Challenges with Latent Variable Approaches to Operationalizing Dimensions of Childhood adversity - a Commentary on Sisitsky et al. (2023)

Abstract

Childhood adversity is a common, powerful risk factor for future mental and physical health problems. Appropriately aggregating and categorizing discrete adverse experiences into cumulative indices remains an active area of research and debate in the field of developmental psychopathology. In a recent article by Sisitsky et al. (2013), they report using confirmatory factor analysis and latent profile analysis to categorize participants into "profiles" of childhood adversity. In this commentary, we argue that categorizing dimensions or profiles of childhood adversity based on the tendency of adversities to co-occur is misaligned with dimensional models of adversity, which derive underlying dimensions from their tendency to cumulatively predict certain outcomes rather than adversity co-occurrence. We summarize the history of theories and methods of operationalizing childhood adversity that led to the development of the dimensional model of adversity and psychopathology. We then explain why latent variable approaches, which simply reflect the tendency of adverse experiences to co-occur, are inappropriate for validating any given approach.

Keywords: Childhood adversity; Deprivation; Dimensional model; Factor analysis; Threat.