Epidemiologic investigations designed to test a hypothesized cause-effect relation by modifying the supposed causal factor (s) in the study population.