Linear relationships in epidemiology describe a direct, proportional relationship between two variables. This means that as one variable increases or decreases, the other variable changes in a consistent manner. For instance, an increase in exposure to a risk factor might lead to a proportional increase in the incidence of a disease.