Impact of COVID-19 pandemic on mental health: An international study

RESEARCH ARTICLE Impact of COVID-19 pandemic on mental health: An international study Andrew T. GlosterID 1‡ *, Demetris Lamnisos 2 , Jelena LubenkoID 3 , Giovambattista PrestiID 4 , Valeria SquatritoID 4 , Marios Constantinou 5 , Christiana Nicolaou 6 , Savvas Papacostas 7 , Go ¨kc¸en Aydın 8 , Yuen Yu ChongID 9 , Wai Tong Chien 9 , Ho Yu Cheng 9 , Francisco J. Ruiz 10 , Maria B. Garcia-Martin 11 , Diana P. Obando-Posada 11 , Miguel A. Segura-Vargas 10 , Vasilis S. Vasiliou 12 , Louise McHugh 13 , Stefan Ho ¨fer 14 , Adriana Baban 15 , David Dias Neto 16 , Ana Nunes da Silva 17 , JeanLouis Monestès 18 , Javier Alvarez-Galvez 19 , Marisa Paez-Blarrina 20 , Francisco Montesinos 21 , Sonsoles Valdivia-Salas 22 , Dorottya OriID 23 , Bartosz Kleszcz 24 , Raimo Lappalainen 25 , Iva Ivanović 26 , David GosarID 27 , Frederick Dionne 28 , Rhonda M. Merwin 29 , Angelos P. KassianosID 30‡ , Maria KareklaID 30‡ 1Division of Clinical Psychology & Intervention Science, Department of Psychology, University of Basel, Basel, Switzerland, 2Department of Health Sciences, European University Cyprus, Nicosia, Cyprus, 3Psychological Laboratory, Faculty of Public Health and Social Welfare, Riga Stradin¸s ˇUniversity, Riga, Latvia, 4Kore University Behavioral Lab (KUBeLab), Faculty of Human and Social Sciences, Kore University of Enna, Enna, Italy, 5Department of Social Sciences, School of Humanities and Social Sciences, University of Nicosia, Nicosia, Cyprus, 6Department of Nursing, Cyprus University of Technology, Limassol, Cyprus, 7Cyprus Institute of Neurology and Genetics, Nicosia, Cyprus, 8Department of Psychological Counseling and Guidance, Faculty of Education, Hasan Kalyoncu University, Gaziantep, Turkey, 9The Nethersole School of Nursing, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong, 10 Department of Psychology, Fundacio ´n Universitaria Konrad Lorenz, Bogotà, Columbia, 11 Faculty of Psychology, University of La Sabana, Chı ´a, Columbia, 12 School of Applied Psychology, University College Cork, Cork, Ireland, 13 School of Psychology, University College Dublin, Dublin, Ireland, 14 Medical University Innsbruck, Innsbruck, Austria, 15 Department of Psychology, Babeş-Bolyai University (UBB), Cluj-Napoca, Romania, 16 Instituto Superior de Psicologia Aplicada (ISPA), Instituto Universita ´rio; APPsyCI—Applied Psychology Research Center Capabilities & Inclusion, Lisboa, Portugal, 17 Faculdade de Psicologia, Alameda da Universidade, Universidade de Lisboa, Lisboa, Portugal, 18 LIP/PC2S, Universite ´Grenoble Alpes, Grenoble, France, 19 Department of Biomedicine, Biotechnology and Public Health, University of Cadiz, Cadiz, Spain, 20 Instituto ACT, Madrid, Spain, 21 Department of Psychology, European University of Madrid, Madrid, Spain, 22 Department of Psychology and Sociology, University of Zaragoza, Zaragoza, Spain, 23 Vadaskert Child and Adolescent Psychiatric Hospital, Budapest, Hungary, 24 Private Pratice, Poland, 25 Department of Psychology, University of Jyva ¨skyla ¨, Jyva ¨skyla ¨, Finland, 26 Clinic for Psychiatry, Clinical Center of Montenegro, Podgorica, Montenegro, 27 Ljubljana University Medical Centre, Ljubljana, Slovania, 28 De ´partement de Psychologie, Universite ´du Que ´bec àTrois-Rivières, Trois-Rivières, Canada, 29 Department of Psychiatry and Behavioral Science, Duke University, Durham, North Carolina, United States of America, 30 Department of Psychology, University of Cyprus, Nicosia, Cyprus ‡ ATG, MK and AK designed and implemented the study together. AK and MK should be considered joint senior authors. *[email protected] Abstract Background The COVID-19 pandemic triggered vast governmental lockdowns. The impact of these lockdowns on mental health is inadequately understood. On the one hand such drastic changes in daily routines could be detrimental to mental health. On the other hand, it might not be experienced negatively, especially because the entire population was affected. PLOS ONE PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 1 / 20 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Gloster AT, Lamnisos D, Lubenko J, Presti G, Squatrito V, Constantinou M, et al. (2020) Impact of COVID-19 pandemic on mental health: An international study. PLoS ONE 15(12): e0244809. https://doi.org/10.1371/journal. pone.0244809 Editor: Joel Msafiri Francis, University of the Witwatersrand, SOUTH AFRICA Received: October 3, 2020 Accepted: December 16, 2020 Published: December 31, 2020 Copyright: ©2020 Gloster et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: This work was supported by grants from the Swiss National Science Foundation awarded to Andrew T. Gloster (PP00P1_ 163716/1 & PP00P1_190082). The funder provided support in the form of salaries for authors [ATG], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of Methods The aim of this study was to determine mental health outcomes during pandemic induced lockdowns and to examine known predictors of mental health outcomes. We therefore surveyed n = 9,565 people from 78 countries and 18 languages. Outcomes assessed were stress, depression, affect, and wellbeing. Predictors included country, sociodemographic factors, lockdown characteristics, social factors, and psychological factors. Results Results indicated that on average about 10% of the sample was languishing from low levels of mental health and about 50% had only moderate mental health. Importantly, three consistent predictors of mental health emerged: social support, education level, and psychologically flexible (vs. rigid) responding. Poorer outcomes were most strongly predicted by a worsening of finances and not having access to basic supplies. Conclusions These results suggest that on whole, respondents were moderately mentally healthy at the time of a population-wide lockdown. The highest level of mental health difficulties were found in approximately 10% of the population. Findings suggest that public health initiatives should target people without social support and those whose finances worsen as a result of the lockdown. Interventions that promote psychological flexibility may mitigate the impact of the pandemic. Introduction The COVID-19 global pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-COV-2) virus triggered governmentally mandated lockdowns, social distancing, quarantines and other measures in the interest of public health. The mandated lockdowns abruptly and dramatically altered people’s daily routines, work, travel, and leisure activities to a degree unexperienced by most people living outside of war zones. Simultaneously, the highly contagious, yet invisible virus transformed previously neutral situations to perceived potentially dangerous ones: social interaction, touching one’s face, going to a concert, shaking someone’s hand, and even hugging grandparents. Given these changes and looming threat, increases in anxiety and depression can be expected [1]. Indeed, common psychological reactions to previous quarantines include post-traumatic symptoms, confusion, and anger [2], though these data stem from quarantines of specific regions or a subgroup of exposed people, such as medical professionals. It therefore remains an empirical question whether such patterns are consistent when entire populations across the globe are simultaneously affected. For most people, it stands to reason that governmentally mandated lockdowns decrease their activity levels and the number of stimuli experienced compared to pre-lockdown levels. The impact of reducing activities, stimuli and routines on the population is unknown, but various analogue situations can be used to make predictions, like death of a spouse [3]; hearing loss [4]; job loss [5]; long duration expeditions [6]; poor acculturation [7]; and even ageing when combined with loneliness [8]. Each of these situations is associated with increases in psychological distress. This reduction of stimulations may lead to boredom and reductions in reinforcement, which has been associated with depression [9]. The sum total of these PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 2 / 20 these authors are articulated in the ‘author contributions’ section. One of the authors is employed by a commercial affiliation: Private Pratice, Poland. This affiliation provided support in the form of salaries for authors [BK], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: One of the authors is employed by a commercial affiliation: Private Pratice, Poland. This affiliation provided support in the form of salaries for author BK, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. This does not alter our adherence to PLOS ONE policies on sharing data and materials. No other authors have competing interests to declare. literatures, and some evidence from country specific studies on COVID-19 suggests that for some people, the mental distress in the form of stress, depression, and negative affect are likely reactions to the lockdown; therefore, people’s wellbeing is likely to suffer. Indeed, increased loneliness, social isolation, and living alone are associated with increased mortality [10]–the exact effect that mandated lockdown and social distancing rules aimed to counteract. Alternately, the planned slowing down of daily routines can be beneficial. For example, vacations and weekends are highly sought-after–if not always achieved–periods of relaxation and stress reduction [11]. Likewise, some religious and spiritual traditions encourage simplicity, mindfulness, and solitude with the goal of increasing wellbeing [12]. It is therefore conceivable that for some people the lockdown could offer a reprieve from daily hassles and stress and even lead to increases in wellbeing. It is therefore equally important to identify protective factors that can buffer against the negative effects of the lockdown. Although nearly all people around the globe have been subject to some form of lockdown measures to contain the COVID-19 response, variations exist with respect to how each person is confined, even within a single country. For instance, during the COVID-19 pandemic some people were allowed to go to work, whereas others were required to work exclusively from home. For various reasons, some people had difficulty obtaining some basic supplies. Further, some were thrust into the situation of taking care of others (e.g., children, due to closing of schools). Finally, some people lost income as a result of the lockdown, and this is a known risk-factor for poor mental health [13,14]. Finally, a lockdown may be experienced differently the longer it continues and potentially when in confined spaces [2]. All of these lockdown-specific features may have an impact on one’s mental health, but to date it remains inadequately explored. As the risk of the pandemic continues, it is important to understand to what degree the virus-induced uncertainty and the lockdown-induced changes in daily routines impact stress, depression, affect, and wellbeing. Towards this end, it is important to identify factors that can mitigate potential negative psychological effects of pandemics and lockdowns. Various social and psychological factors have been identified in other contexts that may also help build resilience in large-scale pandemics such as COVID-19. On the social level, one such candidate is social support, which has repeatedly been found to positively impact mental health and wellbeing [15–18]. Another social factor is the family climate and family functioning, which clearly impacts people’s mental health [19,20]. Psychological factors such as mindfulness and psychologically flexible response styles (as opposed to rigid and avoidant response styles) are behavioral repertoires that have previously been shown to buffer the impact of stress and facilitate wellbeing [21–24]. Given the scope of the COVID-19 pandemic, it is crucial to better understand how a pandemic and associated lockdowns impact on mental health. Thus, the aim of this study was to determine mental health outcomes and to examine known predictors of outcomes to identify psychological processes and contextual factors that can be used in developing public health interventions. It can be assumed, but remains untested, that those with risks in social-demographic factors, living conditions, social factors and psychological factors have more severe reactions to the lockdown. We therefore tested whether outcomes of stress, depression, affect, and wellbeing were predicted by country of residence, social demographic characteristics, COVID-19 lockdown related predictors, social predictors, and psychological predictors. Methods Participants The inclusion criteria were �18 years of age and ability to read one of the 18 languages (English, Greek, German, French, Spanish, Turkish, Dutch, Latvian, Italian, Portuguese, PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 3 / 20 Finnish, Slovenian, Polish, Romanian, Hong Kong, Hungarian, Montenegrin, & Persian.). There were no exclusion criteria. People from all countries were eligible to participate. Procedure Ethics approval was obtained from the Cyprus National Bioethics Committee (ref.: EEBK EP 2020.01.60) followed by site approvals from different research teams involved in data collection. All participants provided written informed consent prior to completing the survey (computer-based, e.g., by clicking “yes”). A population based cross-sectional study was conducted in order to explore how people across the world reacted to the COVID-19. The anonymous online survey was distributed using a range of methods. Universities emailed the online survey to students and academic staff and also posted the survey link to their websites. In addition, and in order to broaden the sample to older age groups and to those with different socio-demographic characteristics, the survey was disseminated in local press (e.g., newspapers, newsletters, radio stations), in social media (e.g., Facebook, Twitter, etc.), in professional networks, local hospitals and health centers and professional groups’ email lists (e.g., medical doctors, teachers, engineers, psychologists, government workers), and to social institutions in the countries (e.g., churches, schools, cities/townships, clubs, etc.). Data were collected for two months between 07th April and 07th June 2020. The majority of countries where data were collected had declared a state of emergency for COVID-19 during this time. Measures Well validated and established measures were used to assess constructs. When measures did not already exist in a language, they were subject to forward and backward translation procedures. Well-validated measures of predictors and outcomes and items measuring COVID-19 related characteristics were selected after a consensus agreement among the members of this study. Predictors Country. Respondents’ countries were coded and entered as predictors. Socio-demographic status. Participants responded to questions related to their sociodemographic characteristics including their age, gender, country of residence, marital status, employment status, educational level, whether they have children as well as their living situation. Lockdown variables. Participants responded to questions related to lockdown including length of lockdown, whether they need to leave home for work, any change in their finances, whether they were able to obtain basic supplies, the amount of their living space confined in during the lockdown. They were also asked whether they, their partner, or a significant other was diagnosed with COVID-19. Social factors. Social factors were measured using the Brief Assessment of Family Functioning Scale (BAFFS; [25]) and the Oslo Social Support Scale (OSSS; [26]). The BAFFS items are summed to produce a single score with higher scores indicating worse family functioning. The OSSS items are summed up and provide three levels types of social support: low (scored 3–8), moderate (scored 9–11) and high (scored 12–14). Psychological factors. Psychological factors including mindfulness and psychological flexibility. Mindfulness was measured using the Cognitive Affective Mindfulness Scale (CAMS; [27]). The CAMS produces a single score with higher scores indicating better PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 4 / 20 mindfulness qualities. Psychological flexibility (e.g., hold one’s thoughts lightly, be accepting of one’s experiences, engage in what is important to them despite challenging situations) was measured using the Psyflex scale [28]. The Psyflex produces a single score with higher scores indicating better psychological flexibility qualities. Outcomes Stress. Stress was measured using the Perceived Stress Scale (PSS; [29]). The PSS assesses an individual’s appraisal of how stressful situations in their life are. Items ask about people’s feelings and thoughts during the last month. A total score is produced, with higher scores indicating greater overall distress. Depression. Depressive symptomatology was assessed using two items from the disengagement subscale of the Multidimensional State Boredom Scale (MSBS; [30]). These items assessed wanting to do pleasurable things but not finding anything appealing (i.e., boredom), as well as wasting time. Based on concepts of reinforcement deprivation (i.e., lack of access to or engagement with positive stimuli) that is known to contribute to depression, we added an item that measured how rewarding or pleasurable people found the activities that they were engaging in (i.e., reinforcement). Higher scores indicated higher depressive symptomatology. Positive affect/ negative affect. The Positive And Negative Affect Scale (PANAS) was used to measure affect [31]. The original version of the questionnaire was used with five additional items: bored, confused, angry, frustrated and lonely. All items were scored on a 5-point Likert type scale, ranging from 1 = very little/not at all to 5 = extremely and summed up so that higher scores in the positive-related items indicating higher positive affect and higher scores in the negative-related items indicating higher negative affect. In order to capture additional dimensions of negative affect believed to be relevant to the COVID-19 lockdowns, we additionally added five items: bored, confused, angry, frustrated, lonely. Wellbeing. Wellbeing was assessed using the Mental Health Continuum Short Form (MHC-SF; [32]); which assesses three aspects of wellbeing: emotional, psychological, and social. The MHC-SF produces a total score and scores for each of the three aspects of wellbeing. The MHC-SF can also be scored to produce categories of languishing (i.e., low levels of emotional, psychological, and social well-being), flourishing (i.e., high levels of emotional psychological and social well-being almost every day), and moderately mentally healthy (in between languishing and flourishing). Statistical analysis The mean and standard deviation was calculated for dependent variables that follow the normal distribution while the median and interquartile range (IQR) were computed for non-normally distributed data. Bivariable association between an outcome variable and each predictor was investigated with ANOVA test for categorical predictor and univariable linear regression for numerical predictor. Linear mixed-effect model with random effect for country was performed to consider simultaneously several predictors in the same model and to account for the variation in outcome variable between countries. Four separate linear mixed-effect models were used for each outcome variable, one for each set of socio-demographic, lockdown, social and psychosocial predictors and multicollinearity for each set of predictors was investigated with the variation inflation criterion (VIF). Standardized regression coefficients were computed as effect size indices to measure the strength of the association between predictor variables and outcome variables. The comparison between the country mean and overall mean for each outcome variable was estimated though a linear regression model with dependent variable the mean centering outcome and predictor the country. Cohen’s d effect size of the PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 5 / 20 standardize difference between country mean and the overall mean was computed as a measure of the magnitude of the difference between the two means. The whole sample was used in linear mixed-effect models while for the comparison of country mean to the overall mean was used the sample from countries with sample size �100. The R packages lme4 and effect sizes were used for fitting the linear mixed effect model and to compute the standardized regression coefficients of the linear mixed effect models [33]. Significance test and confidence intervals were calculated at a significance level of 0.05. The following cut-off values were used for the evaluation of the effect sizes: ‘tiny’ �0.05, ‘very small’ from 0.05 to �0.10, ‘small’ from 0.10 to �0.20, ‘medium’ from 0.20 to �0.30, ‘large’ from 0.30 to �0.40 and ‘very large’ >0.40 [34]. Results Descriptive Participants were n = 9,565 people from 78 countries. See supporting information for a participation flowchart (S1 Appendix). The countries with the largest samples were: Latvia (n = 1285), Italy (n = 962), Cyprus (n = 957), Turkey (n = 702), Switzerland (n = 550), Hong Kong (n = 516), Colombia (n = 485), Ireland (n = 414), Austria (n = 368), Romania (n = 339), Portugal (n = 334), France (n = 313), Spain (n = 296), Germany (n = 279), Hungary (n = 273), Greece (n = 270), USA (n = 268), Finland (n = 157), Montenegro (n = 147), Poland (n = 135), United Kingdom (n = 100), Slovenia (n = 77), and Canada (n = 60). The remaining countries are listed in the supporting information (S1 Table). Outcome variables The means, standard deviations, and where appropriate percentage of participants within categories of the five outcome variables can be seen in Table 1. Predictor variables Countries. A full list of countries can be found in the supporting information (S1 Table). Socio-demographic status. The mean age was 36.9 (13.3) years. A majority of participants were female (77.7%), approximately a fifth male (22.0%), and small minority identified as other (0.3%). More than half of the respondents were either in a relationship (25.7%) or married (36.1%), almost a third were single (30.8%), and the rest were either divorced (5%), widower (1.1%) or other (1.3%). Participants indicated that they lived: alone (14.6%), with both parents (20.8%), one parent (5.1%), with their own family including partner and children (54.1%), or with friends or roommates (5.5%). Less than half of respondents had children (40.8%). Approximately half of the participants were working full time (53.4%), almost a fifth were working part-time (17.5%), 23.2% were unemployed and a small minority were either on parental leave (2.2%) or retired (3.7%). COVID-19 lockdown variables. At the time of responding, participants were in lockdown or self-isolation for a median of 5.0 (3.0 IQR) weeks. Most people indicated that they had not been infected with COVID-19 (88.0%), a small minority indicated they had been infected (1.4%) and the rest had symptoms but were unsure (10.6%). Similar patterns were seen with reported infection rates of partners (no: 92.2%, yes: 0.7%, unsure: 7.1%) and of people close to them (no: 86.0%; yes: 5.6%; unsure: 8.4%). With respect to leaving the house for work, almost half (47.7%) indicated that this never occurred, 7.7% indicated leaving only once, whereas an almost equal number indicated leaving a couple times per week (23.7%) or more than three times per week (21.0%). Nearly all participants indicated they were able to obtain all the basic supplies they needed (93.5%). Participants reported having a median inner living PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 6 / 20 space of 90.0 square meters (80.0 IQR) and median outdoor space of 20.0 square meters (192.1 IQR). Finally, with respect to finances, more than half indicated that their financial situation remained about the same (57.9%), a minority indicated it improved (8.9%), and a third reported that their finances had gotten worse (33.3%). Social and psychological predictors. Mean values of the other predictors (i.e., social predictors and psychological predictors) can be seen in Table 1. Multivariate analyses Stress. Results of multivariate analyses for the outcome of stress can be seen in Table 2. The largest protective factor against stress was social support (high support vs low support (-3.35, 95%CI, -3.39 to -2.92), with a very large effect size). Positive predictors of stress with large effect sizes were being female (2.42, 95%CI, 2.07 to 2.77) and worsening of finances (2.32, 95%CI, 1.68 to 2.96), whereas psychological flexibility buffered this response (-0.65, 95%CI, -0.69 to -0.62). Higher education levels were also associated with lower levels of stress, with a large effect size (see Table 2). Moderate effect sizes for predictors associated with less stress were older age (-0.13, 95%CI, -0.14, -0.11) and mindfulness (-0.69, 95%CI, -0.74, -0.64). Moderate effect sizes of predictors associated with more stress were worse family functioning (0.98, 95%CI, 0.90, 1.06) and not being able to obtain all basic supplies (1.82 95%CI, 1.12, 2.52). Differences in reported levels of stress across countries were largely negligible, with the exception of two countries that reported higher levels of stress (Hong Kong (2.85, 95%CI, 2.22, Table 1. Descriptive statistics for outcomes and predictors (continuous variables). Mean (M) (SD) % Outcomes Stress 17.1 (7.5) Low 33.0% Moderate 55.9% High 11.1% Depression Symptoms 6.6 (2.3) Lack of Reinforcement 2.1 (1.0) 25.8% a Boredom 2.4 (1.1) 32.4% a Waste Time 2.4 (1.1) 47.4% a Positive Affect 30.0 (8.1) Negative Affect 19.6 (7.8) Wellbeing 41.0 (14.0) Languishing 10.1% Moderately Mentally Healthy 50.0% Flourishing 39.9% Predictors Family Functioning 5.5 (1.9) Social Support 9.9 (2.1) Mindfulness 26.7 (3.2) Psychological Flexibility 21.8 (4.1) Note: Possible ranges of the scales were as follows: Stress (PSS; 0–40); Reinforcement (1–4); Boredom (MSBS; 1–4); Positive Affect (PANAS-P; 10–50); Negative Affect (PANAS-N, plus; 15–75); Wellbeing (MHC-SF; 0–70); Family Functioning (BAFFS; 3–12); Social Support (OSSS; 3–14); Mindfulness (CAMS-R; 0–40); Psychological Flexibility (Psyflex; 6–30) a = % of participants who responded with “lot” or “extremely”. https://doi.org/10.1371/journal.pone.0244809.t001 PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 7 / 20 Table 2. Predictors of stress. PSS Score Coefficient (95% CI) Effect Size (95% CI) Socio Demographic Predictors Sex Male Ref Ref Female 2.42 (2.07, 2.77) 0.34 (0.29, 0.39) Age -0.13 (-0.14, -0.11) -0.23 (-0.26, -0.19) Employment Working (full time) Ref Ref Working (part time) 0.23 (-0.18, 0.64) 0.03 (-0.02, 0.09) Unemployed 1.08 (0.66, 1.51) 0.15 (0.09, 0.21) On parental leave -0.05 (-1.03, 0.94) -0.01 (-0.14, 0.13) Retired -0.39 (-1.25, 0.58) -0.05 (-0.17, 0.07) Education Primary School Ref Ref High School -2.55 (-4.18, -0.92) -0.36 (-0.60, -0.13) Some College/University -2.11 (-3.74, -0.47) -0.30 (-0.53, -0.07) Graduated from College/University -2.49 (-4.10, -0.87) -0.36 (-0.59, -0.13) Master/Postgraduate -2.81 (-4.42, -1.19) -0.40 (-0.63, -0.17) Doctoral level -2.50 (-4.17, -0.82) -0.36 (-0.60, 0.17) Other -1.78 (-3.60, 0.03) -0.26 (-0.51, 0.00) Marital status Single Ref Ref In a relationship/engaged 0.20 (-0.24, 0.64) 0.03 (-0.03, 0.09) Married 0.11 (-0.46, 0.69) 0.02 (-0.06, 0.10) Divorced -0.74 (-1.53, 0.03) -0.10 (-0.21, 0.00) Widower 0.50 (-0.95, 1.95) 0.07 (-0.13, 0.27) Other 0.30 (-0.99, 1.59) 0.04 (-0.14, 0.22) Living situation Live alone Ref Ref Live with parents -0.18 (-0.35, 0.73) -0.03 (-0.10, 0.05) Live with one of parents 0.96 (0.29, 1.80) 0.13 (0.03, 0.24) Live with my own family -0.70 (-1.37, -0.26) -0.10 (-0.17, -0.02) Live with friends/roommates -0.05 (-0.66, 0.81) -0.01 (-0.11, 0.09) Having children Yes Ref Ref No -0.72 (-1.18, -0.26) -0.10 (-0.16, -0.04) Lockdown Predictors Weeks in quarantine/self-isolation 0.05 (-0.00, 0.14) 0.00 (0.00, 0.05) Leave for work No Ref Ref Once only 0.14 (-0.51, 0.79) 0.02 (-0.07, 0.11) A couple of times -0.97 (-1.41, -0.53) -0.13 (-0.19, -0.07) More than 3 times per week -1.03 (-1.50, -0.56) -0.14 (-0.21, -0.08) Changes in finance Have gotten better Ref Ref Stayed about the same 0.54 (-0.06, 1.15) 0.07 (-0.01, 0.15) Have gotten worse 2.32 (1.68, 2.96) 0.31 (0.23, 0.40) (Continued) PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 8 / 20 3.49) and Turkey (2.47, 95%CI, 1.93, 3.02)) and two that reported lower levels of stress (Portugal (-2.50, 95%CI, -3.29, -1.71) and Montenegro (-3.30, 95%CI, -4.49, -2.11)) than the average stress level across all countries. See supporting information for information on each country (S2–S6 Tables). Depression. Results of multivariate analyses for the outcome of depression can be seen in Table 3. The strongest predictor of depression was social support, such that high (-1.30, 95% CI, -1.44, -1.16) and medium levels (-0.73, 95%CI, -0.85, -0.62) of social support were protective against depression (relative to low levels) with a very large and large effect sizes, respectively. The only other large effect size was for psychological flexibility, which also served in a protective manner (-0.20, 95%CI, -0.22, -0.19). Moderate effect sizes of predictors associated with less depression symptoms were also observed for higher education levels (see Table 3). Moderate effect sizes of predictors associated with more depression were worse family functioning (0.29, 95%CI, 0.27, 0.32) and not being able to obtain all basic supplies (0.49, 95%CI, 0.27, 0.70). The amount of depression symptoms reported on average within countries was similar for most countries with the exception of one country with lower reported levels than average with a large effect size (Austria (-0.71, 95%CI, -0.95, -0.47)) and one with higher levels than average with a large effect size (USA (0.85, 95%CI, 0.58, 1.13)). See supporting information for information on each country (S2–S6 Tables). Affect. Results of multivariate analyses for the outcome of affect can be seen in Table 4. With respect to positive affect, social support (high support vs low support (5.69, 95%CI, 5.23, 6.16) and psychological flexibility (0.77, 95%CI, 0.74, 0.81) were both predictors with very large effect sizes. Interestingly, those who left their house more than three times per week had higher levels of positive affect than those that did not leave their house for work (1.68, 95%CI, 1.18, 2.17), with a medium effect size. Higher education levels were associated with higher levels of positive affect with a medium to large effect size (see Table 4, PANAS-Positive). Table 2. (Continued) PSS Score Coefficient (95% CI) Effect Size (95% CI) Obtain all basic supplies Yes Ref Ref Nο1.82 (1.12, 2.52 0.24 (0.15, 0.34) Inner living space (m2) -0.02 (-0.03, -0.00) -0.03 (-0.06, -0.01) Outer space (m2) -0.00 (-0.00, 0.00) -0.01 (-0.03, 0.01) Social Predictors Family Functioning Score 0.98 (0.90, 1.06) 0.24 (0.22, 0.26) Perceived Social support Low Ref Ref Moderate -2.09 (-2.45, -1.72) -0.27 (-0.32, -0.23) High -3.35 (-3.79, -2.92) -0.44 (-0.49, -0.38) Psychological Predictors Mindfulness -0.69 (-0.74, -0.64) -0.29 (-0.31, -0.27) Psychological Flexibility -0.65 (-0.69, -0.62) -0.36 (-0.38, -0.34) Note: Effect size interpretation: tiny (0–0.05); very small (0.05–0.10); small (0.10–0.20); medium (0.20–0.30); large (0.30–0.40); very large (>.40); Color code: light green–dark green (tiny–very large positive effect); light red–dark red (tiny–very large negative effect) https://doi.org/10.1371/journal.pone.0244809.t002 PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 9 / 20 flexibility) and risk factors (loss of income and inability to get basic supplies) that affect mental health outcomes. Third, these factors can be used in future public health responses are being made, including those that require large scale lockdowns or quarantines. That is, public health officials should direct resources to identifying and supporting people with poor social support, income loss, and potentially lower levels of education and provide a strategy to mitigate special risks in these subpopulations. The importance of social support needs to be made clear to the public and to the degree possible mechanisms that can contribute to social support should be supported. Further, psychological flexibility is a trainable set of skills that has repeatedly been shown to ameliorate suffering [22,39]; and can be widely distributed with modern technological intervention tools such as digital, internet, or virtual means [40]. We do not claim, however, that psychological flexibility is the only factor that can be used for interventions. Instead, it is a recognized transdiagnostic factor assessed in this study and one that is feasible to be targeted and modified by interventions and prevention [41–43]. This study is limited by several important factors. First, the results are based on cross sectional analysis and correlations. As such, causation cannot be inferred and any delayed impact of the pandemic and lockdown on peoples’ mental health was not captured. Second, all results of this survey were obtained via self-report questionnaires, which can be subject to retrospective response bias. Third, although the sample was large and based on varied recruitment sources, it was not representative of the population and undersampled people who suffered most from the pandemic (i.e., front line health care professionals, people in intensive care, etc.) or people without internet access, etc. Finally, the country-specific incidence rates and lockdown measures differed across countries. These were not assessed, but future studies are encouraged to investigate how such factors impact mental health outcomes. These limitations notwithstanding, based on nearly 10,000 international participants, this study found that approximately 10% of the population was languishing during or shortly after the lockdown period. These finding have implications for public health initiatives. First, officials are urged to attend to, find, and target people who have little social support and/ or whose finances have worsened as a result of the measures. Second, public health interventions are further urged to target psychological processes such as psychological flexibility in general to potentially help buffer other risk factors for mental health. Likewise, availability of social support and information about where to get support and remain connected are needed. These recommendations should become part of public health initiatives designed to promote mental health in general, and should equally be considered when lockdowns or physical distancing are prescribed during a pandemic. Supporting information S1 Table. List of all countries included in the data set. (PDF) S2 Table. Geodemographic predictors for Perceived Stress Scale. (PDF) S3 Table. Geodemographic predictors for MSBS–depression. (PDF) S4 Table. Geodemographic predictors for PANAS positive. (PDF) S5 Table. Geodemographic predictors for PANAS negative. (PDF) PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 16 / 20 S6 Table. Geodemographic predictors for MHCSF—mental health continuum. (PDF) S1 Appendix. Participation flowchart. (PDF) Acknowledgments We wish to thank the following people for their work in helping to implement the study: Spyros Demosthenous, Christiana Karashali, Diamanto Rovania (University of Cyprus); Maria Antoniade (European University of Cyprus); Ioanna Menoikou (Cyprus University of Technology); Elias Ioannou (University of Nicosia); Sonja Borner, Victoria Firsching-Block, Alexander Fenn (University of Basel); Cristīne S ˇneidere, Ingrīda Trups-Kalne, Lolita Vansovica, Sandra Feldmane, (Riga Stradin¸s ˇUniversity); David Nilsson (Lund University); Miguel A. Segura-Vargas (Fundacio ´n Universitaria Konrad Lorenz); Claudia Lenuţa Rus, Catalina Otoiu, Cristina Vajaean (Babes-Bolyai University). We further wish to thank Fabio Coviello and Sonja Borner (University of Basel) for their help in preparing the manuscript. Author Contributions Conceptualization: Andrew T. Gloster, Maria B. Garcia-Martin, Angelos P. Kassianos, Maria Karekla. Data curation: Demetris Lamnisos. Formal analysis: Demetris Lamnisos. Funding acquisition: Andrew T. Gloster. Investigation: Jelena Lubenko, Giovambattista Presti, Valeria Squatrito, Marios Constantinou, Christiana Nicolaou, Savvas Papacostas, Go¨kc¸en Aydın, Yuen Yu Chong, Wai Tong Chien, Ho Yu Cheng, Francisco J. Ruiz, Maria B. Garcia-Martin, Diana P. Obando-Posada, Miguel A. Segura-Vargas, Vasilis S. Vasiliou, Louise McHugh, Stefan Ho¨fer, Adriana Baban, David Dias Neto, Ana Nunes da Silva, Jean-Louis Monestès, Javier Alvarez-Galvez, Marisa PaezBlarrina, Francisco Montesinos, Sonsoles Valdivia-Salas, Dorottya Ori, Bartosz Kleszcz, Raimo Lappalainen, Iva Ivanović, David Gosar, Frederick Dionne, Rhonda M. Merwin. Methodology: Andrew T. Gloster, Angelos P. Kassianos, Maria Karekla. Project administration: Andrew T. Gloster, Angelos P. Kassianos, Maria Karekla. Resources: Jelena Lubenko, Giovambattista Presti, Valeria Squatrito, Marios Constantinou, Christiana Nicolaou, Savvas Papacostas, Go¨kc¸en Aydın, Yuen Yu Chong, Wai Tong Chien, Ho Yu Cheng, Francisco J. Ruiz, Maria B. Garcia-Martin, Diana P. Obando-Posada, Miguel A. Segura-Vargas, Vasilis S. Vasiliou, Louise McHugh, Stefan Ho¨fer, Adriana Baban, David Dias Neto, Ana Nunes da Silva, Jean-Louis Monestès, Javier Alvarez-Galvez, Marisa PaezBlarrina, Francisco Montesinos, Sonsoles Valdivia-Salas, Dorottya Ori, Bartosz Kleszcz, Raimo Lappalainen, Iva Ivanović, David Gosar, Frederick Dionne, Rhonda M. Merwin. Supervision: Andrew T. Gloster, Angelos P. Kassianos, Maria Karekla. Writing – original draft: Andrew T. Gloster, Demetris Lamnisos, Angelos P. Kassianos, Maria Karekla. Writing – review & editing: Andrew T. Gloster, Demetris Lamnisos, Jelena Lubenko, Giovambattista Presti, Valeria Squatrito, Marios Constantinou, Christiana Nicolaou, Savvas PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 17 / 20 Papacostas, Go¨kc¸en Aydın, Yuen Yu Chong, Wai Tong Chien, Ho Yu Cheng, Francisco J. Ruiz, Maria B. Garcia-Martin, Diana P. Obando-Posada, Miguel A. Segura-Vargas, Vasilis S. Vasiliou, Louise McHugh, Stefan Ho¨fer, Adriana Baban, David Dias Neto, Ana Nunes da Silva, Jean-Louis Monestès, Javier Alvarez-Galvez, Marisa Paez-Blarrina, Francisco Montesinos, Sonsoles Valdivia-Salas, Dorottya Ori, Bartosz Kleszcz, Raimo Lappalainen, Iva Ivanović, David Gosar, Frederick Dionne, Rhonda M. Merwin, Angelos P. Kassianos, Maria Karekla. References 1. Presti G, McHugh L, Gloster A, Karekla M, Hayes SC. The dynamics of fear at the time of COVID-19: A contextual behavioral science perspective. Clin Neuropsychiatry. 2020; 17(2):65–71. 2. Brooks SK, Webster RK, Smith LE, Woodland L, Wessely S, Greenberg N, et al. The psychological impact of quarantine and how to reduce it: rapid review of the evidence. Vol. 395, The Lancet. Lancet Publishing Group; 2020. p. 912–20. 3. Kristensen P, Weisaeth L, Heir T. Bereavement and mental health after sudden and violent losses: A Review. Vol. 75, Psychiatry. Guilford Publications Inc; 2012. p. 76–97. https://doi.org/10.1521/psyc. 2012.75.1.76 PMID: 22397543 4. Lawrence BJ, Jayakody DMP, Bennett RJ, Eikelboom RH, Gasson N, Friedland PL. Hearing loss and depression in older adults: A systematic review and meta-analysis. Gerontologist. 2020. https://doi.org/ 10.1093/geront/gnz009 PMID: 30835787 5. van Eersel JHW, Taris TW, Boelen PA. Reciprocal relations between symptoms of complicated grief, depression, and anxiety following job loss: A cross-lagged analysis. Clin Psychol. 2020 Mar; 6. Alfano CA, Bower JL, Cowie J, Lau S, Simpson RJ. Long-duration space exploration and emotional health: Recommendations for conceptualizing and evaluating risk. Vol. 142, Acta Astronautica. Elsevier Ltd; 2018. p. 289–99. 7. Berry JW, Hou F. Immigrant acculturation and wellbeing in Canada. Can Psychol. 2016 Nov; 57 (4):254–64. 8. Shankar A, Rafnsson SB, Steptoe A. Longitudinal associations between social connections and subjective wellbeing in the English Longitudinal Study of Ageing. Psychol Health. 2015 Jun; 30(6):686–98. https://doi.org/10.1080/08870446.2014.979823 PMID: 25350585 9. Kanter JW, Busch AM, Weeks CE, Landes SJ. The nature of clinical depression: Symptoms, syndromes, and behavior analysis. Vol. 31, Behavior Analyst. Association for Behavior Analysis; 2008. p. 1–21. 10. Holt-Lunstad J, Smith TB, Baker M, Harris T, Stephenson D. Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review. Perspect Psychol Sci. 2015 Mar; 10(2):227–37. https:// doi.org/10.1177/1745691614568352 PMID: 25910392 11. Ryan RM, Bernstein JH, Brown KW. Weekends, work, and well-being: Psychological need satisfactions and day of the week effects on mood, vitality, and physical symptoms. J Soc Clin Psychol. 2010 Jan; 29 (1):95–122. 12. Koenig HG. Research on religion, spirituality, and mental health: A review. Vol. 54, Canadian Journal of Psychiatry. Canadian Psychiatric Association; 2009. p. 283–91. https://doi.org/10.1177/ 070674370905400502 PMID: 19497160 13. Paul KI, Moser K. Unemployment impairs mental health: Meta-analyses. J Vocat Behav. 2009 Jun; 74 (3):264–82. 14. Lund C, Myer L, Stein DJ, Williams DR, Flisher AJ. Mental illness and lost income among adult South Africans. Soc Psychiatry Psychiatr Epidemiol. 2013 May; 48(5):845–51. https://doi.org/10.1007/ s00127-012-0587-5 PMID: 23007296 15. Phillips DR, Siu OL, Yeh AGO, Cheng KHC. Informal social support and older persons’ psychological well-being in Hong Kong. J Cross Cult Gerontol. 2008 Mar; 23(1):39–55. https://doi.org/10.1007/ s10823-007-9056-0 PMID: 18228121 16. Golden J, Conroy RM, Bruce I, Denihan A, Greene E, Kirby M, et al. Loneliness, social support networks, mood and wellbeing in community-dwelling elderly. Int J Geriatr Psychiatry. 2009 Jul; 24(7):694– 700. https://doi.org/10.1002/gps.2181 PMID: 19274642 17. Gaymu J, Springer S. Living conditions and life satisfaction of older Europeans living alone: A gender and cross-country analysis. Ageing Soc. 2010 Oct; 30(7):1153–75. PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 18 / 20 18. Panayiotou G, Karekla M. Perceived social support helps, but does not buffer the negative impact of anxiety disorders on quality of life and perceived stress. Soc Psychiatry Psychiatr Epidemiol. 2013 Feb; 48(2):283–94. https://doi.org/10.1007/s00127-012-0533-6 PMID: 22711064 19. Phillips TM. The Influence of Family Structure Vs. Family Climate on Adolescent Well-Being. Child Adolesc Soc Work J. 2012 Apr; 29(2):103–10. 20. Schrodt P, Witt PL, Messersmith AS. A Meta-Analytical Review of Family Communication Patterns and their Associations with Information Processing, Behavioral, and Psychosocial Outcomes. Commun Monogr. 2008 Sep; 75(3):248–69. 21. Chiesa A, Serretti A. Mindfulness-based stress reduction for stress management in healthy people: A review and meta-analysis. J Altern Complement Med. 2009 May; 15(5):593–600. https://doi.org/10. 1089/acm.2008.0495 PMID: 19432513 22. Gloster AT, Meyer AH, Lieb R. Psychological flexibility as a malleable public health target: Evidence from a representative sample. J Context Behav Sci. 2017; 6(2):166–71. 23. Gu J, Strauss C, Bond R, Cavanagh K. How do mindfulness-based cognitive therapy and mindfulnessbased stress reduction improve mental health and wellbeing? A systematic review and meta-analysis of mediation studies. Vol. 37, Clinical Psychology Review. Elsevier Inc.; 2015. p. 1–12. https://doi.org/10. 1016/j.cpr.2015.01.006 PMID: 25689576 24. Panayiotou G, Karekla M, Mete I. Dispositional coping in individuals with anxiety disorder symptomatology: Avoidance predicts distress. J Context Behav Sci. 2014 Oct; 3(4):314–21. 25. Mansfield AK, Keitner GI, Sheeran T. The Brief Assessment of Family Functioning Scale (BAFFS): a three-item version of the General Functioning Scale of the Family Assessment Device. Psychother Res. 2019 Sep; 29(6):824–31. https://doi.org/10.1080/10503307.2017.1422213 PMID: 29351729 26. Dalgard O. Community health profile as tool for psychiatric prevention. Promot Ment Heal. 1996; 5:681– 95. 27. Feldman G, Hayes A, Kumar S, Greeson J, Laurenceau JP. Mindfulness and emotion regulation: The development and initial validation of the Cognitive and Affective Mindfulness Scale-Revised (CAMS-R). J Psychopathol Behav Assess. 2007 Sep; 29(3):177–90. 28. Gloster AT, Block VJ, Klotsche J, Villanueva J, Rinner MTB, Benoy C, et al. Psy-Flex: A Contextually Sensitive State Measure of Psychological Flexibility. In review. 29. Cohen S. Perceived stress in a probability sample of the United States. In: The social psychology of health. Thousand Oaks, CA, US: Sage Publications, Inc; 1988. p. 31–67. (The Claremont Symposium on Applied Social Psychology.). 30. Fahlman SA, Mercer-Lynn KB, Flora DB, Eastwood JD. Development and validation of the multidimensional state boredom scale. Assessment. 2013 Feb; 20(1):68–85. https://doi.org/10.1177/ 1073191111421303 PMID: 21903683 31. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psychol. 1988; 54(6):1063. https://doi.org/10.1037//0022-3514. 54.6.1063 PMID: 3397865 32. Keyes CLM, Wissing M, Potgieter JP, Temane M, Kruger A, van Rooy S. Evaluation of the mental health continuum-short form (MHC-SF) in Setswana-speaking South Africans. Clin Psychol Psychother. 2008 May; 15(3):181–92. https://doi.org/10.1002/cpp.572 PMID: 19115439 33. Bates D, Ma ¨chler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015 Oct; 67(1):1–48. 34. Funder DC, Ozer DJ. Evaluating Effect Size in Psychological Research: Sense and Nonsense. Adv Methods Pract Psychol Sci. 2019 Jun; 2(2):156–68. 35. Keyes CLM. The mental health continuum: From languishing to flourishing in life. J Health Soc Behav. 2002; 43(2):207–22. PMID: 12096700 36. Keyes CLM, Grzywacz JG. Health as a Complete State: The Added Value in Work Performance and Healthcare Costs. J Occup Environ Med. 2005 May; 47(5):523–32. https://doi.org/10.1097/01.jom. 0000161737.21198.3a PMID: 15891532 37. Keyes CLM. Promoting and Protecting Mental Health as Flourishing: A Complementary Strategy for Improving National Mental Health. Am Psychol. 2007 Feb; 62(2):95–108. https://doi.org/10.1037/0003066X.62.2.95 PMID: 17324035 38. Ni MY, Yao XI, Leung KSM, Yau C, Leung CMC, Lun P, et al. Depression and post-traumatic stress during major social unrest in Hong Kong: a 10-year prospective cohort study. Lancet. 2020 Jan; 395 (10220):273–84. https://doi.org/10.1016/S0140-6736(19)33160-5 PMID: 31928765 39. Gloster AT, Klotsche J, Ciarrochi J, Eifert G, Sonntag R, Wittchen HU, et al. Increasing valued behaviors precedes reduction in suffering: Findings from a randomized controlled trial using ACT. Behav Res Ther. 2017 Apr; 91:64–71. https://doi.org/10.1016/j.brat.2017.01.013 PMID: 28160720 PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 19 / 20 40. Karekla M, Stella, Savvides N, Gloster A. An Avatar-Led Intervention Promotes Smoking Cessation in Young Adults: A Pilot Randomized Clinical Trial. Ann Behav Med. 2020; https://doi.org/10.1093/abm/ kaaa013 PMID: 32383736 41. Thompson E, Destree L, Albertella L, Fontenelle L. Internet-Based Acceptance and Commitment Therapy: A Transdiagnostic Systematic Review and Meta-Analysis for Mental Health Outcomes. Behav Ther. 2020 Jul; 42. Levin ME, MacLane C, Daflos S, Seeley JR, Hayes SC, Biglan A, et al. Examining psychological inflexibility as a transdiagnostic process across psychological disorders. J Context Behav Sci. 2013 Dec; 3 (3):155–63. 43. Dindo L, Van Liew JR, Arch JJ. Acceptance and Commitment Therapy: A Transdiagnostic Behavioral Intervention for Mental Health and Medical Conditions. Vol. 14, Neurotherapeutics. Springer New York LLC; 2017. p. 546–53. https://doi.org/10.1007/s13311-017-0521-3 PMID: 28271287 PLOS ONE COVID-19 impact on mental health PLOS ONE | https://doi.org/10.1371/journal.pone.0244809 December 31, 2020 20 / 20