The aim of this study was to identify risk determinants leading to early childhood caries (ECC) and visible plaque (VP) in toddlers. Data for mother-child pairs participating in the Growing Up in Singapore towards Healthy Outcomes (GUSTO) birth cohort were collected from pregnancy to toddlerhood. Oral examinations were performed in 543 children during their clinic visit at 24 months to detect ECC and VP. Following logistic regression, ECC and VP were jointly regressed as primary and secondary outcomes, respectively, using the bivariate probit model. The ECC prevalence was 17.8% at 2 years of age, with 7.3% of children having a VP score >1. ECC was associated with nighttime breastfeeding (3 weeks) and biological factors, including Indian ethnicity (lower ECC rate), higher maternal childbearing age and existing health conditions, maternal plasma folate <6 ng/mL, child BMI, and the plaque index, while VP was associated with psychobehavioral factors, including the frequency of dental visits, brushing frequency, lower parental perceived importance of baby teeth, and weaning onto solids. Interestingly, although a higher frequency of dental visits and toothbrushing were associated with lower plaque accumulation, they were associated with increased ECC risk, suggesting that these established caries-risk factors may be a consequence rather than the cause of ECC. In conclusion, Indian toddlers may be less susceptible to ECC, compared to Chinese and Malay toddlers. The study also highlights a problem-driven utilization pattern of dental services (care sought for treatment) in Singapore, in contrast to the prevention-driven approach (care sought to prevent disease) in Western countries.

Despite major progress in caries control measures in the last few decades, dental caries in permanent dentition remains the most prevalent global condition [Kassebaum et al., 2015] and early childhood caries (ECC) has been reported to be the most common chronic childhood disease in the USA [US Surgeon General, 2000]. According to NHANES survey data, the prevalence of ECC significantly increased from 24% (1988-1994) to 28% (1999-2004) in 2- to 5-year-old American children [Dye et al., 2007]. Furthermore, a polarized distribution of ECC has been observed in the USA, with 25% of the population accounting for nearly 80% of the disease burden [Kaste et al., 1996]. Identification of the risk factors for ECC may be critical for timely caries prevention in toddlers.

Studies have reported high levels of cariogenic bacterial species [Zhou et al., 2012], diet [Palmer et al., 2010], oral hygiene status [Tanzer et al., 2001], insufficient maternal vitamin D levels during pregnancy [Schroth et al., 2014], and low socioeconomic status [Reisine and Psoter, 2001], among others, as risk determinants associated with ECC. Frequently, visible dental plaque has been demonstrated to be an important caries risk indicator in infants and toddlers [Zhou et al., 2012]. However, the contribution of early life risk factors to both visible plaque and ECC concomitantly in infants/toddlers remains uncertain.

Singapore is a small city-state in South-East Asia, with a multiethnic population (namely the Chinese, Malays, and Indians) displaying similar distributions in literacy levels and types of residential dwellings. This unique setting may be an ideal community in which to investigate the ethnic differences in ECC development. In spite of being a 100% fluoridated community in the past 6 decades, a 2005-2006 community-based study observed 40% caries prevalence and 43.6% incidence rates among 3- to 6-year-old children in Singapore [Gao et al., 2009], even worse than the aforementioned data from the USA, which currently has water fluoridation coverage of 74.7%, according to the CDC (Centres for Disease Control and Prevention) 2014 data. Previously, a skewed distribution of caries lesions has been demonstrated in Singaporean children, aged 3-5 years, showing 16% of high-risk children carrying 78% of the disease burden [Gao et al., 2009]. To trace the possible reasons for such a polarized disease distribution, the present study aimed to identify biopsychosocial and behavioral risk factors for both ECC (as the primary outcome) and visible plaque (VP; as a secondary outcome), using data collected from pregnancy until 2 years of age.

Study Design

Children and their biological mothers were recruited through the Growing Up in Singapore towards Healthy Outcomes (GUSTO) birth cohort study [Soh et al., 2014]. Healthy pregnant women in their first trimester (n = 1,237) were enrolled between June 2009 and September 2010, who intended to deliver at KK Women's and Children's Hospital (KKH) or National University Hospital (NUH) - the 2 major public maternity hospitals in Singapore - and who planned to reside in the country for the subsequent 5 years. Expectant mothers receiving chemotherapy or psychotropic drugs or who had type 1 diabetes mellitus were excluded from the study. Ethical approval was obtained from the Centralized Institutional Review Board (CIRB) of SingHealth (reference 2009/280/D) and the Domain-Specific Review Board (DSRB) of the Singapore National Healthcare Group (reference D/09/021). Written informed consent was obtained from all participants upon recruitment.

Data Collection

Interviewer-administered questionnaires were used to collect general and health-related information during pregnancy: the first visit by 14 weeks of gestation (recruitment) and a second at 26-28 weeks. During the latter visit, expectant mothers underwent blood tests, including a 75-g oral glucose tolerance test. The World Health Organization (WHO) criterion of a 2-h postload plasma glucose level >7.8 mmol/L was used to diagnose gestational diabetes [Alberti and Zimmet, 1998]. The cut-off points for adequate plasma micronutrient levels were selected on the basis of published studies: vitamin B6 (pyridoxal phosphate) = 20 nmol/L [Lui et al., 1985]; folate = 6 ng/mL [Herbert, 1987]; vitamin B12 = 300 pg/mL [Allen, 2008], and vitamin D = 50 nmol/L [Malabanan et al., 1998]. In addition to anthropometric measurements, information on sociodemographics, lifestyle, and maternal/infant health was collected during these visits. As maternal anthropometry differs across populations, the WHO criterion for Asian populations was used to classify women's body mass index (BMI) into 4 groups: underweight (<18.5); normal (18.5-22.9); overweight (23-27.5), and obese (>27.5) [WHO Expert Consultation, 2004].

Prenatal, labor/delivery and birth details, including child's sex, weight, gestational age, and neonatal complications were extracted from hospital case notes. For this study, infants were classified as small for gestational age if the birth weight was below the 10th percentile [Mikolajczyk et al., 2011]. Weights were measured using calibrated digital scales for both mothers (SECA 803, SECA Corp., Hamburg, Germany) and their children (SECA 334, SECA Corp.). Breast- and bottle-feeding practices, dietary patterns (including weaning), and child's oral/general health were ascertained through a detailed interviewer-administered questionnaire during each study visit. Children were followed up postnatally at 3 weeks and subsequently at 3- or 6-month intervals until 18 months of age before their 24-month clinical visit.

An interviewer-administered oral health questionnaire was administered by either a registered dentist or trained clinic staff during the 24-month visit, from March 2012 to June 2013, to collect information on the children's oral health practices (feeding patterns, diet habits, oral hygiene measures, and nonnutritive habits) and parental knowledge and attitudes on oral health. Children (n = 543) were examined by a single dentist, calibrated against a gold-standard examiner with representative cases (intraclass correlation coefficient >0.80). Children refrained from food, drink, and toothbrushing for at least 1 h before any oral examination.

Dental examinations, using the knee-to-knee position, were conducted using torchlights. Teeth were assessed mainly by visual inspection, using plane surface mouth mirrors, and aided by tactile inspection when deemed necessary. Autoclaved blunt WHO probes were used to avoid damaging sound enamel surfaces. Prior to the drying/cleaning the tooth surfaces for dental examination (using the ISO-2950 notation system), the oral hygiene status was determined using the modified Silness-Löe plaque index [Silness and Löe, 1964]. Six index primary teeth (55, 52, 64, 75, 72, and 84) were evaluated based on a 4-point scale, with the oral hygiene status considered as “excellent” to “good” for average plaque scores (of 6 teeth) less than 1. When a designated tooth had not yet erupted, an alternative tooth in the same quadrant was taken as a substitute. Teeth were considered present in the oral cavity when any portion of the tooth was visible. A modified International Caries Detection and Assessment System (ICDAS II) was used for ECC detection [Ismail et al., 2007]. Since a certain amount of difficulty and uncertainty has previously been reported in differentiating between active and inactive lesions in a single appointment [Ekstrand et al., 2005], the lesions detected in the present study cannot be accurately ascertained to be active lesions only. However, satisfactory interexaminer and intraexaminer reliability scores were noted, with scores for codes 1-2 being 0.76 and 0.95, respectively, and for codes 3-6 being 1.00 and 1.00, respectively. Since visual examination by ICDAS criteria requires the cleaning of teeth for debris/plaque removal, to detect underlying caries lesions, the first examiner recorded the plaque index and hence, inter-/intraexaminer reliability for plaque scores could not be conducted. Sterile gauze was used to remove debris and dry tooth surfaces. No additional detection methods, including radiographs, were used owing to constraints of the field settings and ethical considerations.

Statistical Analysis

The initial sample size calculation was based on the hypothesized odds ratio (OR) and base line proportion of positive outcome. We assumed that the minimal detectable effect size (OR) would be 2.0, and the proportion of ECC (primary outcome of interest) was set at 20%. Utilizing a test for 2 proportions with the 2-sided Fisher exact test, a total of 378 subjects were calculated to be required after taking into account a 10% attrition rate. This was met with the target recruitment of the GUSTO study set at 1,200 pregnant women (with an ethnic distribution of approx. 60% Chinese, 23% Malay, and 17% Indian) and a total of 587 children were recruited for the oral health study during the 24-month clinic visit. The primary outcome variable “caries present” for each child was defined as the presence of white spot/cavitated lesions observed on any tooth surface (ICDAS II codes 1-6) during the 24-month clinic visit. The secondary outcome variable “high plaque index” was defined as a score >1 (moderate to abundant accumulation of plaque within the gingival pocket and/or tooth and gingival margin) calculated from the Silness-Löe plaque index conducted during the same visit.

The statistical analysis was performed using STATA software version 14. Potential risk determinants for ECC were screened using bivariate analysis (with a p = 0.1 as the cut-off point) in order to assess each variable individually and not to rule out any clinically important factors (Table 1). To choose among competing models, the preferred logistic regression model was selected based on the log likelihood ratio, numbers of significant variables in the model, as well as the Hosmer-Lemeshow goodness-of-fit test. The seemingly unrelated bivariate probit model (SUBP) was used to investigate the independent risk determinants for both the primary (ECC) and secondary categorical outcomes (VP), in which the probability of ECC manifestation and high plaque index were modelled jointly using maximum likelihood estimates with robust standard error estimates. Additionally, the statistically significance of the likelihood-ratio test of rho was assessed to ensure the model was properly estimated. The sociodemographic, prenatal, perinatal, postnatal, and 24-month oral health-related variables were also adjusted in the multivariate SUBP model. A backward stepwise regression procedure was employed.

Table 1

Potential risk determinants of early childhood caries

Potential risk determinants of early childhood caries
Potential risk determinants of early childhood caries

The Wald χ2 test, which is reflected by the statistical significance of p, was used to determine whether the final fitted model would be best estimated jointly in a recursive manner or not. Additionally, the statistical significance of the likelihood-ratio test of rho was assessed to ensure the model was properly estimated with the statistical method utilized. The modified Hosmer-Lemeshow goodness-of-fit test was also applied to check the SUBP model. The final effect size of independent significant risk factors was quantified using the OR estimate and respective associated 95% confidence intervals (CI). A 2-sided p value <0.05 was considered statistically significant.

Sample Characteristics

Out of the 1,237 GUSTO participants, a total of 543 children underwent oral examinations during the 24-month visit. Examination records with incomplete data (for Silness-Löe plaque index and/or ICDAS scoring) were excluded from analysis (n = 8). A total of 535 children were included in the analysis (274 boys and 261 girls), of which 54.8% were Chinese, 28.8% Malay, and 16.4% Indian. The ethnic distribution was representative of the main GUSTO cohort, including 55.8% Chinese, 26.2% Malay, and 18.0% Indians. There were no statistically significant sociodemographic differences between the analyzed and missing/dropout subjects. Initially, we used the full data set available for the statistical modelling, but only 293 data points (n = 293) were on the fitted line, as the rest of them (with missing data points) contributed to error terms in the model (not fitted portion). Hence, a total of 296 mother-child pairs were excluded from the final fitted model. As the result of the Hosmer-Lemeshow goodness of fit test was not significant (p = 0.719), we accepted the hypothesis that our model is well fitted. The prevalence of ECC in the study population at 2 years was 17.8%, with 3.9% of children exhibiting 1 or more cavitated lesions. The mean Silness-Löe plaque index was 0.22 ± 0.45, with 7.3% of children having a score above 1. For the noncavitated lesions (ICDAS), the mean numbers of affected teeth and surfaces were 0.50 ± 1.47 and 0.63 ± 2.26, respectively. The upper central incisors (tooth 51 = 10.1% and tooth 61 = 9.9%) were the most commonly affected (Fig. 1a), followed by the upper lateral incisors (tooth 52 = 6% and tooth 62 = 5.6%) and lower first molars (tooth 74 = 3.4% and tooth 84 = 3.9%), with as high as 7.9% of buccal (incisors) and 3.0% of occlusal (molars) surfaces affected (Fig. 1b).

Fig. 1

ECC distribution per tooth type (a) and tooth surface (b).

Fig. 1

ECC distribution per tooth type (a) and tooth surface (b).

Close modal

For the cavitated lesions, the mean numbers of affected teeth and surfaces were 0.12 ± 0.84 and 0.19 ± 1.43, respectively. The upper central incisors (tooth 51 = 3% and tooth 61 = 2.2%) were the most commonly affected (Fig. 1a), followed by upper lateral incisors (tooth 52 = 0.9% and tooth 62 = 1.5%) and upper first molars (tooth 54 = 0.7% and tooth 64 = 0.9%), with as high as 1.5% of buccal (incisors) and 0.4% of occlusal (molar) surfaces affected (Fig. 1b).

Identification of Caries Risk Determinants

Several significant caries risk and protective factors were identified via the conventional stepwise logistic regression procedure (Table 2). Some variables in the final model with p > 0.05 (e.g., nighttime bottle feeding), were added to the models since they are well-established risk factors/confounders for either or both outcomes. The risk of caries at 2 years of age was associated with Chinese ethnicity, increased maternal age (34 years and above), nighttime breastfeeding (within the first 3 weeks of life), higher child BMI, child's dental visits (from 9 to 24 months of age), daily brushing frequency of child's teeth (at least once per day), and 2-year high plaque index (>1; all p < 0.05). In contrast, a higher maternal education level (undergraduate education and above), maternal health condition prior to/during pregnancy (immune system disorders being the most prevalent, at 34.6%), and those with normal prenatal plasma folate levels (≥6 ng/mL) were protective factors against ECC (all p < 0.05).

Table 2

Identification of caries risk determinants in toddlers using multivariate logistic regression (final-fitted model)

Identification of caries risk determinants in toddlers using multivariate logistic regression (final-fitted model)
Identification of caries risk determinants in toddlers using multivariate logistic regression (final-fitted model)

All ECC risk determinants identified via the conventional logistic model remained statistically significant in the bivariate probit model (Table 3), except maternal education level (p > 0.05 for both ECC and VP). Additionally, a higher plaque accumulation (plaque index above 1) was demonstrated in children who had a delayed weaning onto solids (9 months), child's dental visits, child's brushing frequency, and parental belief of the importance of baby teeth (all p < 0.05). Although the daily brushing frequency of a child's teeth (≥ once per day) and a child's dental visits from 9 to 24 months of age were associated with a lower plaque score, these 2 factors were associated with an increased ECC risk.

Table 3

Identification of caries risk determinants in toddlers using seemingly unrelated bivariate probit regression (final-fitted model)

Identification of caries risk determinants in toddlers using seemingly unrelated bivariate probit regression (final-fitted model)
Identification of caries risk determinants in toddlers using seemingly unrelated bivariate probit regression (final-fitted model)

This study identified potential ECC risk determinants by jointly modelling both ECC and VP in 2-year-old children using a life-course approach. The rate of ECC (including white-spot lesions) in 2-year-old children (17.8%), observed in this study is lower compared to ECC in Switzerland (25.3%) [Menghini et al., 2008] and China (27.6%) [Zhou et al., 2011]. The low cavitation rate (3.9%) seemingly corroborates with previously reported findings on caries susceptibility as being low during the first posteruptive year but rising rapidly 2-3 years following tooth eruption [Carlos and Gittelsohn, 1965]. Moreover, the intraoral caries distribution pattern observed in the present study is parallel with the previously reported data [Gao et al., 2009], confirming the important role of site specificity in the disease pathogenic pathway.

Interestingly, the findings of our study showed the predominance of biological factors associated with ECC in contrast to the psychobehavioral factors associated with VP in 2-year-old children. These findings, in addition to a previous study showing the association of both biological and psychobehavioral factors with ECC in 3- to 6-year-old Singaporean children [Gao et al., 2010], indicate the major role of biological risk factors early in life.

In the present study, Indian toddlers were shown to be at a lower risk for ECC development when compared to Chinese toddlers, although no significant differences were observed in their plaque index. Preschoolers of Indian ethnicity have been reported to have a lower caries rate than Malay preschoolers [Gao et al., 2010]. Ethnic differences in maternal dietary practices during pregnancy were observed in another study of the same GUSTO cohort [Chen et al., 2014], in which Indian mothers were observed to consume more cheese and yoghurt compared to the other 2 ethnic groups, which may be one of the reasons associated with lower caries experience in Indian children [Tanaka et al., 2012]. Moreover, ethnic differences in micronutrient levels during pregnancy [Loy et al., 2015], also observed in our GUSTO cohort previously, may potentially play a role in caries susceptibility; nevertheless, future studies to elucidate the specific mechanism(s) are warranted.

Our findings demonstrated that a preexisting maternal health condition was protective towards ECC development in 2-year-old children. A previous study reported lower ECC experience in 3- to 6-year-old Singaporean children with existing health conditions [Gao et al., 2010]. It is speculated that the health conditions of mother or child may heighten the health consciousness and enhance certain health behaviors [Graham et al., 2016]. The present study also showed children born to mothers with a higher childbearing age (≥34 years) to have higher caries rates when compared to their counterparts. Older mothers have been reported to be associated with infants with low birthweight [Martin et al., 2010] and low birthweight has been shown to be a risk factor for enamel defects in primary dentition [Vello et al., 2010], possibly increasing the caries susceptibility observed in our study.

Among all the maternal micronutrients investigated, we observed insufficient maternal plasma folate levels (<6 ng/mL) to be a risk factor for ECC in the offspring. It is suggested that lower folate concentrations during pregnancy may perturb DNA methylation by modulating the synthesis of methionine and its derivative, S-adenosyl-methione [Bergen et al., 2012], causing higher homocysteine concentrations and increased salivary oxidative stress, potentially contributing to the initiation and progression of dental caries in mother and subsequently in infants/toddlers [Kumar et al., 2011]. It is speculated that a window period may potentially exist during which folate levels may be crucial for optimizing the caries-protective effects of the developing immune system and fetal tissue factors during early life.

A higher BMI (at age 2 years) was associated with caries experience in 2-year-old children in our study. This link has been controversial in the literature thus far. While positive correlations between caries risk and higher BMI have been reported [Vazquez-Nava et al., 2010], other studies demonstrated no association [Hong et al., 2008; Wu et al., 2013] or even an inverse relationship [Sanchez-Perez et al., 2010; Norberg et al., 2012]. Additionally, it has been postulated that there may be a U-shaped trend rather than a linear one, as the prevalence of dental caries may be higher among individuals with extremely higher or lower BMI [Hooley et al., 2012]. As the relationship between dental caries and childhood obesity may be age specific, further studies may be needed to thoroughly characterize this potential association.

A higher frequency of nighttime breastfeeding (during the first 3 weeks) was identified as a risk factor in this study. A plausible explanation may be that nighttime breast-feeding within the first 3 weeks of life may enhance the vertical transmission/ inoculation of cariogenic bacteria during the predentate stage [Plonka et al., 2012], paving the way for cariogenic biofilm to be established at a later stage and thus increasing the child's future susceptibility to dental caries. Previous studies investigating the overall impact of breastfeeding on dental caries have yielded contradicting results, with some studies showing a significant age-specific association [Tham et al., 2015] and others showing inconsistent results [Valaitis et al., 2000]. Similarly, our data did not identify any association between ECC and nocturnal breastfeeding (>9 months), overall breastfeeding, and nighttime bottle-feeding. The relationship between breastfeeding and dental caries may be nonlinear and age specific [Valaitis et al., 2000; Tanaka and Miyake, 2012], partially due to the timing of tooth eruption and the calcium-phosphate composition of breast milk [Featherstone, 2008], which may decrease with progressive lactation, thus increasing its cariogenic properties. Moreover, prolonged breastfeeding practices, resulting in delayed weaning onto solids, may result in increased plaque accumulation as shown in our study.

Interestingly, the frequency and quantity of sweet confectionery or beverages was unrelated to ECC at 2 years of age in the present study. A similar finding has previously been observed in 2-year-old Chinese children [Zhou et al, 2012]. We speculate that since ECC develops as a precavitation condition, which takes time to manifest itself in the oral cavity, the snack intake quantity and frequency may not be directly associated with ECC as early as 2 years of age. As illustrated in the Tables, ECC in toddlers is more associated with biological factors, with environmental and/or behavioral factors affecting plaque deposition mainly at this age, which may subsequently lead to the development of caries lesions at a later age.

The American Academy of Pediatrics (AAP) advocates oral assessment in high-risk infants to be carried out by a dentist no later than 6 months after eruption of the first tooth, or by 12 months of age, whichever comes first [Hale, 2003]. Early dental visits have been associated with low caries experience in preschool children in the US [Beil et al., 2014]. However, only 2.6% of the GUSTO toddlers had been to a dentist prior to this study, and only 6% of the primary caregivers of children during the 2-year clinic visit reported the period within 1.5 years to be a suitable timing for the first dental visit. It is likely that parents consider infant teeth to be of low importance and would only seek dental treatment for their children upon symptoms of pain, fever, and/or suspected conditions. Moreover, history of previous dental visits has been shown to be associated with higher ECC experience in preschool children in China [Zhou et al., 2012], similar to the findings of our study. Additionally, higher dental visits were shown to significantly reduce the VP accumulation in the present study, indicating that clinical visits were actually a result of a child having ECC, rather than a risk factor, and denoting a problem-driven utilization of dental services among Asians compared to the prevention-based dental visits in other developed countries like the USA [Ramos-Gomez et al., 2010].

A strong correlation has been reported in the literature between oral hygiene habits and caries prevalence in primary dentition, where the higher frequency of tooth brushing has been linked to lower caries experience in the child [Grindefjord et al., 1995; Gibson and Williams, 1999]. However, in our study, although the child's brushing frequency was associated with reduced plaque accumulation (OR = 0.62), it was associated with increased caries risk (OR = 1.88). It is likely that caregivers reinforced the frequent brushing in children after seeing their decayed teeth or diagnosis/warning from dentists, parallel to the association between dental visits and ECC experience. In this study, these 2 established caries risk protective factors turned out to be the consequences and thus, indicators of disease rather than the cause of the disease. Furthermore, the presence of dental plaque has been suggested to be an indicator of early stages of ECC and its severity [Declerck et al., 2008; Parisotto et al., 2010]. In our study, VP was found to be a strong risk factor for ECC development at 2 years of age. These findings emphasize the need for effective plaque control measures to be in place at an early age to prevent ECC development in young children, which may be influenced by the primary caregivers' perception of their importance [Finlayson et al., 2007].

One of the strengths of this study is the utilization of a bivariate probit model to estimate ECC, together with VP accumulation, a known clinical caries risk indicator in young children. Secondly, the disease characteristics and distribution were described in 3 Asian ethnic groups, which reflect more than 40% of the global population. The study had its limitations as well. As with most of the longitudinal cohort studies with substantial missing data across time points, some of the subgroup analysis may be underpowered in our study. Our findings in a multiethnic Asian cohort sample might not be generalizable to other populations.

In conclusion, our study identified Indian toddlers to have significantly lower caries rates compared to Chinese and Malays. This study also highlighted the specific problem-driven utilization of dental services in Singapore, in contrast to the prevention-driven approach in Western countries.

This study was funded by Singapore's National Medical Research Council (NMRC/CIRG/1341/2012: R-221-000-059-511) and supported by the Singapore National Research Foundation under its TCR Flagship Programme on Developmental Pathways to Metabolic Disease (NMRC/TCR/004-NUS/2008 and NMRC/TCR/012- 438 NUHS/2014). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Special thanks are extended to the GUSTO subdomains for their great support; Rahul Nair, David Koh, Bindu Karunakaran, Chan Yiong Huak, Tan Pei Ting, Pang Wei Wei, Izzuddin Bin Mohd Aris, and Priyangi Alwis for their valuable contributions. K.G. is supported by the National Institute for Health Research through the NIHR Southampton Biomedical Research Centre and by the European Union's Seventh Framework Programme (FP7/2007-2013), projects EarlyNutrition and ODIN under grant agreement numbers 289346 and 613977. The continuous help of the home visitors and the clinical team from the KK Women's and Children's Hospital and the National University Hospital is greatly appreciated.

K. Godfrey, Y.S. Lee, and Y.S. Chong have received reimbursement for speaking at conferences sponsored by companies selling nutritional products. K. Godfrey and Y.S. Chong are part of an academic consortium that has received research funding from Abbot Nutrition, Nestec, and Danone. The other authors declare that they have no competing interests.

The study design was conceived and designed by R.Y., Y.S.L, M.F-F.C, K.K, S.M.S, K.G, Y.S.C, and C-Y.H. The clinical examination was performed by C.U.L. Data analysis was performed by L.W.K and C.U.L. The manuscript was prepared by C.U.L, A.C.K, and C-Y.H. All authors reviewed and revised the final manuscript.

Alberti KG, Zimmet PZ: Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998;15:539-553.
Allen LH: Causes of vitamin B12 and folate deficiency. Food Nutr Bull 2008;29:S20-S37.
Beil H, Rozier RG, Preisser JS, Stearns SC, Lee JY: Effects of early dental office visits on dental caries experience. Am J Pub Health 2014;104:1979-1985.
Bergen NE, Jaddoe VW, Timmermans S, Hofman A, Lindemans J, Russcher H, Raat H, Steegers-Theunissen RP, Steegers EA: Homocysteine and folate concentrations in early pregnancy and the risk of adverse pregnancy outcomes: the Generation R Study. BJOG 2012;119:739-751.
Carlos JP, Gittelsohn AM: Longitudinal studies of the natural history of caries. II. A life-table study of caries incidence in the permanent teeth. Arch Oral Biol 1965;10:739-751.
Chen LW, Low YL, Fok D, Han WM, Chong YS, Gluckman P, Godfrey K, Kwek K, Saw SM, Soh SE, Tan KH, Chong MF, van Dam RM: Dietary changes during pregnancy and the postpartum period in Singaporean Chinese, Malay and Indian women: the GUSTO birth cohort study. Public Health Nutr 2014;17:1930-1938.
Declerck D, Leroy R, Martens L, Lesaffre E, Garcia-Zattera MJ, Vanden Broucke S, Debyser M, Hoppenbrouwers K: Factors associated with prevalence and severity of caries experience in preschool children. Community Dent Oral Epidemiol 2008;36:168-178.
Dye BA, Tan S, Smith V, Lewis BG, Barker LK, Thornton-Evans G, Eke PI, Beltran-Aguilar ED, Horowitz AM, Li CH: Trends in oral health status: United States, 1988-1994 and 1999-2004. Vital Health Stat 2007;11:1-92.
Ekstrand KR, Ricketts DN, Longbottom C, Pitts NB: Visual and tactile assessment of arrested initial enamel carious lesions: an in vivo pilot study. Caries Res 2005;39:173-177.
Featherstone JD: Dental caries: a dynamic disease process. Aust Dent J 2008;53:286-291.
Finlayson TL, Siefert K, Ismail AI, Sohn W: Maternal self-efficacy and 1-5-year-old children's brushing habits. Community Dent Oral Epidemiol 2007;35:272-281.
Gao XL, Hsu CY, Loh T, Koh D, Hwamg HB, Xu Y: Dental caries prevalence and distribution among preschoolers in Singapore. Community Dent Health 2009;26:12-17.
Gao XL, Hsu CY, Xu Y, Hwarng HB, Loh T, Koh D: Building caries risk assessment models for children. J Dent Res 2010;89:637-643.
Gibson S, Williams S: Dental caries in pre-school children: associations with social class, toothbrushing habit and consumption of sugars and sugar-containing foods. Further analysis of data from the National Diet and Nutrition Survey of children aged 1.5-4.5 years. Caries Res 1999;33:101-113.
Graham M, Uesugi K, Olson C: Barriers to weight-related health behaviours: a qualitative comparison of the socioecological conditions between pregnant and post-partum low-income women. Matern Child Nutr 2016;12:349-361.
Grindefjord M, Dahllof G, Nilsson B, Modeer T: Prediction of dental caries development in 1-year-old children. Caries Res 1995;29:343-348.
Hale KJ: Oral health risk assessment timing and establishment of the dental home. Pediatrics 2003;111:1113-1116.
Herbert V: Making sense of laboratory tests of folate status: folate requirements to sustain normality. Am J Hematol 1987;26:199-207.
Hong L, Ahmed A, McCunniff M, Overman P, Mathew M: Obesity and dental caries in children aged 2-6 years in the United States: National Health and Nutrition Examination Survey 1999-2002. J Public Health Dent 2008;68:227-233.
Hooley M, Skouteris H, Boganin C, Satur J, Kilpatrick N: Body mass index and dental caries in children and adolescents: a systematic review of literature published 2004 to 2011. Syst Rev 2012;1:57.
Ismail AI, Sohn W, Tellez M, Amaya A, Sen A, Hasson H, Pitts NB: The International Caries Detection and Assessment System (ICDAS): an integrated system for measuring dental caries. Community Dent Oral Epidemiol 2007;35:170-178.
Kassebaum NJ, Bernabe E, Dahiya M, Bhandari B, Murray CJ, Marcenes W: Global burden of untreated caries: a systematic review and metaregression. J Dent Res 2015;94:650-658.
Kaste LM, Selwitz RH, Oldakowski RJ, Brunelle JA, Winn DM, Brown LJ: Coronal caries in the primary and permanent dentition of children and adolescents 1-17 years of age: United States, 1988-1991. J Dent Res 1996;75:631-641.
Kumar D, Pandey RK, Agrawal D: An estimation and evaluation of total antioxidant capacity of saliva in children with severe early childhood caries. Int J Paediatr Dent 2011;21:459-464.
Loy SL, Lek N, Yap F, Soh SE, Padmapriya N, Tan KH, Biswas A, Yeo GS, Kwek K, Gluckman PD, Godfrey KM, Saw SM, Muller-Riemenschneider F, Chong YS, Chong MF, Chan JK: Association of maternal vitamin D status with glucose tolerance and caesarean section in a multi-ethnic Asian cohort: the Growing Up in Singapore towards Healthy Outcomes Study. PLoS One 2015;10:e0142239.
Lui A, Lumeng L, Aronoff GR, Li TK: Relationship between body store of vitamin B6 and plasma pyridoxal-P clearance: metabolic balance studies in humans. J Lab Clin Med 1985;106:491-497.
Malabanan A, Veronikis IE, Holick MF: Redefining vitamin D insufficiency. Lancet 1998;351:805-806.
Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Mathews TJ, Osterman MJ: Births: final data for 2008. Natl Vital Stat Rep 2010;59:3-71.
Menghini G, Steiner M, Thomet E, Roos M, Imfeld T: Caries prevalence in 2-year-old children in the city of Zurich. Community Dent Health 2008;25:154-160.
Mikolajczyk RT, Zhang J, Betran AP, Souza JP, Mori R, Gulmezoglu AM, Merialdi M: A global reference for fetal-weight and birthweight percentiles. Lancet 2011;377:1855-1861.
Norberg C, Hallstrom Stalin U, Matsson L, Thorngren-Jerneck K, Klingberg G: Body mass index (BMI) and dental caries in 5-year-old children from southern Sweden. Community Dent Oral Epidemiol 2012;40:315-322.
Palmer CA, Kent R, Loo CY, Hughes CV, Stutius E, Pradhan N, Dahlan M, Kanasi E, Arevalo Vasquez SS, Tanner ACR: Diet and caries-associated bacteria in severe early childhood caries. J Dent Res 2010;89:1224-1229.
Parisotto TM, Steiner-Oliveira C, Duque C, Peres RC, Rodrigues LK, Nobre-dos-Santos M: Relationship among microbiological composition and presence of dental plaque, sugar exposure, social factors and different stages of early childhood caries. Arch Oral Biol 2010;55:365-373.
Plonka KA, Pukallus ML, Barnett AG, Walsh LJ, Holcombe TH, Seow WK: Mutans streptococci and lactobacilli colonization in predentate children from the neonatal period to seven months of age. Caries Res 2012;46:213-220.
Ramos-Gomez FJ, Crystal YO, Ng MW, Crall JJ, Featherstone JD: Pediatric dental care: prevention and management protocols based on caries risk assessment. J Calif Dent Assoc 2010;38:746-761.
Reisine ST, Psoter W: Socioeconomic status and selected behavioral determinants as risk factors for dental caries. J Dent Educ 2001;65:1009-1016.
Sanchez-Perez L, Irigoyen ME, Zepeda M: Dental caries, tooth eruption timing and obesity: a longitudinal study in a group of Mexican schoolchildren. Acta Odontol Scand 2010;68:57-64.
Schroth RJ, Lavelle C, Tate R, Bruce S, Billings RJ, Moffatt MEK: Prenatal vitamin D and dental caries in infants. Pediatrics 2014;133:e1277-e1284.
Silness J, Löe H: Periodontal disease in pregnancy. II. Correlation between oral hygiene and periodontal condition. Acta Odontol Scand 1964;22:121-135.
Soh SE, Tint MT, Gluckman PD, Godfrey KM, Rifkin-Graboi A, Chan YH, Stunkel W, Holbrook JD, Kwek K, Chong YS, Saw SM: Cohort profile: Growing Up in Singapore towards Healthy Outcomes (GUSTO) birth cohort study. Int J Epidemiol 2014;43:1401-1409.
Tanaka K, Miyake Y: Association between breastfeeding and dental caries in Japanese children. J Epidemiol 2012;22:72-77.
Tanaka K, Miyake Y, Sasaki S, Hirota Y: Dairy products and calcium intake during pregnancy and dental caries in children. Nutr J 2012;11:33.
Tanzer JM, Livingston J, Thompson AM: The microbiology of primary dental caries in humans. J Dent Educ 2001;65:1028-1037.
Tham R, Bowatte G, Dharmage SC, Tan DJ, Lau MX, Dai X, Allen KJ, Lodge CJ: Breastfeeding and the risk of dental caries: a systematic review and meta-analysis. Acta Paediatr (Oslo) 2015;104:62-84.
US Surgeon General: Oral health in America: a report of the Surgeon General. J Calif Dent Assoc 2000;28:685-695.
Valaitis R, Hesch R, Passarelli C, Sheehan D, Sinton J: A systematic review of the relationship between breastfeeding and early childhood caries. Can J Public Health 2000;91:411-417.
Vazquez-Nava F, Vazquez-Rodriguez EM, Saldivar-Gonzalez AH, Lin-Ochoa D, Martinez-Perales GM, Joffre-Velazquez VM: Association between obesity and dental caries in a group of preschool children in Mexico. J Pub Health Dent 2010;70:124-130.
Vello MA, Martinez-Costa C, Catala M, Fons J, Brines J, Guijarro-Martinez R: Prenatal and neonatal risk factors for the development of enamel defects in low birth weight children. Oral Dis 2010;16:257-262.
WHO Expert Consultation: Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004;363:157-163.
Wu L, Chang R, Mu Y, Deng X, Wu F, Zhang S, Zhou D: Association between obesity and dental caries in Chinese children. Caries Res 2013;47:171-176.
Zhou Y, Lin HC, Lo EC, Wong MC: Risk indicators for early childhood caries in 2-year-old children in southern China. Aust Dent J 2011;56:33-39.
Zhou Y, Yang JY, Lo EC, Lin HC: The contribution of life course determinants to early childhood caries: a 2-year cohort study. Caries Res 2012;46:87-94.
Copyright / Drug Dosage / Disclaimer
Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.