Short Sleep and Circulating Adipokine Concentrations: Does the Fat Hit the Fire?

THE DISCOVERY OF LEPTIN, A METABOLIC HORMONE PRODUCED BY FAT CELLS, HELPED TRANSFORM THE LIMITED VIEW OF ADIPOSE TISSUE AS PASSIVE FUEL storage to its current status as an active endocrine organ involved in the control of food intake and energy metabolism.^1,2 In addition to leptin, human fat cells secrete a variety of bioactive proteins (known as adipokines), including tumor necrosis factor-α, interleukin-6, adiponectin, retinol binding protein-4 (RBP4), and visfatin (pre-B cell colony-enhancing factor), which have autocrine and paracrine effects in adipose tissue, and systemic hormonal activity in brain, liver, muscle, pancreas, blood vessels, and other metabolic targets.² The accumulation of too much fat in obesity is associated with increased infiltration of adipose tissue by inflammatory cells, enhanced production of pro-inflammatory and decreased production of anti-inflammatory adipokines, and elevated circulating markers of chronic inflammation.^1,2 There is an evolving understanding that this chronic pro-inflammatory state of the adipose tissue is related to ectopic deposition of fat, lipotoxicity and systemic insulin resistance, and contributes to the damaging effects of obesity in the pathogenesis of type 2 diabetes and cardiovascular disease. That understanding reminds one of the old idiom that “when the fat hits the fire, trouble breaks out.”

Self-reported short sleep (≤ 6 h/day) has been associated with increased incidence of diabetes, possibly related to adverse effects of chronic sleep curtailment on insulin secretion and action (i.e., development of pancreatic beta-cell insufficiency and systemic insulin resistance).³ Since experimental sleep restriction is accompanied by increased markers of inflammation,^4–6 the existence of a similar pro-inflammatory response in chronic short sleepers has been hypothesized to contribute to the link between insufficient sleep, insulin resistance, and incident diabetes.³ In this issue of SLEEP, Hayes and colleagues⁷ obtained cross-sectional data from 561 participants in the Cleveland Family Study, all of whom completed 1 night of laboratory polysomnography between 23:00 and 06:45 followed by fasting blood sampling at 07:00, to assess the possibility that reduced sleep duration may be associated with changes in leptin, visfatin, and RBP4 as potential modifiers of insulin sensitivity. These data extend the findings of Patel et al.⁸ on the association of self-reported and polysomnographically-measured sleep time with various markers of inflammation in the Cleveland Family Study. In analyses adjusted for age, gender, race, waist circumference, apnea hypopnea index, and diagnosis of hypertension and diabetes, each 1-hour reduction in total sleep time during overnight polysomnography was associated with a 6% increase in morning leptin and 14% increase in visfatin concentrations.

The observation that shorter overnight sleep is accompanied by higher leptin concentrations is particularly interesting. Influential early experiments found lower leptin, higher ghrelin, and increased hunger and appetite in young men exposed to sleep curtailment and caloric restriction (1500 kcal/day for the average 75 kg study participant) at the time of sampling.⁹ Lower leptin concentrations were also seen during a period of sleep restriction in young men whose caloric intake has been reduced by ˜30% the day before sampling (10 kcal/kg breakfast replaced by 1.2 kcal/kg bolus of intravenous glucose).¹⁰ Supported by observational data from the Wisconsin Sleep Cohort showing a positive association between leptin and self-reported (but not polysomnographic) sleep time, and a negative association between ghrelin and polysomnographic (but not self-reported) sleep time,¹¹ these data have given rise to the popular notion that insufficient sleep triggers key hormonal signals of “famine in the midst of plenty” to cause excessive food intake and weight gain. In reality, only recently have human volunteers been exposed to sleep restriction while truly in the midst of plenty (i.e., given access to adequate or excess amounts of self-selected calories). Consistent with the results of Hayes et al.,⁷ short-term sleep restriction in women resulted in increased leptin concentrations,^12–14 while studies in men found no independent effects of sleep on leptin.¹⁵ Experiments combining 2 weeks of sleep restriction with over- or underfeeding also showed that sleep loss did not affect the corresponding rise and fall in leptin concentrations, whereas ghrelin concentrations increased only in the presence of negative, but not positive, energy balance.^16,17 In short, the latest experimental data suggest that the early reports of lower leptin and higher ghrelin concentrations during experimental sleep restriction^9,10 did not reflect the presence of “famine in the midst of plenty,” but rather the ability of sleep loss to amplify the human neuroendocrine response to acute caloric restriction.³ If so, sleep deprived humans may defend their energy balance much more vigorously against interruptions in their nutrient supply.¹⁷

The difference between the results of Hayes et al.,⁷ who found an association between higher leptin and shorter polysomnographic sleep time, and the Wisconsin Sleep Cohort,¹¹ which did not detect a significant association, underscore the inherent limitations of such studies. Most vexing among them is the notable inconsistency between self-reported and polysomnographic sleep time, which gives rise to different conclusions depending on which measure of sleep exposure is used in analyses.^8,11 Since structured laboratory polysomnography can modify the usual sleep duration of individuals, while self-reports of habitual sleep time harbor significant systemic bias,¹⁸ it is difficult to know what the true association between adipokine concentrations and chronic sleep duration was in these studies. Ongoing development and use of minimally-disruptive portable devices for reliable assessment of human sleep under free-living conditions will be crucial for our improved ability to measure habitual sleep duration.

The high prevalence of diabetes, hypertension and obstructive sleep apnea in the Cleveland Family Study presents additional analytical challenges. The lack of control for differences in fasting blood glucose and the use of various glucose-lowering agents among diabetic participants leaves room for residual confounding (e.g., potential differences among users of oral insulin secretagogues, insulin sensitizers, exogenous insulin or other therapies were not analyzed). Similar considerations apply to the decision to make alternative adjustments for BMI or waist circumference,⁷ as opposed to controlling for both simultaneously,⁸ and the possibility that short sleepers may be taking a greater number of and higher-dose antihypertensive agents with different metabolic and adrenergic properties. Finally, the well-known increase in frequency and severity of respiratory events during REM sleep in many sleep apnea sufferers may have contributed to the association between morning adipokine concentrations and overnight sleep (particularly REM) time.

At the end of the day, the most important question remains: Metaphorically speaking, does the body fat of people with insufficient sleep “hit the fire” causing trouble to break out? Experiments in model organisms and humans indicate the existence of bi-directional relationships between sleep and fat metabolism.^3,17 Based on their data, Hayes et al.⁷ speculate that chronic sleep curtailment may cause dysregulation of adipokine release, systemic inflammation, and insulin resistance. Whether this is a plausible hypothesis is difficult to assess without additional analyses of the relationship of leptin and visfatin with measures of insulin sensitivity and markers of systemic inflammation in the Cleveland Family Study.⁸ In contrast to the poorly characterized effects of visfatin on insulin signaling in humans, adiponectin has important insulin sensitizing properties,² but no significant association with sleep duration was found in the Wisconsin Sleep Cohort.¹¹ Likewise, when target organs respond adequately to its action, higher plasma leptin can enhance lipid storage and metabolism without injury to non-adipose tissues.¹ Therefore, knowledge of the effects of chronic sleep curtailment on the responsiveness of various metabolic targets to key adipokine signals will be crucial for our progress in this field. Ultimately, most definitive answers about the role of adipose tissue dysfunction and inflammation in the human response to chronic sleep loss will come from prospective observations and studies combining behavioral and pharmacological interventions with analyses of fat biopsies, in vivo and in vitro imaging and metabolic testing, and well-defined clinical outcomes.

DISCLOSURE STATEMENT

Dr. Penev has received research support from Sepracor.