Evidence Library · Sleep & Recovery
Sleep & Recovery · AXIOM SELENE
Sleep Optimization: Evidence-Based Fundamentals (and What the Marketing Gets Wrong)
CBT-I has strong RCT evidence for insomnia. Consistent sleep timing and adequate duration have solid observational backing. Many marketed sleep gadgets and supplements do not. Here is what the research actually supports — and where the honest gaps remain.
Evidence grade
Moderate evidence — Some RCTs or robust observational evidence
The Fundamentals vs. the Noise
Sleep research divides cleanly into two categories: well-replicated findings with strong evidence, and a large commercial ecosystem of products, apps, and supplements with weak or no controlled trial support. This article focuses on the former. The basics — consistent sleep timing, adequate duration, and structured behavioural intervention when sleep is genuinely impaired — carry far more evidential weight than any tracking device, cooling pad, or supplement stack. That is not dismissiveness toward technology; it is an honest read of where the controlled trials point.
CBT-I: The Highest-Evidence Intervention for Chronic Insomnia
Cognitive Behavioural Therapy for Insomnia (CBT-I) is a structured multicomponent programme that combines sleep restriction, stimulus control, sleep hygiene education, and cognitive restructuring. It is the only sleep intervention with a strong clinical guideline recommendation from the American Academy of Sleep Medicine — placing it above all pharmacological alternatives as the first-line treatment for chronic insomnia disorder.
CBT-I typically involves four to eight sessions with a trained provider, though digital and self-guided versions have also shown benefit in trials. The treatment gains appear durable: follow-up data suggest continued improvement without ongoing intervention.
- ✓ Proven
- The American Academy of Sleep Medicine (2021) issued a STRONG recommendation that clinicians use multicomponent CBT-I as first-line treatment for chronic insomnia disorder in adults — the highest guideline tier, based on 49 randomised controlled trials.
- ✓ Proven
- In a meta-analysis of 49 RCTs, CBT-I produced a 33% higher remission rate versus control conditions, and an effect size of 0.95 on the Insomnia Severity Index — a clinically meaningful improvement.
🅰 Edinger et al. 2021 — 'Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline' (Journal of Clinical Sleep Medicine, strong recommendation, 49 RCTs)— AASM strong recommendation: clinicians should use multicomponent CBT-I as first-line treatment for chronic insomnia disorder in adults. Based on 49 RCTs demonstrating clinically meaningful improvements.
🅰 Edinger et al. 2021 — 'Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine systematic review, meta-analysis, and GRADE assessment' (Journal of Clinical Sleep Medicine)— Meta-analysis of 49 RCTs. CBT-I showed 33% higher remission rates vs. control, effect size 0.95 for insomnia severity (ISI). Components include stimulus control, sleep restriction, and cognitive restructuring.
Sleep Duration: The U-Shaped Risk Curve
Large meta-analyses consistently show a U-shaped relationship between nightly sleep duration and cardiovascular and metabolic health outcomes. Both insufficient and excessive sleep are associated with elevated risk — with the lowest risk concentrated in the 7–8 hours range for most adults. The National Sleep Foundation and CDC both recommend 7–9 hours for adults aged 18–64.
Importantly, the vast majority of this evidence is observational: tracking what people already do rather than randomising them to sleep durations. This means confounding is a real concern — sicker people may sleep more, and high-stress individuals may sleep less. The association is consistent and large enough across studies to be taken seriously, but 'associated with elevated risk' is not the same as 'causes elevated risk.'
- ~ Probable (incomplete evidence)
- Meta-analyses show that sleeping fewer than 6 hours per night is associated with a 20–32% higher likelihood of developing hypertension compared to sleeping 7–8 hours, across large observational cohorts.
- ~ Probable (incomplete evidence)
- Short sleep (<6 h) and long sleep (>9 h) are both associated with elevated coronary heart disease risk: approximately 1.48× and 1.38× respectively, compared to 7–8 h sleepers, in pooled meta-analyses.
🅰 Covassin & Singh 2016 — 'Sleep Duration and Cardiovascular Disease Risk' (Sleep Medicine Clinics) — U-shaped relationship between sleep duration and cardiovascular outcomes across large meta-analyses— Observational meta-analysis — association, not proven causation. Confounding by health status is a limitation acknowledged in the review.
🅰 Covassin & Singh 2016 — 'Sleep Duration and Cardiovascular Disease Risk' (Sleep Medicine Clinics) — U-shaped relationship between sleep duration and cardiovascular outcomes across large meta-analyses— Observational data across multiple meta-analyses. Causal direction is not proven; long sleep may be a marker of underlying illness rather than a cause of poor outcomes.
Consistent Sleep Timing: Circadian Alignment Matters
Beyond duration, the regularity of sleep timing appears independently relevant to health. 'Social jetlag' — the mismatch between biological sleep timing and socially imposed schedules (typically measured as the difference in sleep midpoint between weekdays and weekends) — is associated with metabolic and cardiovascular markers in population studies.
This is important context for understanding why simply sleeping enough hours is not the whole story. Irregular sleep timing, even at sufficient total hours, disrupts circadian entrainment — the biological alignment of the body's clocks with the external light-dark cycle. Consistent wake and sleep times are a central component of CBT-I's stimulus control technique, and they are among the most universally supported sleep hygiene recommendations.
- ~ Probable (incomplete evidence)
- In a population-representative Czech study (n = 1,601 blood samples), social jetlag of ≥ 0.65 hours was associated with higher total cholesterol and LDL cholesterol, particularly in adults over 50 — independent of total sleep duration.
🅰 Sládek et al. 2023 — 'Population-representative study reveals cardiovascular and metabolic disease biomarkers associated with misaligned sleep schedules' (Sleep, n = 1,601 blood samples)— Observational, cross-sectional design. Association only — does not establish that reducing social jetlag causally improves cholesterol levels.
Evening Light and Blue Light: Real Mechanism, Overstated Consumer Claims
The underlying mechanism is well-established: light in the blue-wavelength range (460–480 nm) activates retinal cells that signal the brain's master clock to inhibit melatonin secretion. When this happens in the hours before bed, it suppresses the natural melatonin rise that facilitates sleep onset and can delay circadian phase. This is physiology, not speculation.
What is less certain is whether commercially available blue-blocking glasses produce meaningful sleep improvements in healthy people who do not have sleep disorders. A 2020 systematic review and meta-analysis found small-to-medium objective effects from wearing blue-blocking glasses, with wide confidence intervals that crossed zero for some outcomes. Crucially, benefits were substantially larger for individuals who already had sleep disorders than for healthy sleepers. Marketing claims that normalise blue-blocking glasses as an essential daily tool for everyone outrun the current evidence.
- ~ Probable (incomplete evidence)
- A 2020 systematic review and meta-analysis of 12 RCTs found that blue-blocking glasses produced small-to-medium objective improvements in sleep efficiency (Hedges g = 0.31, 95% CI −0.05 to 0.66) — a confidence interval that does not exclude zero at the objective level.
🅰 Shechter et al. 2020 — 'Interventions to reduce short-wavelength (blue) light exposure at night and their effects on sleep: A systematic review and meta-analysis' (SLEEP Advances)— Small-to-medium effect with confidence interval crossing zero on objective actigraphy measures. Self-reported sleep quality showed a larger effect (g = −1.25), which may reflect expectation bias. Evidence is stronger for people with existing sleep disorders.
Melatonin Supplements: Honest About the Evidence
Melatonin is widely sold as a sleep aid and widely assumed to work for general insomnia. The evidence is more nuanced. For chronic insomnia disorder, a 2022 review of systematic reviews found mixed results when melatonin was compared to placebo, with some guidelines — including the VA/DoD clinical practice guideline — recommending against its use for this condition.
Where melatonin has better-supported utility is in circadian phase-shifting: helping the body clock adapt to jet lag or shift work by timing supplementation relative to the new schedule. This is a mechanistically different use case from treating primary insomnia, and the two are often conflated in consumer contexts.
For short-term use, the safety profile is generally considered favourable. What is lacking is convincing controlled evidence that melatonin meaningfully treats chronic insomnia in otherwise healthy adults, compared to either placebo or CBT-I.
- ~ Probable (incomplete evidence)
- A 2022 review (NCBI Bookshelf) found 'mixed results on the clinical effectiveness of melatonin for insomnia when compared to placebo,' and noted that the VA/DoD clinical practice guideline recommends against melatonin use for chronic insomnia disorder.
🅱 Hamel & Horton 2022 (CADTH) — 'Melatonin for the Treatment of Insomnia: A 2022 Update' (NCBI Bookshelf / StatPearls)— Narrative review of systematic reviews and guidelines, not a primary meta-analysis. Evidence quality for melatonin in chronic insomnia is mixed; stronger evidence exists for circadian phase-shifting applications such as jet lag.
Sleep Environment: Room Temperature, Noise, and Darkness
Core body temperature naturally declines during the transition to sleep. A cooler sleep environment is thought to support this drop, and most sleep medicine guidance recommends keeping the bedroom between approximately 18–20 °C (65–68 °F) for general adults. A 2023 observational study of older adults (n = 50, ~11,000 person-nights, published in Science of the Total Environment) found that sleep efficiency declined 5–10% as ambient temperature rose above 25 °C (77 °F) — and that considerable individual variation exists.
The temperature-sleep relationship has a plausible mechanism and consistent expert consensus behind it, but controlled trial evidence in general populations (rather than older adults or specific clinical groups) is limited. Darkness and minimising noise during sleep are similarly recommended with strong mechanistic rationale but relatively sparse large-scale RCT evidence compared to CBT-I. They belong in a foundational sleep hygiene protocol, with the honest note that the evidence base is not as deep as that for CBT-I or sleep duration.
Sources in this section
- 🅱 Lan et al. 2023 — 'Optimal Sleep Temperature for Seniors Is Between 68 to 77 Degrees Fahrenheit' (Science of the Total Environment, observational study, n = 50 older adults, ~11,000 person-nights) — Observational study in older adults — findings may not generalise to younger populations. Substantial between-individual differences in optimal temperature were noted. Published in a peer-reviewed journal but a small sample.
What we don't yet know
Honesty about gaps in the evidence is what distinguishes us from most wellness media.
- CBT-I evidence is strong for diagnosed chronic insomnia disorder. Whether structured CBT-I meaningfully benefits people with subclinical sleep difficulties (who don't meet diagnostic criteria) is less well-studied — though sleep hygiene components are broadly reasonable.
- The sleep duration evidence is almost entirely observational. The optimal duration for specific populations — athletes, people with shift-work histories, individuals in tropical climates where ambient temperature and light cycles differ substantially from study populations — has not been established in controlled trials.
- Most sleep research has been conducted in Western (predominantly North American and European) populations. Whether findings about circadian entrainment, light exposure thresholds, or optimal temperature apply equally to people living in tropical climates such as Phuket — where light intensity, humidity, and heat differ substantially — is not established.
- Exercise is robustly associated with improved sleep quality in observational research and small RCTs, but the optimal type, timing, and dose for sleep benefit are not established. Evening high-intensity exercise may delay sleep onset in some individuals, but the evidence here is mixed.
- The evidence for consumer sleep tracking devices improving objective sleep outcomes (rather than just measuring sleep) is limited. Awareness of sleep data may help some individuals, but it can also increase sleep anxiety ('orthosomnia'), potentially worsening outcomes.
- Many marketed 'sleep stack' supplements (magnesium glycinate, L-theanine, ashwagandha, valerian, glycine) have plausible mechanisms and some small trial evidence, but no large-scale RCTs confirm meaningful effects for chronic sleep difficulties in generally healthy adults. This article does not evaluate them.
- Long-term effects of sleep restriction therapy — a CBT-I component that involves temporarily limiting time in bed and initially worsens sleep before improving it — in vulnerable populations (e.g., people with bipolar disorder or severe fatigue conditions) are not well characterised.
All sources
🅰 Primary
Edinger et al. 2021 — 'Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline' (Journal of Clinical Sleep Medicine, strong recommendation, 49 RCTs)AASM strong recommendation: clinicians should use multicomponent CBT-I as first-line treatment for chronic insomnia disorder in adults. Based on 49 RCTs demonstrating clinically meaningful improvements.
🅰 Primary
Edinger et al. 2021 — 'Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine systematic review, meta-analysis, and GRADE assessment' (Journal of Clinical Sleep Medicine)Meta-analysis of 49 RCTs. CBT-I showed 33% higher remission rates vs. control, effect size 0.95 for insomnia severity (ISI). Components include stimulus control, sleep restriction, and cognitive restructuring.
🅰 Primary
Covassin & Singh 2016 — 'Sleep Duration and Cardiovascular Disease Risk' (Sleep Medicine Clinics) — U-shaped relationship between sleep duration and cardiovascular outcomes across large meta-analysesReview of meta-analyses (>3 million participants). Observational data — association, not proven causation. U-shaped risk curve: lowest risk at 7–8 h. Short sleep defined as <6 h, long sleep as >9 h.
🅱 Credible secondary
Hamel & Horton 2022 (CADTH) — 'Melatonin for the Treatment of Insomnia: A 2022 Update' (NCBI Bookshelf / StatPearls)Narrative review of systematic reviews and guidelines. Finds 'mixed results' for melatonin vs. placebo in chronic insomnia. VA/DoD guideline recommends against melatonin for chronic insomnia disorder. Evidence stronger for circadian phase-shifting (jet lag) than for primary insomnia.
🅰 Primary
Shechter et al. 2020 — 'Interventions to reduce short-wavelength (blue) light exposure at night and their effects on sleep: A systematic review and meta-analysis' (SLEEP Advances)12 RCTs on blue-blocking glasses. Objective sleep efficiency effect size g = 0.31 (small-to-medium, CI crossing zero). Subjective PSQI effect size g = −1.25 (large). Benefits appeared greater in individuals with existing sleep disorders than in healthy sleepers.
🅰 Primary
Sládek et al. 2023 — 'Population-representative study reveals cardiovascular and metabolic disease biomarkers associated with misaligned sleep schedules' (Sleep, n = 1,601 blood samples)Observational — association, not proven causation. Social jetlag ≥ 0.65 h correlated with higher total cholesterol and LDL in adults over 50. Cross-sectional design limits causal inference.
This article reviews published research on sleep and is for educational purposes only. It is not medical advice and does not address individual health circumstances. If you have a sleep disorder, chronic insomnia, or a medical condition affecting sleep, consult a qualified healthcare provider. CBT-I should be delivered by or under the supervision of a trained clinician.
Last verified: 2026-06-28 · ← Evidence Library