Hill criteria
Reading File
Searching the Web
Bradford Hill criteria causality epidemiology
Finding Sources
Reading File
Reading File
Here is a complete breakdown of the Bradford Hill Criteria (1965), the nine guidelines for judging whether an observed association between an exposure and a disease is likely to be causal, drawn from authoritative medical textbooks.
Bradford Hill Criteria for Causation
First proposed by Sir Austin Bradford Hill in his 1965 paper "The Environment and Disease: Association or Causation?" (Proc R Soc Med. 1965;58:295-300), these criteria provide a framework for moving from association to causation in observational epidemiology.
| # | Criterion | Explanation |
|---|---|---|
| 1 | Strength of Association | The stronger the association (e.g., high relative risk or odds ratio), the more likely it is causal. A weak association is more easily explained by bias or confounding. |
| 2 | Consistency | The association is replicated across different populations, settings, and study designs. If multiple independent studies reach the same conclusion, chance is a less likely explanation. |
| 3 | Specificity | A single putative cause produces a single specific effect. This is the weakest criterion - many diseases have multiple causes. |
| 4 | Temporality | The exposure must precede the outcome (risk factor predates disease onset). This is the only strictly necessary criterion - without it, causation is impossible. |
| 5 | Biological Gradient | A dose-response relationship exists: increasing exposure increases disease risk, and reducing exposure reduces risk. |
| 6 | Plausibility | The association is biologically plausible - consistent with known mechanisms of disease. |
| 7 | Coherence | The association is compatible with existing theory and knowledge in the field. It does not seriously conflict with known facts about the natural history of the disease. |
| 8 | Experimental Evidence | The factor under investigation is modifiable by an appropriate experimental approach (e.g., removing the exposure reduces disease). Randomized trials or natural experiments support causality. |
| 9 | Analogy | An established cause-and-effect relationship exists for a similar exposure or disease, making the proposed association more credible by analogy. |
Key Points
- Temporality is the only obligatory criterion - all others are supporting evidence that increases the probability of causality.
- These are "viewpoints" or "considerations," not a rigid checklist. Hill himself said satisfying more criteria strengthens the case but none (except temporality) is individually sufficient.
- In complex multifactorial diseases (like CKD, cancer, cardiovascular disease), a risk factor may not fulfill all nine criteria yet still be genuinely causal.
- The criteria were originally applied to the smoking-lung cancer association, where they were convincingly met across all nine dimensions.
Mnemonic: "SCSTBPCAE" or remember the 9 as:
Strength - Consistency - Specificity - Temporality - Biological gradient - Plausibility - Coherence - Analogy - Experimental evidence
(Source: Brenner and Rector's The Kidney, adapted from Hill AB, 1965)
This is a shared conversation. Sign in to Orris to start your own chat.