Cross-sectional associations between reduced sleep times and increased weight have been appreciated for some time (3
). Although a recent study provided longitudinal data suggesting that reduced sleep is associated with an increased risk for future obesity, the small sample size and selection of subjects on the basis of psychopathology potentially limit the generalizability of those results (9
). Another study found that the increase in body mass index over 10 years was greater in short-duration sleepers, but the finding was not statistically significant (8
). In the present study, we have demonstrated, in a sample of middle-aged women up to 65 years of age, that a habitual sleep time of less than 7 hours predicts increased future weight gain independent of baseline weight.
Sleeping 5 hours or less led to about a 1-kg increase in weight over 16 years compared with sleeping 7–8 hours. Although this increase may appear small, even modest weight gain can have important health effects. Relative to maintaining a stable weight, a 5-kg weight gain nearly doubles the risk of developing diabetes (18
). Furthermore, usual sleep times of less than 7 hours are associated with a substantial increase in the risk of major (≥15 kg) weight gain and incident obesity.
How might sleep affect body weight? Hypothalamicpituitary functions, including those that influence eating, energy balance, and metabolism, are strongly tied to circadian rhythms and are highly integrated with sleep regulatory processes. Disruption of the circadian clock can have important metabolic effects. Knock out of the Clock
gene, an integral component of the circadian clock, in mouse models leads to obesity and metabolic dysfunction (19
). Neurons in the hypothalamus producing orexin are involved in the regulation of both feeding and wakefulness, and aberrations in orexin-mediated pathways may provide one link between sleep and obesity. Experiments in lean humans have found that short-term sleep restriction has important effects on cortisol, glucose tolerance, and growth hormone secretion (20
Ultimately, in order to cause weight gain, reduced sleep must increase caloric intake and/or reduce energy expenditure. Short-term sleep restriction lowers levels of the satiety-promoting hormone, leptin, increases levels of the appetite-promoting hormone, ghrelin, and increases subjective ratings on appetite and hunger (11
). An association among reduced sleep time, depressed leptin, and increased ghrelin has also been found in a large, cross-sectional study, suggesting that these effects may persist with long-term sleep restriction (22
). In our cohort, however, we did not find a relation between short-duration sleep and increased caloric intake. In fact, increased dietary consumption was associated with greater sleep durations.
Another possible explanation for the observed relation is that the fatigue produced by limiting sleep may reduce physical activity and thus energy expenditure, thereby predisposing to weight gain. However, we found little relation between sleep duration and reported physical activity. Although effects on voluntary activity cannot be conclusively excluded because of the limitations of questionnaire-based measurements of activity, this finding is consistent with that from the study by Hasler et al. (9
), which also found that changes in purposeful physical activity could not explain the sleep-weight association.
Other mechanisms influencing energy expenditure exist. The largest contributor to energy expenditure in a sedentary person is the basal metabolic rate. Recent work has suggested that sleep can affect the metabolic rate through effects on thermogenesis. Sleep restriction has been demonstrated in rodent models to influence the expression of uncoupling proteins, which regulate metabolic rate through thermogenesis (23
). Further work on how sleep can impact the metabolic rate in humans is clearly needed.
In addition, recent work suggests that involuntary movements such as posture (sitting vs. standing) and fidgeting can have important effects on weight regulation. Studies have found that increased involuntary activity in humans is associated with an ability to resist weight gain despite over-feeding and that lean individuals expend 350 kcal more per day on this type of activity than do the obese (25
). Interestingly, in rats, orexin appears to increase this type of energy expenditure (27
Other potential explanations for our findings include the possibility of a common cause of both sleep duration and weight gain. Although we attempted to control for potential confounders, we cannot exclude the possibility of residual confounding through variables not considered (such as corticosteroid use) or through incomplete adjustment. For example, the effect of depression on both sleep and weight could not be completely accounted for, as the first measurements of depression in this cohort occurred after the sleep question was asked. Finally, because sleep duration was assessed at only one time point, we cannot definitively exclude the possibility of reverse causation.
Several limitations to this work should be noted. First, all measurements were based on questionnaire response and so are liable to misclassification. However, the misclassification of sleep duration is likely nondifferential with respect to weight gain, so any resulting bias should be toward the null hypothesis, leading to underestimation of the true magnitude of association between sleep and weight. Usual sleep time is likely to have varied somewhat over the 16-year period of follow-up, but this also would likely bias the study toward a null result. Finally, it should be noted that this cohort was limited to middle-aged women in the nursing profession and primarily Caucasian. Further research will be needed before generalizing these findings to other populations, such as men, or other age groups.
Overall, our findings suggest that short-sleep duration is associated with a modest increase in future weight gain and incident obesity. These findings have the important implication that increasing sleep time among those sleeping less than 7 hours per night may represent a novel approach to obesity prevention. Given the inadequacies of current obesity therapy, the ability of public health initiatives focusing on sleep hygiene to slow the epidemic of weight gain should be investigated, as should the mechanisms linking sleep duration to obesity.