Reading the evidence: a survival guide for health claims

Why this skill matters most

Health information is a flood, and much of it is wrong, exaggerated, or selling something. Being able to gauge the weight of a claim is arguably the most valuable wellness skill there is, because it protects you from both wasting money on the useless and being harmed by the risky. This page is the lens we try to apply to every other topic on the site, and it is one you can apply yourself.

The core idea is simple: confidence is not evidence. A bold testimonial, a slick site, or a single dramatic study can all feel persuasive while resting on very little. What matters is whether multiple, well-designed studies in actual humans point the same way, and whether the people making the claim have an interest in your believing it.

A quick hierarchy of evidence

Not all evidence carries equal weight. Near the bottom sit anecdotes and testimonials, then laboratory and animal studies, which can suggest ideas but often do not translate to people. Above those are observational human studies, which can reveal associations but struggle to prove cause. Stronger still are randomized controlled trials, which test a single variable, and at the top are systematic reviews and meta-analyses that pool many good trials.

A practical rule follows from this. A single small study, especially one in cells or animals or with a tiny number of people, is a hint, not a conclusion. Be far more confident when several well-designed human trials agree and a careful review summarizes them in the same direction. Headlines tend to do the opposite, trumpeting one preliminary study as if it settled the matter.

Correlation, causation, and common traps

The most frequent error in health reporting is treating correlation as causation. Two things moving together does not mean one causes the other; a third factor may drive both, or the direction may be reversed. Observational findings that people who do X are healthier are interesting but cannot, on their own, prove that X makes you healthy. Good evidence accounts for this; hype ignores it.

Other traps recur often enough to memorize. Cherry-picked studies that ignore contradicting ones. Surrogate outcomes, a change in a lab number, presented as if it were a real health benefit. Tiny effects inflated by dramatic language. And the absence of a control group, which makes it impossible to know whether something worked or people would have improved anyway. Spotting these turns most marketing claims transparent.

Red flags and a sensible default

Certain signals reliably mark claims to distrust. Promises that something cures many unrelated conditions. Reliance on testimonials instead of data. Secret or proprietary formulas. Attacks on mainstream medicine paired with a product to sell. Urgency and scarcity tactics. And undisclosed conflicts of interest, where the person citing the evidence profits from your belief. Any one of these warrants real skepticism; several together are close to a verdict.

When the evidence is genuinely unclear, which is often, a reasonable default is caution proportional to risk: gentle, low-risk, low-cost options can be reasonable to try with modest expectations, while anything expensive, risky, or used in place of proven care demands much stronger proof. And for anything that depends on your health specifics, the most reliable move is still to bring the claim to a clinician who can weigh it against your situation. Knowing what you do not know is part of reading evidence well.