To determine the influence of circadian rhythmicity and sleep on the 24-h leptin diurnal variations, plasma leptin levels were measured at 10-min intervals over 24 h in seven normal subjects, once during nocturnal sleep, and once after an 8-h shift of sleep. The subjects were submitted to constant conditions (continuous enteral nutrition and bed rest in controlled chambers). Body temperature and plasma glucose and insulin levels were measured simultaneously. During nighttime sleep, leptin levels increased to a maximum (109.9 +/- 2.5% of the 24-h mean) and then decreased to reach a nadir in the late afternoon. The mean diurnal variation was 18.0 +/- 3.8% of the 24-h mean. In the daytime sleep condition, leptin levels rose during the night of deprivation to a maximum of 104.7 +/- 2.3% of the 24-h mean, decreased to a minimum around 0700 h, and then rose again during diurnal sleep (108.4 +/- 3.1% of the 24-h mean); the mean diurnal variation was 13.4 +/- 3.6% of the 24-h mean. ANOVA revealed a significant interaction between time of day and sleep effects (P < 0.05). The diurnal and the sleep-related variations of plasma leptin mirrored those of body temperature and roughly paralleled those of plasma glucose and insulin; the amplitudes of the diurnal leptin variations were significantly correlated with the amplitudes of the diurnal body temperature variations (P < 0.05). Plasma leptin levels also displayed irregular pulses of low amplitude (mean duration, 70 min) that were not affected by sleep, but were associated with a significant decrease in glucose and insulin levels (P < 0.01). These results demonstrate that under continuous enteral nutrition, plasma leptin levels are modulated by both a slight circadian component and sleep, which interact under normal conditions, and suggest that leptin is implicated in circadian thermoregulatory adjustments.