Effect of hormone therapy on lean body mass, falls, and fractures: 6-year results from the Women's Health Initiative hormone trials

Abstract

Objective: Loss of lean body mass with aging may contribute to falls and fractures. The objective of this analysis was to determine if taking postmenopausal hormone therapy (or HT: estrogen plus progestogen therapy or estrogen therapy alone) favorably affects age-related changes in lean body mass and if these changes partially account for decreased falls or fractures with HT.

Methods: Participants randomly assigned to either estrogen plus progestogen therapy (n = 543) or control (n = 471) and estrogen therapy alone (n = 453) or control (n = 474) and receiving dual-energy x-ray absorptiometry scans to estimate body composition during the Women's Health Initiative were evaluated. Falls and fracture occurrence were obtained by annual self-report. Fractures were confirmed by a clinical chart review.

Results: At 6 years postrandomization, lean body mass was not different between HT and control groups. Although lean body mass positively influenced bone mineral density, independent of HT status, the preserved lean body mass observed in the HT arms in the first 3 years did not significantly contribute to models evaluating HT influence on falls and fractures between years 3 and 6. Women taking at least 80% of their medication in the HT arms demonstrated fewer falls compared with placebo; this difference was not attributable to change in lean body mass.

Conclusions: Despite early preservation of lean body mass with HT (3 y), HT did not ameliorate long-term (6 y) loss in lean body mass with aging.

Figure 1

Change in kilograms of total body fat and lean body mass from baseline to year 3. ET Active n= 367, Placebo n=395; EPT Active n=438, Placebo n= 398. *p<0.05 for comparisons between active ET or EPT and respective placebo.

Figure 2

Change in kilograms of total body fat and lean body mass between years 3 and 6. ET Active n= 306, Placebo n=330; EPT Active n=369, Placebo n= 322. *p<0.05; ^†p=0.06 for comparisons between active ET or EPT and respective placebo.