Transient Hypogonadotropic Hypogonadism Caused by Critical Illness*

Abstract

The effects of acute severe illness on pituitary-gonadal function were determined in 35 men and 19 women, including 12 who were postmenopausal. Seventeen men and 5 women had traumatic brain injury which resulted in coma. Twelve postmenopausal and 2 premenopausal women had intracranial vascular accidents. Eleven men had myocardial infarctions, while 7 men underwent elective surgery. Serial plasma samples were examined for testosterone (men), percentage of ultrafiltrable testosterone (men), estradiol (women), sex hormone-binding globulin, LH, and FSH. In men, mean testosterone levels fell by 271 +/- 72 (+/- SE), 202 +/- 63 and 195 +/- 75 ng/dl within 24 h of brain injury, myocardial infarction, or elective surgery, representing decreases of 55%, 43%, and 58%. Further declines occurred in the first and third groups to mean nadirs of 93 +/- 16 and 117 +/- 5 ng/dl, respectively. During recovery of neurological function there was no correlation between the testosterone level and the degree of neurological impairment; testosterone levels eventually returned to normal (627 +/- 77 ng/ml). The percentage of ultrafiltrable testosterone and sex hormone-binding globulin did not change in any group. Although significant decreases in mean immunoreactive LH and FSH levels were found after head trauma, and decreases in FSH were found in the men after surgery, these changes occurred after the decline in testosterone. Despite the fall in basal gonadotropin levels in the head trauma group, there were no significant differences in the gonadotropin responses to GnRH (100 micrograms) in 4 patients during their acute illness or recovery. LH, FSH, and estradiol levels in the premenopausal women were significantly lower on the second day of brain injury (LH, 10.3 +/- 4.7 vs. 3.5 +/- 0.6 mIU/ml; FSH, 3.8 +/- 1.9 vs. 1.4 +/- 0.8 mIU/ml, estradiol, 200 +/- 41 vs. 102 +/- 16 pg/ml) and remained suppressed for 7 days. Gonadotropin levels also fell in the postmenopausal women within 24 h; reductions in LH of 74% and in FSH of 62% were present by day 7 of study. We conclude that both men and women who are critically ill uniformly develop temporary hypogonadotropic gonadal insufficiency regardless of their illness. In men, it is manifested by low testosterone levels, while a comparable decrease in estradiol is present in women. The low testosterone concentrations are not due to reduced sex hormone-binding capacity. Based upon our data in postmenopausal women, hypogonadotropism also occurs in the presence of nonfunctioning gonads. Although our studies do not completely establish the pathophysiology of this disorder, they suggest a suprapituitary origin.