Getting on top of sleep

Abstract

There is no doubt that sleep is good for health. Yet, understanding how sleep exerts its healing effects has been a formidable challenge.

“A good laugh and a long sleep are the best cures in the doctor's book.” While slightly exaggerated, the Irish proverb has some wisdom as sleep is well known to have curative powers not only to help the body to fight an infection but also generally for remaining healthy and fit. No surprise then that science has been looking at how sleep heals or, vice versa, if and how lack of sleep contributes to disease and illness. Research has in fact shown that night shift work with unnatural sleep patterns, jet lag, or lifestyles that result in mistiming of sleep disrupt the human transcriptome through alteration of circadian rhythms (Archer et al, 2014), which mediates many of the negative health impacts of sustained sleep disruption. There is also a highly topical dimension for the field given widespread reports of sleep disruption resulting from fear and anxiety associated with COVID‐19 or fallout from lockdown measures, which may have detrimental effects on public health.

While the insight that sleep and health are intricately linked has been well established by science for decades, further research to unravel the underlying mechanisms and unpicking cause from effect had almost stalled. The problems are numerous, starting with defining poor or restricted sleep in the first place, to measuring it reliably and reproducibly on a large scale without relying on self‐reporting, to untangling sleep or the lack thereof from the many other factors that influence health and disease. So far, many of the more detailed findings that relate sleep patterns to various molecular or metabolic markers have been confined to animal models or laboratory settings, making them hard to generalize to the wider population. It will require large‐scale longitudinal studies in humans, following changing patterns of sleep over time, to obtain reliable data on the impacts of sustained sleep restriction.

… many of the more detailed findings that relate sleep patterns to various molecular or metabolic markers have been confined to animal models or laboratory settings, making them hard to generalize to the wider population.

Sleep and risk of Alzheimer's

Notwithstanding, there has been some progress lately in identifying detailed markers of poor sleep and connecting them to specific diseases or conditions, sometimes through a combination of animal and human studies. There has also been some progress overcoming the disconnect between laboratory and general population studies by complementing self‐reporting with more measurements from clinical‐grade instruments. Some of the most significant advances have been made over the link between sleep and Alzheimer's disease (AD). In fact, AD patients often suffer from impaired sleep patterns in addition to memory loss, but there is now growing evidence of a causal relationship between both.

A first step came with the finding that sleep mediates clearance of metabolites from the brain, including amyloid beta, the peptides that are the main components of the amyloid plaques in the brains of AD patients (Xie et al, 2013). This led to the suggestion that sleep's restorative function may at least partly result from enhanced removal of potentially neurotoxic waste products that accumulate in the central nervous system during waking time.

But this was hardly a full smoking gun, merely an association. However, the knot tightened when a more recent study identified a direct association between just a single night of deprived sleep and accumulation of Aβ in the brain (Shokri‐Kojori et al, 2018). It found that the Aβ increases were associated with worsening mood following sleep deprivation, without being related to any genetic risk for Alzheimer's disease. The authors also found that chronically disrupted sleep over a longer time period led to an inverse relationship between baseline Aβ levels and the number of hours slept at night by the study participants.

Even this did not establish a clear causative link mediated by a coherent mechanism. A further advance came last year when a study showed that sleep does affect phosphorylation of the tau protein leading to the aggregation process involved in several neurodegenerative diseases, including AD (Barthélemy et al, 2020). Tau, which occurs chiefly in neurons, is normally highly soluble but highly phosphorylated aggregates accumulate characteristically in these diseases. “Hyperphosphorylation of tau promotes the formation of neurofibrillary tangles in neurons and eventually cell death that causes the cognitive dysfunction in Alzheimer's disease (AD),” explained Brendan Lucey, author on that paper and Head of Sleep Medicine at the Washington University School of Medicine in St Louis, MO, USA. “In this paper, we found that sleep deprivation increases the tau phosphorylation rate at threonine‐217 (pT217). pT217 appears to increase very early in AD with amyloid plaque formation.” She added that there was still further work necessary to elucidate the underlying mechanistic link: “We don't yet fully understand how this works, but the findings are supported by recent work in mouse models that sleep‐wake activity affects protein phosphorylation in the brain.”

In fact, there is still uncertainty over whether poor sleep can actually cause AD and so truly be considered a risk factor. “It is still debatable whether poor sleep is causally linked to development of AD, but the association between poor sleep and AD may be bidirectional as many individuals with AD show sleep disturbances,” said Ehsan Kojori from the University of Texas at Dallas, USA. He cited a 2020 study that presented evidence for such a bidirectional relationship mediated by the accumulation of Aβ and tau (Wang & Holtzman, 2020). On the one hand, poor sleep seems to increase such accumulation, but this in turn leads to further sleep disruption.

A link between sleep and cancer?

The situation is similar for other conditions, given that sleep, or its disruption, is known to play some role in a wide range of diseases. This follows simply from the observation that sleep is ultimately regulated by the same circadian mechanisms as the immune system and associated with inflammatory processes that underly infectious, allergic and auto‐immune conditions, as well as cancer.

Such a link between cancer and sleep has long been assumed but it was hard to pin down. Yet, one of the largest meta studies to date that examined data from 65 studies involving 1,550,524 participants and 86,201 cancer cases has cast doubt on such a link (Chen et al, 2018). This study still raised more questions than it answered, finding that neither short nor long sleep duration increased overall cancer risk directly. However, short sleep duration was associated with cancer risk among Asians, while long sleep duration significantly increased the risk of colorectal cancer generally. The broad conclusion was that long‐term randomized controlled trials and well‐designed prospective studies are needed to establish causality and elucidate the mechanism underlying the association between sleep duration and cancer risk.

The impact of sleep on metabolism

The story is clearer for metabolic conditions as there is broad agreement that sleep plays a key role in maintenance of homeostasis. Earlier studies had examined the role of sleep in controlling insulin sensitivity, since its disruption is fundamental to many metabolic conditions. But these mostly focused on the well‐known role of insulin in glucose metabolism, such as a 2010 study reporting that experimentally induced sleep loss in healthy volunteers decreased insulin sensitivity without adequate compensation in beta‐cell function (Morselli et al, 2010). The result was impaired glucose tolerance and increased risk of type 2 diabetes.

Earlier studies had examined the role of sleep in controlling insulin sensitivity, since its disruption is fundamental to many metabolic conditions.

But the role of insulin goes beyond carbohydrate metabolism and affects lipids, proteins and minerals as well, since it coordinates overall use of body fuels. Disruption of insulin signaling can therefore result in other undesirable outcomes beyond type 2 diabetes, for example, obesity from disordered lipid metabolism.

This was the focus of a 2019 study which found that just four nights of sleep restriction suppressed the postprandial lipemic response and made people feel less satiated upon eating fatty meals, craving for more (Ness et al, 2019). This too is a nuanced story, according to Kelly Ness, sleep specialist at the University of Washington in Seattle, USA, and lead author. In one sense, by lowering production of triglycerides which contribute to cardiac disease for example, sleep deprivation appears superficially beneficial. “We did not find that sleep restriction has harmful effects on postprandial lipemia, in fact, we found that postprandial triglycerides were decreased in sleep restriction compared to baseline, whereas elevated postprandial lipemia is associated with cardiovascular disease,” she commented. “This finding seems, on the surface, to be an improvement in metabolism, but it was likely driven by increased insulin production because sleep restriction had impaired participants' insulin sensitivity. The story, therefore, is nuanced. In the short‐term, sleep loss very clearly impairs insulin sensitivity. The pancreas produces more insulin to compensate, in the context of a mixed meal, which causes faster clearance of meal lipids from the blood. What we do not know yet is what kind of adaptations the system may undergo in the long‐term to explain the association between chronic short sleep and cardiometabolic diseases.”

Making further progress depends to a large extent on improving calibration of poor sleep patterns outside laboratory conditions, which remains a challenge for the field, as Margeaux Gray from the Sleep, Health and Society Laboratory Biobehavioral Health, Pennsylvania State University, USA, observed. “Clinic‐bound research with highly controlled protocol factors, such as environmental exposures and dietary constraints, produces reliable results for those conditions and is necessary to demonstrate causality,” she commented. “Replication in an ecologically valid environment is a worthy challenge for this and other research because we cannot assume that under different circumstances the results would extrapolate without evaluating those alternate circumstances. For example, a major focus in our laboratory right now is adapting our sleep‐related data collection into the field so that we can evaluate our interventions more practically.”

Gray pointed out that some progress has been made to overcome limitations of self‐reported sleep. “One way is to use objective estimates from a combination of activity and biometric data that our ambulatory devices are capable of measuring,” she explained. “Another alternative is using ambulatory polysomnographic monitoring equipment that is clinical‐grade outside the laboratory or clinic, for example in the participant's home.” Polysomnographic (PSG) monitoring involves recording of various body functions and their changes during sleep. These include brain activity (EEG), eye movements (EOG), skeletal muscle activation (EMG), and heart rhythm (ECG).

Therapeutic interventions

Whatever mechanisms are used, converting associations into root causes remains a challenge even in controlled laboratory settings, according to Ness. “Proving the mechanisms underlying the associations between chronic inadequate sleep and disease is very complex, nuanced, and multifactorial,” she commented. Nonetheless, it is not always necessary to know whether sleep is a root cause in order to exploit it as a therapeutic intervention. Providing there is some bidirectional element, interventions to improve sleep could help reverse or at least delay an associated condition. A good starting point is simply to persuade or help “under slept” people to take more, according to Orfeu Buxton, who directs the Sleep, Health & Society Collaboratory at Pennsylvania State University Social Science Research Institute in the USA. “Then multiple aspects of sleep can be optimized, including regular duration, quality, timing, and regularity,” Buxton explained. “In those with insomnia, cognitive behavioral therapy for insomnia (CBTI) is preferred.”

Providing there is some bidirectional element, interventions to improve sleep could help reverse or at least delay an associated condition.

Buxton argued that such measures could help to counter the multiple comorbidities that emerge from sleep loss. “These are thought to synergize in a long‐term negative way for health and well‐being,” he said. “It is clear based on the epidemiological and experimental evidence that short sleep duration is causally related to chronic disease, per consensus panels and meta‐analyses.” The situation appears to be different for getting more sleep than is needed, which seemed not to have any causal link with conditions. “For example, depression is associated with both long sleeping and CVD (cardiovascular disease), but it's thought the best evidence is that depression causes them both, not long sleep causing CVD. Or, that CVD is already present, leading to fatigue and longer sleeping, that is the patient is already sick, so is sleeping longer to heal,” Buxton explained.

He added that there is evidence that therapeutic implications in the form of behavioral change can be effective for optimizing health and well‐being. “As with diet and exercise, there is no magic pill, it is a lifestyle of sleep health, good diet, and adequate physical activity that supports health, and interventions focused on those can be effective, and are not fads,” he said. “For example, dietary choices can negatively affect the quality or quantity of sleep, then being in a sleep‐deprived state predisposes you to eating more calorie‐dense food the next day. Over a lifetime, little choices and effects can add up to excess weight or increased susceptibility to cardiometabolic diseases.” But this also offer opportunities for interventions and turning a negative into a positive feedback loop through sleep, although this can be hard to achieve sustainably in practice. “Behavioral change is very difficult and we are only just beginning to understand how to do this optimally,” Buxton added.

Mental health

The role of sleep in mental health has come to the fore during the ongoing COVID‐19 pandemic given the clear evidence that fear and anxiety disrupt sleep. One of the larger studies so far examined associations between COVID‐related stress, sleep quality, and mental health among 2,541 community adults, 1,969 from Israel and 572 in the USA (Coiro et al, 2021). The authors report high levels of depression and anxiety symptoms with declining sleep quality, clearly linked to COVID‐related stressors. 89% of participants reported changes in sleep patterns following the onset of the pandemic. The study concludes that evidence‐based interventions were urgently needed to improve mental health, including possibly therapies to address sleep deficits. The authors also note that longitudinal research was needed to assess whether the observed high rates of distress and sleep difficulties represented just an immediate short‐term response to the pandemic that would attenuate over time, or a more chronic response requiring sustained intervention.

The role of sleep in mental health has come to the fore during the ongoing COVID‐19 pandemic given the clear evidence that fear and anxiety disrupt sleep.

One interesting finding that did emerge clearly was that while Israelis and Americans displayed similar changes in mental health and sleep, they were much more pronounced in the latter. This could reflect cultural differences in assessment but another possible explanation is that Israeli adults had become more accustomed to stress as a result of their country's recent history and had developed greater resilience. At any rate, this disparity might yield clues over how to develop resilience against future major sources of stress with sleep playing a critical role.

There is no doubt now that sleep does play a pivotal part as a mediator in multiple diseases, including mental illness, and can be a point of intervention to alleviate some of these. There is also no doubt that sustained sleep deprivation or disruption can be causal for a number of conditions, as has been revealed from those studies of nightshift workers for example. Yet, the clear mechanistic link between sleep and health remains largely elusive; and that may be partly because too little is known about sleep's restorative power.

EMBO reports (2021) 22: e52957.