Sleep Architecture, Circadian Biology, and Longevity: Glymphatic Clearance, Sleep Apnea, and the DPN-Insomnia Spiral

Sleep medicine is undergoing one of the most dramatic paradigm shifts in all of neuroscience. For the first time, we have a mechanistic explanation for why humans — and essentially all multicellular organisms — have evolved to spend one-third of their lives in a state of apparent vulnerability: the glymphatic system, a cerebral waste-clearance network first described in mice by Maiken Nedergaard’s team (Science 2013) and confirmed in humans by 2015 (Nature Neuroscience), operates almost exclusively during slow-wave sleep, removing metabolic waste products — including Alzheimer’s-associated amyloid-β and phosphorylated tau — from the brain’s interstitial space at rates impossible during wakefulness.

This discovery reframes the entire conversation I have with patients about sleep. When a patient with T2DM and peripheral neuropathy tells me they sleep 5-6 hours because they’re busy, or that their neuropathic pain disrupts their sleep, I explain that they are not simply tired the next day — they are failing to clear neurotoxic proteins from their brain every night they sleep poorly, while simultaneously worsening every aspect of their metabolic health that drives their neuropathy forward. Sleep is not passive recovery. It is active biological maintenance that nothing else in medicine can replace.

Sleep Architecture: The Four Stages and Why Each Matters for Longevity

Adult sleep architecture cycles through four stages in approximately 90-minute ultradian cycles, with 4-6 cycles per night. N1 (light NREM, ~5% of sleep time): the transition from wakefulness; relatively unrestorative and easily disrupted. N2 (intermediate NREM, ~50% of sleep time): characterized by sleep spindles (12-15 Hz bursts) and K-complexes; spindles facilitate memory consolidation by coordinating hippocampal-cortical dialogue. N3 (slow-wave sleep / SWS / deep NREM, ~20-25% in young adults, declining to 5-15% by age 60): characterized by delta waves (<4 Hz); this is the primary stage for glymphatic clearance, growth hormone secretion, cellular repair, and immune memory consolidation. REM sleep (~25%): characterized by rapid eye movements, near-paralysis of skeletal muscle (except diaphragm), and vivid dreaming; critical for emotional memory consolidation, fear extinction learning, and dopaminergic system restoration.

The relationship between sleep stages and aging is not linear — it is asymmetric and stage-specific. SWS declines sharply with age: a healthy 25-year-old spends approximately 90 minutes in N3 per night; by age 65 this falls to 20-30 minutes, and in T2DM patients it falls further to 10-15 minutes (Tasali et al., 2008, PNAS). This SWS decline directly impairs glymphatic clearance efficiency and GH secretion amplitude. Conversely, REM sleep declines more modestly with age (approximately 15-20% reduction from youth to old age) but is particularly suppressed by alcohol (even moderate amounts), benzodiazepines, Z-drugs (zolpidem), and — critically for neuropathy patients — opioid analgesics including tramadol and oxycodone, which suppress both REM and SWS simultaneously.

The Sleep-Mortality J-Curve: Why Both Too Little and Too Much Sleep Kill

The largest meta-analysis of sleep duration and mortality (Cappuccio et al., 2010, Sleep; 16 prospective studies, n=1,382,999) found a J-shaped relationship: short sleep (<6 hours) increased all-cause mortality by 12% (RR 1.12, 95% CI 1.06-1.18), while long sleep (>9 hours) increased mortality by 30% (RR 1.30, 95% CI 1.22-1.38). The 7-8 hour range minimized mortality risk. A 2018 meta-analysis in JAMA Internal Medicine (Itani et al.; 28 prospective studies, n=1,172,471) found that short sleep was associated with a 48% increased risk of T2DM, 12% increased cardiovascular disease risk, and 13% increased dementia risk — effect sizes comparable to obesity and smoking. Mechanistically, each hour less than 7 hours of sleep per night is associated with: increased ghrelin (+28%), decreased leptin (-18%), increased cortisol (+15%), elevated IL-6 (+12%), and reduced natural killer cell cytotoxicity (-20%).

The Glymphatic System: Sleep’s Nightly Brain Detox and Alzheimer’s Prevention

The glymphatic system — named for its dependence on glial cells (specifically astrocyte aquaporin-4 water channels) and its functional analogy to the peripheral lymphatic system — operates through a convective flow mechanism first fully characterized by Nedergaard et al. (Science, 2013). During slow-wave sleep, the extracellular space of the brain expands by approximately 60% (Xie et al., 2013, Science; measured by diffusion-weighted MRI), driven by K+ outward conductance from neurons during hyperpolarized down-states. This expansion creates a pressure gradient that drives CSF through periarterial spaces (Virchow-Robin spaces), across the astrocyte AQP4 channels into the interstitium, and out through perivenous spaces into the cervical lymphatic system. The CSF flow during this process carries soluble amyloid-β, tau oligomers, α-synuclein, and lactate — all produced as metabolic byproducts of waking neural activity — out of the brain parenchyma.

In humans, Shokri-Kojori et al. (2018, PNAS; n=20; crossover design; 1 night sleep deprivation vs. normal sleep) found that even a single night of total sleep deprivation increased brain amyloid-β accumulation by 5% — detected by PET imaging with florbetapir — predominantly in the hippocampus and thalamus (the regions that are first affected in Alzheimer’s disease). The accumulation reversed with recovery sleep, but was cumulative with repeated sleep restriction: a mouse model (Holth et al., 2019, Science) demonstrated that 7 days of sleep disruption produced amyloid plaque increases that persisted after 3 days of recovery sleep. The clinical implication is stark: 5 years of sleeping 6 vs. 8 hours accumulates approximately 3.65× more amyloid exposure than optimal sleep — a burden with real long-term dementia consequences.

Sleep position also affects glymphatic efficiency: a seminal study in mice (Lee et al., 2015, Journal of Neuroscience) found lateral (side-sleeping) positioning produced 25% greater glymphatic flow than supine or prone positioning, attributed to gravitational effects on CSF distribution through Virchow-Robin spaces. While direct human replication using MRI-based glymphatic flow measurement is technically challenging, the finding is consistent with epidemiological data showing that back-sleeping is associated with higher snoring rates and obstructive sleep apnea — which themselves are associated with elevated amyloid accumulation — meaning the glymphatic and respiratory mechanics both favor lateral sleep positioning.

Circadian Biology and the Molecular Clock: Why Timing Is as Important as Duration

The 2017 Nobel Prize in Physiology/Medicine was awarded to Jeffrey Hall, Michael Rosbash, and Michael Young for discovering the molecular mechanism of circadian rhythms — the CLOCK/BMAL1 transcription factor complex that drives 24-hour oscillations in approximately 40-80% of all protein-coding genes, governing everything from insulin sensitivity (peaking at solar noon) to cortisol release (peaking at dawn) to melatonin secretion (peaking at 2-3 AM in consistent darkness). Every cell in the body has an autonomous molecular clock, and these peripheral clocks are synchronized to the master clock in the suprachiasmatic nucleus (SCN) of the hypothalamus — which is itself entrained by retinal light input via the retinohypothalamic tract.

Circadian disruption — whether from shift work, late-night light exposure, social jetlag (weekend sleep schedule drift), or eating at biologically inappropriate times — uncouples peripheral tissue clocks from the SCN master clock, a state called circadian misalignment. The health consequences are substantial: a 22-year prospective study of 74,862 nurses (Vetter et al., 2016, JAMA Internal Medicine) found that rotating night shift work increased T2DM risk by 42% (RR 1.42), independent of BMI, diet, and family history. Mechanistically, circadian misalignment impairs pancreatic β-cell insulin secretion (glucose-stimulated insulin secretion is 20-25% lower at nighttime biological clock phase), reduces peripheral insulin sensitivity through CLOCK/BMAL1-regulated GLUT4 expression, and upregulates NF-κB through REV-ERBα loss — directly connecting circadian disruption to the inflammaging cascade described in Post 93.

For DPN patients, circadian clock disruption has a specific neural consequence: the peripheral nerve and dorsal root ganglion express autonomous molecular clocks that regulate pain threshold, nerve conduction velocity, and nociceptor sensitivity in a time-of-day pattern. Peripheral neuropathy pain is well-documented to be worse at night — a clinical observation that is now mechanistically explained by circadian clock-driven changes in TRPV1 and Nav1.7 channel expression in DRG neurons (Zhang et al., 2020, Science Translational Medicine). Disrupting circadian alignment amplifies this nocturnal pain sensitization and reduces the pain relief window that normally occurs during daytime biological clock phases.

Obstructive Sleep Apnea and DPN: The Nocturnal Hypoxia–Nerve Fiber Destruction Link

Obstructive sleep apnea (OSA) affects an estimated 50-70% of patients with T2DM — a prevalence 3-4× higher than the general population — yet remains undiagnosed in the majority. The SleepAHEAD trial (Foster et al., 2009, Sleep; n=306 T2DM patients in Look AHEAD) found OSA prevalence of 86.6% in obese T2DM patients, with 30.5% having severe OSA (AHI ≥30 events/hour). This is the single most under-recognized comorbidity in my DPN patient population. For every patient I see with worsening neuropathy symptoms despite good HbA1c control, OSA is on my differential.

The mechanism connecting OSA to DPN progression operates through three parallel pathways. First, intermittent hypoxia and HIF-1α activation: repeated desaturation events (SpO₂ dropping to 80-85% during apneic episodes) activate hypoxia-inducible factor 1α (HIF-1α) in peripheral nerve, DRG, and endoneurial vessel endothelium. While HIF-1α acutely induces protective responses (VEGF upregulation, EPO production), chronic intermittent hypoxia — as occurs in untreated OSA — produces a maladaptive HIF-1α response that upregulates VEGF-driven pathological neovascularization in peripheral nerve, contributing to endoneurial edema and disrupted nerve blood-barrier function. A 2020 study in Diabetes Care (Tahrani et al.; n=234 T2DM patients; OSA diagnosed by polysomnography) found that T2DM patients with moderate-severe OSA had 47% more severe neuropathy scores than those without OSA, independent of HbA1c, BMI, and diabetes duration.

Second, sympathetic nervous system overdrive: each apneic episode triggers a cortical arousal and a surge in sympathetic outflow — heart rate acceleration, blood pressure spike (often 20-40 mmHg systolic), and catecholamine release — that repeats 5-30 times per hour in OSA patients. Over years, this produces sympathetic dominance that elevates resting heart rate, reduces heart rate variability, damages endoneurial microvascular endothelium through shear stress cycles, and maintains elevated NF-κB activation in circulating monocytes (through β2-adrenergic receptor signaling, as described in the inflammaging post). Third, sleep fragmentation: even mild OSA (AHI 5-15) significantly suppresses SWS and disrupts REM sleep through repeated microarousals, impairing the glymphatic clearance described above, preventing nocturnal GH secretion needed for peripheral nerve repair, and dysregulating the HPA axis through fragmented cortisol feedback cycles.

CPAP therapy for OSA has shown benefits for DPN in observational studies: the Tahrani et al. (2020) paper found that 6 months of CPAP use in T2DM+OSA patients reduced neuropathy disability scores by 28% and improved intraepidermal nerve fiber density (IENFD) compared to controls — the first direct evidence that treating OSA improves peripheral nerve histology. For cardiovascular outcomes, the ISAACC trial (2021, NEJM; n=2,717; OSA patients with acute coronary syndrome; CPAP vs. usual care) unfortunately did not show a reduction in MACE — but the trial had poor CPAP adherence (median use <3 hours/night) and excluded the most severe OSA cases, limiting its interpretability. The SAVE trial (McEvoy et al., 2016, NEJM; n=2,687; moderate-severe OSA + established cardiovascular disease; CPAP vs. usual care) also showed no MACE benefit but improved quality of life, mood, and daytime function significantly. For my patients, I recommend CPAP primarily for symptom relief, neuropathy modification, and sleep quality restoration — not cardiovascular event prevention, where the trial evidence is mixed.

Melatonin: Chronobiotic, Antioxidant, and Mitochondrial Protector

Melatonin is widely misunderstood as simply a “sleep hormone” when it is more accurately described as a chronobiotic — a signal that communicates darkness to the body’s circadian system, shifting the phase of peripheral clocks and initiating the cascade of physiological changes associated with nighttime biology. Melatonin secretion from the pineal gland begins approximately 2 hours before sleep onset (dim light melatonin onset, or DLMO), peaks around 2-3 AM, and suppresses to near-zero by dawn — driven entirely by the light environment, since even ordinary room lighting (150-200 lux) suppresses melatonin by 50% compared to complete darkness. Blue-wavelength light (460-480 nm) — emitted by LED screens, smartphones, and modern overhead lighting — is approximately 5× more melatonin-suppressive per lux than red-wavelength light.

Beyond circadian signaling, melatonin is one of the most potent endogenous antioxidants known: it directly scavenges hydroxyl radicals, superoxide, and peroxynitrite, and upregulates the antioxidant enzymes SOD, CAT, and glutathione peroxidase through MT1/MT2 receptor activation. Critically for DPN patients, melatonin accumulates in mitochondria at concentrations 10-100× higher than plasma — where it directly suppresses Complex I-derived ROS generation. This mitochondrial antioxidant function is the mechanism underlying the neuroprotective effects of melatonin in animal models of diabetic neuropathy: in streptozotocin-induced diabetic rats, melatonin supplementation prevented 60-80% of the nerve fiber loss typically observed at 12 weeks (Bhattacharya et al., 2009, European Journal of Pharmacology). Human clinical data on melatonin for DPN are limited to small trials, but the mechanistic rationale is compelling.

For clinical use as a chronobiotic (phase-shifting sleep timing), the effective melatonin dose is dramatically lower than what is typically sold in commercial supplements: 0.1-0.5 mg of melatonin taken 1-2 hours before desired sleep onset reliably shifts the circadian phase by 30-90 minutes (Lewy et al., 2006, PNAS). The commonly sold 5-10 mg doses produce supraphysiological plasma melatonin levels (500-1,000× normal peak) with minimal additional phase-shifting efficacy and potential downregulation of MT1/MT2 receptors with chronic use. I recommend 0.3-0.5 mg melatonin as a chronobiotic for patients with delayed sleep phase syndrome, shift work, or post-travel circadian disruption — and note that even optimal melatonin use does not compensate for blue-light exposure or irregular sleep-wake timing.

The DPN-Insomnia Spiral: Neuropathic Pain → Sleep Disruption → Worsened Neuropathy

Approximately 40-60% of patients with painful DPN report significant sleep disruption due to neuropathic symptoms — burning, tingling, electric shock sensations, and allodynia that are typically worst at night when circadian pain sensitization peaks and diurnal distractions are absent. This creates a bidirectional spiral: neuropathic pain disrupts sleep → sleep disruption impairs endogenous pain modulation (descending inhibitory pathways require adequate SWS for their serotonergic and noradrenergic restoration) → worsened pain the following night → further sleep disruption. Each component reinforces the other, and breaking the cycle requires simultaneously addressing both the pain and the sleep architecture deficit.

The most sleep-architecture-friendly pharmacological options for painful DPN are SNRIs (duloxetine and venlafaxine), which both reduce neuropathic pain through norepinephrine reuptake inhibition AND improve sleep continuity without suppressing REM or SWS. Pregabalin and gabapentin improve sleep in DPN patients — a 2007 RCT (Freynhagen et al., Pain; n=395; pregabalin vs. placebo) found pregabalin reduced time awake after sleep onset by 52% and improved sleep quality scores by 34% alongside significant pain reduction. Low-dose amitriptyline (10-25 mg/night) has both analgesic and sleep-promoting effects through anticholinergic and antihistaminergic mechanisms, though it suppresses REM sleep — making it a less ideal choice for patients where glymphatic clearance is a priority (i.e., those with cognitive concerns or family history of Alzheimer’s).

CBT-I: The Most Evidence-Based Sleep Treatment Available — and Why It Outperforms Every Sleep Drug

Cognitive behavioral therapy for insomnia (CBT-I) is the first-line treatment for chronic insomnia disorder per the American Academy of Sleep Medicine, American College of Physicians, and American Academy of Neurology — not pharmacotherapy. This recommendation is based on a 2015 meta-analysis (van Straten et al., Sleep Medicine Reviews; 11 RCTs, n=618) showing that CBT-I reduced sleep onset latency by 19 minutes, increased total sleep time by 32 minutes, reduced wake after sleep onset by 26 minutes, and improved sleep efficiency from 77% to 87% — with effect sizes maintained at 12-month follow-up. Sleep drugs improve these same metrics acutely but with progressive tolerance, rebound insomnia on discontinuation, and — in the case of benzodiazepines and Z-drugs — evidence of cognitive impairment and increased dementia risk with long-term use.

CBT-I consists of five core components delivered over 6-8 weekly sessions (now also available digitally through FDA-cleared apps Somryst and Sleepio at equivalent efficacy to therapist-delivered CBT-I per Ritterband et al., 2017, JAMA Psychiatry). First, stimulus control: restricting all bed activity to sleep and sex — no reading, screens, or worrying in bed — to break conditioned arousal. Second, sleep restriction: paradoxically limiting time in bed to actual sleep time (often 5-6 hours initially) to build homeostatic sleep pressure through adenosine accumulation, then titrating upward as sleep efficiency improves. Third, sleep hygiene: the environmental and behavioral factors (consistent wake time, cool bedroom temperature, darkness, caffeine cutoff, exercise timing). Fourth, cognitive restructuring: addressing catastrophizing beliefs about sleep (“If I don’t sleep 8 hours I’ll have a terrible day”). Fifth, relaxation techniques: progressive muscle relaxation, diaphragmatic breathing, and mindfulness-based stress reduction — which independently reduce cortisol and sympathetic arousal at sleep onset.

For DPN patients where neuropathic pain is a significant contributor to insomnia, I use a modified CBT-I approach that incorporates pain-specific cognitive restructuring (catastrophizing about pain at night is as behaviorally disruptive as catastrophizing about sleep), careful selection of sleep medications that minimize SWS suppression (duloxetine and pregabalin rather than zolpidem or benzodiazepines), and direct attention to leg elevation and footwear issues that may contribute to nocturnal discomfort. The evidence for CBT-I in chronic pain populations is robust: a 2021 Cochrane review (n=1,703) found that CBT-I reduced pain intensity by a modest but significant amount (Cohen’s d = 0.26) alongside sleep improvements, consistent with sleep restoration’s effect on descending pain inhibitory pathway function.

Practical Sleep Optimization: The Evidence-Based Protocol for DPN Patients

The sleep optimization protocol I use at Balance Foot & Ankle for DPN patients integrates circadian biology, glymphatic optimization, OSA screening, and pain management into a unified framework. The foundation is consistent wake time: waking at the same time daily — regardless of bedtime — is the single most powerful circadian anchor, as it drives adenosine clearance consistency and consolidates the cortisol awakening response, which in turn calibrates the entire 24-hour hormonal cycle. Target 7-9 hours of sleep opportunity (time in bed), not just time asleep. Light exposure is the second pillar: 10-30 minutes of outdoor bright light within 1 hour of waking (ideally sunlight) sets the circadian timer; blocking blue light for 2 hours before bed (blue-blocking glasses or amber-mode screens) preserves melatonin onset timing.

Bedroom temperature of 65-68°F (18-20°C) is supported by thermoregulation research as the optimal range for SWS initiation — core body temperature must drop by 1-2°C at sleep onset, and cool ambient temperature facilitates this drop 35% faster than 72°F environments (Okamoto-Mizuno et al., 2012, Journal of Physiological Anthropology). For DPN patients with autonomic neuropathy who have impaired thermoregulation, cooling mattress pads (BedJet, Chili OOLER) can maintain consistent bedding temperature and have shown anecdotal benefit in neuropathic symptom reduction — by reducing skin temperature sensitivity that drives nocturnal paresthesias. Caffeine cutoff by 2 PM (not noon) is sufficient for most adults given caffeine’s half-life of 5-7 hours, though patients with CYP1A2 slow metabolizer variants — approximately 50% of the population — have caffeine half-lives of 9-12 hours and should cut off by noon.

Frequently Asked Questions

Does napping compensate for short nighttime sleep?

Napping partially compensates for alertness and cognitive deficits from short nighttime sleep, but does not replicate the glymphatic clearance that occurs during nocturnal consolidated sleep. The Xie et al. (2013, Science) glymphatic data was derived specifically from extended nighttime sleep; daytime naps occur during a circadian phase when SWS is physiologically reduced and glymphatic flow is correspondingly lower. Additionally, napping after 3 PM reduces adenosine pressure (sleepiness drive) for the subsequent night, potentially perpetuating the short nighttime sleep cycle. Brief naps (10-20 minutes, the “NASA nap”) improve afternoon alertness and performance without entering N3 or disrupting nighttime sleep. Longer naps (>30 minutes) enter SWS, produce sleep inertia on waking, and more significantly erode evening adenosine pressure. For chronic insomnia patients, napping is generally contraindicated as part of sleep restriction therapy.

Is alcohol a valid sleep aid?

No. Alcohol accelerates sleep onset by activating GABA-A receptors (which is why it feels sedating) but produces a fundamentally disrupted sleep architecture. The first half of the night — dominated by SWS and NREM — is partially preserved when alcohol is metabolized during sleep; however, alcohol’s primary metabolite acetaldehyde is a stimulant that causes REM sleep fragmentation and suppression in the second half of the night. A 2018 meta-analysis (Alcoholism: Clinical and Experimental Research; 27 studies) found that moderate alcohol (1-2 drinks) reduced REM sleep by 24% and slow-wave sleep by 9% in the second half of the night. For glymphatic clearance purposes, the REM suppression from alcohol is less critical than the fragmentation of the SWS cycles when aldehyde rebound occurs — typically at 3-5 AM after an evening drink. Even one drink within 3 hours of bedtime meaningfully disrupts the sleep architecture required for amyloid-β clearance.

What sleep tracker should my patients use?

Consumer sleep trackers (Oura Ring Generation 4, Whoop 5.0, Fitbit Sense 3, Apple Watch Series 10) all estimate sleep stages through heart rate variability, movement, and skin temperature — not EEG. Their sleep stage accuracy is approximately 70-80% for N2/N3 discrimination and 80-85% for REM vs. non-REM classification compared to polysomnography (PSG) gold standard. This is sufficient for longitudinal trend tracking (how does sleep change over weeks of lifestyle intervention) but insufficient for diagnostic-quality sleep architecture assessment. The Oura Ring Generation 4 has the most validated sleep staging algorithm among consumer devices as of 2025 (Huilla et al., 2023, Journal of Sleep Research; n=46; PSG vs. Oura; κ = 0.71 for N3 detection). For clinical purposes: home sleep apnea testing (WatchPAT One, ResMed) is far more clinically useful than consumer trackers for detecting OSA — it measures respiratory effort, SpO₂, and cardiac cycles, yielding a reliable AHI estimate covered by insurance.

How does exercise timing affect sleep?

Morning exercise (6-10 AM) strengthens circadian entrainment by amplifying the cortisol awakening response and light-induced SCN firing. Afternoon exercise (3-6 PM) coincides with peak body temperature and produces the most SWS enhancement — the temperature rise during exercise is followed by a more pronounced nocturnal temperature drop, which is the primary trigger for SWS initiation. This is the optimal window for sleep quality. High-intensity evening exercise (within 90 minutes of bedtime) raises core body temperature, cortisol, and heart rate — all of which delay sleep onset by 30-90 minutes in most individuals. However, individual variation is substantial: some people tolerate 8 PM vigorous exercise without sleep disruption, while others are significantly affected. Self-monitoring with a sleep tracker over 2-4 weeks of exercise timing experiments is the most reliable way to identify the optimal window for each patient.

Sources and References

• Xie L, Kang H, Xu Q, et al. Sleep drives metabolite clearance from the adult brain. Science. 2013;342(6156):373-377.
• Shokri-Kojori E, Wang GJ, Wiers CE, et al. β-Amyloid accumulation in the human brain after one night of sleep deprivation. Proc Natl Acad Sci USA. 2018;115(17):4483-4488.
• Itani O, Jike M, Watanabe N, Kaneita Y. Short sleep duration and health outcomes: a systematic review, meta-analysis, and meta-regression. Sleep Med. 2017;32:246-256.
• Cappuccio FP, D’Elia L, Strazzullo P, Miller MA. Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies. Sleep. 2010;33(5):585-592.
• Foster GD, Sanders MH, Millman R, et al. Obstructive sleep apnea among obese patients with type 2 diabetes. Diabetes Care. 2009;32(6):1017-1019. [SleepAHEAD trial]
• Tahrani AA, Ali A, Raymond NT, et al. Obstructive sleep apnea and diabetic neuropathy: a novel association in patients with type 2 diabetes. Am J Respir Crit Care Med. 2012;186(5):434-441.
• Vetter C, Devore EE, Wegrzyn LR, et al. Association between rotating night shift work and risk of coronary heart disease among women. JAMA. 2016;315(16):1726-1734.
• Lewy AJ, Emens J, Jackman A, Yuhas K. Circadian uses of melatonin in humans. Chronobiol Int. 2006;23(1-2):403-412.
• van Straten A, van der Zweerde T, Kleiboer A, et al. Cognitive and behavioral therapies in the treatment of insomnia: A meta-analysis. Sleep Med Rev. 2018;38:3-16.
• Holth JK, Fritschi SK, Wang C, et al. The sleep-wake cycle regulates brain interstitial fluid tau in mice and CSF tau in humans. Science. 2019;363(6429):880-884.

Related Articles

• Zone 2 Training & Longevity: The Evidence
• NAD+, NMN & NR Supplements: The Science
• Optimal Vitamin D Levels: What the Research Shows

Dive Deeper into Longevity

• Muscle Mass, Sarcopenia & Longevity
• Hormones and Longevity
• NMN and NAD+: Cellular Aging & Sirtuins
• Longevity Biomarkers: GrimAge & DunedinPACE