Effects of androgenic steroids at doses used by athletes were studied in a canine model system in which dosage, diet, and activity were controlled. Dogs were treated with 19-nortestosterone (200 mg/wk intramuscularly) or vehicle and were studied at 18 (n = 4 in steroid and vehicle) or 32 (n = 6 in steroid and n = 4 in vehicle) days. A laparotomy was performed under general anesthesia 17 days before experimentation, and catheters were placed in an artery, portal vein, and hepatic vein. Studies consisted of an equilibration (120 minutes) and a control (40 minutes) period and a three-step immunoreactive insulin euglycemic clamp (1, 2, and 15 mU/kg.min). Step 1 was 150 minutes, and steps 2 and 3 were 90 minutes. Data were collected during the last 30 minutes of each step. Glucose and leucine kinetics were assessed with 3H-glucose and 14C-leucine. Plasma glucose in steroid and vehicle groups was 104 +/- 5 (mean +/- SE) versus 108 +/- 3 mg/dL and 100 +/- 5 versus 107 +/- 4 mg/dL at 18 and 32 days. Glucose turnover was similar at 18 days in steroid and vehicle groups (3.9 +/- 0.3 v 3.6 +/- 0.3 mg/kg.min, respectively), but was elevated in the steroid group at 32 days (5.4 +/- 0.5 v 3.2 +/- 0.4 mg/kg.min). Glucose infusion rates were lower in the steroid group with 15 mU/kg.min immunoreactive insulin at 32 days (15.0 +/- 1.1 v 21.2 +/- 1.4 mU/kg.min). Immunoreactive insulin-independent glucose utilization (Rd) was unaffected at 18 days of steroid treatment, but was increased by almost fourfold at 32 days.(ABSTRACT TRUNCATED AT 250 WORDS)