Back problems among postmenopausal women taking estrogen replacement therapy: the study of osteoporotic fractures.

Abstract

Cross-sectional and prospective. To investigate the association between estrogen replacement therapy use, back pain, and back function in a large cohort of elderly women. Several studies have suggested that women who use estrogen replacement therapy may be more likely to experience back pain than those who do not. However, the relationships between estrogen replacement therapy, back pain, and impaired back function have not been clearly delineated. At baseline information on estrogen replacement therapy use, functional status, back pain and function, and general lifestyle variables was obtained from 7209 elderly white women (mean age 71 years)enrolled in the Study of Osteoporotic Fractures. Lateral radiographs of the lumbar and thoracic spine were taken at baseline and at the third clinic visit, an average of 3.7 years after the baseline visit. Bone mineral density at the hip and spine was measured approximately 2 years after baseline. Follow-up information on back pain and function was also obtained at the third clinic visit. A total of 1039 (14.4%) women were using estrogen replacement therapy at baseline, 2016(28.0%) reported former use, and 4154 (57.6%) had never used estrogen replacement therapy. Compared with never-users, a statistically significant higher percentage of current estrogen users reported clinical back pain (52.7% vs. 43.4%) and back impairment (12.3% vs. 9.2%) at baseline and at the follow-up visit (pain 50.8% vs. 41%; impairment 16.0% vs. 12.1%). This occurred despite a higher prevalence of vertebral fractures in never-users of estrogen at the baseline visit. Current and former estrogen users without vertebral fractures had statistically significant higher likelihoods of having back pain and back dysfunction at both the baseline and third follow-up visit. The increased likelihood of back pain and back impairment in current and former estrogen users remained despite statistical adjustment for age, vertebral fracture, body mass index,smoking history, parity, exercise, arthritis, and diabetes in multivariate models. The relative risk (95%confidence interval) for impaired back function in former and current users at follow-up was 1.1 (0.9, 1.3) and 1.6 (1.3, 2.0), respectively. Our results indicate that postmenopausal estrogen use is associated with an increased likelihood of back pain and impaired back function in elderly white women.