nine curve (AUC) was greater after aspartame ingestion
hour for 8 h) in healthy adults revealed the mean in solution compared to capsules. However, in blinded
peak Phe/LNAA of 0.22 0.05 (Stegink et al., 1989a). clinical studies, large doses are typically administered
Phe/LNAA values as high as 0.125 0.045 have been as divided doses. When lower and more relevant doses
reported (Fernstrom et al., 1979; Stegink and Filer, (20 mg/kg body wt) were used (Burns et al., 1990), there
1996a) in healthy adults under various dietary condi- were much smaller differences between plasma pheny-
tions, and a mean Phe/LNAA two standard deviations lalanine pharmacokinetics after capsules compared to
above this value would be 0.215. Thus, the mean peak solution. This dose is similar to those used in focused
Phe/LNAA observed when about 70 mg/kg body wt as- clinical studies. Thus, capsules provided an appropriate
partame was given over 8 h was only slightly above dosage form for blinded clinical studies.
the normal range despite the fact that the aspartame Aging and aspartame metabolism. Puthrasingam
dose was well above current use levels (about 3 mg/kg et al. (1996) evaluated the metabolism of aspartame
body wt at the 90th percentile). In addition, the effect of in elderly (6580 years) subjects compared to young
aspartame on Phe/LNAA was amplified since subjects subjects. Subjects were given 40 mg/kg body wt aspar-
were fasted the entire day; thus, the normal ingestion tame, and pharmacokinetics (Cmax and AUC) of plasma
of protein that would have modulated the increase in phenylalanine and tyrosine concentrations were deter-
Phe/LNAA was absent. mined. There were significantly higher Cmax and AUC
The consumption of aspartame at levels more re- for phenylalanine in the elderly group compared to
flective of actual consumption (i.e., 4 mg/kg body wt) the young subjects; however, there were significant de-
does not significantly affect the Phe/LNAA (Stegink creases in volume of distribution and clearance in the
et al., 1987c). Further, such effects on Phe/LNAA are elderly, which accounted for these differences. The au-
not unique to aspartame (see Phenylalanine and Neu- thors concluded that the differences were modest and
rochemistry below). Ingestion of carbohydrate, such as were not indicative of a need to change the ADI for as-
sugar, has similar effects on the Phe/LNAA (Martin- partame in the elderly.
DuPan et al., 1982; Burns et al., 1991; Wolf-Novak et al.,
1990). Insulin-mediated uptake of large neutral amino
Studies in Individuals Heterozygous for
acids (especially the branched-chain amino acids) after
Phenylketonuria (PKUH)
carbohydrate ingestion leads to decreased total plasma
concentrations of the other large neutral amino acids As individuals with the rare, genetic disease
relative to phenylalanine and thus an increase in the phenylketonuria (PKU) cannot properly metabolize
Phe/LNAA. Further, despite significant increases in phenylalanine, there was interest in whether individ-
Phe/LNAA after very large doses of aspartame, no uals heterozygous for PKU (carriers) would be able to
consistent effects on brain function have been found in metabolize the phenylalanine derived from aspartame
animal studies and controlled clinical studies (see so that plasma concentrations did not reach levels seen
Phenylalanine and Neurochemistry; Evaluation of As- in untreated PKU (about 120 mol/dl or higher). PKUH
S22 BUTCHKO ET AL.
moles/dL
of aspartame, repeated doses of aspartame, and aspar- 30
tame given with a protein meal in PKUH.
Single-bolus dosing studies in PKUH. After a dose 20
Normal Range
of 10 mg/kg body wt aspartame in PKUH (Stegink
et al., 1987c; Stegink and Filer, 1996b), plasma pheny- 10
products, which are labeled to inform these individu- an MSG-type response after consuming either sucrose
als that the product contains phenylalanine. However, or aspartame. Plasma aspartate concentrations were
questions have arisen regarding the safety of aspar- not significantly different after aspartame loading than
tame ingestion by PKU individuals no longer on a low- after sucrose loading. The results of this study indicate
phenylalanine diet. that subjects who allegedly experienced idiosyncratic
Caballero et al. (1986) evaluated plasma phenylala- reactions to MSG did not experience similar symptoms
nine concentrations at baseline and 1 h after aspartame after ingestion of either aspartame or sucrose.
(10 mg/kg body wt) in healthy subjects, PKU heterozy- In a randomized, crossover study, Stegink et al. (1982)
gotes, mild hyperphenylalaninemics, and PKU homo- evaluated the effect of aspartame (34 mg/kg body wt)
zygotes not being treated with a low-phenylalanine plus MSG (34 mg/kg body wt) ingestion on plasma
diet. Baseline and post-dose plasma phenylalanine con- and erythrocyte amino acids concentrations in healthy
centrations were 4.45 1.29 vs 5.80 0.95 mol/dl in adults fed a high-protein (1 g/kg body wt) meal com-
healthy subjects; 6.88 1.38 vs 8.23 1.73 mol/dl in pared to the meal alone. The addition of aspartame plus
PKU heterozygotes; 41.2 20.8 vs 41.3 1.84 mol/dl MSG had little effect on plasma or erythrocyte concen-
in mild hyperphenylalaninemics; and 136.9 24.0 vs trations of either aspartate or glutamate beyond those
132.3 21.0 mol/dl in PKU homozygotes. from the meal alone. In another similar study (Stegink
Wolf-Novak et al. (1990) evaluated plasma pheny- et al., 1983b), subjects were given a high-protein (1 g/kg
lalanine concentrations after ingestion of a 355-ml body wt) meal, a high-protein meal with 150 mg/kg body
serving of 100% aspartame-sweetened beverage (about wt MSG, and a high-protein meal with 150 mg/kg body
4 mg/kg body wt) in healthy subjects and adoles- wt MSG plus 23 mg/kg body wt aspartame. Ingestion of
cent PKU individuals who were no longer on a low- MSG significantly elevated plasma glutamate plus as-
phenylalanine diet. The aspartame-sweetened bever- partate concentrations above those from the meal alone.
age had no significant effect compared to baseline on The addition of aspartame to the meal with MSG did
plasma phenylalanine concentrations in healthy sub- not increase plasma glutamate plus aspartate concen-
jects (5.39 0.470 mol/dl) or in adolescent PKU sub- trations above those after the meal with MSG.
jects (150 23 mol/dl). Mackey and Berlin (1992) also Stegink et al. (1987a) also evaluated the effect of MSG
evaluated the effect of aspartame in a 355 ml serv- and aspartame on plasma amino acid concentrations
ing of beverage vs a sugar-sweetened beverage on when given a low-protein meal. Study conditions in-
plasma phenylalanine and tyrosine concentrations in cluded the meal alone, the meal plus MSG (50 mg/kg
healthy subjects and PKU individuals no longer on a body wt), and the meal plus MSG (50 mg/kg body wt)
phenylalanine-restricted diet. There were no signifi- and aspartame (34 mg/kg body wt). Plasma glutamate
cant increases in plasma phenylalanine and tyrosine or aspartate or glutamate plus aspartate concentra-
concentrations in healthy subjects. Likewise, in PKU tions were not affected by the meal alone but were sig-
individuals, plasma phenylalanine and tyrosine concen- nificantly increased after MSG and MSG plus aspar-
trations did not differ significantly from baseline. tame conditions. The AUC for plasma aspartate plus
The results of these studies demonstrated that aspar- glutamate concentrations, however, did not differ be-
tame (equal to that in up to 1 liter of beverage sweet- tween the MSG compared to MSG plus aspartame con-
ened with 100% aspartame consumed as a single bolus) ditions. The results of these studies demonstrated that
does not increase plasma phenylalanine concentrations symptoms of Chinese restaurant syndrome elicited
above the already very elevated levels in mild hyper- by MSG were not elicited by aspartame. In addition,
phenylalaninemics and untreated PKU individuals. the results do not support the contention that aspar-
tame consumed with MSG results in rapid and danger-
ous rises in plasma aspartate and glutamate concentra-
Metabolism after Administration of Aspartame Plus
tions.
Monosodium Glutamate (MSG)
The ingestion of large amounts of MSG may pro-
Conclusions
duce a variety of symptoms (e.g., burning sensation,
facial pressure, and chest pain) in sensitive individ- Numerous studies have evaluated the metabolism
uals, usually referred to as Chinese restaurant syn- of aspartame and the pharmacokinetics of its
drome. Because of aspartames aspartate content and componentsaspartate, phenylalanine, and methanol.
the structural similarity between glutamate and aspar- Such studies have involved healthy adults, infants,
tate, concern about aspartame consumption along with children, and adolescents, PKU heterozygous adults,
MSG in foods suggested that aspartame might preci- PKU homozygous individuals, and individuals sensi-
pitate symptoms in subjects allegedly sensitive to MSG. tive to MSG. Studies included acute-dose, repeated-
Stegink et al. (1981b) gave MSG-sensitive subjects as- dose, and long-term dose regimens. In healthy adults
partame (34 mg/kg body wt) or sucrose (1 g/kg body wt) and children, even after enormous doses, aspartame
using a randomized, double-blind, crossover experimen- does not result in plasma concentrations of its com-
tal design. No test subject reported symptoms typical of ponents that are of safety concern. Plasma aspartate
S24 BUTCHKO ET AL.
concentrations remained within the normal range even partame loading on plasma and erythrocyte free amino acid con-
after bolus doses of aspartame as high as 200 mg/kg. centrations in one-year-old infants. J. Nutr. 113, 15911599.
The doses of aspartame used in studies with the other Filer, L. J., Jr., Baker, G. L., and Stegink, L. D. (1984). Aspartame in-
gestion by human infants. In Aspartame: Physiology and Biochem-
subpopulations have ranged from 4 to 100 mg/kg body istry (L. D. Stegink and L. J. Filer, Jr., Eds.), pp. 579591. Marcel
wt. Plasma phenylalanine concentrations observed af- Dekker, New York.
ter large-bolus and repeated doses of aspartame were Filer, L. J., Jr., and Stegink, L. D. (1988). Effect of aspartame on
comparable to the normal postprandial range in both plasma phenylalanine concentration in humans. In Dietary Pheny-
healthy adults and PKUH and well below those in un- lalanine and Brain Function (R. J. Wurtman and E. Ritter-Walker,
treated PKU. Blood methanol concentrations were not Eds.), pp. 1840. Birkhauser, Boston, MA.
detectable after aspartame doses as high as 34 mg/kg Filer, L. J., Jr., and Stegink, L. D. (1989). Aspartame metabolism in
normal adults, phenylketonuric heterozygotes, and diabetic sub-
body wt when ingested as a single bolus or about jects. Diabetes Care 12, 6774.
70 mg/kg body wt when administered as eight di- Frey, G. H. (1976). Use of aspartame by apparently healthy children
vided doses at hourly intervals. Regardless of dose, and adolescents. J. Toxicol. Environ. Health 2, 401415.
blood formate concentrations did not change from base- Hooper, N. M., Hesp, R. J., and Tieku, S. (1994). Metabolism of aspar-
line levels after aspartame administration. Evalua- tame by human and pig intestinal microvillar peptidases. Biochem.
tions of urinary formate excretion after single bo- J. 298, 635639.
lus doses of aspartame as high as 200 mg/kg body Karim, A., and Burns, T. (1996). Metabolism and pharmacokinetics
of radiolabeled aspartame in normal subjects. In The Clinical Eval-
wt demonstrated the bodys ability to rapidly meta-
uation of a Food Additive: Assessment of Aspartame (C. Tschanz,
bolize methanol and excrete formate. Thus, even when H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 5766.
administered at levels well above 90th percentile av- CRC Press, Boca Raton, FL.
erage daily consumption (Butchko and Kotsonis, 1991, Knopp, R. H., Brandt, K., and Arky, R. A. (1976). Effects of aspar-
1996), the risk of adverse effects from aspartame or its tame in young persons during weight reduction. J. Toxicol. Environ.
metabolic components is negligible, strongly supporting Health 2, 417428.
the safety of aspartame under its intended conditions Koch, R., Shaw, K. N. F., Williamson, M., and Haber, M. (1976). Use
of aspartame in phenylketonuric heterozygous adults. J. Toxicol.
of use. Environ. Health 2, 453457.
Kotsonis, F. N., and Hjelle, J. J. (1996). The safety assessment of
REFERENCES aspartame: Scientific and regulatory considerations. In The Clin-
ical Evaluation of a Food Additive: Assessment of Aspartame
Burgert, S. L., Andersen, D. W., Stegink, L. D., Takeuchi, H., (C. Tschanz, H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.),
and Schedl, H. P. (1991). Metabolism of aspartame and its L- pp. 2341. CRC Press, Boca Raton, FL.
phenylalanine methyl ester decomposition product by the porcine Leon, A. S., Hunninghake, D. B., Bell, C., Rassin, D. K., and Tephly,
gut. Metabolism 40, 612618. T. R. (1989). Safety of long-term large doses of aspartame. Arch.
Burns, T. S., Stargel, W. W., and Hurwitz, A. (1990). Bioavailability Intern. Med. 149, 23182324.
of phenylalanine and aspartate from aspartame (20 mg/kg) in cap- Lipton, W. E., Li, Y.-N., Younoszai, M. K., and Stegink, L. D. (1991).
sules and solution. Metabolism 39, 12001203. Intestinal absorption of aspartame decomposition products in adult
Burns, T. S., Stargel, W. W., Tschanz, C., Kotsonis, F. N., and Hurwitz, rats. Metabolism 40, 13371345.
A. (1991). Aspartame and sucrose produce a similar increase in the Mackey, S. A., and Berlin, C. M., Jr. (1992). Effect of dietary aspartame
plasma phenylalanine to large neutral amino-acid ratio in healthy- on plasma concentrations of phenylalanine and tyrosine in normal
subjects. Pharmacology 43, 210219. and homozygous phenylketonuric patients. Clin. Pediatr. 31, 394
Butchko, H. H., and Kotsonis, F. N. (1991). Acceptable daily intake 399.
vs actual intake: The aspartame example. J. Am. Coll. Nutr. 10, Martin-Du Pan, R., Mauron, C., Glaeser, B., and Wurtman, R. J.
258266. (1982). Effect of various oral glucose doses on plasma neutral amino
Butchko, H. H., and Kotsonis, F. N. (1996). Acceptable daily in- acid levels. Metabolism 31, 937943.
take and estimation of consumption. In The Clinical Evaluation Oppermann, J. A., Muldoon, E., and Ranney, R. E. (1973). Metabolism
of A Food Additive: Assessment of Aspartame (C. Tschanz, H. H. of aspartame in monkeys. J. Nutr. 103, 14541459.
Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 4353. CRC Puthrasnigam, S., Heybroek, W. M., Johnston, A., Maskrey, V., Swift,
Press, Boca Raton, FL. C. G., Turner, P., Abrams, S. M. L., and Jackson, S. H. D. (1996).
Caballero, B., Mahon, B. E., Rohr, F. J., Levy, H. L., and Wurtman, R. J. Aspartame pharmacokinetics: The effect of ageing. Age Ageing 25,
(1986). Plasma amino acid levels after single-dose aspartame con- 217220.
sumption in phenylketonuria, mild hyperphenylalaninemia, and Ranney, R. E., and Oppermann, J. A. (1979). A review of metabolism
heterozygous state for phenylketonuria. J. Pediatr. 109, 668671. of the aspartyl moiety of aspartame in experimental animals and
Curtius, H., Endres, W., and Blau, N. (1994). Effect of high-protein man. J. Environ. Pathol. Toxicol. 2, 979985.
meal plus aspartame ingestion on plasma phenylalanine concen- Ranney, R. E., Oppermann, J. A., Muldoon, E., and McMahon, F. G.
trations in obligate heterozygotes for phenylketonuria. Metabolism (1976). Comparative metabolism of aspartame in experimental an-
43, 413416. imals and humans. J. Toxicol. Environ. Health. 2, 441451.
Fernstrom, J. D., Wurtman, R. J., Hammarstrom-Wiklund, B., Rand, Stegink, L. D. (1984). Aspartame metabolism in humans: Acute dos-
W. M., Munro, H. N., and Davidson, C. S. (1979). Diurnal variations ing studies. In Aspartame: Physiology and Biochemistry. (L. D.
in plasma concentrations of tryptophan, tyrosine and other neutral Stegink and L. J. Filer, Jr., Eds.), pp. 509553. Marcel Dekker, New
amino acids: Effect of dietary protein intake. Am. J. Clin. Nutr. 32, York.
19121922. Stegink, L. D. (1987). The aspartame story: A model for the clinical
Filer, L. J., Jr., Baker, G. L., and Stegink, L. D. (1983). Effect of as- testing of a food additive. Am. J. Clin. Nutr. 46, 204215.
ASPARTAME: REVIEW OF SAFETY S25
Stegink, L. D., and Filer, L. J., Jr. (1996a). Effects of aspartame in- Stegink, L. D., Wolf-Novak, L. C., Filer, L. J., Jr., Bell, E. F., Ziegler,
gestion on plasma aspartate, phenylalanine, and methanol concen- E. E., Krause, W. L., and Brummel, M. C. (1987c). Aspartame-
trations in normal adults. In The Clinical Evaluation of a Food sweetened beverage: Effect on plasma amino acid concentrations
Additive: Assessment of Aspartame (C. Tschanz, H. H. Butchko, in normal adults and adults heterozygous for phenylketonuria.
W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 6786. CRC Press, J. Nutr. 117, 19891995.
Boca Raton, FL. Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1988). Repeated
Stegink, L. D., and Filer, L. J., Jr. (1996b). Effects of aspartame in- ingestion of aspartame-sweetened beverage: Effect on plasma
gestion on plasma aspartate, phenylalanine, and methanol concen- amino acid concentrations in normal adults. Metabolism. 37, 246
trations in potentially sensitive populations. In The Clinical Eval- 251.
uation of a Food Additive: Assessment of Aspartame (C. Tschanz, Stegink, L. D., Filer, L. J., Jr., Bell, E. F., Ziegler, E. E., and Tephly,
H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 87113. T. R. (1989a). Effect of repeated ingestion of aspartame-sweetened
CRC Press, Boca Raton, FL. beverage on plasma amino acid, blood methanol, and blood formate
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1977). Effect of concentrations in normal adults. Metabolism 38, 357363.
aspartame and aspartate loading upon plasma and erythrocyte Stegink, L. D., Filer, L. J., Jr., Baker, G. L., Bell, E. F., Ziegler, E. E.,
free amino acid levels in normal adult volunteers. J. Nutr. 107, Brummel, M. C., and Krause, W. L. (1989b). Repeated ingestion
18371845. of aspartame-sweetened beverage: Effect on plasma amino acid
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1979a). Plasma, concentrations in individuals heterozygous for phenylketonuria.
erythrocyte and human milk levels of free amino acids in lactating Metabolism 38, 7884.
women administered aspartame or lactose. J. Nutr. 109, 21732181. Stegink, L. D., Filer, L. J., Jr., Bell, E. F., Ziegler, E. E., Tephly,
Stegink, L. D., Filer, L. J., Jr., Baker, G. L., and McDonnell, J. E. T. R., and Krause, W. L. (1990). Repeated ingestion of aspartame-
(1979b). Effect of aspartame loading upon plasma and erythrocyte sweetened beverages: Further observations in individuals
amino acid levels in phenylketonuric heterozygotes and normal heterozygous for phenylketonuria. Metabolism 39, 10761081.
adult subjects. J. Nutr. 109, 708717. Stegink, L. D., Filer, L. J., Jr., Brummel, M. C., Baker, G. L., Krause,
Stegink, L. D., Filer, L. J., Jr., Baker, G. L., and McDonnell, J. E. W. L., Bell, E. F., and Ziegler, E. E. (1991). Plasma amino acid
(1980). Effect of an abuse dose of aspartame upon plasma and concentrations and amino acid ratios in normal adults and adults
erythrocyte levels of amino acids in phenylketonuric heterozygous heterozygous for phenylketonuria ingesting a hamburger and milk
and normal adults. J. Nutr. 110, 22162224. shake meal. Am. J. Clin. Nutr. 53, 670675.
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1981a). Plasma and Stegink, L. D., Lindgren, S. D., Brummel, M. C., Stumbo, P. J., and
erythrocyte concentrations of free amino acids in adult humans Wolraich, M. L. (1995). Erythrocyte L-aspartyl-L-phenylalanine
administered abuse doses of aspartame. J. Toxicol. Environ. Health hydroxylase activity and plasma phenylalanine and aspartate
7, 291305. concentrations in children consuming diets high in aspartame.
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1981b). Effect of Am. J. Clin. Nutr. 62, 12061211.
aspartame and sucrose loading in glutamate-susceptible subjects. Stern, S. B., Bleicher, S. J., Flores, A., Gombos, G., Recitas, D.,
Am. J. Clin. Nutr. 34, 18991905. and Shu, J. (1976). Administration of aspartame in non-insulin-
Stegink, L. D., Brummel, M. C., McMartin, K., Martin-Amat, G., dependent diabetics. J. Toxicol. Environ. Health 2, 429439.
Filer, L. J., Jr., Baker, G. L., and Tephly, T. R. (1981c). Blood Tephly, T. R. (1991). The toxicity of methanol. Life Sci. 48, 10311041.
methanol concentrations in normal adult subjects administered Tephly, T. R., and McMartin, K. E. (1984). Methanol metabolism and
abuse doses of aspartame. J. Toxicol. Environ. Health 7, 281290. toxicity. In Aspartame: Physiology and Biochemistry (L. D. Stegink
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1982). Effect of and L. J. Filer, Jr., Eds.), pp. 111140. Marcel Dekker, New York.
aspartame plus monosodium L-glutamate ingestion on plasma and Tutelyan, V. A., Gapparov, M. M., Virovets, O. A., Kravchenko, L. V.,
erythrocyte amino acid levels in normal adult subjects fed a high Vasilyev, A. V., Sokolov, A. I., Konovalova, L. S., Vasilevskaya, L. G.,
protein meal. Am. J. Clin. Nutr. 36, 11451152. Shlygin, G. K., Tarasova, I. B., Vorobeychik, T. A., and Kuzmina,
Stegink, L. D., Brummel, M. C., Filer, L. J., Jr., and Baker, G. L. Y. E. (1989). New data on the metabolism and physiological
(1983a). Blood methanol concentrations in one-year-old infants mechanism of aspartame. Vopr. Pitan. 1, 2025.
administered graded doses of aspartame. J. Nutr. 113, 16001606. Tutelyan, V. A., Kravchenko, L. V., and Kuzmina, E. E. (1990). The ef-
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1983b). Plasma fect of aspartame on the activity of rat liver xenobiotic-metabolizing
amino acid concentrations in normal adults fed meals with added enzymes. Drug Metab. Dispos. 18, 223225.
monosodium L-glutamate and aspartame. J. Nutr. 113, 18511860. Wolf-Novak, L. C., Stegink, L. D., Brummel, M. C., Persoon, T. J.,
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1987a). Plasma Filer, L. J., Jr., Bell, E. F., Ziegler, E. E., and Krause, W. L. (1990).
amino acid concentrations in normal adults ingesting aspartame Aspartame ingestion with and without carbohydrate in phenyl-
and monosodium L-glutamate as part of a soup/beverage meal. ketonuric and normal subjects: Effect on plasma concentrations of
Metabolism 36, 10731079. amino acids, glucose and insulin. Metabolism 39, 391396.
Stegink, L. D., Filer, L. J., Jr., Bell, E. F., and Ziegler, E. E. (1987b). Wolraich, M. L., Lindgren, S. D., Stumbo, P. J., Stegink, L. D.,
Plasma amino acid concentrations in normal adults adminis- Appelbaum, M. I., and Kiritsy, M. C. (1994). Effects of diets high in
tered aspartame in capsules or solution: Lack of bioequivalence. sucrose or aspartame on the behavior and cognitive performance
Metabolism 36, 507512. of children. N. Engl. J. Med. 330, 301307.
Aspartate and Excitotoxicity
Introduction neuronal necrosis (Olney, 1976; Reynolds et al., 1976),
Aspartate and glutamate are some of the most com- presumably as a result of elevated blood aspartate
mon amino acids in the diet. For example, a 100-g por- concentrations.
tion of chicken provides about 2600 mg of aspartate, However, aspartame doses that result in plasma as-
whereas 355 ml of beverage sweetened 100% with as- partate concentrations less than 100 mol/dl do not
partame provides only about 70 mg of aspartate (Fig. 1). produce hypothalamic neuronal necrosis, even in the
As noted under Intake of Aspartame vs the Acceptable acutely sensitive neonatal mouse. For example, effects
Daily Intake, aspartame (90th percentile intake) pro- are seen only when the plasma concentration of aspar-
vides only about 2 and 1% of the usual daily dietary tate exceeds 110 mol/dl, the plasma concentration of
intake of aspartate by adults and 4-year-old children, glutamate exceeds 75 mol/dl, or the combined gluta-
respectively. Based on metabolism studies with aspar- mate plus aspartate concentrations exceed 128 mol/dl
tame, it is simply not possible for a human to consume (Finkelstein et al., 1983; Daabees et al., 1985).
enough aspartame in products to produce a significant Olney and co-workers (Olney and Sharpe, 1969;
increase in plasma aspartate concentrations. In addi- Olney et al., 1972) also reported this phenomenon in
tion, both aspartic acid and glutamic acid are readily neonatal nonhuman primates administered large bolus
and rapidly metabolized and utilized by the body for doses of glutamate. This observation, however, could
energy and for incorporation into proteins. not be reproduced by a number of other scientists
Nonetheless, some authors (Olney, 1982; Reif-Lehrer, with either glutamate or aspartame at high dosages
1976) speculated that the aspartate constituent of (Newman et al., 1973; Wen et al., 1973; Abraham et al.,
aspartame, especially when consumed with foods con- 1975; Reynolds et al., 1976, 1979, 1980). For exam-
taining monosodium glutamate (MSG), would cause an ple, aspartame (2000 mg/kg body wt) administered to
increase in the combined plasma concentrations of as- neonatal nonhuman primates, with or without added
partate and glutamate, which might pose a risk of focal MSG (1000 mg/kg body wt), did not produce neuronal
brain lesions. Blaylock (1993) has continued to raise necrosis even though plasma aspartate and glutamate
this speculation. concentrations were markedly elevated (Reynolds et al.,
1976, 1980).
Studies in Animals
Studies in Humans
Very high plasma concentrations of the abundant,
naturally occurring dicarboxylic amino acids, aspar- As described in detail under the section Metabolism of
tate and glutamate, are known to induce hypothala- Aspartame, Stegink and co-workers (1977, 1979) found
mic neuronal necrosis in neonatal rodents (Olney, 1969; no significant changes in plasma concentrations of as-
Lemkey- Johnson and Reynolds, 1974; Okaniwa et al., partate after bolus doses of 34 or 50 mg/kg body wt
1979; Finkelstein et al., 1983; Daabees et al., 1985). As- aspartame. Significant increases in plasma aspartate
partame administered in amounts ranging from 1000 concentrations were observed after aspartame doses of
2500 mg/kg body wt to neonatal mice also produced 100 mg/kg body wt (equivalent to the amount of aspar-
tame in approximately 12 liters of 100% aspartame-
3000 2578 sweetened beverage for a 60-kg individual) or higher
2500 (Stegink et al., 1981a). Even with acute administration
of 200 mg aspartame/kg body wt, mean peak plasma as-
2000
partate concentrations (0.76 0.57 mol/dl) remained
1500 well below those associated with toxicity in neonatal
mg
1000
948
rodents (110 mol/dl). Evaluations of the effect of large
500
doses (34, 50, and 100 mg/kg body wt) of aspartame
133
72 (Filer et al., 1983) in 1-year-old infants demonstrated
0
100% Chicken* No-fat Milk* Banana*
that infants metabolize aspartame as well as adults.
Aspartame (100 g, Roasted) (355 ml) (Medium) Further, daily administration of large doses of aspar-
Beverage
(355 ml)
tame (75 mg/kg body wt/day) over 24 weeks had no
significant effect on mean fasting plasma aspartate
FIG. 1. Aspartic acid content of various foods and beverages com- concentrations (Leon et al., 1989). Thus, even extreme
pared to a 100% aspartame-sweetened beverage. Only a small amount doses of aspartame are metabolized rapidly and do not
of aspartate is provided by aspartame in products compared to that pose a risk of adverse effects from aspartate.
provided by common foods and beverages. Source: U.S. Department
of Agriculture, Agricultural Research Service, 2001. USDA Nutrient The ingestion of large amounts of MSG has been
Database for Standard Reference, Release 14. Nutrient Data Labo- claimed to produce a variety of symptoms (e.g., burning
ratory home page, http://www.nal.usda.gov.fnic.foodcomp. sensation, facial pressure, and chest pain) in sensitive
S26
ASPARTAME: REVIEW OF SAFETY S27
individuals, usually called Chinese restaurant syn- tate or glutamate or glutamate plus aspartate to those
drome. Concern was expressed regarding the consump- associated with toxicity in neonatal rodents. Further-
tion of aspartame with MSG in foods (Olney, 1982; Reif- more, given the totality of the data in primates, even
Lehrer, 1976), suggesting that aspartame might cause if such increased plasma concentrations were possible
symptoms in subjects thought to be sensitive to MSG after very large amounts of aspartame, these doses are
because of the structural similarity between glutamate of questionable relevance to consumption of aspartame
and aspartate and the similar neurotoxicity observed at by humans.
extremely large doses in animal studies. It was also sug-
gested that the combined plasma concentrations of glu- REFERENCES
tamate and aspartate (glutamate plus aspartate) may
be an indicator of the risk of toxicity. Abraham, R., Swart, J., Golberg, L., and Coulston, F. (1975). Elec-
However, Stegink et al. (1981b) demonstrated that tron microscopic observations of hypothalami in neonatal rhesus
subjects who allegedly experienced idiosyncratic reac- monkeys (Macaca mulatta) after administration of monosodium-L-
glutamate. Exp. Mol. Pathol. 23, 203213.
tions to MSG do not experience similar symptoms after
Blaylock, R. L. (Ed.) (1993). Excitotoxins: The Taste That Kills. Health
ingestion of either aspartame or sucrose. Stegink et al. Press, Santa Fe, NM.
(1982) also demonstrated that ingestion of aspartame Daabees, T. T., Finkelstein, M. W., Stegink, L. D., and Applebaum, A.
(34 mg/kg body wt) plus MSG (34 mg/kg body wt) with a E. (1985). Correlation of glutamate plus aspartate dose, plasma
high protein (1 g/kg body wt) meal in healthy adults had amino acid concentration and neuronal necrosis in infant mice.
little effect on plasma concentrations of either aspartate Food Chem. Toxicol. 23, 887893.
or glutamate beyond those from the meal alone. In an- Filer, L. J., Jr., Baker, G. L., and Stegink, L. D. (1983). Effect of as-
other study, Stegink et al. (1983) demonstrated that in- partame loading on plasma and erythrocyte free amino acid con-
centrations in one-year-old infants. J. Nutr. 113, 15911599.
gestion of MSG (150 mg/kg body wt) with a high-protein
Finkelstein, M. W., Daabees, T. T., Stegink, L. D., and Applebaum, A.
meal significantly increased mean peak plasma gluta- E. (1983). Correlation of aspartate dose, plasma dicarboxylic amino
mate plus aspartate concentrations above those from acid concentration and neuronal necrosis in infant mice. Toxicology
the meal alone; the addition of aspartame (23 mg/kg 29, 109119.
body wt) to the meal with MSG did not further increase Lemkey-Johnston, N., and Reynolds, W. A. (1974). Nature and extent
mean peak plasma glutamate plus aspartate concen- of brain lesions in mice related to ingestion of monosodium gluta-
trations over the meal plus MSG. Stegink et al. (1987) mate. J. Neuropathol. Exp. Neurol. 33, 7497.
also evaluated the effect of MSG (50 mg/kg body wt) and Leon, A. S., Hunninghake, D. B., Bell, C., Rassin, D. K., and Tephly,
T. R. (1989). Safety of long-term large doses of aspartame. Arch.
MSG (50 mg/kg body wt) plus aspartame (34 mg/kg body Intern. Med. 149, 23182324.
wt) on plasma amino acid concentrations when given a Meldrum, B. (1993). Amino acids as dietary excitotoxins: A contribu-
low-protein meal. The addition of MSG or MSG plus as- tion to understanding neurodegenerative disorders. Brain Res. Rev.
partame to the meal resulted in significant increases in 18, 293314.
plasma glutamate, aspartate, and glutamate plus as- Newman, A. J., Heywood, R., Palmer, A. K., Barry, D. H., Edwards,
partate concentrations over the meal alone. Although F. P., and Worden, A. N. (1973). The administration of monosodium
L-glutamate to neonatal and pregnant rhesus monkeys. Toxicology
the addition of aspartame to MSG resulted in a small
1, 197204.
but significant increase in plasma aspartate concentra-
Okaniwa, A., Hori, M., Masuda, M., Takeshita, M., Hayashi, N., and
tions over the meal plus MSG, the area under the curve Wada, I. (1979). Histopathological study on effects of potassium
(AUC) for plasma aspartate plus glutamate concentra- aspartate on the hypothalamus of rats. J. Toxicol. Sci. 4, 3145.
tions did not differ between the MSG compared to MSG Olney, J. W. (1969). Brain lesions, obesity, and other disturbances in
plus aspartame conditions. The results of these studies mice treated with monosodium glutamate. Science 164, 719721.
demonstrate that it is simply impossible for a human Olney, J. W. (1976). Brain damage and oral intake of certain amino
to ever consume enough aspartame or MSG and aspar- acids. Adv. Exp. Biol. Med. 69, 497506.
tame together to raise plasma concentrations to those Olney, J. W. (1982). The toxic effects of glutamate and related com-
pounds in the retina and the brain. Retina 2, 341359.
associated with neurotoxicity in neonatal mice. Thus,
Olney, J. W., and Sharpe, L. G. (1969). Brain lesions in an infant
the data do not support the contention that aspartame rhesus monkey treated with monosodium glutamate. Science 166,
consumed with MSG would result in rapid and danger- 386388.
ous rises in plasma aspartate and glutamate concentra- Olney, J. W., Sharpe, L. G., and Feigin, R. D. (1972). Glutamate-
tions. induced brain damage in infant primates. J. Neuropathol. Exp. Neu-
rol. 31, 464488.
Reif-Lehrer, L. (1976). Possible significance of adverse reactions to
Conclusion glutamate in humans. Fed. Proc. 35, 22052212.
Thus, based on extensive metabolism data for exci- Reynolds, W. A., Bulter, V., and Lemkey-Johnston, N. (1976). Hypotha-
lamic morphology following ingestion of aspartame or MSG in the
tatory amino acids in humans (reviewed by Meldrum, neonatal rodent and primate: A preliminary report. J. Toxicol. En-
1993), it is not possible for a human to consume enough viron. Health 2, 471480.
aspartame or MSG and aspartame together in food and Reynolds, W. A., Lemkey-Johnston, N., and Stegink, L. D. (1979). Mor-
drink products to raise plasma concentrations of aspar- phology of the fetal monkey hypothalamus after in utero exposure
S28 BUTCHKO ET AL.
to monosodium glutamate. In Glutamic Acid: Advances in Biochem- Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1981b). Effect of
istry and Physiology (L. J. Filer, Jr., S. Garattini, M. R. Kare, W. A. aspartame and sucrose loading in glutamate-susceptible subjects.
Reynolds, and R. J. Wurtman, Eds.), pp. 217229. Raven Press, New Am. J. Clin. Nutr. 34, 18991905.
York. Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1982). Effect of as-
Reynolds, W. A., Stegink, L. D., Filer, L. J., Jr., and Renn, E. (1980). As- partame plus monosodium L-glutamate ingestion on plasma and
partame administration to the infant monkey: Hypothalamic mor- erythrocyte amino acid levels in normal adult subjects fed a high
phology and plasma amino acid levels. Anat. Rec. 198, 7384. protein meal. Am. J. Clin. Nutr. 36, 11451152.
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1977). Effect of as- Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1983). Plasma
partame and aspartate loading upon plasma and erythrocyte free amino acid concentrations in normal adults fed meals with added
amino acid levels in normal adult volunteers. J. Nutr. 107, 1837 monosodium L-glutamate and aspartame. J. Nutr. 113, 18511860.
1845. Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1987). Plasma
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1979). Plasma, erythro- amino acid concentrations in normal adults ingesting aspartame
cyte and human milk levels of free amino acids in lactating women and monosodium L-glutamate as part of a soup/beverage meal.
administered aspartame or lactose. J. Nutr. 109, 21732181. Metabolism. 36, 10731079.
Stegink, L. D., Filer, L. J., Jr., and Baker, G. L. (1981a). Plasma and Wen, C.-P., Hayes, K. C., and Gershoff, S. N. (1973). Effects of dietary
erythrocyte concentrations of free amino acids in adult humans supplementation of monosodium glutamate on infant monkeys,
administered abuse doses of aspartame. J. Toxicol. Environ. Health weanling rats, and suckling mice. Am. J. Clin. Nutr. 26, 803
7, 291305. 813.
Phenylalanine and Neurochemistry
Introduction partame on brain function, e.g., headaches, seizures,
Phenylalanine is an essential amino acid required for and changes in behavior, cognition, and mood. The
normal growth and development and maintenance of underlying hypothesis was that aspartame, as a source
life. However, individuals with the rare genetic disease of phenylalanine (Phe) without the other large neu-
phenylketonuria (PKU) must be placed on special di- tral amino acids (LNAAs) (i.e., tyrosine, tryptophan,
ets severely restricted in phenylalanine content shortly valine, leucine, isoleucine, methionine, histidine) that
after birth in order to avoid the markedly increased compete for transport across the bloodbrain barrier,
plasma phenylalanine concentrations associated with would increase the serum ratio of phenylalanine to the
well-known sequelae of PKU, including mental retar- other large neutral amino acids (Phe/LNAA), thereby
dation or various degrees of cognitive impairment. For selectively increasing brain phenylalanine concentra-
healthy individuals and individuals heterozygous for tions. It was further hypothesized that such increased
PKU (PKUH), there is no need for dietary restriction of entry of phenylalanine into the brain, along with de-
phenylalanine. creased entry of tyrosine and tryptophan, may result
As aspartame is a source of phenylalanine, questions in disturbances in brain neurotransmitter concentra-
arose regarding the impact of aspartame on plasma tions (Wurtman, 1983). Such neurochemical changes
phenylalanine concentrations (see Metabolism of As- had been reported in rodents (Yokogoshi et al., 1984),
partame section). From the metabolism studies with and there were several case reports describing a poten-
aspartame, amounts of aspartame about 10 times the tial association between seizures and aspartame con-
amount that people consume at the 90th percentile re- sumption (Wurtman, 1985; Walton, 1986).
sult in plasma phenylalanine concentrations within the Upon dietary ingestion of protein and its digestion,
normal postprandial range in both healthy individuals amino acids are absorbed across the gastrointestinal
and PKUH individuals. mucosa into the portal circulation. In the liver, the
In addition, the typical diet provides much greater enzyme phenylalanine hydroxylase converts phenylal-
amounts of phenylalanine than provided by aspartame. anine to tyrosine. In humans, most of the dietary phe-
For example, a serving of no-fat milk provides about six nylalanine passes through the portal circulation and
times more phenylalanine than an equivalent volume the liver unchanged into the systemic circulation. Once
of beverage sweetened 100% with aspartame (Fig. 1). in the systemic circulation, phenylalanine is taken up
Further, aspartame provides only about 3 and 1% of across the bloodbrain barrier into the central nervous
daily dietary intake of phenylalanine for an adult and system through a transport system that is relatively
a 4-year-old child, respectively, at the 90th percentile specific for LNAAs (Fernstrom and Wurtman, 1972).
of aspartame intake (see Intake of Aspartame vs the Therefore, the relative concentrations of these amino
Acceptable Daily Intake section). acids, together with their specific affinity constants to
Nonetheless, some claims regarding aspartame have the carrier system, will ultimately determine how much
centered on possible effects of phenylalanine from as- of a given amino acid will enter and leave the brain.
In that context, more plasma phenylalanine com-
pared to the other LNAAs will increase Phe/LNAA. Un-
1600 der these circumstances, phenylalanine may thereby in-
1400 terfere with the availability of tyrosine and tryptophan
1200 1130 in the brain and also may act as a competitive inhibitor
1000 of the enzyme tyrosine hydroxylase (Ikeda et al., 1967).
824
mg
(from 4.45 1.29 to 5.80 0.95 mol/dl) and in PKU serotonin, 5-hydroxyindole acetic acid (5-HIAA), nora-
heterozygotes (from 6.88 1.38 to 8.23 1.73 mol/dl) drenaline, -aminobutyric acid (GABA), or dopamine
but did not increase the already elevated concentrations and its metabolites homovanillic acid (HVA) or 3,4-
in mild hyperphenylalaninemics and untreated PKU dihydroxyphenylacetic acid (DOPAC).
homozygotes. Baseline Phe/LNAA ratios in these sub- In humans, Romano et al. (1989) also evaluated the
jects were 0.100, 0.170, 0.849, and 4.212, respectively. effect of doses of aspartame that are more relevant to
One hour after administration of aspartame, these val- human consumption (0.83 and 8.3 mg/kg body wt or the
ues were 0.137, 0.214, 0.892, and 3.847, respectively. amount in up to 1 liter of beverage sweetened 100% with
The small increases observed in the controls and PKU aspartame for a 60-kg individual), with a low or high
heterozygotes, although statistically significant, lack carbohydrate meal, on plasma LNAA concentrations.
biological relevance. There was no significant difference when aspartame
Fluctuations in Phe/LNAA occur as a result of dietary was given alone or with carbohydrate. To evaluate
manipulations or normal dietary variability. For ex- this issue further, Wolf-Novak et al. (1990) compared
ample, sugar or other carbohydrates result in changes the effect on plasma large amino acids concentrations
in Phe/LNAA (Martin-Du Pan et al., 1982; Stegink and Phe/LNAA in humans after unsweetened beverage,
et al., 1987; Wolf-Novak et al., 1990; Burns et al., 1991). beverage with 60 g carbohydrate (the amount of carbo-
In this case, the consequent insulin release, which is hydrate in a snack, such as a piece of cake with ice
known to enhance the transport of amino acids across cream), beverage with aspartame (200 mg or approxi-
muscle membranes, decreases plasma amino acid con- mately the amount in one 355 ml can of beverage), and
centrations, especially the LNAAs valine, leucine, and beverage with 60 g of carbohydrate plus 200 mg of as-
isoleucine. Phenylalanine from aspartame will increase partame in both normal adults and PKU individuals.
the numerator, whereas insulin will result in a decrease There were no effects of beverage with aspartame on
in the denominator of Phe/LNAA. The net effect after plasma phenylalanine or other LNAA concentrations.
equisweet amounts of sugar and aspartame is similar However, the beverage with carbohydrate, with or with-
in magnitude. Thus, even if there are small changes out aspartame, resulted in decreases in plasma con-
in Phe/LNAA after aspartame, such changes are not centrations of valine, isoleucine, and leucine due to the
unique to aspartame. insulin response after carbohydrate, resulting in an in-
Considering the above data and assuming that there crease in Phe/LNAA. The authors concluded that the
is no dramatic effect due to accommodation, the thresh- changes in Phe/LNAA were due to carbohydrate and
old for adverse effects, based on the Phe/LNAA ra- not aspartame. Stegink et al. (1990) evaluated the ef-
tio, can be estimated to be substantially higher than fect of aspartame (40 mg/kg body wt) vs aspartame plus
that observed in mild hyperphenylalaninemics, i.e., sucrose (1.2 g/kg body wt) on plasma amino acid concen-
0.85. Mild hyperphenylalaninemics are clinically nor- trations and Phe/LNAA. Plasma phenylalanine concen-
mal and do not require phenylalanine-restricted diets. trations and 4-h AUC increased after both treatments
Thus, as with absolute plasma phenylalanine concen- but did not differ significantly. Mean peak Phe/LNAA
trations, the aspartame dosage required to elevate the did not differ between treatments, and the 4-h AUC was
Phe/LNAA ratio above that associated with adverse only slightly but significantly higher after aspartame
effects far exceeds that possible from consumption of plus sucrose vs aspartame. The authors concluded that
products sweetened with aspartame. simultaneous ingestion of aspartame and sucrose had
It was also hypothesized that the most prolonged dis- only minor effects on plasma Phe/LNAA. Thus, based
turbance of the Phe/LNAA ratio would likely occur after on both animal and human data, there is no significant
the consumption of large doses of aspartame together effect of aspartame at doses relevant to actual consump-
with a load of carbohydrates (Yokogoshi et al., 1984). tion on Phe/LNAA even when given with carbohydrate.
In rodent studies, it was observed that large doses Another important consideration in the overall im-
of aspartame may attenuate carbohydrate-induced in- pact of a given phenylalanine load in studies in ani-
creases in brain tryptophan and serotonin concentra- mals is the fact that interspecies differences exist in
tions (Yokogoshi et al., 1984; Fernstrom et al., 1986). the extent of its hydroxylation to tyrosine. In rodents,
However, Romano et al. (1989) evaluated the effect the rate of liver conversion of phenylalanine to tyro-
of both acute and subacute dosing of aspartame (up sine is greater than the rate of conversion in humans.
to 250 mg/kg body wt/day), with and without carbo- Therefore, in rodents, more of the dietary phenylala-
hydrate, on plasma and brain LNAA and brain neu- nine is presented to the systemic circulation as tyro-
rotransmitter concentrations in rats. Although there sine when compared to humans, which may result in
were changes in plasma and brain amino acid concen- an increased uptake of tyrosine across the bloodbrain
trations in the acute dosing paradigm, there were no dif- barrier and thereby raise the concentration and avail-
ferences in brain LNAA concentrations after aspartame ability of tyrosine. Wurtman and Maher (1987) postu-
compared to aspartame plus carbohydrate in the suba- lated that because rodents metabolize phenylalanine
cute study. There were no effects of aspartame on brain to tyrosine faster than humans and as tyrosine may act
ASPARTAME: REVIEW OF SAFETY S31
as an antidote to the effects of phenylalanine on brain hensive review of the studies on aspartame and neuro-
neurochemistry, brain phenylalanine flux may be better chemistry has been published by Lajtha et al. (1994).
estimated from a new ratio, Phet /Pheo /Tyrt /Tyro , which Most of these studies were acute dosing paradigms
is based on relative changes in the phenylalanine (Phe) (Fernstrom et al., 1983; Yokogoshi et al., 1984;
and tyrosine (Tyr) concentrations. From this, the au- Yokogoshi and Wurtman, 1986; Coulombe and Sharma,
thors estimated that there may be a 60-fold difference 1986; Perego et al., 1988; Garattini et al., 1988;
in this rate between rodents and humans, thus ques- Romano et al., 1989, 1990; Dailey et al., 1989, 1991;
tioning whether the doses of aspartame used in animal Freeman et al., 1990; Diomede et al., 1991; Helali et al.,
studies of effects of aspartame on brain neurotransmit- 1996; Goerss et al., 2000), and a few were subacute
ters and brain function were sufficiently large enough (Romano et al., 1989; Helali et al., 1996) or subchronic
to account for this difference in metabolism. However, (Sharma and Coulombe, 1987; Reilly et al., 1989a;
Fernstroms (1989) evaluation of this hypothesis re- Dailey et al., 1991). In the acute dosing studies, the
vealed that it is not consistent with well-established doses varied from 13 to 2500 mg/kg body wt. Most stud-
notions of competitive LNAA transport. He concluded, ies were oral/gavage-dosing paradigms, whereas some
An analysis of the logic behind this postulation reveals (e.g., Goerss et al., 2000) used intraperitoneal (ip) dos-
there to be no basis for accepting the higher dose con- ing. In the subacute dosing studies (914 days), doses
version of 60 between rat and man. ranged up to 250 mg/kg body wt/day. In the subchronic
Further, in a study done to evaluate this hypothesis, dosing studies (up to 30 days), doses of aspartame
it was shown that rodents require a dose of aspartame ranged up to 863 mg/kg body wt/day.
only two to six times higher than humans to produce In most of these studies, concentrations of mono-
similar increases in plasma Phe/LNAA (Hjelle et al., amine neurotransmitters and their metabolites were
1992). Thus, even with this difference in metabolism, measured in structures such as the cortex, hip-
the doses of aspartame (up to 3000 mg/kg body wt) pocampus, striatum, midbrain, thalamus, and nucleus
used in the animal studies of effects of aspartame on accumbens. Whole brain concentrations of the neuro-
brain neurotransmitters, seizures, behavior, etc., are transmitters and metabolites were evaluated in other
very large compared to doses evaluated in humans studies. While isolated effects of aspartame on brain
(up to 200 mg/kg body wt) as a bolus and 1000 times neurotransmitter or neurotransmitter metabolite con-
90th percentile average daily consumption of aspar- centrations were noted in a few studies (Yokogoshi and
tame (3 mg/kg body wt) in the United States (Butchko Wurtman, 1986; Coulombe and Sharma, 1986; Sharma
and Kotsonis, 1991). and Coulombe, 1987; Dailey et al., 1989, 1991; Freeman
In addition to the studies in animals and hypothe- et al., 1990; Helali et al., 1996; Goerss et al., 2000), it was
ses, Koeppe et al. (1991) evaluated amino acid uptake clear that these effects were neither consistent nor re-
into the brain in humans using positron emission scan- producible. For the most part, despite large increases in
ning. These authors demonstrated an 11.5% decrease brain phenylalanine concentrations and increases in
in amino acid transport rate constant into the brain af- Phe/LNAA, about half of the few reported changes
ter a large (34 mg/kg body wt) oral dose of aspartame in brain concentrations of neurotransmitters or their
(about 10 times 90th percentile daily consumption of metabolites were increases rather than the hypothe-
aspartame). Based on this, the authors concluded that sized decreases.
there would be little measurable transport change un- In addition, in 30-day subchronic studies (Reilly et al.,
der more relevant conditions of aspartame use. 1989b, 1990; Reilly and Lajtha, 1995a), aspartame (up
to 500 mg/kg body wt/day) had no effect on adrener-
gic, serotoninergic, or dopaminergic receptor kinetics in
Studies of Aspartame and Adrenergic, Serotoninergic,
young adult rats or in rats exposed to aspartame dur-
and Dopaminergic Neurotransmitter Systems
ing gestation and lactation. In an acute in vivo voltam-
A number of animal studies have been done to de- metry study in freely moving rats (Perego et al., 1988;
termine whether plasma phenylalanine concentration Garattini et al., 1988), recordings for DOPAC in the
increases secondary to aspartame loading may result striatum and the nucleus accumbens and for 5-HIAA
in changes in concentration or turnover of the neuro- in these same brain areas as well as the hippocampus
transmitters, norepinephrine, dopamine, or serotonin, were no different in rats given 1000 mg/kg body wt of
and their metabolites in whole brain or specific re- aspartame compared to control. Further, a study us-
gions of the brain (Table 1) (Fernstrom et al., 1983; ing transstriatal dialysis demonstrated no significant
Yokogoshi et al., 1984; Yokogoshi and Wurtman, 1986; effect of phenylalanine (100 and 450 mg/kg body wt) on
Coulombe and Sharma, 1986; Sharma and Coulombe, release of dopamine or levels of its metabolites in either
1987; Perego et al., 1988; Garattini et al., 1988; Reilly rat or baboon striatum (Arvin et al., 1992). Such stud-
et al., 1989a, 1990; Romano et al., 1989, 1990; Dailey ies are considered more reliable in evaluating changes
et al., 1989, 1991; Freeman et al., 1990; Diomede et al., in neurotransmission than steady state neurotransmit-
1991; Helali et al., 1996; Goerss et al., 2000). A compre- ter concentrations. All these studies failed to show a
S32 BUTCHKO ET AL.
TABLE 1
Brain Neurotransmitter and Neurotransmitter Metabolite Concentrations
in Animals after Aspartame Dosing
Norepinephrine Fernstrom et al. (1983) Yokogoshi and Wurtman (1986) Yokogoshi and Wurtman (1986)
Yokogoshi and Wurtman (1986) Coulombe and Sharma (1986) Helali et al. (1996)
Sharma and Coulombe (1987) Dailey et al. (1991)
Perego et al. (1988)
Romano et al. (1989)
Reilly et al. (1989a)
Dailey et al. (1989)
Freeman et al. (1990)
Reilly et al. (1990)
Diomede et al. (1991)
Helali et al. (1996)
Dopamine Fernstrom et al. (1983) Yokogoshi and Wurtman (1986)
Yokogoshi and Wurtman (1986) Coulombe and Sharma (1986)
Sharma and Coulombe (1987) Dailey et al. (1991)
Garattini et al. (1988)
Perego et al. (1988)
Romano et al. (1989)
Reilly et al. (1989a)
Dailey et al. (1989)
Romano et al. (1990)
Freeman et al. (1990)
Reilly et al. (1990)
Diomede et al. (1991)
Goerss et al. (2000)
Serotonin Fernstrom et al. (1983) Sharma and Coulombe (1987) Sharma and Coulombe (1987)
Yokogoshi et al. (1984) Goerss et al. (2000) Dailey et al. (1989)
Coulombe and Sharma (1986) Dailey et al. (1991)
Perego et al. (1988)
Garattini et al. (1988)
Romano et al. (1989)
Reilly et al. (1989a)
Reilly et al. (1990)
Freeman et al. (1990)
Diomede et al. (1991)
5HIAAa Yokogoshi et al. (1984) Goerss et al. (2000) Dailey et al. (1989)
Coulombe and Sharma (1986) Sharma and Coulombe (1987)
Romano et al. (1989) Goerss et al. (2000)
Perego et al. (1988)
Freeman et al. (1990)
Diomede et al. (1991)
Fernstrom et al. (1983)
Garattini et al. (1988)
Reilly et al. (1989a)
Reilly et al. (1990)
Goerss et al. (2000)
HVAa Fernstrom et al. (1983) Coulombe and Sharma (1986)
Sharma and Coulombe (1987)
Garattini et al. (1988)
Perego et al. (1988)
Romano et al. (1989)
Reilly et al. (1989a)
Freeman et al. (1990)
Romano et al. (1990)
Reilly et al. (1990)
Diomede et al. (1991)
Goerss et al. (2000)
Dopa Fernstrom et al. (1983)
ASPARTAME: REVIEW OF SAFETY S33
TABLE 1Continued
DOPACa Fernstrom et al. (1983) Yokogoshi and Wurtman (1986) Yokogoshi and Wurtman (1986)
Perego et al. (1988) Coulombe and Sharma (1986) Sharma and Coulombe (1987)
Garattini et al. (1988) Freeman et al. (1990)
Romano et al. (1989)
Reilly et al. (1989a)
Romano et al. (1990)
Reilly et al. (1990)
Diomede et al. (1991)
Goerss et al. (2000)
VMAa Sharma and Coulombe (1987) Coulombe and Sharma (1986)
MHPGa Yokogoshi and Wurtman (1986) Yokogoshi and Wurtman (1986)
Garattini et al. (1988)
Romano et al. (1989)
GABAa Romano et al. (1989) Helali et al. (1996)
Helali et al. (1996)
Epinephrine Yokogoshi and Wurtman (1986) Yokogoshi and Wurtman (1986) Yokogoshi and Wurtman (1986)
Helali et al. (1996) Helali et al. (1996)
Aspartate Reilly and Lajtha (1995a) Helali et al. (1996) Reilly and Lajtha (1995a)
Reilly and Lajtha (1995b) Reilly and Lajtha (1995b)
Helali et al. (1996)
Glutamate Helali et al. (1996) Reilly and Lajtha (1995a)
Reilly and Lajtha (1995b)
a 5HIAA, 5-hydroxyindole acetic acid; HVA, homovanillic acid; DOPAC, 3,4-dihydroxyphenylacetic acid; VMA, 3-methoxy-4-
treatment effect of large doses of aspartame or pheny- phenylalanine to tyrosine, and thereby more closely
lalanine. mimic the human condition. Although the concentra-
The potential effect of aspartame on neurotransmit- tion of retinal phenylalanine increased about sixfold,
ter systems in the developing brain, which was thought there were no significant changes in tyrosine concen-
to be more sensitive, was also studied. Rats were trations and no changes in the rate of tyrosine hydrox-
exposed to aspartame administered to dams through- ylation or dopamine concentrations in the retina. Thus,
out gestation and lactation. No effects were observed enormous doses of aspartame did not inhibit or stim-
on adrenergic, serotoninergic, or dopaminergic neuro- ulate the in vivo rate of tyrosine hydroxylation in the
transmitter concentrations or neurotransmitter recep- retina.
tor kinetics in the dams or the weanlings (Reilly et al.,
1990; Reilly and Lajtha, 1995a).
Studies of the Aspartate and Glutamate
In an in vitro study, Fountain et al. (1988) noted that
Neurotransmitter Systems
aspartame, aspartic acid, phenylalanine, and pheny-
lalanine methyl ester potentiated the response of hip- Helali et al. (1996) observed increased brain aspar-
pocampal CA 1 pyramidal cells; however, there was no tate concentrations in mice after acute aspartame dos-
effect on inhibitory systems or the processes underly- ing with 1000 mg/kg body wt but no changes in brain
ing synaptic plasticity. (Suppression of inhibitory pro- glutamate or GABA concentrations. However, with sub-
cesses is commonly associated with convulsants.) The acute dosing (100 mg/kg body wt/day for 14 days),
authors concluded that aspartame and its moieties are these authors found no changes in brain aspartate
not likely to produce significant neurotoxic effects at or glutamate concentrations but a significant reduc-
physiologically realistic acute doses. tion in GABA. In another study, Romano et al. (1989)
Further, regarding the conjecture that phenylalanine found no effect of aspartame on rat cortical GABA con-
from aspartame may inhibit tyrosine hydroxylase ac- centrations, with or without a carbohydrate-enriched
tivity and thereby impact brain monoaminergic neu- diet, over 14 days. From an in vitro study, Pan-Hou
rotransmitters, Fernstrom et al. (1991) evaluated the et al. (1990) suggested that aspartame, and especially
effects of large doses of aspartame up to 1500 mg/kg its aspartate component, inhibited glutamate binding
body wt on the rate of tyrosine hydroxylation in rat to N-methyl-D-aspartate (NMDA) receptors in a dose-
retina. In this study, rats were pretreated with - dependent manner, altering the affinities for the recep-
chlorophenylalanine to inhibit hepatic metabolism of tors without altering Vmax . From another in vitro study,
S34 BUTCHKO ET AL.
Sonnewald et al. (1995) reported a dose-dependent in- sumption in phenylketonuria, mild hyperphenylalaninemia, and
crease in calcium influx and leakage of lactate dehydro- heterozygous state for phenylketonuria. J. Pediatr. 109, 668671.
genase from murine brain cell cultures, which he felt Coulombe, R. A., Jr., and Sharma, R. P. (1986). Neurobiochemical
alterations induced by the artificial sweetener aspartame (Nu-
confirmed the hypothesis of Pan-Hou regarding effects traSweet). Toxicol. Appl. Pharmacol. 83, 7985.
on the NMDA receptor. However, the relevance of such Dailey, J. W., Lasley, S. M., Mishra, P. K., Bettendorf, A. F.,
in vitro work is questionable since brain cells are not Burger, R. L., and Jobe, P. C. (1989). Aspartame fails to facilitate
exposed to intact aspartame after consumption given pentylenetetrazol-induced convulsions in CD-1 mice. Toxicol. Appl.
its rapid metabolism to its three components before ab- Pharmacol. 98, 475486.
sorption. In contrast, Reilly and Lajtha (1995a,b) eval- Dailey, J. W., Lasley, S. M., Burger, R. L., Bettendorf, A. F., Mishra,
uated the effect of aspartame 500 mg/day for 30 days P. K., and Jobe, P. C. (1991). Amino acids, monoamines and
audiogenic-seizures in genetically epilepsy-prone rats: Effects of
on NMDA and total glutamatergic receptor binding ki- aspartame. Epilepsy Res. 8, 122133.
netics in adult dams and weanlings, which had been ex- Diomede, L., Romano, M., Guiso, G., Caccia, S., Nava, S., and
posed to aspartame throughout gestation and lactation. Salmona, M. (1991). Interspecies and interstrain studies on the in-
Although there were small but reversible decreases in creased susceptibility to metrazol-induced convulsions in animals
brain concentrations of glutamic acid (cortex and hip- given aspartame. Food Chem. Toxicol. 29, 101106.
pocampus) and aspartic acid (cortex only) in weanlings, Fernstrom, J. D. (1989). Oral aspartame and plasma phenylalanine:
Pharmacokinetic difference between rodents and man, and rele-
there were no effects on the kinetics of binding to the
vance to CNS effects of phenylalanine. J. Neural Transm. 75, 159
NMDA receptor and total glutamate binding in the cere- 164.
bral cortex and hippocampus after perinatal exposure Fernstrom, J. D., and Wurtman, R. J. (1972). Brain serotonin content:
to high doses of aspartame. Physiological regulation by plasma neutral amino acids. Science
178, 414416.
Fernstrom, J. D., Fernstrom, M. H., and Gillis, M. A. (1983). Acute
Conclusion effects of aspartame on large neutral amino acids and monoamines
Numerous studies have been done to evaluate the in rat brain. Life Sci. 32, 16511658.
hypothesis that aspartame, as a source of phenylala- Fernstrom, J. D., Fernstrom, M. H., and Grubb, P. E. (1986). Effects
of aspartame ingestion on the carbohydrate- induced rise in trypto-
nine without the other large neutral amino acids that phan hydroxylation rate in rat brain. Am. J. Clin. Nutr. 44, 195205.
compete with it for entry into the brain, may result in Fernstrom, J. D., Fernstrom, M. H., and Massoudi, M. S. (1991).
changes in brain neurotransmitters, which may thereby In vivo tyrosine hydroxylation in rat retina: Effect of aspartame in-
alter brain function. It was clear from these studies that gestion in rats pretreated with p-chlorophenylalanine. Am. J. Clin.
there was no consistent pattern of such effects nor was Nutr. 53, 923929.
there evidence of reproducibility. For the most part, de- Fountain, S. B., Hennes, S. K., and Teyler, T. J. (1988). Aspartame
exposure and in vitro hippocampal slice excitability and plasticity.
spite large increases in brain phenylalanine concentra-
Fundam. Appl. Toxicol. 11, 221228.
tions, about half of the few reported changes in brain
Freeman, G., Sobotka, T., and Hattan, D. (1990). Acute effects of as-
catecholamine concentrations were increases rather partame on concentrations of brain amines and their metabolites
than the hypothesized decreases. Further, there was no in selected brain regions of Fischer-344 and SpragueDawley rats.
effect on receptor kinetics after very large doses of as- Drug Chem. Toxicol. 13, 113133.
partame in either adult animals or weanlings exposed Garattini, S., Caccia, S., Romano, M., Diomede, L., Guiso, G.,
to aspartame throughout gestation and lactation or on Vezzani, A., and Salmona, M. (1988). Studies on the susceptibil-
ity to convulsions in animals receiving abuse doses of aspartame.
release of neurotransmitters. There also were no con- In Dietary Phenylalanine and Brain Function (R. J. Wurtman and
sistent effects on brain glutamate, aspartate, or GABA E. Ritter-Walker, Eds.), pp. 131143. Birkhauser, Boston.
concentrations or on binding to NMDA receptors. Thus, Goerss, A. L., Wagner, G. C., and Hill, W. L. (2000). Acute effects of
it is concluded that consumption of aspartame does not aspartame on aggression and neurochemistry of rats. Life Sci. 67,
adversely affect neurochemical function of the brain. 13251329.
Helali, N. Y., El-Kashef, H., Salem, H., Gamiel, N., and Elmazar,
M. M. A. (1996). The effect of aspartame on seizure susceptibil-
REFERENCES ity and the anticonvulsant action of ethosuximide, valproate and
phenytoin in mice. Saudi Pharm. J. 4, 149156.
Arvin, B., Chapman, A. G., and Meldrum, B. S. (1992). Extracellular Hjelle, J. J., Dudley, R. E., Marietta, M. P., Sanders, P. G., Dickie, B. C.,
levels of monoamines in striatum of baboon and rat following oral Brisson, J., and Kotsonis, F. N. (1992). Plasma concentrations and
phenylalanine. Soc. Neurosci. Abstr. 18, 24. pharmacokinetics of phenylalanine in rats and mice administered
Burns, T. S., Stargel, W. W., Tschanz, C., Kotsonis, F. N., and Hurwitz, aspartame. Pharmacology 44, 4860.
A. (1991). Aspartame and sucrose produce a similar increase in the Ikeda, M., Levitt, M., and Udenfriend, S. (1967). Phenylalanine as
plasma phenylalanine to large neutral amino acid ratio in healthy- substrate and inhibitor of tyrosine hydroxylase. Arch. Biochem. Bio-
subjects. Pharmacology 43, 210219. phys. 120, 420427.
Butchko, H. H., and Kotsonis, F. N. (1991). Acceptable daily intake Koeppe, R. A., Shulkin, B. L., Rosenspire, K. C., Shaw, L. A., Betz,
vs actual intake: The aspartame example. J. Am. Coll. Nutr. 10, A. L., Mangner, T., Price, J. C., and Agranoff, B. A. (1991). Effect of
258266. aspartame-derived phenylalanine on neutral amino acid uptake in
Caballero, B., Mahon, B. E., Rohr, F. J., Levy, H. L., and Wurtman, R. J. human brain: A positron emission tomography study. J. Neurochem.
(1986). Plasma amino acid levels after single-dose aspartame con- 56, 15261535.
ASPARTAME: REVIEW OF SAFETY S35
Lajtha, A., Reilly, M. A., and Dunlop, D. S. (1994). Aspartame con- doses of aspartame on serotonin and its metabolite in various re-
sumption: Lack of effects on neural function. J. Nutr. Biochem. 5, gions of the mouse brain. Food Chem. Toxicol. 25, 565568.
266283. Sonnewald, U., Muller,
T., Unsgard, G., and Peterson, S. B. (1995).
Maher, T. J., and Wurtman, R. J. (1987). Possible neurologic effects of Effects of aspartame on 45 Ca influx and LDH leakage from nerve
aspartame, a widely used food additive. Environ. Health Perspect. cells in culture. NeuroReport 6, 318320.
75, 5357. Stegink, L. D., and Filer, L. J., Jr. (1996a). Effects of aspartame in-
Martin-Du Pan, R., Mauron, C., Glaeser, B., and Wurtman, R. J. gestion on plasma aspartate, phenylalanine, and methanol concen-
(1982). Effect of various oral glucose doses on plasma neutral amino trations in normal adults. In The Clinical Evaluation of a Food
acid levels. Metabolism 31, 937943. Additive: Assessment of Aspartame (C. Tschanz, H. H. Butchko,
Mller, S. E. (1991). Effect of aspartame and protein, administered W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 6786. CRC Press,
in phenylalanine-equivalent doses, on plasma neutral amino acids, Boca Raton, FL.
aspartate, insulin and glucose in man. Pharmacol. Toxicol. 68, 408 Stegink, L. D., and Filer, L. J., Jr. (1996b). Effects of aspartame in-
412. gestion on plasma aspartate, phenylalanine, and methanol concen-
Pan-Hou, H., Suda, Y., Ohe, Y., Sumi, M., and Yoshioka, M. trations in potentially sensitive populations. In The Clinical Eval-
(1990). Effect of aspartame on N-methyl-D-aspartate- sensitive uation of a Food Additive: Assessment of Aspartame (C. Tschanz,
3
L-[ H]glutamate binding sites in rat brain synaptic membranes. H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 87113.
Brain Res. 520, 351353. CRC Press, Boca Raton, FL.
Perego, C., DeSimoni, M. G., Fodritto, F., Raimondi, L., Diomede, L., Stegink, L. D., Brummel, M. C., Persoon, T. J., Filer, L. J., Jr., Bell,
Salmona, M., Algeri, S., and Garattini, S. (1988). Aspartame and E. F., and Zeigler, E. E. (1990). Effect of sucrose on the metabolic
the rat brain monoaminergic system. Toxicol. Lett. 44, 331339. disposition of aspartame. Am. J. Clin. Nutr. 52, 335341.
Reilly, M. A., and Lajtha, A. (1995a). Aspartame does not affect amin- Stegink, L. D., Wolf-Novak, L. C., Filer, L. J., Jr., Bell, E. F.,
ergic and glutamatergic receptor kinetics in rat brain. Ann. N.Y. Zeigler, E. E., Krause, W. L., and Brummel, M. C. (1987).
Acad. Sci. 757, 311313. Aspartame-sweetened beverage: Effect on plasma amino acid con-
Reilly, M. A., and Lajtha, A. (1995b). Glutamatergic receptor kinetics centrations in normal adults and adults heterozygous for phenylke-
are not altered by perinatal exposure to aspartame. Neurochem. tonuria. J. Nutr. 117, 19891995.
Int. 26, 217222. Walton, R. G. (1986). Seizure and mania after high intake of aspar-
Reilly, M. A., Debler, E. A., and Lajtha, A. (1990). Perinatal exposure tame. Psychosomatics 27, 218, 220.
to aspartame does not alter aminergic neurotransmitter systems in Wolf-Novak, L. C., Stegink, L. D., Brummel, M. C., Persoon, T. J.,
weanling rat brain. Res. Commun. Psychol. Psychiatry Behav. 15, Filer, L. J., Jr., Bell, E. F., Ziegler, E. E., and Krause, W. L. (1990).
141159. Aspartame ingestion with and without carbohydrate in phenylke-
Reilly, M. A., Debler, E. A., Fleischer, A., and Lajtha, A. (1989a). tonuric and normal subjects: Effect on plasma concentrations of
Chronic aspartame ingestion does not alter cerebral levels of amin- amino acids, glucose, and insulin. Metabolism 39, 391396.
ergic neurotransmitters and related amino acids in the rat brain. Wurtman, R. J. (1983). Neurochemical changes following high-dose
Res. Commun. Psychol. Psychiatry Behav. 14, 287303. aspartame with dietary carbohydrates. N. Engl. J. Med. 309, 429
Reilly, M. A., Debler, E. A., Fleischer, A., and Lajtha, A. (1989b). Lack 430.
of effect of chronic aspartame ingestion on aminergic receptors in Wurtman, R. J. (1985). Aspartame: Possible effect on seizure suscep-
rat brain. Biochem. Pharmacol. 38, 43394341. tibility. Lancet 2, 1060.
Romano, M., Casacci, F., De Marchi, F., Pacei, T., Esteve, A., Wurtman, R. J., and Maher, T. J. (1987). Effects of oral aspartame
Lomuscio, G., Mennini, T., and Salmona, M. (1989). Effects of aspar- on plasma phenylalanine in humans and experimental rodents.
tame and carbohydrate administration on human and rat plasma J. Neural Transm. 70, 169173.
large neutral amino acid levels and rat brain amino acid and Yokogoshi, H., and Wurtman, R. J. (1986). Acute effects of oral or par-
monoamine levels. J. Nutr. 119, 7581. enteral aspartame on catecholamine metabolism in various regions
Romano, M., Diomede, L., Guiso, G., Caccia, S., Perego, C., and of rat brain. J. Nutr. 116, 356364.
Salmona, M. (1990). Plasma and brain kinetics of large neutral Yokogoshi, H., Roberts, C. H., Caballero, B., and Wurtman, R. J.
amino acids and of striatum monoamines in rats given aspartame. (1984). Effects of aspartame and glucose administration on brain
Food Chem. Toxicol. 28, 317321. and plasma levels of large neutral amino acids and brain 5-
Sharma, R. P., and Coulombe, R. A., Jr. (1987). Effects of repeated hydroxyindoles. Am. J. Clin. Nutr. 40, 17.
Safety of Methanol from Aspartame and the Diet
Introduction than that produced by ethanol, followed by a la-
Aspartame metabolism yields approximately 10% tent period of 1030 h. After this asymptomatic period,
methanol by weight, and the safety of methanol from a syndrome develops that is characterized specifically
aspartame has been evaluated in numerous studies. by visual impairment (ranging from blurring to total
The extensive database of safety studies with aspar- loss of vision) and severe metabolic acidosis. In severe
tame includes studies in both laboratory animals and cases, CNS depression and coma can ensue, with death
humans, and 10% of the extremely large doses adminis- possible from respiratory depression.
tered can be considered to be methanol. The battery of There is a distinct lack of correlation between blood
preclinical toxicology studies consists of numerous tox- methanol levels and the degree of toxicity, suggest-
icity studies, including four 2-year or lifetime studies in ing that methanol itself is not a toxic agent. The ex-
rodents and including one in rats with in utero exposure istence of a latent period between the exposure to
followed by 104 weeks of testing, as well as reproduc- methanol and the onset of symptoms further suggests
tion, teratology, and mutagenicity studies (Kotsonis and that a toxic metabolite is accumulating during the la-
Hjelle, 1996). The no-observed-effect dose of aspartame tent period (Tephly, 1991). Initial experimental stud-
was at least 4000 mg/kg body wt /day as established by ies in animals failed to identify the toxic metabolite(s),
the Joint FAO/WHO Expert Committee on Food Addi- because they were conducted in nonprimate species
tives (1980), the Scientific Committee for Food (1985), that do not develop the syndrome of methanol poison-
and the Health Protection Branch of Health and Wel- ing that is seen in humans. However, monkeys have
fare Canada (1979). This dose of aspartame provides been shown to be susceptible to methanol in the same
400 mg/kg body wt/day of methanol and would be the way as humans (Martin-Amat et al., 1977). Studies us-
equivalent of a 60-kg human consuming 24,000 mg of ing fomepizole (4-methylpyrazole) as an inhibitor of
methanol every day over a lifetime. methanol metabolism in the monkey have confirmed
When evaluating potential toxicity of any substance, that methanol is toxic because of its conversion to a
the actual dose of the substance is of major importance. toxic metabolite (McMartin et al., 1975).
In the case of aspartame, a number of intake stud- Methanol is initially metabolized via alcohol dehy-
ies have been done over the past 15 years. The most drogenase (in humans and monkeys) to formaldehyde, a
extensive evaluation of aspartame intake was in the highly reactive chemical. Formaldehyde is then rapidly
United States from 1984 to 1992 where intake data degraded to formate by several enzyme systems, in-
were evaluated over 14-day periods from about 5000 cluding the specific formaldehyde dehydrogenase com-
people per year. In the United States, aspartame intake plex ubiquitously distributed among various tissues
by the general population at the 90th percentile is about (Uotila and Koivusalo, 1974). Because of its extensive
3.0 mg/kg body wt/day (Butchko and Kotsonis, 1991, metabolism and high reactivity, formaldehyde is elimi-
1994, 1996; Butchko et al., 1994; Butchko and Stargel, nated from the blood very rapidly in all studied species,
2001), well below the acceptable daily intake (ADI) in with half-lives in the range of 12 min (McMartin et al.,
the United States (50 mg/kg body wt/day) and in Europe 1979). In fact, no formaldehyde has been measured in
(40 mg/kg body wt/day). Taking into account differences blood or tissues of methanol-poisoned humans (Alha
in methodologies, the results of intake studies from et al., 1958) or monkeys. In monkeys, parallel studies
other countries, including a number of European coun- with infusions of small doses of formaldehyde showed
tries, have been remarkably consistent with the U.S. that the measurement techniques were sufficiently
data (Hinson and Nicol, 1992; MAFF, 1990, 1995; Bar sensitive to detect formaldehyde in blood and tissues
and Biermann, 1992; Chambolle et al., 1994; Garnier- (McMartin et al., 1979). Thus, methanol-induced form-
Sagne et al., 2001; Virtanen et al., 1988; Leclercq et al., aldehyde accumulation in body fluids is not likely.
1999; Heybach and Ross, 1989; Bergsten, 1993; Hulshof Formaldehyde is completely and quantitatively con-
and Bouman, 1995; National Food Authority Australia, verted to formate. In humans and in monkeys, the accu-
1995; Toledo and Ioshi, 1995). Thus, methanol exposure mulation of formate has been shown to be responsible
from aspartame at the 90th percentile of aspartame in- for the production of metabolic acidosis in methanol poi-
take in the United States is only 0.3 mg/kg body wt/day soning (McMartin et al., 1975; Sejersted et al., 1983). In
or about 1/131/16 of the amount of methanol provided patients who have been intoxicated with methanol and
by aspartame at the ADIs in the US and Europe. in experimental studies in monkeys, the increase in the
anion gap in the blood can be completely explained by
the accumulation of formate. In species that are not
Pathogenesis of Methanol Toxicity: The Importance
sensitive to methanol such as the rat, formate accumu-
of Formate Accumulation
lation and metabolic acidosis do not occur (McMartin
The classic pattern of methanol toxicity includes an et al., 1975). Hence, the species difference in methanol
initial central nervous system (CNS) depression weaker sensitivity is related to the capacity to accumulate
S36
ASPARTAME: REVIEW OF SAFETY S37
in individuals with liver disease after a bolus dose of of investigation for a number of years with a range of
aspartame (15 mg/kg body wt) compared to placebo methanol concentrations reported from different stud-
(Hertelendy et al., 1993). In children with attention ies. In addition, other sources of dietary methanol in
deficit disorder given aspartame (34 mg/kg body wt) or a healthy diet included filbert nuts, legumes, and veg-
placebo for 2 weeks on each treatment (Shaywitz et al., etables that are not typically used for juices, such as
1994), there was no detectable blood methanol and no potato, onion, Brussels sprouts, celery, and parsnip
differences in plasma formate or urinary formate con- (WHO, 1997).
centrations after aspartame vs placebo. As noted by several authors, the highly variable
In a study of long-term exposure to large doses of methanol content of juices depends not only on the type
aspartame (Leon et al., 1989), healthy adult subjects of fruit but also on the ripeness of the fruit as well as
were given 75 mg/kg body wt of aspartame or placebo the type of processing and storage time (Bindler et al.,
daily for 6 months (i.e., the amount of aspartame in 1988; Kirchner and Miller, 1957; Kirchner et al., 1953).
about 10 liters of beverage sweetened with 100% as- In addition, Wucherpfennig et al. (1983) point out that
partame for an adult). This very large dose of aspar- methanol content of juices should be addressed from
tame provides 7.5 mg/kg body wt/day methanol. During the standpoint of total methanol content, meaning
the study with measurements at 6-week intervals, most existing (i.e., free) methanol in the juice plus the
blood methanol concentrations were below the limit of potential (i.e., releasable) methanol that is made
detection in both the aspartame and the placebo groups. available as pectin is broken down by enzymes or dur-
The number of subjects with detectable blood methanol ing storage or digestion. For example, Wucherpfennig
concentrations was similar in both groups. Thus, there and co-workers reported for tomato juice, the exist-
was no accumulation of methanol from long-term in- ing methanol content of 159 mg/liter in addition to po-
gestion of these very large doses of aspartame. In ad- tential methanol content of 142 mg/liter, resulting in
dition, blood formate concentrations were not signifi- a total methanol content of 301 mg/liter. In this ex-
cantly increased after aspartame. Further, evaluation ample, the total methanol exposure is about two times
of 24-h urine collections revealed no increase in urinary the nominal concentration of free methanol in tomato
formate excretion after aspartame compared to placebo juice, which suggests that the methanol concentrations
or in urinary formate to creatinine ratio, indicating no reported in juices in other studies may be underesti-
significant increase in formate formation during high- mates of dietary methanol exposure.
dose, long-term aspartame intake. Consistent with the approach of Wucherpfennig and
co-workers, a report from a panel of experts published
Dietary Exposure to Methanol by the World Health Organization (WHO, 1997) dis-
cussed that dietary methanol can arise from fresh fruits
The amount of methanol released from dietary ex-
and vegetables as free methanol, methyl esters of fatty
posure to aspartame is less than normal dietary ex-
acids, or methoxy groups on polysaccharides such as
posure to methanol from fruits, vegetables, and juices
pectin. Sommer (1962) and Gruner et al. (1994) have
(Butchko and Kotsonis, 1989, 1991). For example, a
reported that pectins are metabolized in the human
serving of tomato juice provides about six times more
GI tract with consequent release and absorption of
methanol than an equivalent volume of beverage sweet-
methanol. Therefore, it is likely that a healthy diet actu-
ened 100% with aspartame (Fig. 1). Safe dietary expo-
ally safely provides greater amounts of methanol than
sure to methanol from fruits and vegetables and their
would appear to be the case from the reported methanol
juices and from alcoholic beverages has been the subject
contents of juices.
Finally, Taucher et al. (1995) calculated the lower
limit of the actual rate of methanol production in hu-
120 107 mans after eating fruit by evaluating methanol in hu-
100
man breath after fruit consumption. These authors es-
timated methanol production in the human body of
80 65 up to 0.1 g/hr, showing that because of fruit con-
mg/355 ml
. . . the tolerable (safe) level of exposure to methanol The enormous doses of aspartame used in the
is 7.1 to 8.4 milligrams per kilogram body weight per metabolism studies in humans are about 1065 times
day (mg/kg body weight/day), or approximately 426 to the 90th percentile average daily intake of aspartame.
504 mg/person/day for a 60 kg adult. Further, FDA also Further, the amount of methanol derived from aspar-
considered metabolic capacity when evaluating safe lev- tame is less than the amounts derived from fruits and
els of methanol intake in humans. FDA (1988) stated, vegetables and their juices and alcoholic beverages in
An adult human can metabolize up to 1500 milligrams the normal diet. According to the FDA (1988), An
of methanol per hour with no adverse symptoms or adult human can metabolize up to 1500 milligrams of
effects. Thus, the capacity for methanol metabolism methanol per hour with no adverse symptoms or ef-
in humans is far greater than estimated dietary in- fects. Thus, the capacity for methanol metabolism in
take. In the case of methanol exposure from aspartame, humans is far greater than estimated dietary intake
0.3 mg/kg body wt/day at the 90th percentile, the safe from all sources.
level as established by FDA is about 25 times that from FDA evaluated the safe levels of methanol intake
aspartame. and concluded, . . . the tolerable (safe) level of expo-
sure to methanol is 7.1 to 8.4 milligrams per kilo-
Evaluation of Recent Issues Regarding Methanol gram body weight per day (mg/kg body weight/day), or
Safety from Aspartame approximately 426 to 504 mg/person/day for a 60 kg
adult (FDA, 1994, 1996). This amount of methanol is
Trocho et al. (1998) concluded from a study in rats
about 25 times greater than the amount of methanol
that aspartame may be hazardous because formalde-
(0.3 mg/kg body wt) provided by aspartame to the diet
hyde adducts from aspartame may accumulate in
at the 90th percentile intake.
tissue proteins and nucleic acids. However, according
to Tephly (1999), the dose of aspartame used in the
study (20 mg/kg body wt = 2 mg of methanol/kg body wt) REFERENCES
would not yield blood methanol concentrations outside
control values. Further, the administration of aspar- Alha, A. R., Raehallio, J., and Muckula, A. L. (1958). Detection of
tame at 200 mg/kg body wt (equal to that in a single bo- methanol poisoning with special consideration of the estimation of
lus of about 25 liters of beverage sweetened 100% with formic acid in solid viscera, blood and urine: Investigation of 11
fatal cases. Ann. Med. Exp. Fenn. 36, 444451.
aspartame) to adult humans results in no detectable
Bar, A., and Biermann, C. (1992). Intake of intense sweeteners in
increase in blood formate concentrations (Stegink Germany. Z. Ernahrungswiss. 31, 2539.
et al., 1981). Administration of [14 C]methanol itself at Bergsten, C. (1993). Intake of Acesulfame-K, Aspartame, Cyclamate
3000 mg/kg body wt to monkeys produces no detectable and Saccharin in Norway, SNT Report 3. Norwegian Food Control
[14 C]formaldehyde in body fluids and tissues (McMartin Authority, Oslo, Norway.
et al., 1979), while there is ample accumulation of for- Bindler, F., Voges, E., and Laugel, P. (1988). The problem of methanol
mate. An alternative explanation for tissue incorpora- concentration admissible in distilled fruit spirits. Food Addit. Con-
tam. 5, 343351.
tion of label from [14 C]aspartame as described by Trocho
Black, K. A., Eells, J. T., Noker, P. E., Hawley, C. A., and Tephly, T. R.
et al. (1998) would be incorporation into amino acids
(1985). Role of hepatic tetrahydrofolate in the species difference in
and nucleotides via one-carbon moieties from the folate- methanol toxicity. Proc. Natl. Acad. Sci. USA 82, 38543858.
dependent metabolism of formate. The lack of formalde- Butchko, H. H., and Kotsonis, F. N. (1989). Aspartame: Review of
hyde accumulation at very high doses of methanol recent research. Comments Toxicol. 3, 253278.
question considerably the conclusion that formaldehyde Butchko, H. H., and Kotsonis, F. N. (1991). Acceptable daily intake
adducts are forming from low doses of methanol (de- vs actual intake: The aspartame example. J. Am. Coll. Nutr. 10,
rived from high doses aspartame). Thus, Tephly (1999) 258266.
concluded, the normal flux of one-carbon moieties Butchko, H. H., and Kotsonis, F. N. (1994). Postmarketing surveil-
lance in the food industry: the aspartame case study. In Nutritional
whether derived from pectin, aspartame, or fruit juices
Toxicology (F. N. Kotsonis, M. Mackey, and J. Hjelle, Eds.), pp. 235
is a physiologic phenomenon and not a toxic event. 250. Raven Press, New York.
Butchko, H. H., and Kotsonis, F. N. (1996). Acceptable daily in-
Conclusion take and estimation of consumption. In The Clinical Evaluation
of A Food Additive: Assessment of Aspartame (C. Tschanz, H. H.
The safety of methanol and methanol derived from Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 4353. CRC
aspartame has been extensively evaluated. Based on Press, Boca Raton, FL.
metabolism studies, it is not possible for a human to Butchko, H. H., and Stargel, W. W. (2001). Aspartame: Scientific eval-
ever consume enough aspartame in products to raise uation in the postmarketing period. Regul. Toxicol. Pharmacol. 34,
blood formate concentrations. Because formate is the 221233.
toxic metabolite of methanol, the lack of formate accu- Butchko, H. H., Tschanz, C., and Kotsonis, F. N. (1994). Postmarket-
ing surveillance of food additives. Regul. Toxicol. Pharmacol. 20,
mulation after very large doses of aspartame indicates 105118.
that the conversion of aspartame to methanol is not Chambolle, M., Collerie de Borely, A., Renault, C., and Volatier, J.
sufficient to induce any toxicity from methanol or its (1994). Intake of Intense Sweeteners in France (19891992). CRE-
metabolites. DOC, Departement Prospective de la Consommation, Paris.
S40 BUTCHKO ET AL.
Eells, J. T., Salzman, M. M., Lewandowski, M. F., and Murray, McMartin, K. E., Makar, A. B., Martin-Amat, G., Palese, M., and
T. G. (1996). Formate-induced alterations in retinal function in Tephly, T. R. (1975). Methanol poisoning. I. The role of formic acid
methanol-intoxicated rats. Toxicol. Appl. Pharmacol. 140, 58 in the development of metabolic acidosis in the monkey and the
69. reversal by 4-methylpyrazole. Biochem. Med. 13, 319333.
Food and Drug Administration (FDA) (1988). Direct food additives McMartin, K. E., Martin-Amat, G., Makar, A. B., and Tephly, T. R.
Food additives permitted for direct addition to food for human con- (1977). Methanol poisoning. V. Role of formate metabolism in the
sumption: Dimethyl dicarbonate. Fed. Regist. 53, 4132541329. monkey. J. Pharmacol. Exp. Ther. 201, 564572.
Food and Drug Administration (FDA) (1994). Food additives permit- McMartin, K. E., Martin-Amat, G., Noker, P. E., and Tephly, T. R.
ted for direct addition to food for human consumption: Dimethyl (1979). Lack of a role for formaldehyde in methanol poisoning in
dicarbonate. Fed. Regist. 59, 53175320. the monkey. Biochem. Pharmacol. 28, 645649.
Food and Drug Administration (FDA) (1996). Food additives permit- Ministry of Agriculture, Fisheries and Food (MAFF) (1990). Intakes
ted for direct addition to food for human consumption: Dimethyl of Intense and Bulk Sweeteners in the UK 19871988, Food Surveil-
dicarbonate. Fed. Regist. 61, 2678626788. lance Paper No. 29, London, UK.
Garnier-Sagne, I., Leblanc, J. C., and Verger, P. (2001). Calculation of Ministry of Agriculture, Fisheries and Food (MAFF) (1995). Survey of
the intake of three intense sweeteners in young insulin-dependent the Intake of Sweeteners by Diabetics, Food Surveillance Informa-
diabetics. Food Chem. Toxicol. 39, 745749. tion Sheet. Food Safety Directorate, London, UK.
Gruner, O., Bilzer, N., and Liebmann, J. (1994). Zur methanolbildung Murray, T. G., Burton, T. C., Rajani, C., Lewandowski, M. F., Burke,
in vitro und in vivo. Blutalkohol. 31, 228232. J. M., and Eells, J. T. (1991). Methanol poisoning: A rodent model
Health and Welfare Canada (1979). Proposal on Aspartame, Informa- with structural and functional evidence for retinal involvement.
tion Letter No. 564. Health Protection Branch, Ottawa. Arch. Ophthalmol. 109, 10121016.
Hertelendy, Z. I., Mendenhall, C. I., Rouster, S. D., Marshall, L., and National Food Authority Australia (1995). Survey of Intense Sweet-
Weesner, R. (1993). Biochemical and clinical effects of aspartame ener Consumption in Australia: Final Report. Canberra, Australia.
in patients with chronic, stable alcoholic liver disease. Am. J. Gas- Nicholls, P. (1976). The effects of formate on cytochrome aa3 and on
troenterol. 88, 737743. electron transport in the intact respiratory chain. Biochim. Bio-
Heybach, J. P., and Ross, C. (1989). Aspartame consumption in a rep- phys. Acta 430, 1329.
resentative sample of Canadians. J. Can. Diet. Assoc. 50, 166170. Noker, P. E., Eells, J. T., and Tephly, T. R. (1980). Methanol toxicity:
Hinson, A. L., and Nicol, W. M. (1992). Monitoring sweetener con- Treatment with folic acid and 5-formyl-tetrahydrofolic acid. Alco-
sumption in Great Britain. Food Addit. Contam. 9, 669680. hol. Clin. Exp. Res. 4, 378383.
Hulshof, K. F. A. M., and Bouman, M. (1995). Use of Various Types Scientific Committee for Food (SCF) (1985). Food-science and Tech-
of Sweeteners in Different Population Groups, 1992 Dutch National niques, Reports of the Scientific Committee for Food, sixteenth se-
Food Consumption Survey. TNO Nutrition and Food Research In- ries. Commission of the European Communities, Luxembourg.
stitute, Netherlands. Sejersted, O. M., Jacobsen, D., Ovrebo, S., and Jansen, H. (1983).
Joint FAO/WHO Expert Committee on Food Additives (JECFA) Formate concentrations in plasma from patients poisoned with
(1980). Toxicological Evaluation of Certain Food Additives, WHO methanol. Acta Med. Scand. 213, 105110.
Food Additive Series No. 15, pp. 1886. JECFA, Geneva. Seme, M. T., Summerfelt, P., Neitz, J., Eells, J. T., and Henry, M. M.
Kirchner. J. G., and Miller, J. M. (1957). Volatile water-soluble and (2001). Differential recovery of retinal function after mitochondrial
oil constituents of Valencia orange juice. Agric. Food Chem. 5, 282 inhibition by methanol intoxication. Invest. Ophthalmol. Vis. Sci.
291. 42, 834841.
Kirchner, J. G., Miller, J. M, Rice, R. G., Keller, G. J., and Fox, M. M. Shaywitz, B. A., Sullivan, C. M., Anderson, G. M., Gillespie, S. M.,
(1953). Volatile water-soluble constituents of grapefruit juice. Agric Sullivan, B., and Shaywitz, S. E. (1994). Aspartame, behavior, and
Food Chem. 1, 510512. cognitive function in children with attention deficit disorder. Pedi-
Kotsonis, F. N., and Hjelle, J. J. (1996). The safety assessment of atrics 93, 7075.
aspartame: Scientific and regulatory considerations. In The Clin- Sommer, H. (1962). The physiological fate of methylalcohol re-
ical Evaluation of A Food Additive: Assessment of Aspartame leased from pectin. Industrielle Obst. Gemuseverwertung.
47, 172
(C. Tschanz, H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), 173.
pp. 2341. CRC Press, Boca Raton, FL. Stegink, L. D., Brummel, M. C., McMartin, K., Martin-Amat, G., Filer,
Leclercq, C., Berardi, D., Sorbillo, M. R., and Lambe, J. (1999). Intake L. J., Jr., Baker, G. L., and Tephly, T. R. (1981). Blood methanol
of saccharin, aspartame, acesulfame K and cyclamate in Italian concentrations in normal adult subjects administered abuse doses
teenagers: Present levels and projections. Food Addit. Contam. 16, of aspartame. J. Toxicol. Environ. Health 7, 281290.
99109. Stegink, L. D., Brummel, M. C., Filer, L. J., Jr., and Baker, G. L.
Leon, A. S., Hunninghake, D. B., Bell, C., Rassin, D. K., and Tephly, (1983). Blood methanol concentrations in one-year-old infants ad-
T. R. (1989). Safety of long-term large doses of aspartame. Arch. ministered graded doses of aspartame. J. Nutr. 113, 16001606.
Intern. Med. 149, 23182324. Stegink, L. D., Filer, L. J., Jr., Bell, E. F., Zeigler, E. E., and Tephly,
Makar, A. B., and Tephly, T. R. (1976). Methanol poisoning in the T. R. (1989). Effect of repeated ingestion of aspartame-sweetened
folate-deficient rat. Nature 261, 715716. beverage on plasma amino acid, blood methanol, and blood formate
Martin-Amat, G., Tephly T. R., McMartin, K. E., Makar, A. B., Hayreh, concentrations in normal adults. Metabolism 38, 357363.
M. S., Hayreh, S. S., Baumbach, G., and Cancilla, P. (1977). Methyl Stegink, L. D., Filer, L. J., Jr., Bell, E. F., Zeigler, E. E., Tephly,
alcohol poisoning. II. Development of a model for ocular toxicity in T. R., and Krause, W. L. (1990). Repeated ingestion of aspartame-
methyl alcohol poisoning using the rhesus monkey. Arch. Ophthal- sweetened beverages: Further observations in individuals het-
mol. 95, 18471850. erozygous for phenylketonuria. Metabolism 39, 10761081.
Martin-Amat, G., McMartin, K. E., Hayreh, S. S., Hayreh, M. S., and Taucher, J., Lagg, A., Hansel, A., Vogel, W., and Lindinger, W. (1995)
Tephly, T. R. (1978). Methanol poisoning: Ocular toxicity produced Methanol in human breath. Alcohol. Clin. Exp. Res. 19, 1147
by formate. Toxicol. Appl. Pharmacol. 45, 201208. 1150.
ASPARTAME: REVIEW OF SAFETY S41
Tephly, T. R. (1991). The toxicity of methanol. Life Sci. 48, 10311041. from human liver: Purification, properties and evidence for the for-
Tephly, T. R. (1999). Comments on the purported generation of form- mation of glutathionine thiol esters by the enzyme. J. Biol. Chem.
aldehyde and adduct formation from the sweetener aspartame. Life 249, 76537663.
Sci. 65, 157160. Virtanen, S. M., Rasanen, L., Paganus, A., Varo, P., and Akerblom,
Trocho, C., Pardo, R., Rafecas, I., Virgili, J., Remesar, X., Fernandez- H. K. (1988). Intake of sugars and artificial sweeteners by adoles-
Lopez, J. A., and Alemany, M. (1998). Formaldehyde derived from cent diabetics. Nutr. Rep. Int. 38, 12111218.
dietary aspartame binds to tissue components in vivo. Life Sci. 63, World Health Organization (WHO) (1997). Environmental Health
337349. Criteria 196: Methanol. WHO, Geneva.
Toledo, M. C. F., and Ioshi, S. H. (1995). Potential intake of intense Wucherpfennig, K., Dietrich, H., and Bechtel, J. (1983). Vorhandener,
sweeteners in Brazil. Food Addit. Contam. 12, 799808. gesamter und potentieller methylalkoholgehalt von fruchtsaften.
Uotila L., and Koivusalo, M. (1974). Formaldehyde dehydrogenase Flussiges Obst. 8, 348354, 367369.
Postmarketing Surveillance: Evaluation of Anecdotal
Reports of Health Effects
Introduction ations, insomnia, and dizziness. About a quarter of the
reports were gastrointestinal in nature and included
In the 1940s and 1950s, when many new drugs were
abdominal pain, nausea, diarrhea, and vomiting (CDC,
being developed and marketed, it became apparent that
1984; Bradstock et al., 1986).
the full spectrum of adverse reactions was not always
The CDC concluded, Despite great variety overall,
apparent until many patients had been exposed to the
the majority of frequently reported symptoms were mild
drug over a period of time (Faich, 1986). Thus, it was
and are symptoms that are common in the general pop-
concluded that a postmarketing surveillance system to
ulace (CDC, 1984). There were no specific clinical syn-
document and evaluate spontaneous reports of any ad-
dromes that suggested a causal relationship with aspar-
verse reactions associated with marketed pharmaceu-
tame. The CDC concluded that focused clinical studies
ticals would be useful to expand the safety profile doc-
would be the way to address the issues raised by the
umented from the extensive preapproval studies.
anecdotal reports.
Shortly after aspartames marketing in the United
States, a number of anecdotal reports of health effects, Evaluation by Food and Drug Administration ARMS
which some consumers related to their consumption
of aspartame-containing products, were documented. Unlike the case of pharmaceuticals, where most in-
Not unexpectedly, the numbers and types of these re- formation is received from physicians, information re-
ports were influenced by negative media stories (CDC, garding food additives is largely reported by consumers.
1984; Butchko et al., 1996; Butchko and Stargel, 2001). In the case of aspartame, about 70% of the reports in
The NutraSweet Company developed a postmarketing ARMS were provided by The NutraSweet Company. Re-
surveillance system for aspartame, based on the prin- ports to ARMS are categorized based on the severity of
ciples used for postmarketing surveillance of pharma- symptoms and on the consistency and frequency with
ceuticals, to document and evaluate these anecdotal re- which they occur. The FDA concluded that there is no
ports (Butchko and Kotsonis, 1989, 1994; Butchko et al., reasonable evidence of possible public health harm
1994, 1996, 2001; Butchko and Stargel, 2001, 2002). and no consistent or unique patterns of symptoms re-
Further, data from this system were shared with the ported with respect to aspartame that can be causally
U.S. FDA and the Centers for Disease Control (CDC), linked to its use (Tollefson et al., 1988).
as discussed below. In a 1995 FDA report on aspartame, a total of 7232
Following the approval of aspartame in carbonated consumer reports had been received since marketing
beverages in 1983, there was an increase in the re- with only about 11% classified as serious. Headache was
porting of adverse health events allegedly associated the most common symptom reported, followed by dizzi-
with the consumption of aspartame-containing prod- ness, mood changes, and nausea/vomiting (Table 1). The
ucts. This led the FDA to request the CDC to evaluate report noted that the numbers of reports from con-
these reports (CDC, 1984; Bradstock et al., 1986). In sumers regarding aspartame had declined since the
addition, the FDAs Center for Food Safety and Applied
Nutrition (CFSAN) started its own process, the Adverse TABLE 1
Reaction Monitoring System (ARMS), in 1985 to moni- Anecdotally Reported Symptomsa
tor accounts of health problems anecdotally associated
with consumption of foods, food and color additives, and Reported symptom Percentage of complaints
vitamin/mineral supplements (Tollefson, 1988; Tollef-
Headache 19.0
son et al., 1988). Dizziness/poor equilibrium 7.5
Change in mood 6.7
Evaluation by the Centers for Disease Control Vomiting or nausea 6.6
Abdominal pain and cramps 4.7
CDC analyzed over 500 reports with about half un- Change in vision 3.7
dergoing detailed follow-up and evaluation. Most of Diarrhea 3.4
the reports came from white females aged 2160 years Seizures and convulsions 3.0
Memory loss 2.6
and were randomly distributed throughout the United
Fatigue, weakness 2.5
States, with one exception. As aspartame had been sub- Other neurological 2.4
jected to substantial negative media coverage in Ari- Rash 2.3
zona, there were proportionately more reports from that Sleep problems 2.1
state. While a variety of different symptoms were re- Hives 2.0
Change in heart rate 1.9
ported, two-thirds fell into the neurologic/behavioral
category, consisting mostly of headache, mood alter- a Source: FDA (1995).
S42
ASPARTAME: REVIEW OF SAFETY S43
60
The NutraSweet Company System for Health
40 Report Evaluation
20 The NutraSweet Companys postmarketing surveil-
0 lance system, which continued for 12 years after mar-
Jan* Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec keting in the United States, was a collaborative ef-
* Negative series on aspartame aired on CBS TV on January 16-18, 1984
fort between the Consumer Center, where the staff
was responsible for data collection, documentation, and
FIG. 1. Negative media reports regarding aspartame had a follow-up, and the Clinical Research Group, where
marked impact on the number of anecdotal reports of health effects al- physicians provided medical expertise for evaluation of
legedly associated with aspartame received by The NutraSweet Com-
pany in early 1984. Adapted from Butchko et al. (1996).
the reports. Reports were largely received directly from
consumers via telephone or letter. As noted in the CDC
and FDA reports, symptoms allegedly associated with
peak in 1985 and further stated, In summary, the num- aspartame tended to be mild and were also ones that
ber of adverse reaction complaints received by the FDA were common in the general population (Butchko and
and the nature of these reports in terms of demographic Kotsonis, 1989, 1994, 1996; Butchko et al., 1994, 2001;
distribution, severity, strength of association with the Butchko and Stargel, 2001, 2002).
product, and symptoms remain comparable to those As expected, the negative media stories and result-
from previous analyses (FDA, 1995). ing controversy about aspartame in the early to mid-
FDA also separately analyzed 251 reports of seizure 1980s had a significant impact on the number of anec-
anecdotally associated with aspartame consumption dotal reports (Fig. 1). The number of reports increased
that had been received through ARMS from 1986 markedly during that time, and, as the controversy de-
1990. FDA concluded that approximately half were creased in the late 1980s and early 1990s, the num-
highly unlikely to be related to aspartame (Tollefson ber of reports declined while consumption of aspartame
and Barnard, 1992). Furthermore, the FDA could not did not (Fig. 2). In addition, there was a striking re-
exclude the possibility that the remaining reports had lationship between media reports and the types of re-
not simply occurred by chance. FDA concluded that the ports from consumers (Butchko et al., 1996; Butchko
anecdotal reports did not support the claim that the and Stargel, 2001).
occurrences of the seizures were linked to consumption As there are more than 100 million aspartame users
of aspartame (Tollefson and Barnard, 1992). It was fur- in the United States alone, it is inevitable that some of
ther concluded that the data did not suggest the need them will experience medical ailments temporally as-
for a controlled clinical study to evaluate this issue. In a sociated with consumption of an aspartame-containing
subsequent analysis (Tollefson, 1993) of the 265 reports product simply by chance. A temporal association does
of seizures received from January 1986 through October not necessarily imply a causal relationship.
1000 Ex-U.S.
U.S.
750
# of Reports
500
250
0
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
FIG. 2. Anecdotal reports of health effects thought to be related to aspartame peaked in the mid-1980s, shortly after widespread aspartame
marketing and declined in the late 1980s. The vast majority of the reports were received from U.S. consumers. Adapted from Butchko et al.
(1996).
S44 BUTCHKO ET AL.
Butchko, H. H., and Stargel, W. W. (2002). Postmarketing surveillance Roberts, H. J. (1990). Aspartame (NutraSweet): Is It Safe? Charles
in the food industry: the aspartame case study. In Nutritional Toxi- Press, Philadelphia.
cology (F. N. Kotsonis and M. Mackey, Eds.), second ed., pp. 287315. Rolla, A. R. (1990). Aspartame (NutraSweet): Is it safe? (book review).
Taylor & Francis, London/New York. N. Engl. J. Med. 323, 14951496.
Caseley, J., and Dixon, W. (2001). Aspartame: In perspective. Nutr. Schiffman, S. (1989). Reply to H. J. Roberts, Adverse reactions to
Food Sci. 31, 2326. aspartame (letter). J. Appl. Nutr. 41, 40.
Centers for Disease Control (CDC) (1984). Evaluation of Consumer Squillacote, D. (1999a). Aspartame (NutraSweet): No Danger. The
Complaints Related to Aspartame Use. CDC, Atlanta, GA. Multiple Sclerosis Foundation, www.msfocus.org/aspar.htm.
Faich, G. A. (1986). Adverse-drug-reaction monitoring. N. Engl. J. Squillacote, D. (1999b). Aspartame: Come in from the Ledge. The Mul-
Med. 314, 15891592. tiple Sclerosis Foundation, www.msfacts.org/aspart2.htm.
Food and Drug Administration (FDA) (1995). Summary of Ad- Tollefson, L. (1988). Monitoring adverse reactions to food additives in
verse Reactions Attributed to Aspartame. FDA, Washington, the U.S. Food and Drug Administration. Regul. Toxicol. Pharmacol.
DC. 8, 438446.
Food Standards Agency (FSA) (2000). Food Safety Informa- Tollefson, L., Barnard, R. J., and Glinsmann, W. H. (1988). Monitor-
tion Bulletin No. 117, www.foodstandards.gov.uk/maff/archive/ ing of adverse reactions to aspartame reported to the U.S. Food
food/bulletin/2000/no117/ asparte.htm. and Drug Administration. In Dietary Phenylalanine and Brain
Henkel, J. (1999). Sugar substitutes: Americans opt for sweetness and Function (R. J. Wurtman and E. Ritter-Walker, Eds.), pp. 317337.
lite. FDA Consumer NovemberDecember, 1216. Birkhauser, Boston.
Roberts, H. J. (1988a). Neurological, psychiatric, and behavioral re- Tollefson, L., and Barnard, R. J. (1992). An analysis of FDA passive
actions to aspartame in 505 aspartame reactors. In Dietary Pheny- surveillance reports of seizures associated with consumption of as-
lalanine and Brain Function (R. J. Wurtman and E. Ritter-Walker, partame. J. Am. Diet. Assoc. 92, 598601.
Eds.), pp. 373376. Birkhauser, Boston. Tollefson, L. (1993). Multiple chemical sensitivity: Controlled scien-
Roberts, H. J. (1988b). Reactions attributed to aspartame-containing tific studies as proof of causation. Regul. Toxicol. Pharmacol. 18,
products: 551 cases. J. Appl. Nutr. 40, 8594. 3243.
Evaluation of Aspartame and Headaches
Introduction may cause headache or the act of chewing itself cannot
be ruled out as causative factors. In a letter to the
Headaches are one of the most common human ail-
editor, Watts (1991) discussed his headaches, which he
ments. Although in the majority of cases the symp-
felt were related to aspartame.
toms are more uncomfortable than disabling, severe
A single-subject (N = 1) study with multiple cross-
headaches can seriously impair the ability to carry on
overs was conducted by Frasca and Aldag (1988) to
normal activities. At ambulatory care centers in the
determine whether aspartame use was a factor in in-
United States, headache is the seventh most common
creased headache frequency experienced by a 27-year-
complaint, accounting for about 18 million visits per
old woman who had also started using oral contracep-
year (Linet et al., 1989). The majority of headache suf-
tives. When both aspartame and the oral contraceptive
ferers, however, do not seek medical treatment, es-
were discontinued, headache frequency diminished. To
pecially those from low-income groups, those whose
determine if aspartame played a role in her headaches,
headaches can be treated easily with nonprescription
a double-blind, single-patient study was done, which
drugs, and those whose headaches are relatively mild
involved four 2-week randomized crossover periods of
and short in duration rather than disabling or long-
challenge with aspartame and placebo with a 1-week
lasting. Thus the incidence of headache is probably
washout period between each 2-week trial period. No
greatly underreported.
significant difference was found in frequency, duration,
or severity of headache between the aspartame and
Anecdotal Reports of Headache and Aspartame
placebo periods.
After aspartame marketing, there were anecdotal
reports from some consumers that aspartame may be Questionnaire Evaluations
associated with a variety of health effects. These com-
plaints were evaluated by the Centers for Disease Con- Lipton et al. (1989) distributed a questionnaire
trol (CDC) (CDC, 1984; Bradstock et al., 1986), as well to patients seen at a headache clinic to determine
as later by the U.S. Food and Drug Administration their opinions about whether three dietary factors
(FDA) (Tollefson, 1988; Tollefson et al., 1988). Headache alcohol, aspartame, and carbohydratestriggered their
accounted for about 20% of all symptoms reported. Nei- headaches. Carbohydrates were included as a nega-
ther CDC nor FDA was able to identify a specific set tive control because they are believed not to precipitate
of symptoms associated with aspartame use that would headaches. Of the 171 persons who completed the ques-
constitute a public health hazard. Most of the symptoms tionnaire, 49.7% reported alcohol, 8.2% reported as-
reported were mild and were also common in the gen- partame, and 2.3% reported carbohydrates as possible
eral population. However, they emphasized that care- headache precipitants. They concluded that aspartame
fully designed, focused clinical trials would be the way may be a dietary trigger of headache in some people. A
to determine definitely if there was a causal link be- questionnaire, however, cannot establish a link between
tween aspartame consumption and any specific adverse aspartame consumption and headache; it only reflects
effect. consumers opinions, which may be biased by the exten-
sive media reports and widespread speculation regard-
Case Reports and Reports of Single-Subject Tests ing aspartame that happened at the same time.
aspartame or a matching placebo. Of the 11 subjects somewhat sure and not sure groups. There were
remaining at the end of the study, six had greater fre- no significant differences in the duration or intensity
quency of migraine occurrence during the aspartame of headaches or other adverse experiences after aspar-
consumption phase than during the placebo phase, re- tame compared to placebo. From these results, the au-
sulting in a significantly higher mean incidence of mi- thors concluded that their study provided evidence of a
graine after aspartame vs placebo; however, there were subset of individuals among those who attributed their
no statistically significant differences in the intensity headaches to aspartame that are particularly suscepti-
or duration of migraine after aspartame compared with ble to aspartame.
placebo. Further, there were no statistically significant Roberts (1995) felt that this study confirmed his case
differences between the aspartame and placebo phases reports of headache, which he attributes to aspartame.
in frequency, intensity, duration, and associated symp- However, the study was criticized on several grounds
toms of tension and unclassified headaches. The re- by Levy et al. (1995) and Schiffman (1995). Schiffman
searchers concluded that aspartame consumption can (1995) discussed that the study, as reported in the jour-
increase the frequency and may extend the duration of nal, appeared to have been a post hoc analysis of the
headache in some persons with a history of migraine. principle authors dissertation, which had not shown
This study was criticized (Schiffman, 1988) because any difference between aspartame and placebo in the
the many factors in the subjects daily environment (in- mean occurrence of headache in all subjects (P = 0.19)
cluding diet) that could be related to headache inci- or among those who were very sure that aspartame
dence were not controlled and, thus, cannot be elimi- had caused their headaches (P = 0.11). In the new anal-
nated as possible alternative explanations for apparent ysis reported in the published article, the authors also
small differences in headache frequency between treat- changed the way that headaches were countedin the
ment and placebo groups. Further, data from only 11 dissertation, they evaluated mean number of headaches
of the 25 subjects were reported, and the effects on fre- whereas in the paper, they evaluated the number of
quency of headaches can be attributed largely to data days that headache occurred. However, this requires
from only 23 subjects. Furthermore, the high dropout an assumption that all days with headache are inde-
rate (8 subjects) and the necessity of excluding 6 other pendent of the other days, when it is well known that
subjects from analysis may have biased the represen- headaches often occur in clusters.
tativeness of the remaining eleven (Schiffman, 1988; In addition, the results from only one subject of the
Amery, 1988). Finally, the statistical tests used to com- 32 enrolled largely accounted for any difference be-
pare mean headache frequencies, intensities, and du- tween aspartame and placebo. Levy et al. (1995) rean-
rations were inappropriate for a crossover study. Thus, alyzed the data in the Van Den Eeden et al. study and
given these and other limitations, no reliable conclu- concluded that the study results were largely due to
sions can be drawn from this study. the results of one outlier subject. Subject 25 reported
In another outpatient study, Van Den Eeden et al. headaches on 13 of the days on aspartame but on only
(1994) did a randomized, double-blind, crossover study 2 of the 21 placebo days. Using the methodology of van
with individuals who had self-identified headaches af- Den Eeden and co-workers, but with this outlier ex-
ter using aspartame. Of the 32 people entered into the cluded, the odds ratios were not statistically different
study, only 18 completed the full protocol, 7 completed between aspartame and placebo. When the data were
part of the protocol before withdrawing due to adverse reanalyzed using a random effects logistic regression
experiences, 3 withdrew for other reasons, 2 were lost to model, Levy et al. found no statistically significant dif-
follow-up, 1 was withdrawn for noncompliance with the ference between aspartame and placebo in all subjects
study protocol, and 1 withdrew with no reason given. or those who were very sure that aspartame caused
This study consisted of a 1-week baseline period their headaches.
and four 1-week periodstwo each with aspartame
(30 mg/kg body wt/day) and two with placebo, with a
Study in a Clinical Research Unit
washout day between treatment periods. Prior to the
study, subjects were asked whether they were very To evaluate whether aspartame was associated with
sure, somewhat sure, or not sure if aspartame headache under controlled conditions, Schiffman and
was associated with their headaches. The authors re- colleagues (1987) conducted a clinical study using sub-
ported that there was a statistically significant differ- jects who had complained to the FDA or to The Nu-
ence (P = 0.04) in the number of days with headache traSweet Company that they were convinced that as-
after aspartame compared to placebo. In addition, sub- partame had caused their headaches within 24 h after
jects who were very sure that aspartame was asso- consuming one or more products containing aspartame.
ciated with their headache prior to the study had sta- The subjects were studied as inpatients at the Duke
tistically greater number of days with headaches after University Medical Center in Durham, North Carolina,
aspartame compared to placebo (P < 0.001), whereas over a 6-day period, allowing close and detailed moni-
there was no statistically significant differences in the toring of their status as well as elimination of variables,
S48 BUTCHKO ET AL.
such as diet, schedule, and activity, which may also im- Another comment was that administering the treat-
pact headaches. ments and measuring results over a 24-h period with
The experimental design was a randomized, double- only a 24-h washout period between the crossover
blind, placebo-controlled, crossover study. On the third phases does not allow detection of long-term, cumula-
and fifth days of hospitalization, each subject was given tive effects or those that require the subject to reach
either aspartame (three doses totaling 30 mg/kg body a threshold level through continued exposure to as-
wt) or a placebo in capsules. For an adult, this total partame (Elsas, 1988; Steinmetzer and Kunkel, 1988).
daily dosage would be equivalent to the amount of as- However, the alleged phenomenon being tested essen-
partame contained in about 4 liters of a beverage sweet- tially involves a short-term response, and reactions oc-
ened 100% with aspartame. curring within 24 h are of primary interest (Schiffman
Twenty-six of the 40 subjects experienced headache et al., 1988). In addition, the results of pharmacoki-
during the study period. Eight subjects had headache netic studies have demonstrated that aspartame is
while receiving aspartame but not placebo, whereas rapidly metabolized to its constituent amino acids and
12 had headaches while receiving placebo but not as- methanol (Stegink, 1987), which enter the metabolic
partame. Six subjects had headaches with both treat- pool (Ranney et al., 1976). Aspartame is not absorbed
ments. Including those 6 subjects, the headache inci- as an intact molecule and cannot accumulate. Thus, an
dence was 35% with aspartame consumption and 45% effect of aspartame on headache because of aspartame
with placebo. Furthermore, the severity, time between accumulation is not plausible.
challenge and onset, and duration of the headaches In addition, Elsas (1988) alleged that there was
were not statistically significantly different after as- 50% less absorption of aspartame from capsules (used
partame vs placebo (Fig. 1). In addition, there were no for blinding purposes) compared to solution. However,
statistically significant differences in the occurrence of there is no significant difference in the extent of pheny-
any of the other adverse experiences reported by these lalanine from aspartame given in capsules vs solu-
subjects when aspartame was compared with placebo. tion (Burns et al., 1990). He also commented that the
It was concluded that the headache incidence after a study failed to test a mechanistic hypothesis. However,
large amount of aspartame was equivalent to that after Schiffman et al. (1987) did discuss a possible mecha-
placebo. nism as circulating concentrations of catecholamines,
This study was criticized by Elsas (1988), Lipton not aspartame, were associated with headache.
et al. (1988), and Steinmetzer and Kunkel (1988). First,
it was asserted that studying subjects in a controlled Long-Term Study
environment rather than their natural environment
may have affected the results of the study. However, The safety of long-term use of large doses of as-
the only way to control for other variables, which may partame, extensively studied before the product was
affect headache, is in such a controlled environment released for marketing, was reconfirmed through a
(Schiffman et al., 1988). In fact, the inability to con- 24-week, randomized, double-blind, placebo-controlled,
trol extraneous factors has compromised other studies parallel-group study involving 108 normal, healthy sub-
evaluating the role of aspartame in headache. jects (Leon et al., 1989). Subjects were administered
three doses of aspartame or placebo each day with their
meals, corresponding to a daily dosage of about 75 mg/
100 10 kg body wt, which is 1.5 times the acceptable daily in-
Aspartame
90
Aspartame
Placebo 9 Placebo
take (ADI) of 50 mg/kg body wt in the United States or
80 8 about 1.9 times the ADI in the European Union. For an
70 7 adult, this dosage is equivalent to the daily consump-
60 6
tion of about 10 liters of a 100% aspartame-sweetened
50 5
beverage and about 25 times the average daily aspar-
40 4
tame intake (90th percentile) by the U.S. general popu-
30 3
lation (Butchko and Kotsonis, 1991).
The most frequent adverse experience was headache,
20 2
but the incidence rate in the aspartame group was not
10 1
significantly different from that in the placebo group.
0 0
Incidence Severity Onset After Duration The authors concluded that consumption of this very
(Percent) (Analog Scale First Dose (Hours)
By Patient) (Hours) large dose of aspartame daily for 6 months was not
associated with any significant changes in the numer-
FIG. 1. Mean (SEM) headache incidence (percentage), severity,
time of onset after the first dose, and duration after aspartame com- ous clinical or laboratory parameters or adverse expe-
pared to placebo in 40 subjects who had reported headache associated riences in healthy adults. In addition, the results of the
aspartame. Source: Schiffman et al. (1987). numerous other human studies done with aspartame
ASPARTAME: REVIEW OF SAFETY S49
do not indicate any association between aspartame and Frasca, M. A., and Aldag, J. C. (1988). The single-patient clinical trial.
headache (Tschanz et al., 1996). Am. Fam. Phys. 37, 195199.
Gaull, G. E. (1986). Response to Johns Migraine provoked by aspar-
tame (letter). N. Engl. J. Med. 315, 456.
Conclusion
Johns, D. R. (1986). Migraine provoked by aspartame (letter). N. Engl.
The controversy regarding aspartame and headache J. Med. 315, 456.
started shortly after marketing when many of the anec- Koehler, S. M., and Glaros, A. (1988). The effect of aspartame on mi-
graine headache. Headache 28, 1014.
dotal consumer reports claimed to be associated with
Leon, A. S., Hunninghake, D. B., Bell, C., Rassin, D. K., and Tephly,
consumption of aspartame were headaches. Headache T. R. (1989). Safety of long-term large doses of aspartame. Arch.
was associated with aspartame from a questionnaire, Intern. Med. 149, 23182324.
as well as single case reports. As headache is also one Levy, P. S., Hedeker, D., and Sanders, P. (1995). Aspartame and
of the most common symptoms in the general popula- headache (letter). Neurology 45, 16311632.
tion, it was not possible to decide with certainty whether Linet, M. S., Stewart, W. F., Celentano, D. D., Ziegler, D., and Sprecher,
aspartame was in fact a cause of headache based on M. (1989). An epidemiologic study of headache among adolescents
anecdotal information. Thus, several studies were done and young adults. JAMA 261, 22112216.
to evaluate this issue. The results of two outpatient Lipton, R. B., Newman, L. C., and Solomon, S. (1988). Aspartame and
headache (letter). N. Engl. J. Med. 318, 1200.
studies appear to indicate an association between as-
Lipton, R. B., Newman, L. C., Cohen, J. S., and Solomon, S. (1989).
partame and headache. However, statistical issues with Aspartame as a dietary trigger of headache. Headache 29, 90
results that rely on only a few subjects prevent draw- 92.
ing any valid conclusions from these studies. Another Ranney, R. E., Oppermann, J. A., Muldoon, E., and McMahon, F. G.
study, which was done in a clinical research unit with (1976). Comparative metabolism of aspartame in experimental an-
individuals who were convinced that aspartame caused imals and humans. J. Toxicol. Environ. Health 2, 441451.
their headaches, allowed control over the confounding Roberts, H. J. (1995). Aspartame and headache (letter). Neurology 45,
cofactors that were present in an outpatient setting and 1631.
found no difference between the occurrence of headache Schiffman, S. S. (1988). Aspartame and headache (letter). Headache
28, 370.
after aspartame and placebo. Finally, numerous human
Schiffman, S. (1995). Aspartame and headache (letter). Neurology 45,
studies have been done with aspartame, and in none of 1632.
these studies, including a high-dose, long-term study by Schiffman, S. S., Buckley, C. E., Sampson, H. A., Massey, E. W.,
Leon et al. (1989), was there an association between as- Baraniuk, J. N., Follett, J. V., and Warwick, Z. S. (1987). Aspartame
partame and headache in adverse experience profiles. and susceptibility to headache. N. Engl. J. Med. 317, 11811185.
Thus, the weight of the scientific evidence indicates that Schiffman, S. S., Buckley, C. E., Sampson, H. A., Massey, E. W.,
aspartame does not cause headache. Baraniuk, J. N., Follett, J. V., and Warwick, Z. S. (1988). Aspartame
and headache (letter). N. Engl. J. Med. 318, 12011202.
Stegink, L. D. (1987). The aspartame story: A model for the clinical
REFERENCES testing of a food additive. Am. J. Clin. Nutr. 46, 204215.
Steinmetzer, R. V., and Kunkel, R. S. (1988). Aspartame and headache
Amery, W. K. (1988). More on aspartame and headache. Headache 28, (letter). N. Engl. J. Med. 318, 1201.
624.
Tollefson, L. (1988). Monitoring adverse reactions to food additives in
Blumenthal, H. J., and Vance, D. A. (1997). Chewing gum headaches. the U.S. Food and Drug Administration. Regul. Toxicol. Pharmacol.
Headache 37, 665666. 8, 438446.
Bradstock, M. K., Serdula, M. K., Marks, J. S., Barnard, R. J., Crane, Tollefson, L., Barnard, R. J., and Glinsmann, W. H. (1988). Monitor-
N. T., Remington, P. L., and Trowbridge, F. L. (1986). Evaluation of ing of adverse reactions to aspartame reported to the U.S. Food
reactions to food additives: The aspartame experience. Am. J. Clin. and Drug Administration. In Dietary Phenylalanine and Brain
Nutr. 43, 464469. Function (R. J. Wurtman and E. Ritter-Walker, Eds.), pp. 317337.
Burns, T. S., Stargel, W. W., and Hurwitz, A. (1990). Bioavailability Birkhauser, Boston.
of phenylalanine and aspartate from aspartame (20 mg/kg) in cap- Tschanz, C., Butchko, H. H., Stargel, W. W., and Kotsonis, F. N. (Eds.)
sules and solution. Metabolism 39, 12001203. (1996). The Clinical Evaluation of A Food Additive: Assessment of
Butchko, H. H., and Kotsonis, F. N. (1991). Acceptable daily intake Aspartame. CRC Press, Boca Raton, FL.
vs actual intake: The aspartame example. J. Am. Coll. Nutr. 10, Van Den Eeden, S. K., Koepsell, T. D., Longstreth, W. T., van Belle,
258266. G., Daling, J. R., and McKnight, B. (1994). Aspartame ingestion
Centers for Disease Control (CDC) (1984). Evaluation of Consumer and headaches: A randomized crossover trial. Neurology 44, 1787
Complaints Related to Aspartame Use. CDC, Atlanta, GA. 1793.
Elsas, L. J. (1988). Aspartame and headache (letter). N. Engl. J. Med. Watts, R. S. (1991). Aspartame, headaches and beta blockers (letter).
318, 1201. Headache 31, 181.
Evaluation of Aspartame and Seizures and Electroencephalograms (EEGs)
Introduction findings (Reynolds et al., 1984). There were too few
monkeys in the study, one monkey had severe birth de-
Case reports of seizures anecdotally associated
fects, and all animals in the mid- and high-dose groups
with aspartame consumption (Wurtman, 1985; Walton,
contracted Shigella infections at various times during
1986; Roberts, 1988; Eshel and Sarova-Pinhas, 1993)
the study. Further, there was uncertainty if the mon-
stimulated interest regarding whether aspartame may
keys had ad libitum access to water, which raised the
potentially be associated with seizures. In addition,
question whether hyperosmolality or generalized hy-
anecdotal reports of seizures, which some consumers re-
peraminoacidemia from dehydration may have caused
lated to their consumption of aspartame, were reported
the seizures rather than exposure to phenylalanine.
through passive surveillance and evaluated by the U.S.
To address this uncertainty, another study was done
FDA (Tollefson, 1988) and the Centers for Disease Con-
by Reynolds et al. (1984) with 20 infant monkeys given
trol (CDC, 1984; Bradstock et al., 1986). Tollefson and
up to 3000 mg/kg body wt/day aspartame or 1650 mg/
Barnard (1992) from FDA also specifically analyzed the
kg body wt/day of phenylalanine or control added to
anecdotal reports on seizures associated with aspar-
the diet for 9 months. There were no clinical seizures
tame received from January 1986 through December
observed, and there was no evidence of epileptiform
1990. For approximately one-half (49%) of the 251 re-
activity on EEGs performed at the start of the study
ports received, it was considered highly unlikely that
and at 4 and 9 months. After completion of this study,
there was any association with aspartame. Many of
these same monkeys were evaluated through learning
the remaining cases had contributing factors (e.g., un-
tests, including measures of object discrimination,
derlying or concurrent disease or medications) which
pattern discrimination, and oddity learning (Suomi,
have been previously associated with seizures. The
1984). Overall, the results did not provide any evidence
authors pointed out that only double-blind controlled
of either deficits in learning performance or hearing
studies could definitely determine causality. However,
capabilities in these young macaque monkeys.
based on a review of the surveillance data for aspar-
tame, they concluded that the findings did not merit
a scientific study on the association between aspar- Studies in Animal Seizure Models
tame and seizures in humans. In a subsequent analysis Aspartame was evaluated in several established an-
(Tollefson, 1993) of reports from January 1986 through imal models of seizure. The studies of mice and rats
October 1992, there were then 265 reports of seizure examined the effects of phenylalanine or aspartame on
which some consumers related to their consumption of audiogenic seizures, electroshock seizures, chemically
aspartame. Again, FDA concluded, the anecdotal case induced seizures, and kindling. In light of the hypoth-
reports of seizures received by FDA do not meet the esis that aspartame may promote seizures due to an
criteria for causality. effect on central neurotransmission, the studies done
Nonetheless, given the seriousness of the claims of in genetically epilepsy-prone rats (GEPRs) are espe-
seizure associated with aspartame, numerous studies cially relevant in evaluating whether aspartame causes
were completed to evaluate whether aspartame or its seizures because seizures in these animals are related
phenylalanine component has an effect on seizure sus- to deficits in central noradrenergic and serotonergic
ceptibility in various animal models of seizure. In addi- neurotransmitter systems (Dailey et al., 1991). Further,
tion, several studies in humans have been completed. photically induced myoclonus studied in the epileptic
baboons is similar from an electrophysiological stand-
Studies in Infant Monkeys point to generalized seizures in humans. Seizures in
these animals also may be affected by changes in brain
Early studies had reported that infant monkeys fed a monoamine concentrations (Meldrum et al., 1989).
diet containing 3000 mg/kg body wt phenylalanine/day
experienced grand mal seizures and lasting intellectual Chemically induced models. Pinto and Maher (1988)
deficits. To evaluate whether the phenylalanine pro- administered aspartame doses of 200, 500, 1000, and
vided by aspartame could cause such effects, a pilot 2000 mg/kg body wt to nonepileptic mice 1 h prior to
study with seven newborn rhesus monkeys was initi- pentylenetetrazol (PTZ) treatment and noted a signif-
ated (Rao et al., 1972); unfortunately, only incomplete icant increase in the number of mice convulsing with
data are available from this study due to the untimely the 1000 and 2000 mg/kg body wt aspartame doses.
death of the principal investigator, H. A. Waisman. The same response was noted with administration of
Based on incomplete data with doses ranging up to equimolar amounts of phenylalanine. Guiso and co-
about 3600 mg/kg body wt/day, it was noted that some workers (1988; Garattini et al., 1988) observed an in-
monkeys experienced seizures. However, there were crease in PTZ-induced convulsions in rats with doses of
several defects in the study that may account for these 750 and 1000 mg/kg body wt aspartame and 250 and
S50
ASPARTAME: REVIEW OF SAFETY S51
500 mg/kg body wt phenylalanine administered 1 h be- the aspartame dose groups. The authors concluded that
fore PTZ. As a follow-up to their work with PTZ in rats, in utero exposure to aspartame has no epileptogenic po-
Guiso et al. (1991) reported that tyrosine (1000 mg/kg tential in the offspring in the fluorothyl model.
body wt) protected against the potentiation by aspar- Chin and Woodbury (1988) reported in abstract form
tame on lowering seizure threshold in PTZ-treated an effect of aspartame on bicuculline-induced seizures
rats. Another study from the same laboratory (Diomede in DBA/2 mice; however, a subsequent study also re-
et al., 1991) demonstrated interspecies differences in ported in abstract form did not confirm effects of aspar-
the susceptibility to PTZ-induced seizures in animals tame with bicuculline-induced seizures in DBA/2 mice
given aspartame. They evaluated two strains of mice, (Bettendorf et al., 1989). Further, no effects of oral as-
one strain of guinea pig, and one strain of rat. At doses partame administration were observed on quinolinic
up to 2000 mg/kg body wt given orally to mice and via acid-induced seizures in rats (Guiso et al., 1988) or on
the ip route in guinea pigs, there were no significant theophylline (Zhi and Levy, 1989) and lidocaine seizure
effects; however, effects were seen in rats. models in CD-1 mice (Kim and Kim, 1987; Kim et al.,
In another study in rats, Tilson et al. (1989) found no 1988).
effect of 1000 mg/kg body wt aspartame on PTZ-induced
seizures. Likewise, in a study in mice using doses of Electroshock-induced seizures. With the exception
aspartame up to 500 mg/kg body wt at pretreatment of one report in mice in abstract form (Chin and
intervals of 0.5, 1, and 2 h, there was no effect of as- Woodbury, 1988), there was no effect of aspartame in
partame on seizure threshold after PTZ (Nevins et al., the electroshock paradigm in mice reported by other
1987). Further, a study with PTZ-induced seizures in investigators (Nevins et al., 1987) and in two strains of
the CD-1 mouse (Dailey et al., 1989), using even larger GEPRs with doses up to 3000 mg/kg body wt (Jobe et al.,
doses of aspartame than used in other studies (up to 1992). In addition, experiments in SpragueDawley
2500 mg/kg), demonstrated no effect of aspartame on rats (Guiso et al., 1988) with doses of aspartame of
the convulsive dose in 50% of the animals (CD50 ) of PTZ. 1000 mg/kg body wt failed to modify tonic hindlimb
This same laboratory (Dailey et al., 1988) reported in extension induced by electroshock. Tilson et al. (1989)
abstract form no effect of aspartame on PTZ-induced evaluated the effect of 1000 mg/kg body wt of aspar-
seizures in GEPRs. More recently, Helali et al. (1996) tame on electroshock-induced seizures in Fischer 344
reported that 1000 mg/kg body wt aspartame given rats and found no effect on the number of animals de-
orally did not alter PTZ-induced seizures in mice. How- veloping tonic seizures after electroshock.
ever, subacute oral dosing (100 mg/kg body wt/day for Kindling. Tilson et al. (1989) also evaluated the ef-
14 days) apparently increased the PD50 (median pro- fect of aspartame (1000 mg/kg body wt) on the rate
tective dose) for ethosuximide, sodium valproate, and of kindled seizures in which one group of rat pups
phenytoin given ip, a condition of questionable rele- were dosed on days 313 (neonatal) and another group
vance to oral dosing of these drugs in humans. were dosed on days 2135 (postweaning). The rate of
It is interesting that, although an acute dose of kindling was tested in both groups of rats at 90 days
7501000 mg/kg body wt aspartame on PTZ-induced of age. There were no effects of aspartame on kindled
seizures in rats significantly increased the number of seizures in either group. In another study to evalu-
rats having tonicclonic seizures (Garattini et al., 1988; ate aspartame and kindled seizures, Cain et al. (1989)
Guiso et al., 1988), these authors found no such ef- gave rats up to 200 mg/kg body wt aspartame in the
fect when the dose of aspartame (1000 mg/kg body group evaluated with amygdala-kindled seizures (the
wt) was given in three divided doses over 2 h, after amygdala has the lowest afterdischarge threshold of
a meal, or ingested overnight with food and drinking any structure in rat brain) and up to 2000 mg/kg body
water (2000 mg/kg body wt). This led Garattini and co- wt aspartame in the group with hippocampal-kindled
workers to conclude, All these data taken together in- seizures. Despite these large doses of aspartame, there
dicate that the potentiation of metrazol [pentylenete- was no effect on afterdischarge threshold or seizure
trazol] convulsions by aspartame has no relevance to strength.
the practical use of aspartame even at abuse doses.
Pinto and Maher (1988) reported that aspartame Seizures in susceptible animal strains. Dietary
(10002000 mg/kg) and an equimolar amount of pheny- phenylalanine was reported over 25 years ago to facil-
lalanine (560 mg/kg body wt) decreased latency to itate audiogenic seizures in DBA/2 mice (Schlesinger
fluorothyl-induced convulsions in CD-1 mice. In con- et al., 1969; Truscott, 1975), and subsequently it was
trast, Sperber et al. (1995) demonstrated no effect of reported in abstract form that aspartame increases au-
in utero exposure to aspartame (either 500 or 750 mg/kg diogenic seizure intensity in DBA/2 mice (Chin and
body wt/day) in guinea pigs. At 30 days of age, offspring Woodbury, 1988). Subsequent studies, however, have
were exposed to fluorothyl. Compared to a nontreated found no effect of aspartame (at doses up to 2000 mg/kg
control group, there were no statistically significant dif- body wt) on audiogenic seizures in DBA/2 mice (Jobe
ferences in either clonic or tonic seizures in either of et al., 1989a,b). Also, in a study with two strains of
S52 BUTCHKO ET AL.
model, any effects appear to be largely situational and induced seizures in genetically epilepsy-prone rats. Epilepsia 29,
species specific as no effects were seen when the dosage 651.
was given as divided doses, after a meal, or ingested Dailey, J. W., Lasley, S. M., Mishra, P. K., Bettendorf, A. F.,
Burger, R. L., and Jobe, P. C. (1989). Aspartame fails to facilitate
overnight with food and water. There was also no effect pentylenetetrazol-induced convulsions in CD-1 mice. Toxicol. Appl.
of aspartame on seizures in nonhuman primates. Pharmacol. 98, 475486.
In addition to the animal studies, several studies Dailey, J. W., Lasley, S. M., Burger, R. L., Bettendorf, A. F., Mishra,
were done in humans. The study by Camfield and co- P. K., and Jobe, P. C. (1991). Amino acids, monoamines and audio-
workers (1992) suggested an effect of aspartame in chil- genic seizures in genetically epilepsy-prone rats: Effects of aspar-
dren with absence seizures on the time per hour spent tame. Epilepsy Res. 8, 122133.
in spike-wave discharges. However, the study lacked a Diomede, L., Romano, M., Guiso, G., Caccia, S., Nava, S., and
Salmona, M. (1991). Interspecies and interstrain studies on the in-
true placebo, as the sucrose used as a placebo could have creased susceptibility to metrazol-induced convulsions in animals
itself confounded the results of the study. Further, the given aspartame. Food Chem. Toxicol. 29, 101106.
study had serious methodological issues because of the Elsas, L. J., and Trotter, J. F. (1988). Changes in physiological con-
inadequate length of baseline period for EEG activity. centrations of blood phenylalanine produce changes in sensitive
Individuals who were convinced that aspartame was parameters of human brain function. In Dietary Phenylalanine
and Brain Function (R. J. Wurtman and E. Ritter-Walker, Eds.),
responsible for their seizures and children with seizure
pp. 187195. Birkhauser, Boston.
disorders were also evaluated. The results of these stud-
Epstein, C. M., Trotter, J. F., Averbook, A., Freeman, S., Kutner, M. H.,
ies, using doses of aspartame more than 10 times the and Elsas, L. J. (1989). EEG mean frequencies are sensitive indices
90th percentile average daily aspartame intake in the of phenylalanine effects on normal brain. Electroencephalogr. Clin.
United States, along with the results of other studies Neurophysiol. 72, 133139.
in healthy individuals and PKUH, demonstrated no ef- Eshel, Y., and Sarova-Pinhas, I. (1993). Aspartame and seizures (let-
fects of aspartame on clinical seizures or EEGs. Thus, ter). Neurology 43, 21542155.
when considering the totality of the data, the evidence is Garattini, S., Caccia, S., Romano, M., Diomed, L., Guiso, G., Vezzani,
A., and Salmona, M. (1988). Studies on the susceptibility to con-
clear that aspartame, even in amounts greatly exceed- vulsions in animals receiving abuse doses of aspartame. In Dietary
ing the 90th percentile intake, is not a proconvulsant Phenylalanine and Brain Function (R. J. Wurtman and E. Ritter-
and does not affect seizure susceptibility. Walker, Eds.), pp. 131143. Birkhauser, Boston.
Guiso, G., Caccia, S., Vezzani, A., Stasi, M. A., Salmona, M., Romano,
M., and Garattini, S. (1988). Effect of aspartame on seizures in
REFERENCES various models of experimental epilepsy. Toxicol. Appl. Pharmacol.
96, 485493.
Anderson, G. M., Novotny, E. J., Jr., and Shaywitz, B. A. (1996). Guiso, G., Diomede, L., Romano, M., Caccia, S., Sarati, S., and
Evaluation of seizures. In The Clinical Evaluation of a Food Ad- Salmona, M. (1991). Effect of tyrosine on the potentiation by aspar-
ditive: Assessment of Aspartame (C. Tschanz, H. H. Butchko, W. W. tame and phenylalanine of metrazol-induced convulsions in rats.
Stargel, and F. N. Kotsonis, Eds.), pp. 205216. CRC Press, Boca Food Chem. Toxicol. 29, 855857.
Raton, FL. Helali, N. Y., El-Kashef, H., Salem, H., Gamiel, N., and Elmazar, M. M.
Bettendorf, A. F., Dailey, J. W., Lasley, S. M., and Jobe, P. C. (1989). A. (1996). The effect of aspartame on seizure susceptibility and the
Aspartame fails to facilitate bicuculline-induced seizures in DBA/2 anticonvulsant action of ethosuximide, valproate and phenytoin in
mice. Soc. Neurosci. Abstr. 15, 47. mice. Saudi Pharm. J. 4, 149156.
Bradstock, M. K., Serdula, M. K., Marks, J. S., Barnard, R. J., Crane, Jobe, P. C., and Dailey, J. W. (1993). Aspartame and seizures. Amino
N. T., Remington, P. L., and Trowbridge, F. L. (1986). Evaluation of Acids 4, 197235.
reactions to food additives: The aspartame experience. Am. J. Clin. Jobe, P. C., Bettendorf, A. F., Lasley, S. M., and Dailey, J. W. (1989a).
Nutr. 43, 464469. Multiple exposures to audiogenic stimuli in DBA/2 mice: Lack of
Butchko, H. H., and Kotsonis, F. N. (1991). Acceptable daily intake effect of aspartame. Epilepsia 30, 653.
vs actual intake: The aspartame example. J. Am. Coll. Nutr. 10, Jobe, P. C., Bettendorf, A. F., Lasley, S. M., and Dailey, J. W. (1989b).
258266. Audiogenic seizures in DBA/2 mice are not facilitated by aspartame.
Butchko, H. H., and Stargel, W. W. (2001). Aspartame: Scientific eval- Soc. Neurosci. Abstr. 15, 47.
uation in the postmarketing period. Regul. Toxicol. Pharmacol. 34, Jobe, P. C., Lasley, S. M., Burger, R. L., Bettendorf, A. F., Mishra,
221233. P. K., and Dailey, J. W. (1992). Absence of an effect of aspartame
Cain, D. P., Boon, F., and Bevan, M. (1989). Failure of aspartame on seizures induced by electroshock in epileptic and non-epileptic
to affect seizure susceptibility in kindled rats. Neuropharmacology rats. Amino Acids 3, 155172.
28, 433435. Kim, K. C., and Kim, S. H. (1987). Studies on the effect of aspartame
Camfield, P. R., Camfield, C. S., Dooley, J. M., Gordon, K., Jollymore, and lidocaine interaction in central nervous system in mice. Fed.
S., and Weaver, D. F. (1992). Aspartame exacerbates EEG spike- Proc. 46, 705.
wave discharge in children with generalized absence epilepsy: A Kim, K. C., Tasch, M. D., and Kim, S. H., (1988).The effect of as-
double-blind controlled study. Neurology 42, 10001003. partame on 50% convulsion doses of lidocaine. In Dietary Pheny-
Centers for Disease Control (CDC) (1984). Evaluation of Consumer lalanine and Brain Function (R. J. Wurtman and E. Ritter-Walker,
Complaints Related to Aspartame Use. CDC, Atlanta, GA. Eds.), pp. 127130. Birkhauser, Boston.
Chin, P., and Woodbury, D. M. (1988). Effects of aspartame (ASM) and Meldrum, B. S., Nanji, N., and Cornell, R. G. (1989). Lack of effect of
its metabolites on seizure susceptibility in mice. Pharmacologist 30, aspartame or of L-phenylalanine on photically induced myoclonus
A119. in the baboon, Papio papio. Epilepsy Res. 4, 17.
Dailey, J. W., Lasley, S. M., Bettendorf, A. F., Burger, R., and Jobe, Nevins, M. E., Arnolde, S. M., and Haigler, H. J. (1987). Aspartame:
P. C. (1988). Aspartame does not facilitate pentylenetetrazol- Lack of effect on convulsant thresholds in mice. In Amino Acids in
ASPARTAME: REVIEW OF SAFETY S55
Health and Disease: New Perspectives (S. Kaufman, Ed.), pp. 437 J., and Schomer, D. L. (1998). Aspartame: Neuropsychologic and
449. Liss, Inc., New York. neurophysiologic evaluation of acute and chronic effects. Am. J.
Pinto, J. M. B., and Maher, T. J. (1988). Administration of aspartame Clin. Nutr. 68, 531537.
potentiates pentylenetetrazole- and fluorothyl-induced seizures in Suomi, S. J. (1984). Effects of aspartame on the learning test perfor-
mice. Neuropharmacology 27, 5155. mance of young stumptail macaques. In Aspartame: Physiology and
Rao, K. S., McConnell, R. G., and Waisman, H. A. (1972). SC-18862: Biochemistry (L. D. Stegink and L. J. Filer, Jr., Eds.), pp. 425445.
52 Week Oral Toxicity Study in the Infant Monkey, G. D. Searle and Dekker, New York.
Co. research report submitted to FDA Record, Aspartame 75F-0355 Tilson, H. A., Thai, L., Zhao, D., Sobotka, T. J., and Hong, J. S. (1989).
(File E-32). Oral administration of aspartame is not proconvulsant in rats.
Reynolds, W. A., Bauman, A. F., Stegink, L. D., Filer, L. J., Jr., and NeuroToxicology 10, 229238.
Naidu, S. (1984). Developmental assessment of infant macaques Tollefson, L. (1988). Monitoring adverse reactions to food additives in
receiving dietary aspartame or phenylalanine. In Aspartame: Phys- the U. S. Food and Drug Administration. Regul. Toxicol. Pharmacol.
iology and Biochemistry (L. D. Stegink and L. J. Filer, Jr., Eds.), pp. 8, 438446.
405423. Marcel Dekker, New York. Tollefson, L. (1993). Multiple chemical sensitivity: Controlled scien-
Roberts, H. J. (1988). Aspartame (NutraSweet)-associated epilepsy. tific studies as proof of causation. Regul. Toxicol. Pharmacol. 18,
Clin. Res. 36, 349A. 3243.
Rowan, A. J., Shaywitz, B. A., Tuchman, L., French, J. A., Tollefson, L., and Barnard, R. J. (1992). An analysis of FDA passive
Luciano, D., and Sullivan, C. M. (1995). Aspartame and seizure surveillance reports of seizures associated with consumption of as-
susceptibility: Results of a clinical study in reportedly sensitive in- partame. J. Am. Diet. Assoc. 92, 598601.
dividuals. Epilepsia 36, 270275. Trefz, F., de Sonneville, L., Matthis, P., Benninger, C., Lanz-Englert,
Schlesinger, K., Schreiber, R. A., Griek, B. J., and Henry, K. R. B., and Bickel, H. (1994). Neuropsychological and biochemical in-
(1969). Effects of experimentally induced phenylketonuria on vestigations in heterozygotes for phenylketonuria during ingestion
seizure susceptibility in mice. J. Comp. Physiol. Psychol. 67, 149 of high dose aspartame (a sweetener containing phenylalanine).
155. Hum. Genet. 93, 369374.
Shaywitz, B. A., and Novotny, E. J., Jr. (1993). Reply to Camfield et al. Truscott, T. C. (1975). Effects of phenylalanine and 5-hydroxy-
Aspartame and seizures (letter). Neurology 43, 630. tryptophan on seizure severity in mice. Pharmacol. Biochem.
Shaywitz, B. A., Anderson, G. M., Novotny, E. J., Ebersole, J. S., Behav. 3, 939941.
Sullivan, C. M., and Gillespie, S. M. (1994). Aspartame has no effect Walton, R. G. (1986). Seizure and mania after high intake of aspar-
on seizures or epileptiform discharges in epileptic children. Ann. tame. Psychosomatics 27, 218, 220.
Neurol. 35, 98103. Wurtman, R. J. (1985). Aspartame: Possible effect on seizure suscep-
Sperber, E. F., Moshe, S. L., and Dow-Edwards, D. L. (1995). Prena- tibility (letter). Lancet 2, 1060.
tal exposure to aspartame and seizure susceptibility. J. Epilepsy 8, Zhi, J., and Levy, G. (1989). Aspartame and phenylalanine do not
5156. enhance theophylline-induced seizure in rats. Res. Commun. Chem.
Spiers, P. A., Sabaounjian, L., Reiner, A., Myers, D. K., Wurtman, Pathol. Pharmacol. 66, 171174.
Evaluation of Aspartame and Behavior, Cognitive Function, and Mood
Introduction between Phe/LNAA and objectively measured behav-
ioral effects.
It was suggested that aspartame might increase the
Tilson et al. (1991) evaluated sensorimotor func-
ratio (Phe/LNAA) of phenylalanine (Phe) to the other
tion, learning, and memory in several experimental
large neutral amino acids (LNAAs) that compete for
paradigms in rats. Neither the acute nor repeated
entry into the brain and that, as such, aspartame
(14 days) administration of aspartame by gavage (up
might affect central neurotransmitter concentrations
to 1000 mg/kg body wt) had a significant effect on rou-
(see Phenylalanine and Neurochemistry section). It was
tine measures of sensorimotor function, including spon-
further suggested that this, in turn, might affect brain
taneous motor activity or acoustic startle reflex and
function, leading to changes in behavior, cognitive func-
prepulse inhibition. They also found no effect of aspar-
tion, or mood. Although no behavioral effects had been
tame (500 or 1000 mg/kg body wt) on acquisition of pas-
noted in animals in the preapproval toxicology test-
sive or active avoidance or a spatial, reference memory
ing, a number of studies in animals, using various test-
task in the Morris water maze. These authors concluded
ing paradigms, were subsequently done to evaluate be-
that large doses of aspartame have no significant effects
havior specifically. In addition, studies have been done
in adult rats when measured by procedures known to
in healthy adults, PKU heterozygous adults (PKUH),
be sensitive to the neurobiological effects of neurotoxi-
healthy children, and children with attention deficit dis-
cants.
order (ADD) or hyperactivity (ADHD) or in children
Goerss et al. (2000) used the standard resident
who were thought to be sugar sensitive.
intruder paradigm to evaluate whether aspartame
(200, 400, and 800 mg/kg body wt given ip) had an
Studies with Animals effect on aggression in rats. According to the authors,
previous studies had demonstrated an inverse relation-
In early studies, several groups examined the effects
ship between aggression and concentrations of brain
of aspartame and phenylalanine on animal behavior.
serotonin. Goerss et al. (2000) similarly noted a signifi-
In studies of learning performance in young stumptail
cant increase in striatal serotonin concentrations after
macaque monkeys, no effects were seen for doses of as-
aspartame that was associated with a decrease in ag-
partame up to 3000 mg/kg body wt given chronically
gressive behaviors in their experimental animals. How-
(Suomi, 1984). Learning tests included measures of ob-
ever, intraperitoneal injection of aspartame used in this
ject discrimination, pattern discrimination, and oddity
study is not relevant to human exposure.
learning and were given to 20 macaques who had re-
LaBuda and Hale (2000) evaluated whether aspar-
ceived a control diet, diet with aspartame (1000, 2000,
tame interfered with the anxiolytic actions of ethanol,
or 3000 mg/kg body wt), or diet with phenylalanine only
as measured by plus-maze performance of rodents. As-
added (1650 mg/kg body wt) for 270 days beginning in
partame (1000 and 2000 mg/kg body wt) was followed
infancy. Overall, the results did not provide any evi-
by either a vehicle or ethanol, all given ip. The aspar-
dence of significant learning performance or hearing
tame had no significant effect on anxiety-related be-
differences between these groups of young macaques.
havior, nor did it interfere with the anxiolytic action of
Several other groups have used different models to
ethanol. However, as noted above, the relevance of ip
evaluate whether aspartame has an effect on behavior
aspartame dosing in animals is questionable.
but some, such as Holder and Yirmiya (1989), used an
ip dosing paradigm. This method of administration is
not relevant to the dietary intake and oral consumption
Studies with Healthy Human Adults
of aspartame by humans.
Mullenix et al. (1991) evaluated spontaneous behav- Ryan-Harshman and co-workers (1987) evaluated
ior in rats 1 h after large doses of aspartame (500 the effect of aspartame (5040 and 10,080 mg) and
and 1000 mg/kg body wt), phenylalanine, and tyro- phenylalanine (840, 2520, and 5040 mg) on energy in-
sine using a computerized pattern recognition system take and macronutrient selection and on feelings of
that identified and classified 13 different behavioral hunger, mood, and arousal in healthy adult males. De-
acts. Motor output was evaluated by the number of be- spite significant increases in plasma phenylalanine con-
havioral initiations, total time that activity was per- centrations and Phe/LNAA (high mean Phe/LNAA was
formed, and a calculation of the behavioral sequence 0.397 after the 10,080-mg dose of aspartame), these au-
structure. Amphetamine was used as a positive control. thors found no effect of aspartame or phenylalanine on
Neither aspartame, nor phenylalanine, nor tyrosine short-term energy and macronutrient intakes or sub-
induced any changes in spontaneous motor behavior jective feelings of hunger, mood, or arousal.
output despite significant changes in Phe/LNAA. The In a study to determine whether aspartame af-
investigators concluded that there was no correlation fected neuropsychological function in 20 normal adults
S56
ASPARTAME: REVIEW OF SAFETY S57
0.40
0.35
15 mg/kg bw/day
Phe/LNAA
0.30
0.25 (1.5 hours post dose)
0.20
0.15
*
0.10
0.05
0.00
Aspartame Placebo Sucrose
0.40
0.35
0.30 45 mg/kg bw/day
Phe/LNAA
* Significantly different from placebo and sucrose treatments. P < 0.05 (planned comparison).
FIG. 1. Phe/LNAA 1.5 h after aspartame, placebo, and sucrose treatments on day 10 in healthy adults. Adapted from Spiers et al. (1998).
Subjects were tested with the Amsterdam Neuropsy- a sweet placebo control (Wolraich et al., 1985; Goldman
chological Tasks battery (de Sonneville, 1999), includ- et al., 1986; Milich and Pelham, 1986; Rosen et al., 1988;
ing tests for visual information processing and exec- Mahan et al., 1988; Wender and Solanto, 1991; Hoover
utive function of varying complexity. This computer- and Milich, 1994).
ized test battery has demonstrated changes in cognitive
Studies with normal children. In a two-phase study
performance to be associated with changes of plasma
with 20 children aged 910 years (Saravis et al., 1990),
phenylalanine concentrations in individuals with PKU
single doses of aspartame (34 mg/kg body wt) and cy-
(Schmidt et al., 1994), even in a very well-controlled
clamate in equivalent sweetness amounts were given
population (Huijbregts et al., 2002). There was no dif-
with carbohydrate in one phase, and, in the other,
ference between either dose of aspartame and placebo
aspartame (approximately 10 mg/kg body wt) was com-
on this battery of neuropsychological tests.
pared with sucrose. Measurements included an asso-
ciative learning task, an arithmetic test, the Childrens
Studies with Children
Depression Inventory, the State Scale of the State-
Many of the studies in children have examined the Trait Anxiety Inventory for Children, and several ob-
behavioral and cognitive effects of aspartame in chil- servational measures. The results suggested that when
dren with attention deficit disorders (ADD), with hy- consumed with carbohydrate, aspartame had no effect
peractivity (ADHD), or with alleged sugar sensiti- on learning, behavior, or mood. The authors reported
vity (Wolraich et al., 1985; Milich and Pelham, decreased minor and gross motor behavior following
1986; Mahan et al., 1988; Hoover and Milich, 1994; sucrose consumption, as compared with aspartame.
Shaywitz et al., 1994b). A number of studies have Goldman et al. (1986) used a double-blind, crossover
evaluated both normal children and children with design to evaluate the effect of sucrose added to juice
ADD, ADHD, or sugar sensitivity (Ferguson et al., on the behavior of eight preschool children. Aspartame
1986; Kruesi et al., 1987; Wender and Solanto, 1991; was used as the sweet placebo control. A decrement
Wolraich et al., 1994), while some studies have focused was found in performance on structured testing and in-
only on normal children (Goldman et al., 1986; Rosen creased inappropriate behaviors were observed dur-
et al., 1988; Saravis et al., 1990). In addition, one study ing free play after the children consumed sucrose but
evaluated behavior in children with seizure disorders not aspartame. Goldman et al. (1986) concluded that
given aspartame (Shaywitz et al., 1994a). Many of the this study provided objective evidence of an effect of
studies with children were primarily concerned with as- sucrose on behavior. In another double-blind, crossover
sessing the effects of sugar while aspartame was used as study in which aspartame was the placebo, Rosen et al.
ASPARTAME: REVIEW OF SAFETY S59
(1988) evaluated the effect of sucrose on behavior in 45 more hyperactive than those in the placebo (aspartame)
preschool and elementary school children. Only small group.
effects were observed and only after sucrose intake. As Shaywitz et al. (1994b) conducted a randomized,
such, the authors concluded that these were not clini- double-blind, placebo-controlled, crossover study with
cally significant. aspartame in children diagnosed with ADD. In this
In a study of the effect of sweetness on newborn cry- study, 15 children with ADD received 34 mg/kg body
ing, Barr et al. (1999) compared the effects of sucrose, wt/day of aspartame and placebo for 2 weeks on each
aspartame, polycose (a nonsweet carbohydrate), and treatment. Several measures of behavior and cogni-
water. Relative to water, both sucrose and aspartame re- tive function were evaluated, including the Matching
duced crying and transiently increased mouthing and Familiar Figures Test, Childrens Checking Task, the
handmouth contact. Polycose had no effect on either Airplane Test, the Wisconsin Card Sorting Test, the
parameter. The authors concluded that newborns re- Multigrade Inventory for Teachers, and the Conners Be-
spond preferentially to sweetness rather than to carbo- havior Rating Scale, as well as the Subjects Treatment
hydrate, but there was no difference between sucrose Emergent Symptom Scale. There were no clinically sig-
and aspartame. nificant differences between aspartame compared with
placebo on behavior or cognition measures in these po-
Studies with ADD, ADHD, or sugar-sensitive children. tentially sensitive children. In another study in 10 chil-
Children with ADD and ADHD represent especially in- dren with seizure disorders, Shaywitz et al. (1994a) also
teresting sample populations. Specifically, it has been evaluated behavior after 2 weeks of aspartame (34 mg/
proposed that the three major monoamine neurotrans- kg body wt/day) compared to 2 weeks of placebo using a
mitters, dopamine, norepinephrine, and serotonin, may modified Connors Behavior Rating Scale. There was no
all play a role in the pathophysiology of these disor- statistically significant difference between aspartame
ders. The possibility that aspartame might affect the and placebo treatments on the behavior ratings made
biosynthesis of these monoamine neurotransmitters led by either the teachers or the parents.
to speculation that aspartame might cause neurochem-
ically mediated changes in ADD- and ADHD-related be- Studies with both normal and sensitive children.
haviors. Ferguson et al. (1986) evaluated behavior and cog-
Wolraich et al. (1985) evaluated sucrose compared nitive function in eight boys and girls ages 5 to 13
to placebo (aspartame) on behavior (playroom obser- years who were believed by their parents to have ad-
vations and examiner ratings) and cognition (learning verse responses to sugar (five with ADD, one with a
and memory tasks) in 16 hyperactive boys. They found neuropsychological disorder, and two with no diagnos-
no effects of sucrose or aspartame on behavior or cogni- able disorder). The study was a randomized, double-
tive function. Milich and Pelham (1986) evaluated the blind, crossover design comparing sugar and aspar-
effect of sugar on behavior and academic performance tame. There was no evidence for a sucrose effect vs
in 16 boys diagnosed with ADD. Aspartame was used aspartame and no change from baseline after either
as the placebo. They found no effect on classroom sucrose or aspartame. These investigators also used a
behavior, noncompliance with adult requests, and peer randomized, double-blind, crossover design to compare
interactions and no effects on academic productivity activity level (actometer, parental and teacher assess-
and accuracy. ments), behavior, fine and gross motor coordination, and
In another study of 16 children who were described visualmotor coordination in 18 normal preschool boys
as overly aggressive, loud, and noncompliant after in- and girls after sugar vs aspartame. When evaluating
gesting sugar, Mahan et al. (1988) found no significant the group data, there was no difference between aspar-
behavior changes using an open-label paradigm. Five tame or sucrose and baseline in behavior, activity level,
children who had changed from baseline on a few of or fine or gross motor coordination; only after sucrose,
the parameters tested were then evaluated more closely was there a difference in comparison to the baseline
with sugar, honey, tapioca, and starch in a double-blind and aspartame conditions on a developmental drawing
challenge that used aspartame as a placebo. No con- task.
sistent effects were found for any of the challenge sub- Kruesi et al. (1987) evaluated aggression and activ-
stances on either learning or attention span. ity after sugar, aspartame, saccharin, and glucose in
To examine the effect of sugar ingestion on mother a double-blind, crossover study in 30 preschool boys
child interactions, Hoover and Milich (1994) studied 35 18 considered sugar reactive by their parents and 12
boys, 5 to 7 years old, who were reportedly sugar sen- healthy boys. There were no effects after any treatment
sitive. Mothers were told that the children would re- on aggression or observers ratings of behavior. There
ceive either sugar or placebo (aspartame) but all sub- were lower actometer counts after aspartame, which
jects, in fact, received only aspartame. Mothers who were not verified by observer ratings. The authors con-
were told that their children were in the experimental, cluded that neither aspartame nor sugar is a clinically
supposedly sugar, group rated their children as being significant cause of disruptive behavior in children.
S60 BUTCHKO ET AL.
In another study of 9 normal children and 17 children adverse effects of aspartame exposure on behavior. Nor
with ADD or ADHD, Wender and Solanto (1991) eval- have they provided any mechanistic or neurophysiolog-
uated the effect of sugar in comparison to aspartame ical basis for any such putative effects.
and saccharin placebos on attention and aggressive be- It is perhaps not surprising, therefore, that controlled
havior. Children with ADHD were significantly more studies in humans have similarly failed to demonstrate
aggressive than were normal children; however, there any adverse behavioral or cognitive effects of aspar-
were no effects of sugar, aspartame, or saccharin on the tame. This body of literature has included studies con-
incidence of aggressive behaviors. Inattention, however, ducted with healthy and potentially vulnerable (PKUH)
was greater in the ADHD group after sugar consump- adults and with healthy and potentially vulnerable chil-
tion compared to either aspartame or saccharin. dren, including children with ADD, ADHD, or alleged
A study of the effects of diets with sucrose, aspartame, sugar sensitivity, or in children with seizure disorders.
and saccharin on behavior and cognitive performance in Various objective tests, self-rating scales, and natural-
normal children has also been reported (Wolraich et al., istic observations have been utilized in protocols inves-
1994). Twenty-three school-age children who had been tigating both acute and chronic exposure to high doses
reported to respond adversely to sugar and 24 normal of aspartame. Overall, a considerable body of human re-
preschool children were given each regimen. The chil- search has demonstrated that aspartame has no effect
dren received a different diet for each of three consecu- on behavior, cognitive function, or mood.
tive 3-week periods; one diet was high in sucrose with
no high-intensity sweeteners; another diet was low in REFERENCES
sugar and contained aspartame; and the other was low
in sugar and contained saccharin. Preschool children Barr, R. G., Pantel, M. S., Young, S. N., Wright, J. H., Hendricks, L. A.,
consumed a mean (SD) of 5600 2100 mg/kg body and Gravel, R. (1999). The response of crying newborns to sucrose:
wt/day sucrose while on the sucrose diet, 38 13 mg/kg Is it a sweetness effect? Physiol. Behav. 66, 409417.
body wt/day aspartame while on the aspartame diet, de Sonneville, L. M. J. (1999). Amsterdam Neuropsychological Tasks:
and 12 4.5 mg/kg body wt/day saccharin while on the A computer-aided assessment program. In Cognitive Ergonomics,
Clinical Assessment and Computer-Assisted Learning: Computers
saccharin diet. The school-age children thought to be in Psychology (B. P. L. M. Den Brinker, P. J. Beek, A. N. Brand, S. J.
sensitive to sugar consumed 4500 1200 mg/kg body Maarse, and L. J. M. Mulder, Eds.), pp. 187203. Swets & Zeitlinger,
wt/day sucrose, 32 8.9 mg/kg body wt/day aspartame, Lisse, The Netherlands.
and 9.9 3.9 mg/kg body wt/day of saccharin while on Dodge, R. E., Warner, D., Sangal, S., and ODonnell, R. D. (1990).
those diets. Behavior and cognitive performance were The effect of single and multiple doses of aspartame on higher cog-
evaluated weekly. For children described as being sugar nitive performance in humans. In 61st Annual Scientific Meeting,
Aerospace Medical Association, p. A30. New Orleans, LA.
sensitive, there was no difference among the three diets
Ferguson, H. B., Stoddart, C., and Simeon, J. G. (1986). Double-blind
on any of 39 behavioral and cognitive parameters. In the challenge studies of behavioral and cognitive effects of sucrose-
normal preschool children, only 4 of the 31 variables dif- aspartame ingestion in normal children. Nutr. Rev. 44, 144150.
fered among the diets, but there was no consistent pat- Goerss, A. L., Wagner, G. C., and Hill, W. L. (2000). Acute effects of
tern in the differences. The parents rating of cognition aspartame on aggression and neurochemistry of rats. Life Sci. 67,
was significantly better with sucrose than either aspar- 13251329.
tame or saccharin, and the grooved pegboard perfor- Goldman, J. A., Lerman, R. H., Contois, J. H., and Udall, Jr., J. N.
(1986). Behavioral effects of sucrose on preschool children. J. Ab-
mance was significantly slower after sucrose compared
norm. Child Psychol. 14, 565577.
to the other sweeteners for both dominant and non-
Holder, M. D., and Yirmiya, R. (1989). Behavioral assessment
dominant hands and the total score. The authors con- of the toxicity of aspartame. Pharmacol. Biochem. Behav. 32,
cluded that neither sucrose nor aspartame in amounts 1726.
exceeding typical dietary levels predictably affects be- Hoover, D. W., and Milich, R. (1994). Effects of sugar ingestion ex-
havior or cognitive function in children. pectancies on motherchild interactions. J. Abnorm. Child Psychol.
22, 501515.
Huijbregts, S. C. J., de Sonneville, L. M. J., Licht, R., van Spronsen,
Conclusion F. J., Verkerk, P. H., and Sergeant, J. A. (2002). Sustained atten-
tion and inhibition of cognitive interference in treated phenylke-
Numerous animal studies have been carried out us- tonuria: Associations with concurrent and lifetime phenylalanine
ing various testing paradigms and very large doses of concentrations. Neuropsychologia 40, 715.
aspartame that have sometimes been hundreds of times Kruesi, M. J. P., Rapoport, J. L., Cummings, E. M., Berg, C. J., Ismond,
the 90th percentile intake for humans in the United D. R., Flament, M., Yarrow, M., and Zahn-Waxler, C. (1987). Effects
States. The only effects observed were after ip dosing, of sugar and aspartame on aggression and activity in children. Am.
J. Psychiatry 144, 14871490.
a paradigm that is not comparable to the dietary con-
LaBuda, C. J., and Hale, R. L. (2000). Anxiety in mice following acute
sumption of aspartame in humans. Taken as a whole, aspartame and ethanol exposure. Alcohol 20, 6974.
and despite significant changes in Phe/LNAA ratios, the Lapierre, K. A., Greenblatt, D. J., Goddard, J. E., Harmatz, J. S., and
findings from the animal research reviewed fail to pro- Shader, R. I. (1990). The neuropsychiatric effects of aspartame in
vide support for what have been purported to be the normal volunteers. J. Clin. Pharmacol. 30, 454460.
ASPARTAME: REVIEW OF SAFETY S61
Lieberman, H. R., Caballero, B., Emde, G. G., and Bernstein, J. G. Sullivan, B., and Shaywitz, S. E. (1994b). Aspartame, behavior, and
(1988). The effects of aspartame on human mood, performance, cognitive function in children with attention deficit disorder. Pedi-
and plasma amino acid levels. In Dietary Phenylalanine and Brain atrics 93, 7075.
Function (R. J. Wurtman and E. Ritter-Walker, Eds.), pp. 196200. Spiers, P., Schomer, D., Sabounjian, L., Lieberman, H., Wurtman, R.,
Birkhauser, Boston. Duguid, J., McCarten, R., and Lyden, M. (1988). Aspartame and hu-
Mahan, L. K., Chase, M., Furukawa, C. T., Sulzbacher, S., Shapiro, man behavior: congnitive and behavioral observations. In Dietary
G. G., Pierson, W. E., and Bierman, C. W. (1988). Sugar allergy Phenylalanine and Brain Function (R. J. Wurtman and E. Ritter-
and childrens behavior. Ann. Allergy 61, 453458. Walker, Eds.), pp. 169178. Birkhauser, Boston.
Milich, R., and Pelham, W. E. (1986). Effects of sugar ingestion on the Spiers, P. A., Sabounjian, L., Reiner, A., Myers, D. K., Wurtman, J.,
classroom and playgroup behavior of attention deficit disordered and Schomer, D. L. (1998). Aspartame: Neuropsychologic and neu-
boys. J. Consult. Clin. Psychol. 54, 714718. rophysiologic evaluation of acute and chronic effects. Am. J. Clin.
Mullenix, P. J., Tassinari, M. S., Schunior, A., and Kernan, W. J. (1991). Nutr. 68, 531537.
No change in spontaneous behavior of rats after acute oral doses Stokes, A. F., Belger, A., Banich, M. T., and Taylor, H. (1991). Effects
of aspartame, phenylalanine, and tyrosine. Fundam. Appl. Toxicol. of acute aspartame and acute alcohol ingestion upon the cognitive
16, 495505. performance of pilots. Aviat. Space Environ. Med. 62, 648653.
Pivonka, E. E. A., and Grunewald, K. K. (1990). Aspartame- or sugar- Stokes, A. F., Belger, A., Banich, M. T., and Bernadine, E. (1994). Ef-
sweetened beverages: Effects on mood in young women. J. Am. Diet. fects of alcohol and chronic aspartame ingestion upon performance
Assoc. 90, 250254. in aviation relevant cognitive tasks. Aviat. Space Environ. Med. 65,
Rosen, L. A., Bender, M. E., Sorrell, S., Booth, S. R., McGrath, M. L., 715.
and Drabman, R. S. (1988). Effects of sugar (sucrose) on childrens Suomi, S. J. (1984). Effects of aspartame on the learning test per-
behavior. J. Consult. Clin. Psychol. 56, 583589. formance of young stumptail macaques. In Aspartame: Physiology
Ryan-Harshman, M., Leiter, L. A., and Anderson, G. H. (1987). Pheny- and Biochemistry (L. D. Stegink and L. J. Filer, Jr., Eds.), pp. 425
lalanine and aspartame fail to alter feeding behavior, mood and 445. Marcel Dekker, New York.
arousal in men. Physiol. Behav. 39, 247253. Tilson, H. A., Hong, J. S., and Sobotka, T. J. (1991). High doses of
Saravis, S., Schachar, R., Zlotkin, S., Leiter, L. A., and Anderson, aspartame have no effects on sensorimotor function or learning
G. H. (1990). Aspartame: Effects on learning, behavior, and mood. and memory in rats. Neurotoxicol. Teratol. 13, 2735.
Pediatrics 86, 7583. Trefz, F., de Sonneville, L., Matthis, P., Benninger, C., Lanz-Englert,
Schmidt, E., Rupp, A., Burgard, P., Pietz, J., Weglage, J., and De B., and Bickel, H. (1994). Neuropsychological and biochemical in-
Sonneville, L. (1994). Sustained attention in adult phenylke- vestigations in heterozygotes for phenylketonuria during ingestion
tonuria: The influence of the concurrent phenylalanine blood level. of high dose aspartame (a sweetener containing phenylalanine).
J. Clin. Exp. Neuropsychol. 16, 681688. Hum. Genet. 93, 369374.
Schomer, D. L., Spiers, P., and Sabounjian, L. A. (1996). Evaluation Wender, E. H., and Solanto, M. V. (1991). Effects of sugar on aggres-
of behavior, cognition, mood, and electroencephalograms in normal sive and inattentive behavior in children with attention deficit dis-
adults and potentially vulnerable populations. In The Clinical Eval- order with hyperactivity and normal children. Pediatrics 88, 960
uation of a Food Additive: Assessment of Aspartame (C. Tschanz, 966.
H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 217233. Wolraich, M., Milich, R., Stumbo, P., and Schultz, F. (1985). Effects
CRC Press, Boca Raton, FL. of sucrose ingestion on the behavior of hyperactive boys. J. Pediatr.
Shaywitz, B. A., Anderson, G. M., Novotny, E. J., Ebersole, J. S., 106, 675682.
Sullivan, C. M., and Gillespie, S. M. (1994a). Aspartame has no Wolraich, M. L., Lindgren, S. D., Stumbo, P. J., Stegink, L. D.,
effect on seizures or epileptiform discharges in epileptic children. Appelbaum, M. I., and Kiritsy, M. C. (1994). Effects of diets high in
Ann. Neurol. 35, 98103. sucrose or aspartame on the behavior and cognitive performance of
Shaywitz, B. A., Sullivan, C. M., Anderson, G. M., Gillespie, S. M., children. N. Engl. J. Med. 330, 301307.
Evaluation of Aspartame and Allergic-Type Reactions
Introduction were reproducible upon double-blind challenge. The
authors concluded that it was difficult to find and study
Allergic-dermatologic reactions attributed to aspar-
subjects thought to be sensitive to aspartame and that
tame ingestion accounted for about 15% of the anec-
blinded oral challenge of patients with presumed aspar-
dotal complaints evaluated by the Centers for Disease
tame sensitivity failed to result in reproducible results.
Control (CDC, 1984). Included as allergic complaints
were rashes, sore throat/mouth, swelling, and itching.
Multicenter Study with Aspartame
In addition to the CDC evaluation, there have been two
case reports of urticaria (Kulczycki, 1986) and two case A multicenter, randomized, double-blinded, placebo-
reports of granulomatous panniculitis (Novick, 1985; controlled, crossover clinical study conducted in clinical
McCauliffe and Poitras, 1991) thought to be related to research units over a 5-day period was designed to eval-
aspartame. However, the pitfalls of single case reports, uate if allergic-type symptoms could be attributable to
even those done under double-blind, placebo-controlled, aspartame (Geha et al., 1993). Subjects were required
crossover conditions, were highlighted by Geha (1992), to have a history of urticaria or angioedema within 12 h
who emphasized the importance of well-controlled clin- of consuming an aspartame-containing product during
ical studies, rather than single case reports, to provide the previous 3 years or a history of chronic urticaria,
valid answers to medically important issues. which resolved without medication upon cessation of
aspartame and recurred upon resumption of aspartame
Animal Studies with Aspartame consumption.
Subjects were given, in random order, (1) capsules
Studies have shown that aspartame is not a direct
containing 25 or 300 mg aspartame, capsules contain-
mast cell or basophil secretagogue, either by in vitro
ing 7.5 mg DKP, and 3.75 mg -aspartame (conversion
methods or by skin testing in vivo (Szucs et al., 1986).
products of aspartame) with 13.75 mg microcrystalline
In addition, studies in rats have shown that aspar-
cellulose and (2) identical placebo capsules containing
tame has no effect on inflammation parameters such
25 or 300 mg microcrystalline cellulose. The aspartame
as carrageenan-induced paw edema, connective tissue
used in the study also contained approximately 0.32%
formation, and adjuvant arthritis (Aspinall et al., 1980).
DKP and 0.35% -aspartame. Conversion products of
Nonetheless, in order to resolve the question of allergic-
aspartame, DKP and -aspartame, were included in
type reactions attributed to aspartame, controlled clin-
the study to determine if allergic-type reactions may be
ical trials were conducted.
due to them rather than aspartame. Meals were stan-
dardized on both treatment days, and no aspartame-
Single-Site Study with Aspartame
containing products were included. In addition, all food
Garriga et al. (1991) conducted a study with placebo- consumed on test days was recorded.
controlled oral challenges in patients who reported sen- Subject recruitment efforts continued for over 4 years
sitivity to aspartame. Although extensive efforts were and included letters to over 4700 allergists in the
made to recruit subjects through local newspaper ad- United States and Canada, advertisements in two ma-
vertisements and interactions with local allergists and jor allergy journals and 11 local allergy and dermatol-
dermatologists, recruitment of suitable study subjects ogy societies, and contacts with 102 individuals who
was extraordinarily difficult compared to the authors had filed complaints with The NutraSweet Company of
experience in other studies of possible food sensitivity. urticaria and/or angioedema thought to be related to
As a result, only 61 initial subjects were identified af- aspartame. However, despite these efforts, only 21 sub-
ter 32 months of enrollment. Upon screening, only 20 of jects, who were convinced that aspartame caused their
these 61 patients had histories and symptoms to war- allergic-type reactions, were enrolled into the study
rant evaluation by oral aspartame challenge. (Table 1). Following challenge, 17 of the 21 subjects had
Prior to oral challenge, each patient underwent skin no allergic-type reactions. Four subjects experienced
testing to a panel of aeroallergens and some foods as allergic-type reactionstwo had reactions following in-
well as both aspartame and diketopiperazine (DKP), a gestion of aspartame but not placebo, while the two oth-
cyclic compound formed from aspartame under condi- ers had reactions following ingestion of placebo but not
tions of extremes in temperature and pH and long stor- aspartame. There was no statistically significant differ-
age times in beverages. Of the 12 patients who enrolled, ence in the incidence of allergic-type reactions between
9 had negative single-blinded challenges. Of the 3 pa- aspartame and placebo challenges (P = 1.00) (Fig. 1). Of
tients with positive reactions to single-blind challenge, the positive reactions, only one individual experienced
1 developed hives, 1 reported throat tightness, and a severe reaction consisting of generalized urticaria,
1 developed rhinitis. However, none of these reactions which persisted for 4 months and required treatment
S62
ASPARTAME: REVIEW OF SAFETY S63
7
quent step increase. An important factor in the over-
*
all increase in malignant childhood brain tumors was
6
an increase in low-grade gliomas, which may be ac-
5
counted for by more ready detection of such tumors
4
by MRI than by CT. The availability of MRI was very
3 restricted prior to 1985. In addition, there was also a
2 change in the classification system for childhood brain
1 tumors during the mid-1980s whereby some tumors
0 (e.g., low-grade gliomas) previously classified as be-
73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 nign were then classified as malignant and therefore
* Aspartame marketing in late 1981
Year subsequently included for the first time in the SEER
database. There were also changes in neurosurgical
FIG. 1. Age-adjusted brain and other CNS tumor rates from 1973 practices (e.g., stereotactic biopsies) in the mid-1980s
to 1996 from the SEER registry (NCI, 1999). Increases in brain tumor that may have led to increased diagnosis and reporting
rates started before aspartame marketing in 1981. Source: NCI, 1999.
of brain tumors in children. These authors attributed
this finding to changes in the mid-1980s in the detec-
tion and/or reporting of brain tumors in children. A sub-
In addition, increases in brain tumor rates have been sequent evaluation of SEER data for childhood brain
primarily in the elderly, especially the very elderly tumors by Linet et al. (1999) supported this conclusion.
(Muir et al., 1994; Greig et al., 1990; Werner et al., 1995; Olney et al. (1996) also purported an increase in the
Davis et al., 1991). For example, Greig et al. (1990) malignancy of brain tumors after aspartame market-
reported a 500% increase in brain tumors in people ing. As with brain tumors in general, increases in highly
85 years and older. This has led to considerable de- malignant glioblastomas were primarily in the elderly
bate regarding whether the reported increases in the in the SEER database (Levy and Hedeker, 1996) as has
elderly represent a real increase in brain tumor rates also been reported by Werner et al. (1995). Thus, in-
or whether other factors account for this observation creases in glioblastomas are largely accounted for by
(Muir et al., 1994; Greig et al., 1990; Werner et al., increases in the elderly and may be the result of the
1995; Davis et al., 1991; La Vecchia et al., 1992; Boyle factors described by Modan et al. (1992). The increase
et al., 1990; Marshall, 1990). This issue was critically in anaplastic astrocytomas referenced by Olney et al.
reviewed by Modan et al. (1992), who concluded that is likely an artifact of recent shifts in classification
the observed increases were not real but were related terminology. Werner et al. (1995) cited wider accep-
to enhanced detection resulting from availability of tance of a three-tier classification system (astrocytoma,
more sophisticated noninvasive diagnostic technology; anaplastic astrocytoma, and glioblastoma) during the
change in the attitude toward care of the elderly; and mid-1980s. Thus, some tumors classified as astrocy-
introduction of support programs such as Medicare that tomas in the early 1980s would have been classified as
facilitate diagnostic procedures in the elderly. anaplastic astrocytomas by the mid- to late 1980s.
Legler et al. (1999) evaluated incidence data for brain
and other central nervous system (CNS) tumors from
the SEER results from 1975 to 1995 and concluded that,
in contrast to earlier reports, there was a level or de- 50
Percent change
clining trend in brain tumor rates for all but the most 40
elderly people. The fact that the elderly largely account
30
for increases in overall brain tumor rates is clear from
a recent SEER report (NCI, 1999). The percent change 20
in brain tumor incidence for all ages from 1973 to 1996 10
was 15.8%; however, for ages under 65 years, the per-
cent change was only 6.0%, whereas it was 46.5% for 0
All ages Under 65 65 and
ages 65 and over during this time period (Fig. 2). As the over
most avid aspartame consumers are young and middle-
aged adults (personal communication, confidential mar- FIG. 2. Percentage of change in brain and other CNS tumor in-
keting data from MRCA Information Services, 1995), cidence for all ages, under age 65 years, and 65 years and older from
1973 to 1996 in the SEER registry (NCI, 1999). Much of the reported
marked increases in brain tumor incidence in the el- increase in brain tumor incidence in the United States has resulted
derly do not correlate with the demographics of aspar- from increases in the elderly population, which does not represent
tame use. avid aspartame consumers. Source: NCI, 1999.
ASPARTAME: REVIEW OF SAFETY S67
Further, if there were truly an increase in the malig- demonstrated that two events occurred during roughly
nancy of brain tumors, one would have expected that the same time period. There is no information available
mortality from brain tumors would have risen. On the regarding whether the individuals who developed brain
contrary, the percent change in mortality for all ages tumors consumed aspartame. For example, one might
from 1975 to 1979 was an increase of 2.5% compared also invoke (a) cellular phone, home computer, and VCR
to a decrease of 2.1% from 1992 to 1996, and 5-year usage; (b) depletion of the ozone layer; or (c) increased
relative survival rates were significantly higher from use of stereo headphones as potentially causative
19891995 compared to 19741976 (NCI, 1999). Thus, agents to argue trends in brain tumors and the chang-
changes in SEER data during the mid-1980s can be ing environment. All such events could potentially be
explained largely by changes in the classification sys- positively correlated with brain tumor incidence, and
tem of brain tumors and the ability to diagnose tumors some or all of these possibilities may or may not have
earlier due to the availability of advanced neuroimag- any biological plausibility to the observed associations.
ing techniques and changes in neurosurgical practices, Recently, Weihrauch et al. (2001) reviewed potential
such as the availability of stereotactic biposies. Declin- carcinogenicity of high-intensity sweeteners. Regard-
ing mortality is not consistent with a true increase in ing the allegations with aspartame, they concluded
brain tumors or increased malignancy of brain tumors. that, despite sensational and unscientific articles in
In Sweden, Hardell et al. (2000) included aspartame (as the lay press and scientific literature, there is no well-
assessed by low-calorie drink consumption) in a case- founded proof of the carcinogenicity of aspartame.
control study of brain tumors and radiology work, med-
ical X-ray investigations, and use of cellular phones.
Evaluation of Carcinogenicity in Animals
Only one of nine comparisons appeared to indicate an
association between low-calorie drink intake and brain The three studies in rats and one study in mice done
tumors. However, not all low-calorie drinks contain as- with aspartame to evaluate carcinogenicity were evalu-
partame, and many are blends of sweeteners. Further, ated by regulatory bodies and expert scientists (JECFA,
the authors stated that the mean age of the cases and 1980; FDA, 1981; Health and Welfare Canada, 1981;
controls was 50 years, and consumption of low-calorie MAFF, 1982; Ishii, 1981; Koestner, 1984, 1997; Cornell
drinks is clearly a more common habit in young sub- et al., 1984); the conclusion is that aspartame is not
jects. The authors concluded that with such a small a carcinogen, even after long-term administration of
number of subjects, these results must be interpreted dosages hundreds to thousands of times greater than
with caution. 90th percentile consumption by humans. The highest
In the United States, Gurney et al. (1997) evalu- dose (8000 mg/kg body wt/day) used in these studies
ated aspartame consumption and the risk of childhood is the equivalent of an adult human consuming about
brain tumors in a case-control study. Case subjects were 1000 liters of beverage sweetened with 100% aspar-
19 years of age or older and were diagnosed with a tame daily over a lifetime. In addition, aspartame was
primary brain tumor between 1984 and 1991. The au- not genotoxic when evaluated in the Ames test, the
thors concluded that children with brain tumors were host-mediated assay in mice and rats, the dominant
no more likely to have consumed aspartame than were lethal mutation assay in rats, and the in vivo cytoge-
children in the control group. There was also no in- netics assay in rats (JECFA, 1980; Kotsonis and Hjelle,
creased risk from maternal consumption of aspartame 1996).
during pregnancy and no evidence of an association of As Olney had done in the late 1970s, Olney et al. in
specific types of brain tumors with aspartame. Gurney 1996 again argued that, in one study (E-33/34) when
and co-workers concluded, . . .it appears unlikely that data from independent dose groups are combined, a
any carcinogenic effect of aspartame ingestion could dose-dependent higher rate of brain tumors can be ob-
have accounted for the recent brain tumor trends as served in aspartame-fed rats compared to control rats.
Olney et al. contend (Gurney, 1997). Further, they argued that the incidence of brain tumors
From an epidemiologists standpoint, Davies et al. in control rats was unreliable in another lifetime rat
(1996) reviewed the paper by Olney et al. (1996) and study, including in utero exposure to aspartame (E-70),
concluded, This report is most disturbing for those who because the observed incidence was higher than Ol-
practice epidemiology (this study was not done by an neys expectations. The underlying basis for such claims
epidemiologist). The take-home message is completely is the incorrect assertion by this long-time aspartame
speculative (as well as volatile), but to undo the damage critic that the background incidence of brain tumors
that has already been done in the media is likely impos- in SpragueDawley (SD) rats is 0.1%; the actual back-
sible. Ross (1998) concluded, From an epidemiologic ground incidence is at least 2030 times higher (FDA,
perspective, the conclusion of the report may well rep- 1981; Koestner, 1980, 1984, 1997; Borzelleca et al.,
resent a classic example of ecologic fallacy [reference 1985; Swenberg, 1986). In actuality, the overall inci-
6 in the paper by Ross], because the Olney et al. study dence of brain tumors in the studies sponsored by G. D.
was a correlative analysis (i.e., ecologic analysis) that Searle and Co. (Searle), the developer of aspartame, was
S68 BUTCHKO ET AL.
identical and well within background incidences in both laboratory as the Searle studies, reported a sponta-
of the contested studies (FDA, 1981). neous brain tumor rate of 2.2%. Further, more detailed
Olneys assertions were evaluated in detail before the histopathologic examinations (eight brain sections per
approval of aspartame in the United States. In 1980, the animal) were done in the Searle studies than was typ-
FDA convened a panel of academic experts, the PBOI ically done (two to three brain sections per animal) at
(FDA, 1981), to evaluate several issues, including Ol- that time or even today. The greater number of brain
neys analysis of brain tumors in rats. In addition, prior sections evaluated increase the chance of finding mi-
to the PBOI, the FDA worked with the Universities croscopic tumors, and thus, the background incidence of
Associated for Research and Education in Pathology tumors would be expected to be greater than in similar
(UAREP), a consortium of nine universities that was studies where smaller numbers of brain sections are ex-
recognized for its expertise in the area of preclinical an- amined. Microscopic tumors actually comprised most of
imal testing, to conduct a detailed audit to evaluate the the brain tumors in the Searle studies (Koestner, 1980,
validity of the Searle studies and to perform an inde- 1984, 1997), leading Koestner to conclude, . . .most of
pendent reassessment of the incidence of brain tumors the tumors would have been missed with a less quali-
in these studies. UAREP and FDA concluded that the fied examination (Koestner, 1997). Based on these fac-
studies were valid for determining whether aspartame tors, FDA considered the background incidence used by
was a carcinogen (FDA, 1981; U.S. General Accounting the PBOI as unduly low and concluded that the techni-
Office, 1987). cally adequate historical control data were consistent
Complicating PBOIs deliberations was the fact that with the rates observed in the Searle studies. Subse-
the incidence of spontaneous brain tumors in Sprague quently, additional information available in the scien-
Dawley rats was not well characterized in the scien- tific literature again confirmed that spontaneous brain
tific literature at that time (FDA, 1981; Koestner, 1980, tumor rates in SpragueDawley rats are quite variable
1984, 1997). Technical differences made precise com- and range from 0 to 11% (Table 1) (Swenberg, 1986;
parisons of available studies difficult. These differences Koestner, 1997; Borzelleca et al., 1985). Today, the rate
included incomplete information on the age of the ani- of spontaneous brain tumors in SpragueDawley rats at
mals at the time of death and evaluation of only macro- 2 years of age is generally accepted to be approximately
scopic tumors in some studies while others included 3% (Koestner, 1997). Regardless of the rates of brain
histopathological examinations to identify microscopic tumors in historical controls, FDA considered that the
tumors, histopathological evaluations done only on ani- most appropriate comparison for treatment groups is
mals exhibiting neurological signs, examination of as concurrent controls within the same study. The rate of
few as one or two sections of brain, and differences brain tumors in the rat study with in utero exposure to
in calculating background incidence. As Koestner con- aspartame (E-70) is consistent with both historical and
cluded, Since brain tumors are an age-related lesion concurrent controls (FDA, 1981).
in rats, and may only be present as microtumors at an Another critical error in Olneys analysis was the in-
age of two years (the time most studies are terminated), appropriate combination of data from independent dose
both age and thoroughness of histological examination groups in an apparent attempt to show a doseresponse
significantly influence the detectable incidence of brain relationship between aspartame and brain tumors in
tumors (Koestner, 1997). the other rat study (E-33/34). The PBOI seriously con-
In its evaluation, the PBOI cited four studies with sidered Olneys post hoc analysis of this study combin-
brain tumor incidence rates ranging from 0.09, 0.6, 0.7, ing data from independent dose groups and concluded
and 3.2% but elected to use a rate of 0.7% as the his- that there did appear to be a dose-related increase in
torical basis for spontaneous brain tumors in Sprague brain tumors in Olneys analysis. However, FDA used a
Dawley rats (FDA, 1981). Thus, based on this histor- standard and accepted analysis of carcinogenicity stud-
ical brain tumor rate, the members of the PBOI were ies including appropriate statistical methods and deter-
unable to rule out a possible carcinogenic effect of as- mined there was clearly no dose-dependent increase in
partame. However, additional studies available in the brain tumors in the two Searle studies (Tables 2 and
FDA administrative record were not considered by the 3). This evaluation was done comparing the concurrent
PBOI, and other studies became available after PBOIs controls versus treatment groups separately for each
deliberations. These studies reported spontaneous inci- sex. Since the medulloblastoma was considered of em-
dences of brain tumors of up to 5% in control rats. bryonal origin and not related to treatment, it was not
FDA concluded that variability in the spontaneous included in the evaluation.
brain tumor rates would be expected among bioassays FDA further enumerated several key errors made by
done at different times, using different procedures, con- the PBOI in assessing the data from the aspartame
ducted at different places and by different people. For carcinogenicity studies (FDA, 1981; U.S. General Ac-
example, the results of a series of studies done by the counting Office, 1987). The Canadian Health Protec-
NCI using animals from the same source as those in the tion Branch concurred with FDA regarding these er-
Searle studies and housed, at least in part, in the same rors (Health and Welfare Canada, 1981). These errors
ASPARTAME: REVIEW OF SAFETY S69
TABLE 1
Variability in Brain Glioma Incidence in Control Male SpragueDawley Rats
(Three Brain Sections Unless Specified)a
TABLE 2
Brain Tumors in the 2-Year Carcinogencity Study with Aspartame
in SpragueDawley Rats (E-33/34)a,b
Dose Number Number and type of brain Number of Number and type of brain
(mg/kg body wt/day) of males tumors in males females tumors in females
0 59 1 (astrocytoma) 59 0
1000 36 1 (astrocytoma) 40 1 (astrocytoma)
1 (astrocytoma) 1 (astrocytoma)
2000 40 1 (astrocytoma) 40 0
4000 40 1 (oligodendroglioma) 40 1 (astrocytoma)
1 (astrocytoma)
1 (astrocytoma)
1 (astrocytoma)
60008000 39 0 38 1 (medulloblastoma)c
1 (astrocytoma)
a From FDA, 1981, E-33/34, E-87, and E-102 (UAREP).
b E-33/34 was originally evaluated with two sections per brain but was re-assessed with eight sections per brain (E-87) and
then by UAREP (E-102).
c Animal died during week 13 of the study (about 16 weeks of age); this medulloblastoma was considered to be embryonal
TABLE 3
Brain Tumors in the 2-Year Carcinogencity Study Including in utero Exposure
with Aspartame in SpragueDawley Rats (E-70)a,b
Dose Number Number and type of brain Number of Number and type of brain
(mg/kg body wt/day) of males tumors in males females tumors in females
0 58 1 (astrocytoma) 57 1 (astrocytoma)
1 (astrocytoma)
1 (astrocytoma)
2000 36 1 (ependymoma) 39 1 (astrocytoma)
1 (astrocytoma)
4000 40 1 (astrocytoma) 40 1 (meningioma)
a From FDA, 1981, E-70, E-87, and E-102 (UAREP).
b Eight sections per brain were evaluated.
S70 BUTCHKO ET AL.
included suggesting a doseresponse relationship based FDA, 1981). These biological criteria include the abil-
on an inappropriate combination of dose groups, misdi- ity to increase tumor incidence beyond expected control
agnosing a primary brain tumor as metastatic when no levels, the demonstration of a doseeffect relationship,
other tissues or organs in the animal had carcinoma, the occurrence of tumors at a younger age, a greater
misunderstanding the expert testimony related to one effect on embryonal and fetal cells than on adult neu-
medulloblastoma and ascribing significance to this roectodermal cells, and a shift to more anaplastic tumor
tumor, making errors in death dates of some animals, types. After thorough analysis of aspartame studies, as-
speculating on the size of tumors and their role in the partame does not meet any of these criteria (Koestner,
cause of death of animals without adequately reviewing 1980, 1984, 1997; FDA, 1981).
the study data, and misunderstanding scientific publi- Further, there is no evidence that aspartame in-
cations relating to the spontaneous incidence of brain creases the malignancy of brain tumors. The tumors
tumors in rats. For example, in E-33/34, the animal with in the aspartame studies were primarily mature, dif-
a medulloblastoma died during Week 13 of the study or ferentiated, and of glial origin. Specifically, none of the
about 16 weeks of age. Medulloblastomas are of embry- tumors were anaplastic; there was no resemblance to
onal origin and very rare in rats. Based on the size of glioblastoma multiforme in humans, as was implied by
the tumor, the tumor was considered to have started in Olney et al. (Koestner, 1997).
the embryonal stage of development. Thus, the tumor The potential mutagenicity and carcinogenicity of
was eliminated from the analysis by FDA as definitely aspartames conversion product, cycloaspartylpheny-
unrelated to aspartame. Supporting this conclusion was lalanine diketopiperazine (DKP), was also extensively
the fact that no such tumors were seen in E-70, where evaluated (see Metabolism of Aspartame section for
animals were transplacentally exposed to aspartame additional information on DKP). DKP was not genotoxic
in utero (FDA, 1981; Koestner, 1980, 1984, 1997). when evaluated in the Ames test, the host-mediated as-
At the time of approval, the FDA Commissioner say in mice and rats, the dominant lethal mutation as-
(FDA, 1981) concluded that the Searle studies (E-33/34 say in rats, and an in vivo cytogenetics assay in rats
and E-70) were negative and stated, . . .the incidence (JECFA, 1980; Kotsonis and Hjelle, 1996). Two-year
rates reported in the Searle studies fall within rea- carcinogenicity studies in rats and mice were done with
sonably expected bounds of spontaneous incidence for doses of DKP up to 3000 mg/kg body wt/day, which is
the type of rat and study size used, and that the pri- over 5000 times estimated 90th percentile daily DKP
mary evaluation of these studies should be between the intake from aspartame in the United States (Kotsonis
treated animals and their concurrent controls. In addi- and Hjelle, 1996). In the 115-week carcinogenicity study
tion, FDA concluded, . . .the data in E-33/34 do not sug- in SpragueDawley rats (E-77/78), DKP was adminis-
gest, in terms of biological significance, a doseresponse tered in the diet to provide dosages of 0, 750, 1500, and
relationship or early tumor onset. 3000 mg/kg body wt/day. Seven sections per brain were
Thus, the results of both rat studies supported the evaluated microscopically, and the incidence of brain
conclusion that aspartame is not carcinogenic. Fur- tumors in males were 2 astrocytomas/56 in controls,
ther, an additional study done in mice (E-75) demon- 1 oligodendroglioma/33 at 750 mg/kg body wt/day, and
strated that aspartame is not carcinogenic (FDA, 1981), 0/32 and 0/31 at 1500 and 3000 mg/kg body wt/day, re-
and subsequently, a third, 2-year study in Wistar rats spectively. In females, the incidence was 0/67, 0/35, and
further confirmed that aspartame is not carcinogenic 0/34 at 0, 750, and 1500 mg/kg body wt/day, respectively,
(Table 4) (Ishii, 1981). and 1 astrocytoma and 1 oligodendroglioma/33 animals
Neurocarcinogens are known to exhibit a number at 3000 mg/kg body wt/day. There was no effect of DKP
of characteristic features (Koestner, 1980, 1984, 1997; on brain tumors or any other evidence of carcinogenicity
TABLE 4
Brain Tumors in the 2-Year Carcinogenicity Study with Aspartame
and Diketopiperazine (DKP) in Wistar Ratsa,b
Dose Number Number and type of brain Number of Number and type of brain
(mg/kg body wt/day) of males tumors in males females tumors in females
(FDA, 1981; JECFA, 1980). There were also no carcino- in vivo conditions, the mutagenicity of these compounds
genic effects in the mouse study (E-76) (Kotsonis and was markedly reduced (Shephard et al., 1993). Fur-
Hjelle, 1996; FDA, 1981). In addition, another study ther, the chemical conditions of the reactions required
(Ishii, 1981) evaluated aspartame with DKP (3:1) in the to produce these products are not unique to aspartame
diet of Wistar rats for 2 years. This study also demon- and are not relevant to aspartame consumption. The
strated no evidence of carcinogenicity of DKP. nitrite concentrations used were at least 10,000 times
In granting the approval of aspartame for its intended greater than actual human gastric levels (Challis et al.,
use in 1981, the FDA Commissioner stated, For all the 1982). As a result, the rate of the reaction reported by
reasons stated above, I conclude that the available data, Shephard et al. (1993) was 100 million times greater
taken as a whole, establish that there is a reasonable than that possible in the human stomach (the reaction
certainty that aspartame and DKP do not cause brain rate is dependent on the square of the nitrite concentra-
tumors in laboratory rats. This conclusion is based on tion) (Flamm, 1997). In fact, Meier et al. (1990), from
studies E-33/34, E-70, and E-77/78, all of which were the same laboratory, concluded that the formation of
considered at the hearing [PBOI]. Additional support nitrosoaspartame under standard conditions of use of
for this conclusion is found in the Japanese study [Ishii], aspartame as a sweetener would be negligible in com-
submitted by Searle after the Board issued its decision. parison with the dietary intake of preformed nitroso
Accordingly, under the acts general safety clause, I find compounds. Furthermore, any possible reaction prod-
that the available data establish the safety of aspar- uct(s) of aspartame would be the same as those formed
tame, in terms of brain tumors, for its proposed use. by common dietary proteins, peptides and/or amino
The Health Protection Branch in Canada (Health and acids, which are present in the diet at far higher con-
Welfare Canada, 1981) also evaluated Olneys asser- centrations than aspartame. As a practical matter, the
tions and came to the same conclusion as FDA, outlining studies show that nitrosation of aspartame in the hu-
a number of errors in the analysis and PBOIs assess- man stomach is negligible and will have no effect on
ment. The Canadian government concluded that there the total body burden of nitrosamines whether they be
is no association between aspartame ingestion and the preformed or formed in the stomach (Flamm, 1997).
incidence of brain neoplasms in the rat.
After approval of aspartame by FDA, the members
Recent Evaluations by Regulatory and
of the PBOI became aware of the additional data and
Government Agencies
FDAs analysis. In a letter dated August 6, 1981 (Nauta,
1981), the Chairman of the PBOI wrote to the FDA The allegations by Olney et al. (1996) that aspartame
Commissioner stating, . . . had we known earlier about may be associated with brain tumors have been evalu-
the reassuring outcome of the recent Japanese [Ishii] ated by scientists at regulatory and government agen-
oncogenicity studies, our recommendation would doubt- cies in several countries. The U.S. National Cancer In-
less have been for unqualified approval . . . we wish to stitute (1997) concluded, a recent analysis of the NCI
express our endorsement of your final decision in this statistics on cancer incidence in the United States does
matter. not support an association between the use of aspar-
tame and an increased incidence of brain tumors. The
U.S. FDA concluded that the analysis does not support
Purported Mechanism
an association between the use of aspartame and in-
A series of experiments done prior to the approval creased incidence of brain tumors (FDA, 1996). The
of aspartame showed no direct evidence of aspartame Committee on Carcinogenicity at the Department of
nitrosation under various harsh chemical conditions Health in the United Kingdom (Department of Health,
(FDA, 1983, 1984). In addition, no N-nitroso compounds 1998; Food Standards Agency, 1999; Caseley and Dixon,
formed in vitro or in vivo in several special studies 2001) arrived at the same conclusion as FDA stat-
done to evaluate the possible nitrosation of DKP (FDA, ing, The Committee concluded that the data published
1983, 1984; JECFA, 1980). Nonetheless, Olney et al. by Olney et al. did not raise any concerns with re-
(1996) claimed that nitrosation studies by Shephard gard to the use of aspartame in the United Kingdom
and co-workers (1993) provide a mechanistic explana- (Department of Health, 1998). The Scientific Commit-
tion for how aspartame could act as a carcinogen (for tee for Food of the European Union concluded, . . .the
additional perspective, see Shephard and Lutz, 1989; data do not support the proposed biphasic increase
Shephard et al., 1991). Olney et al. (1996) cited a re- in the incidence of brain tumors in the USA during
port by Shephard et al. (1993) in which extremely high the 1980s (European Commission, 1997). From their
nitrite concentrations reacted with a variety of amino analysis, the Australia/New Zealand Food Authority
acids or peptides, including aspartame, apparently gen- (ANZFA) concluded, From the extensive scientific data
erated compounds with mutagenic properties when in- available at this stage, the evidence does not support
cubated with bacteria alone. However, when these com- that aspartame is carcinogenic in either animals or hu-
pounds were incubated with liver enzymes to mimic mans. There appears to be no foundation to recent USA
S72 BUTCHKO ET AL.
reports of increased brain tumors in humans (ANZFA, Davies, S. M., Ross, J. A., and Woods, W. G. (Eds.) (1996). Aspartame
1997). and Brain Tumors: Junk Food Science, Causes of Childhood Cancer
Newsletter, 7. University of Minnesota.
Davis, D. L., Ahlbom, A., Hoel, D., and Percy, C. (1991). Is brain cancer
Conclusions mortality increasing in industrial countries? Am. J. Ind. Med. 19,
421431.
In the late 1970s, Olney raised questions regarding
Department of Health (1998). 1996 Annual Report of the Commit-
the results of the carcinogenicity studies with aspar- tees on Toxicity, Mutagenicity, and Carcinogenicity of Chemicals in
tame, suggesting that the data showed an increase in Food, Consumer Products and the Environment, pp. 5657. Depart-
brain tumors in rats. However, upon review of the data, ment of Health, United Kingdom.
scientists in regulatory agencies around the world dis- European Commission (1997). Scientific Committee for Food: Minutes
agreed with Olneys contentions and approved aspar- of the 107th meeting of the Scientific Committee for Food, June 12
13, 1997, pp. 910.
tame. In 1996, Olney et al. again raised this issue, this
Flamm, W. G. (1997). Reply to Olney et al. Increasing brain tumor
time asserting that an apparent increased incidence of
rates: Is there a link to aspartame? (letter). J. Neuropathol. Exp.
human brain tumors in the United States may be asso- Neurol. 56, 105106.
ciated with aspartame. Specifically, Olney et al. claimed Food and Drug Administration (1981). Aspartame: Commissioners
that the rate of brain tumors increased in the United final decision. Fed. Regist. 46, 3828538308.
States concurrent with aspartame approval, that brain Food and Drug Administration (FDA) (1983). Food additives permit-
tumors increased in rats fed aspartame, and that nitro- ted for direct addition to food for human consumption; Aspartame.
sation of aspartame is a putative mechanism whereby Fed. Regist. 48, 3137631382.
aspartame may cause brain tumors. Olney and col- Food and Drug Administration (FDA) (1984). Food additives permit-
ted for direct addition to food for human consumption; aspartame.
leagues claimed a surge in brain tumors in the mid Fed. Regist. 49, 66726682.
1980s based on their analysis of selected data from Food and Drug Administration (FDA) (1992). L. Tollefson letter
the U.S. Surveillance, Epidemiology and End Results (March 20, 1992) to H. J. Roberts regarding his article, Does As-
(SEER) tumor data base from the U.S. National Cancer partame Cause Human Brain Cancer?
Institute (NCI). Food and Drug Administration (1996). FDA Statement on Aspartame,
The NCI, which maintains the SEER database, con- FDA Talk Paper, November 18, 1996.
cluded that Olneys claims were not valid, as did sci- Food Standards Agency, Additives and Novel Foods Division
entific and regulatory bodies in the European Union, (1999). Aspartame. JFSSG. http://www.foodstandards.gov.uk/maff/
archive/food/asp9911.htm.
the United States, and other countries. The scientific
Greig, N. H., Ries, L. G., Yancik, R., and Rapoport, S. I. (1990). In-
consensus is that there is no association between as- creasing annual incidence of primary malignant brain tumors in
partame consumption and an increase in human brain elderly. J. Natl. Cancer Inst. 82, 16211624.
tumor rates, aspartame is not an animal carcinogen, Gurney, J. G., Pogoda, J. M., Holly, E. A., Hecht, S. S., and Preston-
and the speculated biochemical mechanism for car- Martin, S. (1997). Aspartame consumption in relation to childhood
cinogenicity is not relevant to aspartame consumption. brain tumor risk: Results from a case-control study. J. Natl. Cancer
Thus, the evidence is clear that aspartame is not asso- Inst. 89, 10721074.
ciated with brain tumors. Hardell, L., Nasman, A., Pahlson, A., and Hallquist, A. (2000). Case-
control study on radiology work, medical x-ray investigations, and
use of cellular telephones as risk factors for brain tumors. Medscape
Gen. Med. May 4, 2000.
REFERENCES Health and Welfare Canada (1981). Aspartame, Information Letter
No. 602. Health Protection Branch, Ottawa.
Australia New Zealand Food Authority (ANZFA) (1997). Aspartame: Ishii, H. (1981). Incidence of brain tumors in rats fed aspartame.
Information for Consumers, information paper. Toxicol. Lett. 7, 433437.
Borzelleca, J. F., Hogan, G. K., and Koestner, A. (1985). Chronic toxi- Joint FAO/WHO Expert Committee on Food Additives (JECFA)
city/carcinogenicity study of FD&C Blue No. 2 in rats. Food Chem. (1980). Aspartame: Toxicological Evaluation of Certain Food Ad-
Toxicol. 23, 551558. ditives, WHO Food Additive Series No. 15, pp. 1886. JECFA,
Boyle, P., Maisonneuve, P., Saracci, R., and Muir, C. S. (1990). Is the Geneva.
increased incidence of primary malignant brain tumors in the el- Koestner, A. (1980). Prepared Statement: Aspartame Public Board of
derly real? J. Natl. Cancer Inst. 82, 15941596. Inquiry, February 1, 1980, FDA Docket No. 75F-0355.
Butchko, H. (1992). Reply to H. J. Roberts Does aspartame cause Koestner, A. (1984). Aspartame and brain tumors: Pathology issues.
human brain cancer (letter). J. Adv. Med. 5, 257259. In Aspartame: Physiology and Biochemistry (L. D. Stegink and
Caseley, J., and Dixon, W. (2001). Aspartame: In perspective. Nutr. L. J. Filer, Jr., Eds.), pp. 447457. Marcel Dekker, New York.
Food Sci. 31, 2326. Koestner, A. (1997). Reply to Olney et al. Increasing brain tumor
Challis, B. C., Lomas, S. J., and Rzepa, H. S. (1982). A kinetic model for rates: Is there a link to aspartame? (letter). J. Neuropathol. Exp.
the formation of gastric N-nitroso compounds. Acta Cien. Compost. Neurol. 56, 107109.
19, 267282. Kotsonis, F. N., and Hjelle, J, J. (1996). The safety assessment of
Cornell, R. G., Wolfe, R. A., and Sanders, P. G. (1984). Aspartame aspartame: Scientific and regulatory considerations. In The Clin-
and brain tumors: Statistical issues. In Aspartame: Physiology and ical Evaluation of a Food Additive: Assessment of Aspartame
Biochemistry (L. D. Stegink and L. J. Filer, Jr., Eds.), pp. 459479. (C. Tschanz, H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.),
Marcel Dekker, New York. pp. 2341. CRC Press, Boca Raton, FL.
ASPARTAME: REVIEW OF SAFETY S73
La Vecchia, C., Lucchini, F., Negri, E., Boyle, P., Maisonneuve, P., and Nauta, W. J. H. (1981). Letter to FDA Commissioner Hayes, August 6,
Levi, F. (1992). Trends of cancer mortality in Europe, 19551989. 1981.
IV. Urinary tract, eye, brain and nerves, and thyroid. Eur. J. Cancer. Olney, J. W., Farber, N. B., Spitznagel, E., and Robins, L. N. (1996).
28A, 12101281. Increasing brain tumor rates: Is there a link to aspartame? J. Neu-
Legler, J. M., Gloeckler Ries, L. A., Smith, M. A., Warren, J. L., ropathol. Exp. Neurol. 55, 11151123.
Heineman, E. F., Kaplan, R. S., and Linet, M. S. (1999). Brain and Roberts, H. J. (1991). Does aspartame cause human brain cancer?
other central nervous system cancers: Recent trends in incidence J. Adv. Med. 4, 231241.
and mortality. J. Natl. Cancer Inst. 91, 13821390.
Ross, J. A. (1998). Brain tumors and artificial sweeteners? A lesson
Levy, P. S., and Hedeker, D. (1996). Reply to Olney et al. Increasing on not getting soured on epidemiology. Med. Pediatr. Oncol. 30, 78.
brain tumor rates: Is there a link to aspartame? (letter). J. Neu-
ropathol. Exp. Neurol. 55, 1280. Scientific Committee for Food (SCF) (1985). Food-Science and Tech-
niques, Reports of the Scientific Committee for Food, sixteenth se-
Linet, M. S., Ries, L. A. G., Smith, M. A., Tarone, R. E., and Devesa, ries. Commission of the European Communities, Luxembourg.
S. S. (1999). Cancer surveillance series: Recent trends in childhood
cancer incidence and mortality in the United States. J. Natl. Cancer Shephard, S. E., and Lutz, W. K. (1989). Nitrosation of dietary pre-
Inst. 91, 10511058. cursors. Cancer Surv. 8, 401421.
Marshall, E. (1990). Experts clash over cancer data. Science 250, 900 Shephard, S. E., Meier, I., and Lutz, W. (1991). Alkylating potency
902. of nitrosated amino acids and peptides. In Relevance to Human
Cancer of n-Nitroso Compounds, Tobacco Smoke, and Mycotoxins
Meier, I., Shephard, S. E., and Lutz, W. K. (1990). Nitrosation of
(I. K. ONeill, J. Chen, and H. Bartsch, Eds.), pp. 383387. IARC,
aspartic acid, aspartame, and glycine ethylester: Alkylation of
Lyon, France.
4-( p-nitrobenzyl)pyridine (NBP) in vitro and binding to DNA in
the rat. Mutat. Res. 238, 193201. Shephard, S. E., Wakabayashi, K., and Nagao, M. (1993). Mutagenic
activity of peptides and the artificial sweetener aspartame after
Ministry of Agriculture, Fisheries, and Food (MAFF) (1982). Food
nitrosation. Food Chem. Toxicol. 31, 323329.
Additives and Contaminants Committee Report on the Review
of Sweeteners in Food, FAC/REP/34:HMSO 1982. Committee on Smith, M. A., Freidlin, B., Gloeckler Ries, L. A., and Simon, R. (1998).
Toxicity of Chemicals in Food, Consumer Products and the Envi- Trends in reported incidence of primary malignant brain tumors in
ronment. United Kingdom. children in the United States. J. Natl. Cancer Inst. 90, 12691277.
Modan, B., Wagener, D. K., Feldman, J. J., Rosenberg, H. M., and Swenberg, J. A. (1986). Brain tumorsProblems and perspectives.
Feinleib, M. (1992). Increased mortality from brain tumors: A com- Food Chem. Toxicol. 24, 155158.
bined outcome of diagnostic technology and change of attitude to- U.S. General Accounting Office (1987). Food Additive Approval Pro-
ward the elderly. Am. J. Epidemiol. 135, 13491357. cess Followed for Aspartame. Food and Drug Administration.
Muir, C. S., Storm, H. H., and Polednak, A. (1994). Brain and other Weihrauch, M. R., Diehl, V., and Bohlen, H. (2001). Artificial
nervous system tumours. Cancer Surv. 19/20, 36992. sweetenersIs there a carcinogenic risk? Medizinische Klinik. 96,
National Cancer Institute (NCI) (1997). Cancer Facts. Artificial 670675.
Sweeteners, http://cis.nci.nih.gov/fact/pdfdraft/3 risk/fs3 19.pdf. Werner, M. H., Phuphanich, S., and Lyman, G. H. (1995). The increas-
National Cancer Institute (NCI) (1999). SEER Cancer Statistics Re- ing incidence of malignant gliomas and primary central nervous
view 19731996. system lymphoma in the elderly. Cancer 76, 16341642.
Use of Aspartame by Potentially Sensitive Populations
Heterozygotes for Phenylketonuria body wt dose of aspartame is equivalent to a 70 kg
adult consuming approximately 14 liters of beverage
Individuals with the rare genetic disease phenylke-
sweetened with aspartame in a single sitting.
tonuria (PKU) (OMIM 261600) have a severely com-
From the results of the bolus dosing studies, PKU het-
promised ability to metabolize dietary phenylalanine
erozygotes clear phenylalanine at about half the rate
and, consequently, have extreme elevations in plasma
of normal individuals; however, even high intakes of
phenylalanine concentrations. If this disorder is not de-
phenylalanine from aspartame are adequately metab-
tected and dietary restriction of phenylalanine is not
olized by PKUH (Stegink et al., 1979, 1980, 1987). For
begun early in life, a variety of severe developmental
example, a bolus dose of 34 mg/kg body wt aspartame
and neurological sequelae will occur. (For a comprehen-
(Stegink et al., 1979) resulted in mean peak plasma
sive review of PKU, see Scriver and Kaufman, 2001).
phenylalanine concentrations within the normal post-
Individuals heterozygous for PKU (PKUH) are carriers
prandial range. Even after 100 mg/kg body wt aspar-
for this disease and have a somewhat diminished abil-
tame (Stegink et al., 1980), high mean plasma pheny-
ity to metabolize dietary phenylalanine, which results
lalanine concentrations in PKUH (41.7 2.33 mol/dl)
in marginally higher than normal plasma phenylala-
were well below those associated with PKU.
nine concentrations. However, such individuals do not
In multiple dose studies in PKUH (Stegink et al.,
need restriction of dietary phenylalanine and have none
1989, 1990), mean peak plasma phenylalanine con-
of the neurologic sequelae associated with untreated
centrations increased significantly over baseline af-
PKU. Thus, they go undetected until two PKU heterozy-
ter aspartame but approximated the normal postpran-
gous parents reproduce a homozygous PKU offspring.
dial range. The results of these studies demonstrate
Because of the phenylalanine component of aspartame,
that a PKUH could ingest large amounts of aspartame
concerns were raised regarding whether there was po-
at hourly intervals around the clock without increas-
tential vulnerability of PKUH to the alleged effects of
ing plasma phenylalanine concentrations appreciably
aspartame on brain function.
above the postprandial range. Finally, as in healthy
subjects, dietary protein modulates the rise in plasma
Tolerance of aspartame in PKUH. Before approval
phenylalanine concentrations after aspartame inges-
of aspartame, short-term and long-term tolerance of
tion in PKUH (Stegink et al., 1991; Curtius et al., 1994;
aspartame in PKUH were evaluated in a random-
Stegink and Filer, 1996) compared to concentrations
ized, double-blinded manner (Koch et al., 1976). During
that would be expected from the dose of aspartame ad-
the 6-week phase, subjects received increasing (in-
ministered alone.
creased weekly) doses of aspartame (6008100 mg/day);
in the subsequent 21-week phase, subjects received Brain function in PKUH. In two studies from one
1800 mg aspartame/day. During the studies, subjects laboratory (Elsas and Trotter, 1988; Epstein et al.,
were evaluated via physical examinations, including 1989), a small number of PKUH were exposed to
ophthalmologic examinations, plasma phenylalanine 100 mg/kg body wt phenylalanine daily in addition to
and tyrosine concentrations, hematology evaluations, the normal dietary intake of 60 mg/kg body wt. There
biochemical evaluations including tests of liver func- was a suggestion that, when exposed to these large
tion and lipids, thyroid function, glucose and insulin doses of phenylalanine, subjects had a decrease in mean
concentrations, and urine chromatographies. No med- power frequency on their EEGs. However, it is not clear
ical or biochemical changes were found after aspar- that this observation, even if causally related and re-
tame administration compared to placebo during the producible, has any clinical significance.
27 weeks of the study; in addition, plasma phenylala- Thus, a randomized, double-blind, placebo-
nine and tyrosine concentrations remained within the controlled, crossover study (Trefz et al., 1994) was
normal range. The authors concluded that aspartame done in a larger number of PKUH subjects (N = 48)
was well tolerated by PKUH. for 12 weeks duration each with both aspartame and
placebo treatments. During the aspartame treatment
Pharmacokinetics of phenylalanine from aspartame arm, subjects were given either 15 or 45 mg/kg body
in PKUH. The metabolic effects of bolus doses of as- wt/day of aspartame. These doses of aspartame are 5
partame (4, 10, 34, and 100 mg/kg body wt) (Stegink 15 times 90th percentile intake of aspartame (Butchko
et al., 1979, 1980, 1987), repeated doses of aspartame and Kotsonis, 1991). Subjects were tested with
(Stegink et al., 1989, 1990), and aspartame given with a the Amsterdam Neuropsychological Tasks battery (de
protein meal (Stegink et al., 1991; Curtius et al., 1994; Sonneville, 1999), including tests for visual information
Stegink and Filer, 1996) have been extensively eval- processing and executive function of varying complex-
uated in PKUH. (see Metabolism of Aspartame sec- ity. This computerized test battery has demonstrated
tion for further details.) For perspective, the 100 mg/kg changes in cognitive performance to be associated with
S74
ASPARTAME: REVIEW OF SAFETY S75
changes of plasma phenylalanine concentrations in bolus doses of 600 and 1200 mg aspartame and placebo
individuals with PKU (Schmidt et al., 1994), even in in levodopa-treated Parkinsons patients with protein-
a very well-controlled population (Huijbregts et al., sensitive motor fluctuations (Karstaedt and Pincus,
2002). In addition, to evaluate the earlier suggestions 1993). Although there were statistically significant in-
that aspartame may affect the mean power frequency creases in plasma phenylalanine concentrations after
of EEGs, conventional analysis of EEGs as well as the 1200-mg dose of aspartame, there were no statis-
analysis of the spectra by fast Fourier transformation, tically significant effects of aspartame compared with
a more sophisticated analysis than used in the earlier placebo on motor performance.
studies, were done. There were no statistically signifi- In one case study where the lead investigator was
cant differences found on routine EEG analysis, EEG also the single subject, he reported shaking of his arms
spectral analysis, or neuropsychological tests between and legs and various other vague symptoms continu-
either of the aspartame treatments and placebo. In ad- ing for several weeks after a 300-mg aspartame dosage
dition, there was no difference in adverse experiences (Gerrard et al., 1994). However, such symptoms have
after aspartame compared to placebo. not been reported elsewhere or in controlled studies.
Depressed Individuals
Individuals with Dizziness
Walton et al. (1993) reported an increase in the occur-
Review of the consumer anecdotal reports of symp-
rence of adverse experiences after aspartame vs placebo
toms possibly associated with aspartame by the CDC
in a population, which they felt may be potentially vul-
(1984) revealed that dizziness was a frequently reported
nerable to aspartame consumption, i.e., people with
symptom. Gulya and co-workers (1992) asked 53 pa-
monopolar depression. The study design was a double-
tients who visited their clinic complaining of dizziness
blind, placebo-controlled, crossover design where doses
to complete a questionnaire regarding dietary habits
of aspartame of 30 mg/kg body wt/day or placebo were
thought to influence dizziness. Only seven patients, six
given for 7 consecutive days on each treatment. The
of whom consumed aspartame, completed the question-
authors aborted the study after two serious ophthal-
naire. Subjects were then asked to cease aspartame in-
mologic complications were observed. One of these oc-
take for 1 month and keep a log of their episodes of
curred in the placebo arm and one in the aspartame
dizziness. Only one subject appeared to respond to ces-
arm. The authors reported significant differences be-
sation of aspartame intake; however, he did not partic-
tween aspartame and placebo in reported adverse ex-
ipate in the crossover phase where aspartame would
periences. However, the reliability of this study has
have been allowed. In other clinical studies with as-
been questioned (Schomer et al., 1996; Butchko, 1994).
partame, including long-term studies, there was no as-
Since the study was terminated after only 11 of the
sociation of aspartame with dizziness (Tschanz et al.,
proposed 40 subjects completed their participation, too
1996).
few subjects were evaluated to obtain any statistically
valid conclusions. Further, in order to attain sufficient
numbers and a statistically significant result with ad- Individuals with Renal Disease
verse experiences, complaints were consolidated across
Individuals with renal disease are an especially im-
a range of unrelated categories, and subjects who had
portant population as they can have altered amino acid
been dropped for adverse experiences were reentered
profiles and, as many of them are diabetics, would also
for the analyses of complaints. The methodological er-
be likely to consume aspartame, which provides two
rors present in the report and the apparent post hoc
amino acids, aspartic acid and phenylalanine. Gupta
analysis are so serious that it is impossible to draw any
et al. (1989) studied 23 diabetic patients with renal fail-
conclusions from this study.
ure, who were on maintenance hemodialysis, to evalu-
ate the effect of aspartame (10 mg/kg body wt), placebo,
Movement and Motor Disorders
and the postprandial condition on plasma amino acid
In patients with Parkinsons disease with fluctuat- profiles in a randomized, double-blind, crossover study.
ing symptoms, there is an inverse relationship between As expected, plasma phenylalanine concentrations at
plasma LNAA concentrations and motor performance 1 and 2 h after aspartame were significantly higher
as LNAAs compete with levodopa for entry into the than after placebo but were within the normal post-
brain. Thus, some Parkinsons patients achieve a bet- prandial range in these subjects. Plasma concentrations
ter therapeutic response to levodopa if they also are of tyrosine were also higher after aspartame vs placebo
treated with a protein-restricted diet. Because aspar- but were below the normal postprandial range in these
tame provides one of the LNAAs, phenylalanine, it was subjects. There were no adverse experiences associated
suggested that aspartame may adversely affect motor with aspartame. These authors concluded that aspar-
performance in such patients. This was evaluated in a tame is safe for diabetic subjects with chronic renal
double-blind, placebo-controlled, crossover study with failure.
S76 BUTCHKO ET AL.
Individuals with Liver Disease to be safe for individuals with renal disease and indi-
viduals with liver disease.
Phenylalanine has been believed to cause or affect
hepatic encephalopathy in individuals with liver dis-
ease. Consequently, there was concern that such indi- REFERENCES
viduals may be affected by aspartame because of its
amino acid constituents and also that such individu- Butchko, H. H. (1994). Reply to Walton et al. Adverse reactions to
als may not properly metabolize its methanol compo- aspartame: double-blind challenge in patients from a vulnerable
nent. The clinical and biochemical effects of aspartame population (letter). Biol. Psychiatry 36, 206207.
(15 mg/kg body wt), skim (no-fat) milk (10 oz. contain- Butchko, H. H., and Kotsonis, F. N. (1991). Acceptable daily intake
ing phenylalanine in an equimolar amount with aspar- vs actual intake: The aspartame example. J. Am. Coll. Nutr. 10,
258266.
tame dose), and placebo were evaluated in individuals
Centers for Disease Control (CDC) (1984). Evaluation of Consumer
with chronic, alcoholic liver disease in a randomized, Complaints Related to Aspartame Use. CDC, Atlanta, GA.
crossover study (Hertelendy et al., 1993). As expected, Curtius, H., Endres, W., and Blau, N. (1994). Effect of high-protein
aspartame resulted in a significantly greater increase meal plus aspartame ingestion on plasma phenylalanine concen-
in plasma phenylalanine concentrations compared to trations in obligate heterozygotes for phenylketonuria. Metabolism
skim (no-fat) milk and placebo (Cmax , 14.55 7.38, 43, 413416.
10.95 4.95, 8.84 4.55 mol/dl, respectively). There de Sonneville, L. M. J. (1999). Amsterdam neuropsychological tasks:
A computer-aided assessment program. In Cognitive Ergonomics,
were no changes from baseline in plasma or blood as- Clinical Assessment and Computer-Assisted Learning: Computers
partic acid, methanol, formate, or ammonia concentra- in Psychology (B. P. L. M Den Brinker, P. J. Beek, A. N. Brand, S. J.
tions after any of the three treatments. Intensity of Maarse, and L. J. M. Mulder, Eds.), Vol. 6, pp. 187203. Swets &
encephalopathy was evaluated by the portal systemic Zeitlinger, Lisse, The Netherlands.
index (PSE), comprised of assessments of mental sta- Elsas, L. J., and Trotter, J. F. (1988). Changes in physiological con-
tus, asterixis, Reitan-A trail test, blood ammonia con- centrations of blood phenylalanine produce changes in sensitive
parameters of human brain function. In Dietary Phenylalanine
centrations, and EEG activity. There was significant and Brain Function (R. J. Wurtman and E. Ritter-Walker, Eds.),
worsening in PSE, primarily in mental status assess- pp. 187195. Birkhauser, Boston.
ment only after the skim (no-fat) milk treatment. The Epstein, C. M., Trotter, J. F., Averbook, A., Freeman, S., Kutner, M. H.,
authors concluded that aspartame may be used safely and Elsas, L. J. (1989). EEG mean frequencies are sensitive indices
by individuals with chronic, stable liver disease. of phenylalanine effects on normal brain. Electroencephalogr. Clin.
Neurophysiol. 72, 133139.
Conclusion Gerrard, J. W., Richardson, J. S., and Donat, J. (1994). Neuropharma-
cological evaluation of movement disorders that are adverse reac-
Several subpopulations of individuals, including tions to specific foods. Int. J. Neurosci. 76, 6169.
PKUH, individuals with depression, individuals with Gulya, A. J., Sessions, R. B., and Troost, T. R. (1992). Aspartame
dizziness, individuals with Parkinsons disease, indi- and dizziness: preliminary results of a prospective, nonblinded,
prevalence and attempted cross-over study. Am. J. Otol. 13, 438
viduals with renal disease, and individuals with liver 442.
disease, who may be potentially sensitive to aspartame Gupta, V., Cochran, C., Parker, T. F., Long, D. L., Ashby, J., Gorman,
because of its amino acid content, have been studied. M. A., and Liepa, G. U. (1989). Effect of aspartame on plasma amino
Extensive metabolism studies in PKUH using various acid profiles of diabetic patients with chronic renal failure. Am. J.
paradigms have demonstrated that aspartame, even in Clin. Nutr. 49, 13021306.
amounts well above average 90th percentile consump- Hertelendy, Z. I., Mendenhall, C. I., Rouster, S. D., Marshall, L., and
Weesner, R. (1993). Biochemical and clinical effects of aspartame
tion levels, results in plasma phenylalanine concentra-
in patients with chronic, stable alcoholic liver disease. Am. J. Gas-
tions that are considered safe. In addition, tolerance troenterol. 88, 737743.
studies and detailed evaluation of brain function in this Huijbregts, S. C. J., de Sonneville, L. M. J., Licht, R., van Spronsen,
population have shown no adverse effects of aspartame. F. J., Verkerk, P. H., and Sergeant, J. A. (2002). Sustained atten-
A prematurely terminated study in depressed patients tion and inhibition of cognitive interference in treated phenylke-
appeared to indicate that aspartame was associated tonuria: Associations with concurrent and lifetime phenylalanine
concentrations, Neuropsychologia 40, 715.
with more adverse experiences than placebo; however,
Karstaedt, P. J., and Pincus, J. H. (1993). Aspartame use in
methodological flaws and the statistical analysis used Parkinsons disease. Neurology 43, 611613.
raised serious questions regarding the validity of the Koch, R., Shaw, K. N. F., Williamson, M., and Haber, M. (1976). Use
study. In a study in patients with dizziness, only a few of aspartame in phenylketonuric heterozygous adults. J. Toxicol.
subjects completed the questionnaire, and only in one Environ. Health 2, 453457.
subject, who did not complete both treatment arms of Schmidt, E., Rupp, A., Burgard, P., Pietz, J., Weglage, J., and
the study, was there a suggestion that aspartame may De Sonneville, L. (1994). Sustained attention in adult phenylke-
have been associated with his dizziness. No effects of tonuria: The influence of the concurrent phenylalanine blood level.
J. Clin. Exp. Neuropsychol. 16, 681688.
aspartame were seen in individuals with Parkinsons
Schomer, D. L., Spiers, P., and Sabounjian, L. A. (1996). Evaluation
disease, whose symptoms and responsiveness to lev- of behavior, cognition, mood, and electroencephalograms in normal
odopa are sensitive to fluctuations in plasma amino acid adults and potentially vulnerable populations. In The Clinical Eval-
concentrations. Finally, aspartame was demonstrated uation of a Food Additive: Assessment of Aspartame (C. Tschanz,
ASPARTAME: REVIEW OF SAFETY S77
H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 217233. Stegink, L. D., Filer, L. J., Jr., Baker, G. L., Bell, E. F., Ziegler, E. E.,
CRC Press, Boca Raton, FL. Brummel, M. C., and Krause, W. L. (1989). Repeated ingestion
Scriver, C. R., and Kaufman, S. (2001). Hyperphenylalaninemia: of aspartame-sweetened beverage: effect on plasma amino acid
Phenylalanine hydroxylase. In The Metabolic and Molecular Bases concentrations in individuals heterozygous for phenylketonuria.
of Inherited Disease (C. R. Scriver, A. Beaudet, W. S. Sly, and Metabolism 38, 7884.
D. Valle, Eds.), eighth ed., Vol. II, pp. 6671724. McGrawHill, New Stegink, L. D., Filer, L. J., Jr., Bell, E. F., Ziegler, E. E.,
York. Tephly, T. R., and Krause, W. L. (1990). Repeated ingestion of
Stegink, L. D., and Filer, L. J., Jr. (1996). Effects of aspartame in- aspartame-sweetened beverages: Further observations in indi-
gestion of plasma aspartate, phenylalanine, and methanol concen- viduals heterozygous for phenylketonuria. Metabolism 39, 1076
trations in potentially sensitive populations. In The Clinical Eval- 1081.
uation of A Food Additive: Assessment of Aspartame (C. Tschanz, Stegink, L. D., Filer, L. J., Jr., Brummel, M. C., Baker, G. L., Krause,
H. H. Butchko, W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 87113. W. L., Bell, E. F., and Ziegler, E. E. (1991). Plasma amino acid
CRC Press, Boca Raton, FL. concentrations and amino acid ratios in normal adults and adults
Stegink, L. D., Filer, L. J., Jr., Baker, G. L., and McDonnell, J. E. (1979). heterozygous for phenylketonuria ingesting a hamburger and milk
Effect of aspartame loading upon plasma and erythrocyte amino shake meal. Am. J. Clin. Nutr. 53, 670675.
acid levels in phenylketonuric heterozygotes and normal adult sub- Trefz, F., de Sonneville, L., Matthis, P., Benninger, C., Lanz-Englert,
jects. J. Nutr. 109, 708717. B., and Bickel, H. (1994). Neuropsychological and biochemical in-
Stegink, L. D., Filer, L. J., Jr., Baker, G. L., and McDonnell, J. E. (1980). vestigations in heterozygotes for phenylketonuria during ingestion
Effect of an abuse dose of aspartame upon plasma and erythrocyte of high dose aspartame (a sweetener containing phenylalanine).
levels of amino acids in phenylketonuric heterozygous and normal Hum. Genet. 93, 369374.
adults. J. Nutr. 110, 22162224. Tschanz, C., Butchko, H. H., Stargel, W. W., and Kotsonis, F. N., Eds.
Stegink, L. D., Wolf-Novak, L. C., Filer, L. J., Jr., Bell, E. F., Ziegler, (1996). The Clinical Evaluation of A Food Additive: Assessment of
E. E., Krause, W. L., and Brummel, M. C. (1987). Aspartame- Aspartame. CRC Press, Boca Raton, FL.
sweetened beverage: Effect on plasma amino acid concentrations Walton, R. G., Hudak, R., and Green-Waite, R. J. (1993). Adverse
in normal adults and adults heterozygous for phenylketonuria. reactions to aspartame: Double-blind challenge in patients from a
J. Nutr. 117, 19891995. vulnerable population. Biol. Psychiatry 34, 1317.
Endocrine Evaluations with Aspartame
Introduction cant effects of either aspartame or saccharin on plasma
glucose concentrations compared to unsweetened be-
Because individuals with diabetes mellitus would
verage at any time point in either the normal or dia-
likely be enthusiastic consumers of aspartame, there
betic subjects. Specifically, there were no differences
have been a number of studies done to evaluate aspar-
between the three treatments in glucose area under
tames safety in this subpopulation. Studies included
the curve (AUC), Cmax , or tmax in either normal or
both acute and long-term dosing paradigms, with spe-
diabetic subjects (Tables 1 and 2). In both normal and
cial emphasis on potential effects on plasma glucose
diabetic subjects, however, there was a slight decline
and insulin concentrations. In addition, the effect of
in glucose concentrations during the 180-min study
aspartame on plasma glucose and insulin concentra-
period, which was similar among the treatment groups.
tions has been evaluated in nondiabetic subjects in
In normal subjects, the AUC for insulin was greater
some studies. It had been speculated (Rogers et al.,
after aspartame than after saccharin and unsweetened
1988) that sweetness, even without the calories of sugar,
beverage ingestion. However, there were no differences
may induce a cephalic-phase insulin release impacting
among treatments in peak insulin concentrations or the
blood glucose concentrations which may possibly be as-
time to reach the peak. The observed differences were
sociated with increased hunger. Thus, several studies
small and likely related to a decrease from baseline
concentrated on this issue. Because certain amino acids
insulin concentrations after both the unsweetened and
may affect the release of other hormones (e.g., prolactin,
the saccharin beverage and no change from baseline
cortisol, or growth hormone), studies have been under-
after aspartame beverage. In diabetic subjects, there
taken to determine if large doses of aspartame affect
were no differences in insulin concentrations as far as
hormone release.
peak concentrations, time to peak, and AUC. Glucagon
concentrations were variable, but there was no change
Acute Dosing Studies with Aspartame in Normal in the overall magnitude of secretion (Horwitz et al.,
and Diabetic Subjects 1988).
The effect of aspartame on plasma glucose and in-
Several studies have been completed to evaluate the
sulin concentrations was also evaluated in several
effect of acute doses of aspartame in diabetic subjects.
metabolism studies using doses of aspartame ranging
Shigeta et al. (1985) evaluated the effect of 225 mg as-
from approximately 4 to 40 mg/kg body wt (Wolf-Novak
partame or 75 g of oral glucose on plasma glucose and
et al., 1990; Mller, 1991; Stegink et al., 1990). Aspar-
insulin concentrations in subjects with Type 2 diabetes
tame had no effect on plasma insulin concentrations.
mellitus. Aspartame had no effect on plasma glucose
Wolf-Novak et al. (1990) and Stegink et al. (1990) also
or serum insulin concentrations. When aspartame was
reported no effect of aspartame on plasma glucose con-
substituted for sucrose for 3 days in the diets of these
centrations. Although Mller (1991) reported a signifi-
subjects, there was no effect on fasting plasma glucose
cant effect of aspartame on plasma glucose concentra-
concentrations.
tion, such an effect is not consistent with the results
Okuno et al. (1986) studied the effect of 100 g of
of the studies by Wolf-Novak et al. (1990) and Stegink
glucose or 500 mg of aspartame on blood glucose and
et al. (1990). A difference in baseline glucose concen-
insulin concentrations in normal subjects and Type 2
trations between aspartame and the water condition
diabetics. Diabetic subjects were separated into three
likely affected the results in the study by Mller (1991).
groups according to the severity of their fasting hyper-
Overduin and Jansen (1997) used aspartame as the
glycemia. After the oral glucose load, there were the
control in a study to evaluate whether conditioned in-
expected increases in blood glucose and serum insulin
sulin and glucose responses were related to binge eat-
concentrations in proportion to the severity of fasting
ing in healthy females. They used a sixfold pairing of a
hyperglycemia. However, in all groups the ingestion
of 500 mg of aspartame was associated with a decline
TABLE 1
in glycemia which was likely due to fasting, which oc-
curred without observable changes in serum insulin or Plasma Glucose Pharmacokinetics (Mean SE) in
Normal Subjects Given 400 mg Aspartame, 135 mg Sac-
plasma glucagon concentrations.
charin, or Unsweetened Beveragea
Horwitz et al. (1988) studied the effects of single doses
of aspartame (400 mg) and saccharin (135 mg) in bev- Treatment AUC (mg/dl min) Cmax (mg/dl) tmax (min)
erage vs an unsweetened beverage on hormonal and
glycemic responses in normal subjects and subjects with Aspartame 16,664 314 96 2 19 10
Type 2 diabetes. These doses of aspartame and sac- Saccharin 16,179 191 92 1 86
Unsweetened 16,213 174 92 1 19 6
charin are approximately equal to the amount of each
sweetener in 1 liter of beverage. There were no signifi- a Adapted from Horwitz et al. (1988).
S78
ASPARTAME: REVIEW OF SAFETY S79
200
mg/dl
mg/dl
150 10
150
100
100
5
50 50
0 0
0
Week 0 Week 9 Week 18 Week 0 Week 9 Week 18 Week 0 Week 9 Week 18
FIG. 1. Long-term aspartame dosing has no effect on fasting or 2-h postprandial glucose (mean SE) concentrations or on glycohemoglobin
concentrations (% SE) compared to placebo in Type 1 and Type 2 diabetic subjects. Source: Nerhling et al. (1985).
S80 BUTCHKO ET AL.
Glucose Metabolism after Exercise lease effect caused by aspartame, but this finding is not
consistent with the results of other studies discussed
Wouassi et al. (1997) compared metabolic and hor-
later (Hartel et al., 1993; Teff et al., 1995; Abdallah et al.,
monal responses during repeated bouts of brief and
1997).
intense exercise after administration of glucose or
Nguyen et al. (1998) evaluated the effect of single
control (aspartame). Venous plasma lactate increased
doses of aspartame (250 mg) and glucose (75 g) on
significantly after exercise, but no differences were ob-
calcium and oxalate metabolism in a crossover study
served between glucose and aspartame treatments. Ini-
in healthy adults. They reported that urinary calcium
tially, blood glucose concentrations decreased but then
excretion increased after both aspartame and glucose,
increased during exercise in both groups. Plasma in-
which they related to increased calcemia and decreased
sulin concentrations decreased in both groups but ex-
phosphatemia. Aspartame did not alter plasma concen-
hibited a higher peak after glucose treatment than as-
trations of glucose or insulin or urinary oxalate excre-
partame. Glucagon and epinephrine concentrations did
tion. However, these findings from a single-dose study
not change significantly in either group.
are contrary to those observed by Leon et al. (1989,
1999), who reported no differences between blood cal-
In Vitro Study with Rat Pancreatic Islet Cells cium concentrations or urinary calcium excretion after
An in vitro study by Malaisse et al. (1998) evaluated very high doses of aspartame (about 20 times the dose
the effect of different high-intensity sweeteners on in- given by Nguyen et al.) given daily for 6 months com-
sulin release from rat pancreatic islet cells. They found pared to placebo or baseline. The dose of aspartame
no effect of aspartame but did observe increased insulin administered in the study by Leon et al. (75 mg/kg
release with other high-intensity sweeteners that are body wt/day) is the equivalent of a 70-kg individual
known to have bitter components (i.e., saccharin, cycla- consuming the amount of aspartame in over 10 liters
mate, stevioside, and acesulfame-k). of beverage sweetened with 100% aspartame daily for
6 months.
Other Endocrine Evaluations
Cephalic-Phase Insulin Release
Carlson and Shah (1989) evaluated the effect of as-
partame (534 mg) and its constituent amino acids as- It has been suggested that sweetness may induce
partic acid (242 mg) and phenylalanine (300 mg and cephalic-phase insulin release (CPIR), resulting in low-
1000 mg) on serum concentrations of prolactin, corti- ering of blood glucose concentrations with consequent
sol, growth hormone, and insulin, and blood glucose con- increases in hunger (Rogers et al., 1988). To investi-
centrations in healthy individuals. Aspartame (in doses
gate this, Hartel et al. (1993) evaluated five different
equivalent to that present in about 1 liter of beverage), test solutions (containing aspartame, acesulfame-k, cy-
aspartic acid, or phenylalanine (in equimolar amounts) clamate, saccharin, or sucrose) and water in a multiple
did not change the magnitude and pattern of secre- crossover study. Blood was drawn at 0, 5, 10, 15, 30, 60,
tion of prolactin, cortisol, or growth hormone. The high and 120 min after dosing for plasma insulin and blood
dose of phenylalanine (1000 mg/kg body wt) did mod- glucose determinations. As expected, sucrose resulted
estly increase serum prolactin concentrations to a level in a significant increase in plasma insulin and blood
expected after a high-protein mixed meal. There were glucose concentrations compared to the high-intensity
some minor increases in serum glucose and decreases in sweeteners and water, exceeding the normal range be-
serum insulin concentrations in all test groups, which tween 10 and 30 min. Generally, glucose concentrations
were thought to be related to postprandial alterations were within the normal range, and there were no sig-
from breakfast eaten 35 h earlier. Aspartame did not nificant differences in plasma insulin or blood glucose
affect glycemic control, as there was no increased activ- concentrations with the four high-intensity sweeteners
ity of insulin counterregulatory hormones (e.g., cortisol compared to water. The authors concluded that there is
and growth hormone). no cephalic insulin secretion from high-intensity sweet-
In a randomized, double-blind study, Melchior et al. eners.
(1991) gave normal subjects a chocolate drink with su- Teff et al. (1995) did a randomized study using 1- and
crose (50 g) or aspartame (80 mg). Plasma glucose con- 3-min oral exposures in random order to solutions of as-
centrations increased significantly after sucrose; they partame, saccharin, and sucrose and water, which were
also increased after aspartame but were not signif- then expectorated. An apple pie served as a modified
icantly different compared to the fasting condition. sham-feed condition. Blood was drawn at 0 and 1 min,
Plasma insulin concentrations increased after both as- every 2 min for 15 min, and then every 5 min for 15 min.
partame and sucrose compared to fasting. Plasma con- There were no significant increases in plasma insulin
centrations and AUC of -endorphin were significantly concentrations after any of the sweetened solutions, but
higher after aspartame compared to sucrose or fasting. the apple pie resulted in significant increases in plasma
The authors suggested a possible cephalic insulin re- insulin concentrations after both the 1- and the 3-min
ASPARTAME: REVIEW OF SAFETY S81
exposures. Thus, the taste of sweetened solutions alone The American Diabetes Association (2002) has evalu-
was not adequate to stimulate CPIR. ated the use of aspartame in diabetic diets. It found as-
Abdallah et al. (1997) reported the results of a ran- partame to be an acceptable sweetener for diabetics and
domized, double-blind, placebo-controlled study to eval- supports the consumption of aspartame at the accept-
uate CPIR with oral exposure (sucking on the tongue) to able daily intake (ADI) in the United States of 50 mg/
three different tablets: tablets contained 3 g of sucrose, kg body wt, which approximates the aspartame content
18 mg of aspartame plus 3 g polydextrose (a nonsweet in about 67 liters of beverage sweetened with 100% as-
carbohydrate), or 3 g polydextrose as the placebo. Blood partame daily for an adult. Actual aspartame consump-
was drawn at 1-min intervals for 45 min before sucking tion, however, is well below the ADI.
the tablet and for 25 min after the beginning of sucking.
Plasma glucose, insulin, and glucagon concentrations REFERENCES
were not modified after aspartame or sucrose tablets,
suggesting that sweet taste per se was not sufficient for Abdallah, L., Chabert, M., and Louis-Sylvestre, J. (1997). Cephalic
eliciting CPIR. phase responses to sweet taste. Am. J. Clin. Nutr. 65, 737743.
American Diabetes Association (2002). Evidence-based nutrition
Sugar and Aspartame in the Diets principles and recommendations for the treatment and prevention
of Diabetic Subjects of diabetes and related complications Diabetes Care 25, S50S60.
Carlson, H. E., and Shah, J. H. (1989). Aspartame and its constituent
With ongoing interest in nutritional guidelines for amino acids: Effects on prolactin, cortisol, growth hormone, insulin,
diabetics, Colaguiri et al. (1989) undertook a double- and glucose in normal humans. Am. J. Clin. Nutr. 49, 427432.
blind, crossover study for 6 weeks on each treatment Colagiuri, S., Miller, J. J., and Edwards, R. A. (1989). Metabolic ef-
to evaluate the effect of sucrose vs aspartame on the fects of adding sucrose and aspartame to the diet of subjects with
noninsulin-dependent diabetes mellitus. Am. J. Clin. Nutr. 50, 474
diets of well-controlled Type 2 diabetic subjects. Su- 478.
crose or aspartame, given with each meal and in bever-
Hartel, B., Graubaum, H.-J., and Schneider, B. (1993). Effect of
ages between meals, had no effect on glycemic control sweetener solutions on insulin secretion and blood glucose level.
(fasting blood glucose, glycosylated hemoglobin, or in- Ernahrungs-Umschau 40, 152155.
sulin concentrations). In addition, there were no effects Horwitz, D. L., McLane, M., and Kobe, P. (1988). Response to single
of either aspartame or sucrose on serum total choles- dose of aspartame or saccharin by NIDDM patients. Diabetes Care
terol, HDL-cholesterol, or triglyceride concentrations. 11, 230234.
These authors concluded that aspartame was a suitable Leon, A. S., Hunninghake, D. B., Bell, C., Rassin, D. K., and Tephly,
T. R. (1989). Safety of long-term large doses of aspartame. Arch.
sugar substitute for use in diabetics, although sucrose Intern. Med. 149, 23182324.
added as a part of the diabetic diet did not adversely Leon, A. (1999). Comment on effect of large doses of aspartame on
affect glycemic control in well-controlled diabetic urinary calcium excretion (reply to Nguyen et al.). J. Clin. Endocrin.
subjects. Metab. 84, 383384.
More recently, Tepper et al. (1996) evaluated the rela- Malaisse, W. J., Vanonderbergen, A., Louchami, K., Jijakli, H., and
tionship between sweet taste and dietary intake in sub- Malaisse-Lagae, F. (1998). Effects of artificial sweeteners on insulin
release and cationic fluxes in rat pancreatic islets. Cell. Signal. 10,
jects with Type 2 diabetes and controls administered
727733.
sucrose, fructose, and aspartame. Among the various
Melanson, K. J., Westerterp-Plantenga, M. S., Campfield, L. A., and
groups, there were no differences in sweet taste per- Saris, W. H. M. (1999). Blood glucose and meal patterns in time-
ception, pleasantness ratings, daily energy intakes, or blinded males, after aspartame, carbohydrate, and fat consump-
macronutrient composition of diets. Diabetic subjects tion, in relation to sweetness perception. Br. J. Nutr. 82, 437446.
were found to ingest less sucrose but 3.5 times more Melchior, J. C., Rigaud, D., Colas-Linhart, N., Petiet, A., Girard,
aspartame compared to controls. A., and Apfelbaum, M. (1991). Immunoreactive beta-endorphin in-
creases after an aspartame chocolate drink in healthy human sub-
jects. Physiol. Behav. 50, 941944.
Conclusions
Mller, S. E. (1991). Effect of aspartame and protein, administered
Results of the acute dosing studies reveal that aspar- in phenylalanine-equivalent doses, on plasma neutral amino acids,
aspartate, insulin and glucose in man. Pharmacol. Toxicol. 68, 408
tame does not affect the glycemic response in normal or 412.
diabetic subjects. In addition, aspartame does not affect Nehrling, J. K., Kobe, P., McLane, M. P., Olson, R. E., Kamath, S.,
metabolic control when administered to diabetic sub- and Horwitz, D. L. (1985). Aspartame use by persons with diabetes.
jects daily for up to 18 weeks. There are also no effects Diabetes Care 8, 415417.
of aspartame on plasma lipid profiles in subjects with Nguyen, U. N., Dumoulin, G., Henriet, M., and Regnard, J. (1998).
diabetes mellitus. Aspartame also had no effect on hor- Aspartame ingestion increases urinary calcium but not oxalate ex-
mones such as prolactin, growth hormone, and cortisol. cretion in healthy subjects. J. Clin. Endocrin. Metab. 83, 165168.
Several studies have evaluated the effect of aspartame Okuno, G., Kawakami, F., Tako, H., Kashikara, T., Shibamoto, S.,
Yamazaki, T., Yamamoto, K., and Saeki, M. (1986). Glucose toler-
on cephalic-phase insulin release. The results of these ance, blood lipid, insulin and glucagon concentration after single or
studies demonstrated no cephalic-phase insulin release continuous administration of aspartame in diabetics. Diabetes Res.
induced by aspartame. Clin. Pract. 2, 2327.
S82 BUTCHKO ET AL.
Overduin, J., and Jansen, A. (1997). Conditioned insulin and blood Tamura, S., Inoue, W., Kanashige, H., and Sakai, H. (1984). Experi-
sugar responses in humans in relation to binge eating. Physiol. ence on the use of aspartame for a long period in diabetic patients.
Behav. 61, 569575. Rinsho Eiyo (Clin. Nutr.) 64, 515516.
Rogers, P. J., Carlyle, J., Hill, A. J., and Blundell, J. E. (1988). Uncou- Teff, K. L., Devine, J., and Engelman, K. (1995). Sweet taste: Effect
pling sweet taste and calories: comparison of the effects of glucose on cephalic phase insulin release in men. Physiol. Behav. 57, 1089
and the three intense sweeteners on hunger and food intake. Phys- 1095.
iol. Behav. 43, 547552. Tepper, B. J., Hartfiel, L. M., and Schneider, S. H. (1996). Sweet taste
Shigeta, H., Yoshida, T., Nakai, M., Mori, H., Kano, Y., Nishioka, H., and diet in type II diabetes. Physiol. Behav. 60, 1318.
Kajiyama, S., Kitagawa, Y., Kanatsuna, T., Kondo, M., and Otsuki, Wolf-Novak, L. C., Stegink, L. D., Brummel, M. C., Persoon, T. J.,
K. (1985). Effects of aspartame on diabetic rats and diabetic pa- Filer, L. J., Jr., Bell, E. F., Ziegler, E. E., and Krause, W. L., (1990).
tients. J. Nutr. Sci. Vitaminol. 31, 533540. Aspartame ingestion with and without carbohydrate in phenylke-
Stegink, L. D., Brummel, M. C., Persoon, T. J., Filer, L. J., Jr., Bell, tonuric and normal subjects: effect on plasma concentrations of
E. F.,and Zeigler, E. E. (1990). Effect of sucrose on the metabolic amino acids, glucose, and insulin. Metabolism 39, 391396.
disposition of aspartame. Am. J. Clin. Nutr. 52, 335341. Wouassi. D., Mercier. J., Ahmaidi, S., Brun, J. F., Mercier, B., Orsetti,
Stern, S. B., Bleicher, S. J., Flores, A., Gombos, G., Recitas, D., and A., and Prefaut, C. (1997). Metabolic and hormonal responses dur-
Shu, J. (1976). Administration of aspartame in non-insulin depen- ing repeated bouts of brief and intense exercise: Effects of pre-
dent diabetics, J. Toxicol. Environ. Health 2, 429439. exercise glucose ingestion. Eur. J. Appl. Physiol. 76, 197202.
Evaluation of Appetite, Food Intake, and Usefulness
of Aspartame in Weight Control
Introduction products. Researchers have investigated the use of such
nutrient-modified foods to improve the outcome in the
Obesity has become a worldwide epidemic. The World
treatment of obesity. Such research must be well
Health Organization (WHO) has developed a compre-
controlled with careful attention paid to all variables,
hensive analysis of the problem (WHO, 2000). The pri-
including caloric intake and level of exercise. Incor-
mary concern is the health hazard of being overweight
poration of such reduced calorie foods and beverages
or obese and the issues of how best to prevent ex-
into the diet, as part of a multidisciplinary regimen
cessive weight gain and to treat different degrees of
also including exercise and lifestyle changes, can help
overweight and obesity. The prevalence rates for over-
motivated individuals to lose or maintain body weight.
weight and obese people are different in each region,
with the Middle East, Central and Eastern Europe,
and North America having higher prevalence rates. In Safety of Aspartame in Obese Individuals
most countries, women show a greater body mass in- As discussed in other sections, aspartame has under-
dex (BMI in kg/m2 ) distribution with higher obesity gone rigorous safety testing, including comprehensive
rates than do men. Obesity is usually now associated toxicology studies in animals and studies in humans. In
with poverty, even in developing countries. Abdominal addition, both short-(6 weeks) and long- (21 weeks) term
obesity in adults, with its associated enhanced morbid- randomized, double-blind, placebo-controlled studies
ity, occurs particularly in those who had lower birth (Hoffman, 1972, 1973) in obese adults have demon-
weights and early childhood stunting (James et al., strated the safety of aspartame use by this population.
2001). Obese individuals are at an increased risk for Another randomized, double-blind, placebo-controlled
the development of a number of serious diseases, in- 13-week study of aspartame administered in capsules
cluding diabetes, hypertension, cardiovascular disease, to obese adolescents and young adults on weight reduc-
gallbladder disease, and certain types of cancer (U.S. ing diets also showed no differences in weight change,
Department of Health and Human Services, 2001). At plasma triglyceride level, serum cholesterol, or param-
the same time, evidence suggests that moderate weight eters of hematologic, hepatic, or renal function (Knopp
loss (515% of body weight) will reduce health risks et al., 1976).
and medical problems in the majority of obese individ-
uals (Blackburn and Kanders, 1987, 1992; Blackburn,
Epidemiological Evaluation of Weight Change
2001). Thus, the prevention and treatment of obesity
and Artificial Sweetener Use
remains a significant worldwide public health priority
(Khaodhiar and Blackburn, 2001). In 1986, Stellman and Garfinkel suggested a rela-
Despite many billions of dollars being spent yearly tionship between artificial sweetener use and weight
(Begley, 1991) by approximately 23% of men and 40% change over a 1-year period. This was based on the re-
of women trying to lose weight (Horm and Anderson, sults of a study using self-reported body weight change
1993), long-term maintenance of weight loss with either and food intake data from a prospective epidemio-
conventional or nonconventional treatment remains logic cancer study to determine retrospectively whether
disappointing with more than 95% of all dieters regain- there was a relationship between body weight and use of
ing lost weight within 5 years of treatment (Robinson intense sweeteners. The authors concluded that users of
et al., 1993; Wadden, 1993). Recently, Anderson et al. artificial sweeteners were significantly more likely than
(2001) conducted a meta-analysis of 29 reports of long- nonusers to gain weight, regardless of initial weight.
term weight-loss maintenance indicating that weight- However, obese users lost significantly more weight
loss maintenance 4 or 5 years after a structured weight- than did obese nonusers, a fact often overlooked in dis-
loss program averages 3.0 kg or 23% of initial weight cussions of this paper. Further, the authors themselves
loss, representing a sustained reduction in body weight noted that artificial sweetener users who lost weight or
of 3.2%. Although success in weight-loss maintenance whose weight did not increase could have gained weight
has improved over the past decade, much more re- had they not consumed the sweeteners (Stellman and
search is required to enable most individuals to sus- Garfinkel, 1988).
tain the lifestyle changes in physical activity and food This study was criticized (Lavin et al., 1994) for both
choices necessary for successful weight maintenance methodological flaws in experimental design and sta-
(Blackburn, 1999). tistical analysis. The flaws identified included the use
High-intensity sweeteners, such as aspartame, are of data from an unrelated study for which they were not
not a panacea for weight loss or weight maintenance. intended, the failure to correct for bias as a result of con-
However, aspartame can replace the sugar in foods venience sampling, the unvalidated use of data from a
and beverages and thus result in reduced or no calorie homogeneous subpopulation, and the stratification of
S83
S84 BUTCHKO ET AL.
subjects by BMI that was determined from outcome tion to eat regardless of whether the aspartame was
data which was gathered near completion of the 10-year delivered in a solid or liquid medium. Recently, Wilson
retrospective study. In addition, it was inappropriate (2000) compared the effect of plain milk, sucrose-
to compare weight changes in individuals who gained sweetened milk, and aspartame-sweetened milk on
weight vs individuals who lost weight since the outcome mealtime caloric intake in young children. Children
of the study, weight change, determined the populations consumed more sweetened milk than plain milk. How-
being compared. ever, the authors found that young children do not re-
duce caloric intake at a meal to compensate for the
extra calories resulting from sucrose-sweetened milk
Appetite, Hunger, and Food Intake
whereas aspartame increased milk consumption with-
Blundell and Hill (1986, 1987) suggested that as- out providing the extra calories of sucrose-sweetened
partame may stimulate appetite, speculating that the milk. Studies on aspartame, appetite, and food intake
use of aspartame by certain individuals may lead to have been reviewed in detail by Rolls (1991), Renwick
a loss of control over appetite . . . , and contribute to (1994), Drewnowski (1995), and Rolls and Shide (1996).
disordered patterns of eating. However, these stud- As Rolls and Shide (1996) concluded, From evaluation
ies did not measure the effect of aspartame on food of the available data, there is no consistent nor com-
intake, an important omission as food intake can be pelling evidence that the intense sweetener aspartame
dissociated from reported hunger (Mattes, 1990). In increases food intake or body weight.
a later study, researchers from this same laboratory In a series of inpatient studies done in metabolic
(Rogers et al., 1988) reported that subjects had an in- wards, Porikos et al. (1977, 1982) and Porikos and Pi-
crease in hunger after drinking water or solutions of Sunyer (1984) investigated food intake following covert
saccharin, acesulfame-K, and aspartame compared to caloric dilution with aspartame to evaluate whether
glucose solution. However, actual food intake after the normal weight and obese individuals would maintain
high-intensity sweeteners was not significantly differ- their usual caloric intake in response to covert caloric
ent when compared with drinking water alone. They dilution of their diet. In one study with obese subjects
speculated that this paradoxical effect of aspartame (Porikos et al., 1977) where aspartame replaced sucrose
on appetite may be due to cephalic-phase insulin re- (approximately 25% of total calories), energy intake de-
lease (CPIR). However, as discussed in detail under the creased significantly to 77% of baseline intake during
section Endocrine Evaluations with Aspartame, as- the first 3 days and remained at 86% of baseline in-
partame does not cause CPIR. take during the second 3 days. In normal weight vol-
In another study from this same laboratory, Rogers unteers (Porikos et al., 1982), there was a significant
et al. (1990) reported that aspartame ingested without reduction in caloric intake during a 12-day period on
tasting (in capsule form) inhibited hunger and food in- an aspartame-sweetened diet. Intake was lowest in the
take. They speculated that this was due to a postinges- first 3-day period (76% of baseline) with stabilization of
tive inhibitory effect of aspartame on appetite. Black intake at 85% of baseline during the subsequent nine
et al. (1993) reported the results of an experiment com- days of the caloric diluted diet. In the third study of
paring the effects of different volumes of beverage, bev- both obese and normal weight subjects (Porikos and Pi-
erages with aspartame in solution, and beverages with Sunyer, 1984), the investigators again observed a 15%
aspartame in capsules on appetite and food intake. In reduction in energy intake when aspartame was substi-
contrast to the report of Rogers et al. (1990), Black and tuted for sucrose in the diet. While these data suggest
co-workers (1993) reported that aspartame in capsules that there is no short-term physiologic drive to replace
had no effect on appetite. Furthermore, they concluded all calories removed through the use of aspartame, the
that any appetite reduction after consumption of an long-term effects are unclear. In addition, these studies
aspartame-sweetened beverage is likely due to the vol- have been criticized because the baseline diet encour-
ume of the drink and not the aspartame. aged overeating and weight gain and did not reflect the
Many other studies, using various research para- subjects normal diet.
digms, have been done to evaluate the effect of asp- In an outpatient study, Tordoff and Alleva (1990) in-
artame on hunger, appetite and food intake. Replacing vestigated the effect of adding aspartame in a blinded
sucrose with aspartame in foods or beverages has not manner to a normal diet in free-living normal weight
been shown to increase food intake or hunger in chil- subjects. During two 3-week experimental periods, sub-
dren (Anderson et al., 1989; Birch et al., 1989) and has jects received four bottles of beverage (40 oz. or 1135 g)
not been shown to increase food intake in normal weight sweetened with either aspartame or high-fructose corn
(Blundell and Hill, 1987; Rolls et al., 1989, 1990; Black syrup per day. During one 3-week control period, they
et al., 1991; Canty and Chan, 1991; Drewnowski et al., received no experimental drink. Compared with results
1994a,b) or in overweight men and women (Rodin, 1990; during the control period, there was a significant re-
Drewnowski, 1994b). It should be noted that all of these duction in caloric intake in both females and males
studies reported either unchanged or reduced motiva- and decreased body weight of males during the period
ASPARTAME: REVIEW OF SAFETY S85
Non-aspart ame
significant increase in caloric intake and body weight 0
in both sexes. *
Kg
ets of overweight subjects for 10 weeks resulted in in- -10
creases in total energy intake, body weight, fat mass,
and blood pressure. In contrast, body weight, fat mass,
-15
and blood pressure decreased in overweight subjects
who were supplemented with artificial sweeteners.
-20
Week 19 Week 71 Week 156
To evaluate the effect of aspartame on short-term FIG. 1. Mean change in body weight at weeks 19, 71, and 156
control of body weight in obese subjects, Kanders compared to baseline with aspartame and no-aspartame treatment
et al. (1988, 1996) recruited 59 obese (37 kg/m2 ) groups. Adapted from Kanders et al. (1996).
men and women to participate in a pilot study with
a 12-week multidisciplinary diet program. Subjects During the active weight loss period, all subjects
were randomly assigned to consume a balanced deficit also attended weekly 1-h sessions where they were in-
diet (1000 200 kcal/day) with or without aspartame- structed on behavioral and lifestyle strategies to facili-
containing products. Although not statistically signifi- tate weight loss. During 12-month maintenance and the
cant, women in the aspartame group (N = 24) lost 19-month follow-up, the groups met monthly. Through-
3.7 pounds more than women in the no-aspartame out the study, regular exercise, mainly in the form of
group (N = 21). Men showed the opposite trend, with walking, was strongly encouraged. Parameters evalu-
those in the no-aspartame group (N = 7) losing about ated at baseline, 19, 71, and 156 weeks included body
4 lb more than those in the aspartame group (N = 4). weight, aspartame intake, exercise level, and subjective
Forty-six of these subjects (11 males and 35 females) ratings of hunger, desire for sweets, and eating control.
participated in a 1-year follow-up study. Increased lev- One-hundred thirty-six subjects completed the active
els of physical activity, increased consumption of as- weight loss phase; 125 subjects completed the mainte-
partame, and decreased desire for sweets were asso- nance phase; and 86 subjects completed the follow-up
ciated with maintenance of weight loss at the 1-year phase. Subjects in both treatment groups lost a mean
follow-up. In male subjects, aspartame intake at the of approximately 10% of body weight (10 kg) during the
end of follow-up was associated with better weight 19 weeks of active weight loss (Fig. 1). Among subjects
maintenance. Although definitive conclusions could not in the aspartame group, aspartame consumption was
be drawn due to the small sample size, aspartame positively associated with weight loss. The desire for
consumption did not cause weight gain and may be sweets decreased significantly in the aspartame group
beneficial in promoting weight loss and maintenance but not in the no-aspartame group; hunger did not differ
when used as part of a multidisciplinary weight control significantly from baseline in either treatment group,
program. but eating control increased significantly in both treat-
As a follow-up to this pilot study, Blackburn et al. ment groups. In the maintenance phase (weeks 19
(1997) conducted a randomized, controlled, prospective 71), hunger and desire for sweets remained unchanged
clinical study to investigate whether the addition of as- within both treatment groups. Eating control decreased
partame to a multidisciplinary weight control program significantly in both maintenance and follow-up phases
would improve weight loss and long-term control of in both groups, suggesting more uncontrolled eating
body weight in obese women. Initially, 168 obese women during maintenance and follow-up.
aged 20 to 60 years were placed on a nutrient-balanced At the end of the maintenance phase (Week 71), sub-
deficit diet (1000 200 kcal/day) for 3 weeks. At the end jects in the aspartame group experienced a 3.1% mean
of this period, 163 subjects were instructed to continue weight regain, and those in the no-aspartame group
the balanced deficit diet and were randomly assigned regained a mean of 4.9%. By the end of the follow-up
either to consume aspartame-sweetened foods and bev- phase (Week 156), subjects in the aspartame group had
erages during the remaining 16 weeks of active weight regained an additional 2.4%, with a net weight loss
loss phase or to avoid such products. During the 1-year from baseline of 5.1%. In contrast, subjects in the no-
weight maintenance phase and 2-year follow-up peri- aspartame group had a gain of 5.4%, with a net weight
ods, participants were encouraged to continue to con- loss of 0.3% from baseline (Fig. 1). Significant predic-
sume or avoid aspartame-containing products accord- tors of better weight control from baseline to Week
ing to their original group assignment. 156 included increased exercise, increased self-reported
S86 BUTCHKO ET AL.
eating control, and initial treatment group assignment, Blackburn, G. L. (1999). Sweeteners and weight control. In Low-
where aspartame group subjects had an advantage Calorie Sweeteners: Present and Future (A. Corti, Ed.), pp. 7787.
Karger, Basel.
over the no-aspartame group subjects. The authors con-
Blackburn, G. L. (2001). The public health implications of the dietary
cluded that aspartame, as part of a multidisciplinary approaches to stop hypertension trial. Am. J. Clin. Nutr. 74, 12.
weight control program, may facilitate weight control. Blundell, J. E., and Hill, A. J. (1986). Paradoxical effects of an intense
sweetener (aspartame) on appetite. Lancet 1, 10921093.
Conclusions Blundell, J. E., and Hill, A. J. (1987). Artificial sweeteners and the
control of appetite. In Implications for the Eating Disorders in Fu-
Nutritional surveys estimate that added sugars ac- ture of Predictive Safety Evaluation (A. N. Worden, D. V. Parke, and
count for approximately 1520% of calories in the Amer- J. Marks, Eds.), pp. 263281. MTP Press, Boston.
ican diet. High-intensity sweeteners such as aspartame Canty, D. J., and Chan, M. M. (1991). Effects of consumption of caloric
can provide sweetness and palatability without adding vs noncaloric sweet drinks on indices of hunger and food consump-
calories and may facilitate control of body weight in in- tion in normal adults. Am. J. Clin. Nutr. 53, 11591164.
dividuals who are motivated to do so. The few studies Drewnowski, A. (1995). Intense sweeteners and the control of ap-
petite. Nutr. Rev. 53, 17.
that indicated an increased motivation to eat following
Drewnowski, A., Massien, C., Louis-Sylvestre, J., Fricker, J.,
the consumption of aspartame were not replicated by a Chapelot, D., and Apfelbaum, M. (1994a). Comparing the effects of
number of other studies. Further, inpatient investiga- aspartame and sucrose on motivational ratings, taste preferences,
tions of nondieting obese and normal weight individuals and energy intakes in humans. Am. J. Clin. Nutr. 59, 338345.
have demonstrated incomplete caloric compensation af- Drewnowski, A., Massien, C., Louis-Sylvestre, J., Fricker, J.,
ter the covert replacement of sucrose with aspartame. Chapelot, D., and Apfelbaum, M. (1994b). The effects of aspartame
Participation in a multidisciplinary weight control versus sucrose on motivational ratings, taste preferences, and en-
ergy intakes in obese and lean women. Int. J. Obesity 18, 570
program that included the use of aspartame-sweetened 578.
foods and beverages facilitated weight loss, as well as Hoffman R. (1972). Short Term Tolerance of Aspartame by Obese
long-term maintenance of a reduced body weight. Taken Adults, research report submitted by G. D. Searle and Co. to the
together, these results suggest that aspartame, when Food and Drug Administration (E-24).
incorporated into a multidisciplinary weight manage- Hoffman, R. (1973). Long Term Tolerance of Aspartame by Obese
ment program also including exercise and behavior Adults, research report submitted by G. D. Searle and Co. to the
Food and Drug Administration (E-64).
modification, may aid in the long-term control of body
Horm, J., and Anderson, K. (1993). Who in America is trying to lose
weight.
weight? Ann. Intern. Med. 119, 672676.
James P. T., Leach, R., Kalamara, E., and Shayegli, M. (2001). The
REFERENCES world wide obesity epidemic. Obes. Res. 9, 228S233S.
Kanders, B. S., and Blackburn, G. L. (1992). Reducing primary risk
Anderson, G. H, Saravis, S., Schacher, R., Zlotkin, S., and Leiter, L. A. factors by therapeutic weight loss. In The Treatment of the Seriously
(1989). Aspartame: Effect on lunch-time food intake, appetite and Obese Patient (T. A. Wadden and T. B. Van Itallie, Eds.), pp. 213
hedonic response in children. Appetite 13, 93103. 230. Guilford, New York.
Anderson, J. W., Konz, E. C., Frederich, R. C., and Wood, C. L. (2001). Kanders, B. S., Blackburn, G. L., Lavin, P. T., and Kienholz, M. (1996).
Long-term weight-loss maintenance: A meta-analysis of US stud- Evaluation of weight control. In The Clinical Evaluation of a Food
ies. Am. J. Clin. Nutr. 74, 579584. Additive: Assessment of Aspartame (C. Tschanz, H. H. Butchko,
Astrup, A., Raben, A., Vasilaras, T. H., and Mller, A. C. (2002). Su- W. W. Stargel, and F. N. Kotsonis, Eds.), pp. 289299. CRC Press,
crose in soft drinks is fattening: A randomized 10 week study in Boca Raton, FL.
overweight subjects. Am. J. Clin. Nutr. 75, 405S. Kanders, B. S., Lavin, P. T., Kowalchuk, M. B., Greenberg, I., and
Begley, C. E. (1991). Government should strengthen regulation in the Blackburn, G. L. (1988). An evaluation of the effect of aspartame
weight loss industry. J. Am. Diet. Assoc. 91, 12551257. on weight loss. Appetite 11(Suppl. 1), 7384.
Birch, L. L., McPhee, L., and Sullivan, S. (1989). Childrens food in- Khaodhiar, L., and Blackburn, G. L. (2001). Obesity assessment. Am.
take following drinks sweetened with sucrose or aspartame: Time Heart J. 42, 10951101.
course effects. Physiol. Behav. 45, 387395. Knopp, R. H., Brandt, K., and Arky, R. A. (1976). Effects of aspar-
Black, R. M., Tanaka, P., Leiter, L. A., and Anderson, G. H. (1991). Soft tame in young persons during weight reduction. J. Toxicol. Environ.
drinks with aspartame: Effect on subjective hunger, food selection, Health 2, 417428.
and food intake of young adult males. Physiol. Behav. 49, 803 Lavin, P. T., Sanders, P. G., Mackey, M. A., and Kotsonis, F. N. (1994).
810. Intense sweeteners use and weight change among women: A cri-
Black, R. M., Leiter, L. A., and Anderson, G. H. (1993). Consuming tique of the Stellman and Garfinkel study. J. Am. Coll. Nutr. 13,
aspartame with and without taste: Differential effects on appetite 102105.
and food intake of young adult males. Physiol. Behav. 53, 459466. Mattes, R. (1990). Effects of aspartame and sucrose on hunger and
Blackburn, G. L., and Kanders, B. S. (1987). Medical evaluation and energy intake in humans. Physiol. Behav. 47, 10371044.
treatment of the obese patient with cardiovascular disease. Am. J. Porikos, K. P., and Pi-Sunyer, F. X. (1984). Regulation of food intake
Cardiol. 60, 55G58G. in human obesity: Studies with caloric dilution and exercise. Clin.
Blackburn, G. L., Kanders, B. S., Lavin, P. T., Keller, S. D., and What- Endocrinol. Metab. 13, 547561.
ley, J. (1997). The effect of aspartame as part of a multidisciplinary Porikos, K. P., Booth, G., and Van Itallie, T. B. (1977). Effect of covert
weight-control program on short- and long-term control of body nutritive dilution on the spontaneous intake of obese individuals:
weight. Am. J. Clin. Nutr. 65, 409418. A pilot study. Am. J. Clin. Nutr. 30, 16381644.
ASPARTAME: REVIEW OF SAFETY S87
Porikos, K. P., Hesser, M. F., and Van Itallie T. B. (1982). Caloric reg- Rolls, B. J., Laster, L. J., and Summerfelt, A. (1989). Hunger and
ulation in normal weight men maintained on a palatable diet of food intake following consumption of low-calorie foods. Appetite 13,
conventional foods. Physiol. Behav. 29, 293300. 115127.
Renwick, A. G. (1994). Intense sweeteners, food intake, and the weight Stellman, S. D., and Garfinkel, L. (1986). Artificial sweetener use
of a body of evidence. Physiol. Behav. 55, 139143. and one-year weight change among women. Prev. Med. 15, 195
Robinson, J. I., Hoerr, S. L., Strandmark, J., and Mavis, B. (1993). 202.
Obesity, weight loss and health. J. Am. Diet. Assoc. 93, 445449. Stellman, S. D., and Garfinkel, L. (1988). Patterns of artificial sweet-
Rodin, J. (1990). Comparative effects of fructose, aspartame, glucose, ener use and weight change in an American Cancer Society prospec-
and water preloads on calorie and macronutrient intake. Am. J. tive study. Appetite 11 (Suppl.), 8591.
Clin. Nutr. 51, 428435. Tordoff, M. G., and Alleva, A. M. (1990). Effect of drinking soda sweet-
Rogers, P. J., Carlyle, J., Hill, A. J., and Blundell, J. E. (1988). Uncou- ened with aspartame or high-fructose corn syrup on food intake and
pling sweet taste and calories: Comparison of the effects of glucose body weight. Am. J. Clin. Nutr. 51, 963969.
and three intense sweeteners on hunger and food intake. Physiol. U.S. Department of Health and Human Services. (2001). The Sur-
Behav. 43, 547552. geon Generals Call to Action to Prevent and Decrease Overweight
Rogers, P. J., Fleming, H. C., and Blundell, J. E. (1990). Aspartame and Obesity. U.S. Department of Health and Human Services, Pub-
ingested without tasting inhibits hunger and food intake. Physiol. lic Health Service, Office of the Surgeon General, Rockville, MD.
Behav. 47, 12391243. [Available from U.S. GPO, Washington, DC]
Rolls, B. J. (1991). Effects of intense sweeteners on hunger, food in- Wadden, T. A. (1993). Treatment of obesity by moderate and severe
take, and body weight: A review. Am. J. Clin. Nutr. 53, 872878. caloric restriction: Results of clinical research trials. Ann. Intern.
Rolls, B. J., and Shide, D. J. (1996). Evaluation of hunger, food intake, Med. 119, 688693.
and body weight. In The Clinical Evaluation of a Food Additive: Wilson, J. F. (2000). Lunch eating behavior of preschool children: Ef-
Assessment of Aspartame (C. Tschanz, H. H. Butchko, W. W. Stargel, fects of age, gender, and type of beverage served. Physiol. Behav.
and F. N. Kotsonis, Eds.), pp. 275287. CRC Press, Boca Raton, FL. 70, 2733.
Rolls, B. J., Kim, S., and Fedoroff, I. C. (1990). Effects of drinks sweet- World Health Organization (WHO). (2000). Obesity: Preventing and
ened with sucrose or aspartame on hunger, thirst and food intake Managing the Global Epidemic, WHO Obesity Technical Report
in men. Physiol. Behav. 48, 1926. Series 894. WHO, Geneva.
Conclusions
Aspartame is unique among high-intensity sweet- toxicity; phenylalanine and neurochemistry; methanol;
eners because it is metabolized by digestive esterases postmarketing surveillance of anecdotal reports of
and peptidases to three common dietary constituents health effects; allegations of headaches, seizures,
the amino acids, aspartic acid and phenylalanine, and effects on behavior, cognitive function, and mood, as
methanol. These constituents are used by the body in well as allergic-type reactions and brain tumors; safety
the same metabolic pathways as when they are also of aspartame use by potentially sensitive populations;
derived from foods, such as meat, milk, fruits, and veg- effects on endocrine function; and effects on appetite,
etables. Further, these constituents are derived from food intake, and weight control. The weight of the evi-
common foods in much larger amounts than from as- dence from these additional studies further confirmed
partame in foods and beverages. For example, a glass of the results of the earlier studies and expanded the
no-fat milk provides about 6 times more phenylalanine database on aspartame safety. The conclusions of these
and 13 times more aspartic acid and a glass of tomato extensive evaluations are summarized below.
juice provides about 6 times more methanol than an
equivalent volume of beverage sweetened 100% with Preclinical Safety Evaluation of Aspartame
aspartame. Much of the scientific research, both before
The definitive preclinical studies done with aspar-
and after regulatory approval, focused on the safety of
tame demonstrated that aspartame is not toxic, carcino-
these components.
genic, mutagenic, or teratogenic and has no effects on
Prior to the regulatory approvals of aspartame, a com-
reproduction. In addition, there were no effects of aspar-
prehensive battery of toxicology studies was done in an-
tame on the central nervous system, gastrointestinal
imals to evaluate acute, subchronic, and long-term tox-
tract, endocrine system, or reproductive system. From
icity, carcinogenicity, genetic toxicity, and reproductive
the results of the toxicology studies, a no-observed-
toxicity and teratogenicity. These studies demonstrated
effect level of at least 4000 mg/kg body wt and an ADI
that aspartame is not toxic, carcinogenic, mutagenic, or
of 40 mg/kg body wt were established for aspartame in
teratogenic and has no effects on reproduction or de-
Europe and Canada and by JECFA. When aspartame
velopment. Other studies demonstrated aspartame has
was first approved in the United States, the FDA estab-
no effects on the central nervous system, gastrointesti-
lished an ADI of 20 mg/kg body wt for aspartame but
nal tract, endocrine system, and reproductive system
later increased the ADI to 50 mg/kg body wt/day based
or on postnatal developmental in infant primates. In
on the availability of human data.
addition, a number of metabolism and tolerance stud-
Since regulatory approvals of aspartame in the early
ies were done in humans with healthy adults, infants,
1980s, a number of additional animal safety studies
children and adolescents; PKU heterozygotes; obese
were completed. In no cases have subsequent studies
individuals; and individuals with diabetes. This com-
altered original determinations of safety established
prehensive database supporting the safety of aspar-
by regulatory agencies as part of the approval process.
tame has been evaluated by regulatory agencies and
Regulatory agencies regularly review preclinical safety
expert committees around the world and determined
data; these bodies evaluate safety data according to de-
to demonstrate aspartames safety for its intended use
fined standards and internationally accepted biologi-
as a sweetener. During the past 20 years since regula-
cal and statistical criteria. Critics who fault aspartame
tory approvals, aspartame has been safely consumed by
safety data, or those who postulate deficiencies in the
hundreds of millions of people worldwide.
aspartame database, rarely have a background in eval-
Since the marketing of aspartame, a number of sci-
uating safety data and seldom apply internationally ac-
entific issues were raised and thoroughly addressed
cepted criteria in formulating their critiques.
through animal and human studies. As before approval,
these issues centered largely on hypothetical toxicity
Intake of Aspartame vs The Acceptable Daily Intake
of its three metabolic components when given in ex-
tremely large doses and continued to include poten- Aspartame intake has been evaluated in the United
tial neurotoxicity (excitotoxicity) of aspartic acid, po- States and several European countries, as well as
tential effects of phenylalanine on brain function, and Canada, Australia, and Brazil. Average daily aspar-
potential toxicity of the methanol metabolite formate. tame intake at the 90th percentile (eaters only) in the
These issues were resolved to the satisfaction of world- U.S. general population in the last survey was 3 mg/
wide regulatory agencies prior to approval of aspar- kg body wt. Consumption by 2- to 5-year-old children
tame. Nonetheless, additional studies after approval in these surveys ranged from about 2.5 to 5 mg/kg
have included evaluation of additional preclinical safety body wt/day. Although methodologies differed among
paradigms; aspartame intake relative to the accept- the evaluations from various countries, aspartame in-
able daily intake; metabolism; aspartate and excito- take in these surveys is remarkably consistent and is
S88
ASPARTAME: REVIEW OF SAFETY S89
well below the U.S. ADI of 50 mg/kg body wt/day and result in disturbances in brain neurotransmitter con-
that of 40 mg/kg body wt/day for much of the rest of the centrations. However, numerous metabolism studies in
world. adult humans, infants, and PKU heterozygotes demon-
strate that consumption of aspartame in amounts well
Metabolism of Aspartame above typical consumption levels results in plasma
phenylalanine concentrations that are safe.
The metabolism of aspartame has been evaluated
Numerous studies using enormous doses of aspar-
both with doses relevant to actual consumption and at
tame were done in laboratory animals to evaluate the
very large doses highly unlikely ever to be consumed.
hypothesis that very large doses of aspartame might
These studies were done in healthy adults, 1-year-old
alter brain concentrations of neurotransmitters and
children, adult PKU heterozygotes, PKU homozygous
their metabolites. The results of these studies do not
individuals, and individuals with sensitivity to MSG
support this hypothesis, but rather show no repro-
and have included various testing paradigmsacute
ducible effects on various brain neurotransmitter sys-
bolus dosing, repeated dosing, dosing with meals, and
tems after enormous doses of aspartame. Other studies
dosing with monosodium glutamate (MSG). In addi-
show no effects of very large doses of aspartame on the
tion, Phe/LNAA has been calculated in a number of
central release of neurotransmitters and receptor ki-
these studies; this ratio is thought to be predictive of
netics. Despite large increases in plasma Phe/LNAA,
the amount of phenylalanine that enters the brain.
there are no consistent effects on brain chemistry, even
Together, these studies demonstrate that amounts of
with doses of aspartame up to 1000 times the 90th per-
aspartame far exceeding 90th percentile consumption
centile human consumption. Thus, it is not unexpected
result in concentrations of aspartames constituents
that the data are overwhelming that aspartame does
aspartate, phenylalanine, and methanolthat are con-
not affect brain chemistry or function (e.g., headaches,
sidered safe.
seizures, or behavior, cognition, and mood).
Aspartate and Excitotoxicity
Safety of Methanol from Aspartame and the Diet
There was speculation early on that the aspartate
constituent of aspartame, especially when consumed The typical diet, including fruits and vegetables,
with foods containing MSG, may result in an increase juices, wine, and other alcoholic beverages, provides
in the combined plasma concentrations of aspartate greater amounts of methanol than does aspartame in
and glutamate which might pose a risk of focal brain products. Nonetheless, the safety of methanol derived
lesions. Focal brain lesions are reproducible after ex- from aspartame was extensively evaluated in both tol-
tremely high doses of glutamate or aspartate in neona- erance and metabolism studies in humans. A dose of at
tal rodents, but whether these lesions occur in primates least 50 mg/kg body wt aspartame must be ingested to
remains very controversial in that several investigators detect any increase in blood methanol concentrations.
have failed to replicate the single report of this effect in No increases in blood formate concentrations, the toxic
infant primates. However, numerous metabolism stud- metabolite of methanol, are seen even after bolus doses
ies in humans have demonstrated that it is impossible of aspartame up to 200 mg/kg body wt and after long-
for a human ever to consume enough aspartame in prod- term daily doses of 75 mg/kg body wt/day. These doses
ucts, even when combined with MSG, to raise plasma of aspartame are the equivalent of an adult consuming
concentrations of aspartate or glutamate or glutamate about 28 liters of beverage sweetened 100% with as-
plus aspartate to those concentrations associated with partame as a bolus or over 10 liters of beverage daily
toxicity in neonatal rodents. for 6 months, respectively. It is clear that the capacity
for methanol metabolism in humans is far greater than
Phenylalanine and Neurochemistry estimated dietary intake of methanol from aspartame.
Thus, it is impossible for a human ever to consume
Speculation has been made that the phenylalanine enough aspartame in products to raise blood formate
component of aspartame may alter brain function concentrations to levels associated with toxicity.
and, consequently, result in effects such as headaches,
seizures, and behavioral, mood, or cognitive changes.
Postmarketing Surveillance: Evaluation of Anecdotal
Further, it was hypothesized that aspartame, as a
Reports of Health Effects
source of phenylalanine without the other large neu-
tral amino acids that compete for transport across As a first for a food additive, a postmarketing surveil-
the bloodbrain barrier, would selectively increase en- lance program was implemented to document and
try of phenylalanine into the brain by increasing the evaluate anecdotal reports from consumers of adverse
Phe/LNAA ratio. It was further speculated that any health effects allegedly associated with aspartame. The
such increase in entry of phenylalanine, accompanied vast majority of these reports were from U.S. con-
by decreased entry of tyrosine and tryptophan, might sumers; however, reports from other countries were also
S90 BUTCHKO ET AL.
included. Epidemiologists at CDC and FDA concluded of an inadequate baseline period. Studies with individu-
that the reports represented symptoms that were gen- als who were convinced that aspartame was responsible
erally mild and that are common in the general popu- for their seizures and in children with seizure disorders,
lation. Although it is not possible to determine a cause- along with studies in healthy adults and PKU heterozy-
and-effect relationship based on such information, it gotes, demonstrate no effects of aspartame on clinical
was utilized to guide additional research efforts. As a seizures or EEGs. Thus, the evidence is compelling that
result, a number of studies were done in the postmar- aspartame, even in amounts greatly exceeding 90th per-
keting period to evaluate these allegations, including centile intake, is not a proconvulsant and has no effect
headaches, seizures, behavior, cognition, and mood ef- on EEGs.
fects, and allergic-type reactions. Other studies were
done to evaluate issues regarding potential sensitivity
in certain subpopulations. The results of these studies Evaluation of Aspartame and Behavior, Cognitive
demonstrated that, when evaluated under randomized, Function, and Mood
double-blind, placebo-controlled conditions with appro-
A number of studies using various testing paradigms
priate statistical analysis, aspartame is not associated
and very large doses of aspartame (hundreds of times
with adverse health effects.
90th percentile average daily aspartame consumption
in the United States) have been done to evaluate behav-
Evaluation of Aspartame and Headaches
ior, learning, and memory in laboratory animals. The
Headache is the most common symptom anecdo- only observed effects are seen after parenteral (gener-
tally associated with aspartame. A randomized, double- ally ip) dosing, a route of administration not relevant
blind, placebo-controlled study with individuals who to the consumption of aspartame in the diet. Taken as
were convinced that aspartame caused their headaches a whole, the data from animal studies do not provide
was done in a clinical research unit where confounding support for purported effects of aspartame on behav-
variables that are present in an outpatient setting could ior following human consumption. Further, studies in
be controlled. The authors of this study found no differ- humans, including healthy adults and PKU heterozy-
ence between the occurrence of headache after aspar- gous adults, healthy children, and children with ADD,
tame and placebo. In addition, there is no association ADHD, or sugar sensitivity, have evaluated various
between aspartame and headache in numerous other measures of behavior, cognitive performance, and mood
human studies that have been done with aspartame, in- and have consistently demonstrated no effects of aspar-
cluding a high-dose, long-term study. The few published tame. Thus, aspartame, even when given in doses that
reports suggesting an association between aspartame result in significant increases in Phe/LNAA, has no ef-
and headache are either anecdotal in nature, case re- fect on behavior, cognition, or mood.
ports, or outpatient studies that had statistical issues
that prevented drawing any valid conclusions. Thus, the
weight of the scientific data demonstrates that aspar- Evaluation of Aspartame and Allergic-Type Reactions
tame does not cause headache.
Several single case reports suggested that aspartame
is associated with allergic-type reactions. However, sev-
Evaluation of Aspartame and Seizures and
eral clinical studies have been done to evaluate this
Electroencephalograms (EEGs)
issue. In two of these studies with subjects reporting
Numerous studies using various animal models of allergic-type reactions associated with aspartame, ex-
epilepsy and several human studies evaluated allega- tensive recruitment efforts over several years failed to
tions that aspartame is associated with seizures and identify more than about 1020 subjects to enroll in
changes on EEGs. The studies in animals used doses of each of these studies. One of these studies also included
aspartame up to 3000 mg/kg body wt or 1000 times 90th evaluation of allergic-type reactions with diketopiper-
percentile daily intake in the United States (equivalent azine and -aspartame, which are conversion products
to an adult human consuming about 420 liters of a bev- of aspartame. Evaluations of allegations of allergic-type
erage sweetened with 100% aspartame as a single bo- reactions and aspartame in controlled clinical studies
lus). Although some studies using very large bolus doses have consistently demonstrated that aspartame is not
of aspartame in the pentylenetetrazol (PTZ) model in an allergen. Further, animal and in vitro studies demon-
rats indicate an effect of aspartame, these results are strate that aspartame is not a direct mast cell or ba-
not reproducible in other laboratories and appear to sophil secretagogue and has no effect on inflammation
be largely specific to species and study conditions. One parameters such as carrageenan-induced paw edema,
study in children with absence seizures appears to sug- connective tissue formation, and adjuvant arthritis.
gest an effect of aspartame on the time spent per hour in Thus, the weight of evidence demonstrates that aspar-
spike-wave discharges, however, the study lacked a true tame is not associated with allergic-type reactions in
placebo and had serious methodological issues because experimental models or humans.
ASPARTAME: REVIEW OF SAFETY S91
Evaluation of Aspartame and Brain Tumors individuals with Parkinsons disease, whose symptoms
and responsiveness to levodopa are sensitive to fluc-
In the late 1970s, Olney performed a post hoc analysis
tuations in plasma amino acid concentrations. Finally,
on results from long-term bioassays in rats to suggest
aspartame has been demonstrated to be safe for indi-
that aspartame may be associated with brain tumors. In
viduals with renal disease and individuals with liver
one study he combined independent treatment groups
disease.
and then claimed his results showed a doseresponse
relationship between aspartame and brain tumors. Fur-
ther, he asserted that another long-term bioassay in Endocrine Evaluations with Aspartame
rats was invalid because he claimed the incidence of Individuals with diabetes are likely to be enthusi-
brain tumors in controls exceeded what he considered astic consumers of aspartame because of their dietary
to be appropriate for historical controls. Regardless of restrictions. Thus, studies were done to evaluate the ef-
Olneys claims, scientists in regulatory agencies and ex- fect of aspartame on plasma glucose and insulin concen-
pert committees around the world, including the U.S. trations. From the results of the acute dosing studies,
FDA, the Canadian Health Protection Branch, the UK aspartame does not affect the glycemic response in
Committee on Toxicity, and the Joint FAO/WHO Ex- healthy individuals or diabetic subjects. In addition,
pert Committee on Food Additives, who have evaluated aspartame does not affect glycemic control, including
these data according to standard and internationally glycohemoglobin concentrations, when given to dia-
accepted criteria, have concluded that aspartame is not betic individuals over 18 weeks. Aspartame also has
a carcinogen. no effect on other hormones, such as prolactin, growth
In 1996, Olney et al. again asserted that aspartame hormone, and cortisol. Several studies have demon-
may be associated with brain tumors, repeating earlier strated no effect of aspartame on cephalic-phase in-
claims regarding the rat studies and further assert- sulin release, disproving suggestions that aspartame
ing the marketing of aspartame was associated with may cause cephalic-phase insulin release and thereby
an apparent increase in incidence of human brain tu- affect hunger. Thus, aspartame has been found to be
mors in the United States. This newer assertion was an acceptable sweetener for diabetic individuals by The
based on their analysis of selected data from the U.S. American Diabetes Association.
SEER tumor database from the U.S. NCI. However, NCI
has concluded that Olneys claims were not valid, as
have regulatory agencies in the United States, the Eu- Evaluation of Appetite, Food Intake, and Usefulness
ropean Union, the United Kingdom, and Australia/New of Aspartame in Weight Control
Zealand. The scientific consensus is that there is no as- High-intensity sweeteners such as aspartame can
sociation between aspartame consumption and brain provide sweetness and palatability without adding calo-
tumors. ries to the diet and thus may facilitate control of body
weight in motivated individuals. Although a few stud-
Use of Aspartame by Potentially Sensitive Populations
ies indicated an apparent increase in hunger or ap-
Several subpopulations of individuals, including petite following aspartame, a number of other stud-
PKU heterozygotes, individuals with depression, in- ies have not replicated these results. In addition, in-
dividuals with dizziness, individuals with Parkinsons vestigations of nondieting obese and normal weight in-
disease, individuals with renal disease, and individu- dividuals in metabolic wards demonstrate incomplete
als with liver disease, have been studied because of the caloric compensation after the covert replacement of
possibility that these populations may be potentially sucrose with aspartame. Taken together, the results
sensitive to aspartame. Studies in depressed patients of the research demonstrate no compelling evidence
and patients with dizziness did not have an adequate that aspartame has an effect on appetite and food in-
number of subjects completed; thus it is not possible to take. In a prospective weight loss study, a multidisci-
draw any valid conclusions from these studies. Stud- plinary weight control program that included the use of
ies in PKU heterozygotes demonstrate that aspartame, aspartame-sweetened foods and beverages facilitated
even in amounts well above average 90th percentile weight loss, as well as long-term maintenance of a
consumption levels, results in plasma phenylalanine reduced body weight. These results suggest that as-
concentrations that are considered safe. Further, large partame, when incorporated into a multidisciplinary
doses of aspartame given over 12 weeks have no effect weight management program also including exercise
on cognitive function or EEGs compared to placebo in and behavior modification, may aid in the long-term
these individuals. No effects of aspartame are seen in control of body weight.
Summary
The extensive research done with aspartame over- studies suggestive of an apparent effect of aspartame
whelmingly demonstrates it safety. In addition, aspar- are not reproducible by others, are done with dosage
tame has been consumed by hundreds of millions of routes of administration (e.g., ip or in vitro) not rele-
people around the world over the past 20 years, repre- vant to aspartame consumption, or are so flawed from
senting billions of man-years of safe exposure. Nonethe- a methodological or statistical standpoint that valid
less, the safety of aspartame continues to be challenged conclusions cannot be drawn.
by individuals who either ignore the data or are not The scientific evaluation of aspartames safety has
familiar with the totality of the data. The allegations extended well beyond standard safety testing for food
raised regarding aspartame by a few individuals have additives. When the safety data for aspartame are eval-
not been based on controlled scientific studies. Rather, uated as a whole, the weight of scientific evidence is
they have been largely based on anecdote, personal clear that aspartame is safe for its intended uses, and
opinion, conjecture and hypothesis. Further, the few there are no unresolved questions regarding its safety.
S92
Appendix
List of Acronyms and Abbreviations
5-HIAA5-hydroxyindole acetic acid LiCl lithium chloride
-AP -aspartylphenylalanine LNAA large neutral amino acid
ADA American Diabetes Association MAFF UK Ministry of Agriculture,
ADD attention deficit disorder Fisheries & Food
ADHD attention deficit hyperactivity disorder MEDLINE database from U.S. National Library
ADI acceptable daily intake of Medicine
ANZFA Australia/New Zealand Food MRCA MRCA Information Services,
Authority Northbrook, Illinois
ARMS Adverse Reaction Monitoring System MRI magnetic resonance imaging
ATP adenosine triphosphate MSG monosodium glutamate
AUC area under the concentration curve NCI National Cancer Institute
-AP -aspartylphenylalanine NMDA N-methyl-D-aspartate
CDC Centers for Disease Control PBOI Public Board of Inquiry
CFSAN Center for Food Safety and Applied Phe phenylalanine
Nutrition Phe/LNAA plasma ratio of phenylalanine to the
Cmax maximum plasma concentration other large neutral amino acids
CNS central nervous system PKU phenylketonuria
CPD cumulative cell population doubling PKUH heterozygotes for phenylketonuria
CPIR cephalic-phase insulin release PME phenylalanine methyl ester
CSFII Continuing Survey of Food Intakes by POMS profile of mood states
Individuals PSE portal systemic index
CT computed tomography PTZ pentylenetetrazol
DKP diketopiperazine RAST radioallergosorbent test
DNA deoxyribonucleic acid SCF EU Scientific Committee for Food
DOPAC dihydroxyphenylacetic acid SEER U.S. Surveillance, Epidemiology and
DSST Digit Symbol Substitution Test End Results
EEG electroencephalogram SPARTANS cognitive battery of aviation-related
EU European Union information processing tasks
FAO Food and Agriculture Organization SSS Stanford sleepiness scale
FDA U.S. Food and Drug Administration tmax time to mean peak concentration
FETAX Frog Embryo Teratogenesis Assay THF tetrahydrofolate
GABA -aminobutyric acid Tyr tyrosine
GEPR genetically epilepsy-prone rats UAREP Universities Associated for Research
GI gastrointestinal and Education in Pathology
HDL high-density lipoprotein USDA United States Department of
HVA homovanillic acid Agriculture
IgE immunoglobulin E VAMS Visual analogue mood scales
ip intraperitoneal VCR video cassette recorder
JECFA Joint FAO/WHO Expert Committee on WHO World Health Organization
Food Additives Wt Weight
S93