BMC Psychiatry BioMed Central

Research article Open Access

Five-Factor Model personality profiles of drug users
Antonio Terracciano*1, Corinna E Löckenhoff1, Rosa M Crum2, O
Joseph Bienvenu2 and Paul T Costa Jr1,2

Address: 1National Institute on Aging, NIH, DHHS, Baltimore, USA and 2Johns Hopkins Bloomberg School of Public Health, Baltimore, USA
Email: Antonio Terracciano* - [email protected]; Corinna E Löckenhoff - [email protected];
Rosa M Crum - [email protected]; O Joseph Bienvenu - [email protected]; Paul T Costa - [email protected]
* Corresponding author

Published: 11 April 2008 Received: 9 November 2007

Accepted: 11 April 2008
BMC Psychiatry 2008, 8:22 doi:10.1186/1471-244X-8-22
This article is available from: http://www.biomedcentral.com/1471-244X/8/22
© 2008 Terracciano et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract
Background: Personality traits are considered risk factors for drug use, and, in turn, the
psychoactive substances impact individuals' traits. Furthermore, there is increasing interest in
developing treatment approaches that match an individual's personality profile. To advance our
knowledge of the role of individual differences in drug use, the present study compares the
personality profile of tobacco, marijuana, cocaine, and heroin users and non-users using the wide
spectrum Five-Factor Model (FFM) of personality in a diverse community sample.
Method: Participants (N = 1,102; mean age = 57) were part of the Epidemiologic Catchment Area
(ECA) program in Baltimore, MD, USA. The sample was drawn from a community with a wide
range of socio-economic conditions. Personality traits were assessed with the Revised NEO
Personality Inventory (NEO-PI-R), and psychoactive substance use was assessed with systematic
interview.
Results: Compared to never smokers, current cigarette smokers score lower on
Conscientiousness and higher on Neuroticism. Similar, but more extreme, is the profile of cocaine/
heroin users, which score very high on Neuroticism, especially Vulnerability, and very low on
Conscientiousness, particularly Competence, Achievement-Striving, and Deliberation. By contrast,
marijuana users score high on Openness to Experience, average on Neuroticism, but low on
Agreeableness and Conscientiousness.
Conclusion: In addition to confirming high levels of negative affect and impulsive traits, this study
highlights the links between drug use and low Conscientiousness. These links provide insight into
the etiology of drug use and have implications for public health interventions.

Background psychoactive substances, most of them illicit drugs, is also

Drug use is related to adverse health and social outcomes associated with massive social cost beyond the damage to
[1]. Cigarette smoking is a leading cause of preventable the individual users, affecting health care, law enforce-
disability and death in the U.S. and around the world [2], ment, and legal systems [1,5].
increasing the risk of cancer, cardiovascular disease, respi-
ratory and other health problems [3,4]. The use of other

Page 1 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

The high individual and social costs of drug use highlight egorizing traits into "negative affect" (e.g., depression,
the need to study factors related to such behaviors. Even if anxiety), "emotionality" (e.g., extraversion, social disinhi-
personality differences between drug users and non-users bition), and "unconventionality" (e.g., tolerance of devi-
are generally small, these effects can have important clin- ance, non-religiosity). These measures map loosely onto
ical implications due to the large number of people Neuroticism, Extraversion, and Openness, respectively.
involved. Research on the correlates of drug addiction Results suggested that marijuana use was related to high
provides insights for understanding etiology and inform levels of unconventionality, and only weakly to emotion-
prevention policies and cessation programs. For example, ality and negative affect. Another meta-analysis [20]
from a psychiatric perspective, a number of studies have examined the role of a wide range of Conscientiousness-
documented the high comorbidity of drug use with other related measures on health risk behaviors. Across studies,
mental disorders [6-10], which indicate that mood, anxi- a consistent association was found between marijuana use
ety, and personality disorders need to be considered by (as well as other drug use) and low scores on Conscien-
drug treatment specialists to achieve successful interven- tiousness-related traits. Cocaine users are characterized by
tion. The present study contributes to this line of research high scores on Neuroticism-related traits [21,22], such as
by examining the personality traits associated with current depression and impulsivity [23,24], as well as Psychoti-
and lifetime drug use in an economically-diverse commu- cism [21,22], a trait related to low Agreeableness and low
nity sample. Specifically, we attempt to replicate previ- Conscientiousness. Finally, studies of heroin users con-
ously reported associations among personality traits and sistently depict them as high on Neuroticism [25-28].
smoking status and extend the analyses to users of mari- Many studies show an association of heroin use with high
juana, heroin, and cocaine. Comparing personality pro- Extraversion and high Psychoticism, but this association
files, we examine similarities and differences in traits appears to be less robust [26-28]. Inconsistencies in the
associated with a variety of drugs used. Although the high association of personality and drug use are due to several
rate of multiple drug use complicates comparisons across factors, such as differences in the personality measures
substances, results may point to specific traits that under- used, inadequate sample sizes, and socio-cultural differ-
lie the use of a specific drug as well as common factors ences. Most studies use measures that capture only a sub-
across different types of drug use. set of relevant personality traits, and rarely assess all five
major factors and their facets. Studies are also hampered
The present study adopts the well-validated Five-Factor by reliance on small convenience samples. This is particu-
Model of personality [11] which comprehensively covers larly true for studies on the use of illegal drugs. Further,
the five major traits that define human personality across with few exceptions, studies have focused on a single sub-
cultures [12,13]: Neuroticism (N), the tendency to experi- stance at a time, making it difficult to detect common pat-
ence negative emotions such as anxiety and depression; terns across a range of different drugs. This study extends
Extraversion (E), the tendency to be sociable, warm, previous research by examining multiple types of drug use
active, assertive, cheerful, and in search of stimulation; in a large population-based sample while utilizing a well-
Openness to Experience (O), the tendency to be imagina- validated and comprehensive measure of personality that
tive, creative, unconventional, emotionally and artisti- captures both global factors and specific facets of person-
cally sensitive; Agreeableness (A), the dimension of ality.
interpersonal relations, characterized by altruism, trust,
modesty, and cooperativeness; and Conscientiousness Method
(C), a tendency to be organized, strong-willed, persistent, Participants
reliable, and a follower of rules and ethical principles. Participants were drawn from the East Baltimore Epidemi-
Each of these factors is hierarchically defined by specific ologic Catchment Area Study (Baltimore ECA) [29], a
facets, which can provide a more in-depth description of multidisciplinary study which is based on a probability
drug users' personalities. sample of 3,481 East Baltimore residents who were ini-
tially interviewed in 1981 and followed up in 1992–98
In previous studies, cigarette smokers were found to score and in 2004–05. Personality traits were assessed at the
high on facets related to impulsivity and Neuroticism, and two most recent waves. To maximize the sample size,
low on Agreeableness and Conscientiousness [14-16]. cross-sectional analyses were conducted on the last valid
However, in European and Asian studies and some older personality assessments (N = 1,102), 80% of which were
US studies, smokers were also found to score high on obtained in 2004–05. At the time of the personality
Extraversion [16-18]. Compared to cigarette smoking, assessments, age ranged from 30 to 94 years (M = 56.6; SD
there are fewer studies on the personality correlates of ille- = 12.4), and participants had an average of 12.5 years of
gal drug use, and these are based on smaller sample sizes education (SD = 2.6). About 62% of the sample was
and a variety of personality measures. A meta-analysis female; 63% were White/Non-Hispanic, 34% were Black/
[19] examined personality correlates of marijuana use cat- Non-Hispanic, and 3% other or unknown ethnic group.

Page 2 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

To screen out cognitively impaired individuals, partici- current users, about 14% used cocaine daily or almost
pants with Mini Mental State Scores [30] below the cut-off daily, about 43% used it from once or twice a week to
value of 23 were excluded. once or twice a month, and about 43% used it between
once and eleven times a year.
As could be expected, participants who completed the per-
sonality measure at follow-up were those who in 1981 Heroin use was determined by responses to questions on
were younger (35 vs. 54 years old; p < .01) and more edu- whether participants had ever used heroin, even once, and
cated (12.5 vs. 9.7 years of education; p < .01), as com- when they last used it. Of the 1094 individuals without
pared to those who did not complete the personality missing data, about 93% had never tried heroin (n =
assessment (because of mortality, sample loss, or subject 1023), about 6% used it but not in the past-year (n = 62),
refusal). There were no significant differences in the pro- and about 1% used it in the last-year (n = 9). Of the nine
portion of males and females or ethnic groups (Fisher's past-year current users, three individuals used heroin daily
exact test: p > .05). or almost daily, two used it from once or twice a week to
once or twice a month, and four used it between once and
Drug use assessment and prevalence eleven times a year. Given the small number of current
Trained interviewers asked questions about substance use heroin users, and given that eight out of nine current her-
after participants signed the informed consent approved oin users were also current cocaine users, and 87% of the
by the ethics committee. The form assured the confidenti- former heroin users were also former or current cocaine
ality of the answers, and that those were used for research users, we examined the association of personality traits
purposes only. Participation was voluntary, and subjects with cocaine or heroin use. As expected in a population-
could withdraw at any time. Subjects received $20 after based sample, we also found substantial overlap among
the interview. the other drugs used. Of the former and current smokers,
52% had also used marijuana, 21% cocaine, and 8% her-
Cigarette smoking status was determined by responses to oin. About 78% of current and former marijuana users
interview questions, asking participants whether they had had also smoked cigarettes, 40% had used cocaine, and
ever smoked tobacco cigarettes and when they smoked 15% heroin. About 82% of current and former heroin/
their last cigarette. Of the 1,088 participants without miss- cocaine users had also smoked cigarettes and 98% had
ing data, we classified "never smokers" as those who never used marijuana. Demographic information by drug type
smoked (n = 341), "former smokers" as those who is given in Table 1: Across substances, current users were
smoked but not in the last seven days (n = 429), and cur- younger; males were more likely to use marijuana and
rent smokers as those who smoked in the last seven days cocaine/heroin, but there were no sex differences in ciga-
(n = 318). Use ranged from fewer than 10 cigarettes (40%) rette smoking; African-Americans were more likely to use
to over 20 cigarettes a day (16%), with a majority smoking substances; and low education was associated with higher
between 11 and 20 cigarettes a day (44%). use of substances, with the exception of marijuana.

Marijuana use was determined by responses to questions Personality assessment

on whether participants had ever used either marijuana or Participants completed the self-report questionnaire at
hashish, even once, and when they last used marijuana or home or at a participating institution. The Revised NEO
hashish. Of the 1044 individuals without missing data, Personality Inventory [NEO-PI-R, [31]] consists of 240
about 55% of the sample responded that they never tried items answered on a five-point Likert format ranging from
marijuana or hashish (n = 576), 36% used it but not in the strongly disagree to strongly agree. The NEO-PI-R assesses 30
past-year (n = 380), and 8% did use it in the last-year (n = facets, six for each dimension of the FFM (see Table 2 for
88). Among the past-year current users, about 20% used a listing of the 30 facet scales). Raw scores were standard-
marijuana daily or almost daily, about 45% used it from ized as T-scores (M = 50, SD = 10) using combined-sex
once or twice a week to once or twice a month, and about adult norms reported in the Manual [31]. The NEO-PI-R
35% used it between once and eleven times a year. has been translated into several languages and used in
more than 50 cultures [12]. Evidence of convergent and
Cocaine use was determined by responses to questions on discriminant validity is presented in the Manual [31], and
whether participants had ever used cocaine (including all a large literature demonstrates cross-observer agreement
forms of cocaine, such as powder, "crack," freebase, and and prediction of external criteria such as psychological
coca paste) even once, and when they last used any form well-being, health risk behaviors, educational and occu-
of cocaine. Of the 1094 individuals without missing data, pational achievements, coping mechanisms, and longev-
about 82% had never tried cocaine (n = 896), about 16% ity [31,32]. In a previous study [33] we tested the validity
used it but not in the past-year (n = 170), and about 3% of personality assessment in the ECA sample and found
did use it in the last-year (n = 28). Among the past-year

Page 3 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

Table 1: Demographic statistics by drug type.

Group N Age Female African-American Education

Never smoker 341 58 64% 30% 12.9

Former smoker 429 59 60% 30% 12.6
Current smoker 318 52 62% 45% 11.9

Marijuana: never-user 576 62 69% 30% 12.1

Marijuana: former-user 380 50 53% 37% 13.0
Marijuana: current-user 88 48 44% 47% 12.4

Cocaine/heroin: never-user 887 59 65% 32% 12.4

Cocaine/heroin: former-user 178 48 46% 39% 12.9
Cocaine/heroin: current-user 29 46 48% 59% 11.9

Total 1102 57 62% 34% 12.5

Note. Mean age and education expressed in years. For age, ethnicity, and education there were significant differences between never, former, and
current users of cigarettes, marijuana, and cocaine/heroin. Sex differences were seen for marijuana and cocaine/heroin users, but not for cigarette
smoking.

adequate alpha reliabilities, retest-stability, and factor groups are consistent with the findings from previous
structure of the NEO-PI-R scales. studies [14]. Figure 1 plots the full profile of current
smokers against current users of marijuana and cocaine/
Statistical analyses heroin, as well as never user of these substances. The Fig-
All analyses were performed using SPSS 13.0 [34]. For ure presents estimated marginal means after partialling
each drug, we performed a MANCOVA with user status as out the demographic covariates. Former users are
the independent variable, personality factors and facets as excluded.
the dependent variables, and age, sex, education, and eth-
nicity as covariates. Post-hoc comparisons among never, Personality traits and marijuana use
former, and current users groups were based on LSD esti- Compared to never users, current marijuana users scored
mates. The assumption of homogeneity of variance was higher on Openness and lower on Agreeableness and
tested using the Levene statistic, and no large violations Conscientiousness (see Table 3). Former users scored
were found. Effect sizes were estimated using partial η2. intermediate on these three factors. On the facet level, cur-
According to Cohen [35], η2 values of 0.0099, 0.0588 and rent marijuana users scored higher on four facets of Open-
0.1379 correspond to small, medium and large effect ness, particularly Openness to O6:Values and O5:Ideas,
sizes, respectively. and lower on five facets of Conscientiousness, particularly
C3:Dutifulness and C6:Deliberation. Other interesting
Results effects for current marijuana users were the high scores on
Personality traits and smoking status N2:Angry Hostility, N6:Vulnerability, E4:Activity, and
Multivariate analyses of covariance controlling for demo- E5:Excitement Seeking, and the low scores on A4:Compli-
graphic variables indicated significant personality differ- ance, as compared to never users. As depicted in Figure 1,
ences among smoking status groups (see Table 2). the profile of current marijuana users is similar to the pat-
Compared to never smokers, current smokers scored tern found among current smokers, especially for the fac-
higher on Neuroticism and lower on Conscientiousness. ets of Conscientiousness.
Current smokers scored lower on Agreeableness, but this
difference was not significant after controlling for demo- Personality traits and cocaine/heroin use
graphic variables. Former smokers scored intermediate on Compared to never users, current cocaine/heroin users
Neuroticism and Conscientiousness. At the facet level, scored higher on Neuroticism and lower on Conscien-
current smokers were characterized by traits related to the tiousness (see Table 4). Former cocaine/heroin users
construct of impulsivity, (i.e., N5:Impulsiveness, scored lower on Conscientiousness, but did not differ
E5:Excitement-Seeking, C5:Self-Discipline, and C6:Delib- from never-users on Neuroticism. On the facet level, cur-
eration) [36]. Smokers as a group were also high on rent users scored high on all facets of Neuroticism, with
N3:Depression, N6:Vulnerability, and low in A4:Compli- large effect sizes (difference larger than one SD) on
ance, C1:Competence, and C3:Dutifulness. Although the N6:Vulnerability, high on E5:Excitement Seeking, low on
magnitude of the effects is small, the differences among A1:Trust, A2:Straightforwardness, and A4:Compliance,

Page 4 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

Figure 1 profile of current users of tobacco, marijuana, and cocaine/heroin and never users of those substances
NEO-PI-R
NEO-PI-R profile of current users of tobacco, marijuana, and cocaine/heroin and never users of those sub-
stances. Plots show estimated marginal means after controlling for age, sex, ethnicity, and education.

and very low on all facets of Conscientiousness, with dif- ness [8,9,20,37]. Most previous studies have analyzed a
ferences larger than one SD on C1:Competence, limited number of traits (often omitting the crucial Con-
C4:Achievement Striving, and C6:Deliberation. The pro- scientiousness factor) or focused on a single substance,
file of cocaine/heroin current users is illustrated in Figure making it difficult to integrate the body of evidence across
1, which resembles the pattern seen for current smokers, traits or substances. Using a more integrative approach,
but the profile of cocaine/heroin users is more extreme. this study indicates that low Conscientiousness and high
Additional analyses indicated that although the few indi- Neuroticism are consistently associated with tobacco
viduals (n = 9) who used both cocaine and heroin had the smoking, heroin, and cocaine use. Low Conscientiousness
most extreme profile, the individuals who were current is also characteristic of marijuana users, who are average
users of cocaine but not heroin also scored significantly on Neuroticism and high on Openness, a trait that distin-
higher on Neuroticism and lower on Conscientiousness guishes marijuana users from other drug users. The
than never users. present study extends the previous literature by assessing
higher-level personality factors as well as lower-level fac-
Discussion ets. The association among Neuroticism and tobacco/her-
The associations observed in the Baltimore ECA sample oin/cocaine use was found for all six facets of
are consistent with the existing literature, which finds Neuroticism, indicating that multiple aspects of negative
drug users generally high on measures of negative emo- emotionality and psychopathology are involved in this
tionality or psychopathology and low on Conscientious- effect. With the exception of C2:Order, low scores on all

Page 5 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

Table 2: Mean personality traits for never, former, and current cigarette smokers.

NEO-PI-R scales Never-smokers (n = 341) Former-smokers (n = 429) Current-smokers (n = 318) F Partial η2

Neuroticism 48.9 (.51)a 50.7 (.46)b 51.6 (.55)b 6.9** .013

Extraversion 48.8 (.44) 49.3 (.40) 48.4 (.48) 1.0 .002
Openness 45.6 (.44) 46.4 (.39) 46.5 (.47) 1.2 .002
Agreeableness 50.9 (.49) 50.9 (.44) 49.9 (.52) 1.3 .002
Conscientiousness 49.4 (.52)a 49.0 (.46)a 47.0 (.55)b 5.5** .010

N1: Anxiety 49.7 (.46)a 51.0 (.41)b 51.7 (.50)b 4.7** .009
N2: Angry Hostility 49.5 (.50)a 50.7 (.45)a 52.3 (.54)b 6.9** .013
N3: Depression 49.3 (.57)a 50.8 (.51)a 53.1 (.61)b 10.0** .019
N4: Self-consciousness 48.4 (.52)a 50.1 (.46)b 50.0 (.56)b 3.6* .007
N5: Impulsiveness 48.3 (.48)a 49.8 (.43)b 49.7 (.52)b 3.4* .007
N6: Vulnerability 49.3 (.58)a 50.4 (.52)a 52.7 (.63)b 7.9** .015
E1: Warmth 48.9 (.52) 48.4 (.46) 47.2 (.56) 2.4 .005
E2: Gregariousness 50.5 (.50) 51.0 (.45) 49.7 (.54) 1.7 .003
E3: Assertiveness 50.1 (.51) 49.6 (.46) 49.2 (.55) 0.7 .001
E4: Activity 48.1 (.49) 47.3 (.43) 46.9 (.52) 1.5 .003
E5: Excitement-Seeking 46.9 (.47)a 47.7 (.42)a 49.0 (.50)b 4.4* .008
E6: Positive Emotions 48.9 (.51)a 48.9 (.46)a 47.0 (.55)b 3.9* .007
O1: Fantasy 47.8 (.44) 48.5 (.39) 48.3 (.47) 0.7 .001
O2: Aesthetics 48.3 (.48) 49.4 (.43) 49.4 (.52) 1.7 .003
O3: Feelings 46.7 (.46) 47.6 (.41) 47.0 (.50) 1.2 .002
O4: Actions 45.2 (.50) 46.1 (.45) 46.0 (.54) 1.0 .002
O5: Ideas 46.5 (.48) 46.9 (.43) 48.1 (.52) 2.7 .005
O6: Values 45.0 (.45) 45.6 (.40) 44.6 (.48) 1.4 .003
A1: Trust 47.2 (.53) 46.6 (.47) 46.2 (.57) 0.9 .002
A2: Straightforwardness 50.7 (.50) 51.0 (.44) 50.2 (.53) 0.7 .001
A3: Altruism 50.2 (.52)a 51.6 (.46)b 49.9 (.55)a 3.5* .007
A4: Compliance 50.5 (.55)a 48.8 (.49)b 48.4 (.59)b 3.8* .007
A5: Modesty 50.9 (.50) 51.5 (.45) 51.1 (.54) 0.3 .001
A6: Tender-mindedness 52.0 (.50) 52.4 (.44) 52.2 (.53) 0.2 .000
C1: Competence 50.0 (.56)a 49.9 (.50)a 46.7 (.60)b 10.2** .019
C2: Order 47.5 (.47) 47.0 (.42) 46.8 (.51) 0.5 .001
C3: Dutifulness 49.0 (.51)a 48.5 (.45)a 45.9 (.55)b 9.0** .017
C4: Achievement Striving 49.0 (.55)a 48.4 (.49)a 46.4 (.58)b 5.4** .010
C5: Self-Discipline 48.6 (.55)a 47.7 (.49)a 46.3 (.59)b 3.8* .007
C6: Deliberation 53.5 (.53)a 52.6 (.48)a 50.0 (.57)b 10.1** .019

Note. Estimated marginal means (Standard Errors), statistical tests, and effect sizes computed after controlling for age, sex, ethnicity (black, white),
and education (n = 1058).
Means with different superscripts differ significantly at p < .05.
* p < .05; ** p < .01

facets of Conscientiousness were associated with drug use. Cross-sectional association analyses provide limited input
Although Extraversion showed no association with drug on the cause and effect relation between personality traits
use on the factor level, facet-level analyses revealed a con- and drug use. Although individual differences in person-
sistent association between high scores on E5:Excitement- ality traits are particularly stable in adulthood [38-40],
Seeking and all types of drug use. This finding is not sur- some evidence suggests that substance use influences per-
prising given that, together with N5:Impulsiveness (ina- sonality-related variables [41]. Cigarette smoking contrib-
bility to resist cravings), C5:Self-Discipline (limited utes to stress, negative affect states, and the onset of
ability to stay on task), and C6:Deliberation (lack of care- clinical correlates of Neuroticism, such as anxiety and
ful consideration of the consequences of one's actions), depressive disorders [42-44]. Piedmont [45] reported sub-
the E5:Excitement-Seeking facet is an aspect of impulsivity stantial declines in Neuroticism and increases in Agreea-
[36]. The selective association between drug use and this bleness and Conscientiousness in a group of
specific Extraversion facet also suggests that inconsistent polysubstance abusers following a rehabilitation program
findings for Extraversion may be due to Extraversion [26] (but see [28]). Consistently, smoking cessation is
measures that differ in their relative emphasis on the associated with lower Neuroticism scores, a lower level of
excitement-seeking component. stress, and lower risk of anxiety disorders [42,44]. Another

Page 6 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

Table 3: Mean personality traits for never, non-current, and current marijuana users.

NEO-PI-R scales Never-user (n = 576) Former user (n = 380) Current user (n = 88) F Partial η2

Neuroticism 50.0 (.42) 50.4 (.50) 50.9 (1.01) 0.4 .001

Extraversion 49.0 (.37) 48.5 (.44) 49.4 (.88) 0.7 .001
Openness 45.5 (.37)a 46.1 (.44)a 48.8 (.88)b 5.5** .011
Agreeableness 51.6 (.40)a 49.6 (.49)b 48.4 (.98)b 6.3** .012
Conscientiousness 49.7 (.43)a 47.0 (.51)b 47.0 (1.03)b 8.0** .016

N1: Anxiety 50.5 (.38) 50.8 (.46) 50.6 (.92) 0.1 .000
N2: Angry Hostility 50.0 (.41)a 51.2 (.50)ab 53.0 (1.00)b 3.9* .008
N3: Depression 50.4 (.47) 51.4 (.57) 51.2 (1.14) 0.7 .001
N4: Self-consciousness 49.2 (.43) 50.0 (.51) 49.1 (1.03) 0.7 .001
N5: Impulsiveness 48.7 (.40) 49.7 (.48) 50.1 (.96) 1.4 .003
N6: Vulnerability 49.6 (.47)a 51.5 (.57)b 52.2 (1.15)b 3.7* .007
E1: Warmth 49.1 (.43)a 47.1 (.52)b 47.1 (1.03)ab 4.2* .008
E2: Gregariousness 50.8 (.41) 50.1 (.50) 50.4 (1.00) 0.5 .001
E3: Assertiveness 49.7 (.42) 49.4 (.51) 50.7 (1.03) 0.7 .001
E4: Activity 47.4 (.40)a 46.9 (.48)a 50.1 (.97)b 4.6* .009
E5: Excitement-Seeking 47.1 (.39)a 48.5 (.47)b 49.9 (.94)b 4.3* .009
E6: Positive Emotions 49.0 (.42) 47.4 (.51) 48.4 (1.02) 2.3 .005
O1: Fantasy 47.9 (.36) 48.5 (.44) 48.9 (.87) 0.8 .001
O2: Aesthetics 48.5 (.40)a 49.0 (.48)a 51.4 (.96)b 3.9* .008
O3: Feelings 46.6 (.38) 47.3 (.46) 48.6 (.92) 1.9 .004
O4: Actions 45.2 (.42)a 45.9 (.50)a 48.1 (1.01)b 3.2* .006
O5: Ideas 47.2 (.40)a 46.2 (.49)a 49.7 (.97)b 5.7** .011
O6: Values 44.0 (.37)a 46.1 (.44)b 46.9 (.89)b 7.9** .015
A1: Trust 47.3 (.43) 46.5 (.52) 44.7 (1.04) 2.7 .005
A2: Straightforwardness 51.4 (.41) 50.0 (.49) 49.2 (.99) 2.9 .006
A3: Altruism 51.4 (.42)a 49.4 (.51)b 50.4 (1.03)ab 3.8* .008
A4: Compliance 50.4 (.45)a 48.2 (.55)b 47.2 (1.10)b 5.4** .011
A5: Modesty 51.9 (.42) 50.7 (.50) 49.5 (1.01) 2.8 .006
A6: Tender-mindedness 52.5 (.41) 51.6 (.50) 52.2 (.99) 1.0 .002
C1: Competence 50.0 (.47)a 47.7 (.56)b 47.1 (1.13)b 5.0** .010
C2: Order 47.6 (.39) 46.5 (.48) 45.9 (.95) 2.1 .004
C3: Dutifulness 49.2 (.42)a 46.2 (.51)b 46.2 (1.01)b 9.8** .019
C4: Achievement Striving 49.0 (.45)a 46.6 (.55)b 46.1 (1.09)b 6.0** .012
C5: Self-Discipline 48.7 (.46)a 46.5 (.55)b 45.6 (1.10)b 5.3** .010
C6: Deliberation 53.2 (.44)a 50.9 (.54)b 49.9 (1.07)b 6.6** .013

Note. Estimated marginal means (Standard Errors), statistical tests, and effect sizes computed after controlling for age, sex, ethnicity (black, white),
and education (n = 1013).
Means with different superscripts differ significantly at p < .05.
* p < .05; ** p < .01

set of evidence suggests that personality traits are risk fac- Limitations
tors for psychoactive substance use, along with social There are several limitations to consider when interpret-
environment and life experiences [46,47]. For example, in ing the results. This sample is not representative of the
long term longitudinal studies, low Conscientiousness in entire US population, but it was drawn from a probability
childhood predicts cigarette smoking in adulthood sample that included a wide range of socio-economic con-
[48,49]. Longitudinal studies in Europe also suggest that ditions. There may be some misclassification with the cat-
high scores on Neuroticism and Extraversion during ado- egories of never, former, and current-users. For example,
lescence increase the likelihood of being a smoker later in some individuals might be reluctant to disclose their illicit
life [18,50]. A common hypothesis is that individuals drug use. Some might not recall use in the distant past.
with high Neuroticism use drugs to self-medicate [51,52]. Categorizing current users based on self-reported behav-
Finally, third variables might be responsible for the asso- ior during the past year might be too broad. There are
ciation of personality and addictive behaviors. For exam- marked differences in the frequency and quantity of drug
ple, personality traits and cigarette smoking are both use, but the relatively small number of users in the present
highly heritable [53,54], and could be influenced by com- sample does not allow finer distinctions or the use of
mon genetic factors [37]. stricter criteria of addiction. However, preliminary analy-

Page 7 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

Table 4: Mean personality traits for never, non-current, and current cocaine/heroin users.

NEO-PI-R scales Never-user (n = 887) Former user (n = 178) Current user (n = 29) F Partial η2

Neuroticism 50.3 (.32)a 49.7 (.73)a 57.6 (1.77)b 8.8** .016

Extraversion 48.5 (.28) 50.0 (.63) 50.7 (1.54) 2.7 .005
Openness 46.1 (.28) 46.7 (.63) 45.1 (1.54) 0.7 .001
Agreeableness 50.8 (.31) 49.9 (.69) 49.2 (1.69) 1.1 .002
Conscientiousness 49.2 (.32)a 46.8 (.73)b 40.3 (1.77)c 14.7** .027

N1: Anxiety 51.0 (.29)a 49.5 (.66)b 54.9 (1.62)c 5.6** .010
N2: Angry Hostility 50.4 (.31)a 51.4 (.71)a 57.9 (1.73)b 9.1** .017
N3: Depression 50.8 (.36)a 51.0 (.82)a 58.5 (1.99)b 7.2** .013
N4: Self-consciousness 49.6 (.32)a 48.5 (.74)a 54.4 (1.80)b 4.9** .009
N5: Impulsiveness 49.0 (.30)a 50.0 (.69)a 54.0 (1.68)b 4.7** .009
N6: Vulnerability 50.5 (.36)a 50.2 (.82)a 61.3 (2.01)b 14.4** .027
E1: Warmth 48.1 (.32) 48.7 (.74) 49.3 (1.80) 0.4 .001
E2: Gregariousness 50.3 (.31) 51.2 (.71) 49.4 (1.74) 0.9 .002
E3: Assertiveness 49.6 (.32) 50.0 (.73) 47.3 (1.78) 1.0 .002
E4: Activity 47.3 (.30) 48.5 (.69) 46.2 (1.69) 1.7 .003
E5: Excitement-Seeking 47.2 (.29)a 50.0 (.66)b 51.0 (1.62)b 8.7** .016
E6: Positive Emotions 48.4 (.32) 48.0 (.73) 45.7 (1.79) 1.2 .002
O1: Fantasy 48.1 (.28) 48.7 (.63) 48.9 (1.53) 0.5 .001
O2: Aesthetics 48.9 (.30) 49.6 (.69) 50.6 (1.68) 0.8 .002
O3: Feelings 47.0 (.29) 47.3 (.66) 48.7 (1.61) 0.5 .001
O4: Actions 45.5 (.31)a 47.5 (.71)b 42.3 (1.74)a 5.4** .010
O5: Ideas 47.2 (.31) 46.8 (.69) 47.6 (1.69) 0.2 .000
O6: Values 44.7 (.28)a 46.9 (.64)b 45.3 (1.55)ab 4.8** .009
A1: Trust 47.1 (.33)a 45.4 (.74)b 42.6 (1.82)b 4.5* .008
A2: Straightforwardness 50.9 (.31)a 50.2 (.70)a 46.5 (1.72)b 3.3* .006
A3: Altruism 50.6 (.32) 50.7 (.73) 52.1 (1.79) 0.4 .001
A4: Compliance 49.4 (.34)a 49.0 (.78) 45.3 (1.91) 2.3 .004
A5: Modesty 51.2 (.32) 51.2 (.72) 52.2 (1.75) 0.2 .000
A6: Tender-mindedness 52.2 (.31) 52.5 (.71) 52.1 (1.72) 0.1 .000
C1: Competence 49.7 (.35)a 47.4 (.79)b 37.0 (1.92)c 22.3** .041
C2: Order 47.2 (.29) 47.1 (.67) 44.3 (1.63) 1.5 .003
C3: Dutifulness 48.3 (.32)a 46.4 (.73)b 42.5 (1.77)c 7.2** .014
C4: Achievement Striving 48.5 (.34)a 46.7 (.77)b 38.1 (1.87)c 15.9** .029
C5: Self-Discipline 48.1 (.34)a 46.1 (.78)b 39.0 (1.91)c 12.6** .023
C6: Deliberation 52.7 (.33)a 51.0 (.75)a 42.1 (1.84)b 16.6** .031

Note. Estimated marginal means (Standard Errors), statistical tests, and effect sizes computed after controlling for age, sex, ethnicity (black, white),
and education (n = 1062).
Means with different superscripts differ significantly at p < .05.
* p < .05; ** p < .01

ses using different classification criteria produced similar use early in life, which may introduce attrition and other
results, and the main findings are mostly consistent with biases. A review of the literature [20] suggested that stud-
the literature. In addition to self-report ratings, future ies that involve older populations (over the age of 30
studies should use multiple methods for assessing drug years) report weaker association of Conscientiousness-
use and personality traits. related traits and drug use.

Most studies on drug use are conducted in adolescents Finally, some of the findings may be culture-bound [55].
and young adults, who are at life stages associated with For example, the results for smoking closely replicate the
the greater use of drugs. We presented data from an older findings we previously reported from another US cohort
cohort, which has presumably passed the experimenta- of similar age but different socio-economic status. How-
tion age. While this contributes to the scarce literature on ever, studies conducted in Europe [50] and Japan [17]
drug use in later parts of the lifespan, the advantages of a have found Extraversion associated with cigarette smok-
lifespan perspective come at the cost of having fewer cur- ing [16]. Such differences might reflect the different social
rent users in this older cohort. In addition, older cohorts acceptance of smoking across countries.
only include the survivors among those who started drug

Page 8 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

Conclusions: Clinical and social policy References

implications 1. Nutt D, King LA, Saulsbury W, Blakemore C: Development of a
rational scale to assess the harm of drugs of potential mis-
Personality traits are associated with the outcome of ther- use. Lancet 2007, 369(9566):1047-1053.
apeutic interventions. For example, several studies found 2. World Health Organization: The World Health Report 2002 --
Neuroticism, anxiety, and depressive disorders related to Reducing risks, promoting healthy life. Geneva , World Health
Organization; 2002.
poor treatment outcome for nicotine dependence [56]. 3. World Health Organization: The facts about smoking and
Although we found systematic differences between the health. [http://www.wpro.who.int/media_centre/fact_sheets/
personality profiles of substance users and non-users, fs_20020528.htm].
4. U.S. Department of Health and Human Services: The health conse-
there is substantial variability in both groups (e.g., not all quences of smoking: A report of the Surgeon General. Wash-
smokers score high on Neuroticism or low on Conscien- ington , U.S. Department of Health and Human Services, Centers for
Disease Control and Prevention, National Center for Chronic Dis-
tiousness). Individual differences among substance abus- ease Prevention and Health Promotion, Office on Smoking and
ers can play an important role in the choice of treatment Health; 2004.
options [57]. Recently, more attention has been focused 5. Reuter P: What drug policies cost: estimating government
drug policy expenditures. Addiction 2006, 101(3):315-322.
on personality trait effects on the efficacy of different treat- 6. Grant BF, Stinson FS, Dawson DA, Chou SP, Dufour MC, Compton
ment plans [58] to tailor therapeutic interventions to indi- W, Pickering RP, Kaplan K: Prevalence and co-occurrence of
substance use disorders and independent mood and anxiety
vidual needs [59,60]. More research is needed to fully disorders: results from the National Epidemiologic Survey
evaluate how personality assessment can be useful in the on Alcohol and Related Conditions. Arch Gen Psychiatry 2004,
choice of treatment plans. 61(8):807-816.
7. Kessler RC, Chiu WT, Demler O, Merikangas KR, Walters EE: Prev-
alence, severity, and comorbidity of 12-month DSM-IV disor-
Although individual treatments might reduce the rate of ders in the National Comorbidity Survey Replication. Arch
drug abuse, public policy is an important tool for cigarette Gen Psychiatry 2005, 62(6):617-627.
8. Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, Goodwin
smoking and other drug abuse prevention and cessation. FK: Comorbidity of mental disorders with alcohol and other
Because of the low conscientiousness, high impulsivity, drug abuse. Results from the Epidemiologic Catchment
Area (ECA) Study. Jama 1990, 264(19):2511-2518.
and high emotional vulnerability of most drug users, rely- 9. Lasser K, Boyd JW, Woolhandler S, Himmelstein DU, McCormick D,
ing on an individual's resources, without therapeutic Bor DH: Smoking and mental illness: A population-based
intervention, may produce limited results. Evidence- prevalence study. JAMA 2000, 284(20):2606-2610.
10. Bakken K, Landheim AS, Vaglum P: Axis I and II disorders as long-
based interventions such as safer injecting environments term predictors of mental distress: a six-year prospective
are an important adjunct which can reduce drug-related follow-up of substance-dependent patients. BMC Psychiatry
harm [61]. In the case of cigarette smoking, societal pres- 2007, 7:29.
11. McCrae RR, John OP: An introduction to the Five-Factor Model
sure in the form of high taxation, restriction in advertis- and its applications. J Pers 1992, 60(2):175-215.
ing, and interdiction of smoking in public places are cost- 12. McCrae RR, Terracciano A, 78 Members of the Personality Profiles of
Cultures Project: Universal features of personality traits from
effective programs that are reducing the prevalence of the observer's perspective: Data from 50 cultures. J Pers Soc
smoking [62]. Psychol 2005, 88:547-561.
13. Terracciano A, McCrae RR: Cross-cultural studies of personality
traits and their relevance to psychiatry. Epidemiol Psichiatr Soc
Competing interests 2006, 15:176-184 [http://www.psychiatry.univr.it/page_eps/docs/
Paul T. Costa, Jr. receives royalties from the Revised NEO 2006_3_Terracciano.pdf].
14. Terracciano A, Costa PT Jr.: Smoking and the Five-Factor Model
Personality Inventory. The authors declare that they have of personality. Addiction 2004, 99:472-481.
no other competing interests. 15. Paunonen SV, Ashton MC: Big five factors and facets and the
prediction of behavior. J Pers Soc Psychol 2001, 81(3):524-539.
16. Malouff JM, Thorsteinsson EB, Schutte NS: The five-factor model
Authors' contributions of personality and smoking: a meta-analysis. J Drug Educ 2006,
The ECA is an ongoing longitudinal study in which PTC, 36(1):47-58.
OJB, and RMC are active research members that partici- 17. Arai Y, Hosokawa T, Fukao A, Izumi Y, Hisamichi S: Smoking
behaviour and personality: A population-based study in
pate in its conception, design, and coordination. AT con- Japan. Addiction 1997, 92(8):1023-1033.
ceived the current manuscript, performed the statistical 18. Munafo MR, Black S: Personality and smoking status: A longitu-
dinal analysis. Nicotine Tob Res 2007, 9(3):397-404.
analysis, and drafted the manuscript. CEL, PTC, RMC, and 19. Gorman DM, Derzon JH: Behavioral traits and marijuana use
OJB contributed to the conception and draft of the manu- and abuse: a meta-analysis of longitudinal studies. Addict Behav
script. PTC coordinated the work for the manuscript. All 2002, 27(2):193-206.
20. Bogg T, Roberts BW: Conscientiousness and health-related
authors read and approved the final manuscript. behaviors: a meta-analysis of the leading behavioral contrib-
utors to mortality. Psychol Bull 2004, 130(6):887-919.
21. Saiz PA, Gonzalez MP, Paredes B, Martinez S, Delgado JM: Personal-
Acknowledgements ity and use-abuse of cocaine. Addiciones 2001, 13(3):47-59.
This research was supported in part by the Intramural Research Program 22. Kilbey MM, Breslau N, Andreski P: Cocaine use and dependence
of the NIH, National Institute on Aging, and by NIH grants MH47447, in young adults: associated psychiatric disorders and person-
MH50616, and K23-MH64543. ality traits. Drug Alcohol Depend 1992, 29(3):283-290.
23. Ball SA, Schottenfeld RS: A five-factor model of personality and
addiction, psychiatric, and AIDS risk severity in pregnant
and postpartum cocaine misusers. Subst Use Misuse 1997,
32(1):25-41.

Page 9 of 10
(page number not for citation purposes)
BMC Psychiatry 2008, 8:22 http://www.biomedcentral.com/1471-244X/8/22

24. Rosenthal TL, Edwards NB, Ackerman BJ, Knott DH, Rosenthal RH: 47. Kodl MM, Mermelstein R: Beyond modeling: parenting prac-
Substance abuse patterns reveal contrasting personal traits. tices, parental smoking history, and adolescent cigarette
J Subst Abuse 1990, 2(2):255-263. smoking. Addict Behav 2004, 29(1):17-32.
25. Brooner RK, Schmidt CW, Herbst JH: Personality trait charac- 48. Kubicka L, Matejcek Z, Dytrych Z, Roth Z: IQ and personality
teristics of opioid abusers with and without comorbid per- traits assessed in childhood as predictors of drinking and
sonality disorders. In Personality disorders and the Five-Factor Model smoking behaviour in middle-aged adults: a 24- year follow-
of personality 2nd edition. Edited by: Costa PTJ, Widiger TA. Washing- up study. Addiction 2001, 96(11):1615-1628.
ton, DC , American Psychological Association; 2002:249-268. 49. Hampson SE, Goldberg LR, Vogt TM, Dubanoski JP: Forty years on:
26. Tremeau F, Darreye A, Leroy B, Renckly V, Ertle S, Weibel H, teachers' assessments of children's personality traits predict
Khidichian F, Macher JP: [Personality changes in opioid-depend- self-reported health behaviors and outcomes at midlife.
ent subjects in a methadone maintenance treatment pro- Health Psychol 2006, 25(1):57-64.
gram]. Encephale 2003, 29(4 Pt 1):285-292. 50. Harakeh Z, Scholte RH, de Vries H, Engels RC: Association
27. Blaszczynski AP, Buhrich N, McConaghy N: Pathological gam- between personality and adolescent smoking. Addict Behav
blers, heroin addicts and controls compared on the E.P.Q. 2006, 31(2):232-245.
'Addiction Scale'. Br J Addict 1985, 80(3):315-319. 51. McClernon FJ, Hiott FB, Westman EC, Rose JE, Levin ED: Transder-
28. Kornor H, Nordvik H: Five-factor model personality traits in mal nicotine attenuates depression symptoms in nonsmok-
opioid dependence. BMC Psychiatry 2007, 7:37. ers: a double-blind, placebo-controlled trial.
29. Eaton WW, Anthony JC, Gallo J, Cai G, Tien A, Romanoski A, Lyket- Psychopharmacology (Berl) 2006, 189(1):125-133.
sos C, Chen LS: Natural history of Diagnostic Interview Sched- 52. Breslau N, Peterson EL, Schultz LR, Chilcoat HD, Andreski P: Major
ule/DSM-IV major depression. The Baltimore Epidemiologic depression and stages of smoking. A longitudinal investiga-
Catchment Area follow-up. Archives of General Psychiatry 1997, tion. Arch Gen Psychiatry 1998, 55(2):161-166.
54:993-999. 53. Maes HH, Sullivan PF, Bulik CM, Neale MC, Prescott CA, Eaves LJ,
30. Folstein M, Folstein S, McHugh P: Mini-Mental State. A practical Kendler KS: A twin study of genetic and environmental influ-
method for grading the cognitive state of patients for the cli- ences on tobacco initiation, regular tobacco use and nicotine
nician. Journal of Psychiatric Research 1975, 12:189-198. dependence. Psychol Med 2004, 34(7):1251-1261.
31. Costa PT Jr., McCrae RR: Revised NEO Personality Inventory 54. Pilia G, Chen WM, Scuteri A, Orrú M, Albai G, Dei M, Lai S, Usala L,
(NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) pro- Lai M, Loi P, Mameli C, Vacca L, Deiana M, Masala M, Cao A, Najjar
fessional manual. Odessa, FL , Psychological Assessment SS, Terracciano A, Nedorezov T, Sharov A, Zonderman AB, Abecasis
Resources; 1992. G, Costa PT, Lakatta E, Schlessinger D: Heritability of Cardiovas-
32. Terracciano A, Löckenhoff CE, Zonderman AB, Ferrucci L, Costa PT cular and Personality Traits in 6,148 Sardinians. PloS Genetics
Jr: Personality predictors of longevity: Activity, Emotional 2006, 2:e132.
Stability, and Conscientiousness. Psychosomatic Medicine 2008. 55. Hughes JR, Giovino GA, Klevens RM, Fiore MC: Assessing the gen-
33. Löckenhoff CE, Terracciano A, Bienvenu OJ, Patriciu NS, Nestadt G, eralizability of smoking studies. Addiction 1997, 92(4):469-472.
McCrae RR, Eaton WW, Costa PT Jr: Ethnicity, education, and 56. Hooten WM, Ames SC, Vickers KS, Hays JT, Wolter TD, Hurt RD,
the temporal stability of personality traits in the East Balti- Offord KP: Personality correlates related to tobacco absti-
more Epidemiologic Catchment Area Study. Journal of nence following treatment. Int J Psychiatry Med 2005,
Research in Personality in press. 35(1):59-74.
34. SPSS: SPSS 13.0 for Windows edition. Chicago, SPSS Inc; 2004. 57. Brooner RK, Kidorf MS, King VL, Stoller KB, Neufeld KJ, Kolodner K:
35. Cohen J: Statistical power analysis for the behavioral sciences. Comparing adaptive stepped care and monetary-based
second edition. Hillsdale, New Jersey , Lawrence Erlbaum Associates; voucher interventions for opioid dependence. Drug Alcohol
1988. Depend 2007, 88( Suppl 2):S14-23.
36. Whiteside SP, Lynam DR: The Five Factor Model and impulsiv- 58. Berlin I, Covey LS: Pre-cessation depressive mood predicts fail-
ity: using a structural model of personality to understand ure to quit smoking: the role of coping and personality traits.
impulsivity. Pers Individ Differ 2001, 30(4):669-689. Addiction 2006, 101(12):1814-1821.
37. Gilbert DG, Gilbert BO: Personality, psychopathology, and nic- 59. Staiger PK, Kambouropoulos N, Dawe S: Should personality traits
otine response as mediators of the genetics of smoking. be considered when refining substance misuse treatment
Behav Genet 1995, 25(2):133-147. programs? Drug Alcohol Rev 2007, 26(1):17-23.
38. Terracciano A, Costa PT Jr, McCrae RR: Personality Plasticity 60. Miller T: The psychotherapeutic utility of the five-factor
After Age 30. Pers Soc Psychol Bull 2006, 32:999-1009. model of personality: A clinician's experience. Journal of Person-
39. Terracciano A, McCrae RR, Costa PT Jr: Longitudinal trajectories ality Assessment 1991, 57:415-433.
in Guilford-Zimmerman Temperament Survey data in the 61. Rhodes T, Kimber J, Small W, Fitzgerald J, Kerr T, Hickman M, Hol-
Baltimore Longitudinal Study of Aging. J Gerontol B Psychol Sci loway G: Public injecting and the need for 'safer environment
Soc Sci 2006, 61(2):P108-P116. interventions' in the reduction of drug-related harm. Addic-
40. Terracciano A, McCrae RR, Brant LJ, Costa PT Jr: Hierarchical lin- tion 2006, 101(10):1384-1393.
ear modeling analyses of NEO-PI-R scales in the Baltimore 62. Frieden TR, Mostashari F, Kerker BD, Miller N, Hajat A, Frankel M:
Longitudinal Study of Aging. Psychol Aging 2005, 20:493-506. Adult tobacco use levels after intensive tobacco control
41. Stein JA, Newcomb MD, Bentler PM: Personality and drug use: measures: New York City, 2002-2003. Am J Public Health 2005,
Reciprocal effects across four years. Personality and Individual Dif- 95(6):1016-1023.
ferences 1987, 8:419-430.
42. Parrott AC: Nesbitt's paradox resolved? Stress and arousal
modulation during cigarette smoking. Addiction 1998, Pre-publication history
93(1):27-39. The pre-publication history for this paper can be accessed
43. Picciotto MR, Brunzell DH, Caldarone BJ: Effect of nicotine and here:
nicotinic receptors on anxiety and depression. Neuroreport
2002, 13(9):1097-1106.
44. Breslau N, Novak SP, Kessler RC: Daily smoking and the subse- http://www.biomedcentral.com/1471-244X/8/22/pre
quent onset of psychiatric disorders. Psychol Med 2004, pub
34(2):323-333.
45. Piedmont RL: Cracking the plaster cast: Big Five personality
change during intensive outpatient counseling. Journal of
Research in Personality 2001, 35:500-520.
46. Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards
V, Koss MP, Marks JS: Relationship of childhood abuse and
household dysfunction to many of the leading causes of
death in adults. The Adverse Childhood Experiences (ACE)
Study. Am J Prev Med 1998, 14(4):245-258.

Page 10 of 10
(page number not for citation purposes)