PROBLEM DRINKING IN WOMEN--A 20-YEAR NATIONAL STUDY

点击翻译按钮获取中文摘要

This application proposes a national survey of 1,550 women in 2001 to increase knowledge about longitudinal patterns of women's drinking. The survey will include 700 women interviewed in 1981 and 1991, 350 women first interviewed in 1991, and a new sample of 500 women age 21-30 in 2001. (Subsamples of women were also interviewed in 1986 and 1996). Combining the 2001 survey with preceding surveys will produce 20-year cross-sectional data for all age groups, 20-year multiwave longitudinal data from women over age 40 in 2001, and 10-year longitudinal data from women age 31-40 in 2001. Specific aims of the proposed research are to evaluate (1) 20-year trends and age, period, and cohort effects in women's drinking behavior; (2) predictors of 5-, 10- and 20-year age-specific changes in women's drinking behavior; (3) effects on and from women's drinking trajectories across the adult life span; (4) correlates and predictors of heavier drinking among older women, and among women of childbearing age; (5) effects of question formats and interview modes on women's drinking self-reports; (6) links of women's drinking patterns with disordered eating behavior and with use of prescribed psychoactive drugs; and (7) cross-national variations in women's drinking behavior and its antecedents and consequences, using data from an international collaborative project coordinated by our research group. In the 2001 survey, professional female interviewers will conduct 75- minute face-to-face interviews using many questions from previous surveys about drinking patterns, drinking-related problems, changes in work and family roles, depressive symptoms, sexual and reproductive experience, and relationships with significant others. New questions will include a measure of trait impulsivity and additional questions about binge eating, estrogen replacement therapy, antidepressant use, and health problems of older age. Data analysis will include cross- tabular, correlational, and regression analyses; analysis of variance; cluster analysis (of drinking partnerships and drinking trajectories); structural equation modeling (for longitudinal prediction of 2001 drinking patterns); and generalized estimating equations, random regression models, latent transition analysis, and survival analysis (for comparing trends and trajectories and for predicting trajectories). The 2001 survey, combined with data from the 1981, 1986, 1991, and 1996 surveys, will yield the largest, longest-term, and most detailed set of longitudinal and life-historical data yet available about U.S. women's drinking and its antecedents and consequences. Together with findings from the international collaborative gender and alcohol project, issues addressed by the proposed analyses of these data should provide a strong foundation for efforts to improve the prevention and treatment of women's problem drinking in the 21st century.