Vascular depression: A distinct diagnostic subtype?

Abstract

Vascular depression has been proposed as a unique and valid diagnostic subtype on the basis of studies of external (concurrent and predictive) validity. Validating a diagnostic entity on the basis of external validity is problematic, because it presupposes that the construct is well defined (i.e., the proposed features cluster together to define a distinct patient group). Because such evidence has not been obtained, we propose that the next critical step in evaluating this potential subtype is to establish internal (construct) validity and highlight taxometric analysis and latent class cluster analysis as illustrative multivariate statistical techniques that can be used in this effort. The psychometric approach advocated here (despite its inherent assumptions and limitations) might substantially improve on previous diagnostic efforts (e.g., expert consensus), and vascular depression might serve as a prototype for future psychiatric classification.