What cells and mice are actually for

These sit on the second floor of the pyramid, and they earn their place — just not for the job marketing gives them. In-vitro ("in glass") studies test a compound on cells, enzymes or tissue in a dish. Animal studies test it in a living organism, usually mice or rats. They're indispensable for the questions that come before a human trial:

  • Mechanismhow might this molecule work? What does it bind, block or switch on?
  • Safety screening — is it wildly toxic before we ever give it to a person?
  • Dose-finding — roughly what amount does anything at all?

What they are not is proof that a supplement helps a human. And the gap between "worked in a dish" and "works in you" is enormous.

Why the dish lies to you

  • Impossible doses. Cells are often bathed in a concentration of the compound you could never achieve in your blood — sometimes you'd have to drink kilograms. "Kills cancer cells at 50 µM" is meaningless if your bloodstream tops out at 0.5 µM.
  • Absorption (the big one). A molecule can be a superstar in a dish and simply never get into your body — poorly absorbed in the gut, or destroyed by the liver before it reaches your tissues. The dish skips your entire digestive system.
  • Species differences. Mice aren't tiny humans. Countless compounds that cured disease in mice did nothing — or harmed — in people.
  • Isolated vs whole. A cell in a plastic well has no immune system, no other organs, no feedback loops. Bodies are systems; dishes are not.

Put those filters in a row and you get a brutal funnel — the reason "promising in preclinical studies" so rarely becomes "works":

Attrition funnel from in-vitro hits to compounds that help patients Works in a dish (in-vitro) thousands of hits Still works in mice hundreds Absorbed & safe in humans a few dozen Actually helps patients (RCT) a handful
Most "breakthroughs" die on the way down. Each step — surviving a real organism, getting absorbed, being safe, then actually helping in a trial — kills off most candidates. A label quoting the top bar is describing a lottery ticket, not a winner. (Proportions are illustrative of a well-known pattern in drug and nutrient development.)
“…in mice.”

There's a running joke among scientists — an account that simply adds the missing words "IN MICE" to breathless health headlines. It exists because the omission is so common. When you read a stunning result, silently add "in a dish" or "in mice" and see if the excitement survives.

Worked example Curcumin: the in-vitro superstar that barely absorbs

Curcumin (from turmeric) is the poster child. In the dish it's genuinely dazzling — anti-inflammatory, anti-cancer, antioxidant activity across hundreds of in-vitro papers. It's the most-studied "miracle" compound of the last two decades.

Then reality: plain curcumin has very low oral bioavailability — it's poorly absorbed, rapidly metabolised, and quickly cleared, so almost none reaches your bloodstream intact. The spectacular dish concentrations are simply unreachable by eating turmeric or swallowing a capsule. Human RCTs have been far more muted than the lab hype, and the interesting work is now all about formulations that fix the absorption problem — which is a different, humbler claim than "curcumin cures things". A label showing you a wall of in-vitro citations is telling you about the dish, not about you.

None of this means preclinical science is worthless — it's where every real breakthrough is born. It means a supplement resting only on cell and animal data is a hypothesis, not a result. Enjoy the mechanism story; then ask the one question that separates the funnel's top from its bottom: did it work in humans who swallowed it?

What to remember

“Worked in a dish” is a hypothesis, not a result.

  • Cells and mice are for mechanism, safety and dose-finding — the questions before a human trial, not proof a supplement helps you.
  • Impossible doses and poor absorption break most translations. A dish skips your gut and liver entirely.
  • Silently add “in a dish” or “in mice” to any dramatic claim. Then ask the only question that matters: did it work in humans who swallowed it?
Try it · ~1 min

Add the missing words.

  1. Find an "[ingredient] fights/kills/boosts…" claim on a product or in a headline.
  2. Check whether the study behind it was a dish, an animal, or humans.
  3. Re-read the claim with "in a dish" or "in mice" appended. Does it still sound like a reason to buy?

References

1
Nelson KM, et al. The essential medicinal chemistry of curcumin. J Med Chem. 2017;60(5):1620–1637 — curcumin's poor bioavailability and instability. doi:10.1021/acs.jmedchem.6b00975
2
Seok J, et al. Genomic responses in mouse models poorly mimic human inflammatory diseases. PNAS. 2013;110(9):3507–3512 — limits of animal-to-human translation. doi:10.1073/pnas.1222878110

Progress saves automatically in this browser · skip ahead anytime · no streak.