The effects of anorexia nervosa on bone metabolism in female adolescents.

Osteopenia is a frequent, often persistent, complication of anorexia nervosa (AN) in adolescent girls and occurs during a critical time in bone development. Little is known about bone metabolism in this patient population. Therefore, we measured bone density (BMD) and body composition by dual energy x-ray absorptiometry, nutritional status, bone turnover, calcium, and hormonal status in 19 adolescent girls with AN (mean +/- SEM, 16.0+/-0.4 yr) and 19 bone age-matched controls. The mean duration of AN was 19+/-5 months. Spinal (L1-L4) osteopenia was common in AN. Lumbar anterioposterior BMD was more than 1 SD below the mean in 42% of patients, and lateral spine BMD was more than 1 SD below in 63% of patients compared with controls. Lean body mass significantly predicted lumbar bone mineral content (r = 0.75; P < 0.0001) in controls only. In AN, duration of illness was the most significant predictor of spinal BMD (lumbar: r = -0.44; P = 0.06; lateral: r = -0.59; P = 0.008). AN adolescents with mature BA (15 yr and greater) were hypogonadal [estradiol, 16.2+/-1.9 vs. 23.3+/-1.6 pg/mL (P = 0.01); free testosterone, 0.70+/-0.17 vs. 1.36+/-0.14 pg/mL (P = 0.01)] although dehydroepiandrosterone sulfate and urinary free cortisol levels did not differ. Leptin levels were reduced in AN (2.9+/-2.1 vs. 16.5+/-1.8 ng/mL; P < 0.0001). Insulin-like growth factor I (IGF-I) was reduced in AN to 50% of control levels (219+/-41 vs. 511+/-35 ng/mL; P < 0.0001) and correlated with all measures of nutritional status, particularly leptin (r = 0.80; P < 0.0001). Surrogate markers of bone formation, serum osteocalcin (OC) and bone-specific alkaline phosphatase (BSAP), were significantly (P = 0.02) reduced in AN vs. controls (OC, 39.1+/-6.4 vs. 59.2+/-5.2 ng/mL; BSAP, 27.9+/-4.0 vs. 40.6+/-3.4 U/L). The majority of the variation in bone formation in AN was due to IGF-I levels (OC: r2 = 0.72; P = 0.002; BSAP: r2 = 0.53; P = 0.01) in stepwise regression analyses. Bone resorption was comparable in patients and controls. These data demonstrate that bone formation is reduced and uncoupled to bone resorption in mature adolescents with AN in association with low bone density. Lean body mass was a significant predictor of BMD in controls, but not AN patients. The major correlate of bone formation in AN was the nutritionally dependent bone trophic factor, IGF-I. Reduced IGF-I during the critical period of bone mineral accumulation may be an important factor in the development of osteopenia in adolescents with AN.