Role of sleep deprivation in immune-related disease risk and outcomes

Modern societies are experiencing an increasing trend of reduced sleep duration, with nocturnal sleeping time below the recommended ranges for health. Epidemiological and laboratory studies have demonstrated detrimental effects of sleep deprivation on health. Sleep exerts an immune-supportive function, promoting host defense against infection and inflammatory insults. Sleep deprivation has been associated with alterations of innate and adaptive immune parameters, leading to a chronic inflammatory state and an increased risk for infectious/inflammatory pathologies, including cardiometabolic, neoplastic, autoimmune and neurodegenerative diseases. Here, we review recent advancements on the immune responses to sleep deprivation as evidenced by experimental and epidemiological studies, the pathophysiology, and the role for the sleep deprivation-induced immune changes in increasing the risk for chronic diseases. Gaps in knowledge and methodological pitfalls still remain. Further understanding of the causal relationship between sleep deprivation and immune deregulation would help to identify individuals at risk for disease and to prevent adverse health outcomes.

decline is present not only in high-income and developed countries 18 but also in low-income or racial/ethnic minorities 19 , thus representing a worldwide problem.
In addition to fatigue, excessive daytime sleepiness, and impaired cognitive and safety-related performance, sleep deprivation is associated with an increased risk of adverse health outcomes and all-cause mortality [20][21][22][23][24] . Indeed, epidemiological and experimental data support the association of sleep deprivation with the risk of cardiovascular (CV) (hypertension and coronary artery disease) and metabolic (obesity, type 2 diabetes (T2DM)) diseases [24][25][26][27] . In the United States, sleep deprivation has been linked to 5 of the top 15 leading causes of death including cardio-and cerebrovascular diseases, accidents, T2DM, and hypertension 28 . Data also point to a role for sleep deprivation in the risk of stroke, cancer, and neurodegenerative diseases (NDDs) 26,29,30 . Sleep deprivation is also associated with psychopathological and psychiatric disorders, including negative mood and mood regulation, psychosis, anxiety, suicidal behavior, and the risk for depression [31][32][33][34][35][36] .
Both too short or too long sleep durations have been found to be associated with adverse health outcomes and all-cause mortality with an U-shaped relationship [37][38][39] . Although the relation of long sleep duration to adverse health outcomes may be confounded by poor health conditions occurring in older adults 37 , the causal association of sleep deprivation with negative health effects is substantiated by experimental evidence providing biological plausibility 24,40,41 .
Sleep profoundly affects endocrine, metabolic, and immune pathways, whose dysfunctions play a determinant role in the development and progression of chronic diseases [42][43][44] . Specifically, in many chronic diseases, a deregulated/exacerbated immune response shifts from repair/regulation towards unresolved inflammatory responses 45 .
Regular sleep is crucial for maintaining immune function integrity and favoring a homeostatic immune defense to microbial or inflammatory insults 46,47 . Sleep deprivation may result in deregulated immune responses with increased pro-inflammatory signaling, thus contributing to increase the risk for the onset and/ or worsening of infection, as well as inflammation-related chronic diseases.
Here we reviewed the evidence regarding the impact of sleep deprivation on immune-related diseases by discussing the major points as follows: (1) the immune-sleep relationship; (2) the association of sleep deprivation with the development and/or progression of immune-related chronic diseases; and (3) the immune consequences of sleep deprivation and their implications for diseases. Finally, possible measures to reverse sleep deprivation-associated immune changes were discussed.

Basic immune mechanisms of sleep regulation
The discovery of muramyl peptide, a bacterial cell wall component that is able to activate the immune system and induce the release of sleep-regulatory cytokines, primary regulators of the inflammatory system, provided the first molecular link between the immune system and sleep 48 . Thereafter, other microbialderived factors such as the endotoxin lipopolysaccharide (LPS) 49 , as well as mediators of inflammation, such as the cytokines interleukin (IL)-1 and tumor necrosis factor (TNF)-α, prostaglandins (PGs), growth hormone-releasing hormone (GHRH), and growth factors, were recognized as sleep-regulating factors 50 .
Along this line, most animal studies have consistently shown a role in particular for IL-1, TNF-α, and PGD 2 in the physiologic, homeostatic non-rapid eye movement (NREM) sleep regulation, so that the inhibition of their biological action resulted in decreased spontaneous NREM sleep, whereas their administration enhanced NREM sleep amount and intensity, and suppressed rapid eye movement (REM) sleep [51][52][53] . Moreover, the circulating levels of IL-1, IL-6, TNF-α, and PGD 2 are highest during sleep 54 . Their effects are dose-and time-of-day-dependent so that, for instance, low doses of IL-1 enhance NREMS, whereas high doses inhibit sleep 55 . Reciprocal effects may be involved in sleep regulation: for instance, the effects of systemic bacterial products such as LPS may also involve TNF-α 49 . Links exist between IL-1β and GHRH/growth hormone (GH) in promoting sleep so that IL-1 induced GH release via GHRH 56 , and hypothalamic γ-aminobutyric acid (GABA)ergic neurons (promoting sleep) are responsive to both GHRH and IL-1β 57 . Instead, antiinflammatory cytokines, including IL-4, IL-10, and IL-13, inhibited NREM sleep in animal models 58 .
Through these substances, the immune system may signal to the brain and interact with other factors involved in sleep regulation such as neurotransmitters (acetylcholine, dopamine, serotonin, norepinephrine, and histamine), neuropeptides (orexin), nucleosides (adenosine), the hormone melatonin, and the hypothalamus-pituitary axis (HPA) axis. Signaling mechanisms to the brain also involve vagal afferents: for instance, vagotomy attenuates intraperitoneal TNF-α-enhanced NREMS responses 59 .
Cytokines are produced by a vast array of immune cells, including those resident in the central nervous system (CNS), and non-immune cells, e.g., neurons, astrocytes and microglia, and peripheral tissue cells 60,61 . Cytokines interact with the brain through humoral, neural, and cellular pathways, and form a brain cytokine network ( Fig. 1) able to produce cytokines, their receptors, and amplify cytokine signals 50 . Peripheral cytokines reach the brain through different non-exclusive mechanisms, including blood-brain barrier (BBB) disruption 62 , penetration of peripheral immune cells, and via afferent nerve fibers, such as the vagus nerve, a bundle of parasympathetic sensory fibers that conveys information from peripheral organs to the CNS 63 .
In the CNS, cytokines mediate a multiplicity of immunological and nonimmunologic biological functions 64 , such as synaptic scaling, synapse formation and elimination, de novo neurogenesis, neuronal apoptosis, brain development, cortical neuron migration 65 , circuit homeostasis and plasticity 66 , and cortical neuron migration 65 , and complex behaviors, sleep, appetite, aging, learning and memory 65 , and mental health status 67,68 .
A common experimental finding is that after damage to any brain area, if the animal or human survives, sleep always ensues 69 . Recent evidence indicates that sleep is a self-organizing emergent neuronal/glial network property of any viable network regardless of size or location, whether in vivo or in vitro 53,[70][71][72][73] . Several sleep-regulatory substances, e.g., TNF, IL-1, nitric oxide, PGs, and adenosine are all produced within local cell circuits in response to cell use 74,75 .
From this point of view, TNF-α and IL-1 are closely interconnected and play a similar role in the regulation of sleep [76][77][78][79][80][81] . IL-1β and TNF-α self-amplify and increase each other's mRNA expression in the brain 82 . In rats, IL-1 83 and TNF-α 84 mRNAs show diurnal variations in different brain areas, with the highest concentrations recorded during increased sleep propensity and peaks occurring at time of sleep period onset in rats and mice 85 .
Sleep-like states in mixed cultures of neurons and glia are dependent in part on the IL-1 receptor accessory protein (AcP) 69,86 . In the brain, there is an AcP isoform, neuron-specific (AcPb) 87 , whose mRNA levels increase with sleep loss 88,89 . AcPb is anti-inflammatory, whereas AcP is pro-inflammatory 87,88 .
TNF signaling promotes sleep, whereas reverse TNF-α signaling (the soluble TNF receptor) promotes waking 90 . The brain production of TNF-α is neuron activity-dependent 91 . Afferent activity into the somatosensory cortex enhances TNF expression 92 , and in vitro optogenetic stimulation enhances neuronal expression of TNF immunoreactivity 93 .
Peripheral immune activation following acute or chronic infection or inflammatory diseases is marketed by altered cytokine concentrations and profiles, and is transmitted to the CNS initiating specific adaptive responses. Among these, a sleep response is induced and has been hypothesized to favor recovery from infection and inflammation, supposedly via the timely functional investment of energy into the energy-consuming immune processes 54,94 . Accordingly, acute mild immune activation enhances NREM sleep and suppresses REM sleep, whereas severe immune response with an upsurge of cytokine levels causes sleep disturbance with the suppression of both NREM and REM sleep 49,[95][96][97][98] . This sleep change correlates to the course of the host immune response as observed in bacterial and Trypanosoma infections 97,99 . Supportively, the increase in NREM sleep was a favorable prognostic factor for rabbits during infectious diseases 96 .
Immune regulators also mediate the complex interrelation between sleep and the circadian systems 74 . Circadian rhythms in behavior and physiology are generated by a molecular clockwork located in the suprachiasmatic nucleus, i.e., the master circadian pacemaker, and peripheral tissues, and involving the so-called clock genes (Clock, Bmals, Npas2, Crys, Pers, Rors, and Reverbs) 100 . Cytokines, including TNF-α, IL-1β 101,102 , and LPS [103][104][105] , suppress the peripheral and hypothalamic expression of core clock genes and clock-controlled genes, resulting in reduced locomotor activity accompanied by prolonged rest time 101 .
Sleep deprivation and immune-related disease outcomes In the following section, the association between sleep deprivation and risk or outcomes of immune-related disorders, as observed in human studies (mostly observational) and animal experimentations, will be examined. In this context, considering the sleep-immunity relationship, research has also begun to explore whether and how immune deregulation and inflammation may link sleep deprivation with adverse health outcomes.

Infection.
A breakdown of host defense against microorganisms has been found in sleep-deprived animals, as shown by the increased mortality after septic insult in sleep-deprived mice compared with control mice 106 , or by systemic invasion by opportunistic microorganisms leading to increased morbidity and lethal septicemia in sleep-deprived rats 107 . There is growing evidence associating longer periods of sleep with a substantial reduction in parasitism levels 108 and reduced sleep quality with increased risk of infection and poor infection outcome 109,110 . Accordingly, patients with sleep disorders exhibited a 1.23-fold greater risk of herpes zoster than did the comparison cohort 111 . Furthermore, sleep-deprived humans, as those with habitual short sleep (≤5 h) compared with 7-8 h sleep, are more vulnerable to respiratory infections in cross-sectional and prospective studies 112,113 , and after an experimental viral challenge 109,114 . Similarly, compared with long sleep duration (around 7 h), short sleep duration (around 6 h) is associated with an increased risk of common illnesses, including cold, flu, gastroenteritis, and other common infectious diseases, in adolescents 115 .
Compared with non-sleep-deprived mice, REM-sleep-deprived mice failed to control Plasmodium yoelii infection and, consequently, presented a lower survival rate 110 . This was correlated to an impaired T-cell effector activity, characterized by a reduced differentiation of T-helper cells (Th) into Th1 phenotype and following production of pro-inflammatory cytokines, such as interferon (IFN)-γ and TNF-α, and compromised differentiation into T-follicular helper cells (Tfh), essential to B-cell maturation, which therefore resulted to be reduced 110 . Accordingly, both Maf, a Tfh differentiation factor, and T-bet, a pro-Th1 transcription factor, were reduced in the REM-sleep-deprived group 110 . The combination of REM-sleep deprivation and P. yoelii infection resulted in an additive effect on glucocorticoid synthesis, and chemical inhibition of this exacerbated glucocorticoid synthesis reduced parasitemia, death rate, and restored CD4 T-cell, Tfh, and plasma B-cell differentiation in infected sleep-deprived mice 110 , suggesting a role of HPA axis hyperactivation in impairing host immune response under sleep deprivation.
Seep deprivation may exert detrimental effects on sepsisinduced multi-organ damage. Sleep deprivation (3 days) after LPS administration increased the levels of pro-inflammatory cytokines (IL-6 and TNF-α) in the plasma and organs (lung, liver, and kidney), which could be abrogated by subdiaphragmatic vagotomy or splenectomy 14 days prior to LPS administration 116 . Gut microbiota-vagus nerve axis and gut microbiota-spleen axis may play essential roles in post-septic sleep deprivation-induced aggravation of systemic inflammation and multi-organ injuries 116 .
Considering the association between sleep deprivation and immune response to infections, vaccination studies allow to assess the impact of sleep and sleep loss on ongoing immune response and the clinical outcome. Studies in which sleep deprivation (one or few nights) was applied to healthy humans during (mostly after) the immunological challenge of vaccination demonstrate that sleep deprivation reduced both the memory and effector phases of the immune response, as indexed by suppressed antigen-specific antibody and Th cell response compared with undisturbed sleep 117 .
Congruently, habitual (and hence chronic) short sleep duration (<6 h) compared with longer sleep duration was associated with reduced long-term clinical protection after vaccination against hepatitis B 118 . Sleep deprivation did not exert any impairing effect on mice already immunized 119 . From these studies, it seems that sleep supports-and sleep deprivation impedes-the formation of the immunological memory. Potential mechanisms involved in the beneficial effect of normal sleep on the vaccination response include: (i) the sleep-induced reduction in circulating immune cells that most likely accumulate into lymphatic tissues, increasing the probability to encounter antigens and trigger the immune response; (ii) the sleep-associated profile of inflammatory activation towards Th1 cytokines (increased IL-2, IFN-γ, etc.), which may favor macrophage activation, antigen presentation, and T-cell and B-cell activation; (iii) the effect of sleep stage on the formation of immunological memory through specific immune-active hormones: indeed, during slow wave sleep-rich early sleep, the profile of immune-active hormones, characterized by minimum concentrations of cortisol, endowed with antiinflammatory activity, and high levels of GH, prolactin, and aldosterone, which support Th1 cell-mediated immunity, may facilitate the mounting of an effective adaptive immune response to a microbial challenge 54 .
Cancer. Sleep deprivation has increasingly been recognized as a risk factor for impaired anti-tumor response. Epidemiological studies suggest, albeit not consistently 120 , a significant association between short sleep duration and the risk for several cancers, including breast, colorectal, and prostate cancer 29,121-123 . Potential mechanisms underlying this association include a shorter duration of nocturnal secretion of melatonin (putatively due to increased light exposure at night) 124 , which exerts anticancer properties through antimitotic, antioxidant, apoptotic, anti-estrogenic, and anti-angiogenic mechanisms 125 . Melatonin also plays immunomodulatory and anti-inflammatory effects with relevance for its anti-cancer activity, being able to inhibit the proinflammatory nuclear factor-κB (NF-κB)/NLRP3 inflammasome pathways, and to support T/B-cell activation and macrophage function 126 . However, besides melatonin, impaired anti-tumor immune response has been invoked in the sleep deprivationassociated risk for cancer development. A reduced cytotoxic activity of natural killer (NK) cells, which are immune cells with anti-tumor effect, has been reported in 72 h sleep-deprived mice compared with control mice, accompanied by reduced numbers of the cytotoxic cells such as CD8 T cells and NK cells in the tumor microenvironment after chronic sleep deprivation (for 18 h/day during 21 days) in an animal model of experimental pulmonary metastasis 127,128 . In this model, the reduced antitumor immunity of sleep-deprived animals was also indexed by the reduced number of antigen-presenting cells (dendritic cells) in the lymph nodes, as well as by the decreased effector CD4 T-cell numbers and corresponding cytokine profile (decreased IFN-γ), resulting in lowered Th1 response of Th cells, i.e., the most effective immune response against tumors. Therefore, an immunosuppressive environment develops with sleep deprivation, which could translate into an early onset and increased growth rate of cancer 128 or increased mortality 129 .
An integrated meta-analysis of transcriptomic data showed that circadian rhythm-related genes are downregulated and upregulated in the cortex and hypothalamus samples of mice with sleep deprivation, respectively, with downregulated genes associated with the immune system and upregulated genes associated with oxidative phosphorylation, cancer, and T2DM 130 . Several circadian rhythm-related genes were common to both T2DM and cancer, and seem to associate with malignant transformation and patient outcomes 130 .
Hence, although these sleep deprivation-induced immunemediated mechanisms in cancer warrant further confirmation in humans, the importance of the immune function in the antitumor host defense is well recognized 131 , thus suggesting that the impaired immune response after sleep deprivation may represent a plausible mediator of the associated increased risk for cancer as described in animal models and in humans.
Neurodegenerative diseases. NDDs are aging-related diseases that selectively target different neuron populations in the CNS, and include Alzheimer's disease, multiple sclerosis, Parkinson's disease, Huntington's disease, and amyotrophic lateral sclerosis. One prevailing hypothesis is that altered sleep habits and specifically sleep deprivation may be a consequence and frequently a marker of the disease [132][133][134] . However, human and animal studies have also suggested a causative or contributing role for sleep deprivation in the development and/or worsening of neurodegenerative processes [132][133][134] .
Potential pathophysiological mechanisms involve, among others, neuro-immune dysregulation. Indeed, a common feature -and a potential therapeutic target-of NDDs is the chronic activation of the immune system, where aspects of peripheral immunity and systemic inflammation integrate with the brain's immune compartment, leading to neuroinflammation and neuronal damage 135 . Neuroinflammation following sleep deprivation has been studied as a pathogenic mechanism potentially mediating the association between sleep deprivation and neurodegenerative processes. Low-grade neuroinflammation as indexed by heightened levels of pro-inflammatory mediators (e.g., TNF-α, IL-1β, and COX-2) and activation of astrocytes and microglia, main immune cells in the brain, was observed in the hippocampus and piriform cortex regions of the brain of chronic sleep-deprived rats along with neurobehavioral alterations (anxiety, learning, and memory impairments) 136 . The sleep deprivation pro-inflammatory milieu was accompanied by oxidative stress in the brain 137 and BBB disruption with consequent increased permeability to blood components 138 . After acute sleep deprivation, there was a significant increased recruitment of B cells in the mouse brain, which could be important given evidence of B cells involvement in NDDs 139 .
Progressive and chronic aggregations of unique proteins in the brain and spinal cord are hallmarks of NDDs 140 and trigger inflammatory responses, gradual loss of physiological functions of the nerve cells, and cell death 141 . Impaired autophagy in humans, a catabolic process of cytoplasmic components, contributes to the aggregation and accumulation of β-amyloid (Aβ), cytoskeletonrelated protein τ, and synuclein in neuronal cells and tissues 140 . Sleep plays an important role in the clearance of metabolic waste products accumulated during wakefulness and neural activity. Indeed, the Aβ protein is predominantly cleared from the brain during sleep, possibly through the glymphatic pathway. Congruently, acute and chronic experimental sleep deprivation in animals 142,143 and humans 144 resulted in brain Aβ accumulation and plaque formation, a typical pathological change in Alzheimer's disease process, the most common type of dementia. Imaging studies have revealed that healthy humans with selfreported short sleep were more prone to have cerebral Aβ plaque pathology 145 and disruption of deep sleep (slow wave sleep) increases Aβ in human cerebrospinal fluid (CSF) 146 . Likewise, patients with insomnia present higher CSF levels of Aβ 147 .
This pathological Aβ accumulation might reflect disrupted balance of Aβ production and clearance after sleep deprivation. On the one hand, sleep deprivation results in reduced clearance as suggested by clinical studies showing that Aβ levels in CSF are the highest before sleep and the lowest after wakening, whereas Aβ clearance from CSF was impaired by sleep deprivation 148 . Impaired clearance might also derive from disrupted peripheral Aβ transport, as suggested by the sleep deprivation-induced downregulation of low-density lipoprotein receptor-related protein-1 (LRP-1), which promotes Aβ efflux from the brain to the peripheral circulation across the BBB, and elevations of receptor of advanced glycation end products (RAGE), which promotes on the contrary the influx of peripheral Aβ into the brain, thus preventing Aβ clearance 149 . On the other hand, apart from impairing Aβ and τ interstitial fluid clearance, sleep deprivation may also have a role in increasing Aβ and τ exocytosis, thereby increasing CSF Aβ and τ levels 150 . In animals, sleep deprivation also leads to upregulation of β-secretase 1 (BACE-1), the most important enzyme regulating Aβ generation in the brain 142,143,149 , thus opening the hypothesis of increased Aβ production by sleep deprivation. Sleep deprivation-induced neuroinflammatory mediators correlate and could lead to disturbed Aβ clearance and stimulated amyloidogenic pathway 143 , being pro-inflammatory cytokines able to suppress the expression of LRP-1 and to increase RAGE 151 and BACE-1 levels 152 . Likewise, oxidative stress induced by sleep deprivation may also contribute to the neuroinflammatory burden and the increased expression of BACE-1 153 . Furthermore, patients with insomnia, compared with healthy controls, showed decreased serum levels of neurotrophins, including brain-derived neurotrophic factor (BDNF), proteins especially relevant in neuroplasticity, memory and sleep, and this reduction was significantly related to the insomnia severity 154 .
Sleep deprivation is associated with a rapid decline in circulatory melatonin levels, which may be linked to rapid consumption of melatonin as a first-line defense against the sleep deprivation-associated rise in oxidative stress 155 . Melatonin is a potent antioxidant, interacts with BDNF 156 , and promotes neurogenesis and inhibits apoptosis 157 . The neuroprotective potential of melatonin can target events leading to Alzheimer's disease development including Aβ pathology, τ hyperphosphorylation, oxidative stress, glutamate excitotoxicity, and calcium dyshomeostasis 150,158 . Accordingly, melatonin treatment could restore the autophagy flux, thereby preventing tauopathy and cognitive decline in Alzheimer's disease mice 159 .
Patients with Alzheimer's disease have an increased incidence of sleep-disordered breathing 160 . In addition, sleep-disordered breathing is associated with an increased risk of mild cognitive impairment or dementia and with earlier onset of Alzheimer's disease 161 . Sleep-disordered breathing is also associated with altered levels of Alzheimer's disease biomarkers in CSF, including decreased levels of Aβ and elevated levels of phosphorylated τ 162 . Sleep-disordered breathing possibly via hypoxia, inflammation, and sleep disruption/deprivation could contribute to Alzheimer's disease processes, e.g., increase of Aβ production and aggregation, suppression of glymphatic clearance of Alzheimer's disease pathogenic proteins (τ, Aβ) and oxidative stress, inflammation, and synaptic damage 134,163 .
To summarize, the sleep deprivation-associated risk for Alzheimer's disease could be linked to the induction of inflammation in the brain and disorders of systemic innate and adaptive immunity 164 . However, the relationship of sleep deprivation to inflammation in Alzheimer's disease is mostly speculative and needs to be confirmed.
Similar to Aβ in Alzheimer's disease, abnormal levels of αsynuclein are common to Parkinson's disease, the second most common NDDs 165 . Sleep disturbances are not only a common comorbidity in Parkinson's disease, but often precede the onset of classic motor symptoms 166 . The main pathological features of Parkinson's disease are the reduction of dopaminergic neurons in the extrapyramidal nigrostriatal body and the formation of Lewy bodies formed by the aggregation of α-synuclein and its oligomers surrounded by neurofilaments. Due to the degeneration of the dopaminergic neurons, affected people show muscle stiffness, resting tremors, and posture instability; other pathways involved in sleep, cognition, mental abnormalities, and other non-motor symptoms are also affected 167 . Epidemiological studies also suggest that disturbed sleep may increase the risk of Parkinson's disease 168,169 . Such disease-modifying mechanisms may include activation of inflammatory and immune pathways, abnormal proteostasis, changes in glymphatic clearance, and altered modulation of specific sleep neural circuits that may prime further propagation of α-synucleinopathy in the brain 169 . Melatonin could reduce neurotoxin-induced α-synuclein aggregation in mice. Furthermore, melatonin pretreatment reduced neurotoxin-induced loss of axon and dendritic length in dopaminergic neurons through suppression of autophagy activated by CDK5 and α-synuclein aggregation, thereby reducing dyskinesia symptoms in Parkinson's disease animal models 170 . A few reports have shown that melatonin exerts protective effects in several experimental models of Parkinson's disease 171 .
However, although animal experimentations suggest a link between sleep deprivation and immune dysfunction in neurodegenerative processes, no human investigations have yet confirmed the mediating role of immune dysregulation in the association between sleep deprivation and risk or outcomes of NDDs.
Autoimmune diseases. Sleep disturbances are frequently reported in autoimmune diseases, and immunotherapy in patients with autoimmune pathologies results in sleep improvement 172 . However, knowledge of the immunopathology of autoimmune diseases have disclosed new concepts on the impact of sleep deprivation on autoimmune disease process, showing that sleep deprivation can promote a breakdown of immunologic self-tolerance. Human cohort studies found that non-apnea sleep disorders, including insomnia, were associated with a higher risk of developing autoimmune diseases such as rheumatoid arthritis, ankylosing spondylitis, systemic lupus erythematosus, and systemic sclerosis (adjusted hazard ratio: 1.47, 95% confidence interval (CI) 1.41-1.53) 173 Similarly, in relatives of systemic lupus erythematosus patients, and hence at increased risk for systemic lupus erythematosus, self-reported short sleep duration (<7 h/night) was associated with transitioning to systemic lupus erythematosus (adjusted odds ratio: 2.0, 95% CI 1.1-4.2), independent of early preclinical features that may influence sleep duration such as prednisone use, depression, chronic fatigue, and vitamin D deficiency 174 . This role of sleep deprivation as a risk factor for autoimmune diseases is corroborated by animal studies. In mice genetically predisposed to develop systemic lupus erythematosus 175 , chronic sleep deprivation, applied at an age when animals were yet clinically healthy, caused an early onset of the disease, as indexed by the increased number of antinuclear antibodies, without affecting disease course or severity, according to data on proteinuria, a surrogate marker of autoimmune nephritis, and longevity. Several mechanisms have been postulated to explain the link between sleep deprivation and autoimmune disease risk. Sleep deprivation can accelerate disease development through mechanisms including sleep deprivationinduced increased production of several pro-inflammatory cytokines 44,54 , as better discussed below. Indeed, cytokines are synergistically involved in the pathogenesis of autoimmunity, such as IL-6, whose abnormal production results in polyclonal B-cell activation and the occurrence of autoimmune features 176 , and IL-17 and the related Th17-cell response 177 , which require IL-6 for activation 178 and can cause greater amounts of autoantibody production and immune complex formation, or can intensify chronic inflammation by promoting angiogenesis and recruiting of inflammatory cells at inflammation sites as well as cartilage and bone erosion 179 . Furthermore, experimentally sleep-deprived healthy humans showed impaired suppressive activity of CD4 regulatory T cells (Treg), which normally is highest during the night and lowest in the morning 180 . The suppressive function of Treg towards excessive immune response is an important homeostatic mechanism, whose impairment is implicated in autoimmune disease pathogenesis 181 . Hence, sleep deprivation may not be merely an early symptom or a consequence of an autoimmune disease, but may contribute directly to the pathogenesis increasing the susceptibility to develop an autoimmune disease. More studies are warranted in this field.
Metabolic and vascular diseases. Prospective epidemiological evidence associate sleep deprivation (commonly <7 h/night, often <5 h/night) with the incidence of fatal and non-fatal CV outcomes, with a 48% higher risk of coronary heart disease 25 , a 15% higher risk of stroke 182 , and a 12% increased risk of all-cause mortality 37 , which is mainly due to CV causes, according to some authors 183 . In a recent prospective cohort, a low-stable sleep pattern (<5 h sleep/night) during the 4-year follow-up had the highest risk of death and CV events 184 . Short sleep has also been associated with increased subclinical atherosclerotic burden, the dominant underlying cause of CV diseases 185 .
In addition, sleep deprivation increases the risk for obesity (about 55% higher risk) 39,186 , insulin resistance, T2DM (28% higher risk) 38 , and hypertension (21% higher risk) 187 , which are powerful and preventable risk factors for CV diseases. Notably, the risk for diabetes attributable to sleep deprivation is comparable to that of other established traditional cardiometabolic risk factors 188 , thus underscoring the clinical significance of targeting sleep deprivation in the prevention of cardiometabolic diseases. In contrast with normal nocturnal sleep and in particular NREM sleep characterized by a marked decrease in sympathetic activity, catecholamine plasma levels, and blood pressure, experimental sleep deprivation (acute or chronic) is accompanied by increased sympathetic outflow, with consequent higher blood pressure and heart rate, thus providing a pathogenic link between sleep deprivation and hypertension risk [189][190][191][192] .
Regarding the influence of sleep deprivation on metabolic pathways, studies support a plausible causal link between sleep deprivation and the risk of overweight and obesity, possibly mediated by the effect of sleep deprivation on circulating levels of hormones (leptin, ghrelin) controlling hunger, satiety and energy balance, besides other factors intervening during sleep deprivation, including physical inactivity and overfeeding 193 . Furthermore, human experimental evidence with chronic sleep deprivation protocol demonstrate that sleep deprivation may alter glucose metabolism 194 and insulin sensitivity 195 , thus increasing the risk for obesity and T2DM. The reduction in total body insulin sensitivity observed after sleep deprivation (4.5 h per night for 4 days) in healthy subjects was paralleled by impaired peripheral insulin sensitivity, as demonstrated in subcutaneous fat playing a pivotal role in energy metabolism 195 . Considering a more chronic sleep deprivation, reduced insulin sensitivity was reported in overweight adults after 14 days of experimental sleep deprivation (5.5 h per night) compared with 8.5 h per night of sleep 196 , and after habitual curtailment in sleep duration of 1.5 h (<6 h of sleep per night) in healthy young adults with a family history of T2DM 197 .
Although the mechanisms that underlie most associations between short sleep duration and adverse cardiometabolic outcomes are not fully understood, potential causative mechanisms involving immune-inflammatory activation have been postulated. It is indeed well established that the subclinical inflammatory status induced by sleep deprivation has pathogenic implications for metabolic and CV risk factors (glucose metabolism, diabetes, hypertension, atherogenic lipid profile, endothelial dysfunction, and coronary calcification) and outcomes (stroke and coronary heart disease) 24 . Accordingly, most of the markers of systemic and cellular inflammation (leukocyte counts and activation state, cytokines, acute-phase proteins, and adipose tissue-derived adipokines) found to be altered after sleep deprivation have been epidemiologically and pathogenically associated with insulin resistance, T2DM, and vascular complications 198 . In fact, inflammation is an early pathogenic process during the development of obesity and insulin resistance 199 . Many adipose tissue-released inflammatory factors with pro-atherogenic and pro-thrombotic actions have also been regarded as a molecular link between obesity and atherosclerotic CV diseases 200 . Furthermore, chronic inflammatory processes are firmly established as central to the development and clinical complications of CV diseases, form the initiation, promotion and progression of atherosclerotic lesions to plaque instability, and the precipitation of thrombosis, the main underlying cause of myocardial infarction or stroke. Most CV risk factors (adiposity, insulin resistance, T2DM, hypertension, and dyslipidemia) act by inducing or intensifying such underlying inflammatory processes that ultimately promote endothelial dysfunction, altered vascular reactivity, innate and adaptive immune system activation, leukocyte infiltration into the vessel wall, and thus atherogenesis 201 . Experimental sleep deprivation leads to endothelial dysfunction, an early marker of atherosclerosis, as indexed by impaired endothelial-dependent vasodilation or increased levels of endothelial adhesion molecules 191 .
Among the inflammatory markers, besides being a biomarker of future risk for CV diseases and a predictor of clinical response to statin therapy 202 , C-reactiove protein (CRP) has been shown to be involved in the immunologic process that triggers vascular remodeling and atherosclerotic plaque deposition 202 . CRP levels lack diurnal rhythm and its liver production is stimulated by cytokines including IL-6 and IL-17, which are upregulated by sleep deprivation 203 . As such, although limited evidence have found an elevation of circulating CRP following sleep deprivation 204 , CRP is a prototypical inflammatory factor with the potential to mark and-to some extent mediate-CV risk following sleep deprivation. Congruently, elevated and sustained plasma levels of CRP have been observed in healthy humans after prolonged sleep deprivation (5 or 10 nights), in concomitance with increased heart rate 190,203 , lymphocyte pro-inflammatory activation, and production of cytokines (e.g., IL-1, IL-6, and IL-17) 203 . Similarly, the increase in blood pressure and heart rate observed after acute total sleep deprivation (40 h) was accompanied and even preceded by impaired vasodilation and by increased levels of IL-6 and markers of endothelial dysfunction and activation, such as cellular adhesion molecules (E-selectin, ICAM-1, etc.) 191 . The sleep deprivation pro-atherogenic effect in animal model of sleep fragmentation is mediated, at least in part, by reduced hypothalamic release of hypocretin (i.e., orexin), a wake-inducing neuropeptide, which limits the production of leukocytes (monocytes and neutrophils) and atherosclerosis development, and has been inversely associated with the risk of myocardial infarction, heart failure, and obesity 205 . The activation of the sympathetic nervous system (SNS) may be another mechanism for the inflammatory link between sleep loss and atherosclerotic CV disease, because such activation increases the bone marrow release of progenitor cells, the production of innate immune cells (monocytes), and the levels of inflammatory cytokines, and triggers endothelial dysfunction, thereby leading to systemic and vascular inflammation and atherosclerosis 206,207 . Playing a key role in instigating inflammatory responses and promoting atherosclerosis 208 , the sleep deprivation-associated oxidative stress may also contribute to CV risk. It has also been hypothesized a role for melatonin suppression following sleep deprivation in the vascular impairment associated with sleep deprivation, given that melatonin inhibits oxidative stress and cytokine production by immune and vascular cells, and represses atherosclerotic lesion formation in vivo 209 .
Therefore, a significant and consistent association exists between sleep deprivation and cardiometabolic risk and clinical outcomes, with several plausible immune-mediated causative mechanisms explaining this association.
Immune mechanisms linking sleep deprivation and diseases As shown above, sleep deprivation has been found to alter inflammatory immune processes via multiple pathways, which could lead to increased susceptibility to chronic inflammatory diseases (Fig. 2). Most of the current knowledge on immune effects of sleep deprivation come from studies using controlled experimental sleep deprivation protocols, among which chronic partial sleep deprivation, lasting 2-15 days, is that mostly resembling the human condition of chronic insufficient sleep.
Some studies have observed that sleep deprivation, compared with regular nocturnal sleep, leads to increased circulating numbers of total leukocytes and specific cell subsets mainly neutrophils, monocytes, B cells, CD4 T cells, and decreased circulating numbers and cytotoxic activity of NK cells 203,[210][211][212][213] . Other studies, however, found contrasting results, including a decrease in CD4 T cells after sleep deprivation 213,214 , probably due to differences in sleep deprivation protocol, sampling methodologies, and other factors. Sleep deprivation has also shown to alter circadian rhythm of circulating leukocytes 215 , with higher levels during the night and at awakening and a flattened rhythm 210,212 . Additional findings are suggestive of immune deregulation by sleep deprivation, including a decreased neutrophils phagocytic activity 213 , altered lymphocytes adhesion molecule expression 216 , and reduced stimulated production of IL-2 and IL-12, which are important for adaptive immunity 211,217 .
Experimental sleep deprivation has been reported to affect systemic markers of inflammation, with studies showing increased circulating pro-inflammatory molecules (IL-1, IL-6, CRP, TNF-α, and MCP-1); this associated in some studies with a subsequent homeostatic increase in endogenous inhibitors, including IL-1 receptor antagonist and TNF receptors 203,[218][219][220] . In agreement with experimental sleep deprivation, population studies found a direct independent association between habitual short sleep duration (generally < 5 or 6 h) and elevated circulating pro-inflammatory markers, e.g., acute phase proteins (CRP and IL-6), cytokines (TNF-α, IFN-γ, IL-1, etc.), adhesion molecules, and leukocyte counts 183,[221][222][223][224][225] . Furthermore, a reduced NK cell activity 226 and a decline in naive T cells 227 , compatible with reduced immune competence, was reported in association with habitual short sleep. Shortening of leukocyte telomere length, a cellular senescence marker linked with inflammation, was also associated with shorter sleep duration 228,229 .
The reported elevation of systemic inflammation is clinically relevant, because it is suggested to specifically mediate the increased risk of mortality associated with short sleep 23,230,231 and, as observed, the risk for chronic disease development.
Regarding cellular markers of inflammation, some studies found that the ex-vivo LPS-stimulated production of TNFα 232,233 , IL-1β, and IL-6 203,232-234 by human monocytes increased during sleep deprivation but decreased during regular nocturnal sleep 54,203,[232][233][234] . However, other studies reported a decrease of TNF-α production by activated monocytes after sleep deprivation compared with regular nocturnal sleep 203,235 . These contrasting results need further investigations and may depend on differences in the cytokine sensitivity to different sleep deprivation protocols or sampling methods and time. For instance, it seems that partial acute sleep deprivation increased stimulated monocytic TNF-α production 232,233 , whereas more sustained sleep deprivation decreased it 203,235 .
Undisturbed sleep is predominantly characterized by a Th1 polarization of Th cells (expressing IFN-γ, IL-2, and TNF-α), and experimental sleep deprivation in humans leads to a shift from a Th1 pattern towards a Th2 pattern (expressing IL-4, IL-5, IL-10, and IL-13) 217,236 . Accordingly, conditions featured by disturbed sleep with specific deficit in slow wave sleep, as observed in elderly people 237 , alcoholic 238 , and insomnia 239 patients, show a cytokine shift towards Th2. The balance of Th1/Th2 immunity and its shift during sleep deprivation may have crucial implications in anti-microbial and anti-tumor immune responses. Th2 over-activity is known to be involved in some forms of allergic responses, and to increase the susceptibility to infection 240 . Likewise, regarding the anti-tumor immune action, Th1 response supports cytotoxic lymphocytes and tumor cells destruction with the potential of elimination or control of tumor cell growth, so that a type 1 adaptive immune response (increased antigen presentation, IFN-γ signaling, and T-cell receptor signaling) may be associated with an improved survival or prognosis 241,242 . In contrast, Th2 over-response is thought to contribute to tumor development and progression, by limiting cytotoxic T lymphocytes proliferation and by the modulation of other inflammatory cell types 241 .
Several cellular and molecular signaling pathways may be involved in mediating the influence of sleep deprivation on immune and inflammatory functions (Fig. 3). Increased oxidative stress markers and/or decreased antioxidant defense have been found after sleep deprivation [243][244][245] . Sleep shows an antioxidant function, responsible for eliminating reactive oxygen species produced during wakefulness, and contrarily sleep deprivation may cause oxidative stress, which leads to cell senescence, unbalanced local/systemic inflammation, dysmetabolism, and immune derangements 246,247 .
Effects of sleep deprivation on the immune response may derive from the activation of the SNS with the corresponding increase in systemic catecholamines 22,248 . Catecholamines signal to immune cells via adrenergic receptors, which are primarily αand β-adrenergic in myeloid cells and β-adrenergic in lymphocytes 249 . The immune outcome of the sympathetic signaling is complex, and includes both stimulatory and inhibitory effects depending on cell and receptor types, cell development/ activation states, and local microenvironment 249,250 . Some evidence suggest that β-adrenergic signaling inhibits and αadrenergic signaling promotes excessive inflammation under endotoxemia 250 . Activation of α-adrenergic signaling in peripheral tissues induces the upregulation of pro-inflammatory cytokines 250,251 . Sympathetic activation also suppresses the transcription of type I IFNs (IFN-α and IFN-β) genes and interferon response genes, which play a key role in anti-viral immunity 252 , and inhibits via β-adrenergic signaling the antitumor cytotoxicity of T lymphocytes 253 . In vitro β-adrenergic stimulation repressed Th1 response and stimulated Th2 response, with varying effects found in vivo 249,254 . Although the specific role of SNS activation in the immune phenotype associated with sleep deprivation is not clearly established, data suggest a proinflammatory effect of SNS under sleep deprivation. Indeed, chemical sympathectomy has been recently shown to alleviate the inflammatory response following chronic sleep deprivation in mice 255 , and both αand, to a lesser extent, β-adrenergic receptors seem to contribute to the sympathetic regulation of inflammatory responses to sleep deprivation 256 .
Sleep loss and mistimed sleep also led in the blood transcriptome to alteration and reduction in the circadian rhythmicity of gene expression 261,263 , which is an integral part of basic biological processes and homeostasis [264][265][266] .
The activation of the pro-inflammatory NF-κB/Rel family of transcription factors by sleep deprivation, first demonstrated in the late 1990s in mice 267 , and subsequently widely confirmed 233,260,261,[268][269][270][271][272] , is one of the most consistent findings regarding upstream transcriptional regulation. NF-κB induces the expression of genes (e.g., cytokines/chemokines, growth factors, receptors/transporters, enzymes, adhesion molecules) involved in inflammation, immunity, proliferation, and apoptosis 273 , circadian clock activity 274 , and sleep propensity 275 . Potential signals for NF-κB activation under sleep deprivation include increased adenosine levels, oxidative stress, altered metabolism (adiposity and decreased insulin sensitivity), brain proteins/ metabolites (e.g., Aβ), melatonin suppression 276 , circadian clock proteins 277 , and catecholamine surge due to increased sympathetic activity 278 . Given the role of NF-κB in the pathophysiology of inflammatory diseases 273 , its activation under sleep deprivation may be a common pathway for the risk of morbidity and mortality.
The intestinal microbiota is also affected by sleep loss [279][280][281] , showing indices of dysbiosis (increased Firmicutes:Bacteroidetes ratio; decreased diversity and richness), which may affect the immune system 282 , and are similar to those associated with cardiometabolic diseases 45 . Countermeasures for sleep deprivation: effect on immune parameters Although the impact of strategies to improve sleep duration on neurobehavioral performance and alertness after sleep deprivation have been assessed [283][284][285] , sleep deprivation countermeasures to improve immune and inflammatory parameters, and, correspondingly, disease risk and outcomes have been studied to a lesser extent.
Although extension of habitual short sleep did not show to significantly counterbalance the immune consequence of sleep deprivation [286][287][288] , mixed results derive from nighttime recovery sleep following sleep deprivation (Table 1), with limited evidence of effectiveness for specific immune parameters 210,214 , and mostly after multiple consecutive nights of 8 h sleep recovery or with an extended nocturnal sleep duration 212,289 .
Although daytime napping (<20 min) restores alertness, and mental and physical performance without provoking sleep inertia associated with longer nap [290][291][292] , the effects of a short nap on immune/inflammatory parameters after sleep deprivation have yet to be firmly established. Differently form population studies 293 , laboratory studies found immune benefit from nap 218,289,294,295 . Regarding immune-related clinical outcomes, controversy exists, with studies finding no association 296 , inverse associations 297,298 or positive association 296 , and a J-shaped relationship [299][300][301] between napping and CV and metabolic diseases or cancer events and mortality. Whether changes in immune parameters could contribute to the associations between napping and immune-related diseases remains unclear.
Among the strategies to recover sleep deprivation-induced immune changes, cognitive behavior therapy improves sleep outcomes in insomnia and lowers cellular and systemic inflammatory markers 302,303 , and the risk score composed of CV and metabolic risk factors 304 . This highlights the potential role of targeting sleep in reducing the inflammatory risk and the associated chronic diseases.

Summary and concluding remarks
Sleep exerts immune-supportive functions and impairments of the immune-inflammatory system are a plausible mechanism mediating the negative health effects of sleep deprivation, and in particular, its role in the risk and outcomes of chronic diseases such as infections, CV, metabolic and autoimmune diseases, NDDs, and cancer. Caution should be exercised in interpreting cellular and molecular outcomes of sleep deprivation in experimental studies conducted till now as a result of an independent effect of sleep deprivation, because other factors may play a role, including extended wakefulness-associated processes, other features of sleep-wakefulness, their temporal and functional segregation or methodologies of sleep manipulation.
Randomized controlled trials assessing the effect of treatment of sleep deprivation on inflammatory immune dysfunction and/ or health outcomes are needed. Knowledge of inflammatory and immunological signatures in response to sleep curtailment may inform not only on the underlying molecular links, but also contribute to refine risk profiles to be used for developing biomarkers of sleep deprivation and sleep disturbance-related health Fig. 3 Pro-inflammatory molecular pathways induced by sleep deprivation. A schematic model of potential mechanistic pathways linking sleep deprivation and inflammatory immune activation is depicted. Sleep deprivation is associated with activation of the sympathetic nervous system and release of norepinephrine and epinephrine into the systemic circulation, as well as to some extent with impaired hypothalamus-pituitary axis stimulation. These neuromediators may act along with other potential stimuli accumulated following sleep deprivation including reactive oxygen species (ROS), adenosine, metabolic waste products (e.g., β-amyloid) not cleared during normal sleep, gut microbiota dysbiosis leading to altered local and systemic pattern of metabolic products, as well as with changes in the profile of neuro-endocrine hormones, such as prolactin, growth hormone, and altered circadian rhythm of melatonin secretion. In immune cells located in the brain and the peripheral tissues, these stimuli may in concert trigger inflammatory activation, with release of cytokines, chemokines, acute phase protein, etc. via the recruitment of transcriptional regulators of pro-inflammatory gene expression, mainly nuclear factor (NF)-κB, and disturbing the circadian rhythmicity of gene expression of both clock genes and metabolic, immune and stress response genes (see text for further detail). E: epinephrine; NE: norepinephrine; TLR: Toll-like receptor. Arrows indicate stimulation; lines indicate inhibition. The illustrations were modified from Servier Medical Art (http://smart.servier.com/), licensed under a Creative Common Attribution 3.0 Generic License. outcomes, which may also represent potential targets of interventions. Recent metabolomic 305 and transcriptomic 306 studies hold promise in biomarker discovery 306 . These efforts may converge towards a new ground fostering interactions between the sleep research and the medical community to translate scientific knowledge into the clinic, prioritize health issues, and develop strategies and policies for subject risk stratification, to include evidence-based sleep recommendations in guidelines for optimal health and to address sleep hygiene at the individual and the population levels, as a means to prevent the negative health consequences of sleep deprivation. These actions might also foster health literacy and empowerment of individuals to actively better manage their own health and well-being throughout their life course by means of lifestyle, nutritional, and behavioral habits including sleep hygiene 307 .
Conclusively, in the perspective of staying healthy in this rapidly changing society, the sleep-immunity relationship raises relevant clinical implications for promoting sleep health and, as evidenced here, for improving or therapeutically controlling inflammatory response by targeting sleep. This may ultimately translate, in the era of preventive medicine, into addressing sleep as a lifestyle approach along with diet and physical activity to benefit overall public health.
Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability
All data generated or analysed during this study are included in this published article.