Factor Structure, Internal Consistency, and Measurement Invariance of the Eating Pathology Symptoms Inventory (EPSI) in Transgender and Gender-Diverse Adults.

Abstract

Objective: This study aimed to examine the psychometric properties, including factor structure, internal consistency, and measurement invariance of the Eating Pathology Symptoms Inventory (EPSI) in a large national sample of transgender and gender-diverse (TGD) adults.

Method: The sample consisted of 2098 TGD adults-including transgender men (n = 599), transgender women (n = 293), and gender-diverse individuals (n = 1206)-who completed online self-report surveys. Using a subset of the sample, exploratory factor analysis (EFA) was conducted to identify underlying factor structures, followed by confirmatory factor analysis (CFA) to confirm the model fit within each population in the remainder of the sample. Multi-group confirmatory factor analysis (MG-CFA) was used to assess measurement invariance across transgender men, transgender women, and gender-diverse individuals.

Results: The EPSI's eight-factor structure was supported across all three TGD groups with strong model fit: transgender men (CFI = 0.985, RMSEA = 0.047, SRMR = 0.078), transgender women (CFI = 0.948, RMSEA = 0.048, SRMR = 0.085), and gender-diverse individuals (CFI = 0.962, RMSEA = 0.040, SRMR = 0.060). Measurement invariance analyses supported that the EPSI was invariant across the groups, facilitating meaningful group comparisons using the inventory. Internal consistency, assessed using McDonald's omega, was acceptable for all subscales (ω = 0.74-0.94).

Discussion: This study provides psychometric support for the EPSI in TGD groups. The results indicate that the EPSI performs reliably and consistently across these groups for assessing eating pathology among gender minority people.