[Latent structure of depression in the elderly: a taxometric analysis].

This study aimed to investigate the latent structure of depression with a non-clinical sample of elderly, using taxometric analysis. The latter consists of a family of statistical procedures conceived for testing whether a given psychological construct is best represented by categories or dimensions in which individuals vary. The sample consisted of 570 elderly with a mean age of 71.9 years (SD = 7.45), who answered the Brazilian version of the Geriatric Depression Scale, a cognitive test, and demographic questions. Three taxometric procedures were used: Mean Above Minus Below A Cut (MAMBAC), Maximum Eigenvalue (MAXEIG), and Latent Mode (L-mode). Sets of simulated categorical and dimensional data, along with the comparison curve fit indices (CCFI), oriented the study data's interpretation. The results with the three techniques pointed to a better fit with the dimensional format as opposed to the taxonic one, that is, depression represented better as a syndrome in which subjects are distributed along a continuum rather than as a discrete diagnostic entity.