The Bi-factor model of repetitive negative thinking: Common vs. unique factors as predictors of depression and anxiety

Abstract

Background and objectivesDifferent forms of repetitive negative thinking (RNT) have traditionally been conceptualized as being distinctly linked to specific disorders. However, emerging evidence suggests that a common process lies at the core of different RNT manifestations. This common process might also largely explain the link between RNT and psychopathology. To examine the latent factor structure of RNT, we compared three structural-equation models, assuming (a) a common factor across different RNT measures (single-factor model); (b) scale-specific factors for each RNT measures (separate-factor model); and (c) both a common and scale-specific factors (bi-factor model). We additionally tested whether these latent factors predicted depression and anxiety at a follow-up time-point. MethodsA community sample (N = 523) completed an online assessment comprising measures of rumination, worry and content-independent RNT as well as depressive and anxiety symptoms at baseline (t1) and three months later (t2). ResultsThe bi-factor model showed the best fit to the data among the three models. Moreover, the common factor of the bi-factor model significantly predicted depression and anxiety three months later. Next to the common factor, some but not all scale-specific factors additionally predicted symptoms. LimitationsThe study was conducted in a non-clinical sample and the assessment of psychopathology was restricted to depressive and anxiety symptoms. ConclusionsThe findings support transdiagnostic conceptualizations of RNT, which highlight common aspects of different forms of RNT as well as the relevance of RNT across different diagnostic categories.