Like most of the epidemiological studies we need to have control data to determine what "unusual" meant. However, we can’t do the comparison among individuals since everyone is different. Rather, the comparison should be done within the individual. In another words, we will compare the individual’s activity before the disease onset to his/her usual activity. Thus here, we have a "case" component and also a "control" component, but the information of both components will come from the same individual.