Scientific Claims Review T·M
What the label says.
What the evidence says.
A line-by-line audit of a best-selling sleep supplement — five claims taken straight from the product page, each checked against the studies it leans on, and rewritten into a sentence the science can actually hold.
Your science is the best thing about your product. My job is to make sure a customer can trust every sentence that surrounds it.
Prepared by
Teodora Malešević
Scientific Brand Strategist · MSc Molecular Biology
Supplements · Health · Biotech
No. 01
Sample Work
The Method01 — Approach
How I read a claim

Most health copy doesn't fail because the science is weak. It fails because the sentence claims more than the study earned. A real effect in older adults becomes “clinically proven” for everyone. A result from a petri dish becomes a promise about your night's sleep. The distance between those two things is exactly where a brand earns trust or quietly loses it.

Before I judge a single line, I put it through four questions.

i
What kind of study is this? A claim inherits the strength of its evidence. A meta-analysis of randomized trials can carry a strong verb. A mouse study cannot — no matter how promising the mechanism.
ii
Who was it tested on? An effect measured in magnesium-deficient older adults does not automatically transfer to a healthy thirty-year-old. Population is not a footnote; it's half the claim.
iii
How big was the effect, really? “Statistically significant” is not “meaningful.” Seven minutes is a real number and a small one. Both facts belong in the sentence.
iv
Does the dose match the bottle? A benefit shown at one dose says nothing about a bottle holding a fifth of it — or five times it. The label has to earn its own numbers.
The evidence ladder
In vitro
& animal
→
Observational
→
Single
RCT
→
Meta-analysis /
systematic review
Where a claim sits on this ladder decides how strong a verb it is allowed to use.
The goal is never to weaken a claim. It's to find the strongest sentence the evidence can actually hold.
Teodora MaleševićScientific Claims Review · 02
The Findings02 — Claims 01–05
01
Magnesium
bisglycinate
Evidence for claim
Overstated
The claim
“Clinically proven to improve your sleep quality.”
There is real evidence here — but “clinically proven” outruns it. The most-cited meta-analysis pooled just three randomized trials in 151 older adults and found magnesium shortened the time to fall asleep by about 17 minutes — while rating its own evidence low quality [Mah & Pitre, 2021 · BMC Complement. Med. Ther.]. A 2025 placebo-controlled trial in poor sleepers saw only a small drop in insomnia severity, and mostly in people who began with low dietary magnesium [RCT, 2025 · Nature & Science of Sleep]. The effect is modest, concentrated in the deficient, and unproven in the healthy and well-nourished.
What I'd write instead
“If your magnesium runs low — and for many people it does — topping it up may help you fall asleep a little faster. Most studied in adults with low levels or disrupted sleep.”
02
Form ·
Glycinate
Evidence for claim
Unsupported
The claim
“Magnesium glycinate — the most bioavailable form there is.”
Human studies support a weaker, true version: organic forms like glycinate and citrate absorb better than cheap magnesium oxide. They do not crown a single winner — where head-to-head data exist, citrate performs at least as well. The precise figures that circulate in supplement copy (“80–90% absorption”) trace to repetition, not measurement. Glycinate's genuine advantage isn't a bioavailability crown; it's tolerability — gentle on the stomach, without the laxative effect of oxide.
What I'd write instead
“We use magnesium glycinate for a simple reason: your gut absorbs it well and tolerates it easily — none of the stomach upset that comes with cheaper forms.”
Teodora MaleševićScientific Claims Review · 03
The Findings02 — Claims 01–05
03
Melatonin
Evidence for claim
True, but oversold
The claim
“Fall asleep faster, naturally.”
Directionally true, quietly small. Across 19 randomized trials in 1,683 people, melatonin cut the time to fall asleep by about 7 minutes on average [Ferracioli-Oda et al., 2013 · PLoS ONE]. It works best not as a sedative but as a clock-setter — for jet lag or a shifted schedule — and at low doses of 0.3–0.5 mg. Many bottles pack 3–10 mg: more than the evidence calls for, and enough to leave some people groggy the next morning.
What I'd write instead
“Melatonin is your body's night-time signal. In studies it helped people drift off about seven minutes sooner — most useful when your sleep clock is off, and effective at low doses.”
04
L-Theanine
Evidence for claim
Supported (as calm)
The claim
“Melts away stress for deeper, more restorative sleep.”
The first half is well-earned; the second reaches. In a four-week randomized trial, 200 mg of L-theanine a day lowered stress and anxiety scores and improved self-rated sleep [Hidese et al., 2019 · Nutrients]. But the measured effect is about perceived calm, not sedation, and rests on self-report rather than objective sleep recordings. It's an honest relaxation claim wearing a sleep-claim costume.
What I'd write instead
“L-theanine, the calming compound in green tea. In a controlled trial, 200 mg a day helped people feel less stressed and wind down more easily — the mental off-switch, not a knockout.”
Teodora MaleševićScientific Claims Review · 04
The Findings02 — Claims 01–05
05
Apigenin ·
Chamomile
Evidence for claim
Preclinical
The claim
“Apigenin — the clinically studied compound that helps you drift off.”
“Clinically studied” is carrying weight the human data can't yet support. Apigenin's sedative story comes largely from preclinical work — rodent studies showing it binds the same receptors as anti-anxiety drugs. Human sleep trials test chamomile extract, not isolated apigenin, and results are thin: a placebo-controlled pilot in 34 adults with insomnia found no significant change in sleep-diary measures [Zick et al., 2011 · BMC Complement. Altern. Med.]. A promising mechanism — not yet a clinical sleep claim.
What I'd write instead
“Apigenin, the compound behind chamomile's long reputation for calm. The research is still early — mostly laboratory work — so we include it for that traditional role, not a proven sleep effect.”
The pattern underneath
Notice what these five fixes share: not one throws the ingredient away. Every rewrite is still a selling sentence — just one the studies can stand behind. Overstatement isn't only a compliance risk; it's a trust tax. A reader who catches one inflated claim quietly discounts the other four. Precision reads as confidence — and in a category built on trust, confidence is the whole product.
Teodora MaleševićScientific Claims Review · 05
The Verdict03 — Summary
Five claims, scored.
№The claim, in shortIngredientEvidenceVerdict
01Improves sleep qualityMagnesium Overstated
02Most bioavailable formGlycinate Unsupported
03Fall asleep fasterMelatonin True, oversold
04Melts away stressL-Theanine Supported (calm)
05Helps you drift offApigenin Preclinical
What I'd fix first
i
Retire two words. “Clinically proven” and “most bioavailable” are the claims an informed customer — or a regulator — challenges first. Replacing them costs nothing and buys credibility.
ii
Right-size the melatonin. Make the honest low dose the feature: “a gentle 0.5 mg — the dose the research actually supports” outsells a silent 5 mg.
iii
Separate calm from sleep. Let L-theanine own relaxation and melatonin own timing. Two precise claims persuade better than one blurry one.
This is one page of one product. Send me yours — a product page, a label, or the deck you're about to raise on — and I'll reply with the first three things I'd change.
Teodora Malešević
Scientific Brand Strategist · Supplements · Health · Biotech
Product anonymized. Claims are representative of language used across leading sleep supplements. Full citations available on request.