Animal study extrapolation applies findings from laboratory animals to predict human health effects.

Attributable risk is the portion of disease in an exposed population that can be attributed to the exposure.

Background risk is the rate at which a disease occurs in the general population without the exposure at issue.

The Bradford Hill criteria are nine considerations used to evaluate whether an observed statistical association reflects a causal relationship.

A case-control study compares exposure histories between people with a disease and matched people without it.

A cohort study follows exposed and unexposed groups forward to compare disease incidence between them.

A confidence interval expresses the range of values statistically consistent with a study's data, indicating estimate precision.

A dose-response relationship exists where greater exposure produces greater incidence or severity of disease, a central indicator of causation.

Epidemiology is the study of disease patterns in populations, and it supplies the primary scientific evidence of general causation.

In vitro evidence comes from studies on cells or tissues outside a living organism, showing biological mechanism.

A latency period is the interval between exposure and disease manifestation, used to assess causal plausibility.

A meta-analysis statistically combines results from multiple studies to produce a pooled effect estimate.

An odds ratio compares the odds of exposure among diseased and non-diseased groups, used where incidence cannot be measured directly.

Relative risk is the ratio of disease incidence in an exposed population to incidence in an unexposed population.

Statistical significance indicates that an observed association is unlikely to have arisen from chance alone.

A systematic review comprehensively identifies and appraises all studies addressing a question using a predefined protocol.

A threshold dose is the exposure level below which no adverse effect occurs, a contested concept for carcinogens.

Weight of the evidence methodology integrates epidemiological, animal, and mechanistic data to reach a causation conclusion.