We evaluated the effects of recombinant insulin-like growth factor-I (IGF-I) and growth hormone (GH) on calciotropic hormones and bone turnover markers in 16 healthy elderly women 71.9 +/- 1.3 years of age (mean +/- SEM). Subjects consumed a fixed diet providing 1000 mg of calcium and 0.9 g/kg of protein for 10 days before starting baseline 24-h urine and blood collections. Specimens were collected for 6 consecutive days before initiating subcutaneous injections of GH (25 micrograms/kg/day, n = 5) and IGF-I at 60 micrograms/kg b.i.d. (high-dose, n = 5) or at 15 micrograms/kg b.i.d. (low-dose, n = 6) for 28 days. Resorption markers included urine hydroxyproline (OHP), total pyridinolines (PYD), and N-telopeptide; formation markers include osteocalcin, skeletal alkaline phosphatase (sALP), and type I procollagen carboxy-terminal extension peptide (CICP). For each subject, baseline daily turnover markers varied substantially (DV = 16-22%). With GH and high-dose IGF-I, resorption and formation markers increased progressively to maximum levels at day 21. For GH, the increase in day 21 PYD, N-telopeptide, osteocalcin, and CICP was 143, 111, 53, and 81%, respectively (p < 0.96-0.02). For high-dose IGF-I, these increases were 108, 81, 77, and 111% (p < 0.02-0.002). However, with low-dose IGF-I no change was observed in resorption markers while osteocalcin and CICP increased progressively (day 21, % increases = 88 +/- 51, 36 +/- 14). Twenty-four hour urine collections during the last days of baseline and of study drug were taken as six 4 h aliquots. When deoxyPYD was measured on these samples in the low-dose IGF-I group, a significant increase was observed only on the 0800-1200 h aliquot. Serum phosphorus concentrations increased with GH (21.2 +/- 3.3%) and high-dose IGF-I (8.8 +/- 3.6%) by day 21 but actually decreased by day 28 (-9.7 +/- 2.7, p < 0.02) with low-dose IGF-I. Urinary phosphorus excretion decreased with high-dose IGF-I only. Twenty-four hour calcium excretion increased with all treatments. These results indicate that both GH and high-dose IGF-I activate remodeling osteons. By contrast, low-dose IGF-I may directly increase osteoblastic function with only a minimal increase in bone resorption and may therefore provide a useful means to increase bone mass. The results also suggest some of the GH action on renal phosphorus handling represents a direct action of GH on the nephron which does not involve the intermediacy of IGF-I. Finally, even under controlled conditions bone turnover markers exhibit substantial daily variation so that a very large treatment effect will be required for these markers to have clinical utility.