The evidence ladder
Not all 'studies' carry the same weight. Research climbs a ladder, and each rung is a bigger test of real-world usefulness:
1. In vitro ('in glass') — cells, enzymes, or tissues in a dish. Cheap and fast, great for exploring mechanisms.
2. Animal — mice, rats, or other species. Adds a living body, but one that differs from ours in size, metabolism, and lifespan.
3. Human — observational studies, then randomized controlled trials. The only rung that directly tests whether something helps people [1].
Why dish-and-mouse results often fade
A compound that does something striking to cells in a dish may never reach the same concentration in a living person, may be broken down before it acts, or may cause effects that only appear in a whole body. Animal results translate to humans more often than in vitro ones, but still fail frequently — doses used in rodents are often far higher, per kilogram, than anyone would take.
'Studies show' — show what, exactly?
When a label or article says 'studies show,' ask which rung of the ladder it means:
- A petri-dish antioxidant result is a hypothesis, not a health benefit.
- A mouse study suggesting an effect is a reason to run human trials, not a reason to buy.
- A large, well-controlled human trial — ideally repeated — is what justifies confidence.
NCCIH's guidance on reading research makes the same point: preliminary laboratory or animal findings are not evidence that something works in people [1].
A quick mental filter
Before acting on an exciting claim, find the study type. If the headline rests on cells or animals, treat it as 'interesting, unproven.' Save your confidence for consistent human evidence [2].