A man with visible abdominal muscles told me last week that seed oils were the reason Western populations are unhealthy. He said it with total conviction, in a well-lit kitchen, holding a jar of something. The video had 400,000 likes.
He might be right. I doubt it, but he might be. The problem is that I have absolutely no way of knowing, and neither does he, and neither do the 400,000 people who liked it — and the mechanism by which he arrived at that conclusion had nothing to do with evidence and everything to do with the fact that it makes an extremely good sixty-second video.
This is the actual problem with fitness information in 2026. It isn’t that there’s too little. It’s that there’s an enormous surplus, all of it delivered with the same confident tone, and no obvious way to sort the reliable from the ruinous if you don’t already know the answer. Which, if you’re asking the question, you don’t.
I’ve spent an embarrassing number of hours on this — first as a consumer, then as someone who got annoyed enough to start checking claims properly. I have no credentials. What I do have is a set of heuristics that have saved me a lot of wasted years, and I think they’re more useful than any specific piece of advice I could give you.
Start here: nearly everyone is sincere
It’s tempting to assume the internet is full of grifters. Some of it is. But the more useful model is that most people giving bad fitness advice genuinely believe it, and that’s precisely what makes it dangerous.
Here’s how it happens. Someone in their twenties, with good genetics and no injuries, tries a thing. They get in shape — as almost anyone in their twenties will, doing almost anything, consistently. They conclude that the thing worked. They then teach the thing with the passionate certainty of personal experience, and their audience, who are largely forty and injured and have different lives entirely, get nothing from it and conclude they must be broken.
That’s not fraud. It’s just an extremely small sample size being mistaken for a discovery. Almost all fitness advice is somebody generalising confidently from a sample of one.
Once you see that pattern you can’t unsee it, and it makes a lot of contradictory advice suddenly comprehensible. They’re not all lying. They’re all describing what happened to them.
Heuristic one: be suspicious of anything that identifies a single villain
Sugar. Carbs. Fat. Gluten. Seed oils. Cardio. Sitting. Cortisol.
Every few years a new single cause of everything gets nominated, gets a book deal, gets a documentary, and then gets quietly replaced. The pattern is so regular you could set a calendar by it.
The reason single-villain theories are so popular isn’t that they’re true. It’s that they’re satisfying. Human bodies are complicated and outcomes are multi-causal and none of that fits into a video, whereas “here’s the one thing making you fat” fits perfectly and makes people feel like they’ve been let in on something.
Real answers in this field are almost always boring and plural. Why are people less healthy than they were in 1975? Because food is more available and more engineered to be over-eaten, and jobs stopped being physical, and cars replaced walking, and screens replaced everything else, and portions grew, and sleep shrank, and stress rose. That’s the answer. It’s true, it’s dull, and nobody can sell you a supplement for it.
When someone offers you one cause, they’re offering you a story, not a finding.
Heuristic two: check what they’re selling before you check what they’re saying
Not because selling things is wrong — everybody has to eat — but because it tells you which direction the errors will lean.
Someone selling a twelve-week programme has an incentive to make you believe twelve weeks is meaningful. Someone selling supplements has an incentive to convince you that food is insufficient. Someone selling a coaching package has an incentive to make training seem too complicated to do alone. Someone selling a book about a single villain needs that villain to be terrifying.
None of them will consciously distort anything. They don’t need to. Incentives don’t work by making people lie; they work by making certain conclusions feel more obviously correct than others, and by quietly determining which studies get remembered and which get forgotten.
So the useful question isn’t “is this person honest?” It’s “if they were subtly wrong, in which direction would they be wrong?” — and then discount accordingly.
Heuristic three: distrust confidence in proportion to how new the claim is
There is a rough hierarchy of certainty in exercise science, and it’s worth knowing where things sit
At the very solid end: resistance training builds strength and muscle. Cardiovascular exercise improves cardiovascular health. Energy balance determines body mass over time. Sleep is required for recovery. Protein supports muscle repair. Progressive overload works. These are not controversial and they haven’t been for decades.
In the middle: optimal training frequency, ideal rep ranges, meal timing, the relative merits of different programme structures. There’s real research here, the effects are usually small, and reasonable experts disagree.
At the speculative end: most of what’s currently trending. The specific supplement of the moment. Cold exposure protocols. Whichever hormone is being blamed this year. Most of it rests on small studies, animal models, or mechanistic reasoning that hasn’t been tested in actual humans doing actual training.
The tell is that confidence tends to run inversely to certainty. Nobody makes a passionate video about progressive overload, because it’s settled and boring. The passion clusters at the speculative end, where the evidence is thinnest and the claims are newest, because that’s where there’s still room for a novel take.
