Mental disorders following COVID-19 and other epidemics: a systematic review and meta-analysis

Abstract

Abstract COVID-19 has imposed a very substantial direct threat to the physical health of those infected, although the corollary impact on mental health may be even more burdensome. Here we focus on assessing the mental health impact of COVID-19 and of other epidemics in the community. We searched five electronic databases until December 9, 2020, for all peer-reviewed original studies reporting any prevalence or correlates of mental disorders in the general population following novel epidemics in English, Chinese or Portuguese. We synthesised prevalence estimates from probability samples during COVID-19 and past epidemics. The meta-analytical effect size was the prevalence of relevant outcomes, estimated via random-effects model. I 2 statistics, Doi plots and the LFK index were used to examine heterogeneity and publication bias. This study is pre-registered with PROSPERO, CRD42020179105. We identified 255 eligible studies from 50 countries on: COVID-19 (n = 247 studies), severe acute respiratory syndrome (SARS; n = 5), Ebola virus disease (n = 2), and 1918 influenza (n = 1). During COVID-19, we estimated the point prevalence for probable anxiety (20.7%, 95% CI 12.9–29.7), probable depression (18.1%, 13.0–23.9), and psychological distress (13.0%, 0–34.1). Correlates for poorer mental health include female sex, lower income, pre-existing medical conditions, perceived risk of infection, exhibiting COVID-19-like symptoms, social media use, financial stress, and loneliness. Public trust in authorities, availability of accurate information, adoption of preventive measures and social support were associated with less morbidity. The mental health consequences of COVID-19 and other epidemics could be comparable to major disasters and armed conflicts. The considerable heterogeneity in our analysis indicates that more random samples are needed. Health-care professionals should be vigilant of the psychological toll of epidemics, including among those who have not been infected.