Sleep & Anxiety Supplements
Magnesium glycinate and L-theanine are small, low-risk nudges for sleep, though L-theanine does nothing proven for anxiety. Melatonin earns a narrow body-clock role, but new observational data warns against taking it nightly for years, and most other "natural" sleep aids are oversold.
This is the supplement-pillar quick reference. For the full evidence base, mechanisms, dose recommendations, and the addiction question for each of magnesium (glycinate and L-threonate), L-theanine, melatonin, apigenin, ashwagandha, glycine, phosphatidylserine, taurine, valerian, and CBD — see the dedicated Sleep supplements page in the Sleep section.
Quick summary
| Supplement | Evidence | Safety | Addiction risk | Verdict |
|---|---|---|---|---|
| Magnesium glycinate (200–400 mg) | Weak | Excellent | None | Worth trying |
| L-theanine (200–450 mg) | Weak (sleep, self-rated only); none shown (anxiety) | Excellent | None | Worth trying for sleep |
| Magnesium L-threonate (cognition) | Weak (cognition and sleep) — manufacturer-linked trials only | Good | None | Optional; count the magnesium |
| Melatonin (0.5–5 mg for jet lag; 0.3–1 mg to shift the clock) | Moderate (narrow circadian); Weak (general insomnia) | Good short-term; cardiovascular caution long-term | None | Case-specific, not nightly forever |
| Ashwagandha (300–600 mg, root) | Modest, low-certainty (stress/anxiety); suggestive (sleep) | Moderate (liver injury; interactions) | None documented | Selective use |
| Apigenin (~50 mg) | Weak/preliminary | Good | None | Optional |
| Glycine (3 g) | Weak | Excellent | None | Optional |
| Phosphatidylserine (100–400 mg) | No sleep or anxiety trials in adults | Good | None | Skip |
| Valerian | No demonstrated efficacy | Good short-term; long-term unknown | None documented (untested) | Skip |
| Taurine | Not a sleep aid | Good | None | Not for sleep |
| GABA (oral) | Weak/negative | Good | None | Skip |
| CBD | Preliminary | Variable (drug interactions) | Low | Skip absent indication |
On the addiction question, specifically
None of the supplements above produces clinically significant addiction, tolerance, or withdrawal in the doses studied. This is a key advantage over prescription sleep medications:
- Benzodiazepines, Z-drugs, over-the-counter (OTC) antihistamines — benzodiazepines and Z-drugs carry meaningful tolerance, dependence, and withdrawal risks, antihistamines lose their sedating effect within days,[1] and all three carry a falls-and-dementia signal in older adults (the standard avoid-list for older people names them from age 65).
- Dual orexin-receptor antagonists (suvorexant, lemborexant, daridorexant) — newer prescription class showing no tolerance, withdrawal or rebound when stopped. They are still Schedule IV controlled substances, because recreational sedative users report measurable drug-liking — more than placebo, though for daridorexant at its prescribed dose less than with an above-normal dose of zolpidem.
That said:
- "No addiction" doesn't mean "use forever without thinking." A psychological dependence ("I can't sleep without it") can develop with any sleep aid. The strongest sustainable approach is foundational sleep hygiene plus cognitive behavioural therapy for insomnia (CBT-I) if needed, with supplements as occasional adjuncts.
- Long-term safety data (>1–2 years continuous daily use) is limited for most of these. Magnesium has the longest clean track record; melatonin's long-term picture now carries a cardiovascular caveat (see below).
The melatonin caveat worth knowing
A 2025 observational cohort presented at the American Heart Association linked a year or more of melatonin use to higher rates of heart failure and death. It is observational and confounded — heavy long-term users tend to have worse insomnia and more co-morbidity — so it cannot prove cause. But it is enough to favour using melatonin strategically (0.5–5 mg of the immediate-release form for jet lag, a small 0.3–1 mg dose to shift the clock) rather than as an open-ended nightly habit. Going above about 5 mg adds nothing, and even 3 mg can spill into the next morning. Full detail in Sleep supplements.
What to combine, what to avoid
Reasonable combinations
- Magnesium glycinate + L-theanine (evening, 1–2 h before bed) — common stack, both safe, complementary mechanisms (magnesium dampens the brain's main excitatory signal and supports its main inhibitory one, while L-theanine promotes calm, relaxed-alertness brain-wave activity).
- Magnesium glycinate + melatonin, dosed to the purpose, for travel or shift transition — fine, short-term.
Be cautious with
- Cannabidiol (CBD) + benzodiazepines / Z-drugs / selective serotonin reuptake inhibitors (SSRIs) / statins / warfarin — CBD inhibits CYP3A4 and CYP2C9, two major drug-metabolising liver enzymes.
- Ashwagandha + thyroid medications, immunosuppressants, sedatives — can amplify effects; it also modulates CYP3A4.
- Multiple sedatives layered — even "natural" sedatives can have additive effects with each other and with alcohol.
Avoid layering with prescription sleep aids
If you're on a prescription sleep medication, talk with your prescriber before adding any of these supplements. Most are safe, but the conservative approach is to use one strategy at a time.
When supplements aren't enough
Persistent sleep problems (insomnia ≥3 nights/week for ≥3 months) call for:
- CBT-I as first-line — it produces large, durable effects that no supplement matches. See Insomnia treatment.
- Screening for obstructive sleep apnoea (OSA) if any risk factor — see Sleep breathing.
- Evaluation for underlying mood, pain, or medical conditions.
Supplements are adjuncts, not solutions, for chronic insomnia. The same is true of anxiety: ashwagandha has one of the better-studied short-term anxiety signals among supplements, though the certainty is low. An oral lavender-oil capsule, Silexan, rests on larger, more consistent trials: five trials and about 1,200 people with anxiety disorders or milder, below-threshold anxiety, scored on a clinician-rated scale. Nearly all were run by its manufacturer,[2] but an independent comparison of herbal remedies found a similar effect, and none for L-theanine.[3] The evidence covers that one product, not lavender in general. Either way, the autonomic-retraining levers (slow breathing, mindfulness, exercise, social connection) are far higher-leverage long-term. See Stress.