Relationship of plasma leptin to plasma insulin and adiposity in normal weight and overweight women: effects of dietary fat content and sustained weight loss.

Leptin, the product of the human homologue of the ob gene, which is defective in the obese (ob/ob) mouse, may be a humoral regulator of human adiposity. Plasma leptin concentrations were measured by RIA in 19 normal weight [body mass index (BMI) = 24.5 +/- 0.6 kg/m2] and 19 overweight to obese (BMI = 34.7 +/- 1.2 kg/m2) nondiabetic postmenopausal women on sequential controlled weight-maintaining diets containing 31%, 23%, and 14% of energy as fat, each for 4-6 weeks. Thereafter, the subjects ate a very low fat diet (< 15%) ad libitum; plasma leptin and insulin concentrations, BMI, percent body fat (%BF), and resting energy expenditure were determined after 6 and 8 months. Absolute and adiposity-corrected plasma leptin levels were higher in overweight/obese women (37.7 +/- 3.5 ng/mL; 1.01 +/- 0.07 ng.mL-1.%BF-1) than in normal weight women (16.9 +/- 2.2 ng/mL; 0.57 +/- 0.06 ng.mL-1.%BF-1, both P < 0.005 vs. obese), but were not different between the 31%, 23%, and 14% fat diets when body weight was stable. Plasma leptin was highly correlated with BMI (r = 0.81, P < 0.0001), %BF (r = 0.80, P < 0.0001), and fasting plasma insulin (r = 0.61, P < 0.0001). After 8 months on the ad libitum low fat diet, the women had lost an average of 6.9 +/- 1.0% of body mass (-2.0 +/- 0.3 kg/m2, P < 0.0001). In 15 subjects who lost more than 7% of body mass (-12.3 +/- 1.0%), plasma leptin concentrations decreased (-9.6 +/- 1.9 ng/mL, P < 0.0005), and the decrease of plasma leptin per change of adiposity (delta leptin/delta %BF) was greater in overweight/obese women (3.6 +/- 0.5) than in normal weight women (0.9 +/- 0.4, P < 0.01 vs. obese). In 18 other subjects who lost less than 7% of body mass (-2.7 +/- 0.6%), plasma leptin was unchanged (+1.4 +/- 1.4 ng/mL). Overall, the change of plasma leptin was significantly correlated with change of BMI (r = 0.43, P < 0.02), the change of %BF (r = 0.49, P < 0.005), the change of resting energy expenditure (r = 0.40, P < 0.02), and with the change of plasma insulin independently of changes of body adiposity (r = 0.45, P < 0.01). We conclude that plasma leptin concentrations are: 1) not affected by dietary fat content per se; 2) highly correlated with BMI, %BF, and plasma insulin in both overweight/obese and normal weight women; 3) decreased in parallel with plasma insulin after sustained weight loss; and 4) decreased more in overweight/obese than in normal weight women