Establishing causality in epidemiology involves several criteria, often referred to as the Bradford Hill criteria. These include:
1. Strength of Association: Strong associations are more likely to be causal. 2. Consistency: The association is observed in different populations and settings. 3. Specificity: A specific population and disease are associated with the exposure. 4. Temporality: The exposure precedes the disease. 5. Biological Gradient: A dose-response relationship exists. 6. Plausibility: The association is biologically plausible. 7. Coherence: The association is consistent with existing knowledge. 8. Experiment: Experimental evidence supports the association. 9. Analogy: Similar associations have been observed with other exposures and outcomes.