Ruminative response scale for eating disorders: bifactor model and measurement invariance in a Portuguese community sample

Abstract

The Ruminative Response Scale for Eating Disorders (RRS-ED) measures ruminative thought content specifically related to eating disordered themes, assessing two domains of rumination, brooding and reflection. This study aims to examine the factor structure of the RRS-ED in a Portuguese community sample, using correlated two-factor models, unifactorial and bifactor models and test for invariance across sex. A sample of 535 adults (179 male; 356 female) filled out the RRS-ED. A subsample (n=347) answered additional measures of repetitive negative thinking and eating psychopathology. The bifactor model of the RRS-ED provided the best fit, demonstrating a reliable general rumination factor. Also, the bifactor model of the RRS-ED was invariant across sex. RRS-ED showed moderate to strong correlations with negative perseverative thinking and eating psychopathology. Both domain-specific factors of RRS-ED were associated with higher levels of eating psychopathology. Findings indicate that RRS-ED is a reliable and valid measure to assess the ruminative response from the general population in Portugal, showing initial evidence that supports the use of a total score of RRS-ED as an overall measure of rumination, while specific factor scores should be reported with caution. Future studies are needed to replicate the findings and further corroborate the unidimensionality of the RRS-ED.