Sample
Guangzhou is the capital city of the Guangdong province, which is located in Southern China. The climate is hot and long in the summer due to its location in the southern subtropical zone. The summer is 6 months in length (May to October). Guangzhou is a rapidly developing city with a population of approximately 10 million. The gross domestic product (GDP) per capita was USD 9,302 in 2007, and Guangzhou ranked as the third economically developed city in China, following Beijing and Shanghai [5].
A cross-sectional study was conducted in Guangzhou by a stratified multi-stage cluster sampling. A pilot study was performed before the formal study. The prevalence of dental erosion was 20.8% for 175 school children (12-13 years of age) examined from an urban junior high school. A total of 1,500 school children were proposed in our study according to the sample size formula for the estimation of prevalence (α = 0.05) in order to decrease study error. Guangzhou is comprised of 10 urban districts and 2 suburban districts. The population ratio between the urban and suburban regions was approximately 4:1 [5]. Four urban districts and one suburban district were selected by simple random sampling. A list of all junior high schools in these districts was obtained from the local Department of Education. Two junior high schools were selected by simple random sampling in each district. In each school, three classes of grade one students were selected by the same sampling method and 140-160 children were cluster-selected. All of the selected children were 12-13 years of age. Those who were with orthodontic appliances, enamel defect accompanied by a loss of tooth substance, and fractured or missing teeth of the incisors or the first molars were excluded. A total of 1,499 children (774 boys and 725 girls), 12-13 years of age, from 10 junior high schools were invited to participate in the study.
The study protocol was approved by the Research Ethics Committee of Guanghua School of Stomatology of Sun Yat-sen University. An informed consent letter regarding the aim and importance of the study was signed by the children and the parents/guardians before starting the survey, which assured that children participated in the study on their own accord.
Calibration of examiners
Training and calibration exercises were conducted prior to the study. An experienced dental epidemiologist who had a high level of education and experience, was responsible for training and calibrating two examiners. The diagnostic criteria of dental erosion were thoroughly discussed with the examiners. A range of dental erosion levels based on the diagnosis via photographic images was reviewed in the calibration exercise. The epidemiologist also provided instruction for the examiners during the pilot study.
To assess the reproducibility of the diagnostic criteria, approximately 10% of the subjects were re-examined (only the labial surfaces of the upper incisors). In every dental examination section at each school, approximately 15 students in intervals of 10 were re-examined by the same examiner and another examiner. The results of the duplicate examinations were used to estimate the intra- and inter-examiner reliability.
Clinical Examination
The two previously calibrated examiners participated in the clinical examinations and visited the selected schools. The clinical examinations were performed in well-lit classrooms or in shaded places under natural light using plane mouth mirrors and sterilized cotton to remove debris. The central incisors, lateral incisors, and first molars in the upper and lower jaws were examined. The diagnostic criteria of dental erosion proposed by Eccles [6] were used in this study. The index of O'Sullivan [7] was adopted to record the distribution, severity, and amount of affected teeth. This index is especially designed for epidemiologic surveys and for the diagnosis of erosion in children to determine treatment options.
O'sullivan index for measurement of dental erosion:
Site on erosion on each tooth
Code A: Labial or buccal only
Code B: Lingual or palatal only
Code C: Occlusal or incisal only
Code D: Labial and incisal/occlusal
Code E: Lingual and incisal/occlusal
Code F: Multi-surface
Grade of severity (worst score for an individual tooth recorded)
Code 0: Normal enamel
Code 1: Matt appearance of the enamel surface with no loss of contour
Code 2: Loss of enamel only (loss of surface contour)
Code 3: Loss of enamel with exposure of dentine (enamel-dentin junction visible)
Code 4: Loss of enamel and dentine beyond enamel-dentin junction
Code 5: Loss of enamel and dentine with exposure of the pulp
Code 9: Unable to assess (e.g. tooth crowned or large restoration)
Area of surface affected by erosion
Code -: Less than half of surface affected
Code +: More than half of surface affected
Questionnaire
The school children completed a questionnaire at the schools prior to the clinical examination. The questionnaire was designed to reflect the socio-economic status, behavioural factors, and general health involved in the etiology of erosion, as proposed by Lussi [3] and in other studies [8, 9]. The pilot study was carried out to test and refine the questionnaire.
The questionnaire included questions about general information (gender and age), socio-economic status, occupation and education levels of the parents, oral hygiene habits, frequencies of ingesting certain beverage types, amount of acidic drink intake per week (including carbonated drinks, sport drinks, lemon tea, and fruit juices), special drinking habits, general health (including frequency of vomiting and heartburn or nausea in this study), and vitamin C supplements. The frequency of swimming in summer was also included in the questionnaire because of possible lower pH value in the water of swimming pools [3].
Statistical analyses
The study data were entered into a computer using Epidata (version 3.0) and analyzed using SPSS software (version 13.0). Intra- and inter-examiner agreement was evaluated using Cohen's kappa. Descriptive analysis was conducted to describe the prevalence and characteristics of dental erosion. A two-step approach was used to analyze risk factors of dental erosion. First, bivariate analysis was used to test the relationship between dental erosion and the associated factors. Then, a logistic regression analysis was used to analyse the factors that were independently related to the presence of erosion. The variables (P < 0.5) in the bivariate analyses were entered into a logistic regression model in a forward fashion. The level of statistical significance was set at 5%.