Structural differentiation of self-reported depression and anxiety in late life.

Research has shown impressive support for tripartite models of anxiety and depression that include a common factor of negative affect, and the unique factors positive affect and arousal. It is not clear whether this structure extends into later life. The current study used confirmatory factor analysis to model the structural relationship of anxiety and depression in two samples of older adults: a large probability sample (N = 1429) and a smaller convenience sample (N = 210). Across all analyses, a correlated, two-factor, psychometric model was most parsimonious. The tripartite model could be fit to the data, but added no explanatory power; in some cases a one-factor model also fit. The results suggest that there is a unitary factor of "distress" that incorporates anxiety and depression, but that the structure is not consistent with factor structures found in younger samples. Instead, the broad constructs may be represented in a more complex manner among older adults, and are less easily differentiated.