Determinants of serum insulin-like growth factor I in growth hormone deficient adults as compared to healthy subjects

OBJECTIVE Growth hormone status is an important determinant of serum IGF-I but it is well known that hypopituitary adults with pronounced GH-deficiency (GHDA) may exhibit normal IGF-I levels. To elucidate possible causes of this apparent paradox we compared the significance of putative IGF-I predictors in GHDA and normal subjects. DESIGN A cross-sectional study. SUBJECTS Twenty-seven GHDA (9 females, 18 males, mean +/- SE age 44 +/- 1 years) and 27 healthy control subjects (9 females, 18 mates, mean +/- SE age 43 +/- 2 years). RESULTS Serum IGF-I and IGFBP-3 were significantly lower in GHDAs, but a considerable overlap existed (IGF-I (mu g/l) 87 +/- 12 (GHDA) vs 177 +/- 10 (Control) (P < 0.001)). In both Controls and GHDA, IGF-I was higher in males than females (Control: 196 +/- 12 vs 138 +/- (P = 0.004); GHDA: 97 +/- 16 vs 56 +/- 11 (P = 0.05)). In GHDA, males on testosterone substitution had the highest IGF-I concentrations. The molar IGF-I:IGFBP-3 ratio was significantly lower in GHDAs (0.18 +/- 0.01 vs 0.23 +/- 0.02 (P = 0.002)). IGFBP-1 (mu g/l) was significantly elevated in GHDAs (6.28 +/- 1.11 vs 3.07 +/- 0.32 (P < 0.001)) despite comparable fasting insulin levels. Percentage total body fat (TBF, DEXA, waist/hip ratio, and intra-abdominal fat (CT) were all elevated in GHDAs. IGF-I correlated positively with lean body mass (DEXA) and negatively with TBF and IGFBP-1 in both groups. IGF-I correlated negatively with age in CON but not in GHDAs, whereas IGF-I correlated positively with IGFBP-3 only in GHDAs. Multiple regression analysis revealed that age and IGFBP-1 were the only significant predictors of IGF-I in CON, whereas IGFBP-3 and, to a lesser extent TBF, were the only independent predictors of IGF-I in GHDAs. Neither peak stimulated GH, nor physical fitness contributed in any equations in the two groups. CONCLUSIONS 1) IGF-I levels are regulated by several variables in addition to GH status 2) age per se is an independent negative determinant in healthy subjects but not in GHDA 3) It is probable that some cases of paradoxically high IGF-I levels in GHDA are secondary to inappropriately elevated IGFBP-3 levels. 4) in mid-adulthood males have higher IGF-I levels than females and it is likely that testosterone directly stimulates IGF-I. The influence of gender and sex steroids must therefore be accounted for when comparing IGF-I levels between hypopituitary and healthy subjects.