Association of alcohol use with memory decline in middle-aged and older Chinese: a longitudinal cohort study

Background Previous studies on associations of alcohol use with memory decline showed inconclusive results. We examined these associations using longitudinal data from the Guangzhou Biobank Cohort Study (GBCS) and explored whether these associations varied by sex and age group. Methods Memory function was assessed by delayed 10-word recall test (DWRT) and immediate 10-word recall test (IWRT) at both baseline (2003–2008) and follow-up (2008–2012) examinations, expressed as the mean annual change and mean annual rate of change in scores. Memory cognitive impairment was defined by DWRT scores of less than 4. Multivariable linear regression models and restricted cubic spline were used for data analysis. Results Of 14,827 participants without memory cognitive impairment at baseline, 90.2% were never or occasional drinkers, 5% moderate drinkers, 1.5% excessive drinkers, and 3.3% former drinkers. The mean (standard deviation) age was 60.6 (6.6) years old. During an average of 4.1 years follow-up, 1000 (6.7%) participants developed memory cognitive impairment. After adjusting for confounders, compared with never or occasional drinkers, moderate and excessive drinkers had significant decline in DWRT scores ( β , 95% confidence interval ( CI ) = -0.04 (-0.08 to -0.01), and − 0.07 (-0.14 to 0.01), respectively), and IWRT scores ( β , 95% CI = -0.10 (-0.19 to -0.01), and − 0.15 (-0.30 to 0.01), respectively) annually. With respect to the mean annual rate of change, moderate and excessive drinkers also showed greater decline in DWRT scores ( β , 95% CI = -1.02% (-1.87% to -0.16%), and − 1.64% (-3.14% to -0.14%), respectively). The associations did not vary by sex and age group (all P values for interaction ≥ 0.10). Conclusion Compared to never or occasional alcohol use, moderate and excessive alcohol users had greater memory decline and the associations did not vary by sex and age group.