Most people trying to fix their sleep reach for a supplement first. The evidence, laid out honestly, argues for doing that last — not first. Rank sleep interventions by how well they're actually supported in trials, and a clear hierarchy falls out: structured therapy at the top, a handful of hygiene habits in the middle, and supplements doing modest, supporting work at the bottom.
The hierarchy, at a glance
"Sleep hygiene" gets used as a catch-all for everything from cutting caffeine to buying a supplement, which is part of the problem — it flattens interventions with very different levels of evidence into one bucket. Separated out by trial quality and effect size, the tiers look like this.
| Tier | Evidence strength | What the trials show |
|---|---|---|
| CBT-I (structured therapy) | Strong | 20-RCT meta-analysis: 19.0 min faster sleep onset, 26.0 min less time awake overnight, +9.9 points sleep efficiency (total sleep time: not significant). |
| Consistent schedule & light timing | Strong (mechanistic) | Most-replicated circadian finding in the field; no single trial number, but consistent across reviews. |
| Sleep hygiene tips alone | Unquantified — never tested against a control | 15-study review: no trial compared it with sham or no treatment; pre-to-post change averaged 5% sleep efficiency and 2 PSQI points, and it lost to CBT-I by a medium-to-large margin. |
| Supplements (e.g. magnesium) | Modest, low-quality | 3 trials, 151 older adults: 17.4 min faster onset vs placebo; total sleep time not significant; evidence graded low to very low. |
Tier three: sleep hygiene tips alone
"Sleep hygiene" — the standard list of tips like avoiding caffeine late in the day, keeping a cool dark room, and not scrolling in bed — is the advice everyone already knows. A systematic review and meta-analysis in Family Practice pooled 15 studies of sleep hygiene education for poor sleep or insomnia, and the single most important thing about it is a gap in the literature rather than a result: twelve of the fifteen studies compared sleep hygiene education with CBT-I, three with mindfulness-based therapy, and none with sham or no treatment.1
That means the field has never established what sleep hygiene education does relative to doing nothing. The improvements the review reports — averaging about 5% in sleep-diary sleep efficiency and two points on the Pittsburgh Sleep Quality Index — are pre-treatment to post-treatment changes, with small-to-medium effect sizes. People do get better while receiving sleep hygiene education. Whether they get better because of it is not something these 15 studies can answer, because insomnia improves over time and with attention regardless. That's the same reason an uncontrolled before-and-after number is worth so much less than a placebo-controlled one, whether the intervention is a leaflet or a capsule.
What the review does establish clearly: sleep hygiene education was significantly less efficacious than CBT-I, with the difference in effect size ranging from medium to large, and the SHE–CBT-I gap averaging about 8% in sleep efficiency and two PSQI points. Only subjective measures reached significance — objective ones did not. The authors' own conclusion is deliberately unresolved: "unanswered methodological and implementation issues prevent a firm conclusion to be made on whether SHE has a role in a stepped-care model for insomnia in primary care."1
What's actually inside CBT-I — and why the "hygiene" part isn't doing the work
This is the part that gets missed when people treat CBT-I and "sleep hygiene tips" as two flavors of the same thing. CBT-I is a package of educational, cognitive and behavioral strategies. A 2024 component network meta-analysis in JAMA Psychiatry — pooling 241 randomized controlled trials and 31,452 adults (mean age 45.4; 67% women), the largest analysis of its kind — pulled that package apart to see which pieces were actually driving the result.4 Reported as incremental odds ratios for remission:
- Cognitive restructuring: iOR 1.68 (95% CI 1.28–2.20) — the strongest single component.
- Third-wave components (acceptance- and mindfulness-based): iOR 1.49 (1.10–2.03).
- Sleep restriction: iOR 1.49 (1.04–2.13).
- Stimulus control: iOR 1.43 (1.00–2.05) — beneficial, though the lower bound sits exactly on 1.00.
- Sleep hygiene education: not essential — iOR 1.01 (0.77–1.32). Essentially no incremental benefit once the other components are accounted for.
- Relaxation procedures: potentially counterproductive — iOR 0.81 (0.64–1.02). An earlier version of this article listed relaxation training as one of the things making CBT-I work. On this analysis it may do the opposite, and that's worth knowing before buying a "relaxation for sleep" programme.
- Delivery format: in-person, therapist-led programmes were most beneficial — iOR 1.83 (1.19–2.81).
That's a useful, slightly uncomfortable finding for anyone selling "sleep hygiene checklists" as a product: the tips list isn't the active ingredient. The behavioral rules around when you're allowed to be in bed, and the work of restructuring anxious thoughts about sleep, are. Combine that with the Family Practice review's finding that hygiene education has never been tested against no treatment, and the honest summary is that hygiene tips sit at the bottom of this hierarchy not because their effect is small but because nobody has established that they have one on their own.
Tier one: cognitive behavioral therapy for insomnia (CBT-I)
CBT-I isn't a tips list — it's a structured intervention combining stimulus control, sleep restriction and cognitive therapy, usually delivered over several sessions. A meta-analysis of 20 randomized controlled trials in 1,162 adults with chronic insomnia (64% female, mean age 56) found that at the end of treatment, sleep onset latency improved by 19.03 minutes (95% CI 14.12 to 23.93), wake after sleep onset by 26.00 minutes (15.48 to 36.52), and sleep efficiency by 9.91 percentage points (8.09 to 11.73).2 Total sleep time improved by 7.61 minutes, but that one was not statistically significant (95% CI −0.51 to 15.74) — worth saying, because CBT-I's benefit is mostly about falling asleep faster and staying asleep, not about spending more hours in bed asleep. Changes seemed to be sustained at later time points and no adverse outcomes were reported. One limitation the authors flag: trials of insomnia comorbid with medical, sleep or psychiatric disorders were excluded, so the estimates apply most cleanly to uncomplicated chronic insomnia. We go deeper on this in our dedicated CBT-I guide.
Access used to be the main argument against CBT-I: it's harder to find a trained provider than it is to buy a supplement. That gap has narrowed. The 2024 component analysis found in-person delivery still outperformed other formats, but digital and self-guided versions delivering the same core components still beat control. The honest caveat is that not all digital programmes include the full component set; an app that mainly delivers sleep hygiene education is, per the same data, delivering the one piece with no measurable incremental effect.
Tier two: consistent schedule and light exposure
Sitting between the tips list and full CBT-I is a smaller set of habits with genuinely strong, mechanistic backing: keeping a consistent sleep and wake time, and getting bright light earlier in the day while dimming it before bed. This isn't folk wisdom — light is the primary input the body's internal clock runs on. We cover the mechanism in more depth in our deep sleep guide, and the practical version for a shifted clock in jet lag and shift work. The distinction that matters here: this tier earns its place through consistent replication across many studies, not one blockbuster trial — a different, but still solid, kind of evidence. It also costs nothing, which is exactly why it gets less marketing attention than a bottle with an ingredient list on it. And if you're waking at 3am rather than struggling to fall asleep, the stimulus-control logic is the relevant lever.
Where supplementation actually fits
A systematic review and meta-analysis of oral magnesium for insomnia in older adults pooled just three randomized trials in 151 people across three countries. Sleep onset latency was 17.36 minutes shorter after magnesium than placebo (95% CI −27.27 to −7.44; p=0.0006). Total sleep time improved by 16.06 minutes, but that difference was not statistically significant. All three trials were at moderate-to-high risk of bias, and the authors graded the outcomes on low to very low quality evidence, concluding that "the quality of literature is substandard for physicians to make well-informed recommendations."3 An earlier version of this article said the trials were run specifically in people who were already low in magnesium; they weren't — participants were older adults with insomnia, regardless of magnesium status. A newer, larger trial in general adults found a similarly small effect: insomnia severity fell 3.9 points versus 2.3 on placebo, a Cohen's d of 0.2, with all other outcomes null.5 We've written that one up in full in what the new magnesium trial actually found, and the form question in why we use glycinate.
Supplementation is a reasonable, modest assist layered on top of the higher-evidence interventions. It isn't a substitute for them, and no ingredient we sell claims to be. The same goes for melatonin, where the dose-response picture is less impressive than the marketing.
One thing worth ruling out before any of this: if you snore loudly, wake gasping, or feel exhausted no matter how long you spend in bed, none of the four tiers above is the right tool. The STOP-BANG questions take two minutes, and supplements do not treat apnea.
Bottom line
If sleep is a real, ongoing problem rather than an occasional bad night, the evidence-ranked order is: structured CBT-I first (a sleep specialist or a validated CBT-I programme, several of which are now available digitally), consistent light exposure and sleep timing second, basic sleep hygiene habits third — with the honest caveat that their standalone effect has never been measured against a control — and a targeted supplement as a modest layer on top. Reaching for a bottle before doing any of the first three is the one order the evidence doesn't support. Matching your effort to the size of the problem is the whole point: quick habit fixes for occasional rough nights, and the structured, evidence-heavier option for insomnia that's actually persistent.
This article is for informational purposes only and has not been evaluated by the FDA. It is not intended to diagnose, treat, cure, or prevent any disease. Speak with a healthcare provider before starting any new supplement or sleep treatment, especially if you take medication or have an existing health condition.
References
- Chung KF, Lee CT, Yeung WF, Chan MS, Chung EW, Lin WL. Sleep hygiene education as a treatment of insomnia: a systematic review and meta-analysis. Family Practice, 2018;35(4):365-375. pubmed.ncbi.nlm.nih.gov/29194467
- Trauer JM, Qian MY, Doyle JS, Rajaratnam SM, Cunnington D. Cognitive Behavioral Therapy for Chronic Insomnia: A Systematic Review and Meta-analysis. Annals of Internal Medicine, 2015;163(3):191-204. pubmed.ncbi.nlm.nih.gov/26054060
- Mah J, Pitre T. Oral magnesium supplementation for insomnia in older adults: a Systematic Review & Meta-Analysis. BMC Complementary Medicine and Therapies, 2021;21:125. ncbi.nlm.nih.gov/pmc/articles/PMC8053283
- Furukawa Y, Sakata M, Yamamoto R, et al. Components and Delivery Formats of Cognitive Behavioral Therapy for Chronic Insomnia in Adults: A Systematic Review and Component Network Meta-Analysis. JAMA Psychiatry, 2024;81(4):357-367. pubmed.ncbi.nlm.nih.gov/38231522
- Schuster J, Cycelskij I, Lopresti A, Hahn A. Magnesium Bisglycinate Supplementation in Healthy Adults Reporting Poor Sleep: A Randomized, Placebo-Controlled Trial. Nature and Science of Sleep, 2025;17:2027-2040. tandfonline.com
Related reading: CBT-I: The Sleep Treatment With Better Evidence Than Any Supplement.
Proco Magnesium Glycinate provides 275mg of elemental magnesium from 2,500mg of magnesium glycinate across three capsules. Shop Magnesium →