(PS11-C62) Reciprocal Relations Between Symptoms of Depression and Anxiety: Who’s on First?

Introduction: Depressive and anxiety disorders are among the most common psychological disorders. The disorders tend to be highly comorbid with lifetime prevalence estimates of depression being 50-60% among those with an anxiety disorder. Stressful life events increase risk for these disorders. The stress of the COVID-19 pandemic and its various associated stressors has been marked by a substantial increase in these problems. Studying symptoms over the course of the pandemic offers a unique opportunity to assess how depressive and anxiety symptoms may influence one another over time. To study this relationship, we used a random intercept cross-lagged panel model (RI-CLPM) to examine such reciprocal relationships while modeling both autoregressive effects and parsing within- and between-person components.

Method: Participants were 1,200 U.S. adults who completed a series of seven online assessments over a 7.5-month period early in the COVID-19 pandemic. Depressive symptoms were rated using the Quick Inventory of Depressive Symptomatology – Self-Report (QIDS) and anxiety symptoms were rated using the Generalized Anxiety Disorder Scale – 7 (GAD). The first four assessments occurred every two-weeks and the following three occurred monthly. Intraclass correlation coefficients (ICC) were calculated to estimate the proportion of between-person variance for each measure. Reciprocal relationships were examined with RI-CPLM in MPlus, disaggregating the between- and within-person variation in symptoms and modelling change in symptoms over time.

Results: At baseline, moderate or greater symptom severity was reported by 45.8% of participants for anxiety (GAD score > 10) and 35.3% of participants for depression (QIDS score > 11). ICCs showed high between-person variance for GAD (ICC = 0.82) and QIDS (ICC = 0.80). The RI-CLPM of anxiety and depression symptoms achieved a strong fit. Within-person depressive symptoms significantly predicted subsequent anxiety with a median standard estimate of 0.053. Within-person anxiety significantly predicted subsequent depressive symptoms with a median standard estimate of 0.059.

Conclusion: Using a random intercept cross-lagged panel model, our results supported a reciprocal and roughly comparable predictive relationship between anxiety and depressive symptoms and vice versa. The within-person effects were obtained even in the context of largely between-person variability in both measures. Our results suggest a somewhat more important role for depressive symptoms predicting anxiety symptoms, than previous research examining symptoms or disorders that has largely not focused on within-person relationships. Although the optimal timing of assessments is not known, there is reason to believe that relatively short intervals are desirable.