Alcohol consumption in midlife and old age and risk of frailty: Alcohol paradox in a 30-year follow-up study

Background alcohol consumption has many harmful health effects, but also benefits of moderate consumption on frailty have been reported. We examined this relationship longitudinally from midlife to old age. Methods data of reported alcohol consumption in midlife (year 1974) and in old age (years 2000 and 2003) were available of a socioeconomically homogenous sample of 2360 men (born 1919-34, the Helsinki Businessmen Study). Alcohol consumption was divided into zero (N = 131 at baseline), light (1-98 g/week, N = 920), moderate (99-196, N = 593), and high consumption (>196, n = 716). Incidence of phenotypic frailty and prefrailty was assessed in 2000 and 2003. Alcohol consumption (reference 1-98 g/week, adjusted for age, body mass index and smoking) was related to frailty both longitudinally (from 1974 to 2000, and from 2000 to 2003) and cross-sectionally in 2000 and 2003. Results during a 30-year follow-up, high consumption clearly decreased whereas lighter consumption remained stable. High consumption in midlife predicted both frailty (odds ratio = 1.61, 95% confidence interval = 1.01-2.56) and prefrailty (1.42; 1.06-1.92) in 2000, association with zero and moderate consumption was insignificant. Cross-sectionally in 2000, both zero (2.08; 1.17-3.68) and high consumption (1.83; 1.07-3.13) were associated with frailty, while in 2003 only zero consumption showed this association (2.47; 1.25-4.88). Conclusion the relationship between alcohol and frailty is a paradox during the life course. High, not zero, consumption in midlife predicts old age frailty, while zero consumption in old age is associated with frailty, probably reflecting reverse causality.