The Bradford Hill criteria are a set of principles used to evaluate whether an observed association is likely to be causal.
- These criteria provide supportive but not definitive evidence of causality.
- They are frequently tested in epidemiology and biostatistics questions on the USMLE Step 1.
Overview
- Purpose: Help distinguish causation from mere association in observational studies.
- Key point: None of the criteria alone proves causation, but taken together, they strengthen causal inference.
The Nine Criteria
| Criterion | Definition / Key Point | High-Yield Tip |
|---|---|---|
| Strength | Stronger associations provide more evidence for causality. | But strength alone ≠ causation. |
| Consistency | Findings are observed repeatedly in different populations & settings. | Replication = reliability. |
| Specificity | A specific cause leads to a specific effect. | Rare in medicine, but strong when present. |
| Temporality | The cause must precede the effect. | Only mandatory criterion. |
| Biological Gradient | A dose–response relationship exists. | ↑ exposure → ↑ effect. |
| Plausibility | A biological mechanism can explain the association. | Depends on current scientific knowledge. |
| Coherence | Findings fit with known facts about disease & biology. | Should not contradict established science. |
| Experiment | Evidence from interventions (eg, animal studies, RCTs) supports the link. | Strong supportive evidence. |
| Analogy | If similar exposures cause similar effects, inference is stronger. | “By comparison” reasoning. |
Learning Objective:
By the end of this review, you should be able to list and explain the Bradford Hill criteria, recognize that temporality is essential, and apply these principles to determine whether an observed association may be causal in epidemiological studies.
