Moderate Alcohol Use and Cognitive Function in an Elderly Chinese Cohort

To the Editor: Moderate alcohol use usually has health benefits, including better cognitive function in older people,1 although there are exceptions,2 and heavy alcohol use is harmful,3 but most evidence concerning moderate alcohol use comes from observations in Western populations, where moderate alcohol use is typical, and systematic differences between moderate alcohol users and others may lead to residual confounding.4 Evidence from other settings may help disentangle the effects of moderate alcohol use from its social context. In southern China, social use of small amounts of alcohol, unlikely to have significant effects,5 is the typical drinking pattern. The association between social and moderate alcohol use and cognitive function was examined in a large (n=64,278) population-based cohort of older Chinese men (n=21,950) and women (n=42,328) enrolled from July 1998 to December 2001 at all 18 Elderly Health Centers of the Hong Kong Department of Health.6 Participants underwent a standardized health assessment and examination at enrollment, including functional status, lifestyle habits, and socioeconomic position. Alcohol use was assessed based on self-reported frequency and quantity and categorized as never, social (<1/week), weekly social, moderate (≤30 g ethanol/day (men) and ≤20 g ethanol/day (women), excessive (>30 g ethanol/day (men) and >20 g ethanol/day (women)), and exdrinkers. Cognitive function was assessed from the Abbreviated Mental Test (AMT), with cognitive impairment defined as an AMT score less than 8.7 Multivariate logistic and linear regression was used to examine the adjusted associations between alcohol use and cognitive impairment and AMT score, respectively. Whether associations differed according to sex was assessed according to the significance of the interaction terms. Ethics approval was obtained from the ethics committee of the Li Ka Shing Faculty of Medicine, University of Hong Kong. The study complied with the Declaration of Helsinki. Forty-nine percent of the men and 84% of the women were never drinkers. Most of the ever-drinkers (76% of men and 88% of women) reported social (<1/week) alcohol use. Alcohol use had a similar association with cognitive impairment according to sex, but there were sex-specific associations with AMT score (P-value for interaction <.001). Figure 1 (upper panel) shows that, adjusted for age, sex, education, smoking, and exercise, cognitive impairment was less likely in social (<1/week) alcohol users and more likely in social weekly, excessive, and exdrinkers but was unrelated to moderate alcohol use. Figure 1 (lower panel) shows similarly, in sex-specific analysis, that AMT score was higher in social (1</week) alcohol users and lower in ex-drinkers but unrelated to moderate alcohol use, adjusted for age, education, smoking, and exercise. Additional adjustment for depressive symptoms, social isolation, and the presence of cardiovascular disease did not change estimates (data not shown). Odds ratio (OR) of cognitive impairment (upper panel) and sex-specific differences in Abbreviated Mental Test score (AMT; lower panel) according to alcohol use category compared with never drinkers adjusted for sex (if appropriate), age (5-year age groups), education (postsecondary), secondary, primary, no formal education but literate, and illiterate), smoking status (never or ever), and exercise (never, <30 min/d, ≥30 min/d). Consistent with other studies in older people,1 the typical pattern of alcohol use in older Chinese was associated with the absence of cognitive impairment and with higher AMT score, although in this developed non-Western setting, the typical pattern of alcohol use was social (<1/week), not moderate, alcohol use. Moreover, moderate alcohol use, as defined in Western settings, was a less common pattern of alcohol use and was not associated with better cognitive function. The possibility that occasional, social use of, most likely, small quantities of alcohol has a cognitively protective role cannot be excluded, but given the lack of protective effect of higher levels of alcohol use in this study, it is more likely that factors associated with being a typical alcohol user may underlie the commonly observed protective effect of moderate alcohol use for cognitive function8 rather than alcohol per se underlying the association. In Western settings, the definition of moderate alcohol use is often based on the amount considered “statistically normal,”“noninjurious,” or “optimum,”9 and may identify people who are systematically different, and healthier, than others.4 This is particularly likely in older people, who may quit drinking with ill health, consistent with the observed associations between ex-drinking and poor cognition, but will inevitably result in the self-selection of healthy typical alcohol users. Comprehensive identification and measurement of all factors that may differ between typical and atypical alcohol users is difficult to achieve, leaving observation studies open to residual confounding that statistical adjustment cannot resolve. With population aging, health conditions such as cognitive decline, dementia or Alzheimer's disease are anticipated to become more common. In contrast to previous studies, the current study suggests that the apparent positive associations between moderate alcohol use and cognitive function in older people may be noncausal. More research in non-Western populations is warranted to verify the association between moderate alcohol use and cognitive function or other health attributes in older people. We wish to thank the staff of the Elderly Health Centers for their assistance in data collection and entry. Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this paper. This project was funded by the Health Care and Promotion Fund Committee in Hong Kong (Grant: HSRC #S111016). Author Contribution: Au Yeung SL, Leung GM, Schooling CM: study concept and design, analysis and interpretation of data, and preparation of manuscript. Chan WM, Hui YF, Lam TH: study concept and design and acquisition of subjects and data. Sponsor's Role: The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the letter.