Adult Depression
Is the Concept of Melancholic Depression Relevant in the 21st Century: A psychometric analysis of the Sydney Melancholia Prototype Index (SMPI)
John F. Buss, None
Ph.D. Candidate
Indiana University
Indianapolis, Indiana, United States
Lorenzo Lorenzo-Luaces, Ph.D.
Associate Professor
Indiana University
Bloomington, Indiana, United States
Introduction: The relegation of melancholic depression to a specifier status since the Diagnostic and Statistical Manual of Mental Disorders 3rd edition (DSM-III) has been heavily criticized. The Sydney Melancholia Prototype Index Self Report (SMPI-SR) was developed to revive the pre-DSM-III construct of melancholic depression. Despite the popular usage of the SMPI for assessing melancholic depression, few studies have analyzed the SMPI's factor structure.
Objective: A previous study indicated a five-factor for the SMPI-SR. Our objectives were to confirm the five-factor solution of the SMPI and assess the convergent validity of the five factors.
Methods: We recruited 420 participants from an online sample. We used confirmatory factor analyses to analyze the 5-factor structure of the SMPI. Correlations between SMPI factors and related measures were explored. Additionally, we used confirmatory latent profile analyses to determine if assigned classes indicate the existence of a melancholic profile within the sample, contributing to the discussion on the categorical nature of melancholia as a depression state.
Results: Confirmatory factor analysis did not support the existence of the five-factor structure (X^2 = 3188.079, RMSEA = 0.073; 95% CI = .067 - .079, CFI = 0.85, TLI = 0.83). However, we found moderate (i.e., 0.4 -0.7) correlations between the five factors and the measures selected as validators. Confirmatory latent profile analysis did not indicate the existence of a melancholic profile.
Conclusion: Converging evidence from our analyses suggests that the Sydney Melancholia Prototype Index (SMPI) does not distinguish constructs of melancholic and non-melancholic depression. Response patterns indicate variations in clinical severity rather than categorically defined subtypes. Future research should explore alternative endophenotypes for depression, such as general trait neuroticism and elevated early morning cortisol levels.