In a Cox regression, the target variable is the time until the occurrence of an event (for example, illness or death). Censored data are available, i.e. not all participants could be observed for the entire duration of the study. As a result, a Cox regression provides the hazard ratio (HR) calculated using a proportional hazards model. The basic assumption is that the risk in the two groups differs by a certain factor. The interpretation of HR and RR (relative risk) is similar. In the Million Women Study, in addition to HRT use, other influencing factors such as age were taken into account.