Determinants of bone density and prevalence of osteopenia among female runners in their second to seventh decades of age

This is a cross-sectional study of spine and hip hone density (BMD) in 124 female athletes, aged 16-68 years, who trained for at least 3 hs/week. The aim was to document the effects of competitive running on BMD in women over a broad age range, Thirty-three subjects, aged <35 years, were currently o ligo- or amenorrheic and, of the 50 who were >40 years, and who were now me nstruating normally, 13 had previously been oligo- or amenorrheic, Fifty-tw o women <50 years of age had never had disturbed menses. Twenty-four older women were postmenopausal, Women who had never had menstrual disturbance ha d significantly increased bone density at the lumbar spine, femoral neck, a nd femoral trochanter, as compared with young normal European reference dat a (range from +0.4 population SD or T-score units to +1.2 units according t o measurement site and age group). In contrast, young amenorrheic or oligom enorrheic runners had reduced bone density, particularly at the spine (mean T score < -1.1), whereas older runners who previously had disturbed menses , but were now menstruating normally, had bone densities that were similar to sedentary young controls. Postmenopausal runners had bone density values that differed little from sedentary postmenopausal controls matched for ti me since menopause, after adjusting for the runners' lower body weight, Bon e density outcomes were related to candidate explanatory variables. After t aking into account the other variables, age, per se, influenced only the fe moral neck and Ward's area. Years since last exposure to estrogen (at preme nopausal levels) was an important determinant of bone loss at both hip and spine. Body weight had a beneficial influence on the femoral neck region, w hereas (in contrast) height had a positive influence on the lumbar spine. M onths of breastfeeding (totaled for all children) had a modest, positive in fluence, which was larger in the femoral measurement sites. There was no ev idence of an effect of calcium intake or percent body fat on BMD at any sit e independent of these other effects. It is concluded that, with the consis tent presence of normal premenopausal estrogen levels, running at least 3 h s/week substantially improves bone density, particularly at the proximal fe mur. This beneficial effect is reversed in the absence of the consistent pa st and current presence of normal menstrual function. There was no clear be nefit of running seen on BMD in postmenopausal women, but premenopausal vet eran athletes who started running after the age of 30 years were not disadv antaged compared with early starters. (Bone 26:591-598; 2000) (C) 2000 by E lsevier Science Inc. All rights reserved.