Long-Term Effect of Exercise on Bone Mineral Density in Postmenopausal Women.

Abstract

0753 PURPOSE: A four-year follow-up to the Bone, Estrogen, Strength Training Study (BEST) designed to test the effects of exercise with and without hormone replacement therapy on bone mineral density was conducted. METHODS: Total body and regional bone mineral density (BMD) and soft tissue composition were measured twice (averaged for analyses) each year (year 1, 2, 3 and 4) by dual energy x-ray absorptiometry. All subjects received 800 mg/day calcium supplements. One-year data were collected on 266 women comprising 4 groups, HRT with exercise (n = 71), HRT with no exercise (n = 65), no HRT with exercise (n = 71), and no HRT with no exercise (n = 59). For the next 3 years 167 women continued in the study and were measured annually for BMD. High load, low rep weight lifting (70–80% 1RM, 2 sets of 6–8 reps, 8 exercises) and moderate impact aerobic weight-bearing exercise was supervised in the first year. After one year, all subjects exercised by choice and recorded their exercise monthly. The sample included women who continued exercise, women who discontinued exercise, and women who began exercise sometime during the followup period. Within the sample of 167, 30 exercisers and 23 controls remained in the same group all four years. The analysis included exercise attendance as a continuous variable averaged over the four years for the sample of 167 and used a group contrast as a predictor variable for the subsample of 53. RESULTS: BMD adjusted for baseline BMD, HRT status, and weight change increased in the exercise group as compared to controls (n = 30 exercisers, 23 control) for both femur trochanter and lumbar spine (P <.05), but not femur neck. Similar results were found in the larger sample using exercise attendance as a predictor variable. Weight change and exercise were the major contributors to predicted BMD change and the difference between exercise and control progressively increased in both trochanter and lumbar spine over the fouryear period. CONCLUSION: We conclude both trochanter and lumbar spine BMD were responsive to exercise over the four-year period. Supported by NIH Grant R01 AR39559 and Mission Pharmacal.