Study design
The present study is a longitudinal study and part of a register-based cohort study on development of dental caries in children between 3 and 7 years of age. Data sources were based on Swedish national registry data held by the Swedish National Board of Health and Welfare and by Statistics Sweden (SCB). Information on dental caries was collected from data sources at the Public Health Care Administration in Stockholm.
Study population
The study population was created from the population of child residents in Stockholm County, born in 2000–2003 (n = 83 147). The present study evaluated children who had been examined at 3 and 7 years of age in the Public Dental Service, by private practitioners, or at the Department of Dental Medicine, Karolinska Institutet. When the mother was foreign-born, participants were excluded from the study if there was no information on the maternal country of origin, age upon arrival in Sweden, or length of residence in Sweden. The final study sample comprised 63 931 children who had had dental examinations at 3 and 7 years of age.
Variables
Information on age, year of arrival, and maternal country of origin was obtained from the Swedish Registry of Total Population at SCB. We classified countries into four categories for maternal nationality; the first category was Swedish-born mothers. For the other categories we used the HDI developed by the United Nations Development Program. It is a composite index of life expectancy, education level, and per capita income indicators. We used the HDI from 1999, the year before the children in this study were born (HDI 1999). The HDI cut-off values were as follows: low = 0.350-0.449, medium = 0.550-0.699, and high ≥ 0.700 (Table 1).
From the database of The Swedish National Board of Health and Welfare, which maintains the Swedish Medical Birth Registry (MBR), we collected maternal age at childbirth (14-20 years; 21–25 years; 26–30 years; 31–35 years; > 35 years) and family situation (single mothers, cohabitants). The SCB determines disposable household income using an algorithm that considers all household incomes reported to the Swedish Tax Agency, reduced by all taxes, and then divided by consumer units. The range of disposable household income in the study was used by first constructing quintiles and thereafter categorizing them as low (1st and 2nd quintiles), medium (3rd quintile), and high (4th and 5th quintiles). Maternal educational level was classified as ≤ 12 years and > 12 years of education. The length of maternal residence in Sweden at childbirth was calculated as the difference between the child’s birth year and the year that the mother had arrived in Sweden.
A proxy measure of acculturation was maternal age upon arrival in Sweden, classified as < 6, 7-12, 13-19, and ≥ 20 years of age.
Dental caries
Data on manifest caries lesions were collected from clinical and radiographic examinations. The decayed, extracted, and filled primary teeth (deft) index measured the severity of caries experience in children at 3 and 7 years of age. The definition of “manifest caries” as a lesion that clearly extends into the dentin was used in this study [16]. Radiographic examinations in the form of bitewings were only done given indications, to provide more extensive data on dentin caries lesions. The variable caries increment (Δ deft) was defined as the difference between deft at age 3 and deft at age 7 years; the result was then dichotomized as “no caries increment" (Δ deft = 0) or “caries increment” (Δ deft ≥ 1). No permanent teeth were included in the outcomes at 7 years of age.
Statistical analyses
STATA 14 for Windows (Stata Statistical Software; StataCorp LP; College Station, TX, USA) was used for data analysis. Descriptive analyses included relative and absolute frequencies. Differences between categorical variables were assessed using the chi-square test. We used negative binomial regressions to analyze 10 models, estimating incidence rate ratios (IRR) with 95% confidence intervals to assess the association between caries increment in children between 3 and 7 years of age (0: no risk group; 1: risk group). The analyzed models were:
(1) The HDI of the maternal country of origin and maternal age upon arrival in Sweden concurrently, using the Swedish-born population as a reference group (Figure 1).
(A) Evaluating mothers born in low HDI countries: (I.a) the model compares maternal age upon arrival in Sweden, with the mothers who arrived in Sweden before age 7 years as a reference group; (II.a) length of residence in Sweden of mothers, the with mothers who had lived in Sweden for 20 years or more as a reference group; (III.a) the model compares maternal age upon arrival in Sweden, with the mothers who arrived in Sweden before age 7 years as a reference group, adjusted by length of residence in Sweden of mothers.
(B) evaluating mothers born in medium HDI countries: (I.b) the model compares maternal age upon arrival in Sweden, with the mothers who arrived in Sweden before age 7 years as a reference group; (II.b) length of residence in Sweden of mothers, the with mothers who had lived in Sweden for 20 years or more as a reference group; (III.b) the model compares maternal age upon arrival in Sweden, with the mothers who arrived in Sweden before age 7 years as a reference group, adjusted by length of residence in Sweden of mothers.
(C) evaluating mothers born in high HDI countries: (I.c) the model compares maternal age upon arrival in Sweden, with the mothers who arrived in Sweden before age 7 years as a reference group; (II.c) length of residence in Sweden of mothers, the with mothers who had lived in Sweden for 20 years or more as a reference group; (III.c) the model compares maternal age upon arrival in Sweden, with the mothers who arrived in Sweden before age 7 years as a reference group, adjusted by length of residence in Sweden of mothers. All analyses were adjusted for sex of the child, maternal age, family situation, and income. We used the Bayesian Information Criterion (BIC) to assess the overall fit of a model and compare various models.