Observational studyCohort study · Epidemiological study
A study that watches what people already do: powerful for spotting patterns, unable to prove cause on its own.
An observational study measures what happens without assigning anyone to anything. Researchers record exposures such as diet, activity or a biomarker, follow people over time or compare groups, and look for associations with outcomes.
The main designs differ in how they handle time. Cohort studies follow people forward, which is stronger. Case control studies start from an outcome and look backwards. Cross sectional studies capture one moment and cannot establish which came first.
They matter because most nutrition and lifestyle questions cannot ethically or practically be randomised. Nobody can assign people to smoke for thirty years, and few trials can afford to feed thousands of people a controlled diet for a decade, so cohorts carry most of the evidence on how people actually live.
Their weakness is confounding. People who eat more fibre also tend to exercise more, smoke less and have more money, and statistical adjustment can only account for what was measured and measured well. Residual confounding is the reason so many observational findings fail when finally tested in a trial.
Reverse causation is the other trap: early illness changes behaviour, so low cholesterol or low body weight can look like a cause of poor outcomes when it is an early consequence of them.
What strengthens the causal reading is a large effect, a dose response gradient, consistency across different populations and designs, biological plausibility, and agreement with trials or Mendelian randomisation where those exist. Smoking and physical activity clear that bar comfortably.
In practice, treat associated with as exactly that, ask how large the effect was in absolute terms, and expect our evidence tiers to rank a well conducted cohort below a randomised trial for the same claim.