Alcohol consumption and the risk of dementia in the oldest old: The Monzino 80‐plus study

Abstract

AbstractBackgroundWine, beer, spirits or liqueurs are widely popular beverages worldwide. Several studies have investigated the association between alcohol consumption and dementia with inconsistent results, however very few data are available on this relationship in the oldest‐old. Objectives: To investigate the association of alcohol consumption with the development of dementia in advanced age in a prospective population‐based study of 80‐years or older residents in Varese province, Italy (Monzino 80‐plus Study).MethodInformation on alcohol consumption (wine, beer, spirits or liqueurs) was obtained from the subject and/or a primary informant, mostly a family member, using a standardized questionnaire. Diagnosis of dementia fulfilled DSM‐IV criteria. Covariates entered in the fully adjusted logistic and Cox regression models: age, sex, education, diabetes, hypertension, heart failure, myocardial infarction, COPD, ictus, TIA, parkinsonism. Since former drinkers stopped drinking on average 12.4±13.7 (cross‐sectional study) and 16.0±17.0 (longitudinal study) years before interview, they were categorized, together with the never drinkers, in the group of non‐drinkers.ResultAmong the 2,195 participants (mean age: 90.2 years; women: 72.7%) interviewed at first visit, 1,107 (50.4%) were current drinkers (wine, only or together with other alcoholic beverages: 99.8% in cross‐sectional and 100% in longitudinal studies), 318 (14.5%) former drinkers, and 770 (35.1%) never drinkers. Consumers on average currently drank 18.2±15.5 g/day of ethanol, started drinking at 21.4±16.2 years of age, and have been drinking for 68.1±11.2 years. Prevalence of dementia was lower among current drinkers (27.8%) compared to non‐drinkers (47.2%; fully adjusted p<0.0001). There were 1,139 individuals without dementia at first visit with at least one follow‐up available (mean age: 88.4 years; women: 69.3%). In the following 15.5 years, current drinkers (n=675) had a lower occurrence of dementia (43.3%) than non‐drinkers (53.0%; fully adjusted HR: 0.8372, 95% CI: 0.7008‐1.0002, p=0.0503). When current drinkers were divided according to daily intake into two groups (>0‐24 and >24 g/day of ethanol), results of both cross‐sectional and longitudinal analyses gave analogous results.ConclusionAlcohol consumption, mostly of moderate amounts, is common among oldest‐old and is associated with a marginally non‐significant lower risk of incident dementia.