Testosterone Administration in Women With Anorexia Nervosa

Abstract

Complications of anorexia nervosa (AN) include marked bone loss, cognitive dysfunction, and major depressive disorder. A lifetime prevalence of depression as high as 68% has been reported in women with AN. Because testosterone (T) increases bone density and may lessen depression in hypogonadal men, the investigators proposed that short-term testosterone, given in physiological doses, would promote bone formation, reduce depressive symptoms, and improve spatial cognition in women with AN. The study group included 33 women who weighed less than 85% of ideal body weight and whose serum free T was less than the median of the reference range for premenopausal women. All of them had been amenorrheic for 3 months or longer, and all exhibited psychiatric features of AN. Participants were randomized to receive transdermal T in a dose of 150 or 300 μg or placebo for 3 weeks. The 3 groups were similar demographically and clinically at the outset. Treated patients had significant increases in serum total and free T levels. A majority of women had free T levels at 3 weeks that exceeded the upper limit of normal for premenopausal women. There were no changes in levels of estradiol, sex hormone-binding globulin, dehydroepiandrosterone sulfate, or insulin-like growth factor-I. Elevations of serum free T were most marked in higher-weight patients. Levels of C-terminal propeptide of type 1 collagen were higher during T treatment than in placebo recipients and correlated with the change in free T over 3 weeks. No changes in osteocalcin or bone-specific alkaline phosphatase were noted. The Beck Depression Inventory demonstrated improved mood in depressed women who received T compared with those on placebo (Fig. 4). The mean score improved from the severely to the moderately depressed range. There were no changes in the use or dose of antidepressants during the study period. Testosterone treatment appeared to improve spatial cognition as reflected by scores on a test of visuospatial ability (Fig. 5). The transdermal patches were, in general, tolerated well. Hirsutism scores did not change over 3 weeks. T administration did not lead to significant changes in blood lipid or lipoprotein levels. Low-dose T, administered transdermally, holds promise for promoting bone formation and lessening depression in women with AN, but the present findings must be confirmed in larger and longer-lasting randomized, placebo-controlled studies. At present, testosterone is not recommended for these patients.