Skip to main content
. 2016 Dec 3;184(11):847–855. doi: 10.1093/aje/kww112

Figure 1.

Figure 1.

Causal diagram illustrating number of health-care encounters as a confounder of the observed relationship between diabetes and depression. A) Basic causal model, where number of health-care encounters can serve as a confounder of the proposed (dotted line) diabetes-depression relationship. B) Causal model in which number of encounters may serve as a proxy for general illness.