How to read a supplement study without falling for the headline

A headline promises that a new ingredient raised strength by 15 per cent. Three paragraphs later you are on a checkout page. The question almost nobody asks first is: 15 per cent compared with what, in whom, and after how many weeks?

Nowhere in fitness is the gap between what research shows and what a label claims as wide as it is with supplements. Researchers are not the problem. The problem is that one sentence from a paper's conclusion travels easily onto a tub, and from there into a headline.

This is not a list of supplements to buy. It is a method for judging the next claim you read on your own.

What we know reliably

Supplements work best when they fill a genuine gap. The evidence here is consistent and has been repeated for decades across different populations and methods. Correct a real shortfall and you see a clear effect. Add the same thing to someone who already has enough and you usually see nothing. More than enough is not better than enough.

The list of ingredients with strong evidence for training is short, and it barely changes. Creatine monohydrate and caffeine have by far the most independently replicated trials behind them, alongside protein, which is a nutrient rather than a magic powder. We went through who makes that list and why in the piece on supplements with the strongest evidence. Everything outside it rests on thinner ground — which is not the same as saying it does nothing.

Protein works through your daily total, not through the shake. For people lifting regularly, the evidence converges on roughly 1.6 to 2.2 grams per kilogram of bodyweight per day. Powder is simply a convenient way to reach that number when food is not going down.

A single supplement trial is almost always small. Typical studies run 10 to 30 participants for 4 to 12 weeks. Small groups produce noisy results, and effect estimates from small studies run high on average. That pattern is among the best-documented findings in research methodology, and it holds regardless of the topic.

Whoever pays tends to get the prettier result. Manufacturer-funded trials report favourable outcomes more often than independent ones. The pattern shows up across many fields and is consistent. It does not make any given study wrong; it means you should want confirmation from somewhere else before you act.

A cell in a dish and a rat in a cage are not a person under a barbell. A large share of claims about fat burning and recovery come from laboratory models that were never tested on training humans at all.

What is likely, but not settled

Dose and form. Outside creatine and caffeine, the dose-response curve for most ingredients is poorly mapped. Often one dose gets tested, it becomes the standard by repetition, and the question is never revisited.

Duration. Most trials finish inside three months. Whether the benefit persists over a year, whether the body adapts, whether cycling makes any difference — that is thinly researched for almost everything except the oldest ingredients.

Individual response. It is plausible that some people respond much better than others. Our ability to identify those people in advance, from genetics or a blood panel, remains weak. The same caution applies to the effects people attribute to age; we covered what actually shifts in the piece on training after 60.

Timing. Whether you take something 30 minutes before or two hours after a session is probably minor next to what you eat across the day. The evidence leans that way without closing the question.

Combinations. Pre-workout blends with ten ingredients are almost never tested as a whole product. Individual components were tested, at other doses, in other people. Adding them up is not evidence.

The middle tier. Beta-alanine for efforts lasting roughly one to ten minutes, and dietary nitrate from beetroot for endurance, sit on moderate evidence, though the benefit looks smaller in well-trained athletes. Collagen for tendons is interesting and early. Electrolytes are their own case, and the actual state of that evidence is worth reading properly.

What gets reported badly

• „Statistically significant" does not mean „large". It means the difference was unlikely to be chance. Half a kilogram on a bench press can be significant and useless.
• Relative numbers sound bigger than absolute ones. Fifteen per cent of something small is still small.
• „No evidence it works" is not „evidence it doesn't". The first often just means nobody has run a decent trial.
• One study is not science. Science is what survives replication in other labs. How to spot the difference in news coverage is the subject of our piece on headlines about new research.
• The dose in the study is often not the dose in the tub. Proprietary blends do not disclose individual amounts, so no comparison is possible.
• A surrogate marker is not an outcome. A number in your bloodwork is not a heavier lift or a faster 5K.
• „Natural" is not „harmless". Plant extracts can raise heart rate and blood pressure and interact with medication.

What this means for your training and diet

Before supplements are worth a thought, the base has to hold: enough sleep, daily protein and calories matched to the goal, two to four serious sessions a week, and load that climbs over months. The best-supported supplements deliver something in the range of one to a few per cent. That margin only matters once everything underneath it is in order — a point that also runs through the evidence on strength and longevity.

When a headline lands, ask five questions: how many people, for how long, compared with what, how big the difference was in absolute terms, and who paid. If the article answers none of them, you do not have enough to decide anything.

If you still want to try something from the weaker tier, run it as an experiment. One ingredient at a time, 8 to 12 weeks, the same programme throughout, kilograms and reps written down. Without a log you are measuring your mood, not the supplement. And do not judge it by how sore you are the next day — post-training soreness is a poor guide to progress.

If you compete under anti-doping rules, choose products with independent batch testing. Contamination of supplements with banned substances is a documented problem, and the responsibility always falls on the athlete.

When to see a doctor

Get checked if you have symptoms that persist: lasting fatigue, dizziness, a racing heart, swelling, yellowing skin or dark urine. Blood work is the only honest way to establish whether you are actually short of something. Guessing from symptoms usually misses.

Speak to a doctor before taking anything if you are on prescription medication, have kidney or liver disease or high blood pressure, or are pregnant or breastfeeding. The same goes for high doses of fat-soluble vitamins and for the stimulant load in pre-workout formulas.

In short

• Supplements help most when they correct a real shortfall — the strongest and most consistent finding in the field.
• The list of ingredients with solid evidence for training is short and moves slowly.
• Small trials overstate effects, and industry-funded ones come out favourable more often.
• A percentage without an absolute number tells you nothing usable.
• Dose, duration, individual response and ingredient combinations remain thinly researched.
• Sleep, protein, calories and progression decide the outcome; a supplement is the last per cent, not the first step.