Interrelationships of spontaneous growth hormone axis activity, body fat, and serum lipids in healthy elderly women and men.

Aging is associated with decreased growth hormone (GH) secretion and plasma insulin-like growth factor-I (IGF-I) levels, increased total and abdominal fat, total and low-density lipoprotein (LDL) cholesterol, and triglycerides, and reduced high-density lipoprotein (HDL) cholesterol. Similar changes in lipids and body composition occur in nonelderly GH-deficient adults and are reversed with GH administration. To examine whether GH/IGF-I axis function in the elderly is related to the lipid profile independently of body fat, we evaluated GH secretion, serum IGF-I and IGF binding protein-3 (IGFBP-3) levels, adiposity via the body mass index (BMI), waist to hip ratio (WHR), dual-energy x-ray absorptiometry (DEXA), and magnetic resonance imaging (MRI), and circulating lipids in 101 healthy subjects older than 65 years. Integrated nocturnal GH secretion (log IAUPGH) was inversely related (P < .005) to DEXA total and abdominal fat and MRI visceral fat in both genders. Log IAUPGH was inversely related to visceral fat in women (P < .005) and men (P < .0001), but was not significantly related to total fat in either gender. In women, log IAUPGH was related inversely to total and LDL cholesterol and positively to HDL cholesterol (P < .008). In men, log IAUPGH was inversely related to total cholesterol and triglycerides (P < .005). In women, HDL cholesterol was inversely related to the WHR (P < .005). In men, triglycerides were positively related (P < .001) to the WHR and DEXA abdominal and MRI visceral fat. Multivariate regression revealed log IAUPGH, but not DEXA total body fat, to be an independent determinant of total (P < .001 for women and P = .01 for men) and LDL (P < .007 and P = .05) cholesterol in both sexes and of HDL cholesterol (P < .005) and triglycerides (P < .03) in women. Log IAUPGH, but not DEXA abdominal fat, was related to total (P < .005 and P < .03) and LDL (P < .03 and P = .05) cholesterol in both genders and to HDL in women (P < .05). Log IAUPGH, but not MRI visceral fat, was related to total cholesterol (P < .03 and P = .05) in women and men. Age, IGF-I, and IGFBP-3 were not significantly related to any body fat or lipid measures, except for a positive correlation of IGF-I with triglycerides in men. Thus, endogenous nocturnal GH secretion predicts total, LDL, and HDL cholesterol levels independently of total or abdominal fat, suggesting that it is an independent cardiometabolic risk factor in healthy elderly people.