(PS9-B30) Worry and Rumination Differ in Their Prediction of Generalized Anxiety Disorder and Major Depressive Disorder

Background: Worry and rumination are prominent cognitive features of emotional disorders (Harvey et al., 2004). Early theories linked worry with anxiety, and rumination with depression (Nelson & Mazure, 1985; Wells & Matthews, 1994). Recent work, however, has revealed overlapping characteristics and shared outcomes of worry and rumination (e.g., Ehring & Watkins, 2008; Olatunji et al., 2013), raising the question of whether these processes discriminate anxiety from depression or serve as equivalent risk factors for both. To shed new light on this question, we: (a) measured worry, rumination, anxiety, and depression in the same sample; (b) examined longitudinal as well as concurrent associations, to evaluate the predictive power of worry and rumination; and (c) compared the magnitude of these associations using formal tests. To provide a particularly conservative test of discriminability, we operationalized anxiety and depression using generalized anxiety disorder (GAD) and major depressive disorder (MDD), the two disorders with which worry and rumination are most strongly associated (Ehring & Watkins, 2008).

Method: A total of 172 community-dwelling adults were diagnosed with GAD and/or MDD (n = 121) or no psychopathology (n = 51) using the Anxiety and Related Disorders Interview Schedule for DSM-5 (Brown & Barlow, 2014). All participants were assigned separate clinical severity ratings (0–8) for GAD and MDD. Worry and rumination were measured using the Penn State Worry Questionnaire (Meyer et al., 1990) and Ruminative Responses Scale (Treynor et al., 2003), respectively. One year later, all available participants who had been diagnosed with GAD/MDD (n = 92) were re-assessed for GAD and MDD severity. We used Steiger’s t test to test for significant differences between dependent correlations (Cohen & Cohen, 1983; Steiger, 1980) at baseline and across the one-year follow-up.

Results: Worry and rumination each shared large, concurrent correlations and more moderate, prospective correlations with clinician-rated GAD and MDD severity. Tests of differences between correlations revealed that worry was more strongly associated with GAD severity than with MDD severity, both cross-sectionally (r = .74 vs. .51) and prospectively (r = .46 vs. .20), ts > 2.77, ps < .006. By contrast, rumination shared similar associations with GAD and MDD severity, both cross-sectionally (r = .68 vs. .71) and prospectively (r = .30 vs. .32), ts < 0.72, ps > .475.

Conclusion: Worry is associated more strongly with GAD than with MDD. This pattern is evident not only cross-sectionally, as might be expected given worry’s role as a symptom of GAD, but also when worry is used to predict symptoms and functional impairment a year later. These results suggest that worry may serve as a risk or maintaining factor for GAD, and that treatments aimed at reducing cognitive vulnerability for anxiety should preferentially target worry. Conversely, contrary to the conventional theoretical pairing of rumination with depression, rumination did not show greater specificity to MDD relative to GAD, suggesting that rumination may make similar contributions to both disorders. Implications for risk assessment, targeted treatment, and theories of anxiety and depression will be discussed.