Abstract
OBJECTIVE: Although the relative risk from a prospective cohort study is numerically approximate to the odds ratio from a case-control study for a low-probability event, a definite relationship between case-control and cohort studies cannot be confirmed. In this study, we established a different model to determine the relationship between case-control and cohort studies. METHODS: Two analysis models (the cross-sectional model and multiple pathogenic factor model) were established. Incidences in both the exposure group and the nonexposure group in a cohort study were compared with the frequency of the observed factor in each group (diseased and nondiseased) in a case-control study. RESULTS: The relationship between the results of a case-control study and a cohort study is as follows: Pe=(Pd∗m)/(Pc∗m)/(Pd∗m)/(Pn=(m)/(∗PdPc∗m)/(Pd∗m)/(Pe and Pn represent the incidence in the exposed group and nonexposed group, respectively, from the cohort study, while Pd and Pc represent the observed frequencies in the disease group and the control group, respectively, for the case-control study; finally, m)/(. CONCLUSIONS: There is a definite relationship between the results of case-control and cohort studies assessing the same exposure. The outcomes of case-control studies can be translated into cohort study data.