Late Night sleep loss

      Nutrients. 2023 Apr 23;15(9):2035. doi: 10.3390/nu15092035

Late, but Not Early, Night Sleep Loss Compromises Neuroendocrine Appetite Regulation and the Desire for Food

Svenja Meyhöfer 1,2,3,4,†, Rodrigo Chamorro 1,5,†, Manfred Hallschmid 2,6,7, Denisa Spyra 1, Nelli Klinsmann 1, Bernd Schultes 1, Hendrik Lehnert 4,8, Sebastian M Meyhöfer 1,2,4,*, Britta Wilms 1,2,4

Editors: Albert Lecube, David J Mela

    Discussion

In this balanced crossover experiment, we examined the effects of the timing of sleep loss on the regulation of hunger and appetite in humans.

 ‘Late-night sleep loss’, but not ‘early-night sleep loss’, elevated ghrelin concentrations, as well as feelings of hunger and appetite, 

and desire for food during the subsequent morning was observed, whereas leptin concentrations were not affected by acute sleep loss per se nor timing of sleep loss. Our data show a more pronounced effect of acute sleep loss on ghrelin, hunger, appetite, and the desire to eat when sleep is restricted in the second half of the night compared to a restriction of sleep during the first half of the night.

 This underscores the chronobiological role of timed sleep loss in human appetite regulation.

Studies on the effects of sleep restriction (with different timing) on plasma ghrelin and leptin levels have yielded inconsistent results. Our group showed increased ghrelin levels after one night of partial sleep restriction (sleep from 22:30 to 03:30 h) compared to a night with 7 h of sleep [35]. Another study showed that two nights of sleep restriction (sleep from 01:00 to 05:00 h), as compared to two nights with 10 h recovery sleep, resulted in lower leptin and higher ghrelin levels in healthy men, and these changes were associated with increased feelings of hunger and appetite during the subsequent day [20]. 

Furthermore, four nights of sleep restriction (from 01:00 to 05:30 h) increased ghrelin levels with no changes in leptin, and the increase in ghrelin levels correlated with ingested calories from sweet foods [17]. 

Overall, most, but not all [18] data, seem to be aligned with our study, which shows elevated ghrelin levels when sleep loss occurs in the late part of the night. With regard to the measurement of ghrelin levels within the different studies, one needs to point out that some studies reported on total ghrelin, whereas others reported on acylated and deacylated ghrelin, separately, which would have been more informative. In some studies, changes in ghrelin were also correlated with increased energy intake [17,20], 

supporting altered homeostatic control of appetite after sleep loss during the second half of the night. Broussard et al. reported on increased ghrelin levels and caloric intake by more than 300 kcal after four nights of 4.5 h sleep compared to four nights of 8.5 h sleep in lean men. Furthermore, elevated ghrelin levels in the evening were related to higher energy intake, mainly due to carbohydrates (12). Besides ghrelin and leptin, measurements of orexigenic and anorexigenic adipokines, such as Peptide YY (PYY) or Glucagon-like peptide 1 (GLP1), would be relevant to comprehensively assess the effects of timed sleep loss on neuroendocrine appetite regulation.

We showed elevated feelings of hunger and appetite, as well as desire for food, when sleep was restricted, again, in the late part of the night. McNeil et al. evaluated the effects of shortened sleep (~4 h) using an advanced wake time (similar to our ‘late sleep loss’ condition) vs. a delayed bedtime condition (similar to our ‘early sleep loss’ condition) in 18 healthy young men and women [36]. They reported higher fasting appetite and post-breakfast ratings after short sleep with the advanced wake time. Additionally, increased explicit wanting and liking for high-fat relative to low-fat foods after short sleep was observed with advanced wake time vs. the control sleep condition [30]. Another study reported increased energy intake and pre-prandial hunger levels upon one night of 4 h sleep restriction during the second half of the night (from 02:00–06:00 h), as compared to 8 h of regular sleep [14].

Interestingly, a 3.5 h sleep condition using an early night sleep loss protocol (wake until 03:30 h) for three days in healthy men increased hunger and prospective food consumption, as reported by Hibi and co-workers [25]. These results contrast with previous findings from our group, showing that two nights of sleep loss during the first half of the night (wake until 02:45 h) did not affect feelings of hunger and appetite and energy intake in healthy humans [15]. In line with this, our present data also disagree with the study by Hibi et al., as early night sleep loss did not affect ghrelin or hunger feelings compared with late night sleep loss or a control sleep condition. This discrepancy could be related—at least partly—to the different study duration, one and two nights, respectively, in our studies vs. three nights in the study by Hibi and co-workers. It is also worth mentioning that it has been proposed that—in Westernized settings—the assessment of hunger and appetite feelings would reflect the hedonic component, rather than the homeostatic regulation of food intake [36]. Our data showed increased hunger and appetite feelings in the morning (mainly after late night sleep loss) could relate to exacerbated hedonic component of appetite control in the morning after both sleep loss conditions. However, the increased morning ghrelin levels could reflect the adverse effect of a late-night sleep loss for the homeostatic component of energy intake regulation.

Other have also shown increased explicit wanting and liking for high-fat relative to low-fat foods after short sleep with advanced wake time vs. a control sleep condition [30]. Liking and wanting for foods are two separate processes modulating food reward [37]. Whereas liking mirrors the sensory pleasure experience, wanting reflects the motivation related to appetite and, putting forward, the actual desire for food reward. Cerebral structures are differently involved in appetite control, as shown by functional imaging studies. Wanting-related processes are related to ventral pallidum and striatum, whereas liking specific regions are the orbitofrontal cortex, insular cortex, and amygdala, which hints at a neural dissociation of liking and wanting for foods and food-related cues [38]. When addressing the wanting/liking process in the context of sleep restriction, it is relevant to consider that time of day impacts wanting, but not liking, for food. We could show that, both under free living conditions and laboratory settings, liking for energy-dense foods was not different in the morning and evening, whereas wanting was clearly increased in the evening, albeit feelings of hunger and satiety remained unchanged [39]. In our present study, however, liking and wanting were not addressed, which would have allowed for a closer insight into the neuroendocrine appetite regulation, especially in the context of timed sleep loss.

Chronic circadian disruption not on only occurs during insufficient and short sleep, but also during rotating shift work or (social) jetlag and, thus, affects neuroendocrine appetite regulation. Social jetlag is a common form of circadian rhythm disruption and presents in up to 70% of the population [9]. It is defined as the misalignment between individual’s endogenous circadian clock and the actual sleep time [40]. One cross-sectional observational study reported elevated appetite for caloric-dense food in subjects with social jetlag, independent of sleep deprivation compared with subjects with sleep deprivation alone [41]. In addition, ghrelin levels were elevated in those with social jet lag. These data point to a social jetlag-related increase in incentive value of food, together with anticipated pleasure of ingesting these foods. Upon a fMRI study, the same workgroup reported that social jetlag is associated with altered resting-state activity in brain regions related to hedonic feeding, i.e., the reward system. In detail, an increased activity in the putamen, part of the striatum with functions linked to the reward system, was shown in subjects with social jetlag. Importantly, this effect is independent of short sleep duration. Further, subjects with social jetlag reported a higher perceived appetite for high caloric food and had overall detrimental eating habits than those without social jetlag [42]. Thus, data support evidence that subjects with social jetlag are prone to hedonic feeding and, in the long term, to body mass gain.

Under an acute sleep restriction setting, experimental studies report on increased caloric intake due to night-time calories. On the other hand, it has been shown, both in mice and humans, that feeding either at normal eating time while performing a rotating shift work paradigm [43], or having caloric intake matched for the caloric needs while under chronic sleep restriction [44], did not alter circadian rhythms. In the latter study, subjects were exposed to a 32-day inpatient protocol, including 24 cycles of a 20 h forced desynchrony protocol either with or with sleep restriction. There was no impact of short sleep in subjective feelings of hunger that was detectable because hunger rhythm remained stable with a peak in the evening and low levels in the early morning, as already shown by Scheer et al. [45]. Chronic circadian disruption, however, was associated with decreased subjective feelings of hunger, independent of the sleep duration. This study provides deeper insight into the different contributions of chronic circadian disruption per se and short sleep and suggested a less strong impact of short sleep on subjective hunger than reported by some studies [20,35], but not all [15]. Further studies are needed here to evaluate the distinct effects of timing of sleep loss in the context of chronic circadian disruption, since one may speculate that the chronobiological aspects of short sleep differently—at least in part—affects feeling of hunger, as shown in our present study.

We assessed appetite regulation and desire for food after only one night of sleep restriction by four hours compared to control sleep of 8 h. Therefore, we cannot draw any conclusions concerning the effects of longer-lasting sleep restriction or, on the other hand, shifted sleep, as well as social jetlag. Further studies are needed to evaluate the prolonged effects of sleep restriction with altered sleep timing, e.g., using a forced desynchrony protocol [44]. However, the possibility to perform such studies in humans under experimental settings is limited compared to conducting an animal study using a shift work paradigm (e.g.,). As a further limitation, our study population included young, healthy, and normal-weight men. Although we do not expect gender differences, we cannot extrapolate the observed effects on women. In fact, current evidence regarding the metabolic consequences of sleep restriction has been provided from studies conducted mostly in men. However, studies comprising young healthy adults, both men and women, have reported consistent increases in hunger, appetite, energy intake, and susceptibility to food stimuli [46], together with reduced insulin sensitivity, in young adults [19]. Importantly, those studies have used different duration of sleep loss intervention. In females, increased morning leptin levels [22,47], but no changes in hunger scores, have been shown after an acute partial sleep restriction (3 h of sleep) [22]. With longer duration of intervention (four nights), increased energy intake and weight gain have been reported after partial sleep restriction in healthy women [13]. Others, comparing the effects of sleep loss in men and women, have shown increased ghrelin levels after four days of sleep restricted to 4-h only in men [48]. Additionally, after five days of 4 h of sleep (sleep from 04:00 to 08:00 h), men, but not women, showed exacerbated overall and late-night hour energy intake [49]. These last studies, however, are not comparable to our current study, as we assessed a very acute (one night) sleep loss intervention. It is further important to point out that, despite the small number of participants, a parametric test procedure was used. To address sphericity, we used the Green House Geisser approach. Moreover, our results should be confirmed in subjects prone to metabolic diseases, such patients as with obesity and T2D. As discussed above, using more comprehensive methods to assess the regulation of food intake, such as the wanting/liking task, or fMRI, would provide deeper insight into the impact of short sleep and its timing