Critères de causalité: critères de Bradford Hill - II

The biological gradient, also called the dose-response relationship, indicates that increasing exposure should lead to increased risk of the effect. For example, the dustier a workplace, the higher the risk of developing dust allergies. Similarly, higher levels of cigarette smoke inhalation correspond to greater lung cancer risk, demonstrating a clear dose-response pattern.

The plausibility criterion requires that the association between cause and effect must align with current biological knowledge. For instance, the carcinogenic effects of tobacco are biologically plausible given the known toxic constituents in tobacco and their effects on cells. This criterion ensures proposed causal relationships are grounded in established biological understanding.

The coherence criterion states that the proposed cause-and-effect relationship should not contradict established knowledge about the disease's biology and natural progression. For example, the link between HPV infection and cervical cancer aligns coherently with the known progression from infection to cancerous lesions, supporting the causal claim.

The experiment criterion suggests examining causal relationships through appropriate experimentation, though not all causal relationships can be tested this way. Controlled trials, such as those showing decreased lung cancer rates following smoking cessation, illustrate how intervention studies can clarify and support causal links in epidemiological research.

The analogy criterion allows attributing an effect to a cause by analogy with established causal relationships. For example, once thalidomide's teratogenic effects were known, scientists suspected other drugs might also cause congenital disabilities during pregnancy. Similarly, when certain chemicals were known to disrupt the endocrine system, researchers explored whether other substances had similar effects.

Multiple criteria provide a robust framework for evaluating causation across various epidemiological studies. While not definitive or exhaustive, these criteria guide researchers through the complexities of establishing causality in public health investigations. Using multiple criteria together strengthens confidence in causal claims beyond what any single criterion could achieve.

The biological gradient or dose-response relationship is quantitative, showing that increased exposure produces increased effect risk. Other criteria like plausibility and coherence are qualitative, assessing whether causal claims align with biological knowledge. Together, dose-response provides measurable evidence while other criteria ensure the proposed mechanism is scientifically reasonable.