🌙 Sleep better

What actually decides this outcome, in order of size

Four different problems share one word, and the treatments for them are not interchangeable. What decides which one you have:

1. Whether you are short of sleep or bad at it. Short sleep duration is common enough in the adult population to be the modal state rather than the exception Liu 2016, and it is the one version of this problem that no product addresses, because the intervention is time in bed. Extending it works: collegiate athletes given an extended schedule gained 110.9 ± 79.7 minutes of objective sleep and improved timed sprint from 16.2 to 15.5 seconds (P<0.001) Mah 2011.
2. Whether the airway closes. Obstructive sleep apnea is estimated at 936 million adults aged 30 to 69 worldwide, 425 million of them moderate to severe Benjafield 2019. It presents as unrefreshing sleep of normal duration, it is invisible to every blood test on this site, and no sedative improves it.
3. Whether the problem is conditioned arousal rather than a chemical deficit. Across 37 studies, cognitive behavioral therapy for insomnia produced remission in 36.0% of patients against 16.9% of controls, a pooled odds ratio of 3.28 (95% CI 2.30 to 4.68, P<0.001) Wu 2015. Nothing sold for sleep has an effect size like that, and it is a behavioral treatment.
4. Whether the clock is in the wrong place. A delayed circadian phase looks exactly like insomnia at 11 pm and exactly like normal sleep at 3 am. Melatonin is a timing signal for this and the dosing is counterintuitive: maximum phase advances occurred with 0.5 mg taken in the afternoon, 2 to 4 hours before dim light melatonin onset Burgess 2010.
5. Whether the depth is there, which is what makes sleep worth having. Suppressing slow-wave sleep for three nights without shortening total sleep time reduced insulin sensitivity, with the size of the fall correlated to the size of the slow-wave reduction Tasali 2008. Sleep also drives a large increase in interstitial space and in convective clearance of metabolites from the brain Xie 2013. Depth is a physiological process, not a feeling.

The order to run these in, and what has to be true first

Work out which of the four it is, remove the chemical causes, treat the behavior, then time the clock. Sedatives are last because three of the four problems are made worse by them.

1. Triage by pattern before buying anything. Cannot fall asleep goes to Sleep onset — GABAergic & sedative. Falls asleep fine and wakes at 3 am goes to What's keeping you awake — the upstream causes. Sleeps eight hours and wakes unrefreshed goes to Sleep depth, slow-wave & recovery quality and to an apnea question. Sleeps well on the wrong schedule goes to Circadian & melatonergic rhythm.
2. Remove the caffeine before adding anything. Four hundred milligrams taken 0, 3 or 6 hours before bed all measurably disrupted sleep against placebo, and the 6-hour dose cost more than an hour Drake 2013. With an elimination half-life around 5 hours, and substantially longer on estrogen-containing contraceptives, an afternoon coffee is a midnight drug.
3. Then the behavioral treatment, which outperforms the shelf. Stimulus control, sleep restriction and the cognitive component together are what the odds ratio of 3.28 was measured on Wu 2015. It is unglamorous, it is free in several formats, and it is the only intervention on this hub with an effect size of that magnitude.
4. Then the blood work that explains the 3 am waking. Ferritin, because low iron stores are associated with restless legs and periodic limb movements at ferritin values well inside the normal range; TSH (Thyroid-Stimulating Hormone) with Free T4 (Thyroxine); HbA1c (Hemoglobin A1c) for the nocturnal glycemic story; and Progesterone where perimenopause is plausible, since progesterone metabolites are positive allosteric modulators at GABA-A and their loss is felt as sleep loss.
5. Melatonin for timing, at a timing dose. Half a milligram in the afternoon shifts the clock; ten milligrams at bedtime is a sedative dose of a molecule that is not a sedative Burgess 2010. Epitalon and DSIP belong to the same rhythm pathway with far weaker evidence.
6. Glycine, Magnesium, L-Theanine, Apigenin and Magnolia Bark are the onset layer. Glycine is an inhibitory neurotransmitter at its own receptor and lowers core temperature through peripheral vasodilation, which is a plausible onset mechanism; apigenin is a benzodiazepine-site ligand at GABA-A with low affinity; magnolia bark's honokiol and magnolol are GABA-A positive modulators. Ashwagandha has its own insomnia trial and belongs beside them.
7. Trazodone, Doxepin and Phenibut are the sedative tail and they are not equivalent. Low-dose doxepin is a selective H1 antagonist at the doses used for sleep maintenance; trazodone is a 5-HT2A antagonist with antihistamine activity. Phenibut is a GABA-B agonist with documented tolerance, physical dependence and a difficult withdrawal, which places it in a different category of decision from everything else on this hub.

What gets bought for this that cannot move it

Sedation is not sleep, and the difference is measurable. Several agents that reliably produce unconsciousness degrade the architecture that makes sleep restorative. Alcohol is the clearest case: it shortens onset latency, suppresses REM in the first half of the night and produces a rebound with fragmented sleep in the second half. An agent that removes slow-wave sleep while producing eight recorded hours has removed the part that changes insulin sensitivity Tasali 2008.

Melatonin at ten milligrams is the category error of this whole hub. The phase response data shows the timing dose is small and the timing is early Burgess 2010. A large dose at bedtime produces a weak hypnotic effect and no phase shift, which is why it disappoints in exactly the condition it should treat, and the disappointment then gets attributed to the molecule rather than to the protocol.

The category that fails structurally is anything sedative taken against a mechanical airway problem. Sedatives reduce upper airway muscle tone, which is the mechanism that keeps the airway open during sleep. Given the scale of undiagnosed obstructive sleep apnea Benjafield 2019, a meaningful proportion of the people buying sleep aids for unrefreshing sleep are buying something with a mechanism for making their problem worse.

And if the answer is simply not enough time in bed, no page on this site can help. Extending sleep opportunity improved objective sleep by nearly two hours and improved measured performance Mah 2011, and that is a scheduling decision rather than a purchase. If sleep is fine and the fatigue is not, the question is Oxygen carrying, blood sugar & the boring causes or Adrenal, cortisol rhythm & stress-driven fatigue.

How you would know it was working, on a real read-out and a real timescale

The prediction that separates sedation from sleep: a genuine improvement shows up as shorter time to fall asleep and fewer awakenings without a longer time in bed, and as a daytime measure that improves. If the only thing that changed is that you remember less of the night, that is sedation.

• A written sleep diary for two weeks before and two weeks after. Time to sleep onset, number of awakenings, total time in bed and a morning restedness rating. This is the instrument the behavioral treatment is built on Wu 2015 and it is more reliable than any consumer device.
• Ferritin with hs-CRP (High-Sensitivity C-Reactive Protein), and the threshold that matters is not the laboratory's. Restless legs and periodic limb movements are associated with iron stores well inside the reported normal range, so a ferritin in the 20s with a night of leg discomfort is a finding rather than a normal result. CRP goes on the same draw because ferritin rises with inflammation independently of stores.
• TSH (Thyroid-Stimulating Hormone) with Free T4 (Thyroxine) once. Hyperthyroidism produces insomnia and hypothyroidism produces unrefreshing sleep, and the same single draw answers both.
• HbA1c (Hemoglobin A1c) if the waking is consistently at the same hour. A reproducible 3 am waking with sweating or hunger is worth a glycemic explanation before a GABAergic one, and it is one of the few sleep complaints with a blood read-out at all.
• A daytime performance measure, chosen in advance. Reaction time, a timed sprint, or a fixed cognitive task. This is the end point that separates more hours from better hours, and it is what moved in the sleep extension study Mah 2011.

What will fool you. Consumer sleep trackers estimate stages from movement and heart rate rather than measuring electroencephalography, so their slow-wave and REM numbers should not be used to judge a product; the total sleep time and the timing are the parts worth reading. Anything sedative improves the memory of the night more than the night itself. And the first week of any sleep intervention carries a strong expectation effect, which is exactly the week people decide on.

Sources read for these sections

• Wu JQ, et al. Cognitive Behavioral Therapy for Insomnia Comorbid With Psychiatric and Medical Conditions: A Meta-analysis. JAMA Internal Medicine 2015;175(9):1461-72 · PMID 26147487
• Benjafield AV. Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis. Lancet Respiratory Medicine 2019;7(8):687-698 · PMID 31300334
• Burgess HJ. Human phase response curves to three days of daily melatonin: 0.5 mg versus 3.0 mg. Journal of Clinical Endocrinology and Metabolism 2010;95(7):3325-31 · PMID 20410229
• Tasali E. Slow-wave sleep and the risk of type 2 diabetes in humans. Proceedings of the National Academy of Sciences 2008;105(3):1044-9 · PMID 18172212
• Xie L. Sleep drives metabolite clearance from the adult brain. Science 2013;342(6156):373-7 · PMID 24136970
• Mah CD. The effects of sleep extension on the athletic performance of collegiate basketball players. Sleep 2011;34(7):943-50 · PMID 21731144
• Drake C. Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. Journal of Clinical Sleep Medicine 2013;9(11):1195-1200 · PMID 24235903
• Liu Y. Prevalence of Healthy Sleep Duration among Adults - United States, 2014. MMWR Morbidity and Mortality Weekly Report 2016;65(6):137-41 · PMID 26890214