Non-Prescription Sleep Aids

← Back to Chapters

Most people who sleep poorly do not begin at the prescription pad. They begin in a pharmacy aisle, a supplement shop, or a dispensary — and they usually do it without telling anyone. Surveys consistently find that the majority of adults with insomnia symptoms have tried an over-the-counter or "natural" product, and a large fraction use one regularly, yet fewer than half mention it to a clinician. That silence matters: these agents carry real anticholinergic burden, real hepatotoxicity, real drug interactions, and real next-day impairment, and they are frequently stacked on top of prescribed sedatives, antidepressants, and analgesics. A clinician who does not ask specifically — "What do you take to help you sleep, including anything you buy yourself?" — is working from an incomplete medication list. This chapter maps the whole non-prescription landscape by mechanism, states plainly what the evidence does and does not support, and gives the dosing and safety detail needed to counsel patients rather than simply defer to the label.

A word on regulation — why "natural" is not the same as "tested"

In the United States, dietary supplements — which includes melatonin, valerian, magnesium, and every botanical in this chapter — are governed by the Dietary Supplement Health and Education Act of 1994 (DSHEA). Under DSHEA, manufacturers do not have to demonstrate efficacy, do not have to prove safety before marketing, and are not subject to pre-market FDA review of content or purity. The FDA can act only after a product reaches the market and is shown to be adulterated or misbranded. The practical consequences are concrete: label doses are frequently inaccurate, contamination and undeclared ingredients are common, and there is no assurance that two bottles of the "same" product contain the same thing. Independent third-party verification (USP Verified, NSF, ConsumerLab) is the only practical safeguard, and clinicians should recommend it explicitly. Cannabis occupies its own regulatory category — federally a Schedule I substance, legal for medical or adult use under a patchwork of state laws, and subject to state-level (not FDA) testing standards that vary widely in rigor.

Over-the-counter antihistamines (diphenhydramine, doxylamine)

The first-generation antihistamines are the active ingredient in nearly every product labeled "PM," "nighttime," or "sleep." Diphenhydramine (Benadryl, ZzzQuil, and the "PM" analgesic combinations) and doxylamine (Unisom SleepTabs) are ethanolamine-class H₁ antagonists that cross the blood–brain barrier and block central histaminergic wake-promoting signaling. Sedation is a class effect, not a therapeutic refinement — it is the same property that makes these drugs poor daytime allergy agents.

Two facts dominate their clinical profile. First, tolerance to the sedative effect develops rapidly — within about three to four days of continuous use — so their usefulness is largely confined to occasional, short-term situations. Second, they are potently anticholinergic. Beyond the familiar dry mouth, constipation, urinary retention, and blurred vision, central antimuscarinic activity produces next-day grogginess, impaired psychomotor performance, and — in older or cognitively vulnerable patients — confusion and frank delirium. Diphenhydramine and doxylamine each score the maximum on validated anticholinergic burden scales.

Other cautions: additive sedation and anticholinergic load when combined with TCAs, low-potency antipsychotics, bladder antimuscarinics, or opioids; anticholinergic delirium in overdose (which can also cause seizures, wide-complex tachyarrhythmia, and a classic antimuscarinic toxidrome); and the hidden-dose problem of "PM" combination analgesics, where a patient taking acetaminophen-diphenhydramine nightly may be unaware of either ingredient's cumulative exposure. Doxylamine plus pyridoxine (vitamin B₆) is separately used and FDA-approved for nausea of pregnancy, which is worth knowing because pregnant patients sometimes repurpose it — or plain doxylamine — as a sleep aid. Dimenhydrinate and meclizine (antivertigo/antiemetic antihistamines) are occasionally used off-label for sleep and carry the same anticholinergic liabilities.

Melatonin — a chronobiotic, not a sedative

Melatonin is the most-used sleep supplement in the world, and it is also the most misunderstood. Endogenous melatonin is secreted by the pineal gland under suprachiasmatic control, rising after dusk (the "dim-light melatonin onset," DLMO) and signaling biological night. Exogenous melatonin works primarily as a chronobiotic — it shifts the timing of the circadian clock — and only weakly as a direct hypnotic. Understanding that distinction predicts where it helps and where it disappoints.

Where the evidence is strongest: circadian rhythm sleep–wake disorders. Low-dose melatonin taken several hours before the desired bedtime is effective for delayed sleep–wake phase disorder, for jet lag (dosed at the destination's bedtime), and as adjunct for some shift-work and non-24-hour disorders. It has a reasonable evidence base in children with autism spectrum disorder and neurodevelopmental insomnia, where guidelines support a trial after behavioral measures. Where it is weaker: primary insomnia in adults. Meta-analyses show a statistically real but clinically small reduction in sleep-onset latency (on the order of several minutes) and a modest increase in total sleep time — genuinely helpful for some patients, unimpressive for many.

The content problem is not hypothetical. An analysis of commercial melatonin products (JAMA, 2023) found actual melatonin content ranging from 74% to 347% of the labeled amount, and several "melatonin" gummies also contained unlabeled CBD. Because these products are frequently marketed as gummies that look and taste like candy, pediatric melatonin ingestions reported to US poison centers have risen more than five-fold over the past decade, now numbering in the hundreds of thousands of calls, with a small number of hospitalizations and rare deaths. Counsel families to store it like a medication, not a snack, and to buy third-party-verified products.

Safety and interactions: melatonin is generally well tolerated short-term — headache, daytime sleepiness, vivid dreams, and dizziness are the usual complaints. Long-term safety data are limited. It can potentiate other sedatives and may modestly affect glucose regulation and blood pressure; theoretical interactions exist with anticoagulants and with CYP1A2 substrates (fluvoxamine markedly raises melatonin levels). The prescription melatonin-receptor agonists ramelteon (MT₁/MT₂) and tasimelteon (for non-24) are pharmacologic cousins covered in the prescription Sleep Medications and Orexin chapters; they are not OTC.

Herbal and botanical sleep aids

The botanicals cluster mechanistically around GABAergic modulation. Evidence quality is generally low — small trials, heterogeneous preparations, high placebo response — but a few have enough signal and a favorable safety profile to justify a supervised trial in a motivated patient. The overriding safety themes are additive sedation, herb–drug interactions (many are CYP substrates or inhibitors), unpredictable potency, and, for one agent, serious hepatotoxicity.

Minerals and amino acids

Magnesium is the most defensible mineral option. It acts as an NMDA antagonist and GABA-A agonist, and deficiency is genuinely common in older adults, in alcohol use, and with PPI or diuretic use. A frequently cited 8-week trial in older adults with insomnia used ~500 mg elemental magnesium daily with modest improvement in sleep parameters. Well-absorbed organic salts (glycinate, citrate) are preferred; the main adverse effect is dose-dependent diarrhea, and magnesium accumulates in renal impairment — check kidney function before recommending standing doses. Glycine (about 3 g at bedtime) is an inhibitory neurotransmitter that lowers core body temperature and has small trials suggesting improved subjective sleep quality and reduced daytime fatigue. L-theanine (~200 mg), an amino acid from tea, promotes relaxation and alpha-wave activity with a benign safety profile and limited but generally positive sleep data, often as an adjunct.

Tryptophan and 5-HTP are serotonin/melatonin precursors with mixed efficacy data; tryptophan carries the historical caution of eosinophilia–myalgia syndrome (traced to a contaminated batch in 1989, but a reminder of supply-chain risk), and both raise a real theoretical concern for serotonin syndrome when combined with SSRIs, SNRIs, MAOIs, or other serotonergic agents. GABA sold as a supplement crosses the blood–brain barrier poorly and has little credible direct evidence. Zinc is a cofactor in melatonin synthesis with only preliminary sleep data.

Cannabis and cannabinoids

Improving sleep is one of the most common reasons patients give for using cannabis, medically or recreationally, and clinicians now encounter it constantly — in dispensary products, in CBD gummies from a gas station, and in patients quietly substituting it for prescribed hypnotics. It deserves a careful, evidence-based treatment precisely because the marketing has run so far ahead of the data.

The cannabinoids and the endocannabinoid system

Cannabis contains over a hundred cannabinoids; three matter for sleep. Δ9-THC is the principal psychoactive constituent and a partial agonist at CB₁ receptors, which are dense in brain regions regulating sleep–wake state. CBD (cannabidiol) is non-intoxicating with complex, dose-dependent pharmacology (5-HT_1A agonism, indirect endocannabinoid effects, anxiolysis). CBN (cannabinol), a mildly psychoactive oxidative breakdown product of THC, is aggressively marketed as "the sleep cannabinoid" despite almost no controlled human data. The endocannabinoid system does modulate circadian and sleep–wake regulation, which gives biological plausibility to sleep effects — but plausibility is not proof.

THC and sleep architecture: the old story and the new evidence

The intuition that THC "knocks you out" comes largely from small studies in the 1970s–80s that used enormous doses (70–210 mg) and reported reduced sleep-onset latency, increased slow-wave sleep, and REM suppression. Those findings shaped a generation of assumptions. But a 2025 systematic review and meta-analysis of polysomnographic studies tells a more sober story: at the therapeutic doses actually used today (roughly 2.5–15 mg), cannabis administration produced no consistent objective effect on total sleep time, sleep-onset latency, REM, slow-wave sleep, or sleep efficiency. The dramatic REM suppression was largely an artifact of the old, supraphysiologic dosing. Tolerance to any initial sedative effect develops quickly with regular use.

CBD and CBN

CBD is dose-dependent and, for sleep, underwhelming as a direct agent. Low doses have shown, if anything, mild alerting effects in some studies; a nightly dose around 150 mg showed preliminary benefit in one small insomnia trial; and high-dose acute administration produced no measurable effect on the sleep–wake cycle in healthy volunteers. Where CBD helps sleep, it appears to do so largely through anxiolysis. CBN has essentially no rigorous standalone human evidence; the few positive signals come from formulations combining CBN with THC (not CBD), making it impossible to attribute benefit to CBN itself. The "sleep cannabinoid" branding is commercial, not clinical.

The most consistent finding: withdrawal disrupts sleep

Paradoxically, the clearest and most reproducible sleep effect of cannabis is what happens when a regular user stops. Cannabis withdrawal reliably produces reduced total sleep time, prolonged sleep-onset latency, vivid dreams, and REM rebound, often lasting days to weeks. In the 2025 meta-analysis these withdrawal effects were large and consistent — far more robust than any effect of administration. This is clinically decisive: sleep disruption during abstinence is a major driver of relapse and of the "I need it to sleep" belief. Patients who use nightly cannabis for sleep are frequently treating a withdrawal-driven insomnia of the drug's own making — a dependence loop analogous to rebound insomnia from sedative-hypnotics. Framing it this way, without judgment, opens the door to a supervised taper paired with CBT-I.

Sleep apnea, and other specific harms

Do not recommend cannabis for obstructive sleep apnea. The American Academy of Sleep Medicine has issued a position statement against the use of cannabis or synthetic cannabinoids (e.g., dronabinol) for OSA, citing unreliable delivery, absent long-term safety data, and inadequate evidence; sedation may also blunt arousal responses. Other harms to counsel on: next-day cognitive and psychomotor impairment and driving risk; dose-dependent anxiety and, at high doses, paranoia; precipitation or worsening of psychosis in vulnerable individuals; cannabinoid hyperemesis syndrome with chronic heavy use; risks in adolescents (neurodevelopmental) and in pregnancy/lactation (advised against); and the development of cannabis use disorder. Edibles deserve special caution — delayed, unpredictable onset invites re-dosing and overshoot, sometimes producing acute anxiety or cardiac symptoms hours later.

Clinical bottom line on cannabis

The evidence does not support cannabis as a reliable, direct treatment for insomnia, and it should not be recommended for sleep apnea. Where patients report benefit, the mechanism is usually indirect (anxiety, pain, PTSD), tolerance erodes any hypnotic effect, and nightly use sets up a withdrawal-driven insomnia that perpetuates the habit. For patients already using it, a non-judgmental conversation — clarifying the target symptom, checking CYP interactions, and offering CBT-I with a gradual taper — is far more useful than either endorsement or reflexive prohibition.

Special populations

Older adults: avoid first-generation antihistamines (Beers); prefer low-dose melatonin, magnesium (with renal check), and behavioral treatment; watch cumulative anticholinergic burden across the whole regimen. Pregnancy and lactation: most supplements and cannabis are inadequately studied or advised against; behavioral measures are first-line; doxylamine-pyridoxine has pregnancy safety data (for nausea) but sedation and next-day effects still apply. Children: behavioral intervention first; melatonin has the best pediatric evidence (notably in ASD) but demands accurate dosing and safe storage given the ingestion epidemic. Liver disease or alcohol use: avoid kava and be cautious with ashwagandha, valerian combinations, and high-dose CBD. Renal impairment: cap magnesium. Patients on serotonergic drugs: caution with tryptophan/5-HTP; on sedatives, opioids, or alcohol: counsel on additive CNS depression across every agent here.

A practical clinical approach

Ask directly and specifically about everything the patient takes for sleep, including supplements and cannabis, and reconcile it against the prescribed list for anticholinergic load, serotonergic overlap, CYP interactions, and hepatotoxic combinations. Anchor the plan in CBT-I and sleep-health behaviors, since no non-prescription agent substitutes for treating the underlying disorder. When a patient wants to try an OTC agent, steer toward the better-tolerated, better-evidenced options — low-dose melatonin for a circadian or sleep-onset problem, magnesium (renal function permitting), and time-limited use rather than a standing nightly habit — and away from routine first-generation antihistamines, kava, and nightly cannabis. Recommend third-party-verified products to blunt the content-variability problem. And in the older adult, remember that the safest sleep aid is often the deprescription of the ones already on board. The companion OTC Sleep Aid Comparison tool puts these agents side by side by mechanism, evidence grade, onset/duration, and key risks.

References & Further Reading

1. 2023 American Geriatrics Society Beers Criteria Update Expert Panel. American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults. J Am Geriatr Soc. 2023.
2. Gray SL, et al. Cumulative use of strong anticholinergics and incident dementia: a prospective cohort study. JAMA Intern Med. 2015;175(3):401–407.
3. Cohen PA, et al. Quantity of melatonin and CBD in melatonin gummies sold in the US. JAMA. 2023;329(16):1401–1402.
4. Lelak K, et al. Pediatric melatonin ingestions — United States, 2012–2021. MMWR Morb Mortal Wkly Rep. 2022.
5. Auld F, et al. Evidence for the efficacy of melatonin in the treatment of primary adult sleep disorders. Sleep Med Rev. 2017.
6. Suraev AS, et al. Cannabinoid therapies in the management of sleep disorders: a systematic review. Sleep Med Rev. 2020.
7. Cannabis and sleep architecture: a systematic review and meta-analysis. Sleep Med Rev. 2025.
8. Kaul M, et al. Using cannabis and CBD to sleep: an updated review. Curr Psychiatry Rep. 2024.
9. Ramar K, et al. Medical cannabis and the treatment of obstructive sleep apnea: an American Academy of Sleep Medicine position statement. J Clin Sleep Med. 2018;14(4):679–681.
10. Abbasi B, et al. The effect of magnesium supplementation on primary insomnia in elderly: a double-blind placebo-controlled clinical trial. J Res Med Sci. 2012.
11. Leach MJ, Page AT. Herbal medicine for insomnia: a systematic review and meta-analysis. Sleep Med Rev. 2015.
12. Teschke R, et al. Kava hepatotoxicity: a clinical review. Ann Hepatol. (and subsequent DILI case series on ashwagandha, Am J Gastroenterol/J Clin Exp Hepatol).

This chapter is an educational review for clinicians and trainees. Doses are typical ranges from the literature, not individualized recommendations; supplement content and quality vary and are not FDA-verified. Nothing here substitutes for clinical judgment, current product labeling, or treatment of the underlying sleep disorder.