Randomized controlled trial (RCT)

A randomized controlled trial is the strongest ordinary way to prove that a treatment actually works — that it causes an effect rather than just happening to appear alongside one. It has two features baked into the name.

Controlled means there are at least two groups: one gets the compound, another gets a placebo or the standard treatment for comparison. Without that control group, you’d never know whether people improved because of the drug or would have improved anyway.

Randomized means people are sorted into those groups by chance, like a coin flip. That’s the clever part. Random assignment spreads out all the hidden differences — age, health, lifestyle, luck — evenly across both groups, so at the end the main thing separating them is the treatment itself. Often it’s also “double-blind,” meaning neither the patient nor the researcher knows who got what, which stops expectations from coloring the results.

Put together, an RCT is designed to strip away the ways we fool ourselves. It’s why the RCT sits at the top of the evidence ladder, above observational studies and animal work.

Take semaglutide and tirzepatide as two concrete examples. In an RCT, researchers randomly assign participants to receive one of these compounds or a comparison such as a placebo, then measure the same predefined clinical trial endpoint in every group. If the groups differ at the end, randomization and the control make it much more credible that the compound caused the difference instead of age, starting health, expectations, or another hidden factor.

The honest catch for peptides: many popular ones have few or no human RCTs yet, resting instead on animal or observational data. That doesn’t make them worthless — but “backed by an RCT” and “shown in mice” are very different claims, and the gap is worth respecting.