Revisiting the factor structure of the EURO-D, a depressive symptomatology scale: An exploratory graph analysis in 26 European countries

Despite the widely use and multiple validations of the EURO-D scale, its factor structure is still under debate. Exploratory Graph Analysis (EGA), a novel network psychometric method, offers a promising approach to examining dimensionality. Methodology: 45,390 participants (mean age = 71.27, 57.4% women) from 26 European countries. The sample was randomly split into a derivation sample (n = 22,823) and a cross-validation sample (n = 22,567). EGA was applied to the derivation sample to determine the structure of the EURO-D scale, utilizing two estimation methods: Graphical Least Absolute Shrinkage and Selection Operator (GLASSO) and Triangulated Maximally Filtered Graph (TMFG). The identified factor structures were then tested via Confirmatory Factor Analysis (CFA) in the cross-validation sample for model fit. Results: EGA consistently revealed a two-factor structure with minor differences in the placement of suicidality and fatigue items across estimation methods. CFA results confirmed an adequate model fit for both solutions. Conclusion: This study combines exploratory (EGA) and confirmatory (CFA) approaches, supporting a two-factor structure for the EU-RO-D scale with alternative placements for fatigue and suicidality items. Results are discussed in contrast to previous studies reporting two and three-factor solutions with different assignments of these items.