Evidence quality
Confounding
When another factor partly or fully explains an apparent association.
What it means
Definition
Confounding: Confounding occurs when groups differ in ways that affect the outcome. It is a major limitation of observational evidence.
What it can tell you
It can help separate an observed association from a tested causal claim.
What it cannot tell you
It cannot remove confounding or prove causation without stronger study design and analysis.
Headline trap
Why this matters in headlines
People who take supplements may also exercise more, eat differently, or be wealthier, which can confound associations.