Epidemiology of insulin-like growth factor-I in elderly men and women. The Rancho Bernardo Study.

Insulin-like growth factor-I (IGF-I) is abundant in the circulation and has been shown to have a wide array of biologic effects. The authors carried out a cross-sectional community-based study of 420 men and 419 nonestrogen-using postmenopausal women aged 50-97 years to ascertain the within-person and laboratory reliability of IGF-I measurements, and the association of IGF-I with common epidemiologic confounders. There was no evidence of seasonal or diurnal variation. IGF-I decreased linearly with age in both sexes, with significantly higher levels in men than women (134.1 microg/liter vs. 126.9 microg/liter; p = 0.03) [corrected]. In age-adjusted analyses, IGF-I was not associated with height, total or central body fat, lean body mass, current smoking, physical activity, or commonly used medications. By contrast, in both men and women who reported any alcohol use, IGF-I levels were significantly higher compared with those in men and women who reported no alcohol use, and alcohol as a continuous variable showed a significant positive linear trend in men (p = 0.0007). The authors conclude that IGF-I varied significantly only with age, sex, and alcohol use. The minimal number of confounding variables, good reliability, and little intraindividual variation suggest that IGF-I should be suitable for epidemiologic research.