key: cord-0705523-4y1irs8m
authors: Dragioti, Elena; Li, Han; Tsitsas, George; Lee, Keum Hwa; Choi, Jiwoo; Kim, Jiwon; Choi, Young Jo; Tsamakis, Konstantinos; Estradé, Andrés; Agorastos, Agorastos; Vancampfort, Davy; Tsiptsios, Dimitrios; Thompson, Trevor; Mosina, Anna; Vakadaris, Georgios; Fusar‐Poli, Paolo; Carvalho, Andre F.; Correll, Christoph U.; Han, Young Joo; Park, Seoyeon; Il Shin, Jae; Solmi, Marco
title: A large‐scale meta‐analytic atlas of mental health problems prevalence during the COVID‐19 early pandemic
date: 2022-01-09
journal: J Med Virol
DOI: 10.1002/jmv.27549
sha: 009455ecdbbf20f96c5209045ff823ae96fd4297
doc_id: 705523
cord_uid: 4y1irs8m

The COVID‐19 pandemic and related restrictions can impact mental health. To quantify the mental health burden of COVID‐19 pandemic, we conducted a systematic review and meta‐analysis, searching World Health Organization COVID‐19/PsycInfo/PubMed databases (09/29/2020), including observational studies reporting on mental health outcomes in any population affected by COVID‐19. Primary outcomes were the prevalence of anxiety, depression, stress, sleep problems, posttraumatic symptoms. Sensitivity analyses were conducted on severe mental health problems, in high‐quality studies, and in representative samples. Subgroup analyses were conducted stratified by age, sex, country income level, and COVID‐19 infection status. One‐hundred‐seventy‐three studies from February to July 2020 were included (n = 502,261, median sample = 948, age = 34.4 years, females = 63%). Ninety‐one percent were cross‐sectional studies, and 18.5%/57.2% were of high/moderate quality. The highest prevalence emerged for posttraumatic symptoms in COVID‐19 infected people (94%), followed by behavioral problems in those with prior mental disorders (77%), fear in healthcare workers (71%), anxiety in caregivers/family members of people with COVID‐19 (42%), general health/social contact/passive coping style in the general population (38%), depression in those with prior somatic disorders (37%), and fear in other‐than‐healthcare workers (29%). Females and people with COVID‐19 infection had higher rates of almost all outcomes; college students/young adults of anxiety, depression, sleep problems, suicidal ideation; adults of fear and posttraumatic symptoms. Anxiety, depression, and posttraumatic symptoms were more prevalent in low‐/middle‐income countries, sleep problems in high‐income countries. The COVID‐19 pandemic adversely impacts mental health in a unique manner across population subgroups. Our results inform tailored preventive strategies and interventions to mitigate current, future, and transgenerational adverse mental health of the COVID‐19 pandemic.

Since December 2019, a novel coronavirus (SARS-CoV-2), causing the severe acute respiratory syndrome coronavirus 2 (COVID- 19) spread from Wuhan, China, worldwide, becoming a pandemic. 1 As of 01/13/2021, the World Health Organization (WHO) has reported over 90 million confirmed cases of COVID-19 and over 1.9 million deaths. 2 Restrictions, such as social distancing, travel restriction, and quarantine, became necessary to reduce pandemic spread. 3 A large body of evidence exists regarding the physical effects of COVID-19 on different groups of the population, including pregnant women, 4 pediatric patients, 5 or those with pre-existing risk factors for COVID-19. 6 However, comparatively fewer studies have evaluated the mental health consequences of COVID-19. [7] [8] [9] [10] It also remains unclear whether nonclinical risk factors, such as sex, age, incomelevel country data, are associated with adverse mental health consequences. 7, 8, 11 Learning from earlier severe acute respiratory syndrome (SARS-Cov-1/MERS-CoV) epidemics, COVID-19 might heavily impact mental health. During SARS, healthcare workers (HCWs) reported concerns for personal safety and increased anxiety, depression, and psychotic symptoms, 12 and both perceived risk of contracting SARS and quarantine were associated with depression. 12, 13 Working without adequate equipment and training had also detrimental effects on the mental health of HCWs. 14 The general population reported fear of contagion and infecting close contacts, loneliness, and boredom associated with quarantine, 11 as well as anxiety and insomnia. 14 Similar effects on mental health were also observed following the novel Influenza A (H1N1), Ebola, and Middle East respiratory syndrome (MERS) epidemics. 15 Since COVID-19 has spread worldwide, the global negative mental health impact could be much higher. 7, 8 Indeed, converging evidence suggests the COVID-19 pandemic adversely affects mental health across different countries with different income and measures, [16] [17] [18] [19] patient subpopulations, [20] [21] [22] and quarantine status. 23, 24 A recent systematic review found that HCWs in direct contact with COVID-19 patients were at higher risk for depression, anxiety, insomnia, distress, and indirect traumatization than other occupational groups. 25 The risk of psychological distress increased with quarantine duration and social isolation. 15, 23, 24 Furthermore, the likelihood of experiencing mental health concerns was disproportionately increased in those with pre-existing psychiatric disorders, or physical conditions (epilepsy, Parkinson's disease, cancer, etc.). [20] [21] [22] 26, 27 Hence, summarizing evidence on the mental health effects of COVID-19 in the general and specific subpopulations is of paramount importance. Previous evidence synthesis efforts were restricted to specific populations 20, 21, 22, 25, 28, 29 or included mainly studies from Asia (almost 91%), with very few studies from other countries/continents. 9, 10 Three previous meta-analyses have pooled data on the prevalence of mental health outcomes in the general population; however, many studies have been published since the most recent one, and all previous meta-analyses narrowed inclusion criteria to a restricted set of outcomes of interest. [30] [31] [32] The aim of our work is to conduct a focused meta-analysis to summarize the mental health impact of COVID-19 pandemic during the first 6 months of the pandemic, without restrictions on outcomes or population.

This meta-analysis followed a protocol (https://osf.io/3ary9/) and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement recommendations 33 (Supporting Information). Study selection flowchart is provided in Figure S1 .

We searched the WHO COVID-19/PsycInfo/PubMed databases, last search 09/29/2020 (for details, see Supporting Information Methods). Two authors (E. D., K. T.) independently screened title/ abstracts. A third author (M. S.) resolved any disagreement. Coprimary outcomes were the prevalence of anxiety, depression, stress, sleep problems, and posttraumatic symptoms. Secondary outcomes were the prevalence of any other mental health problem (e.g., anger, suicidal ideation, hostility, fear, wellbeing etc.). Since we measured prevalence, outcomes were all categorical (e.g., including percentage/ number of individuals scoring higher than scales' thresholds, or meeting criteria for outcomes).

Included were peer-reviewed observational studies (surveys, crosssectional, case-control, cohort) reporting on primary/secondary outcomes in any population group in countries affected by COVID- 19 36, 37 (Table S1) for cross-sectional studies.

Random-effect meta-analyses of prevalence with 95%CIs (calculated if not reported, see Supporting Information Methods) 38 were performed by the target population, using metaprop packages in Stata. 39, 40 The 95%CIs of the pooled prevalence rate was calculated using the cimethod (exact) and the Freeman Tukey double arcsine transformation (ftt command), which computes the weighted pooled estimate and performs the back transformation on the pooled estimate. 39 This method has properties that make it the clearly preferred option over other choices (e.g., logit transformation). 41 Heterogeneity was calculated as the I 2 . 42 Publication bias was assessed if ≥10 studies using Egger test (p < 0.10), and visual funnel plot inspection. 43, 44 Sensitivity analyses were conducted on studies that reported severe psychosocial/psychiatric symptomatology as defined by cut-off scores or prevalence rates from the original articles, were of high quality, and included representative samples. infection (authors' definition). A univariate random-effects metaregression was applied to investigate factors potentially contributing to the between-study heterogeneity for the main outcomes 45 followed by multivariate analyses with those variables with p < 0.10 in the univariate analyses. We used Stata 13. 46 

Out of 4242 records after removing duplicates, 313 full texts were assessed, 140 of which were excluded (Table S2) , and 173 included ( Figure S1 ; Table S3 ).

Included studies reported on 1237 estimates, from 502,261 participants from 32 countries. Figure 1 The list of instruments used to measure the symptoms and the respective cut-offs are reported in Table S4 .

Only six studies reported data on race and ethnicity, and three studies involved multilanguage surveys (maximum = 2 languages).

Snowball/convenience sampling was the most common recruitment method, and only three studies included representative samples. [49] [50] [51] No study accounted for infection rates. Fifteen studies reported data on confirmed COVID-19 cases, 12 also reported estimates of mental health measurements before the pandemic (seven cross-sectional, one case−control, two cohort, two without specifying design).

However, data for a pre-post meta-analysis were available in only six of those studies. Studies collected data between 02/2020 and 07/ 2020. Thirty-two studies (18.5%) met AHRQ/NOS scale high-quality, 99 (57.2%) moderate, 42 studies (24.3%) low quality (agreement 89%).

Prevalence estimates for primary outcomes across different population groups are reported in Table 1 . In patients with mental disorders and in COVID-19-infected patients, sleep problems and PTSD symptomatology were most prevalent (34% and 32% and 63% and 94%, respectively) without evidence for publication bias for anxiety/ depression symptoms (Figures S2−S5).

In patients with a somatic disorder, anxiety/depression symptoms were most prevalent (31%/37%, respectively). In HCWs, stress and sleep problems were most prevalent (33% and 37%, respectively). Publication bias was detected for anxiety/depression symptoms, stress, and sleep problems in HCWs (Figures S6−S12).

In non-HCW working populations, depression/PTSD symptomatology were most prevalent (22%/14%, respectively). Publication bias emerged for depression symptoms only (Figures S13−S15). In the general population, anxiety symptoms and sleep problems were most prevalent (21% and 35%, respectively). Publication bias affected estimates of anxiety, depression, posttraumatic symptoms, and sleep problems (Figures S16−S22).

In caregivers and family members of people with COVID-19 or young adults in quarantine, anxiety/depression symptoms were most prevalent (42%/21%). In pregnant women, stress and sleep problems were most prevalent (84% and 53%, respectively).

Prevalence estimates for secondary outcomes across different population groups are reported in Table 1 . In patients with mental disorders, behavior problems were most prevalent (77%) and suicidal ideation was the least (12%). In COVID-19-infected patients, fatigue was most prevalent (54%) and miscellaneous (i.e., impaired general F I G U R E 1 Heat map chart of included studies across countries; darker color represents higher number of studies mental health) were the least (44%). In patients with a somatic disorder, hostility was most prevalent (37%) and organic psychosis, personality disorders, schizophrenia, or schizoaffective disorders were the least (1%). In HCWs, fear was most prevalent (71%) and suicidal ideation the least (9%). In non-HCW working populations, fear was most prevalent (29%), and obsessivecompulsive symptoms were the least (2%). In the general population, general outcomes were most prevalent (38%) and behavior problems the least (10%). In caregivers and family members of people with COVID-19 or young adults in quarantine, anger was most prevalent (32%) and hopelessness the least (3%). In pregnant women the only reported secondary outcome was eating disorder with a prevalence of 21%.

Results of sensitivity analyses of severe mental symptoms (as indicated in original articles reporting the prevalence of most severe symptoms) are detailed in Table 2 Results of analyses limited to high-quality studies are detailed in Table S5 . Briefly, the highest mental health problem prevalence was for stress (84%) in pregnant women, followed for prevalence rates 

Subgroup analyses' results are detailed in Tables 2, 3, and Tables S6,   S7 . 

Meta-regression was performed for the main outcomes. Significant moderators of greater adverse impact in the final multivariate metaregression model (Table S8 and 

In this systematic review and meta-analysis, we provide a paramount picture of the mental health impact of the COVID-19 pandemic.

Summarizing evidence on the prevalence of >20 mental health outcomes from 173 studies, across 32 countries and 502,261 participants, we show that the pandemic is substantially associated with a high prevalence of mental health problems globally, but with specific effects across different population groups. Overall, based on only six studies with longitudinal data, an increase between 9% and 31% in the prevalence rates before and after the pandemic were found for anxiety, depression, stress, and sleep problems.

As the negative impact differed across specific groups, 28 In subgroup analyses, COVID-19 seems to adversely affect mental health to a lesser extent in children, adolescents, and older adults, compared with younger (college students) and (middle-aged) adults, although the overall prevalence rates still ranged from 6% to 20% for anxiety/depression symptoms, and sleep problems. To the best of our knowledge, this systematic review and prevalence meta-analysis is the largest and most comprehensive review on the mental health burden across various populations during the COVID-19 pandemic. Our findings widen and complete the findings of two previous meta-analyses, adding more than 100 studies the former one which included 63 studies 30 and around 40 studies to the latter 31 pooling data on a broader set of outcomes. Regarding outcomes, the latter meta-analysis focused only on prevalence rates for depression, anxiety, insomnia, and PTSD, while this meta-analysis focuses on more than five times the number of outcomes. Also, while the prevalence estimates in that work 31 have been provided by HCWs/citizens, our meta-analysis goes more in depth providing specific and fine-grained measures across many strata of the general population. Furthermore, 102 studies were added to a more recent meta-analysis focusing on the prevalence of mental health problems during the pandemic. 62 The added value of such a granular approach is to inform tailored interventions and preventing strategies per population subgroup. Our work is also larger than one further previous Appropriate tests for outcomes with ≥10 studies indicated significant publication bias and small-study effects for many common outcomes, such as anxiety, depression, posttraumatic symptoms, sleep problems, but not all outcomes could be tested owing to low numbers of included studies for some outcomes.

Finally, the present study has several limitations. First and most importantly, study designs, population characteristics, assessment methods and resulting findings were highly heterogeneous. However, meta-analyses evaluating prevalence estimates showed substantially higher heterogeneity than meta-analyses of other effect size metrics. [64] [65] [66] We tried to address this caveat by conducting subgroup analyses and meta-regression analyses based on study-level factors, examining potential sources of heterogeneity. Significant factors for higher heterogeneity were female sex for anxiety; time of the investigation, that is, May 2020 and combinations of various months (i.e., March to April) versus January 2020 for depression; study location, that is, USA versus Asia for stress; and use of unvalidated instruments for sleep problems. Second, most included studies were cross-sectional and used self-report instruments employing online surveys. Importantly, in studies using ICD-10 diagnostic criteria, the prevalence estimates of mental disorders were much lower ranging from 1% to 3%. 47 As a result, no causal inferences can be made, and there is a possibility of recall or other biases related to self-report instruments. Third, data on the change in the prevalence of mental health problems from before the pandemic to the time during the pandemic were restricted to only six studies. Finally, this metaanalysis does not report on effect size estimates based on continuous data, as too few studies employed methods to measure syndromal ICD or DMS disorder prevalences. Fourth, we did not identify any study in Chinese, which might have left out studies published in

Chinese not listed in the databases we searched.

In conclusion, the COVID-19 pandemic is having a concerning impact on mental health globally, adversely affecting diverse symptom clusters in specific at-risk populations differently. More multilanguage and multicountry studies are required, with larger sample sizes, that include multiple subgroups and account for the COVID-19

pandemic course, ask about the mental health status at the time before the COVID-19 pandemic as well as at the time of completing the survey, that also ask about mental healthcare-seeking behaviors, access and adherence to care, and that also collect data in re- 

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

An interactive web-based dashboard to track COVID-19 in real time

Association of public health interventions with the epidemiology of the COVID-19 outbreak in Wuhan, China

COVID-19 in pregnant women and neonates: a systematic review of the literature with quality assessment of the studies

Clinical characteristics of children with COVID-19

Indicators of critical illness and predictors of mortality in COVID-19 patients

Mental health and clinical psychological science in the time of COVID-19: challenges, opportunities, and a call to action

Using social and behavioural science to support COVID-19 pandemic response

Prevalence of psychological morbidities among general population, healthcare workers and COVID-19 patients amidst the COVID-19 pandemic: a systematic review and meta-analysis

COVID-19 pandemic and mental health consequences: systematic review of the current evidence

The psychological impact of quarantine and how to reduce it: rapid review of the evidence

The psychological impact of the SARS epidemic on hospital employees in China: exposure, risk perception, and altruistic acceptance of risk

Depression after exposure to stressful events: lessons learned from the severe acute respiratory syndrome epidemic

The immediate psychological and occupational impact of the 2003 SARS outbreak in a teaching hospital

Occurrence, prevention, and management of the psychological effects of emerging virus outbreaks on healthcare workers: rapid review and meta-analysis

Generalized anxiety disorder, depressive symptoms and sleep quality during COVID-19 outbreak in China: a webbased cross-sectional survey

A nationwide survey of psychological distress among Italian people during the COVID-19 pandemic: immediate psychological responses and associated factors

The depressive state of Denmark during the COVID-19 pandemic

Factors associated with mental health outcomes among health care workers exposed to coronavirus disease 2019

Do psychiatric patients experience more psychiatric symptoms during COVID-19 pandemic and lockdown? A case-control study with service and research implications for immunopsychiatry

Severe psychological distress among patients with epilepsy during the COVID-19 outbreak in southwest China

Early impact of COVID-19 on individuals with eating disorders: a survey of~1000 Individuals in the United States and the Netherlands

The psychological impact of the COVID-19 epidemic on college students in China

The immediate mental health impacts of the COVID-19 pandemic among people with or without quarantine managements

Psychiatric symptomatology associated with depression, anxiety, distress, and insomnia in health professionals working in patients affected by COVID-19: a systematic review with meta-analysis

Incidence of anxiety in Parkinson's disease during the coronavirus disease (COVID-19) Pandemic

Epidemiology of mental health problems among patients with cancer during COVID-19 pandemic

Prevalence of depression, anxiety, and insomnia among healthcare workers during the COVID-19 pandemic: a systematic review and meta-analysis

Prevalence of anxiety in medical students during the COVID-19 pandemic: a rapid systematic review with meta-analysis

Prevalence of mental health problems during the COVID-19 pandemic: a systematic review and metaanalysis

Prevalence of symptoms of depression, anxiety, insomnia, posttraumatic stress disorder, and psychological distress among populations affected by the COVID-19 pandemic: a systematic review and meta-analysis

The prevalence of psychological consequences of COVID-19: a systematic review and meta-analysis of observational studies

Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement

(NOS) for assessing the quality of nonrandomized studies in meta-analyses

Disparities in cancer screening in people with mental illness across the world versus the general population: prevalence and comparative meta-analysis including 4 717 839 people

AHRQ series paper 5: grading the strength of a body of evidence when comparing medical interventions--agency for healthcare research and quality and the effective health-care program

Motivations for weight loss in adolescents with overweight and obesity: a systematic review

Calculating confidence intervals for rates and ratios

Metaprop: a Stata command to perform meta-analysis of binomial data

Meta-analysis in clinical trials

Meta-analysis of prevalence

Measuring inconsistency in meta-analyses

Bias in metaanalysis detected by a simple, graphical test

Quantifying the risk of error when interpreting funnel plots

How should meta-regression analyses be undertaken and interpreted

StataCorp. Stata Statistical Software: Release 13. 2013. College

Mental health of orthopaedic trauma patients during the 2020 COVID-19 pandemic

The ICD−10 classification of mental and behavioural disorders: conversion tables between ICD-8 I-aI-. World Health Organization Division of Mental Health

The association of being in quarantine and related COVID-19 recommended and nonrecommended behaviors with psychological distress in Chinese population

Prevalence of depression symptoms in US adults before and during the COVID-19 pandemic

Mental health before and during the COVID-19 pandemic: a longitudinal probability sample survey of the UK population

Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science

COVID-19 related stress exacerbates common physical and mental pathologies and affects treatment (Review)

Severe mental disorders in complex emergencies

COVID-19 pandemic and its impact on mental health of healthcare professionals

Mental health and the Covid-19 pandemic

Psychological status of medical workforce during the COVID-19 pandemic: a cross-sectional study

Psychological impact of severe acute respiratory syndrome on health workers in a tertiary hospital

Preventing a parallel pandemic-a national strategy to protect clinicians' well-being

COVID-19 and older adults: what we know

Economic, social and political issues raised by the COVID-19 pandemic

Public mental health problems during COVID-19 pandemic: a large-scale meta-analysis of the evidence

Impact of coronavirus syndromes on physical and mental health of health care workers: systematic review and meta-analysis

A systematic review and meta-analysis of prevalence of Escherichia coli in foods of animal origin in Ethiopia. Heliyon

A systematic review and meta-analysis on the prevalence of dementia in Europe: estimates from the highest-quality studies adopting the DSM IV diagnostic criteria

High statistical heterogeneity is more frequent in meta-analysis of continuous than binary outcomes

A large-scale meta-analytic atlas of mental health problems prevalence during the COVID-19 early pandemic