Background: The role of IGF-I in prostate development is currently under thorough investigation since it has been claimed that IGF-I is a positive predictor of prostate cancer. Objective: To investigate the effect of chronic GH and IGF-I de®ciency alone or associated with testosterone de®ciency on prostate pathophysiology in a series of patients with hypopituitarism. Design: Pituitary, androgen and prostate hormonal assessments and transrectal prostate ultrasonography (TRUS) were performed in 30 men with adulthood onset GH de®ciency (GHD) and 30 agematched healthy controls, free from previous or concomitant prostate disorders. Results: Plasma IGF-I levels were signi®cantly lower in GHD patients than in controls (Pearson's coef®cient P < 0.0001). At study entry, 6 of the 13 hypogonadal patients and 7 of the 17 eugonadal patients had plasma IGF-I below the age-adjusted normal range. At study entry, testosterone levels were low in 13 patients (mean 6 S.E.M., 3.8 6 1.0 nmol/l) while they were normal in the remaining 17 (19.4 6 1.4 nmol/l). No difference in prostate-speci®c antigen (PSA), and PSA density was found between GHD patients (either hypo-or eugonadal) and controls, while free PSA levels were signi®cantly higher in eugonadal GHD than in controls (0.4 6 0.04 vs 0.2 6 0.03 mg/l; P < 0.01).No difference in antero-posterior prostate diameter and transitional zone volume (TZV) was observed among groups, while both transverse and cranio-caudal diameters were signi®cantly lower in hypogonadal (P < 0.01) and eugonadal GHD patients (P < 0.05) than in controls. Prostate volume (PV) was signi®cantly lower in hypogonadal GHD patients (18.2 6 3.0 ml) and eugonadal GHD patients (22.3 6 1.6 ml), than in controls (25.7 6 1.4, P < 0.05). The prevalence of prostate hyperplasia (PV > 30 ml) was signi®cantly lower in hypogonadal and eugonadal GHD patients, without any difference between them (15.3% and 5.8%), than in controls (43.3%) (x 2 6:90, P 0:005). No difference was found in PV between patients with normal or de®cient IGF-I levels both in the hypogonadal group (19.9 6 4.7 vs 17.3 6 4.0 ml) and in the eugonadal group (22.6 6 2.3 vs 21.8 6 2.5 ml). When controls and patients were divided according to age (<60 years and >60 years), PV was signi®cantly lower in hypogonadal GHD patients aged below 60 years than in agematched controls (P < 0.01) or eugonadal GHD patients (P < 0.01), without any difference between controls and eugonadal GHD patients. Controls aged above 60 years had signi®cantly higher PV than both hypogonadal and eugonadal GHD patients (P < 0.01). Calci®cations, cysts or nodules were found in 56.7% of patients and in 50% of controls (x 2 0.067, P 0:79). In controls, but not in GHD patients, PV and TZV were correlated with age (r 0:82, r 0:46, P < 0.0001 and P < 0.01 respectively). PV was also correlated with GH (r À0:52, P 0:0026), IGF-I (r À0:62, P = 0.0002) and IGF-binding protein 3 (IGFBP-3) levels (r À0:39, P 0:032) but neither with testosterone or dihydrotestosterone (DHT) levels. In GHD patients T...