Measurement Invariance of Disorder-Specific and Transdiagnostic Measures of Repetitive Negative Thinking

Repetitive negative thinking is conceptualized to be a transdiagnostic process linked to the development and maintenance of psychopathology. Prior research distinguishes between disorder-specific exemplars (worry, rumination) and transdiagnostic measures of repetitive negative thinking with differences across disorders reported. However, establishing the measurement invariance of these measures is necessary to support meaningful comparisons across clinical groups.

Bayesian structural equation modelling was used to assess the approximate invariance of the Ruminative Response Scale, Penn State Worry Questionnaire, and the Repetitive Thinking Questionnaire across individuals with a principal diagnosis of either depressive disorder, social anxiety disorder, or generalized anxiety disorder.

All scales demonstrated approximate measurement invariance across the three disorder groups. The depressive disorder group reported a higher level of rumination than the generalized anxiety disorder group (Δµ = 0.25, 95% Credibility Interval [0.06, 0.45]), with no difference between the generalized anxiety disorder and social anxiety disorder groups. The depressive disorder and generalized anxiety disorder groups did not differ in their levels of trait repetitive negative thinking, but the social anxiety disorder group was markedly lower than the generalized anxiety disorder group (Δµ = -0.21 [-0.37, -0.05]). Similarly, levels of worry did not differ between the generalized anxiety disorder and depressive disorder group but were lower in the social anxiety disorder group than the generalized anxiety disorder group (Δµ = -0.23 [-0.41, -0.06]).

The Ruminative Response Scale, Penn State Worry Questionnaire, and Repetitive Thinking Questionnaire are measuring trait repetitive negative thinking in a consistent manner across individuals with a principal diagnosis of depressive disorder, social anxiety disorder, or generalized anxiety disorder. This supports their use in transdiagnostic contexts and indicates that it is appropriate to directly compare the scores on these measures between diagnostic groups.