Abstract
AIM: To investigate the effects of alcohol on serum glycated albumin (GA) levels in Chinese men. METHODS: A total of 2314 male subjects from the Jinuo ethnic group in China were enrolled. Of these, 986 subjects drank alcohol frequently and 404 subjects did not. Lifestyle information was gathered by using a questionnaire, and measurements of blood pressure, body mass index, blood glucose level, liver function, and kidney function were collected. GA was measured by using an enzymatic method. Frequent drinking was defined as a history of drinking ethanol > 80 g/d within the past two weeks. Nondrinking was defined as no alcohol consumption in the past three months. Subjects with an alcohol intake between 0 and 80 g/d in the past two weeks were included in the drinking-occasionally group. Analysis of variance (ANOVA), correlation analysis, and linear regression were used to evaluate the effects of drinking on serum GA levels. Decision tree regression (DTR) algorithm was used to evaluate the effect of features (variables) on GA levels. RESULTS: We found that male subjects who drank frequently had significantly lower serum GA levels than subjects who did not drink (13.0 ± 1.7 vs. 14.1 ± 3.7, p < 0.05). Spearman's correlation analysis calculated a coefficient of -0.152 between drinking and GA (p < 0.005). Linear regression established that drinking was an independent predictor for GA levels with a standardized regression coefficient of -0.144 (p < 0.05). Decision tree regression showed that the effect of drinking on GA levels (0.0283) is five times higher than that of smoking (0.0057). CONCLUSIONS: Frequent alcohol consumption could result in decreased GA levels in men of the Jinuo ethnic group in China.