Eating disorders, menstrual dysfunction, weight change and DMPA use predict bone density change in college-aged women

Abstract

IntroductionThere are limited longitudinal studies that have evaluated bone mineral density (BMD) changes in college-aged women. Our objective was to simultaneously evaluate factors influencing 4-year BMD change. MethodsThis was a longitudinal cohort study of healthy, physically active women in the US Military Academy (n=91; average age=18.4years). Assessments over four years included: height, weight, calcium intake, physical fitness, menstrual function (annual number cycles), oral contraceptives (OCs) or depot-medroxyprogesterone acetate (DMPA) use, and eating disorder behavior (Eating Disorder Inventory; (EDI)). BMD was measured annually at the lumbar spine and total hip by dual X-ray absorptiometry and calcaneal BMD by PIXI. Slope of 4year BMD change at each skeletal site (spine total hip and calcaneus) was calculated for each woman. ResultsBMD gains occurred at the spine in 50% and the hip in 36% of women. In unadjusted analyses, spine bone gain was positively related to menstrual cycle frequency (p=0.04). Spine and hip BMD loss occurred in those using DMPA (p<0.01) and those with the highest EDI quartile scores (p<0.05). BMD change was unrelated to OC use. Hip and calcaneus BMD decreased with weight loss (average 4.8+2.2lb/year) as compared to those with stable weight/weight gain (p<0.05). In multivariable analysis, spine BMD increase was significantly related to African American (AA) race, normal EDI score and normal menses. Hip BMD increase was related to AA race, weight increase and normal menses. DMPA use was associated with spine, hip, and calcaneus bone loss. ConclusionOn average, BMD may modestly increase in college-aged women, in the absence of risk factors. However, risk factors including subclinical eating disorders, weight loss, menstrual dysfunction and DMPA use can have significant detrimental effects on BMD in young healthy physically active women.