METHODS

The distribution of the osteoporosis in eating disor ders needs to be determined. Decreased bone density has been measured in multiple sites in AN patients including the radius, spine, and femur. Only one study compared these sites and found that the area most involved was the femur ( 7). Since femoral fractures carry the highest morbidity and mortality of osteopo rotic fractures, the distribution pattern of osteoporosis in eating disorders has important future implications for these patients (10). The contributing factors to osteoporosis in AN also are not clearly understood. Physical parameters such as body mass and level of exercise appear to be related (3,7). The role of estrogen deficiency, which predis poses the postmenopausab female to osteoporosis, is controversial (3,4,7). In order to determine the prevalence and distribution of osteoporosis in these eating disorder subgroups and the possible contributing factors, we studied hospital ized eating disorder patients from May 1987 to May 1988 by bone densitometry. Clinical parameters includ ing age, body surface area (BSA), age of onset of illness (AO), length of illness (LOI), exercise, and menstrual status were correlated with the bone density measure ments