ALCOHOL INTAKE AND SURVIVAL IN THE ELDERLY - A 77 MONTH FOLLOW-UP IN THE DUBBO STUDY

Background: A prospective study in non-institutionalised Australian el derly aged 60 years and over commenced in Dubbo, NSW in 1988. Aim: To examine the relationship between all-causes mortality and alcohol inta ke. Methods: The data were derived from a community-based sample compr ising 1236 men and 1569 women followed for a median period of 77 month s. Regular alcohol intake was reported by 78% of men and 52% of women. Eighty-seven per cent of men and 44% of women primarily drank beer. R esults: Death occurred in 305 men and 236 women, 34% and 39% respectiv ely from coronary heart disease (CHD). In a proportional hazards model , the hazard ratio (HR) for all-causes mortality in male drinkers, com pared with abstainers, was 0.75 at one-seven drinks/week, 0.76 at eigh t-14 drinks/week, 0.69 at 15-28 drinks/week and 0.49 at >28 drinks/wee k (p<0.04), an inverse relationship. In female drinkers, HR was 0.78 a t one-seven drinks/week, 0.49 at eight-14 drinks/week (p<0.04) and 0.6 2 at 15-28 drinks/week, potentially a U shaped relationship. The effec t on all-causes mortality could not be attributed to a differential ef fect of beer versus wine/spirit intake. Although the mortality rate wa s lower in those taking any alcohol compared with abstainers, those ta king any alcohol exhibited an increased proportion of deaths due to ca ncer at the expense of a reduced proportion of CHD and stroke deaths. Conclusions: Alcohol intake in the Dubbo elderly appears to be indepen dently associated with a significant increase in life expectancy. Mech anisms underlying the effect may emerge at a longer interval of follow -up.