1
Investigating the effect of national government physical distancing measures on
depression and anxiety during the COVID-19 pandemic through meta-analysis and
meta-regression
João M. Castaldelli-Maia, Megan E. Marziali, Ziyin Lu, Silvia S. Martins
Department of Epidemiology, Mailman School of Public Health, Columbia University, New
York, NY, U.S, 10032.
Corresponding author:
Dr João Mauricio Castaldelli-Maia
Department of Epidemiology, Mailman School of Public Health, Columbia University
722 West 168th street, Rm. 515, New York, NY 10032
Email: jc5572@cumc.columbia.edu
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted September 2, 2020. ; https://doi.org/10.1101/2020.08.28.20184119doi: medRxiv preprint
NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.
2
Abstract (following PRISMA statement)
Background: COVID-19 physical distancing measures can potentially increase the
likelihood of mental disorders. It is unknown whether these measures are associated with
depression and anxiety.
Objectives: To investigate meta-analytic global levels of depression and anxiety during the
COVID-19 pandemic and how implementation of mitigation strategies (i.e. public
transportation closures, stay-at-home orders, etc.) impacted such disorders.
Data sources: Pubmed, MEDLINE, Web of Science, BIOSIS Citation Index, Current
Content Connect, PsycINFO, CINAHL, medRxiv, and PsyArXiv databases for depression
and anxiety prevalences; Oxford Covid-19 Government Response Tracker for the
containment and closure policies indexes; Global Burden of Disease Study for previous
levels of depression and anxiety.
Study eligibility criteria: Original studies conducted during COVID-19 pandemic, which
assessed categorical depression and anxiety, using PHQ-9 and GAD-7 scales (cutoff
≥
10).
Participants and interventions: General population, healthcare providers, students, and
patients. National physical distancing measures.
Study appraisal and synthesis methods: Meta-analysis and meta-regresssion.
Results: In total, 226,638 individuals were assessed within the 60 included studies. Global
prevalence of both depression and anxiety during COVID-19 pandemic were 24.0% and
21.3%, respectively. There was a wide variance in the prevalence of both anxiety and
depression reported in different regions of the world and countries. Asia, and China
particularly, had the lowest prevalence of both disorders. Regarding the impact of mitigation
strategies on mental health, only public transportation closures increased anxiety prevalence.
Limitations: Country-level data on physical distancing measures and previous
anxiety/depression may not necessarily reflect local (i.e., city-specific) contexts.
Conclusions and implications of key findings: Mental health concerns should not be
viewed only as a delayed consequence of the COVID-19 pandemic, but also as a concurrent
epidemic. Our data provides support for policy-makers to consider real-time enhanced mental
health services, and increase initiatives to foster positive mental health outcomes.
Systematic review registration number: https://doi.org/10.17605/OSF.IO/JQGSF
Key-words: COVID-19, depression, anxiety, public transport, social isolation
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted September 2, 2020. ; https://doi.org/10.1101/2020.08.28.20184119doi: medRxiv preprint
3
1. Introduction
COVID-19 is an unprecedented health emergency, affecting millions of individuals across
the globe.
1
SARS-Coronavirus-2, the virus which causes COVID-19, is transmitted person-
to-person via respiratory droplets.
2
In order to prevent and lessen spread, countries began
implementing mitigation strategies, such as: stay-at-home or shelter-in-place orders,
international travel constraints, closure of schools and workplaces, and movement
limitations
3
. Despite being necessary public health measures, researchers have speculated that
these measures could increase feelings of social isolation and loneliness
4
; this is of
importance, as previous studies have demonstrated that social isolation could impact the
likelihood of mental disorders
5
and physical health outcomes
6
. As of yet, it still remains
unclear to what extent the COVID-19 mitigation strategies could impact mental health. Thus,
it is imperative to investigate the levels of mental health disorders and the possible impacts of
social distancing measures on mental health outcomes
7
.
Prior to the pandemic, depression and anxiety were the most prevalent mental health
disorders in the world
8
. These mental health disorders have also been connected to social
isolation during COVID-19 in local studies
9
. During the COVID-19 pandemic, the levels of
such disorders have increased. Pappa et al.
10
conducted a meta-analysis with thirteen studies
that included 33,062 healthcare workers during COVID-19, and reported a prevalence of
23.2% and 22.8% of anxiety and depression, respectively. These prevalences are greater than
those found in the pre-COVID-19 era.
8
Several studies have assessed depression and anxiety
using scales involving self-reporting during the pandemic.
11-70
These studies report a wide
range of prevalence estimates, which appear to be dependent on the sub-population of interest
(i.e., general population, healthcare providers, students, patients), and the geographic location
within which the study is focused.
11-70
There is a need for meta-analytic investigations
generating global prevalence measures for both depression and anxiety during the pandemic,
with additional exploration via subgroup analysis.
Further, there are mixed findings regarding the effect of mitigation strategies on depression
and anxiety during this pandemic. Previous research has demonstrated marked increases in
online search trends for mental health topics (i.e., sleep disturbances, negative thoughts,
anxiety, suicidal ideation) prior to the implementation of stay-at-home orders in the U.S..
71
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted September 2, 2020. ; https://doi.org/10.1101/2020.08.28.20184119doi: medRxiv preprint
4
Further, an online qualitative study evaluated focus groups during the beginning of the social
distancing measures in the U.K., where they found negative impacts on well-being and
mental health after implementation of mitigation strategies.
72
Individuals who had lower pay,
or vulnerable employment, were the most affected.
72
Thus, the effects of these physical
distancing strategies may be time-sensitive. Moreover, there are varying ongoing physical
distancing measures (i.e., school closures, workplace closures, public events cancellations,
restrictions on the size of gatherings, public transport closures, stay-at-home orders,
restrictions on internal movement between cities and regions within a country, and
international travel controls) during different periods, depending on the location.
3
There is a
need to explore whether these strategies have lasting impacts on depression and anxiety,
taking different time of exposure thresholds to such physical distancing measures into
account.
The present study aims to (1) investigate meta-analytic global levels of depression and
anxiety during the COVID-19 pandemic, and (2) explore the effects of these mitigation
strategies on depression and anxiety.
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted September 2, 2020. ; https://doi.org/10.1101/2020.08.28.20184119doi: medRxiv preprint
5
2. Methods
2.1 Study design
We first conducted a meta-analysis of studies related to the COVID-19 pandemic which
assessed depression and anxiety using PHQ-9 and GAD-7 scales. Subgroup analysis for
region of the world, country, type of population, and coverage were also carried out. Then,
we collected national data regarding the implementation of physical distancing measures and
mitigation strategies,
3
along with previous levels of anxiety and depression from a global
database.
8
These data were included in meta-regression models for the investigation of time-
sensitive effects of mitigation strategies on depression and anxiety, adjusted for previous
levels of such disorders and other possible confounders.
2.2. Review Guidelines and Registration
This study followed the PRISMA statement for transparent report of systematic reviews and
meta-analysis
73
and MOOSE guidelines for Meta-analysis Of Observational Studies in
Epidemiology.
74
Figure S1 and S2 respectively present PRISMA and MOOSE checklists
reporting the page of the manuscript in which we consider that each item was addressed. This
study was registered at the Center for Open Science/Open Science Framework.
75
2.2. Search Strategy
We searched Pubmed, MEDLINE, Web of Science, BIOSIS Citation Index, Current Content
Connect, PsycINFO, and CINAHL databases. All searches were conducted with an end date
of July 29
th
, 2020. Search terms used were: ((sars-cov-2 OR coronavir* OR alphacoronavirus
OR betacoronavirus OR COVID OR COVID-19) AND (PHQ-9 or GAD-7)). As this topic is
developing quickly, we accessed pre-print servers medRxiv and PsyArXiv using the above
search terms. We also searched the WHO database which includes COVID literature (cite)
for studies published by the same date, using the following search terms: (PHQ-9 or GAD-7).
In addition to MEDLINE, this database also includes WHO COVID, Elsevier, Lanzhou
University/CNKI, LILACS, and WPRIM databases.
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)
The copyright holder for this preprintthis version posted September 2, 2020. ; https://doi.org/10.1101/2020.08.28.20184119doi: medRxiv preprint