Early experiments in young men exposed to insufficient sleep and reduced caloric intake found lower anorexigenic (leptin) and higher orexigenic (ghrelin) hormone concentrations in association with increased hunger and appetite [16], whereas reports of sleep-deprived volunteers who were given adequate or excess amounts of self-selected calories found either stimulatory or no independent effects of sleep loss on plasma leptin [17–21]. Additional experiments combining 2 weeks of sleep restriction with over- or underfeeding showed that sleep insufficiency did not affect the corresponding rise and fall in leptin, whereas ghrelin increased only in the presence of sleep loss and negative, but not positive, energy balance [12, 14]. These observations support the notion that the early reports of lower leptin and higher ghrelin concentrations reflected the ability of sleep loss to amplify the human neuroendocrine response to caloric restriction (Fig. 8.1), and that sleep-deprived humans have a more vigorous response to threats of negative energy balance [4]. In further agreement with this concept, St-Onge et al. exposed healthy men and women to 4 days with sleep opportunity of 4 vs. 9 h/night and caloric restriction (average daily deficit of ~400 kcal) [15]. When participants slept less during the 4-day period of caloric restriction, their ad lib energy intake on day 5 was ~300 kcal higher than that after the same caloric restriction with an extended sleep opportunity. If operational under long-term free-living conditions, this enhanced response to caloric restriction may undermine the success of dietary weight-loss therapy in individuals with insufficient sleep—a possibility which is consistent with data from early clinical trials and epidemiological observations [22–24].

Sleep-deprived humans also exhibit greater propensity to overeat when given unrestricted access to easily available calories. This increased food intake does not require alterations in circulating orexigenic and anorexigenic hormone concentrations [6, 14, 25] and has been attributed to multiple sleep-loss-related changes in the central mechanisms that regulate human eating behavior, such as altered neuronal responsivity to stimuli from the environment, digestive system, and peripheral metabolic networks; declines in dietary restraint; enhanced reward seeking and desire for calorie-dense foods; and increased motivation and food purchasing behavior [26–29]. Lack of sleep can also lead to excessive energy consumption as a result of longer exposure to environmental stimuli which promote overeating, as well as maladaptive changes in the circadian pattern and timing of daily food intake. Indeed, insufficient sleep has been associated with late night eating combined with irregular meal habits and more snacking between meals [14, 30–36].

In addition to energy, sleep also conserves carbohydrate. Higher respiratory quotient (RQ) measurements following sleep restriction [18] and repeated disruption of sleep [37] suggest that partial sleep loss is associated with use of a greater proportion of energy from carbohydrate. Sleep restriction also caused a shift in substrate utilization toward oxidation of relatively more carbohydrate in overweight and obese adults placed on a 2-week hypocaloric diet [12]. The modest decline in fasting blood glucose and improved insulin economy in this setting [38] resembled the human metabolic adaptation to reduced carbohydrate availability. These findings raise the possibility that increased use of carbohydrate in individuals with insufficient sleep may stimulate hunger and food intake at times of diminishing glucose availability at night and during the late postprandial period. Indeed, some studies suggest that higher RQ predicts future weight gain [31]. In addition, Chaput et al. observed that self-reported short sleepers have more relative hypoglycemia at the end of an oral glucose tolerance test, which also predicted future weight gain [39, 40].

The apparent adaptation in 24-h energy expenditure in response to sleep loss indicates that the additional metabolic cost of extended wakefulness can also be offset by declines in resting (basal metabolic rate) and non-resting (activity-related) energy expenditure. Limited by the reliability of subjective recall and differences in study design and population, cross-sectional analyses of sleep and physical activity have given inconsistent results showing either positive, negative, or no significant association between shortened sleep time and changes in physical activity. Few studies have directly tested the effects of sleep deprivation on the amount and intensity of daily activity outside of the limitations imposed by a room calorimeter. Roehrs et al. found a higher percentage of time spent in inactivity in laboratory settings after one night of total sleep deprivation [41]. Schmid et al. reported that overnight sleep restriction reduced the amount and intensity of free-living activity on the following day [17]. In contrast, Brondel et al. found that a night with restricted sleep was followed by a day with increased food intake and more movement [42], while Bosy-Westphal et al. and Calvin et al. did not find effects of sleep restriction and higher food intake on daily activity [18, 25]. Finally, St-Onge et al. studied healthy lean adults exposed to 5 nights with fixed time-in-bed (4.0 vs. 9.0 h/night) and inadvertent caloric restriction and reported lower percent time spent in very heavy and heavy physical activity and a trend for lower peak activity during the sleep-restricted condition [15]. Interpreting the results of these studies is challenging, since the impact of insufficient sleep on activity-related energy expenditure can differ as individuals adapt to recurrent exposure [43, 44] and brief interventions cannot capture the changes in physical activity of people who exercise only a few times a week. Similarly, the amount of daily activity can change considerably in response to positive or negative energy balance [9].

More recently, free-living activity counts and time spent in sedentary, light, moderate, and vigorous-intensity physical activities were measured by accelerometry in matching groups of participants with habitual sleep <6 vs. ≥6 h/night [45]. Compared to participants who slept ≥6 h/night, short sleepers had 27 % fewer daily activity counts, spent less time in moderate-plus-vigorous physical activity (−43 min/day), and were more sedentary (+69 min/day). To test whether insufficient sleep can be a causal factor for the reduced physical activity in short sleepers, 18 subjects completed 1 week of experimental sleep restriction in the laboratory (time-in-bed 5.5 h/night) and a matching period with 8.5-h nighttime sleep opportunity in randomized crossover fashion [46]. Participants received a carefully controlled weight-maintenance diet and those who exercised regularly were allowed to follow their usual exercise routines. Sleep restriction decreased daily activity by 31 % as participants spent 24 % less time engaged in moderate-plus-vigorous-intensity physical activity and became more sedentary. Importantly, most of the decrease in physical activity during the 5.5-h time-in-bed condition was seen in individuals with regular exercise habits (−39 % vs. −4 % decline in exercisers vs. non-exercisers): on average, they re-allocated 30 min of daily moderate-plus-vigorous-intensity activity to less intense light and sedentary behaviors when their sleep was curtailed. Estimates of energy balance in studies where habitual exercisers were exposed to 2 weeks of treatment with time-in-bed of 5.5 vs. 8.5 h/night suggest that insufficient sleep is accompanied by combined reduction in resting and activity-related energy expenditure of ~250 kcal/day—an amount equivalent to 60 min of moderate physical activity at 3.6 MET for the average study participant. The clinical significance of such reduced energy expenditure is readily apparent, since current guidelines recommend 1 h of daily moderate-intensity physical activity for the prevention of long-term weight gain.