Relationship between growth hormone (GH) status, serum leptin and body composition in healthy and GH deficient elderly subjects.

OBJECTIVE

Growth hormone (GH) secretion declines with age and is affected by body composition. The signal that mediates the latter relationship remains III-defined. Leptin, the protein product of the adipocyte specific ob gene, is thought to accurately reflect fat mass and could therefore be a candidate to influence GH secretion. We have therefore investigated the relationship between GH status, leptin and body composition in normal and GH-deficient elderly subjects.

DESIGN

GH Secretion was assessed by 20-minute sampling over 24 hours and serum leptin concentrations were measured in a single morning, fasted sample.

PATIENTS

Twenty-one GH deficient elderly patients (61-83 years) and 22 gender- and BMI-matched controls (61-88 years).

MEASUREMENTS

Body composition was assessed by dual-energy X-ray absorptiometry (DEXA). GH was measured in an ultrasensitive chemiluminescent assay and serum leptin was determined by radio-immunoassay.

RESULTS

Leptin was correlated with percentage body fat in both sexes (male r = 0.75, female r = 0.89, both P < 0.001). Male patients had increased fat mass (FM) (P < 0.01) and leptin concentrations (P < 0.05) but similar lean mass (LM) compared with controls. However, leptin concentration per unit FM was identical in both groups (P = 0.3). In contrast, female patients had lower LM (P < 0.05) but similar FM to controls, yet their leptin concentration per unit FM was twice that of the controls (P < 0.05). In multiple linear regression (MLR) leptin was determined positively by FM and negatively by LM (controls r2 = 76%; patients r2 = 73%, both P < 0.0001). When controlled for gender, GH secretion in the controls was correlated negatively with leptin (r = -0.68, P < 0.01) and negatively with percentage body fat (r = -0.73, P < 0.01). In MLR, using leptin as a marker of body composition, 66% of the variability in GH secretion in the controls could be explained by gender (38%) and by leptin (28%).

CONCLUSIONS

Both decreased lean mass and increased fat mass raise serum leptin concentrations in normal and growth hormone-deficient elderly subjects. Leptin is therefore a marker of body composition rather than fat mass alone. The influence of body composition on growth hormone secretion in the elderly may be mediated through leptin, acting as a peripheral signal from adipose tissue to decrease GH secretion.