Study population
A cross-sectional study was carried out from August to December 2011 in 24 randomly selected public and private kindergartens in Guangzhou, the capital of Guangdong Province in southern China. Guangzhou has approximately 12 million inhabitants and is divided into 12 administrative districts. The sample was selected from an overall population of about 100,000 5-year-old children in the city, 98% of whom were enrolled in kindergartens.
The required sample size was determined based on the prevalence of dental caries among 5-year-old children in Guangzhou. In the calculation, the estimated prevalence of dental caries was 60%, which was the finding from the latest national oral health survey [20]. The precision of the estimate was set to have a standard error of 1.5%. The minimum sample size to satisfy the requirements was calculated to be 1067 children.
The study protocol was approved by the Ethics Committee of Guanghua School of Stomatology, Sun Yat-sen University. A two-stage sampling technique was employed. The first stage consisted of a simple random selection of six of the 12 administrative districts in Guangzhou. The second stage consisted of a cluster sampling, in which four kindergartens were randomly selected in each of the selected administrative districts. All the 5-year-old children in the selected kindergartens and their main caregivers were invited to take part in this survey. All caregivers were fully informed of the study purpose in writing, and they were given a free choice as to whether or not they participated. Written informed consent for participation was obtained from the caregivers who decided to take part in the study.
To test data collection procedures and ascertain the applicability of the instruments, a pilot study was conducted on 60 children in a kindergarten that was not selected for the survey.
Instruments and measures
Data were collected by means of a questionnaire administered to the caregivers. The questionnaire consisted of five parts: questions on the child’s oral health-related behaviors (sugary snack intake habits, toothbrushing habits, and use of dental services); questions on the caregiver’s oral health knowledge and attitudes; the SOC scale; questions on the caregiver’s oral health-related behaviors; and questions on the child’s demographic and socioeconomic background. The questionnaire was distributed by the kindergarten to the caregivers, who then completed the questionnaire at home and returned it to the kindergarten.
Outcome variables
Four oral health-related behaviors of the 5-year-old children were designated as outcome variables: frequency of sugary snack intake (<once/day vs. ≥once/day); toothbrushing frequency (≤once/day vs. ≥twice/day); having used dental services (yes vs. no); and pattern of dental visit for the children who had visited a dentist (for treating dental problems vs. mainly for check-ups).
Independent variable
The caregiver’s SOC was measured by the short version of the SOC scale (SOC-13) developed by Antonovsky [21]. The scale consists of 13 items, each of which is scored on a Likert scale, ranging from 1 (“very often”) to 7 (“very seldom or never”). The SOC scale measures three dimensions: comprehensibility (five items); manageability (four items); and meaningfulness (four items). After reversing the scores of the five negatively worded items (items 1, 2, 3, 8, and 13), the scores of the 13 items were added to obtain the overall SOC score. Thus, the SOC score could range from 13 to 91, with higher scores indicating stronger SOC. The SOC-13 scale has been translated into many languages, including Chinese, and extensively used in studies in many countries [21, 22]. The Chinese version of SOC-13 scale (C-SOC-13) has been shown to have good reliability and validity as well as discriminatory power [22]. In this sample, Cronbach’s alpha coefficient of the C-SOC-13 scale was 0.80.
Controlling variables
The child’s demographic background (gender, single child, marital status of parents, caregiver) and socioeconomic background (mother’s education and occupation, father’s education and occupation, family income), the caregiver’s oral health-related behaviors (frequency of sugary snack intake, toothbrushing frequency, utilization of dental services, pattern of dental attendance), and oral health knowledge and attitudes were used as controlling variables. Similarly worded questions were used to assess the oral health-related behaviors of the caregiver and those of the child.
The caregiver’s oral health knowledge was measured by four questions about the causes and prevention of tooth decay and periodontal disease, which had been used in a previous study of Guangdong adults [23]. Each correct answer to a question was given a score of 1, and incorrect or “don’t know” answers were scored 0. A maximum of four correct answers was accepted for each question, so the caregiver could score up to 4 points per question. The overall oral health knowledge score was the sum of the scores of the four questions, which could range from 0 to 16, with higher scores indicating better oral health knowledge.
To explore the caregiver’s attitudes toward oral health, eight statements that had been used in a previous study were selected [23]. These related to dental health beliefs and the importance of oral health, retaining natural teeth, and the use of dental services. The response to each statement was “agree,” “disagree,” or “neither.” A dental attitude score was constructed by counting the total number of statements to which the caregiver showed a positive attitude. The final score could range from 0 to 8, with higher scores indicating a more positive attitude toward oral health.
Statistical analysis
Data analysis was carried out using SPSS for Windows (version 16.0). The total SOC score was analyzed as a continuous variable. Initially, associations between the independent variable and the outcome variables were assessed by two-sample t tests. Chi-square tests and t tests were carried out to analyze the relationships between the controlling variables and the outcome variables. Controlling variables with a p value of 0.20 or lower were included in the multiple logistic regression analysis. Multiple logistic regression analysis was performed to examine the relative significance of the effects of the independent variable and the controlling variables on the outcome variables in different caregiver groups. The level of significance for all statistical tests was set at 0.05.