J Epidemiol 2012;22(1):72-77

doi:10.2188/jea.JE20110042

Original Article

Association Between Breastfeeding and Dental Caries

in Japanese Children
Keiko Tanaka and Yoshihiro Miyake
Department of Preventive Medicine and Public Health, Faculty of Medicine, Fukuoka University, Fukuoka, Japan

Received April 11, 2011; accepted September 25, 2011; released online December 10, 2011

ABSTRACT
Background: Studies investigating the impact of breastfeeding on dental caries have produced contradictory
results. This cross-sectional study investigated the relationship between breastfeeding and the prevalence of dental
caries in young Japanese children.
Methods: The study subjects were 2056 Japanese children aged 3 years. Information on breastfeeding was obtained
by means of a questionnaire. Children were classified as having caries if 1 or more deciduous teeth were decayed,
missing, or had been filled at the time of examination.
Results: The prevalence of dental caries was 20.7%. As compared with breastfeeding for less than 6 months,
breastfeeding for 18 months or longer was associated with a significantly higher prevalence of dental caries. The
relation was J-shaped: the adjusted prevalence ratios for less than 6 months, 6 to 11 months, 12 to 17 months, and 18
months or longer were 1.0, 0.79 (95% confidence interval [CI]: 0.60–1.05), 0.86 (95% CI: 0.66–1.13), and 1.66 (95%
CI: 1.33–2.06), respectively (P for linear trend <0.0001, P for quadratic trend <0.0001).
Conclusions: Breastfeeding for 18 months or longer was positively associated with the prevalence of dental caries,
while breastfeeding for 6 to 17 months was nonsignificantly inversely associated with the prevalence of dental caries.

Key words: breastfeeding; cross-sectional studies; dental caries; Japan

association between daytime breastfeeding habits and

INTRODUCTION
prevalence.10
Breastfeeding is promoted as the preferred method These inconsistent and even contradictory results among
of infant feeding and provides a number of advantages, studies are likely mostly due to methodological differences,
including health, nutritional, immunological, develop- such as the use of different cut-off points for breastfeed-
mental, psychological, social, economic, and environmental ing, lack of adjustment for confounding factors, different
benefits.1 definitions of dental caries, and the ages at which outcomes
However, the evidence has been mixed regarding the were assessed. Additional data are required to reach a con-
association between breastfeeding and dental caries.2–18 clusion concerning the association between breastfeeding and
Several studies have reported an inverse association between dental caries. This cross-sectional study investigated the
breastfeeding and dental caries,2–4 while other studies have association between breastfeeding and the prevalence of
failed to show any highly beneficial association.5–13 Moreover, dental caries among young Japanese children.
some studies have demonstrated that breastfeeding was
associated with a higher prevalence of dental caries.2,10,13–18
METHODS
A US cross-sectional study in children aged 2 to 5 years
showed that breastfeeding and its duration were not associated Study population
with an increased prevalence of early childhood caries.5 In a Data for the present study came from the Fukuoka Child
retrospective cohort study of early childhood caries among Health Study (FCHS), a cross-sectional study of the
infants aged 25 to 30 months in Myanmar, the prevalence association between various selected factors and child health
of early childhood caries was greater in children who were problems, such as dental caries and allergic disorders.19–22 In
breastfed more than twice at night, whereas there was no Japan, when children reach age 3 years, the municipality in
Address for correspondence. Keiko Tanaka, DDS, PhD, Department of Preventive Medicine and Public Health, Faculty of Medicine, Fukuoka University,
Fukuoka 814-0180, Japan (e-mail: k-tanaka@fukuoka-u.ac.jp).
Copyright © 2011 by the Japan Epidemiological Association

72
 Tanaka K, et al. 73

which the family currently resides sponsors a physical ex- In a sensitivity analysis, we also used data on dental caries
amination that includes a dental examination, measurement of gathered at 18 months of age, which were recorded by a
height and weight, and an interview with parents or guardians dentist in the Maternal and Child Health Handbook when our
regarding the child’s health. Eligible subjects for the present subjects received their municipal physical examinations at
study were children aged 3 years who received this physical age 18 months. These data were transcribed by parents or
examination at any of the 7 public health centers offering it guardians from the Maternal and Child Health Handbook to
in Fukuoka City, a metropolitan area on Kyushu Island in our questionnaire.
southern Japan with a total population of approximately
1 414 000. During the period from June 2006 to January 2007, Exposure variables and covariates
we were granted permission by the Fukuoka City government Information on breastfeeding duration in months was obtained
to provide our questionnaires directly to parents or guardians from the structured self-administered questionnaire described
of children receiving the physical examination at age 3 years. above. Breastfeeding duration was the period during which
Out of the 8269 eligible children, the parents or guardians the infants received breast milk, regardless of exclusivity. The
of 8064 were provided with a structured self-administered questionnaire also included questions on the sex of the child,
questionnaire, a brief self-administered diet history ques- dental health practice (such as toothbrushing frequency, use
tionnaire (4 pages), and a postage-paid, addressed, return of fluoride, and pattern of dental care), between-meal snack
envelope. The structured self-administered questionnaire habits, maternal smoking during pregnancy, environmental
consisted of 16 pages: 13 pages for the 68 questions and 3 tobacco smoke (ETS) exposure at home, and paternal and
pages for data transcription from the Maternal and Child maternal educational levels. Use of fluoride was defined
Health Handbook. Ultimately, the parents or guardians of as positive if children reported using fluoride toothpaste or
2109 children answered the questionnaires and mailed receiving topical application of fluoride gel at a health center
these materials to the data management center (participation or a dental clinic.
rate = 25.5%). Our research technicians conducted telephone Data from the brief self-administered diet history ques-
interviews with individual participants when it was necessary tionnaire were not used in the current study.
to obtain missing data or clarify implausible responses.
The current study was restricted to subjects who provided Statistical analyses
complete information concerning the variables under study; Breastfeeding duration was classified into 4 categories
this left 2056 children available for analysis (24.9% of all (<6 months, 6–11 months, 12–17 months, and ≥18 months).
eligible children). Permission to perform this study was The following variables were controlled for in the multi-
obtained from the ethics committee of the Faculty of Medicine variable model: the sex of the child, toothbrushing frequency
at Fukuoka University. (<2 or ≥2 times/day), use of fluoride (yes or no), regular dental
check-ups (yes or no), between-meal snack frequency (<1, 1,
Outcome variable or ≥2 times/day), maternal smoking during pregnancy (yes or
During the physical examination, the presence of dental no), ETS exposure at home (yes or no), and paternal and
caries was assessed by visual examination without the use of maternal educational level (<13 years, 13–14 years, and ≥15
radiographs. Dental examination data were recorded by a years). In the sensitivity analysis using data on dental caries
dentist in the Maternal and Child Health Handbook provided at age 18 months, adjustment was made for child sex and
by the municipality during pregnancy, in which data parental educational levels only. Prevalence ratios (PRs) and
pertaining to prenatal checkups, postnatal health conditions 95% confidence intervals (CIs) were estimated using binomial
of both mother and baby, and growth of the child are regression with the log link function.24 Linear trend of the
recorded.23 The first half of the handbook provides space for association between duration of breastfeeding and dental
recording data on health condition and vaccination records, caries was assessed using a log-binomial regression model,
and the latter half provides information for mothers on treating the categories of breastfeeding duration as con-
pregnancy, delivery, and parenting. In our study, the parents or secutive integers. For tests of quadratic trend, we included
guardians of the children had to transcribe dental examination linear and quadratic terms in the model. Two-sided P
data from the Maternal and Child Health Handbook to our values less than 0.05 were considered to indicate statistical
self-administered questionnaire. To facilitate the transcription significance. All statistical analyses were performed using the
of the oral examination data to the self-administered ques- SAS software package version 9.1 (SAS Institute, Cary, NC,
tionnaire, we used exactly the same format as that used for the USA).
records in the Maternal and Child Health Handbook, and
parents or guardians transcribed all of the information,
RESULTS
including symbols. Children were classified as having dental
caries if 1 or more primary teeth had decayed, were missing, Of the 2056 children, 425 (20.7%) had dental caries. The
or had been filled. mean number of dental caries was 0.70. Toothbrushing 2 or

J Epidemiol 2012;22(1):72-77
74 Breastfeeding and Dental Caries

Table 1. Prevalence ratios (PRs) and 95% confidence Table 2. Prevalence ratios (PRs) and 95% confidence
intervals for dental caries in relation to study intervals for dental caries according to duration of
variables (unadjusted analysis) breastfeeding in 2056 children (FCHS, Japan)

Prevalence Crude PR (95% CI) Crude PR Adjusted PR

Prevalence
(95% CI) (95% CI)a
Sex
Female 184/969 (19.0%) 1.00 Breastfeeding duration (months)
<6 85/416 (20.4%) 1.00 1.00
Male 241/1087 (22.2%) 1.17 (0.98–1.39)
6–11 74/498 (14.9%) 0.73 (0.55–0.97) 0.79 (0.60–1.05)
Toothbrushing frequency (times/day)
12–17 90/583 (15.4%) 0.76 (0.58–0.99) 0.86 (0.66–1.13)
<2 252/1244 (20.3%) 1.00 ≥18 176/559 (31.5%) 1.54 (1.23–1.93) 1.66 (1.33–2.06)
≥2 173/812 (21.3%) 1.05 (0.89–1.25) P for linear trend <0.0001 <0.0001
Use of fluoride P for quadratic trend <0.0001 <0.0001
No 54/318 (17.0%) 1.00
Yes 371/1738 (21.4%) 1.26 (0.97–1.63) FCHS, Fukuoka Child Health Study.
Regular dental check-ups
a
Adjusted for sex, toothbrushing frequency, use of fluoride, regular
No 301/1166 (25.8%) 1.00 dental check-ups, between-meal snack frequency, maternal smoking
Yes 124/890 (13.9%) 0.54 (0.45–0.65) during pregnancy, exposure to environmental tobacco smoke at home,
Between-meal snack frequency (times/day) and paternal and maternal educational levels.
<1 78/444 (17.6%) 1.00
1 128/754 (17.0%) 0.97 (0.75–1.25)
≥2 219/858 (25.5%) 1.45 (1.15–1.83)
Table 2 shows the PRs and their 95% CIs for the
Maternal smoking during pregnancy
No 346/1787 (19.4%) 1.00 relationship between breastfeeding duration and the prev-
Yes 79/269 (29.4%) 1.52 (1.23–1.87) alence of dental caries. As compared with less than 6 months
ETS exposure at home of breastfeeding, breastfeeding for 18 months or longer was
No 205/1156 (17.7%) 1.00
Yes 220/900 (24.4%) 1.38 (1.16–1.63) significantly associated with a higher prevalence of dental
Paternal educational level (years) caries. After adjustment for sex, toothbrushing frequency,
<13 145/565 (25.7%) 1.00 use of fluoride, dental check-up history, between-meal snack
13–14 63/307 (20.5%) 0.80 (0.62–1.04)
≥15 217/1184 (18.3%) 0.71 (0.59–0.86)
frequency, maternal smoking during pregnancy, ETS exposure
Maternal educational level (years) at home, and paternal and maternal educational levels, the
<13 166/581 (28.6%) 1.00 positive association remained statistically significant (adjusted
13–14 151/823 (18.4%) 0.64 (0.53–0.78)
PR = 1.66, 95% CI: 1.33–2.06). A J-shaped relationship was
≥15 108/652 (16.6%) 0.58 (0.47–0.72)
observed between breastfeeding duration and dental caries:
ETS, environmental tobacco smoke.
the lowest PR was among children breastfed for 6 to 11
months (adjusted PR = 0.79, 95% CI: 0.60–1.05) and the
highest PR was for children breastfed for 18 months or longer
more times per day was reported in about 40% of the subjects. (adjusted PR = 1.66, 95% CI: 1.33–2.06, P for linear trend
Approximately 85% of children reported fluoride use, either <0.0001, P for quadratic trend <0.0001).
as fluoride toothpaste or topical application of fluoride gel We also conducted a sensitivity analysis using dental caries
at a health center or a dental clinic. Approximately 44% of at 18 months of age as the outcome variable. The prevalence
children received regular dental check-ups. More than 41% of of dental caries at age 18 months was 2.9% (n = 59). After
subjects had between-meal snacks 2 or more times per day. adjustment for sex and paternal and maternal educational
In utero exposure to maternal smoking occurred in 13.1% of levels, breastfeeding duration was significantly and positively
children, and 43.8% were exposed to ETS at home at least associated with dental caries in children at age 18 months. As
once. About 20% of children were breastfed for less than compared with children breastfed for less than 6 months, the
6 months, whereas 27.2% were breastfed for 18 months or adjusted PRs for children breastfed for 6 to 11, 12 to 17, and
longer. 18 months or longer were 1.52 (95% CI: 0.45–5.17), 2.57
Table 1 shows the results of bivariate analyses for selected (95% CI: 0.85–7.82), and 6.45 (95% CI: 2.30–18.11),
covariates. Regular dental check-ups were associated with respectively (P for linear trend <0.0001).
a lower prevalence of dental caries. In contrast, 2 or more
between-meal snacks per day was positively associated with
DISCUSSION
dental caries. With regard to smoking, both maternal smoking
during pregnancy and postnatal ETS exposure were associated The present study found that breastfeeding for 18 months or
with an increased prevalence of dental caries. Children with longer was significantly associated with a higher prevalence
fathers who had more than 15 years of education and those of dental caries. Our results partially agree with those of other
with mothers who had 13 to 14 years, or more than 15 years, studies showing an adverse effect of breastfeeding on dental
of education were less likely to have caries than children caries2,10,13–18 but are at variance with previous findings
with fathers or mothers who had less than 13 years of indicating a null or inverse association between breastfeeding
education. and dental caries.2–13 We observed a J-shaped relationship

J Epidemiol 2012;22(1):72-77
 Tanaka K, et al. 75

between duration of breastfeeding and the prevalence of longer breastfeeding may have confounded the observed
dental caries, ie, a nonsignificant inverse association was association. Mothers who breastfeed longer may be more
found between breastfeeding for 6 to 17 months and the likely to sleep with their child and breastfeed freely during
prevalence of dental caries. Nevertheless, in a sensitivity the night.17 A cross-sectional study of Brazilian preschool
analysis using data on dental caries at age 18 months, a children showed that the prevalence of early childhood caries
nonsignificant positive association was observed between was higher in children who were breastfed at night after
breastfeeding for 6 to 17 months and the prevalence of dental age 12 months than in children who stopped breastfeeding
caries, while breastfeeding for 18 months or longer was before age 12 months.18
significantly positively associated with the prevalence of Our study has several strengths. The study subjects were
dental caries. With regard to dental caries that had developed similar in age and geographic background, which likely
before age 18 months, the potential beneficial effects of reduced potential confounding produced by unmeasured
breastfeeding for 6 to 17 months might not be detected. factors related to age and geographic background. Data on
Alternatively, the results of the sensitivity analysis might have dental caries were obtained by means of oral examinations by
arisen by chance. dentists, and we were able to control for a variety of potential
A systematic review suggested that breastfeeding for confounding factors.
longer than 1 year, as well as nighttime breastfeeding This study also has some limitations that should be
after the eruption of teeth, is associated with some forms of considered. Of the 8269 eligible subjects in Fukuoka City,
early childhood caries, although the lack of methodological only 2057 (24.9%) were included in this analysis, so selection
consistency and the inconsistent definitions of caries and bias may be a factor. We were unable to assess differences
breastfeeding used in previous studies make it difficult to draw between participants and nonparticipants because, except
definitive conclusions.25 For example, some studies compared for age, no information on the personal characteristics of
breastfeeding for 13 months or longer with a period shorter nonparticipants was available. Thus, our subjects cannot be
than 13 months.7,16 Some studies investigated only the associ- considered representative of Japanese children in the general
ation between breastfeeding at night and dental caries.6,10,18 population, and the present findings should not be generalized.
One study compared outcomes according to whether In fact, the educational levels of parents in the present study
breastfeeding was ever performed or never performed.12 were higher than those of the general population. According
Inconsistency in the definition of dental caries, such as to the 2000 population census of Japan, the proportions of men
whether it comprises filled or missing teeth16; decayed, aged 35 to 39 years in Fukuoka City with less than 13 years,
missing, or filled teeth4–6,8,17; 2 or more decayed, missing, 13 to 14 years, 15 years or more, and unknown years of
or filled labial or palatal surfaces of primary incisors11; education were 39.6%, 8.0%, 43.3%, and 9.1%, respectively.28
or cavitated, filled, or missing smooth surfaces in primary The corresponding figures among fathers in the present
maxillary anterior teeth,18 also reduces the comparability of study were 27.5%, 14.9%, 57.6%, and 0.0%, respectively.
the available studies. Moreover, many studies were unable to The proportions of women 30 to 34 years of age in Fukuoka
control for important confounding factors such as oral hygiene City with less than 13 years, 13 to 14 years, 15 years or more,
practices,4,6–8,10,11,16,18 exposure to fluoride,4–8,10,11,14,16,18 and and unknown years of education were 41.3%, 34.4%, 16.1%,
socioeconomic status.4,8,10–12,16,18 and 8.3%, respectively.28 The corresponding figures among
We do not have a definitive explanation regarding the mothers in the present study were 28.3%, 40.1%, 31.7%, and
mechanisms that underlie our observations. Several minerals 0.0%, respectively. In addition, the prevalence of dental caries
in breast milk, such as phosphate and calcium, help protect in the study population (20.7%) was lower than that in a
tooth enamel. The mineral composition of breast milk changes sample of 3-year-old Japanese children assessed in a 2005
with advancing lactation, which may affect its cariogenic survey of dental diseases (24.4%).29
properties. In a longitudinal investigation, a significant decline The data on dental caries used in the present study were
was observed in the levels of phosphate and calcium in breast gathered during routine examinations by a number of dentists
milk over time, with concentrations of phosphate in breast at public health centers. The dentists were given detailed
milk at 3, 6, and 26 weeks of lactation of 14.7, 12.7, criteria for performing the examination but were not
and 10.7 mg/100 ml, respectively.26 Corresponding values for specifically trained so as to ensure standardization of their
calcium concentrations in breast milk during those time examinations. The unstandardized nature of the examinations
frames were 25.9, 27.7, and 24.8 mg/100 ml, respectively.26 could lead to nondifferential misclassification of caries and
Another possible explanation is that the prophylactic effects thus bias the results toward the null, that is, toward a lack of
of breast milk against dental caries (through the transfer association between breastfeeding and early childhood caries.
of maternal protective elements such as immunoglobulins, Moreover, because parents or guardians of the children
lactoferrin, and casein27 from mother to infant) decline due transcribed the data gathered at the dental examinations
to the gradual depletion of these elements after prolonged from their Maternal and Child Health Handbook to our self-
lactation.9 Alternatively, some unknown factors related to administered questionnaire, we cannot exclude the possibility

J Epidemiol 2012;22(1):72-77
76 Breastfeeding and Dental Caries

that transcription errors occurred. Nevertheless, misclassi-

fication of outcome is unlikely to differ across categories of REFERENCES
breastfeeding status. If this had occurred, the consequence 1. Gartner LM, Morton J, Lawrence RA, Naylor AJ, O’Hare D,
would have been an underestimation of values in our results. Schanler RJ, et al. Breastfeeding and the use of human milk.
Breastfeeding duration was assessed 3 years after the birth of Pediatrics. 2005;115:496–506.
the children, and this delay could have led to recall bias. 2. Sayegh A, Dini EL, Holt RD, Bedi R. Caries prevalence and
Most questionnaires, however, were completed by the child’s patterns and their relationship to social class, infant feeding and
mother, and it is unlikely that a mother would not remember oral hygiene in 4–5-year-old children in Amman, Jordan.
the breastfeeding history of her child. In a validation study, Community Dent Health. 2002;19:144–51.
breastfeeding duration reported by the mother has been shown 3. Saraiva MC, Bettiol H, Barbieri MA, Silva AA. Are intrauterine
growth restriction and preterm birth associated with dental
to be accurate for 20 years or even longer after the birth
caries? Community Dent Oral Epidemiol. 2007;35:364–76.
of a child.30 Even if errors in measurement of breastfeeding
4. al-Dashti AA, Williams SA, Curzon ME. Breast feeding, bottle
duration occurred, misclassification is unlikely to differ across feeding and dental caries in Kuwait, a country with low-fluoride
exposure categories. If this had occurred, the results would be levels in the water supply. Community Dent Health. 1995;
biased toward the null. 12:42–7.
Although we have adjusted our analyses for numerous 5. Iida H, Auinger P, Billings RJ, Weitzman M. Association
potential confounders, such as toothbrushing frequency, use of between infant breastfeeding and early childhood caries in the
fluoride, and parental educational levels, residual confounding United States. Pediatrics. 2007;120:e944–52.
effects cannot be ruled out. Additionally, it is possible that 6. Mohebbi SZ, Virtanen JI, Vahid-Golpayegani M, Vehkalahti
our results remain confounded by other potentially important MM. Feeding habits as determinants of early childhood caries
factors such as frequency of bedtime breastfeeding, age at in a population where prolonged breastfeeding is the norm.
introduction of foods and fluids other than breast milk, Community Dent Oral Epidemiol. 2008;36:363–9.
7. Campus G, Solinas G, Strohmenger L, Cagetti MG, Senna A,
and dietary intake (eg, sugar intake). Our study design was
Minelli L, et al. National pathfinder survey on children’s oral
cross-sectional, so the results of this analysis should not
health in Italy: pattern and severity of caries disease in 4-year-
be interpreted as providing evidence of a cause-effect olds. Caries Res. 2009;43:155–62.
relationship between breastfeeding and dental caries. 8. Rosenblatt A, Zarzar P. Breast-feeding and early childhood
In conclusion, we found a J-shaped association between caries: an assessment among Brazilian infants. Int J Paediatr
breastfeeding duration and the prevalence of dental caries Dent. 2004;14:439–45.
among young Japanese children. Our findings indicate that 9. Hallett KB, O’Rourke PK. Social and behavioural determinants
breastfeeding for 18 months or longer is associated with of early childhood caries. Aust Dent J. 2003;48:27–33.
a higher prevalence of dental caries. From the perspective 10. van Palenstein Helderman WH, Soe W, van ’t Hof MA. Risk
of avoiding any harmful impacts of breastfeeding on infant factors of early childhood caries in a Southeast Asian population.
dental health, the ideal duration of breastfeeding might J Dent Res. 2006;85:85–8.
be shorter than 18 months, given that a nonsignificant 11. Roberts GJ, Cleaton-Jones PE, Fatti LP, Richardson BD, Sinwel
RE, Hargreaves JA, et al. Patterns of breast and bottle feeding
inverse association was observed between breastfeeding for
and their association with dental caries in 1- to 4-year-old South
6 to 17 months and the prevalence of dental caries. Further,
African children. 2. A case control study of children with nursing
prospective, studies with clearly defined breastfeeding caries. Community Dent Health. 1994;11:38–41.
variables (eg, exclusive breastfeeding duration and 12. Ollila P, Larmas M. A seven-year survival analysis of caries
breastfeeding frequency) that also account for confounding onset in primary second molars and permanent first molars in
factors need to be undertaken before a definitive conclusion different caries risk groups determined at age two years. Acta
can be reached. Odontol Scand. 2007;65:29–35.
13. Du M, Bian Z, Guo L, Holt R, Champion J, Bedi R. Caries
patterns and their relationship to infant feeding and socio-
ACKNOWLEDGMENTS economic status in 2–4-year-old Chinese children. Int Dent J.
The authors would like to thank Drs. Masako Sakamoto, 2000;50:385–9.
Yasuko Arase, and Miki Nagano, the Fukuoka City 14. Vázquez-Nava F, Vázquez RE, Saldivar GA, Beltrán GF,
Almeida AV, Vázquez RC. Allergic rhinitis, feeding and oral
Government, and Prof. Tomoko Shirota of Nakamura
habits, toothbrushing and socioeconomic status. Effects on
Gakuen University, for their valuable support. Additionally,
development of dental caries in primary dentition. Caries Res.
we would like to thank all of the children and parents 2008;42:141–7.
who participated in this study. This study was supported by 15. Weber-Gasparoni K, Kanellis MJ, Levy SM, Stock J. Caries
Health and Labour Sciences Research Grants for Research on prior to age 3 and breastfeeding: a survey of La Leche League
Allergic Disease and Immunology from the Ministry of members. J Dent Child (Chic). 2007;74:52–61.
Health, Labour and Welfare, Japan. 16. Schluter PJ, Durward C, Cartwright S, Paterson J. Maternal self-
Conflicts of interest: None declared. report of oral health in 4-year-old Pacific children from South

J Epidemiol 2012;22(1):72-77
 Tanaka K, et al. 77

Auckland, New Zealand: findings from the Pacific Islands Epidemiol. 1998;27:91–5.
Families Study. J Public Health Dent. 2007;67:69–77. 25. Valaitis R, Hesch R, Passarelli C, Sheehan D, Sinton J. A
17. Hallett KB, O’Rourke PK. Early childhood caries and infant systematic review of the relationship between breastfeeding and
feeding practice. Community Dent Health. 2002;19:237–42. early childhood caries. Can J Public Health. 2000;91:411–7.
18. Azevedo TD, Bezerra AC, de Toledo OA. Feeding habits and 26. Greer FR, Tsang RC, Levin RS, Searcy JE, Wu R, Steichen JJ.
severe early childhood caries in Brazilian preschool children. Increasing serum calcium and magnesium concentrations in
Pediatr Dent. 2005;27:28–33. breast-fed infants: Longitudinal studies of minerals in human
19. Tanaka K, Miyake Y, Sasaki S. The effect of maternal smoking milk and in sera of nursing mothers and their infants. J Pediatr.
during pregnancy and postnatal household smoking on dental 1982;100:59–64.
caries in young children. J Pediatr. 2009;155:410–5. 27. Hamosh M. Protective function of proteins and lipids in human
20. Tanaka K, Miyake Y, Sasaki S. Intake of dairy products and the milk. Biol Neonate. 1998;74:163–76.
prevalence of dental caries in young children. J Dent. 2010; 28. Statistics Bureau, Ministry of Public Management, Home
38:579–83. Affairs, Posts and Telecommunications, Japan. 2000
21. Tanaka K, Miyake Y, Sasaki S. Association between Population Census of Japan. Vol. 3-2-40: Labour Force Status
breastfeeding and allergic disorders in Japanese children. Int J of Population, Industry (Major Group) of Employed Persons,
Tuberc Lung Dis. 2010;14:513–8. and Education. Fukuoka-ken. Tokyo: Statistics Bureau,
22. Tanaka K, Miyake Y. Association between prenatal and Ministry of Public Management, Home Affairs, Posts and
postnatal tobacco smoke exposure and allergies in young Telecommunications, Japan; 2002.
children. J Asthma. 2011;48:458–63. 29. Japanese Society for Dental Health. Statistics of Oral Health
23. Takayanagi K, Iwasaki S, Yoshinaka Y. The role of the Maternal 2007. Tokyo: Ishiyaku Publishers; 2007.
and Child Health Handbook system in reducing perinatal 30. Kark JD, Troya G, Friedlander Y, Slater PE, Stein Y. Validity of
mortality in Japan. Clin Perform Qual Health Care. 1993; maternal reporting of breast feeding history and the association
1:29–33. with blood lipids in 17 year olds in Jerusalem. J Epidemiol
24. Skov T, Deddens J, Petersen MR, Endahl L. Prevalence Community Health. 1984;38:218–25.
proportion ratios: estimation and hypothesis testing. Int J

J Epidemiol 2012;22(1):72-77