The sample
This study analyzed a nationally representative data from the China Health and Retirement Longitudinal Study (CHARLS), using a steady-state design, from 450 villages or communities in 150 counties or districts of 28 provinces. The baseline survey was conducted in 2011, through a four-stage, stratified, cluster probability sampling design [41]. The surveys in 2013, 2014 and 2015 contained the assessments of social, economic and health status, and recruited additional individuals who had just become 45 years old. Further details are provided elsewhere [42], this study combined data from the baseline survey and life history in rounds 2, 3 and 4 surveys. The exclusion criteria were as follows: (i) those individuals who did not report an edentulous status; (ii) age was less than 45 years old; and (iii) those with missing data.
The sample size was calculated based on the prevalence of Chinese edentulism derived from the SAGE (8.0-9.0%) [1], according to the calculation formula below [43], the marginal error was within 2% with 95% confidence level, P=0.5, thus the estimated maximum sample size was 692. (see Formula 1 in the Supplementary Files)
Measure
Edentulism was assessed based on the response question: “Have you lost all of your teeth?”, the variable had two options (yes or no).
Basic information included sex and age which were categorized into four groups: “45-59”, “60-69”, “70-79”, and “80 or more”.
Childhood SES
The education of parents consisted of “illiterate”, “elementary school”, “middle school”, and “high school or above”.
The occupation of the parents was categorized into “farmer” and “non-agricultural”.
The parents working status was divided into “all of childhood” and “part of childhood”.
Relationships with parents was based on the question: “How would you rate your relationships with your female/male guardian when you were growing up?” were classified as “excellent”, “very good”, “good”, or “poor”.
Quarrels and fights between parents were classified as “often”, “sometimes”, “not very often”, or “never”.
Parents divorced during childhood was divided into “yes” or “no”.
Care from mother to her child was based on the question: “How much effort did your female guardian put into watching over you?”, which was classified into “a lot”, “some”, “a little”, or “none at all”.
Love and affection from mother to her child was based on the question: “How much love and affection did your female guardian give you while you were growing up?” was classified as “often”, “sometimes”, “rarely”, or “never”.
Father’s health-related behaviors (drinking/smoking) were divided into “yes” or “no”.
Neighbors’ help was based on the question: “Were the neighbors of the place where you lived as a child willing to help each other out?” included “willing to”, “somewhat willing to”, or “unwilling to”.
Neighborhood relationship was based on the question: “Were the neighbors of the place where you lived as a child very close-knit?” included “very close-knit”, “somewhat close-knit”, or “not close-knit”.
Financial situation of the family was based on the question: “When you were a child under 17 years old, compared to the average family in the same community/village at that time, how was the financial situation of your family?”, which was classified into “a lot better off than them”, “somewhat better off than them”, “same as them”, “somewhat worse off than them”, “a lot worse off than them”.
Food availability was based on the question: “When you were a child before under 17 years old was there ever a time when your family did not have enough food to eat?” included “Yes” or “No”.
Primary residence was divided into “rural” or “urban”.
Medical convenience was based on the question: “Are you satisfied with the quality, cost, and convenience of local health care” included “yes” or “no”.
Vaccination in childhood was based on the question: “Before you were 15 years old (including 15 years old), did you receive any vaccinations?” included “yes” or “no”.
Adulthood SES
Adulthood SES was assessed by adult educational achievements [44], divided into four groups (illiterate, elementary school, middle school, or high school and above).
Statistical analysis
The differences in edentulism by age, sex, childhood SES and adulthood SES were analyzed using the Chi-Square tests. This study used principal component analysis (PCA) to determine the main factors of childhood SES. The participants with missing data on variables of interest were excluded. In order to evaluate the representativeness of the sample, the age and sex were compared between the excluded and included participants [45]. There was no statistically significant difference between these two groups.
Binary logistic regression was used to examine the associations between childhood SES and edentulism. Age and sex were adjusted in model 1; adulthood SES was subsequently adjusted in model 2. The odds ratio (OR) and the 95% confidence interval (CI) are presented. The Hosmer-Lemeshow goodness-of-fit test was used to evaluate the multivariable model fit. The statistical significance was considered as two-sided P < 0.05. All analyses were conducted using the Stata 14 software (Stata Corp. LLC, College Station, Texas, USA).