The aim of this study was to determine whether serum leptin concentrations are reduced in response to short-term energy restriction in centrally obese individuals with impaired glucose tolerance or non-insulin-dependent diabetes mellitus. Twenty African Americans [16 females and 4 males, 44 +/- 7 y (x +/- SD), 107.2 +/- 23.8 kg, 39 +/- 7% body fat] consumed a 7-d energy-restricted diet (4.03 +/- 0.72 MJ/d) of whole foods. Oral-glucose-tolerance tests (OGTTs) were performed before and immediately after the diet to assess changes in serum leptin, glucose, and insulin concentrations. Baseline leptin concentration correlated significantly with percentage body fat (r = 0.80), body mass index (r = 0.72), fat mass (4 = 0.64), waist-height ratio (r = 0.6), body weight (r = 0.59, all P < 0.01), waist circumference (r = 0.49), and basal insulin concentration (r = 0.48, both P < 0.05). Seven days of energy restriction resulted in significant reductions (P < 0.005) in leptin (-6.1 +/- 8.4 micrograms/L), basal glucose (-0.9 +/- 0.8 mmol/L), OGTT glucose area under the curve (-158 +/- 164 mmol/L), and basal insulin concentration (-34 +/- 69 pmol/L, P < 0.05). In addition, there was a trend for a reduction in OGTT insulin area under the curve (-15,567 +/- 3,658 pmol/L, P = 0.05), and a tendency for basal insulin and leptin to change together (r = 0.41, P = 0.07). Despite the weight loss of 3.1 +/- 1.3 kg (P < 0.0001), the loss of fat mass was calculated to be only -1.0 +/- 0.1 kg. These results suggest that negative energy balance or improved insulin action was responsible for the changes in leptin, glucose, and insulin concentrations. In summary, short-term energy restriction effectively reduced serum leptin concentrations and improved glucose tolerance and insulin action in obese individuals with impaired or diabetic glucose tolerance.