Risks of Bone Loss Persist Long After Diagnosis

An increasing body of research has revealed the deleterious effects of eating disorders upon bone density. Danish researchers at the Osteoporosis Clinic, Aarhus University, Aarhaus, Denmark, report that the risk of fractures persisted after the initial diagnosis among 4394 retrospectively studied patients with bulimia nervosa (BN) anorexia nervosa (AN), and eating disorders not otherwise specified (EDNOS). The increased risk of fractures years after diagnosis indicated permanent damage to the skeleton (Int J Eat Disord 2002;32:301).

Age at diagnosis was important

Except for fractures of the femur, patients with AN did not have an increased fracture risk before they were diagnosed. However, a significant increase in fracture risk was noted after diagnosis of AN, especially more than one year later.

Age at diagnosis was also important for patients with AN that was diagnosed after they were 20 years of age, for this group had a higher incidence rate ratio (IRR) compared to controls before diagnosis compared to those diagnosed before the age of 20 years. The IRR was established as a ratio of bone density between patients and controls. Each patient was compared with three age-, gender-, and social-status-matched controls.

Among patients with BN, the risk of any fracture was increased up to 10 years before diagnosis, but returned to normal more than one year after diagnosis. The EDNOS group had a significantly increased risk of any fracture both before and after diagnosis.

The increased fracture risk after diagnosis among the anorexic patients may indicate one of two things—first, that the anorexic state permanently damaged the skeleton, which was worsened by the age-related decline in bone mineral density or, second, that the treatment was not successful.

The observation in both AN and BN patients that the relative facture risk of diagnosis was increased with increasing age at diagnosis might suggest that the young skeleton has a better potential for recovery after treatment than does the older skeleton. Another explanation might be that treatment was delayed for the older subjects, leading to more-pronounced adverse effects on the skeleton.

The femoral neck was a target in patients with AN

The increase in fracture risk in AN was 1.98.Fracture risk at the femoral neck was 7.17. In the spine, the risk was 3.49. A loss of fat on the hips would mean a decrease in the shock-absorbing potential of this “fat cushion,” according to the authors, and the increase in femoral neck fractures before diagnosis suggests that this site may be more susceptible to the effects of malnutrition and reduction in soft tissue padding than other sites on the skeleton.

Among those with BN, there was a nonsignificant increase in several types of fractures, but only the increase in the total number of fractures was statistically significant. The authors think one explanation may be that in BN—in contrast to AN—the disorder may go unnoticed for a long time because patients do not present with any alarming signs of excessive weight loss; this might also be true for those with EDNOS.