To investigate whether previously reported increased levels of insulin-like growth factor-binding protein-1 (IGFBP-1) in GH-deficient patients only reflect decreased levels of insulin or are elevated in relation to insulin, diurnal profiles of IGFBP-1 and insulin were determined in plasma from patients with GH levels below 0.2 microgram/L throughout 24 h (n = 23) and compared to profiles from patients with insulin-dependent diabetes mellitus (IDDM; n = 9) and healthy subjects (n = 12). Samples were drawn using a continuous withdrawal technique at 20-min intervals. Levels of IGF-I were determined in one morning sample. As in healthy subjects, serum IGFBP-1 displayed a diurnal variation in GH-deficient as well as in IDDM patients, with lowest levels in the afternoon and evening and a rise with maximum levels during the night and morning. Fasting and 24-h mean levels of IGFBP-1 were significantly higher in GH-deficient patients [61 +/- 12 (P < 0.01) and 39 +/- 6 micrograms/L (P < 0.01), respectively] and IDDM patients [72 +/- 18 (P < 0.01) and 45 +/- 9 micrograms/L (P < 0.01), respectively] compared to those in healthy subjects (27 +/- 4 and 18 +/- 2 micrograms/L, respectively). Fasting levels of IGFBP-1 correlated to 24-h mean values of IGFBP-1 in all groups separately and in the combined group (r = 0.931; P < 0.001). An inverse relationship was found between IGFBP-1 and insulin in GH-deficient patients, both between 24-h mean values (r = -0.756; P < 0.001) and between fasting values (r = -0.721; P < 0.001). Corresponding values for healthy subjects were r = -0.548; P = 0.065 and r = -0.712; P < 0.01, respectively, whereas in IDDM patients the relationship was nonsignificant. Moreover, in GH-deficient patients, the diurnal mean levels of IGFBP-1 were inversely correlated to IGF-I (r = -0.477; P < 0.05) and body mass index (r = -0.450; P < 0.05). When insulin was taken into account, a tendency for a negative correlation between IGFBP-1 and IGF-I (P = 0.054) remained, whereas the relationship to body mass index disappeared. However, IGFBP-1 levels were elevated in relation to insulin levels in GH-deficient patients compared to healthy subjects (F = 48.7; P < 0.001 and F = 32.5; P < 0.001, diurnal mean values and fasting values, respectively). The majority of the IDDM patients had values in the same range as the GH-deficient patients.(ABSTRACT TRUNCATED AT 400 WORDS)