Sleep is the most consequential wellness variable I have tracked. The quality of my sleep affects everything else — mood, energy, focus, stress tolerance, even physical recovery. Most sleep advice is generic ("sleep more!"). I wanted to know what specifically would actually improve my sleep.
Over six months, I tested six different interventions, each for at least three weeks, with consistent tracking. Here is the honest report on what worked.
The setup
Tracked daily:
- Bedtime and wake time
- Subjective sleep quality 1-5
- Times woke during night
- Energy on waking 1-5
- Daytime energy average 1-5
Baseline (3 weeks before any intervention):
- Average sleep: 6.8 hours
- Sleep quality: 3.2/5
- Wake-time energy: 2.9/5
- Daytime energy: 3.3/5
1. Cool bedroom (65°F / 18°C)
The protocol: kept bedroom at 65°F overnight using a smart thermostat. Previously had been around 70°F.
What changed:
- Average sleep: 7.0 hours (slight increase)
- Sleep quality: 3.5 (notable improvement)
- Wake-time energy: 3.2 (slight improvement)
- Times waking during night: noticeably fewer
Verdict: Real and easy. Required minimal effort, produced consistent improvement. The temperature was cooler than felt comfortable initially but became normal within a week.
2. No screens 1 hour before bed
The protocol: no phone, laptop, or TV after 9 PM (for a 10:30 PM bedtime). Allowed: paper books, conversation, music.
What changed:
- Average sleep: 7.2 hours (continued improvement)
- Sleep quality: 3.7 (notable improvement)
- Wake-time energy: 3.4 (notable improvement)
- Time to fall asleep: noticeably faster
Verdict: Hard but worth it. The first week was uncomfortable; I had been using screen time as wind-down. Found that reading was a better wind-down once I committed to it.
3. Magnesium glycinate (400mg before bed)
The protocol: 400mg of magnesium glycinate 30 minutes before bed.
What changed:
- Average sleep: 7.2 hours (no significant change)
- Sleep quality: 3.8 (slight improvement)
- Wake-time energy: 3.5 (slight improvement)
- Subjective "deeper sleep" feeling: yes, but hard to measure objectively
Verdict: Modest improvement. Cheap (about $0.20/dose). Worth continuing given the cost-benefit.
4. Consistent wake time (including weekends)
The protocol: wake at the same time (6:30 AM) every day, including weekends. Previously slept in until 8 or 9 on weekends.
What changed:
- Average sleep: 7.0 hours (slightly less, due to consistent earlier wake)
- Sleep quality: 3.9 (improvement)
- Wake-time energy: 3.7 (notable improvement)
- Sunday-Monday transition: dramatically easier
Verdict: Counterintuitively powerful. The weekend "catch-up sleep" had been making weekdays harder. Consistent timing produced better baseline rhythm.
5. Morning sunlight exposure (15+ minutes)
The protocol: outside in natural light within 30 minutes of waking, for at least 15 minutes. No sunglasses.
What changed:
- Average sleep: 7.1 hours
- Sleep quality: 4.0 (notable improvement)
- Wake-time energy: 3.8 (continued improvement)
- Falling asleep at night: noticeably easier
Verdict: The most pleasantly surprising intervention. Free, took 15 minutes, produced real improvement in nighttime sleep onset. The circadian rhythm research is real.
6. No caffeine after 2 PM
The protocol: no coffee, tea, or other caffeine after 2 PM. Previously had been okay with afternoon coffee.
What changed:
- Average sleep: 7.2 hours
- Sleep quality: 4.1 (continued improvement)
- Wake-time energy: 3.9
- Times waking during night: continued reduction
Verdict: Caffeine's 6-hour half-life is real. The afternoon coffee had been affecting sleep more than I realized.
The end-of-experiment numbers (with all 6 interventions in place)
- Average sleep: 7.3 hours (up from 6.8)
- Sleep quality: 4.1 (up from 3.2)
- Wake-time energy: 3.9 (up from 2.9)
- Daytime energy: 4.0 (up from 3.3)
The interventions stacked. Each one alone produced modest improvement. Together they produced substantial improvement.
The honest ranking by impact
If I had to do them again in order of impact per effort:
- Morning sunlight (highest impact, lowest effort)
- No screens before bed (high impact, moderate effort)
- Cool bedroom (high impact, one-time setup)
- Consistent wake time (moderate impact, requires social adjustment)
- No caffeine after 2 PM (moderate impact, requires habit change)
- Magnesium supplement (lowest impact, lowest effort)
What I learned about sleep
1. Multiple interventions stack. Each one alone produced 5-10% improvement in measurable sleep quality. Combined, they produced 25-30% improvement.
2. Environment matters more than I thought. Temperature, light, screen exposure — environmental factors dominated supplement and timing factors.
3. The circadian rhythm work is real. Morning sunlight and consistent wake time produced larger improvements than I expected. The research on circadian rhythm regulation appears to be accurate.
4. The "sleep hygiene" generic advice mostly worked. Conventional sleep wisdom is conventional because it generally works.
5. Tracking matters. Without measurement, I would not have noticed the cumulative improvement. The numbers told me what subjective experience could not.
What I did NOT find useful
Things I tried but abandoned:
- Sleep tracking apps with detailed sleep stage analysis. The data was noisy and inactionable. Better to track simple metrics.
- White noise machines. Did nothing for me. Maybe useful for noisier environments.
- Specific bedtime routines (lavender, tea, etc.). Pleasant but not measurably impactful.
- Weighted blankets. Tried one. Liked it but did not see measurable sleep improvement.
This is not advice
Sleep affects different people differently. The interventions that worked for me may not work for you. The point of writing this down is the methodology — test specific changes, measure outcomes, keep what works.
If you have sleep issues, talk to a doctor. Some sleep problems require medical evaluation rather than lifestyle changes.
The takeaway
Sleep is the single highest-leverage wellness intervention available to most people. The interventions above produced substantial improvement at modest cost.
If you want better sleep, start with morning sunlight exposure (free, 15 minutes, immediate effect). Then add the others incrementally as the impact compounds. Most people can substantially improve their sleep within 60 days through environmental and behavioral changes.
