Impact of gender and androgen status on IGF-I levels in normal and GH-deficient adults.

Abstract

The regulation of IGF-I levels is complex and not only dependent on GH status, as the diagnostic sensitivity of serum IGF-I levels for GH deficiency (GHD) in adults is low. Other GH-related parameters have so far not proven to be of additional diagnostic value in GHD adults. In the present study we evaluated the impact of gender and androgen status on IGF-I levels and the diagnostic value of IGF-I and GH-related parameters in a population of adult hypopituitary patients and age- and gender-matched healthy subjects. A cross-sectional study. Fifty-nine GHD patients (40 males, mean age 39.3+/-1.7 (s.e.m.) years, and 19 females, mean age 41.9+/-2.6 years) and 69 healthy subjects (42 males, mean age 36. 7+/-1.5 years, and 27 females, mean age 38.9+/-2.1 years). IGF-I levels were low in the GHD patients (91+/-7 vs 173+/-7 microgram/l, P<0.001), and lower in female patients than in male (68+/-10 vs 100+/-8 microgram/l, P=0.03). In the control group there was no gender-related difference in IGF-I levels (males: 178+/-8, females: 164+/-12 microgram/l, P=0.23). IGF-II and IGF-binding protein-3 (IGFBP-3) were also decreased in GHD without any gender-related differences. GH-binding protein (GHBP) levels were increased in the patient group. The diagnostic sensitivity (%) of IGF-I, IGF-I/GHBP, IGF-I/IGFBP-3, and of the combination of IGF-I plus IGF-II (both low or one normal and one low), was higher in female patients than in male (IGF-I: 57.8 vs 22.0, P<0.0001; IGF-I/GHBP: 84.2 vs 48.8, P=0. 002; IGF-I/IGFBP-3: 36.8 vs 7.3 P=0.001; IGF-I+IGF-II: 77.8 vs 52.6, P=0.01). Testosterone levels were reduced in the female patients compared with female controls (0.5+/-0.3 vs 2.1+/-0.2nmol/l, P<0.001). Forward regression analyses revealed that IGFBP-3 was a significant predictor of IGF-I levels in both patients and healthy subjects. In a combined analysis of both patients and controls, sex hormone-binding globulin (SHBG) level was the main contributor as an explanatory variable. Gender and prolactin also predicted IGF-I in patients, whereas SHBG and estradiol were significant predictors only in the control group. (i) Levels of IGF-I, and of IGF-I/IGFBP-3 and IGF-I/GHBP ratios are lower in females compared with male adult GHD patients. (ii) IGF-I/GHBP has a high diagnostic sensitivity of adult GHD, in particular in women. (iii) We hypothesize that the gender difference in IGF-I levels among adult GHD patients are causally related to the very low androgen levels observed among females.