Hormone replacement therapy and body size: How much does lifestyle explain?, ,

OBJECTIVE: Our purpose was to assess the association of hormone replacement therapy with body size. STUDY DESIGN: A total of 1658 randomly selected women aged 45 to 64 years in four regions (two provinces in eastern Finland, one southwestern province, and the capital area) were studied in a 1992 population survey. Linear regression analysis was used to assess the use of hormone replacement therapy as a determinant of body size, measured with the body mass index, the waist/hip ratio, and the body fat percentage and adjusted by sociodemographic (area and education) and lifestyle (diet, smoking, physical activity, and alcohol use) factors. The regression coefficients of hormone replacement therapy use obtained from the model when including sociodemographic and lifestyle variables (β1a[Hormone replacement therapy use] = −0.98) were compared with the coefficients obtained when these variables were excluded (β1b[Hormone replacement therapy use] = −1.36). The percentage of change (β1b − β1a/β1b%) in the regression coefficients denoted the variability explained by sociodemographic and lifestyle characteristics. RESULTS: Hormone replacement therapy users (28%, n = 463) had higher education, were more often from the capital area, had a significantly higher healthy diet factor score, and were leaner than nonusers. Use of hormone replacement therapy remained a significant determinant of body mass index, waist/hip ratio, and body fat percentage after adjusting for sociodemographic and lifestyle factors. Of the difference in body mass index between hormone replacement therapy users and non-users, 28% were explained by lifestyle and sociodemographic risk factors, respectively, 31% of the difference by use of the waist/hip ratio and 42% by use of the body fat percentage. CONCLUSIONS: The difference in body size between hormone replacement users and nonusers may not be the result of self-selection. The use of hormone replacement therapy was inversely associated with body size in menopause. (Am J Obstet Gynecol 1998;178:66-73.)