Mechanisms of Osteoporosis in Adult and Adolescent Women with Anorexia Nervosa*

TLDR

Osteoporosis is a known consequence of anorexia nervosa in adults, yet the mechanisms of bone loss remain unclear and studies of bone mass before peak bone mass are lacking. The study investigates the causes of bone loss in AN, the impact of adolescent onset on bone mass, and the role of endogenous cortisol excess. Researchers compared cortical and trabecular bone in 26 AN women using radial photon absorptiometry and spinal CT, and measured urinary cortisol/creatinine clearance. Adult AN patients exhibited marked spinal osteopenia, with earlier amenorrhea onset and longer duration linked to lower bone density, and elevated cortisol levels correlated negatively with bone mass, indicating that adolescent onset, prolonged amenorrhea, and cortisol excess contribute to severe osteopenia.

Abstract

Osteoporosis is a known consequence of anorexia nervosa (AN) in adults, but the mechanism of bone loss is not established, and there have been no studies of bone mass in women developing AN before attaining peak bone mass. To investigate the causes of bone loss in AN and to determine the consequences of developing AN during adolescence on bone mass, we compared the effects of AN on cortical and trabecular bone in 26 women with AN (19 adults, 18–42 yr old, and 7 adolescents, 14.9–17.0 yr old) using direct radial photon absorptiometry and both single and dual energy spinal computed tomography. The adult AN patients had marked spinal osteopenia [mean bone density, 120 ± 28 (±sd) mg K2HPO4/cm3] compared with age-matched normal women (mean, 176 ± 26 mg K2HPO4/cm3; P = 0.0001), which was severe (>2 sd below the normal mean) in 50% of patients. Adult AN patients with the onset of amenorrhea before age 18 yr had significantly (P = 0.04) lower spinal bone density than those developing amenorrhea later (mean, 103 ± 25 vs. 129 ± 25 mg K2HPO4/cm3) independent of other variables correlated with bone density (duration of amenorrhea or urinary cortisol excretion). We found a negative correlation between spinal bone density and duration of amenorrhea (P = 0.006; r = −0.53). To determine the contribution of endogenous cortisol excess to the pathogenesis of the osteoporosis, urinary cortisol was measured. Urinary cortisol/creatinine clearance was significantly (P = 0.001) higher in AN patients than in normal women (mean, 2.7 ± 1.2 vs. 1.0 ± 0.3 nmol/L) and was negatively correlated (P = 0.03; r = −0.43) with bone mass. We conclude that AN affects trabecular bone and that both the onset of AN before attainment of peak bone mass and prolonged duration of amenorrhea result in more severe osteopenia. In addition to the significant contribution of hypogonadism, the cortisol excess that occurs in AN patients may contribute to the development of osteoporosis in these patients.