An observed dose-response relationship strengthens a causal hypothesis. Dosage and duration are often interchanged in study design, and both may cause a gradient in disease frequency. People with longer exposure time may have a greater time period in which to develop the disease.