Anorexia Nervosa

Calcif Tissue Int. 2007 Sep;81(3):174-82.

Evaluation of bone loss and its mechanisms in anorexia nervosa.
Legroux-G?ot I, Vignau J, D'Herbomez M, Collier F, Marchandise X, Duquesnoy B, Cortet B.

Department of Rheumatology, University Hospital of Lille, H?ital Roger Salengro, 59037 Lille c?ex, France. i-legroux@chru-Lille.fr

The purpose of this cross-sectional study was to assess the extent of and mechanisms involved in bone loss in anorexia nervosa patients. We compared 113 anorexia nervosa patients (mean age 25 +/- 8 years, mean duration of disease 5.7 +/- 6.1 years) with 21 age-matched controls. Mean duration of amenorrhea was 3.2 +/- 4.7 years. We measured serum calcium and phosphate; bone remodeling markers (osteocalcin, bone-specific alkaline phosphatase [BSAP], serum crosslaps [CTX], and carboxyl-terminal telopeptide of type I collagen [ICTP]); follicle-stimulating hormone and luteinizing hormone levels; and estradiol (ultrasensitive assay), cortisol, urinary free cortisol, thyroid function, prolactin, and nutritional factors (insulin-like growth factor I [IGF-I], IGF binding protein 3 [IGFBP3]). In controls, only bone remodeling markers and nutritional factors were measured. Osteodensitometry was also performed on both patients and controls. Weight and body mass index (BMI) were significantly lower in anorexia nervosa patients than in controls (P < 0.0001). No significant differences were observed in biological indicators except for IGF-I, which was lower in anorexia nervosa patients (0.9 +/- 0.4 UI/mL) than in controls (1.5 +/- 0.4 UI/mL) (P < 0.0001). Densitometric measurements at three sites were significantly lower in anorexia nervosa patients and correlated with duration of disease and amenorrhea and with IGF-I at the hip only (P < 0.01). In the study population, osteoporosis was observed in 24 patients (21%) and osteopenia in 54 patients (48%). Patients with osteoporosis were significantly older and had longer disease and amenorrhea durations; lower weight and BMI; higher alkaline phosphatase, BSAP, and osteocalcin; and lower serum ICTP, IGF-I, and IGFBP3. All of these differences were significant and remained so even after multiple adjustments were made, except for IGF-I (P = 0.21). When multivariate analysis was performed, we found that age at onset of amenorrhea, weight, alkaline phosphatase, urinary free cortisol, and serum estradiol concentration accounted for 54% of the variance in spinal bone mineral density (BMD). Duration of amenorrhea, alkaline phosphatase, and weight explained 46.6% of the variance in femoral neck BMD. Duration of amenorrhea, IGF-I, and ICTP levels accounted for 38.6% of the variance observed in total hip BMD. The etiology of bone loss in patients with anorexia nervosa is multifactorial. Hypoestrogenia alone cannot account for this loss, and nutritional factors, IGF-I concentrations in particular, seem to play an important role.