The Impact of COVID-19 Infection and Characterization of Long COVID in Adolescents With Anxiety Disorders: A Prospective Longitudinal Study

Diversity & Inclusion Statement: We worked to ensure sex and gender balance in the recruitment of human participants. We worked to ensure race, ethnic, and/or other types of diversity in the recruitment of human participants. We worked to ensure that the study questionnaires were prepared in an inclusive way. One or more of the authors of this paper self-identifies as a member of one or more historically underrepresented racial and/or ethnic groups in science. We actively worked to promote sex and gender balance in our author group. We actively worked to promote inclusion of historically underrepresented racial and/or ethnic groups in science in our author group. The author list of this paper includes contributors from the location and/or community where the research was conducted who participated in the data collection, design, analysis, and/or interpretation of the work. While citing references scientifically relevant for this work, we also actively worked to promote sex and gender balance in our reference list. One or more of the authors of this paper self-identifies as a member of one or more historically underrepresented sexual and/or gender groups in science. Diversity & Inclusion Statement: We worked to ensure sex and gender balance in the recruitment of human participants. We worked to ensure race, ethnic, and/or other types of diversity in the recruitment of human participants. We worked to ensure that the study questionnaires were prepared in an inclusive way. One or more of the authors of this paper self-identifies as a member of one or more historically underrepresented racial and/or ethnic groups in science. We actively worked to promote sex and gender balance in our author group. We actively worked to promote inclusion of historically underrepresented racial and/or ethnic groups in science in our author group. The author list of this paper includes contributors from the location and/or community where the research was conducted who participated in the data collection, design, analysis, and/or interpretation of the work. While citing references scientifically relevant for this work, we also actively worked to promote sex and gender balance in our reference list. One or more of the authors of this paper self-identifies as a member of one or more historically underrepresented sexual and/or gender groups in science. While the coronavirus disease 2019 (COVID-19) pandemic has profoundly impacted pediatric mental health, the impact of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection on youth with anxiety disorders has not been prospectively examined. Further, there are limited prospective data on post-acute sequelae COVID-19, including symptoms that constitute the long COVID neuropsychiatric syndrome.1Rogers J.P. Chesney E. Oliver D. et al.Psychiatric and neuropsychiatric presentations associated with severe coronavirus infections: A systematic review and meta-analysis with comparison to the COVID-19 pandemic.Lancet Psychiatry. 2020; 7: 611-627https://doi.org/10.1016/S2215-0366(20)30203-0Abstract Full Text Full Text PDF PubMed Scopus (1450) Google Scholar In December 2019, we began a longitudinal study of adolescents aged 12-17 years with DSM-5 primary anxiety disorders treated with either duloxetine or escitalopram. The study was approved by the institutional review board, and all parents/guardians and adolescents provided informed consent and assent. Diagnoses were confirmed by a board-certified child and adolescent psychiatrist using the Mini-International Neuropsychiatric Interview for Children and Adolescents (MINI-KID).2Sheehan D.V. Sheehan K.H. Shytle R.D. et al.Reliability and validity of the Mini International Neuropsychiatric Interview for Children and Adolescents (MINI-KID).J Clin Psychiatry. 2010; 71: 313-326https://doi.org/10.4088/jcp.09m05305whiCrossref PubMed Scopus (0) Google Scholar At each follow-up visit, a child and adolescent psychiatrist reviewed affective and anxiety symptoms during the prior interval, week by week, with the adolescent and their caregiver(s), using the methodology of the Longitudinal Interval Follow-up Examination (LIFE).3Leon A.C. Solomon D.A. Mueller T.I. et al.A brief assessment of psychosocial functioning of subjects with bipolar I disorder: The LIFE-RIFT. Longitudinal Interval Follow-up Evaluation-Range Impaired Functioning Tool.J Nerv Ment Dis. 2000; 188: 805-812https://doi.org/10.1097/00005053-200012000-00003Crossref PubMed Scopus (90) Google Scholar Each follow-up interview included all items from the Generalized Anxiety Disorder-7 (GAD-7)4Mossman S.A. Luft M.J. Schroeder H.K. et al.The Generalized Anxiety Disorder 7-item scale in adolescents with generalized anxiety disorder: Signal detection and validation.Ann Clin Psychiatry. 2017; 29: 227-234PubMed Google Scholar and Quick Inventory of Depressive Symptomatology (QIDS)5Bech P. Fava M. Trivedi M.H. Wisniewski S.R. Rush A.J. Factor structure and dimensionality of the two depression scales in STAR∗D using level 1 datasets.J Affect Disord. 2011; 132: 396-400https://doi.org/10.1016/j.jad.2011.03.011Crossref PubMed Scopus (42) Google Scholar scales at each week and a weekly clinician-rated Clinical Global Impressions-Severity (CGI-S) scale.6Guy W. CGI Clinical Global Impressions. ECDEU Assessment Manual for Psychopharmacology, revised (DHEW Publ. No. ADM 76-338). National Institute of Mental Health, Rockville, MD1976: 217-222Google Scholar We examined the longitudinal course of anxiety, including following laboratory-confirmed SARS-CoV-2 infection in affected adolescents. In patients with COVID-19, we examined central nervous system (CNS) post-acute sequelae COVID-19–related symptoms (hypersomnia, irritability, concentration, motivation/involvement, and energy) from the QIDS to assess long COVID. To examine differences in anxiety symptoms before and after SARS-CoV-2 infection, the distribution of the difference in means for GAD-7/QIDS items and CGI-S were computed using Monte Carlo simulation draws from a Student t distribution for the mean symptom scores before and after COVID-19. Mean differences, 95% credible intervals, and Bayesian p values were computed from the simulation draws. For the CGI-S and GAD-7 scores, a longitudinal model allowing for individual effects was estimated as: yit=αi+δCOVIDit+βt+uit, where COVIDit is an indicator for post-COVID observations and t is a linear trend. This measures the average impact of COVID-19 on scores over time, controlling for variation across individuals over the sampling period. Further, the high dimensionality and sampling frequency (1,308 observations) increases our statistical power to identify effects even within a small sample of patients (N = 26). In this longitudinal effects model, the COVID indicator variable captures the impact of contracting COVID-19 and reduces confounding from aggregate effects of the pandemic (eg, isolation, SARS-CoV-2 community prevalence) that varied over time. This approach minimizes the possibility that the effect of COVID-19 infection within individuals is spuriously related to time variation. Twenty-six patients contributed 1,308 weeks of data (1,019 patient-weeks from patients who did not contract COVID-19, 289 patient-weeks from patients with COVID-19 infection, 139 patient-weeks from patients after recovery from COVID-19). Patients had a mean (SD) age of 14.3 (1.64) years, with 14.8% male patients and 3.8% Black patients, and were followed for a mean (SD) 50.3 (30.4 weeks) (range, 7-96 weeks; median, 55.5 weeks). All patients met criteria for generalized anxiety disorder, 19% had separation anxiety disorder, 78% had social anxiety disorder, 30% had panic disorder, 26% had attention-deficit/hyperactivity disorder, and 37% had a history of a depressive disorder. All patients were treated with either duloxetine (30-120 mg/day) or escitalopram (5-20 mg/day) at study entry. All patients with SARS-CoV-2 infection were unvaccinated and experienced mild symptoms; none were hospitalized, received monoclonal antibodies, or developed multisystem inflammatory syndrome in children. After SARS-CoV-2 infection, anxiety symptoms significantly worsened for all GAD-7 subscales (Figure 1A), and patients had significant syndromic worsening, reflected by CGI-S scores (p < .001). SARS-CoV-2 infection was not qualitatively associated with worsening medication adherence, and patients with and without COVID-19 had similar use of psychotherapy (p = .159). SARS-CoV-2 infection was generally associated with within-patient worsening in long COVID symptoms other than hypersomnia (Figure 1B). Finally, all but 1 patient with COVID-19 achieved remission, defined by a CGI-S score ≤2 before SARS-CoV-2 infection. Removing the patient who did not remit did not significantly change the magnitude of any of the observed effects. This prospective study of the longitudinal impact of COVID-19 in pediatric anxiety disorders reveals that COVID-19 is associated with worsening anxiety symptoms and a disquieting 33% worsening in syndromic severity. Further, these data raise the possibility that, in anxious youth, COVID-19 is associated with a surfeit of neuropsychiatric symptoms. How these findings relate to neurotropic aspects SARS-CoV-2, inflammatory processes, or psychosocial factors remains to be determined. Additionally, the symptomatic worsening in this already anxious population may relate to myriad psychosocial effects of COVID-19 infection (eg, isolation from postinfection quarantine, family stress). Beyond this, the relation between COVID-19–related worsening anxiety and long COVID is likely complex and characterized by significant endogeneity. Adolescents with more severe anxiety may experience greater or more severe long COVID symptoms, or COVID-19–related worsening anxiety may potentiate long COVID symptoms. Ultimately, understanding the neuropsychiatric effects of COVID-19 and long COVID (ie, post-acute sequelae of COVID-19) may inform novel treatments or optimization of existing interventions. This is one of very few prospective studies using patient-level data in anxious adolescents, with more than half of the patients followed for approximately 52 weeks. Our approach controls for confounding from aggregate effects of the pandemic, although unmeasured factors may have contributed to worsening following SARS-CoV-2 infection, including having other infected family members, knowing someone who was hospitalized or died due to COVID-19, and experiencing pandemic-related impacts on school and food security. Further, while sample size, focus on adolescents, and sample homogeneity may limit the generalizability of the study results, these longitudinal data have high dimensionality, and there are >130 observations/patient over time, which represents a time-series trial for each patient. Thus, we observe the same patient before and after contracting COVID-19; combining the time series and cross-patient dimensions, we have >1,100 observations from patients before COVID-19 and approximately 150 observations from patients after COVID-19. Further, for the analyses involving patients who contracted COVID-19, our analyses control for most patient characteristics before and after COVID-19. Finally, despite our use of frequent interviews with patients and caregivers to improve the accuracy of the reported symptoms and a statistical approach that provides high precision in small samples, additional studies with larger samples and longer follow-up are needed.