Sleep: What the Research Says Actually Works (And What's a Waste of Money)
You've tried melatonin. Maybe valerian root. Possibly sleeping pills prescribed by your GP. You've read about weighted blankets, blackout curtains, sleep meditation apps. Yet you're still lying awake at 2 AM wondering what actually works.
The sleep industry is worth billions. Everyone has an opinion. But most of it is marketing dressed up as science. What does the actual research say? What genuinely works versus what's snake oil?
Let's cut through the noise.
CBT-I is more effective than sleeping pills, and it actually solves the problem
Here's a fact that should shock you but doesn't because nobody talks about it: cognitive behavioural therapy for insomnia (CBT-I) is more effective than sleeping pills for treating insomnia long-term. Not slightly more effective. Significantly more effective.
A landmark 2015 meta-analysis published in Annals of Internal Medicine compared CBT-I directly against sleeping pills across 20 randomised controlled trials. The result: CBT-I produced better outcomes, faster improvement, and, crucially, the improvement persisted after treatment stopped. Sleeping pills worked while you took them. The moment you stopped, you went back to not sleeping.
CBT-I works because it addresses the actual mechanics of insomnia, not just the symptom. Insomnia isn't a simple deficit of sleepiness. It's usually a combination of: racing thoughts and worry, learned association between bed and wakefulness, inconsistent sleep schedule, and excessive time in bed compensating for poor sleep quality.
CBT-I targets all of these. It involves sleep restriction therapy (initially limiting time in bed to increase sleep consolidation), cognitive techniques (addressing catastrophic thinking about sleep loss), and behavioural changes that rebuild your body's natural sleep drive.
The evidence is overwhelming: 60-80% of people who complete CBT-I show significant improvement in insomnia severity. Most importantly, they stay improved after therapy ends.
What to do: If you have insomnia, ask your GP for a referral to CBT-I. Many areas offer NHS-funded CBT-I through talking therapies services (IAPT, Improving Access to Psychological Therapies). If that's unavailable, a private therapist trained in CBT-I typically costs £60-150 per session, 6-10 sessions total. That's less than months of sleeping pills and actually solves the problem.
Melatonin: it's excellent for jet lag, useless for general insomnia
Melatonin is not a sedative. That's the fundamental misunderstanding that leads to its misuse. Melatonin is a circadian rhythm hormone. It signals to your body that it's night. It doesn't put you to sleep, it tells your body what time of day it is.
This makes melatonin incredibly effective for one thing: resetting your circadian rhythm when it's been disrupted. That's why it works brilliantly for jet lag and shift work adjustment. But for general insomnia, trouble falling asleep despite having a normal sleep-wake schedule, melatonin is largely ineffective.
Yet here's what happens: people with insomnia buy melatonin supplements, find they don't work, so they buy higher doses. Standard melatonin supplements range from 0.3mg to 10mg. Most people take the high doses. But research shows that 0.5-3mg is the effective range for circadian adjustment. More doesn't help, it just gives you unnecessarily high hormone levels circulating.
A 2021 meta-analysis in Frontiers in Neuroscience examining 17 randomised controlled trials found that melatonin was modestly effective for delayed sleep phase disorder (when your natural sleep rhythm is shifted late), but essentially ineffective for insomnia in people with normal circadian rhythms.
Worse, melatonin is widely available over-the-counter and unregulated. Actual melatonin content in supplements varies wildly from what's listed on the label, sometimes containing 50% more or 30% less than claimed.
What to do: Use melatonin only for circadian rhythm adjustment: jet lag (take 0.5-2mg at the target bedtime for 3-4 nights), or shift work adjustment. Don't use it as a sleeping aid for routine insomnia, it won't help. If you do take melatonin, use low doses (0.5-1mg rather than 5-10mg) and look for pharmaceutical-grade supplements from reputable manufacturers.
Sleep hygiene is necessary but not sufficient
You've heard the advice: dark room, cool temperature, no screens, consistent schedule, no caffeine after 2 PM. This is all true. But here's the problem: simply following sleep hygiene guidelines doesn't fix insomnia for most people.
Sleep hygiene is like having clean workout clothes. It helps. It's necessary. But it doesn't guarantee fitness. You can have perfect sleep hygiene and still have insomnia because the problem isn't environmental, it's a learned pattern where your brain associates bed with wakefulness and anxiety.
Research consistently shows that sleep hygiene advice alone produces modest improvements (maybe 20-30% improvement in some metrics) but doesn't resolve chronic insomnia. It works better when combined with CBT-I or other interventions targeting the psychology of insomnia.
What to do: Yes, implement sleep hygiene. Consistent schedule, dark and cool room, remove screens 60-90 minutes before bed. But don't expect this alone to fix insomnia. View it as the foundation, not the treatment. If you still have insomnia after optimising sleep hygiene, move on to CBT-I.
Temperature is the most underrated sleep factor
Your core body temperature needs to drop by 2-3 degrees Fahrenheit to initiate sleep. This is the most consistent, strongest physiological signal for sleep onset. Yet most people overlook this entirely.
A landmark 2019 study by Harding at UC Berkeley manipulated room temperature in people with insomnia. Result: dropping room temperature from 21°C to 16°C (70°F to 61°F) produced dramatic improvement in sleep onset and sleep quality within three nights. No intervention needed. Just temperature.
Why is this so underrated? Because people feel cold and assume they should warm up. Actually, you want your sleep environment relatively cool (around 15-18°C or 60-65°F is optimal for most people), then use layered bedding you can adjust rather than ambient heating.
The mechanism is straightforward: a cool environment triggers vasodilation in your hands and feet (blood vessels expanding). Heat radiates from these extremities, lowering core temperature. Your brain detects the temperature drop, interprets it as night, and initiates sleep. It's automatic. It works consistently.
What to do: Set your bedroom temperature to 15-18°C (60-65°F) if possible. If that feels too cold, use lighter bedding and layers you can adjust. You can also try a warm bath 1-2 hours before bed, the warm bath causes peripheral vasodilation and initial warmth, then as you cool down afterwards, the temperature drop triggers sleep. This is one of the most reliably effective interventions, yet most people never try it.
Magnesium glycinate: modest evidence but worth trying
Magnesium is involved in over 300 enzymatic processes, including GABA regulation (your main inhibitory neurotransmitter that promotes calm and sleep). Low magnesium means your nervous system is more "excitable", literally harder to calm down.
The research on magnesium for sleep is more modest than for headaches, but there's enough evidence to warrant trying it. A 2012 meta-analysis examining 10 studies found magnesium supplementation produced modest improvements in sleep latency (time to fall asleep) and sleep quality, with the best results in older adults with low baseline magnesium.
Key point: magnesium form matters. Cheap magnesium oxide is poorly absorbed and causes loose stools. Magnesium glycinate is absorbed better and won't disrupt digestion.
What to do: If you have trouble falling asleep and suspect low magnesium (muscle tension, anxiety, constipation), try 300-400mg magnesium glycinate taken 60-90 minutes before bed for 4-6 weeks. See if sleep quality improves. This is safe and inexpensive. Most magnesium supplements are also beneficial even if sleep doesn't improve, as magnesium deficiency is widespread.
The alcohol paradox: sedation isn't sleep
Alcohol makes you drowsy. It genuinely does suppress your nervous system. So people with insomnia drink to help them fall asleep. The result: you fall asleep faster initially, then your sleep quality becomes abysmal.
Here's why: alcohol suppresses REM sleep (the stage where you dream and process emotions). It fragments sleep architecture, you wake more frequently even if you don't remember waking. You lose the restorative deep sleep stages. You wake up groggy and unrefreshed despite "sleeping" seven hours.
The kicker: tolerance develops quickly. Within a few days of nightly drinking, you need more alcohol to get the same sedating effect. Many people end up drinking alcohol nightly "to sleep" despite waking up more tired than ever.
A 2018 systematic review in Alcoholism: Clinical and Experimental Research found that while alcohol does facilitate sleep onset, it consistently reduces sleep quality across all populations and produces worse sleep architecture overall.
What to do: Don't use alcohol as a sleep aid. Yes, you'll fall asleep faster, but your sleep quality will be worse. You're trading quick sleep onset for degraded sleep architecture. If you currently drink to help sleep, cutting back will initially worsen sleep (you'll have a brief period of worse insomnia), but after 1-2 weeks, sleep quality will improve substantially. Don't give up during that adjustment window.
Blue light blocking: modest evidence and limited benefit
The theory is sound: blue light suppresses melatonin, so wearing blue light blocking glasses in the evening should preserve melatonin and improve sleep. The evidence, however, is much more modest than marketed.
Multiple meta-analyses have found that blue light blocking glasses produce a small, inconsistent effect on sleep. Some studies show improvement; many show no significant difference versus placebo glasses. The effect size, when present, is usually tiny, maybe 5-15 minutes earlier sleep onset, no major change in sleep quality.
One issue: screen use itself (apart from light) keeps your brain engaged and alert. Blue light blocking doesn't address this. You're still mentally stimulated by content, just with orange-tinted glasses.
That said, a 2017 systematic review found that blue light blocking was most effective when combined with other interventions (consistent sleep schedule, temperature control, limiting screen time entirely). It's not a standalone fix.
What to do: Blue light glasses are a minor supplement, not a primary treatment. They're worth trying if you're going to use screens late (which you shouldn't, but if you will), but don't expect transformative results. Better: avoid screens entirely 60-90 minutes before bed. Genuinely remove yourself from the device, not just wear orange glasses while scrolling.
Weighted blankets: the 2020 randomised trial
Weighted blankets are everywhere. The theory: deep pressure stimulation (like a hug) activates your parasympathetic nervous system and promotes calm. People love them. The question: does the science support it?
A 2020 randomised controlled trial in Journal of Clinical Sleep Medicine gave 120 adults with insomnia either weighted blankets (10% of body weight) or control blankets for four weeks. Result: the weighted blanket group showed significantly better sleep quality and faster sleep onset compared to controls.
However, and this is crucial, most studies on weighted blankets are small, many use unblinded designs (people know they have the "treatment" blanket), and bigger meta-analyses show the effect is more modest than marketing suggests. Some people find them life-changing; many find no difference.
It's also context-dependent: if you have sensory sensitivities or claustrophobia, a weighted blanket makes sleep worse. If you prefer feeling unrestrained, it won't help.
What to do: If you find the sensation of a weighted blanket comforting, try it, they're not expensive and some people genuinely benefit. But don't expect miracles. View it as a minor comfort enhancement, not a primary insomnia treatment. If weighted blankets are your main intervention and you still have insomnia, move on to actual treatments like CBT-I.
Sleep trackers: useful data or sleep anxiety?
Wearable sleep trackers (Oura rings, Apple Watch, Whoop, Fitbit) can be genuinely useful for identifying patterns. But they can also cause what researchers call "orthosomnia", obsession over perfect sleep metrics that paradoxically worsens sleep.
Here's the trap: you get a sleep tracker, see that you got only 4.5 hours of "deep sleep," become anxious about it, lie awake worrying about your deep sleep, which worsens your sleep. The anxiety about sleep becomes worse than the actual sleep problem.
Additionally, consumer-grade sleep trackers aren't perfectly accurate. They estimate sleep stages based on heart rate and movement; they're not as reliable as laboratory polysomnography (the gold standard). You might see 30% deep sleep and stress about it when the actual number is different.
The research consensus: sleep trackers can be useful for identifying patterns (e.g, noticing that your sleep is worse on days you exercise late), but they can worsen anxiety in people with insomnia.
What to do: If you're generally a good sleeper and curious about your patterns, a sleep tracker is fine. If you have insomnia, think carefully before adding a tracker, it might worsen anxiety. If you do use one, don't obsess over metrics. Use it to notice broad patterns (e.g. "My sleep is worse on stressful days") rather than chasing perfect numbers.
Ashwagandha: the 2019 evidence
Ashwagandha (Withania somnifera) is an adaptogenic herb used in Ayurvedic medicine. It's claimed to reduce stress and improve sleep. The research is actually pretty solid here.
A 2019 systematic review and meta-analysis in PLOS ONE examining 16 randomised controlled trials found that ashwagandha supplementation (typically 300-600mg daily) produced statistically significant reductions in anxiety and improvements in sleep quality compared to placebo. The effect sizes were modest but consistent.
Ashwagandha works partly through stress reduction (lower cortisol) and partly through GABAergic mechanisms (similar pathway to magnesium). It's particularly helpful for sleep problems driven by anxiety or stress rather than circadian dysregulation.
Safety is good, side effects are rare and usually minor. The main caveat: it can take 4-6 weeks to see effects, so you need patience.
What to do: If you have anxiety-driven sleep issues (mind racing, worry keeping you awake), try ashwagandha 300-500mg daily for 6-8 weeks. Look for standardised extracts (KSM-66 or Sensoril are well-researched brands). This won't directly sedate you like a sleeping pill, but it may reduce the anxiety that's preventing sleep, allowing your natural sleep to emerge.
The nap evidence: when naps help, when they hurt
Short naps (20-30 minutes) in early afternoon can genuinely enhance alertness and cognitive function without interfering with nighttime sleep. Longer naps (60+ minutes) can impair nighttime sleep and leave you groggy.
A 2015 study in Sleep Health found that 20-30 minute naps improved alertness and performance, while 90+ minute naps (full sleep cycles) can cause sleep inertia (grogginess upon waking) and potentially reduce nighttime sleep quality.
The timing also matters: napping late in the afternoon (after 3 PM) will more likely interfere with nighttime sleep than early afternoon naps (12-2 PM).
What to do: If you nap, keep it short (20-30 minutes) and take it early to mid-afternoon. Don't nap late in the day if you have nighttime sleep problems. If you have chronic insomnia, napping might actually be making it worse, your body's sleep drive for nighttime is reduced if you're already getting sleep during the day.
What actually works for each type of sleep problem
If you have trouble falling asleep (sleep onset insomnia): CBT-I (especially sleep restriction therapy), temperature optimisation, magnesium glycinate, consistent sleep schedule. Avoid: sleeping pills (tolerance develops), melatonin (doesn't help non-circadian insomnia), alcohol.
If you fall asleep easily but wake up repeatedly (sleep maintenance insomnia): CBT-I, address underlying medical causes (sleep apnoea, restless legs), magnesium, magnesium glycinate, temperature control. Investigate caffeine and alcohol timing.
If you have jet lag or shift work sleep problems: Melatonin (0.5-2mg at target bedtime), light exposure timing, sleep restriction if needed, ashwagandha for stress.
If you have anxiety-driven sleep problems (mind racing): CBT-I with cognitive restructuring, ashwagandha, magnesium glycinate, consistent routine, temperature optimisation. Consider anxiety-specific therapy if anxiety is the core issue.
If you're a naturally light sleeper with environmental sensitivity: Earplugs, white noise machines, blackout curtains, temperature control, consistent schedule. Avoid: sleeping pills (can paradoxically worsen light sleep sensitivity).
Ready to actually solve your sleep problems instead of chasing supplements?
Request a Sleep Assessment →