CORRELATES OF POSTMENOPAUSAL ESTROGEN USE AND TRENDS THROUGH THE 1980S IN 2 SOUTHEASTERN NEW-ENGLAND COMMUNITIES

Correlates and prevalence of current postmenopausal estrogen use were examined in two southeastern New England communities for the period 19 81 through 1990. Data were obtained from five biennial cross-sectional household health surveys, with current use of postmenopausal estrogen determined by inspection of medication bottles. Analysis included 3,2 79 women aged 40-64 years, of whom 2,215 (67.6%) were postmenopausal. The prevalence of estrogen use among all postmenopausal women increase d from 5.3% (95% confidence interval (Cl) 3.2-7.4) in 1981-1982 to 10. 9% (95% Cl 7.5-14.4) in 1989-1990. Among women with surgical menopause , prevalence increased from 11.4% (95% Cl 6.6-16.2) to 20.3% (95% Cl 1 3.6-27.0), while use among women with natural menopause increased from 1.5% (95% Cl 0.04-2.9) to 3.5% (95% Cl 0.74-6.2). Logistic regression was used to compute age-adjusted prevalence of estrogen use according to demographic characteristics and cardiovascular risk factors. Postm enopausal estrogen use was positively associated with income, and a po sitive trend with education was suggested. Estrogen use was positively associated with high density lipoprotein cholesterol and was inversel y associated with body mass index. Nonsmokers, women who exercised at least once per week, and women who reported having their cholesterol c hecked in the past year were more likely to use estrogen. These findin gs suggest that estrogen users have a more healthy profile than do non users. In addition, the prevalence of postmenopausal estrogen use appe ars to be substantially lower in women with natural versus surgical me nopause throughout the period studied.