Predicting Anxiety and Depression During the COVID-19 Pandemic
Author: Brooklynn Bailey
Publisher:
Published: 2021
Total Pages: 0
ISBN-13:
DOWNLOAD EBOOKObjective: Substantial increases in symptoms of anxiety and depression have been reported among the general public during the coronavirus disease 2019 (COVID-19) pandemic. Prior to the pandemic, individual risk factors for emotional disorders have been identified by a considerable body of research. However, COVID-19 has brought exposure to a variety of new stressors, including risk of infection and stress and isolation related to infection control procedures. Well studied transdiagnostic vulnerability factors as well as novel risk factors for anxiety and depression need to be examined in this context. This study utilized machine learning regularization and variable selection methods to identify key individual risk factors for anxiety and depressive symptoms over the course of the pandemic. Methods: Adults in the United States (N = 1,200) completed seven online self-report assessments over a 4.5-month period (data collected April 24-October 3, 2020). Anxiety and depressive symptoms were assessed at each time point using the Generalized Anxiety Disorder Scale-7 and the Quick Inventory of Depressive Symptomatology. Cumulative symptom severity across the assessment period was calculated using area under the curve (AUC). A machine learning approach to elastic net regularized regression was used to select predictors of 1) cumulative anxiety severity and 2) cumulative depression severity among a set of 68 sociodemographic, psychological, and pandemic-related baseline variables. Results: Using established cut scores at the initial assessment, nearly half of participants met criteria for probable generalized anxiety disorder (n = 564, 47.0%) and a quarter met criteria for probable major depressive disorder (n = 306, 25.5%). For cumulative anxiety severity, 52 selected baseline variables predicted 79.7% of the variance (predicted R2). Cumulative anxiety severity was most strongly explained by perceived stress, depressive symptom reactivity, and brooding. For cumulative depression severity, the combination of 46 selected baseline variables predicted 76.2% of the variance (predicted R2). Comorbid generalized anxiety, health anxiety, and poor sleep contributed most strongly to predicting depression. There were considerable similarities in the risk factors identified across models. Depressive symptom reactivity, full-time employment, and living with someone age 60+ were among the most important predictors of greater symptom severity in both models. Important psychological risk factors that represent potential targets for interventions include depressive symptom reactivity, brooding, instrumental social support, and hopelessness. Discussion: Many established risk factors for anxiety and depression emerged as important in the context of the pandemic, including greater stress and comorbid psychopathology. Contrary to pre-pandemic research, full-time employment and being married/cohabiting were related to greater risk of anxiety and depression. Several coping strategies which may be adaptive in other contexts, such as use of instrumental social support, were associated with symptoms over the study period. Characteristics associated with COVID-19 risk (e.g., living with someone age 60+, having a chronic health condition) also emerged as risk factors for anxiety and depression. Findings suggest that although many risk factors for internalizing psychopathology appear to generalize to the pandemic, differential relationships and novel contributors need to be considered. Findings highlight risk factors for internalizing psychopathology during the pandemic and suggest possible treatment targets for psychological interventions.