The sleep stack on trial: magnesium threonate, theanine and apigenin against placebo
See more of CureMed in your Google results.
Three popular bedtime supplements were tested together against dummy pills for the first time. In healthy people who already slept fairly well, they made no measurable difference to sleep in two weeks.
In its first placebo-controlled trial, the bedtime stack of magnesium L-threonate, L-theanine and apigenin did no better than placebo on sleep: not on the questionnaires and not on a sleep-tracking ring, over two weeks in 103 healthy adults. The group that improved took ashwagandha and rhodiola in the morning. The trial was funded by the company that made the supplements, and it did not include people with insomnia.
- What was tested: 145 mg of magnesium (as 2 g magnesium L-threonate), 200 mg L-theanine and 50 mg apigenin every evening, matching the doses listed in Huberman Lab's 2021 sleep toolkit.
- Who: 103 healthy, physically active adults (average age 35) in four groups. Double-blind, placebo-controlled, two weeks of supplements.
- Bedtime stack against placebo: no significant difference in sleep disturbance (p = 0.48), in the Pittsburgh sleep score, or in the ring's sleep quality score, sleep time, time to fall asleep and sleep efficiency.
- The morning blend (ashwagandha 600 mg + rhodiola 300 mg) lowered perceived stress (p = 0.034) and sleep-related impairment at week two (p = 0.032).
- Read it fairly: healthy sleepers, about 25 per group, two weeks, gaps in the ring data, and a funder that sells the product.
- If you already sleep well, this stack has not earned its price. If you have insomnia, this trial does not answer your question.
CureMed ranks on published human evidence, and screens drug interactions first.
What was tested
The bedtime stack, dose by dose
| Ingredient | Dose in the trial | Dose in Huberman Lab's sleep toolkit (2021) |
|---|---|---|
| Magnesium L-threonate | 2 g, giving 145 mg of magnesium | 145 mg magnesium threonate (or 200 mg magnesium bisglycinate) |
| L-theanine | 200 mg | 100–400 mg |
| Apigenin | 50 mg | 50 mg |
The trial also tested a morning blend, 600 mg of ashwagandha with 300 mg of rhodiola, on its own and together with the bedtime stack. That is why there were four groups: morning blend only, bedtime stack only, both, and placebo for both.
The trial at a glance
| What the paper reports | |
|---|---|
| Design | Randomised, double-blind, placebo-controlled, four groups |
| People | 115 randomised, 103 finished; 53 women; average age 35; healthy and physically active |
| Group sizes | Both: 25 · morning blend: 26 · bedtime stack: 25 · placebo: 27 |
| Length | One baseline week, two weeks of supplements, one week after stopping |
| Sleep measures | Two questionnaires (PROMIS, Pittsburgh Sleep Quality Index) and a sleep-tracking ring worn every night |
| Funding | Momentous LLC, which produced the supplements, and a US National Institutes of Health training grant |
What it found
Each supplement against placebo
| Outcome | Bedtime stack | Morning blend |
|---|---|---|
| Perceived stress | No effect (p = 0.68) | Lower (p = 0.034) |
| Sleep disturbance (questionnaire) | No effect (p = 0.48) | No significant effect (p = 0.09) |
| Sleep-related impairment (questionnaire) | No effect reported | Lower at week two (p = 0.032) |
| Pittsburgh sleep score, overall | No difference between groups | No difference between groups |
| Ring: sleep quality score, sleep time, time to fall asleep, sleep efficiency | No effect | No effect |
| Ring: time awake during the night | No improvement reported in the results | Lower (p = 0.047) |
On every sleep measure, the bedtime stack was indistinguishable from the dummy pills. The authors put it in one line: morning-only supplementation was enough to produce improvements, "whereas evening-only supplementation was insufficient".
Taking both was no better than the morning blend alone. The authors suggest that lowering daytime stress may have done more for sleep than pills aimed at sleep itself.
How to read it fairly
- These were healthy sleepers. The authors say low sleep disturbance at the start may have left little room to improve. The trial says nothing about people with insomnia.
- It was short and small: two weeks of supplements, about 25 people per group.
- The ring data had gaps. The authors report technical problems with the device, which weakens that part of the result.
- The funder made the products. Momentous LLC paid for the trial and produced both supplements. The result favoured one of its products and not the other.
- It is one trial, and the first of this combination. The authors note that the three had not been studied together before.
What to do with this
- If you already sleep well, skip the stack. Its first controlled test found nothing a placebo did not also deliver.
- If you take it and feel it helps, that fits this trial too: several measures improved over time in every group, placebo included, which the authors call consistent with the placebo response.
- If you have real insomnia, a supplement is not the first step. Clinical guidelines put cognitive behavioural therapy for insomnia first.
- Do not read this as a green light for ashwagandha. Two people in the morning group stopped because of side effects (diarrhoea in one, dizziness and muscle weakness in the other), and the US National Center for Complementary and Integrative Health advises against it in pregnancy and breastfeeding, before surgery, and with thyroid or autoimmune disorders. It can interact with thyroid, diabetes, blood-pressure and sedative medicines, and rare cases of liver injury have been reported.
- Add one thing at a time, for a problem you actually have, and check it against what you already take.
Frequently asked questions
Does the Huberman sleep stack work?
Not in its first placebo-controlled trial. In 103 healthy adults, two weeks of magnesium L-threonate (145 mg of magnesium), L-theanine (200 mg) and apigenin (50 mg) at bedtime did no better than placebo on sleep questionnaires or on a sleep-tracking ring. The trial did not include people with insomnia, so it cannot say what the stack does for them.
Does apigenin help you sleep?
No placebo-controlled trial of apigenin on its own for sleep turned up in our PubMed search on 6 October 2026. It was one of three ingredients in the bedtime stack tested in 2026, and that stack did not beat placebo on any sleep measure over two weeks in healthy adults. Chamomile extracts, which contain apigenin, have their own trials, which this page does not cover.
Does magnesium threonate help you sleep?
In the 2026 trial it was given with theanine and apigenin, and the combination did not improve sleep compared with placebo in healthy adults. The trial did not test magnesium threonate on its own.
Who paid for the sleep stack trial?
Momentous LLC, which also produced the supplements used in it, together with a US National Institutes of Health training grant for one of the authors. The authors reported no conflict of interest.
What did improve sleep in the trial?
The morning blend of ashwagandha (600 mg) and rhodiola (300 mg). That group reported lower perceived stress and less sleep-related impairment at week two, and the ring recorded less time awake during the night. Adding the bedtime stack on top did not improve on it.
Is ashwagandha safe to take for sleep?
It may be safe in the short term (up to three months), according to the US National Center for Complementary and Integrative Health, but it is not risk-free. In this trial two people taking it stopped because of side effects. The same agency advises against it in pregnancy and breastfeeding, before surgery, and with thyroid or autoimmune disorders, notes interactions with thyroid, diabetes, blood-pressure and sedative medicines, and reports rare cases of liver injury.
What should I try first for insomnia?
For insomnia that lasts, clinical guidelines put cognitive behavioural therapy for insomnia first, ahead of any pill. CureMed ranks the over-the-counter sleep aids by their trial evidence on a separate page.
Keep reading
- Best sleep aids, ranked by evidence.
Melatonin, magnesium, glycine and the rest, graded on randomised trials.
- How we grade evidence
Why one small trial moves a grade less than people expect.
- Free stack check
Screen your supplements against everything else you take.
More in Supplements
- The best vitamins for longevity, ranked by human trial evidence
Vitamin D if your level is low, B12 if you are at risk, magnesium for most. Graded only on human randomised trials, including the two that showed harm.
- Best sleep aids, ranked by evidence.
Diphenhydramine stops working in about three nights. Melatonin is modest for insomnia, strong for jet lag. Every OTC sleep aid ranked on its human trials.
- Which OTC brands pharmacists recommend most — and where the evidence disagrees
Pharmacists name the same OTC brands every year. For immune support and joint health, the trials behind the two biggest sellers do not back how people use them.
- OTC cold and flu remedies, ranked by evidence
The FDA says the decongestant in most cold products does not work. Zinc works only above a dose most lozenges miss. Every OTC cold remedy ranked on its trials.
- PCC1 and the grape-seed longevity pill: what the study actually showed
PCC1 extended mouse lifespan by 9.4% — by injection, starting in old age. Here is the full study, the human data that exists, and what the capsules leave out.
- Taurine and longevity: what the human evidence actually shows
The study that made taurine famous extended lifespan in mice, not humans. What the actual human trials show, the dose studied, and who should be careful.