Moderate

The sleep shelf is three metres wide. The evidence is not.

Eleven ingredients, two that hold up, and a gap between studied dose and sold dose that nobody puts on the label.

What's on this page

Walk into any pharmacy and the sleep shelf is three metres wide. The trial literature behind it is not.

What the trials actually measured

Most trials used a questionnaire rather than a sleep lab.1 That is a real finding about how people described their nights, and it is not the same as a finding about sleep architecture.

The magnesium picture is narrower still: the effect concentrates in people whose intake was low to begin with.2

Why the product may not be the product in the study

If a label does not name the form and the amount per serving, there is no basis for assuming the trial results transfer to it.

Where that leaves the grade

Two ingredients hold up. One is worth a conversation.3 The rest of the shelf is marketing.

References

  1. Placeholder A, Placeholder B. (2019). Melatonin and sleep onset latency: a meta-analysis. Journal of Placeholder Sleep Research
  2. Placeholder C. (2021). Magnesium supplementation in adults with low intake. Placeholder Nutrition Reviews
  3. Placeholder D, Placeholder E. (2020). L-theanine and subjective relaxation: a systematic review. Placeholder Journal of Nutrition

One email a week. Graded, not hyped.

What changed in the evidence, what got downgraded, and what quietly turned out to be nothing.

No spam, and we never sell your address.

Related

All guides
Limited1 min

Valerian root

Studied for decades with stubbornly inconsistent results. Extracts vary so much between manufacturers that trials are not really testing the same substance.

Moderate1 min

Glycine

A handful of small trials found faster sleep onset and better next-day ratings, plausibly through a drop in core temperature. Small, consistent, not yet repeated at scale.

None1 min

Proprietary sleep blends

A proprietary blend lists ingredients without amounts, so you cannot tell whether a studied dose is present or a trace. It is not that these fail — it is that they cannot be evaluated.