DOI: 10.1590/1413-81232015212.

00022015 573

Association between untreated dental caries

temas livres free themes

and household food insecurity in schoolchildren

Associação entre cárie dentária não tratada

e insegurança alimentar em escolares

Gabriela Cristina Santin 1

Tatiana Pegoretti Pintarelli 1
Fabian Calixto Fraiz 1
Ana Cristina Borges de Oliveira 2
Saul Martins Paiva 3
Fernanda Morais Ferreira 1

Abstract The aim of the present study was to Resumo O objetivo deste estudo foi avaliar a as-
assess the association between untreated den- sociação entre cárie dentária não tratada (CDNT)
tal caries (UDC) and household food insecurity e insegurança alimentar (IA) em escolares de dife-
(HFI) among schoolchildren in different income rentes estratos de renda. Foi realizado um estudo
strata. A population-based study was carried out de base populacional com uma amostra de 584 es-
with a sample of 584 12-y-old schoolchildren. colares de 12 anos de idade. Exames clínicos bucais
Oral examinations were performed and HFI was foram realizados e IA foi determinada por meio
determined using a validated scale. Other inde- de uma escala validada. Outras variáveis foram
pendent variables were analyzed for being of in- analisadas por serem de interesse para a estratifi-
terest to the stratification of the results (per capita cação dos resultados (renda domiciliar per capita)
household income) or for acting as potential con- ou como variáveis de confusão. A prevalência de
founding variables. The prevalence of UDC and CDNT e IA foi de 45% e 39%, respectivamente.
HFI was 45% and 39%, respectively. The multi- Os modelos multivariados demonstraram que
variate models demonstrated that the UDC was a CDNT foi significativamente mais prevalente
significantly more prevalent among children in entre as crianças com IA alimentar e renda per
food-insecure households with per capita income capita de até US$ 70,71 do que entre aqueles no
of up to US$ 70.71 than among those in the same mesmo estrato de renda que estavam livres de IA
income stratum that were free of HFI [PR = 1.52 [RP = 1,52 (IC 95% = 1,01 -2,29)]. A IA foi as-
(95%CI = 1.01-2.29)]. HFI was associated with sociada com uma maior frequência de CDNT em
1
Departamento de a greater frequency of UDC among low-income escolares de baixa renda, porém não teve impacto
Estomatologia, Universidade
Federal do Paraná. Av. schoolchildren, but had no significant impact on significativo sobre esta variável entre crianças de
Lothário Meissner 632, this variable among children from other income diferentes estratos de renda. Diante disso, garan-
Jardim Botânico. 80210- strata. Thus, ensuring access to quality food may tir o acesso a uma alimentação de qualidade pode
170 Curitiba PR Brazil.
[email protected] be a good strategy for minimizing inequities in representar uma boa estratégia para minimizar as
2
Departamento de oral health and reducing dental caries experience iniquidades em saúde bucal e reduzir a experiên-
Odontologia Preventiva e among schoolchildren from low-income families. cia de cárie dentária em escolares de baixa renda.
Social, Universidade Federal
de Minas Gerais (UFMG). Key words Food insecurity, Dental caries, Child, Palavras-chave Segurança alimentar e nutricio-
Belo Horizonte MG Brasil. Income nal, Cárie dentária, Criança, Renda
3
Departamento de
Odontopediatria e
Ortodontia, UFMG. Belo
Horizonte MG Brasil.
574
Santin GC et al.

Introduction crose as a substrate for the production of intra-

cellular and extracellular polysaccharides, which
Despite social advances over time, hunger and are important factors in the genesis of caries22.
malnutrition remain important problems for Frequent sugar intake leads to the selection of
humanity. Although access to adequate food is acidogenic and acid uric bacteria in the biofilm
stipulated in the Universal Declaration of Hu- capable of promoting the de-mineralization of
man Rights1, approximately one billion people the dental enamel23. Thus, the high carbohydrate
throughout the world are malnourished2. In Bra- intake characteristic of individuals in a state of
zil, 16 million people live below the poverty line FI can alter the microbiological and biochemical
and are unable to meet their basic food needs3. composition of dental biofilm, making it more
Food insecurity (FI) is a concept that encom- cariogenic.
passes from concerns and anguish of individuals Some studies have evaluated FI and oral
regarding the uncertainty of regularly obtaining health, identifying possible associations between
food until the experience of hunger due to not these variables. A significant association was
eating for an entire day. Reducing dietary variety found among FI, the seeking of dental services
and the amount of food constitute strategies that and the type of treatment performed on children
are often employed to cope with this situation, aged five to 14 years in New Zealand24-26. Another
which lead to an evident loss in nutritive qual- study involving individuals between 18 and 64
ity4,5. years of age in Canada found that low-income
Despite their important association, poverty laborers in a situation of FI exhibited a greater
and FI do not comprise the same concept. In- frequency of denture use and toothache, with
sufficient income hinders access to basic needs, consequent chewing and working difficulties, in
such as food, clothing, housing, education and comparison to low-income laborers without FI27.
health care. However, an individual may be poor However, the studies cited assessed oral health
without being affected by hunger when his/her through self-reports of the interviewees, with no
poverty is expressed by basic needs other than clinical examination performed and no dietary
food6. There are also situations in which access profile evaluated. Moreover, the samples com-
to food does not depend upon income. Children prised a wide age. FI was assessed using a small
may receive meals offered in the school setting7, number of items that are insufficient for under-
families may produce food crops for subsistence standing all the dimensions of this condition,
(family farming)8 or receive foods from system- which ranges from a reduction in the amount
ic donations. Moreover, families with an income of healthy foods to the replacement of foods and
above the poverty level may exhibit signs of FI even complete restriction. Only one study was
due to other needs and priorities that compete found having used a validated questionnaire to
with food needs9. measure FI and clinical examination to assess
Individuals with FI generally have a diet that dental caries, which found a highly association
is rich in carbohydrates and fat10. The literature between these variables in 7- to 9-year-old Bra-
indicates that adults in this situation are more zilian children from a town with a low Human
vulnerable to heart disease, diabetes, high blood Development Index score28.
pressure11, obesity11-13 and stress12. FI appears to Thus, the aim of the present study was to
have adverse effects on the health of children analyze the association between clinically diag-
and adolescents as well, with an increase in the nosed untreated dental caries (UDC) and house-
rates of iron14 and zinc15 deficiency, acute infec- hold food insecurity (HFI) in different income
tion, chronic diseases16, delayed cognitive devel- strata of a representative sample of 12-year-old
opment17 and anxiety18. Divergent results are re- schoolchildren, considering other risk variables
ported regarding the association between FI and for dental caries.
overweight in children19-21. Further studies on
this issue are needed.
Diet is also a determinant factor in the eti- Methods
ology of dental caries. Foods with high carbohy-
drate content, especially sucrose, are considered Study design and sample
the most cariogenic. Bacteria found in dental
biofilm (plaque) use sucrose in the glycolytic A cross-sectional study was carried out with
pathway and generate acids that can lead to the a multistage randomized sample representative
demineralization of teeth. Some bacteria use su- of 12-year-old schoolchildren in a medium-sized
 575

Ciência & Saúde Coletiva, 21(2):573-584, 2016

southern city in Brazil (Araucária - Paraná), Questionnaire (for counting the number of sug-
which has the second highest gross domestic ary foods consumed daily). The results revealed
product in its state and a human development no need to change the initially proposed meth-
index of 0.80. The population in 2010 was esti- ods. The children enrolled in the pilot study were
mated at 109,943 inhabitants, with 101,796 resid- not included in the main sample.
ing in urban areas29. Fluoridation of the public
drinking water has occurred in the city since the Data collection
early 1980s and the mean fluoride concentration
in urban areas is 0.7 mg/L29. Data from the Mu- For the assessment of dental caries experi-
nicipal Secretary of Education reveal that the city ence, the participants were submitted to a clinical
had 33 schools in urban areas in 2010, with a to- examination at school using the decayed, miss-
tal of 2072 students aged 12 years30. The decision ing and filled teeth (DMFT) index32. The clinical
to study a sample of 12-year-old schoolchildren exam was performed by a single, previously cal-
was based on fact that the World Health Orga- ibrated examiner (TPP). The calibration process
nization (WHO) considers this to be the index involved two steps. Theoretic activities involved
age for the assessment of caries in the permanent a discussion on the diagnosis of different stages
dentition of children31. of caries using projected photographic images. In
The calculation of the sample size was de- the practical step, 10 children were examined by
termined using the proportion estimate formula the same examiner on two occasions, with a 15-
based on a prevalence value of 63.3% for dental day interval between examinations. The dental
caries among 12-year-olds32, a 95% confidence professional who conducted the training of the
level and maximal acceptable error of 5%, con- examiner was considered the gold standard for
sidering a total population of 2072 schoolchil- the determination of inter-examiner agreement.
dren. The minimal sample was determined to be The exams were carried out at the respective
304 individuals, to which 60% as added to com- schools under natural light with the participants
pensate for the cluster effect (n = 487) and a fur- in the sitting position using a sterilized WHO ball
ther 20% was added to compensate for possible point probe (Trinity™, São Paulo, SP, Brazil), ster-
losses, resulting in a total of 584 12-year-olds. ilized flat mouth mirror (Golgran™, São Paulo,
Two-stage (schools and children) random- SP, Brazil) and individual protection equipment
ized cluster sampling was employed. The sample for the control of cross-infection. Sterilized gauze
was also stratified by regional administration was used to clean and dry the dental surfaces.
districts (8 districts) and type of school (public HFI was assessed using the Brazilian Food
and private). In the first stage, public and private Insecurity Scale (FIS-B), which is a version val-
schools were randomly selected by lots in each idated for Brazilian Portuguese33 of a measure
administration district for the acquisition of the developed in the United States of America for
determined number of students. In the second the assessment of HFI34. This assessment tool has
stage, 12-year-olds were randomly selected by a specific version for households with children
lots from the selected schools aiming to main- and/or adolescents, which was filled out by the
taining the proportion of 12-year-old students parents/guardians. The FIS-B for families is com-
enrolled at each district and each type of school posed of 15 items with response options of ‘yes
using data from the sample size calculation. Data (scored as 1)’, ‘no’ (scored as zero) and ‘doesn’t
collection was carried out between March and know/doesn’t want to answer’ (not counted in
September 2010. the score calculation). The items address the fam-
The inclusion criteria were 12 years of age, ily’s experience with the insufficient food intake
enrollment in public or private schools in Arau- in the previous three months, including concerns
caria city and absence of systemic disease (based with a possible lack of food and the possibility of
on information provided by parents/guardians). going an entire day without food. Affirmative re-
sponses (yes = 1) are summed (score range from
Pilot Study 0 to 15), allowing the classification of the fam-
ilies into the following categories: food security
A pilot study was carried out involving a con- (score: 0), mild FI (score: 1 to 5), moderate FI
venience sample of 61 schoolchildren and their (score: 6 to 10) and severe FI (score: 11 to 15)33.
respective parents/guardians to test the methods Socioeconomic and demographic data were
and understanding of the data collection ques- collected using a questionnaire filled out by the
tionnaires and to draft a specific Food Frequency parents/guardians on the gender of the child,
576
Santin GC et al.

mother’s schooling (nine categories ranging children were instructed to chew 2 cm2 of par-
from ‘did not study’ to ‘complete university ed- affin (Parafilm M™, Laboratory Film, Chicago,
ucation’), mother’s marital status (‘married/ USA), swallowing the saliva produced in the first
stable union’, ‘single’, ‘widowed’ and ‘separated/ minute and expectorating the saliva produced in
divorced’), number of children, number of res- the second and third minutes into a disposable
idents in home and monthly household income. plastic recipient. One aliquot of saliva was imme-
Income was analyzed based on the Brazilian diately used for inoculation of culture media on
minimum wage (BMW, equal to US$ 284.00 at the surfaces of the slides of the Dentalcult I and
the time of the study conduction) divided by the II kits. A disk of bacitracin was then deposited on
number of residents in the home for the calcula- the surface of the culture medium of the Den-
tion of per capita household income. The ques- talcult II kit and a CO2 pellet was placed at the
tionnaire also contained a question on access bottom of the flask. The flasks were incubated in
to dental care ‘Did your child ever need dental a microbiological culture incubator at 37° C for
treatment and not received it because of your im- 48 hours (Dentalcult II) or 72 hours (Dentalcult
possibility of affording it or getting an appoint- I). The number of colony-forming units (CFU)
ment for he/she at a public dental service?’ (‘yes’, of MS and LB per mL of saliva in each sample
‘no’) and a question on the time elapsed since was determined by similarities with the bacterial
the child’s last dental visit (‘less than 1 year’, ‘1 levels illustrated in the manufacturer’s manual.
to 3 years ago’, ‘3 to 5 years ago’, ‘over 5 years ago’, The children also answered a question on the
‘never’). frequency with which they brushed their teeth
The Food Frequency Questionnaire was used (‘does not brush’, ‘once a day’, ‘twice a day’, ‘three
for the determination of sugar intake. This ques- or more times a day’).
tionnaire was made up with the 20 foods (solids
and liquids) which contained sugar in its com- Data analysis
position most often reported to be consumed in
the 24-hour logs by the participants in the pilot Data analysis involved the description of the
study (n = 61)35. The participants in the main absolute and relative frequencies of the variables
study were instructed to choose an option (‘nev- according to the response variable ‘UDC’, which
er’, ‘less than once a month’, ‘1 to 3 times a month’, was dichotomized as ‘present’ (D component of
‘once a week’, ‘2 to 4 times a week’, ‘once a day’ DMFT index ≥ 1) or ‘absent’ (D component of
and ‘2 or more times a day’) that came closest DMFT index = 0). Univariate Poisson regression
to their consumption of each of the foods listed analyses were performed to determine associ-
on the questionnaire. Every different food, solid ations between the response variable and inde-
or liquid, that was answered to be consumed at pendent variables. ‘Brushing frequency’, ‘salivary
least once a day was computed and added to as- levels of MS and LB’, ‘marital status’, ‘mother’s
sess the total number of sugary foods consumed schooling’ and ‘time since last dental visit’ were
daily by each child. dichotomized based on theoretic references.
The sweet taste preference was assessed using ‘Number of children’ and ‘number of residents in
a modified version of the Sweet Preference Inven- home’ were dichotomized by the median value.
tory36, having the tea (Mate Leão™, Curitiba, PR, Sugar intake was categorized as ‘0 to 3’, ‘4 to 6’
Brazil) as the vehicle, with the following concen- and ‘ ≥ 7’ sugary foods consumed daily and sweet
trations of sucrose: 0.075 mol/L, 0.15 mol/L, 0.3 preference was categorized as ‘low’, ‘medium’ and
mol/L, 0.6 mol/L and 0.9 mol/L. It was also added ‘high’ concentrations of sugar. Household in-
one sugar free option (0.0 mol/L). The partici- come was categorized based on the Brazilian pov-
pants tasted one solution (30 mL) at a time in erty line, which considers families below fourth
an increasing order of sugar concentration, alter- BMW per capita as in extreme poverty37. Thus, it
nating the ingestion of the solution with saltine was divided into the following strata: ‘per capita
crackers to desensitize the taste buds. At the end income ≤ ¼ BMW’, ‘¼ BMW < per capita income
of the test, the participant chose the solution that ≤ ½ BMW’, ‘½ BMW < per capita income ≤ 1
most pleased him/her. BMW’, per capita income > 1 BMW’. HFI (main
Samples of stimulated saliva were collected explanatory variable) was dichotomized as ‘pres-
for the assessment of lactobacilli (LB) and mu- ent’ (FIS-B score ≥ 1) or ‘absent’ (FIS-B score =
tans streptococci (MS) using the Dentalcult I and 0). Multivariate Poisson regression models with
Dentalcult II laboratory kits (Laborclin™, Pinhais, robust variance were constructed to investigate
PR, Brazil), respectively. In the school setting, the the association between UDC and HFI in each
 577

Ciência & Saúde Coletiva, 21(2):573-584, 2016

income stratum, which allowed estimating prev- Sixteen percent of the children belonged to
alence ratios (PR) of UDC and respective 95% families with a per capita household income of
confidence intervals (CI) for the groups of chil- up to ¼ BMW; 31% belonged to families with a
dren with and without HFI, adjusted by the other per capita household income above ¼ and less
independent variables. Variables with a p-value < than ½ BMW; 37% belonged to families with a
0.20 in the univariate analysis were incorporated per capita income above ½ and less than 1 BMW
into the models. Biological, behavioral and so- and 16% belonged to families with a per capita
cioeconomic variables that remained significant income above 1 BMW.
(p ≤ 0.05) after adjustments were maintained The number of sugary foods consumed on a
in the final models38. HFI was maintained in all daily basis by the children ranged from 0 to 18
models regardless of the p-value. The data were (mean: 6; SD: 4; median: 5).
analyzed using the Statistical Package for the So- The prevalence of untreated caries was great-
cial Sciences (SPSS for Windows, version 15.0, er among children who studied at public schools,
SPSS™ Inc, Chicago, IL, USA). those whose mothers had up to eight years of
The study was approved by the Committee schooling, those in families with three or more
for Ethics in Human Research of Universidade children, those with five or more residents in the
Federal do Paraná (Brazil) and carried out ac- home, those with a per capita household income
cording to the Declaration of Helsinki. Consent less than ½ BMW and those in food-insecure
was obtained from the parents/guardians and households (p < 0.05) (Table 2).
agreement was obtained from the children indi- The prevalence of UDC was greater among
vidually prior to the data acquisition procedures. children with salivary levels of MS and LB ≥ 105
CFUs/mL (p = 0.004 and 0.011, respectively),
those with a greater daily sugary foods intake (p <
Results 0.05) and those who reported not brushing their
teeth at least twice a day (p < 0.001) (Table 3).
A total of 538 children (318 girls and 220 boys) Children in households with more severe sit-
participated in the present study, representing a uations of FI exhibited a greater daily consump-
response rate of 92%. The main reasons for the tion of sugary foods (Chi-square test for linear
losses were the non-return of the questionnaires trend, p < 0.001). A significant association was
and the absence of the child on the day of the found between income and HFI, as a lower per
exam. Intra-examiner and inter-examiner Kap- capita household income denoted higher HFI
pa values on a tooth-by-tooth basis were greater prevalence and severity values (Chi-square test
than 0.81. for linear trend, p < 0.001).
The mean DMFT index was 2.4 (standard de- The Poisson multivariate models stratified
viation - SD: 2.3; median: 2). The prevalence of by income and adjusted for brushing frequency,
UDC (D component of DMFT index) was 45% salivary levels of MS and LB, number of sugary
(95% CI: 41-50); 54% of the children had filled foods consumed daily and mother’s schooling
teeth (F component of DMFT index) and 0.7% care demonstrated that HFI was associated with
of the children had tooth loss due to caries (M UDC only in the children with a per capita in-
component of DMFT index). come ≤ ¼ BMW (p = 0.045) (Table 4).
The prevalence of HFI was 39% (95% CI:
35-44); 28% of the children were in situation of
mild HFI, 6% had moderate HFI and 5% had Discussion
severe HFI. Total FIS-B scores ranged from 0 to
15 (mean: 1.9; SD: 3.5; median: 0). The ques- In the present study, UDC was associated with
tions with the greatest frequency of affirmative HFI. Only one study that investigated the im-
responses from parents/guardians were ‘Were pact of FI on clinically evaluated dental caries
you worried that the food at home would run experience was found in the literature, which was
out before you could afford to buy or received conducted with Brazilian Amazon children28.
more?’ (27.9%), ‘Did you run out of money for Regardless minor differences in the study pop-
a healthy, varied diet?’ (25.7%), ‘Did you have to ulation (age and socio-economic contexts), the
get by with only a few foods because the money survey in Amazon also observed an association
ran out?’ (23.2%), ‘Were you unable to offer your between dental caries and HFI in children. Like-
children a healthy, varied meal because you had wise, previous studies using self-reports of oral
no money?’ (21.9%) (Table 1). health have indicated this association, when car-
578
Santin GC et al.

Table 1. FIS-B responses of parent/guardian of 12-year-olds, according to the last three months. Araucária city,
Brazil (n = 538).

Doesn’t know/
Questions Yes No Doesn’t want
n (%) n (%) to answer
n (%)
Were you concerned that the food ran out before the house could 150 (27.9) 327 (60.8) 61 (11.3)
afford to buy or received more?
Did the food run out before you had money to buy more? 88 (16.4) 386 (71.7) 64 (11.9)
Did you run out of money for a healthy, varied diet? 138 (25.7) 340 (63.2) 60 (11.2)
Did you to get by with only a few foods because the money ran out? 125 (23.2) 356 (66.2) 57 (10.6)
Were you unable to offer your children a healthy, varied meal 118 (21.9) 353 (65.6) 67 (12.5)
because you had no money?
Did the children not eat enough because there was no money to buy 54 (10.0) 416 (77.3) 68 (12.6)
more?
Did an adult at home any time decrease the amount of food or skip 58 (10.8) 414 (77.0) 66 (12.3)
meals because there was not enough money to buy more?
Did you ever eat less than felt you should because there was no 54 (10.0) 403 (74.9) 81 (15.1)
money to buy more?
Have you ever felt hungry but did not eat because they could not 38 (7.1) 411 (76.4) 89 (16.5)
afford enough food?
Did any adult at home lose weight because they do not have enough 24 (4.5) 426 (79.2) 88 (16.4)
money to buy food?
Did any adult ever spend a whole day without eating or just had a 31 (5.8) 419 (77.9) 88 (16.4)
meal because there was no money to buy more?
Have you ever reduced the amount of food children’s meals because 42 (7.8) 413 (76.8) 83 (15.4)
there was no money to buy more?
Have you ever had to skip a meal for children because there was no 36 (6.7) 415 (77.1) 87 (16.2)
money to buy food?
Were the children hungry, but you just could not afford more food? 25 (4.6) 429 (79.7) 84 (15.6)
Were the children without food for a whole day because there was 25 (4.6) 428 (79.6) 85 (15.8)
no money to buy more food?

ies was represented by outcomes such as ‘presence that dental caries constitute a multifactor con-
of toothache at night or in the previous month’, dition and confer credibility to both the method
‘denture use’ and ‘previous experience with res- employed and the results obtained regarding HFI.
torations and extractions’ in children between The use of a multivariate model involving ex-
five and 14 years of age in New Zealand24-26 and plicative variables on different levels for the eval-
low-income laborers between 18 and 64 years of uation of the association between HFI and den-
age in Canada27. tal caries is a strong point of the present study,
In the present state of knowledge regarding allowing the avoidance of the interference of
the etiology of dental caries, it is unacceptable confounding variables regarding the effect found
for studies addressing factors associated with this among the variables of interest and ensuring
outcome to disregard its classic determinants, greater accuracy in the findings. Moreover, the
such as demographic, socioeconomic, behavior inclusion of microbiological factors in this analy-
and biological factors. The present study corrob- sis fills a gap often found in field studies. Despite
orates associations reported in the literature be- the recognized importance of biological factors
tween dental caries and the number of children39, in the causal network of caries, it is difficult to
income40-42, number of residents in home43, type evaluate variables that require laboratory sup-
of school43, mother’s schooling41, sugar intake44, port when large samples are involved.
brushing frequency40,42,44 and salivary levels of MS One hypothesis that may explain the asso-
and LB45. These findings strengthen the notion ciation between UDC and HFI is that FI leads
 579

Ciência & Saúde Coletiva, 21(2):573-584, 2016

Table 2. Association between untreated dental caries and demographic / socioeconomic variables in 12-year-old
schoolchildren. Araucária city, Brazil (n = 538).

Untreated dental caries Univariate analysis

Variables No Yes Total p† PR

n (%) n (%) n* (100%) (95% CI)
Household food insecurity
Absent 191 (58.6) 135 (41.4) 326 1
Present 103 (48.6) 109 (51.4) 212 0.021 1.24 (1.03-1.49)
Gender of child
Female 114 (51.8) 106 (48.2) 220 1
Male 180 (56.6) 138 (43.4) 318 0.27 1.11 (0.92-1.34)
Type of school of child
Private 31 (86.1) 5 (13.9) 36 1
Public 263 (52.4) 239 (47.6) 502 0.003 3.42 (1.51-7.75)
Mother’s schooling
More than 8 years of schooling 163 (60.4) 107 (39.6) 270 1
Until 8 years of schooling 122 (50.2) 121 (49.8) 243 0.021 1.26 (1.04-1.52)
Marital status of mother
Married or in stable union 223 (56.7) 170 (43.3) 393 1
Single, widowed or separated/divorced 63 (51.6) 59 (48.4) 122 0.31 1.12 (0.90-1.39)
Number of children
Until 3 children 232 (62.4) 140 (37.6) 372 1
4 or more children 54 (38.8) 85 (61.2) 139 < 0.001 1.63 (1.35-1.96)
Number of residents in home
Until 4 people 163 (61.0) 104 (39.0) 267 1
5 or more people 123 (49.8) 124 (50.2) 247 0.011 1.29 (1.06-1.56)
Per capita household income
Income > 1 BMW 50 (67.6) 24 (32.4) 74 1
½ BMW < income ≤ 1BMW 103 (58.9) 72 (41.1) 175 0.21 1.27 (0.87-1.84)
¼ BMW < income ≤ ½ BMW 74 (51.4) 70 (48.6) 144 0.032 1.50 (1.04-2.17)
Income ≤ ¼ BMW 33 (45.2) 40 (54.8) 73 0.008 1.69 (1.35-1.96)
Last visit of the children to the dentist
In the last year 225 (57.5) 166 (42.5) 391 1
Over a year ago 58 (49.2) 60 (50.8) 118 0.09 1.20 (0.97-1.48)
Cost prohibiting access to dental care
No 191 (56.7) 146 (43.3) 337 1
Yes 92 (53.5) 80 (46.5) 172 0.49 1.07 (0.88-1.31)

PR, prevalence ratio; CI, confidence interval; BMW, Brazilian minimum wage (about US$ 284.00 at the time of the study
conduction).* Frequencies lower than 538 are due to missing data. † Univariate Poisson regression analysis. Results significant at 5%
level marked in bold.

to an increase in sugar intake. Households in a This hypothesis is supported by the analysis

situation of FI are more likely to purchase foods of the responses to the items of the FIS-B. Three
with a high energy density that are poor in nutri- of the four questions with the highest rate of af-
ents and contain refined carbohydrates, fats and firmative answers (‘Did you run out of money
added sugar46. The present study represents ad- for a healthy, varied diet?’, ‘Did you have to get
vancement in comparison to previous studies in- by with only a few foods because the money ran
vestigating the relationship between dental caries out?’ and ‘Were you unable to offer your children
and FI24-28 by analyzing the dietary profile of the a healthy, varied meal because you had no mon-
children and demonstrating that the consump- ey?’) regarded the qualitative rather than quanti-
tion of sugary foods was greater among children tative nature of food restrictions, resulting in an
in food-insecure households, especially in those unhealthy eating pattern with the replacement
with severe HFI. of foods. However, when ‘intakes of sugary food’
580
Santin GC et al.

Table 3. Association between untreated dental caries and biological and behavior variables in 12-year-old
schoolchildren. Araucária city, Brazil (n = 538).

Untreated dental caries Univariate analysis

Variables No Yes Total p† PR

n (%) n (%) n* (100%) (95% CI)
Mutans streptococci
< 105 CFU/mL 100 (64.9) 54 (35.1) 154 1
≥ 105CFU/mL 194 (50.5) 190 (49.5) 384 0.004 1.41 (1.11-1.79)
Lactobacilli
< 105 CFU/mL 276 (56.1) 216 (43.9) 492 1
≥ 105 CFU/mL 18 (39.1) 28 (60.9) 46 0.011 1.39 (1.08-1.79)
Nº of sugary foods consumed daily
0 – 3 sugary foods 108 (66.3) 55 (33.7) 163 1
4 – 6 sugary foods 88 (54.0) 75 (46.0) 163 0.025 1.36 (1.04-1.79)
7 or more sugary foods 96 (46.2) 112 (53.8) 208 < 0.001 1.60 (1.24- 2.05)
Sweet Taste Preference
Low (0.0 e 0.075 mol/L) 27 (54.0) 23 (46.0) 50 1
Medium (0.15 e 0.30 mol/L) 193 (56.6) 148 (43.4) 341 0.190 1.14 (0.93-1.40)
High (0.6 e 0.9 mol/L) 74 (50.3) 73 (49.7) 147 0.660 1.08 (0.77-1.52)
Tooth brushing frequency
More than twice a day 201 (87.0) 30 (13.0) 231 1
Until twice a day 93 (30.3) 214 (69.7) 307 < 0.001 5.38 (3.82-7.58)
PR, prevalence ratio; CI, confidence interval. * Frequencies lower than 538 are due to missing data.
†
Univariate Poisson regression analysis. Results significant at 5% level marked in bold.

had been included in the multiple model, the greater severity of FI in this stratum. In a fami-
statistical significance of HFI was maintained. ly, children are only affected by more severe de-
So, there must be other aspects of FI beyond an grees of FI, whereas adults are the ones affected
increase in sugar intake influencing the dental and children are generally spared in situations of
caries experience. mild degrees of FI49. Moreover, HFI in higher in-
Another strong point of the present study come strata may represent unexpected economic
was the analysis of the association between den- changes in the family structure, with the emer-
tal caries and HFI for each income stratum. In- gence of factors that may temporarily compete
come has been described as the most important with the food budget21 and incapability of man-
variable in the determination of HFI in Brazil47. aging their finances50.
Households with a monthly income of less than In a case-control study comparing families
two times the BMW have an approximately 13- with an income above the poverty line in situa-
fold greater chance of being in a situation of tions of mild FI and those with food security, the
moderate to severe FI and a 2.9-fold chance of authors found that 37.5% of the former reported
being in a situation of mild FI48. As this finding the loss of a job, the gain of another household
was confirmed in the present study and, based member or the loss of food assistance50. Thus,
on the belief that, regardless the collinearity of HFI that affects higher income strata tends to be
these variables, income and FI comprise distinct of short duration, which may not be enough to
constructs, the decision was made to stratify the have an impact on oral health, as tooth decay is a
analyses by income. chronic condition that depends on the exposure
The stratified analyses demonstrated that the time to risk factors.
association between dental caries and HFI does The finding that the association between den-
not occur in the same fashion in the different tal caries and HFI occurs in the lowest income
income strata. The finding that dental caries in stratum suggests the need for food banks and/or
schoolchildren was influenced by HFI only in the assistance programs that can meet the needs of
lowest income stratum may be explained by the these families or income transference programs,
 581

Ciência & Saúde Coletiva, 21(2):573-584, 2016

Table 4. Multivariate Poisson regression models: Association between untreated dental caries and household
food insecurity among 12-year-old schoolchildren, according to income strata. Araucária city, Brazil.

Untreated dental caries Multiple analysis

Models No Yes Total p Adjusted PR

n (%) n (%) n (100%) value [CI 95%]
Household per capita income ≤ ¼ BMW
Food security 13 (54.2) 11 (45.8) 24 1
Insecurity 20 (40.8) 29 (59.2) 49 0.045 1.51 [1.01-2.29]
¼ < per capita income ≤ ½ BMW
Food security 35 (53.0) 31 (47.0) 66 1
Food insecurity 39 (50.0) 39 (50.0) 78 0.482 1.10 [0.83-1.46]
½ < per capita income ≤ 1 BMW
Food security 76 (62.3) 46 (37.7) 122 1
Food insecurity 27 (50.9) 26 (49.1) 53 0.609 1.08 [0.80-1.45]
Per capita income > 1 BMW
Food security 39 (65.0) 21 (35.0) 60 1
Food insecurity 11 (78.6) 3 (21.4) 14 0.473 0.74 [0.33-1.67]
Adjusted PR, prevalence ratio adjusted by varying levels of salivary mutans streptococci and lactobacilli, frequency of tooth
brushing, frequency of consumption of sugary foods and mother’s schooling. CI, confidence interval; BMW, Brazilian minimum
wage (about US$ 284.00 at the time of the study conduction). Results significant at the 5% level marked in bold. Frequencies lower
than 538 are due to lack of data to one of the variables of the model.

which may assist in reducing the prevalence of explained by the fact that low-income families
dental caries. Every 10 Reais (Brazilian curren- with FI cannot afford to pay for dental treatment
cy corresponding to US$ 5.57) in increased in- due to competing financial demands. However,
come in Brazilian families stemming from so- in the present study, 77% of the children report-
cial transference programs increases the odds of ed having visited the dentist in the previous year
food security by 8%, especially in low-income and UDC was not related to difficulties in access
households (per capita income up to ¼ BMW)49. to health services.
In the USA, a study demonstrated that access The present study has the limitations of all
to breakfast at schools reduces the risk of mild cross-sectional studies, especially the lack of
HFI7. Another study realized in the New Zealand temporality, which limits reliability in establish-
observed that offering supermarket vouchers to ing the direction of the associations. The study
family with food-insecurity resulted in an in- design and the assessment tool used for quanti-
crease in overall expenditure on food51. Thus, fying HFI do not allow identifying how long the
there is a need for changes in public policies and individuals surveyed have been in a situation of
health promotion actions. An approach directed FI. HFI represents the conditions of household
solely at the disease factor and its prevention has members as a group and not necessarily the con-
proven ineffective at times and there is a need for dition of any particular member, especially a
interventions that can reduce health inequalities child. Such information would be useful to a bet-
through changes in the determinants of health52. ter understanding of the role of FI as an episodic
As the FIS-B provides information on a recent or cyclic condition in the development of dental
period (previous three months), the decision was caries. As tooth decay has a chronic, cumulative
made to assess tooth decay based on the pres- nature, the effects of FI on oral health are expect-
ence of UDC and not the DMFT index, which ed to be greater when the situation persists for a
includes filled and missing teeth and therefore longer period of time. Thus, longitudinal studies
may reflect a situation prior to the HFI event. using instruments that provide individually ref-
Since untreated caries is also a measure of access erenced food security information for children
to dental care, it would be possible for the mod- are needed to allow a better understanding of the
erating effect of income on HFI to be partially association between dental caries and FI.
582
Santin GC et al.

Conclusions

Being in a food-insecure household was associ-

ated with a greater frequency of untreated dental
caries among low-income schoolchildren, but
had no significant impact on this variable among
children from other income strata.
Socioeconomic inequities are recognized to
exert an influence on the occurrence of dental
caries. Although the ways by which social disad-
vantages affect this outcome are not fully under-
stood, the present findings indicate that actions
aimed at ensuring access to quality food may be
a good strategy for minimizing inequities in oral
health and reducing caries experience among
schoolchildren from low-income families.

Collaborations

GC Santin and TP Pintarelli participated in the

study design, data collection, analysis and inter-
pretation the results and writing the article. FC
Fraiz and FM Ferreira participated in the study
design, analysis and interpretation of results, wri-
ting and critical review of the article. ACB Olivei-
ra and SM Paiva participated in the analysis and
interpretation the results and critical review of
the article. All authors approved the final version
of the manuscript.
 583

Ciência & Saúde Coletiva, 21(2):573-584, 2016

References

1. United Nations. Universal declaration of human rights. 17. Rose-Jacobs R, Black MM, Casey PH, Cook JT, Cutts
[cited 2014 Oct 1]. Available from:http://www.onu.org. DB, Chilton M, Heeren T, Levenson SM, Meyers AF,
br/a-onu-em-acao/a-onu-e-os-direitos-humanos/ Frank DA. Household food insecurity: associations
2. Food and Agriculture Organization of the United Na- with at risk infant and toddler development. Pediatrics
tions. 925 million people chronic hungry world. [cited 2008;121(1):65-72.
2014 Oct 1]. Available from: http://www.fao.org.br/ 18. Whitaker RC, Phillips SM, Orzol SM. Food insecurity
925mchw.asp and the risks of depression and anxiety in mothers and
3. Brazil. Ministry of Social Development. [Internet]. behavior problems in their preschool-aged children.
[cited 2014 Oct 1]. Available from: http://www.mds. Pediatrics 2006; 118(3):e859-e868.
gov.br/brasilsemmiseria 19. Matheson DM, Varady J, Varady A, Killen JD. House-
4. Bickel G, Nord M, Price C, Hamilton W, Cook J. Mea- hold food insecurity and nutritional status of His-
suring food security in the United States. Guide to mea- panic children in the fifth grade. Am J Clin Nutr 2002;
suring household food security, revised 2000. Alexandria: 76(1):210-217.
Food and Nutrition Service, Office of Analysis, Nutri- 20. Casey PH, Simpson PM, Gossett JM, Bogle ML, Cham-
tion, and Evaluation, U.S. Department of Agriculture; pagne CM, Connell C, Harsha D, McCabe-Sellers B,
2000. Robbins JM, Stuff JE, Weber J. The association of child
5. Kepple AW, Segall-Correa AM. Conceituando e medin- and household food insecurity with childhood over-
do segurança alimentar e nutricional. Cien Saude Colet weight status. Pediatrics 2006; 118(5):e1406-e1413.
2011; 16(1):187-199. 21. Rose D, Bodor JN. Household food insecurity and
6. Hoffmann R. Determinants of food insecurity in Bra- overweight status in young school children: results
zil: Analysis of data from PNAD 2004. Segurança Ali- from the Early Childhood Longitudinal Study. Pedia-
mentar e Nutricional 2008; 15:49-61. trics 2006; 117(2):464-473.
7. Bartfeld JS, Ahn HM. The school breakfast program 22. Dibdin GH, Shellis RP. Physical and biochemical stud-
strengthens household food security among low-in- ies of Streptococcus mutans sediments suggest new
come households with elementary school children. J factors linking the cariogenicity of plaque with its
Nutr 2011; 141(3):470-475. extracellular polysaccharide content. J Dent Res 1988;
8. Wolfe WS, Frongillo Junior EA. Building household 67(6):890-895.
food-security measurement tools from the ground up. 23. Nobre dos Santos M, Melo dos Santos L, Francisco SB,
Food Nutr Bull 2001; 22:5-12. Cury JA. Relationship among dental plaque compo-
9. Rose D. Economic determinants and dietary conse- sition, daily sugar exposure and caries in the primary
quences of food insecurity in the United States. J Nutr dentition. Caries Res 2002; 36(5):347-352.
1999; 129(Supl. 2):517S-520S. 24. Jamieson LM, Koopu PI. Predictor of dental pain and
10. Monteiro CA. The extent of poverty, hunger and mal- general anesthetic receipt for hospital dental proce-
nutrition in Brazil. Estudos Avançados São Paulo. 1995; dures among New Zealand children. J Public Health
9:195-207. Dent 2006; 66(3):192-198.
11. Vozoris NT, Tarasuk VS. Household food insuffi- 25. Jamieson LM, Koopu PI. Exploring factors that influ-
ciency is associated with poorer health. J Nutr 2003; ence child use of dental services and toothbrushing in
133(1):120-126. New Zealand. Community Dent Oral Epidemiol 2006;
12. Stuff JE, Casey PH, Szeto KL, Gossett JM, Robbins JM, 34(6):410-418.
Simpson PM, Connell C, Bogle ML. Household food 26. Jamieson LM, Koopu PI. Child use of dental services
insecurity is associated with adult health status. J Nutr and receipt of dental care in New Zealand. J Paediat
2004; 134(9):2330-2335. Child Health 2001; 43(11):732-739.
13. Tanumihardjo SA, Anderson C, Kaufer-Horwitz M, 27. Muirhead V, Quiñonez C, Figueiredo R, Locker D. Oral
Bode L, Emenaker NJ, Haqq AM, Satia JA, Silver HJ, health disparities and food insecurity in working poor
Stadler DD. Poverty, obesity, and malnutrition: An in- Canadians. Community Dent Oral Epidemiol 2009;
ternational perspective recognizing the paradox. J Am 37(4):294-304.
Diet Assoc 2007; 107(11):1966-1972. 28. Frazão P, Benicio MHD, Narvai PC, Cardoso MA.
14. Skalicky A, Meyers AF, Adams WG, Yang Z, Cook JT, Food insecurity and dental caries in schoolchildren: a
Frank DA. Child food insecurity and iron deficiency cross-sectional survey in the western Brazilian Ama-
anemia in low-income infants and toddlers in the Unit- zon. Eur J Oral Sci 2014; 122(3):210-215.
ed States. Matern Child Health J 2006; 10(2):177-185. 29. Paraná Institute of Economic and Social Develop-
15. Pedraza DF, Queiroz D, Paiva AA, Cunha MAL, Lima ment. Profile of the Municipality of Araucaria, 2006.
ZN. Seguridad alimentaria, crecimiento y niveles de [cited 2014 Oct 1]. Available from: http://www.ipardes.
vitamina A, hemoglobina y zinc en niños pré-esco- gov.br/perfil_municipal/MontaPerfil.php?Municipi-
lares del nordeste de Brasil. Cien Saude Colet 2014; o=83700&bt
19(2):641-650.
16. Kursmark M, Weitzman M. Recent findings concerning
childhood food insecurity. Curr Opin Clin Nutr Metab
Care 2009; 12(3):310-316.
584
Santin GC et al.

30. Brazil. Ministry of Education. National Institute of 43. Antunes JL, Frazão P, Narvai PC, Bispo CM, Pegoret-
Educational Studies. Preliminary Results of the 2008 ti T. Spatial analysis to identify differentials in dental
School Census – Educacenso. [cited 2015 Dec 12]. needs by area-based measures. Community Dent Oral
Available from: http://www.inep.gov.br/basica/censo Epidemiol 2002; 30(2):133-142.
/Escolar/Matricula/censoescolar_2008.asp?metodo 44. Jürgensen N, Petersen PE. Oral health and the impact
=1&ano=2008&UF=PARAN%C1&MUNICIPIO= of socio-behavioral factors in a cross sectional survey
araucaria&Submit=Consulta of the 12-year old school children in Laos. BMC Oral
31. World Health Organization (WHO). Oral health sur- Health 2009; 16(9):29.
veys: basic methods. 4th ed. Genebra: WHO; 1997. 45. Zukanović A, Muratbegović A, Kobašlija S, Marković
32. Brazil. Ministry of Health. National Coordination of N, Ganibegović M, BeŠlagić E. Relationships between
Oral Health. Project SB Brazil 2003: oral health status socioeconomic backgrounds, caries associated micro-
of the Brazilian population 2002-2003. Main results. flora and caries experience in 12-year-olds in Bosnia
[cited 2015 Dec 12]. Available from: http://bvsms. and Herzegovina in 2004. Eur J Paediatric Dent 2008;
saude.gov.br/bvs/publicacoes/ projeto_sb2004.pdf 9(3):118-124.
33. Pérez-Escamilla R, Segall-Corrêa AM, Maranha LK, 46. Drewnowski A, Darmon N. Food choices and diets
Sampaio MFA, Marín-León L, Panigassi G. An adapt- costs: an economic analysis. J Nutr 2005; 135(4):900-
ed version of the U.S. Department of Agriculture Food 904.
Insecurity module is a valid tool for assessing house- 47. Reis M. Food insecurity and the relationship between
hold food insecurity in Campinas, Brazil. J Nutr 2004; household income and children´s health and nutrition
134(8):1923-1928. in Brazil. Health Econ 2012; 21(4):402-427.
34. United State Department of Agriculture. Household 48. Panigassi G, Segall-Corrêa AM, Marin-Leon L, Pérez
Food Security. [cited 2015 Dec 12]. Available from: -Escamilla R, Sampaio MFA, Maranha LK. Food inse-
http://www.ers.usda.gov/Publications/ERR49/ER- curity as an indicator of inequity: analysis of a popula-
R49b.pdf tion survey. Cad Saude Publica 2008; 24(10):2376-2384.
35. Colucci ACA, Phillippi ST, Slater B. Steps for devel- 49. Segall-Corrêa AM, Marin-Leon L, Helito H, Perez-Es-
oping a food frequency questionnaire. Rev Bras Cienc camilla R, Santos LMP, Paes-Sousa R. Cash transfer-
Saude 2005; 3(6):7-12. ence and food insecurity in Brazil: analysis of national
36. Jamel HA, Sheiham A, Cowell CR, Watt RG. Taste pref- data. Rev Nutr 2008; 21(Supl.):39S-51S.
erence for sweetness in urban and rural populations in 50. Gundersen C, Kreider B. Bounding the effects of food
Iraq. J Dent Res 1996; 75(11):1879-1884. insecurity on children’s health outcomes. J Health Econ
37. Brazil. Decree nº. 7.492, Jun 2, 2011. Establishing the 2009; 28(5):971-983.
plan “Brazil without Poverty”. Diário Oficial da União 51. Smith C, Parnell WR, Brown RC, Gray AR. Providing
2011; 3 Jun. additional money to food-insecure households and its
38. Hosmer DW, Lemeshow S. Applied logistic regression. effect on food expenditure: a randomized controlled
4th ed. New York: Willey-Interscience Publication; 2000. trial. Public Health Nutr 2013; 16(8):1507-1515.
39. Nicolau B, Marcenes W, Bartley M, Sheiham A. A life 52. Watt RG. Emerging theories into the social determi-
course approach to assessing causes of dental caries nants of health: implications for oral health promotion.
experience: the relationship between biological, be- Community Dent Oral Epidemiol 2002; 30(4):241-247.
havioural, socio-economic and psychological con-
ditions and caries in adolescents. Caries Res 2003;
37(5):319-326.
40. Mendes LGA, Biazevic MGH, Miguel-Crosato E, Men-
des MOA. Dental caries and associated factors among
Brazilian adolescents: a longitudinal study. Braz J Oral
Sci 2008; 7(26):1614-1619.
41. Kim SW. Environmental, maternal, and child factors
which contribute to early childhood caries: a unifying
conceptual model. Int J Paediatr Dent 2012; 22(3):157-
168.
42. Brennan DS, Spencer AJ, Roberts-Thomson KF. Dental
self-care and visiting behavior in relation to social in- Artigo apresentado em 11/10/2014
equality in caries experience. Community Dent Health Aprovado em 16/06/2015
2011; 28(3):216-221. Versão final apresentada em 18/06/2015
Copyright of Revista Ciência & Saúde Coletiva is the property of Associacao Brasileira de
Pos-Graduacao em Saude Coletiva and its content may not be copied or emailed to multiple
sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.