"Sleep hygiene" is a grab-bag term. The reality is a short list of interventions with large effects and a long list of interventions with small effects. Use the short list first.
Tier 1: large effect, easy to do
Consistent timing. Variable bedtime independently predicts cardiometabolic disease and depressive symptomatology, separately from total sleep duration ( Besedovsky et al. 2019 ). Target: bedtime within ±30 minutes night-to-night. Wake time within ±30 minutes on free days vs workdays (large "social jetlag" is a real metabolic stressor).
Bedroom temperature 16-19°C (60-67°F). Cool rooms shorten sleep onset latency and increase time in slow-wave sleep. The mechanism: core body temperature naturally drops ~1°C during sleep onset; a cool room facilitates the drop.
Caffeine cutoff. Caffeine half-life averages ~5 hours, with significant genetic variability (CYP1A2 fast vs slow metabolizers). For a 200 mg afternoon coffee at 3pm, 100 mg is still active at 8pm. Cut caffeine by 2pm if bedtime is 10pm. Some users need 10am as the cutoff; a 2-week experiment with no caffeine after 10am tells you which bucket you're in.
No alcohol within 4 hours of bed, ideally zero on training days. Alcohol is sedating (faster sleep onset) and sleep-disruptive (fragmented REM, increased awakenings in the second half of the night). Net recovery cost is large. For resistance trainees, even moderate drinking attenuates muscle protein synthesis post-workout.
Tier 2: moderate effect, still easy
Morning bright light 10+ minutes. Anchors circadian phase via SCN input. Outdoor light (10,000+ lux on a sunny morning, 1,000+ on overcast) is an order of magnitude brighter than indoor lighting (100-300 lux). Stepping outside with coffee for 10 minutes after waking produces reliable circadian entrainment.
Evening light <50 lux, 2 hours pre-bed. Bright indoor lighting in the evening suppresses melatonin secretion and delays circadian phase. Dim the lights 2 hours before bed. Warm color-temperature screens (f.lux, Night Shift) help but don't fully replace lower ambient light levels.
Regular aerobic exercise, NOT within 2 hours of bed. Cardio done >4 hours pre-bed improves sleep depth and total time. Done within 2 hours of bed, it can delay onset by 30+ minutes for some.
Tier 3: small-to-moderate effect
Meal timing. Large meals within 3 hours of bed impair sleep quality in controlled studies. Light protein or minimal food the last 3 hours is the conservative recommendation. Body composition goals aside, the sleep case alone justifies it.
Magnesium glycinate 300-400 mg pre-bed. See the magnesium glycinate entry. Effect on sleep is small but real for people with low dietary magnesium; larger for those with documented deficiency.
Melatonin 0.3 mg 30-60 min pre-bed. Physiologic dose. Higher doses (3-10 mg) are pharmacologic and often produce next-day grogginess without better sleep. See the melatonin compound entry.
Cognitive behavioral therapy for insomnia (CBT-I). The most-evidenced treatment for chronic insomnia, outperforms sleep medications at follow-up. Apps (Somryst, CBT-i Coach) deliver it reasonably well. Ong 2014 (Sleep, n=54) showed mindfulness-based interventions also produce durable insomnia improvement ( Ong et al. 2014 ).
Tier 4: weak or overrated
- Blue-blocking glasses in the evening. Modest effect beyond dimming the lights themselves. Not harmful; not the game-changer proponents suggest.
- Mouth tape. Good for snorers and mouth-breathers; not a general sleep-quality intervention for the average sleeper.
- Sleep gummies (melatonin + L-theanine + ashwagandha stacks at 5+ mg melatonin). High melatonin dose often counterproductive. The stack rarely outperforms the 0.3 mg melatonin monotherapy.
- White noise. Helps some, not others. Try a free app first before buying a $150 machine.
Stacking protocol
| Phase | Dose | Frequency | Notes |
|---|---|---|---|
| Bedtime consistency | +/- 30 min night-to-night | daily | Hardest to install; biggest payoff. |
| Bedroom temp | 16-19°C (60-67°F) | all night | Cooling mattress pad is cheaper than AC upgrade. |
| Caffeine cutoff | 200 mg max, by 14:00 | daily | Adjust later only if you're a confirmed slow CYP1A2 metabolizer. |
| Morning light | 10+ min outdoor | daily within 60 min of waking | Overcast counts. |
| Evening dim | <50 lux, 2h pre-bed | daily | Dim switch + warm bulbs. |
| Alcohol limit | Zero within 4h of bed | daily | Ideally zero on training days. |
| Melatonin (if needed) | 0.3 mg | 30-60 min pre-bed | Use for travel/jet lag; chronic use not required |
Counter-view
Guzey 2019 critique of Walker's "Why We Sleep" flagged several overstatements in the popular sleep literature ( Walker 2017 ); some claims about precise effect sizes (e.g., "short sleep causes cancer" framings) are stronger than the data supports. The conservative read is directional: short sleep is bad, consistent sleep is good, specific quantitative claims vary in strength. Matthew Walker himself has since updated several positions; the honest practitioner treats the popular summaries as directional.
