How phew thinks about sleep science
We're a supplement brand, which means we're allowed to make a lot of claims and not many of them have to be true. We took the opposite approach. Here's what we believe, why we believe it, and what would change our mind.
On dose. The strongest published evidence for melatonin's effect on sleep-onset latency is at 1 mg, per EFSA's authorized health claim (EFSA Journal 2011;9(6):2241). Your body produces 0.1–0.3 mg of melatonin per night on its own. Phew is at the 1 mg dose. Aisle products that ship 5–10 mg don't have a research-backed rationale for doing so — they're dosed for what feels strong at the shelf, not for what works in studies.
On botanicals. Valerian, lavender, chamomile, and hibiscus all have research bases of varying strength. Valerian and chamomile have the most evidence for sleep-related effects, both via GABAergic activity. Lavender has the strongest body of work for mild anxiety. Hibiscus has the lightest research footprint of the four; it's in the formula for flavor and mild antioxidant content, not for a specific sleep mechanism. We're honest about each in the ingredient pages.
What we don't claim. Phew does not treat insomnia. Phew does not treat anxiety. Phew is not a substitute for sleep hygiene, therapy, or medication. We don't use symptom-reduction language because we don't have the studies to back it. If we ever do, we'll say so with citations.
How we evaluate research. Three things we look for: randomized controlled trials over observational studies; doses similar to what's in the product; outcome measures that match the claim (sleep-onset latency for a sleep-onset claim, not “sleep quality” which is everything-and-nothing). If a study uses 10 mg in elderly cancer patients, we don't extrapolate to “1 mg helps healthy adults fall asleep.” Different dose, different population, different question.
What would change the formula. If a higher-quality trial showed 0.3 mg outperforms 1 mg in healthy adults, we'd dose down. If a botanical we use stopped meeting our standardization spec, we'd swap suppliers. If a meta-analysis suggested one of our botanicals doesn't move the needle at our dose, we'd consider replacing it. The formula isn't a contract; it's a current best answer.
Three tests for evaluating any supplement. When deciding whether to take something nightly: (1) Is the dose specific or hidden behind a proprietary blend? (2) Does the marketing claim something more specific than what the research actually shows? (3) Is the source disclosed, or is it “third-party tested” without naming the third party? Phew answers yes / no / working on it to these three. We'd answer them the same way for any product, ours or not.