Alcohol drinking and diabetes risk among older adults in rural Guangzhou

Abstract

Background: Observation studies indicated that potential benefit was associated with moderate alcohol drinking and type 2 diabetes mellitus. However, this association might not be reliable as bias and residual confounding could hardly be avoided in the nature of these studies. And Randomized Control Trial (RCT) is not feasible due to ethical issue. Southern China might be a suitable place to investigate the association between moderate alcohol drinking and diabetes for its unique social drinking pattern. With rapid urbanization and ageing in China, the prevalence of diabetes among the older adults in rural area is increasing. Guangzhou is a capital city in southern China, which is highly urbanized. It could function as a predictor for the trends and the burden of the disease, which could provide reference for the other inland cities. Methods: A cross-sectional survey with a multistage stratified random sampling design was conducted from March 2011 to April 2012 in rural Guangzhou. 4550 older adults in rural Guangzhou were sampled, receiving health examination and standardized survey. Binary logistic regression was used to adjust the possible confounders examine the association between different drinking pattern and diabetes. Results: In total, 4029 subjects were included in the study. 15.0% of older adults in rural Guangzhou reported they were current (422/606) or ex-drinkers (184/606). The prevalence of diabetes was 19.3% among the older population in rural Guangzhou. In the adjusted association of alcohol drinking pattern and diabetes risk among older adults in rural Guangzhou, diabetes was associated with occasional drinking (Odd Ratio(OR) =0.554, 95% Confident Interval(CI) 0.318 to 0.966), moderate drinking (OR=0.235, 95%CI 0.113 to 0.487), ex-drinker (OR=2.358, 95%CI 1.699 to 3.328) adjusted for sex, age, education, smoking. Conclusions: Occasional and moderate drinking might have potential benefit for diabetes. While the ex-drinkers quitting drinking might be driven by their poor health condition. However, recall bias, survival bias and residual confounders might still need to be further adjusted in this study. The association between them still needs further investigate and cautious conclusion should be made.