Physical disease is commonly considered a risk factor for depression among older adults. However, this pattern is not consistently supported, and a theoretical framework for understanding such a relationship has not been articulated. P.M. Lewinsohn, H. Hoberman, L. Teri, and M. Hautzinger's (1985) integrative model of depression predicts that disease will be a risk factor for depression only when disease results in functional impairment, and that impairment in the absence of disease is also a risk factor for depression. The authors tested these predictions in a community-based sample of older adults followed longitudinally and found that functional impairment was a significant risk factor for depression, regardless of disease status. Disease was not a significant predictor of major depression, nor did it interact with impairment to predict depression.