Determinants of early childhood caries among pre-school children

¹School of Dentistry, Hanoi Medical University, Hanoi, Vietnam, ²University of Medicine and Pharmacy, Vietnam National University, Hanoi, Vietnam, ³Da Nang University of Medical Technology and Pharmacy, Da Nang, Vietnam, ⁴Dong Da General Hospital, Hanoi, Vietnam, ⁵School of Preventive Medicine and Public Health, Hanoi Medical University, Hanoi, Vietnam

^&Corresponding author
Tran Thi My Hanh, School of Dentistry, Hanoi Medical University, Hanoi, Vietnam

Introduction    

Early childhood caries (ECC) is one of the most prevalent chronic conditions affecting young children worldwide and remains a leading cause of pain, infection, impaired nutrition, and reduced quality of life [1,2]. ECC reflects a complex interaction between biological susceptibility, behavioural exposures, and broader social and environmental determinants. According to the definition aligned with the American Academy of Pediatric Dentistry and the World Health Organization, ECC is diagnosed when any primary tooth surface in a child aged ≤71 months is decayed (cavitated or non-cavitated), missing due to caries, or filled as a result of caries experience [2-5]. Severe early childhood caries (S-ECC) represents a more advanced and clinically consequential form of the disease and is defined by age-specific dmfs thresholds or the presence of any smooth-surface caries in children younger than three years [2,6,7].

Globally, ECC affects a substantial proportion of preschool children, with estimates indicating that approximately 46-57% of children experience the disease, and markedly higher rates are consistently reported in socially and economically disadvantaged populations [1,3,8]. Considerable geographic variation has been documented, with pooled ECC prevalence of 61.7% reported in an Iranian meta-analysis spanning three decades (Jabbarian et al. in 2025) and national survey data indicating 53.2% prevalence with a mean dmft of 2.7 among children aged 1-5 years [4]. In contrast, ECC prevalence has been reported as low as 19% in Southern Italy, with S-ECC accounting for 2.7% of cases [9], while substantially higher prevalence has been observed in other settings, such as 75% among 5-year-olds in southwest China (mean dmft 4.7) [10] and approximately 45% caries experience with mean dmft 2.51 among Tanzanian preschoolers [11]. These differences highlight the influence of contextual factors, including diet, oral hygiene behaviours, access to dental care, and socio-economic conditions.

Despite the growing body of international evidence on ECC prevalence and determinants, data from Vietnam remain limited and fragmented, particularly regarding disease severity and the role of family-level factors and parental knowledge, attitudes, and practices. Existing studies in similar low- and middle-income contexts indicate that ECC is strongly shaped by modifiable behaviours such as sugar consumption, oral hygiene practices, and feeding patterns, as well as by parental education, health literacy, and residential environment [1,4,12]. However, there is a lack of comprehensive evidence integrating clinical measures of caries severity with family and caregiver characteristics in Vietnamese preschool children. Therefore, this study aimed to (1) estimate the prevalence and severity of early childhood caries among 4-year-old children in Hanoi and (2) examine factors that are independently associated with the presence of ECC and with severe ECC (S-ECC). Specifically, the study addressed the following research questions: what is the prevalence and severity profile of ECC among preschool children in Hanoi? And what factors are independently associated with the severity of disease, particularly S-ECC?

Methods     

Study design: this cross-sectional descriptive study investigated early childhood caries and associated parental knowledge, attitudes, and practices among preschool children.

Setting: the study was conducted from December 2024 to October 2025 at three kindergartens in Hanoi, Vietnam: Yen Vien Town Kindergarten and Anh Duong Kindergarten (Phu Dong Commune) and 10-10 Kindergarten (Tuong Mai Ward). Clinical and methodological supervision was provided by the Institute of Odonto-Stomatology Training, Hanoi Medical University.

Participants: the study population comprised 4-year-old children enrolled in the selected kindergartens and their parents or primary caregivers. Children were eligible if they were aged 4 years, cooperative during examination, and had written parental consent and school approval. Children with acute systemic illness or acute oral conditions at the time of examination were excluded. Parents or caregivers were eligible if they were primarily responsible for the child's daily care and consented to participation; those with acute illness or unable to complete the questionnaire were excluded. Participants were identified through school enrollment lists provided by the participating kindergartens. All eligible children within the selected age group were invited to participate, and recruitment was conducted consecutively during the study period. Information sheets and consent forms were distributed to parents through classroom teachers, and only children with completed consent forms and paired caregiver questionnaires were included in the final analysis. This school-based consecutive sampling approach was used to maximise participation and minimise selection bias within the study setting.

Variables: child variables included sex and dental caries status. Family variables included residential area, number of children, birth order, parental occupation, parental education, and household income. Parental knowledge, attitude and practice (KAP) scores related to oral health care for primary teeth were assessed. Early childhood caries (ECC) was operationally defined as the presence of at least one decayed (cavitated or non-cavitated), missing (due to caries), or filled tooth surface in any primary tooth in a child aged ≤71 months, consistent with definitions from the American Academy of Pediatric Dentistry and World Health Organization. Severe early childhood caries (S-ECC) was defined based on age-specific criteria, including the presence of extensive caries lesions or dmfs values exceeding established thresholds for severe disease in preschool children. Dental caries outcomes included dmft/dmfs indices and caries prevalence, defined as the presence of at least one decayed tooth. Severe early childhood caries was analysed as a separate outcome.

Data sources/measurement: clinical examinations were performed using standardised dental instruments under natural light with supplementary illumination when required. Teeth were cleaned with cotton before examination, and caries were diagnosed according to International Caries Classification and Management System criteria. Specifically, dental caries detection followed the International Caries Classification and Management System (ICCMS), which incorporates principles derived from ICDAS for identifying both non-cavitated and cavitated lesions. Lesions were recorded at the tooth and surface levels and subsequently used to calculate dmft and dmfs indices. Examinations were conducted by two calibrated dentists assisted by recorders. A structured data collection form included a clinical examination sheet and a parental questionnaire adapted from validated instruments. The questionnaire comprised four sections: socio-demographic characteristics, oral health knowledge, attitudes, and practices. Parental knowledge, attitude, and practice (KAP) regarding oral health care for primary teeth were assessed using a structured questionnaire adapted from validated instruments. Knowledge, attitude, and practice scores were calculated using predefined scoring schemes with maximum scores of 14, 10, and 23, respectively. These scores were treated as continuous variables representing the level of parental knowledge, attitudes, and practices, and no categorical cut-offs (e.g., adequate vs. inadequate) were applied. Examination results and treatment recommendations were recorded in school health records and communicated to parents. Questionnaires were distributed through classroom teachers and collected after three days; only paired clinical-questionnaire records were included in the analysis.

Bias: measurement bias was minimised through examiner training and calibration sessions, standardised examination procedures, and use of established diagnostic criteria. Inter-examiner reliability was assessed by duplicate examinations in 5-10% of children, yielding Kappa values of 0.75-0.85, indicating good to very good agreement. Selection bias was reduced by including all eligible children with parental consent in the selected schools, although potential bias related to school-based sampling may remain.

Study size: sample size was calculated using a single-proportion formula assuming early childhood caries prevalence of 82%, 95% confidence level, and 5% absolute precision, yielding a minimum required sample of 227 children. A total of 296 eligible children were examined, and their corresponding parents or caregivers were invited to participate.

Quantitative variables: continuous variables included KAP scores and dmft/dmfs indices and were summarised as means with standard deviations. Categorical variables included sex, caries presence, and socio-demographic characteristics and were expressed as frequencies and percentages. Caries prevalence was treated as a binary outcome variable. KAP scores were analysed as continuous predictors and categorised when appropriate.

Statistical methods: data were entered and analysed using SPSS version 20.0. Descriptive statistics summarised participant characteristics and outcome measures. Associations between categorical variables were assessed using the chi-square test. Univariable and multivariable logistic regression analyses were performed to identify factors associated with early childhood caries and severe early childhood caries, with results reported as odds ratios and 95% confidence intervals. Statistical significance was defined as p < 0.05.

Ethical consideration statement: the study protocol was reviewed and approved by the Scientific Council and Ethics Committee of the Institute of Odonto-Stomatology Training, Hanoi Medical University (Code: 20/2025/QD-DHYHN). Written informed consent was obtained from parents or legal guardians before participation, and permission was granted by participating schools. All procedures complied with established ethical standards for biomedical research involving human participants. Personal information was kept confidential and used solely for research purposes. Children diagnosed with dental caries were referred for appropriate counselling and treatment.

Results     

Demographic characteristics Among the 296 participants, children were almost evenly distributed between urban (51.0%) and suburban (49.0%) areas, with a slightly higher proportion of boys (53.7%) than girls (46.3%). Most caregivers were aged 35 years or older (52.2%) or 25-34 years (46.4%), and the majority had a university or postgraduate education (70.6%). Self-employed caregivers accounted for the largest occupational group (46.6%), followed by civil servants (31.4%). More than half of households reported a monthly income above 10 million VND (52.4%). Parents were the primary accompanying persons for children's oral health care (95.3%). Most families had two children (60.1%), and nearly half of the children were second-born (49.3%) (Table 1).

Prevalence and severity profile of early childhood caries (Research question 1): among the 296 preschool children examined, early childhood caries (ECC) was highly prevalent, affecting 238 children (80.4%). Severe early childhood caries (S-ECC) was identified in 133 children (44.9%), indicating that nearly half of the affected children had advanced disease (Table 2). Lesion stage analysis showed that 52.4% of children presented with early-stage lesions, while 24.0% had moderate-stage lesions and 4.0% had extensive lesions. When stages were combined, 80.4% of children exhibited lesions ranging from early to extensive stages, whereas 28.1% had lesions extending from moderate to extensive stages. The mean dmft index was 6.23 ± 5.84, with untreated decay (dt = 5.95 ± 5.34) accounting for the largest component. Missing and filled teeth contributed minimally (mt = 0.08 ± 0.59; ft = 0.20 ± 0.88). Similarly, the mean dmfs index was 9.96 ± 13.13, dominated by decayed surfaces (ds = 9.33 ± 11.43). These findings indicate that ECC in this population is both highly prevalent and clinically severe, with a predominance of untreated lesions and substantial cumulative tooth surface involvement.

Factors independently associated with severe early childhood caries (Research question 2): to address the second research question, multivariable analysis was performed to identify factors independently associated with the severity of disease, using S-ECC as the outcome variable (Table 3 and Table 4). Children who were third-born or later had a significantly higher likelihood of severe ECC compared with first-born children (OR = 4.48, 95% CI: 1.07-18.76). In addition, higher parental attitude scores were associated with increased odds of S-ECC (OR = 1.16 per unit increase). Children living in suburban areas had more than twice the odds of severe disease compared with those living in urban areas (OR = 2.35). All other variables, including family size, caregiver characteristics, household income, and parental knowledge and practice scores, were not significantly associated with S-ECC (p > 0.05). These findings demonstrate that, although ECC was widely distributed across the study population, the severity of disease was more strongly associated with specific family-level and contextual factors, particularly birth order and residential environment.

Discussion     

This study highlights a clear epidemiological pattern: although early childhood caries (ECC) was highly prevalent across the study population, only severe early childhood caries (S-ECC) showed significant associations with family-level and contextual factors. This suggests that in high-prevalence settings, determinants of disease severity may be more informative than determinants of disease presence.

The prevalence of ECC (80.4%) and S-ECC (44.9%) observed in this study exceeds global estimates for children aged 36-71 months (57.3%) [3] and is higher than reports from Iran (53.2-61.7%) [4,13] and Tanzania (approximately 45%) [11], while being comparable to high-risk populations such as southwest China (75.0%) [10]. These findings indicate that ECC in this population represents not only a widespread condition but also a substantial burden of advanced disease. The high proportion of children with moderate to extensive lesions further suggests ongoing disease progression. The proportion of S-ECC is also higher than that reported in European populations such as Southern Italy (2.7%) [9] and Germany (9.5%) [14], but is comparable to high-risk clinical settings such as France [12].

The mean dmft (6.23) and dmfs (9.96) values in this study exceed those reported in Tanzania (mean dmft 2.51) [11], Iran (approximately 2.7) [4], and longitudinal cohorts in Australia (3.41) [15], and are comparable to high-risk populations in China (4.7 ± 4.6) [10]. Importantly, the dominance of the decayed component, with minimal contributions from missing and filled teeth, reflects a high level of untreated disease. This pattern is consistent with findings in severe ECC populations in Italy and Germany [9,14]. The high dmfs values further indicate extensive surface-level destruction, aligning with studies emphasising dmfs as a sensitive indicator of disease severity and cumulative burden [5,16,17].

In this study, no independent factors were associated with ECC occurrence. This finding is consistent with evidence that in high-prevalence settings, ECC becomes widely distributed across subgroups, limiting the ability to detect differences based on socio-demographic or caregiver characteristics [3,13]. Under such conditions, analysis of disease presence alone may have limited discriminatory value.

In contrast, several factors were significantly associated with S-ECC, supporting the importance of analysing disease severity. Birth order was a strong determinant, with third-born or later children having higher odds of severe disease. This finding is consistent with studies from France and Sweden, which report increased caries risk with higher birth order [12,18]. These patterns likely reflect reduced parental attention and preventive capacity as family size increases.

Parental attitude score was also associated with S-ECC. Although higher attitude scores are generally expected to be protective, this association may reflect increased awareness following disease onset rather than effective prevention. Similar findings have been reported in clinical populations, where knowledge and attitudes do not necessarily translate into preventive practices or timely care utilisation [19-21].

Residential area was another significant factor, with children living in suburban areas having higher odds of severe disease. This is consistent with evidence showing that children in less advantaged or peri-urban settings experience higher disease burden due to disparities in access to dental care and preventive services [4,11,22,23].

Overall, these findings indicate that while ECC is highly prevalent across all subgroups, progression to severe disease is more strongly associated with family structure and residential context. This has important implications for prevention strategies. Interventions should prioritise early detection and management of lesions, particularly among children from larger families and those living in suburban or disadvantaged areas, and should focus on translating parental knowledge and attitudes into effective preventive practices and improved access to care.

This study has several limitations. The cross-sectional design limits causal inference. Parental knowledge, attitude and practice data were self-reported and subject to recall and social desirability bias. The absence of detailed behavioural indicators, such as toothbrushing practices, dietary habits, and fluoride exposure, may also limit the comprehensiveness of the analysis. In addition, the study was conducted in selected kindergartens in Hanoi, which may limit generalizability to other regions. However, the use of standardised diagnostic criteria and examiner calibration enhances the reliability of the findings.

Conclusion     

In conclusion, early childhood caries and severe early childhood caries remain prevalent among preschool children in the study setting, with a substantial burden of untreated disease. While no independent factors were associated with early childhood caries overall, severe disease was significantly related to higher birth order, parental attitudes, and residential area. These findings highlight the importance of distinguishing disease severity in epidemiological analyses and support the need for targeted, context-sensitive preventive strategies to reduce the progression and consequences of early childhood caries.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Conception and study design: Tran Thi My Hanh and Ha Ngoc Chieu. Data collection: Ha Ngoc Chieu, Truong Manh Nguyen and Dao Thi Hang Nga. Data analysis and interpretation: Nguyen Manh Cuong and Nguyen Thu Giang. Manuscript drafting: Ha Ngoc Chieu. Manuscript revision: Tran Thi My Hanh, Luu Van Tuong, Pham Thi Hanh Quyen, Le Thi Thuy Linh, Nguyen Dinh Phuc and Nguyen Hoang Thanh. All authors read and approved the final version of the manuscript.

Acknowledgments     

The authors thank the administrators, teachers, children, and parents of Yen Vien Town Kindergarten, Anh Duong Kindergarten, and 10-10 Kindergarten for their cooperation and participation. The authors also acknowledge the Institute of Odonto-Stomatology Training, Hanoi Medical University, for clinical supervision and technical support during data collection.

Tables     

Table 1: socio-demographic characteristics of preschool children, caregivers, and family context in Hanoi, Vietnam, 2024–2025 (n = 296)

Table 2: prevalence and severity profile of early childhood caries and dmft/dmfs indices among 4-year-old children in Hanoi, Vietnam, 2024-2025 (n = 296)

Table 3: socio-demographic and parental factors associated with the presence of early childhood caries among preschool children in Hanoi, Vietnam, 2024–2025 (n = 296), using univariable logistic regression

Table 4: socio-demographic and parental factors associated with severe early childhood caries among preschool children in Hanoi, Vietnam, 2024-2025 (n = 296), using multivariable logistic regression

References