Insulin‐Like Growth Factor‐1

To the Editor: It is well known that insulin-like growth factor (IGF)-1 levels decrease with age,1 but in the absence of a true deficiency, it is not clear why so much variability exists in circulating levels of serum IGF-1 nor what contributes to this variability, aside from growth hormone levels.2 Cross-sectional studies are inconsistent regarding the associations between serum IGF-1 and lean tissue mass and bone mineral density (BMD) in the elderly, and there are few longitudinal studies addressing this issue in healthier older adults. Serum levels of IGF-1 in a well-functioning cohort of black and white men and women from the Memphis and Pittsburgh areas, aged 70 to 79 at baseline, were examined. A random sample (n=625) of participants was selected from the Health, Aging, and Body Composition Study, and IGF-1 in serum that had been frozen at −80°C after a fasting blood draw was measured. IGF-1 was separated from its binding proteins using acid-ethanol cryoprecipitation and measured using an IGF-binding protein-blocked radioimmunoassay (ALPCO, Windham, NH). The mean intra-assay coefficient of variation (CV) for the 20 batches run was 2.3% (range=0.43–5.87%), and the interassay CV was 3.2%. Total lean and fat mass and regional BMD were measured using dual-energy x-ray absorptiometry (DXA) at baseline and again 2 years later. Abdominal visceral fat and thigh muscle and fat areas were measured using computed tomography (CT) scanning. Thigh intermuscular fat was calculated using the deep fascial plane and muscle areas as guidelines. Thigh muscle density, as an indicator of fatty infiltration into muscle, was calculated using mean attenuation values in Hounsfield Units excluding intermuscular fat. Participants were divided into sex-specific groups by 25th and 75th percentile cutpoints for IGF-1. Those cutpoints were 94 ng/L and 148 ng/L for men and 77 ng/L and 127 ng/L for women, respectively. Analysis of covariance was used to examine the cross-sectional relationship between various body composition components and IGF-1 levels in 609 participants with valid IGF-1, CT, and BMD measurements (Table 1). Linear regression was then used to examine the association between baseline IGF-1 (log-transformed) and the prospective 2-year change in total lean body mass (n=514), total fat mass (n=528), and total hip BMD (n=526) from DXA with adjustment for age, sex, race, study site, thigh length (from CT), and baseline levels of lean body mass, total fat mass, and hip BMD for the respective analyses. IGF-1 was not associated with change in lean body mass (P=.12), fat mass (P=.27), or hip BMD (P=.56). Further adjustment for smoking, alcohol consumption, exercise level, or prevalent diseases at baseline did not affect these results, nor did further modeling using the 25th and 75th percentile cutpoints for IGF-1. With the exception of higher indices of thigh muscle mass and muscle density, variations in IGF-1 were not related to markers of body composition determined using DEXA or CT scanning at baseline or to longitudinal changes in body composition. The results, similar to those of many previous studies,3–5 raise the question: what does the variation within the normal range of IGF-1 in elderly adults relate to or predict? Despite the lack of consistency in prior results, studies continue to examine the body composition/IGF-1 relationship. It may be that the known role of IGF-1 in growth and development of lean mass in earlier ages is separate and distinct from its role in older ages. Numerous studies have linked IGF-1 levels in adulthood to incidence of cancer,6 and there are also data supporting its role in glucose homeostasis.7 Perhaps future work examining associations between IGF-1 with other physiological markers or diseases may shed light on the role of IGF-1 in the elderly.