[Multilevel logistic regression analysis on hypercholesterolemia related risk factors among adults in China]

Objective: To investigate the prevalence of hypercholesterolemia among Chinese adults in different geographic areas, and to analyze the related factors. Methods: China Chronic and Non-Communicable Disease and Risk Factor Surveillance was conducted in 2013, based on 298 counties/districts in 31 provinces of Chinese mainland. The adults aged 18 years old were randomly selected using multi-stage stratified clustering sampling method. Information on chronic disease and risk factors was collected using face-to-face questionnaire interview and physical measurement. Blood samples were collected by local staffs. Serum total cholesterol (TC) was determined using standard method in a central laboratory. After excluding 565 participants missing key variables and 1 558 participants with abnormal TC values, a total of 174 976 participants were included. Weighted prevalence of hypercholesterolemia was calculated. Hypercholesterolemia related individual or geographic determinants were defined using multilevel logistic regression. Results: The prevalence of hypercholesterolemia in Chinese adults age 18 years old and above was 6.9% (95%CI: 6.4%-7.3%), ranged from 3.0% (95%CI: 2.5%-3.4%) in the northwest of China to 14.2% (95%CI: 12.9%-15.5%) in the south (χ(2)=183.42, P<0.001). The prevalence of hypercholesterolemia was higher in 6 provinces including Tianjin, Liaoning, Fujian, Guangdong, Guangxi, and Hainan (≥9.0%), but lower in 7 provinces including Shanxi, Shannxi, Gansu, Qinghai, Ningxia, Xinjiang, and Tibet (<3.9%). There was a 1.27 fold variation in hypercholesterolemia prevalence by provincial level, and 72.5% of the geographical variation in hypercholesterolemia prevalence was account for by area-level determinants. With multilevel logistic analysis, the individual risk factors associated with hypercholesterolemia included aging, higher education level or annual household income per capita, regular drinking, too much red meal intake, inactivity, overweight or obesity. For geographic factors, residents living at south China, counties/districts with higher urbanization rates, higher education level or lower standardized death rates were more likely to have hypercholesterolemia (all P<0.05). Conclusion: The prevalence of hypercholesterolemia was high in Chinese adults, it was different between regions and related with characteristics of population, individual behaviors and geographical regions.