Long-term outcomes of drinking: a 20-year longitudinal study of men.

This paper examines long-term changes in drinking, with regard to: (a) the degree to which overall patterns of drinking have shifted or remained across a 20-year period; (b) whether these patterns vary on the basis of age; and (c) the degree to which distinct patterns of drinking may be differentially subject to mortality and/or nonresponse. Data for this investigation are from a 20-year prospective follow-up study of two general population surveys. The first of these was originally interviewed in 1964 and consists of interviews with 405 males aged 23 and older; the second, originally interviewed in 1967, consists of interviews with 786 males aged 21-59. The results indicate that while consumption was modestly associated with mortality from all causes, no significant relationship was observed between consumption and non-response. The results also indicate that as respondents aged 20 years, mean levels of alcohol consumption remained stable. This was true despite the fact that when individual respondents did change their drinking, they were more likely to decrease their consumption than increase it. These results do not support conclusion drawn from cross-sectional studies that aging modifies consumption patterns.