Supplement evidence tiers: from S-tier to one rat study

By Marcus Renshaw ·

Every supplement claim comes with an implied confidence level, and most people never check whether it is earned. “Studies show” covers everything from a meta-analysis of thousands of humans to a single experiment on twenty rats. The supplement industry loves this blur, because the weakest evidence and the strongest evidence get described in the same breathless language. The fix is a tier system. Before you spend money or swallow anything, ask what tier the evidence sits in, and price your confidence to match. Here is how I sort my own stack, from the things I would defend to a skeptic down to the ones that rest on a rodent and a hope.

Supplement evidence runs from S-tier, large human randomized trials, down to a single animal study extrapolated to people. Creatine and omega-3 sit near the top with hundreds of trials. Urolithin A is mid-tier: real human data, but small and short. Fadogia agrestis is bottom-tier, one 2005 rat study plus a toxicity signal. Match your confidence to the tier.

What separates S-tier from the bottom?

Four questions do most of the sorting. Was it tested in humans or only animals? Was it a randomized controlled trial or an anecdote? How many people, for how long? And who paid for it, the seller or an independent body? A supplement earns a high tier by clearing all four: repeated human RCTs, decent sample sizes, independent replication, and a plausible mechanism. It sits at the bottom when the entire case is one small animal study that a brand quietly scaled up to a human dose.

The trap is that tier and marketing are unrelated. The loudest claims often sit on the thinnest evidence, because a compound with strong data does not need hype and a compound with weak data has nothing else. So the volume of the claim tells you nothing. The tier does.

S-tier: what has real human evidence behind it

The top shelf is small and boring, which is how you know it is real. Creatine monohydrate is the clearest case: hundreds of human trials, a long safety record, and a well-mapped mechanism. It earns its place for strength, and, at a higher dose, has a growing second case for the brain, not just the biceps. The sports-nutrition position stand documents no kidney harm in healthy adults up to the highest doses studied. That is what S-tier looks like: you can hand a skeptic the evidence and it holds.

Omega-3 (EPA and DHA) is close behind for specific uses. The triglyceride-lowering effect is well established across randomized trials, with meta-analyses showing meaningful reductions at 1 to 4 grams a day. Worth noting honestly, its record on preventing cardiovascular events is genuinely mixed, with some large trials positive and others null, so even an S-tier ingredient is S-tier for a particular claim, not for everything a label prints. That distinction, strong evidence for this outcome, weak for that one, is most of what tiering teaches you.

The middle: promising but thin

Mid-tier is where honest people disagree, and where most of the interesting stuff lives. Urolithin A is a good example. There is real human data: a randomized trial reported about a 12% gain in muscle strength and improved muscle endurance versus placebo, with plausible mitochondrial mechanisms behind it. That is far better than most of the shelf.

But the trials are small, short, and several are tied to the company selling it, and the primary endpoint of peak power did not always hit. So it can be worth a look, but the confidence belongs at mid-tier: promising, not proven, until independent long-term data confirms the early signal. That is the correct posture for a whole class of compounds, the ones with a couple of decent human studies and a commercial interest attached. You can run them; just do not file them next to creatine.

The bottom: one rat study wearing a lab coat

Then there is the floor, and fadogia agrestis is the textbook case. Its reputation as a testosterone booster traces to essentially one 2005 rat study in which an aqueous stem extract raised testosterone in male rats over a few days. That is the entire foundation. There are no published human trials showing it works or that it is safe at the doses people take.

Worse, the follow-up rodent work found harm. The same research group reported that higher doses caused derangement of testicular tissue in rats, along with markers of liver and kidney toxicity. And the doses people actually swallow were never the doses tested, so buyers are extrapolating a short-term rat result to long-term human use, in the dark, with a toxicity signal blinking. This is the bottom tier made concrete: a single animal study, no human data, and a reason for active concern. It is exactly the kind of thing I will not put in my body, and exactly the kind of thing sold hardest.

How to read a supplement claim yourself

You do not need to memorize studies, you need a reflex. When you see a supplement claim, run the tier check before the credit card. Human or animal? Randomized or anecdotal? How big and how long? And follow the money on the studies. A retrospective association from real people (the way I treat nattokinase and arterial plaque) is stronger than a rat study but weaker than a randomized trial, and it deserves that middle confidence, no more.

This is the same discipline as deciding which supplements to drop, pointed at the buying decision instead of the quitting one. A bloated stack is usually bloated with bottom-tier ingredients that entered on a headline and never earned their place. Tiering on the way in keeps the stack honest so you have less to subtract later.

How I tier my own stack

My core is deliberately S-tier and dull: creatine, omega-3, a short list of things with real human trials behind them, held at doses inside the studied range. The mid-tier items I do run, like nattokinase, are in but flagged in my own head as provisional, kept only as long as I can tell they are doing something and the independent evidence keeps pointing the right way. The bottom tier, the fadogia end of the shelf, I skip entirely, because “one rat study and a toxicity signal” is not a bet, it is a dare. None of this is medical advice, and anything that touches your hormones or your labs deserves bloodwork and a clinician, not a supplement forum.

Pricing your confidence to the evidence is the actual skill here, and it only works if you also track what each thing does once it is in you. Start the logging habit on one cheap number and build from there; the free BP tracker is a zero-cost on-ramp for making that a habit before you ever spend on the exotic stuff.

FAQ

What makes a supplement high evidence versus low evidence?

Four things: whether it was tested in humans rather than only animals, whether the studies were randomized controlled trials rather than anecdotes, how many people were studied for how long, and whether the research was independent of the seller. High-tier supplements clear all four. Bottom-tier ones usually rest on a single small animal study scaled up to a human dose.

Is creatine actually proven to work?

Yes, it is one of the best-evidenced supplements available. Hundreds of human randomized trials support its effect on strength and lean mass, it has a long safety record with no evidence of kidney harm in healthy people up to high doses, and it has a clear mechanism. There is also a growing, separate case for a higher dose benefiting the brain.

Does urolithin A have human evidence?

Yes, but it is mid-tier. Randomized trials in humans show gains of roughly 12% in muscle strength and improvements in muscle endurance versus placebo, with plausible mitochondrial mechanisms. The trials are small and short, some are tied to the company selling it, and a primary power endpoint did not always hit, so it is promising rather than proven.

Is fadogia agrestis safe?

There is no good evidence that it is. Its reputation rests on a single 2005 rat study, and there are no published human trials for either effectiveness or safety at the doses people take. Follow-up rat studies reported testicular damage and markers of liver and kidney toxicity at higher doses. That combination, no human data plus a toxicity signal, is a reason for real caution.

How should I judge a “studies show” supplement claim?

Ask what the studies actually were. Human or animal? Randomized or observational? How large and how long? Independent or seller-funded? “Studies show” can mean a meta-analysis of thousands of people or one experiment on a few rats, and the marketing language is identical either way. The volume of the claim tells you nothing; the tier of the evidence tells you everything.

Are seller-funded supplement studies worthless?

Not worthless, but they deserve extra skepticism and independent confirmation before you upgrade your confidence. A single manufacturer-funded trial can be well run and still benefit from selective reporting or a friendly design. Treat it as a starting signal that needs replication by researchers with no financial stake, not as a settled result.

Sources