COVID-19 pandemic and lockdown measures impact on mental health among the general population in Italy. An N=18147 web-based survey

Abstract

Background
The psychological impact of the COronaVIrus Disease 2019 (COVID-19) outbreak and lockdown measures on the Italian population are unknown.
The current study assesses rates of mental health outcomes in the Italian general population three to four weeks into lockdown measures and explores the impact of COVID-19 related potential risk factors.
Methods
A web-based survey spread throughout the internet between March 27th and April 6th 2020. 18147 individuals completed the questionnaire, 79.6% women.
Selected outcomes were post-traumatic stress symptoms (PTSS), depression, anxiety, insomnia, perceived stress and adjustment disorder symptoms (ADS). Seemingly unrelated logistic regression analysis was performed to identify COVID-19 related risk factors.
Results
Respondents endorsing PTSS, depression, anxiety, insomnia, high perceived stress and adjustment disorder were 6604 (37%), 3084 (17.3%), 3700 (20.8%), 1301 (7.3%), 3895 (21.8%) and 4092 (22.9%), respectively. Being woman and younger age were associated with all of the selected outcomes. Quarantine was associated with PTSS, anxiety and ADS. Any recent COVID-related stressful life event was associated with all the selected outcomes. Discontinued working activity due to the COVID-19 was associated with all the selected outcomes, except for ADS; working more than usual was associated with PTSS, Perceived stress and ADS. Having a loved one deceased by COVID-19 was associated with PTSS, depression, perceived stress and insomnia.
Conclusion
We found high rates of negative mental health outcomes in the Italian general population three weeks into the COVID-19 lockdown measures and different COVID-19 related risk factors. These findings warrant further monitoring on the Italian population s mental health.

Full text

COVID-19 pandemic and lockdown measures impact on mental 1 health among the general population in Italy. An N=18147 web-based 2 survey. 3
4
Rodolfo Rossi1, Valentina Socci2*, Dalila Talevi2, Sonia Mensi3, Cinzia Niolu1,4, Francesca 5 Pacitti2, Antinisca Di Marco5, Alessandro Rossi2, Alberto Siracusano1,4, Giorgio Di 6 Lorenzo1,4,6. 7
1Chair of Psychiatry, Department of Systems Medicine, University of Rome Tor Vergata, Rome, 8 Italy 9
2Department of Biotechnological and Applied Clinical Sciences, University of L’Aquila, 10 L’Aquila, Italy 11
3Department of Anesthesiology, Intensive Care and Emergency Medicine, Fondazione Policlinico 12 Universitario A. Gemelli IRCCS, Rome, Italy 13
4Psychiatry and Clinical Psychology Unit, Fondazione Policlinico Tor Vergata, Rome, Italy 14
5Department of Information Engineering, Computer Science and Mathematics, University of 15 L’Aquila, L’Aquila, Italy 16
6IRCCS Fondazione Santa Lucia, Rome, Italy. 17
18
*Correspondence: 19 Valentina Socci 20 valentinasocci@gmail.com 21 22
Abstract 23
Background 24
The psychological impact of the COronaVIrus Disease 2019 (COVID-19) outbreak and lockdown 25 measures on the Italian population are unknown. 26
The current study assesses rates of mental health outcomes in the Italian general population three 27 to four weeks into lockdown measures and explores the impact of COVID-19 related potential risk 28 factors. 29
Methods 30
A web-based survey spread throughout the internet between March 27th and April 6th 2020. 18147 31 individuals completed the questionnaire, 79.6% women. 32
Selected outcomes were post-traumatic stress symptoms (PTSS), depression, anxiety, insomnia, 33 perceived stress and adjustment disorder symptoms (ADS). Seemingly unrelated logistic 34 regression analysis was performed to identify COVID-19 related risk factors. 35
Results 36
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

Respondents endorsing PTSS, depression, anxiety, insomnia, high perceived stress and adjustment 37 disorder were 6604 (37%), 3084 (17.3%), 3700 (20.8%), 1301 (7.3%), 3895 (21.8%) and 4092 38 (22.9%), respectively. Being woman and younger age were associated with all of the selected 39 outcomes. Quarantine was associated with PTSS, anxiety and ADS. Any recent COVID-related 40 stressful life event was associated with all the selected outcomes. Discontinued working activity 41 due to the COVID-19 was associated with all the selected outcomes, except for ADS; working 42 more than usual was associated with PTSS, Perceived stress and ADS. Having a loved one 43 deceased by COVID-19 was associated with PTSS, depression, perceived stress and insomnia. 44
Conclusion 45
We found high rates of negative mental health outcomes in the Italian general population three 46 weeks into the COVID-19 lockdown measures and different COVID-19 related risk factors. These 47 findings warrant further monitoring on the Italian population’s mental health. 48
49
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

Background 50
The psychological impact of the COronaVIrus Disease 2019 (COVID-19) outbreak and related 51 lockdown measures among the Italian population are unknown. The COVID-19 pandemic is a 52 global health emergency that could potentially have a serious impact on public health, including 53 mental health (World Health Organization, 2020a; Xiang et al., 2020). Since clusters of atypical 54 pneumonia of unknown etiology were discovered in the city of Whuan, Hubei province, in late 55 December 2019, the viral disease has continued to exponentially spread throughout China and 56 worldwide. Italy has been the first European country that had to face the pandemic. On March 9th 57 2020, lockdown measures were enforced by the government on entire national territory. 58 Lockdown measures included travel restrictions, the mandatory closure of schools, nonessential 59 commercial activities and industries. People were asked to stay at home and socially isolate 60 themselves to prevent being infected. 61
As previously reported, health emergencies such as epidemic can lead to detrimental and long-62 lasting psychosocial consequences, due to disease related fear and anxiety, large-scale social 63 isolation, and the overabundance of (mis)information on social media and elsewhere (Dong and 64 Bouey, 2020). At the individual level, epidemics are associated with a wide range of psychiatric 65 comorbidities including anxiety, panic, depression and trauma-related disorders (Tucci et al., 66 2017). The psychosocial impact of health emergences seems to be even higher during quarantine 67 measures (Brooks et al., 2020). Quarantine has been associated with high stress levels 68 (DiGiovanni et al., 2004), depression (Hawryluck et al., 2004), irritability and insomnia (Lee et 69 al., 2005). Furthermore, being quarantined is associated with acute stress (Bai et al., 2004) and 70 trauma-related (Wu et al., 2009) disorders, particularly in specific at-risk populations such as 71 health workers (Lai et al., 2020). 72
Concerning the COVID-19 pandemic, a study on 1210 respondents in China found rates of 30% 73 of anxiety and 17% of depression (Wang et al., 2020). Further, in a nationwide survey including 74 more than 50.000 Chinese respondents, almost 35% of the participants reported trauma-related 75 distress symptoms, with women and young adults showing significantly higher psychological 76 distress (Qiu et al., 2020). 77
Together, these findings strongly suggest the need to accurately and timely assess the magnitude 78 of mental health outcomes in the general population exposed to COVID-19 pandemic, with 79 particular regard to the implementation of preventive and early interventions strategies for those at 80 higher risk. However, no study to date has investigated mental health outcomes and associated 81 risk factors in the Italian population. This could be of additional relevance considering the 82 implementation of the strict lockdown and social distancing measures imposed on the entire 83 national territory. 84
The aim of the current study was to assess rates of mental health outcomes in the Italian general 85 population three to four weeks into lockdown measures and to explore the impact of COVID-19 86 related potential risk factors. This study aims at providing evidence that could potentially inform 87 subsequent research strategies and mental health delivery in Italy and Europe. 88
89
Methods 90
Study Design 91
A cross-sectional web-based survey design was adopted. Approval for this study was obtained 92 from the local IRB at University of L’Aquila. On-line consent was obtained from the participants. 93
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

Participants were allowed to terminate the survey at any time they desired. The survey was 94 anonymous, and confidentiality of information was assured. 95
Data on mental health were collected between March 27th and April 6th 2020 using an on-line 96 questionnaire spread throughout the internet, using sponsored social network advertisement 97 together with a snowball recruiting technique. The investigated timeframe corresponds to the 98 contagion peak in Italy, according to epidemiogical data confirmed by the World Health 99 Organization (World Health Organization, 2020). The survey was developed using the free 100 software Google Forms®. 101
102
Participants 103
All Italian citizens 18 years were eligible. A total of 18147 individuals completed the 104 questionnaire, of which 14447 (79.6%) women, median age was 38 (IQR=23). Because of the 105 web-based design, no response rate could be estimated as it was not possible to estimate how 106 many persons were reached by social network advertisement. 107
108
Mental health outcomes 109
Post-Traumatic Stress Symptoms (PTSS), depression, anxiety, insomnia, perceived stress and 110 adjustment disorder symptoms (ADS) were assessed using the Italian versions of the following 111 instruments and cut-offs or scoring: 112
• the Global Psychotrauma Screen, post-traumatic stress symptoms subscale (GPS-PTSS) 113 (Olff et al. in press): PTSS were considered of clinical relevance if more than 3 out of five 114 5 symptoms were reported as present; 115
• the 9-item Patient Health Questionnaire (PHQ-9) (Spitzer et al., 1999), using the cut-off 116 for severe depression at 15; 117
• the 7-item Generalized Anxiety Disorder scale (GAD-7) (Spitzer et al., 2006), using the 118 cut-off for severe anxiety at 15; 119
• the 7-item Insomnia Severity Index (ISI) (Morin et al., 2011), using the cut-off at 22 for 120 severe insomnia; 121
• the 10-item Perceived Stress Scale (PSS) (Cohen and Hoberman, 1983), using a quartile 122 split to separate the higher quartile from the remaining participants; 123
• the International Adjustment Disorder Questionnaire (IADQ) (Shevlin et al., 2020), using 124 the standard scoring system. IADQ comprises a brief checklist of potentially stressful 125 events, such as financial, work, health or housing problems. The IADQ checklist was 126 modified in order to ascertain if the reported problem was due to COVID-19. ADS were 127 rated as present if a stressful life event correlated to COVID-19 was present, together with 128 preoccupation and failure to adapt symptoms and a relevant impact on global functioning. 129
130
Independent variables 131
Standardized age, gender and region of residence (Northern Italy: Aosta Valley, Piedmont, 132 Liguria, Lombardy, Trentino-Alto Adige, Veneto, Friuli-Venezia Giulia, Emilia-Romagna; 133 Central Italy: Tuscany, Umbria, Marche, Lazio; Southern Italy: Abruzzo, Molise, Apulia, 134 Campania, Basilicata, Calabria, Sicily and Sardinia) were inserted as independent variables. 135
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

Region of residence was inserted in order to account for the different incidence of COVID-19 136 among Italian regions. COVID-19 related independent variables were: 1) being under quarantine 137 either because infected or in close proximity to infected people; 2) any changes in working 138 activity compared to “working as usual” (e.g., smart-working, working activity discontinued due 139 to lockdown measures, higher workload due to COVID-19); 3) having a loved one infected, 140 hospitalized or deceased due to COVID-19; 4) any stressful events comprised in the IDAQ 141 checklist, purposely modified in order to capture only stressful events due to COVID-19. The 142 IADQ checklist comprises 8 questions about any potential stressful life event occurred in the 143 recent past, with a yes/no response, including financial, working, educational, housing, 144 relationship, own or loved one’ s health and caregiving problems. In order to separate COVID-19 145 related stressful life events from non-COVID-19 related events, responses to the checklist were 146 modified as follows: “no”; “yes”; “yes, due to COVID-19”. Responses were collapsed in a binary 147 variable where 1=“any stressful life evet only if due to COVID-19” and 0=“no stressful life events 148 or presence of a stressful life event not due to COVID-19”. 149
150
Confounders 151
A history of childhood trauma and any previous mental illness, as assessed by the dedicated GPS 152 module; education level, occupation (employed, unemployed, student, retired) and being in a 153 relationship. 154
155
Statistical Analysis 156
Frequency analysis were performed in order to ascertain the prevalence of each outcome, 157 separately for Northern, Central and Southern Italy. 158
A seemingly-unrelated multivariate logistic regression model was fitted in order to explore the 159 impact of the proposed covariates and confounders on the selected outcomes. Seemingly unrelated 160 regression models are systems of equations that allow to jointly model several outcomes, 161 assuming correlation among their errors. Because of the very low missing data rates (<3%), 162 missing data were treated with listwise deletion in regression analysis. 163
Data analysis was performed using Stata v. 16® (StataCorp). Seemingly unrelated logistic 164 regression was performed using the -suest - postestimation command after running a panel of 165 logistic regressions. 166
167
Results 168
Socio-demographic characteristics of the sample, along with rates of mental health outcomes, are 169 reported in Table 1. Of the 18147 respondents, 6666 (37.14%) reported 3/5 PTSS, with a median 170 total GPS symptom score of 7 (IQR=6, range 0-17); 3099 respondents (17.3%) reported severe 171 depressive symptoms, with a PHQ total median score of 8 (IQR=6, range 0-17); 3732 (20.8%) 172 respondents reported severe anxiety symptoms, with GAD median score of 8 (range 0-21, 173 IQR=10); 1306 (7.3%) respondents reported severe insomnia symptoms, with ISI median total 174 score of 10 (range 0-28, IQR=12); PSS total score median was 25 (range 4-44, IQR=13), 75th 175 percentile was 31, with 3933 (21.9%) respondents scoring above this threshold; 4129 (23.0%) 176 respondents reported a IADQ scoring compatible with the suspect of a presence of an adjustment 177 disorder. 178
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

Seemingly unrelated logistic regression analyses are reported in Table 2. Being a woman was 179 associated with all of the selected outcomes (PTSS: OR=2.12 [1.94, 2.31]; depression: OR=1.39 180 [1.24, 1.56]; anxiety: OR=1.77 [1.59, 1.97]; perceived stress: OR=2.06 [1.85, 2.30]; insomnia: 181 OR=1.50 [1.26, 1.78]; adjustment disorder: OR=1.64 [1.45, 1.84]). Younger age was associated 182 with PTSS, depression, anxiety and perceived stress (respectively: OR=1.49 [1.39, 1.60]; 1.55 183 [1.42, 1.69]; 1.72 [1.59, 1.87]; 1.76 [1.62, 1.90]). Compared to Northern Italy, participants from 184 Southern Italy showed higher odds of all of the selected outcomes, except for ADS (PTSS: 185 OR=1.36 [1.26, 1.47]; depression: OR=1.25 [1.13, 1.37]; anxiety: OR=1.29 [1.18, 1.41]; 186 perceived stress: OR=1.20 [1.10, 1.32]; insomnia: OR=1.41 [1.24, 1.62]). Being under quarantine 187 because infected or in close proximity to infected people was associated with PTSS, Anxiety and 188 ADS (respectively: OR=1.74 [1.21,2.49]; 1.52 [1.05,2.22]; 2.28 [1.44,3.61]). Having experienced 189 a stressful life event due to COVID-19, as assessed by the modified IADQ checklist, was 190 associated with all of the selected outcomes (PTSS: OR=1.46 [1.37,1.56]; depression: OR=1.58 191 [1.45,1.72]; anxiety: OR=1.64 [1.51,1.78]; perceived stress: OR=1.82 [1.68,1.97]; insomnia: 192 OR=1.58 [1.40,1.79]). OR of IADQ-Checklist on ADS was not estimated due to the perfect 193 prediction, because having an IADQ checklist event is a prerequisite for having a suspected 194 Adjustment Disorder. Working activity discontinued due to COVID-19 was associated with all of 195 the selected outcomes except for ADS (PTSS: OR=1.15 [1.05,1.27]; depression: OR=1.40 196 [1.23,1.59]; anxiety: OR=1.16 [1.03,1.31]; perceived stress: OR=1.19 [1.06,1.34]; insomnia: 197 OR=1.22 [1.03,1.46]), while working more than usual due to the COVID-19 was associated with 198 PTSS, perceived stress and ADS (respectively: OR= 1.42 [1.18,1.71]; 1.71 [1.38,2.12]; 1.39 199 [1.04,1.87]). Having a loved one deceased by COVID-19 was associated with PTSS (OR=1.68 200 [1.30,2.16]), depression (OR=1.41 [1.03,1.93]), perceived stress (OR=1.34 [1.01, 1.78], insomnia 201 (OR=1.74 [1.18, 2.54]), while having a loved one diagnosed with COVID-19 was associated with 202 PTSS (OR=1.22 [1.05, 1.42]). 203
204
Discussion 205
In this study, we report for the first time on the mental health outcomes related to COVID-19 206 outbreak and related lockdown measures on the general population in Italy. To the best of our 207 knowledge, this is the first study to report on mental health outcomes related to the COVID-19 208 outbreak in Europe on such a large sample size. This study shows relatively high rates of PTSS, 209 Depression, Anxiety, Insomnia, Perceived stress and ADS, with young women having higher odds 210 of endorsing a mental health outcome. These outcomes were associated with a number of COVID-211 19-related risk factors, including being under quarantine, having a loved one deceased by COVID-212 19, working activity discontinued due to lockdown measures, or experiencing other stressful 213 events (i.e. working, financial, relationship or housing problems) due to the pandemic or 214 lockdown measures. These findings were adjusted for previous psychiatric illness and a history of 215 childhood trauma, suggesting that the COVID-19 pandemic is exerting an independent effect on 216 the population mental health. 217
218
Previous literature 219
Compared to an early report on the mental health outcomes related to COVID-19 in China on 220 1210 respondents (Wang et al., 2020), we found lower rates of anxiety, similar rates of depression 221 and higher levels of perceived stress, notwithstanding differences in assessment tools. The 222 negative association with age and the positive association with female gender was confirmed, 223 suggesting that young women may be at heightened risk for mental disorders. Compared to 224 another large web-based survey from China on 52730 respondents that evaluated peritraumatic 225
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

stress-related symptoms, we found similar rates of PTSS (Qiu et al., 2020). Another study on 285 226 participants from hardest-hit Hubei province found substantially lower rates of PTSS, around 7% 227 (Liu et al., 2020). Such disparities could be due to different assessment tools used and differences 228 in sample size. A study on 7143 medical students in China (Cao et al., 2020) found severe anxiety 229 rates, assessed as GAD 15, to be 0.9%, compared to our 20.9%. This inconsistence could be due 230 to the particular population investigated, having a high education level. Indeed, higher education 231 was associated with better outcomes in our study. Furthermore, cultural, social and health care 232 system differences between China and Italy could explain differences in reported mental health 233 outcomes. 234
Coherently with previous reports from China, female gender (Liu et al., 2020; Qiu et al., 2020; 235 Wang et al., 2020) and younger age (Qiu et al., 2020; Wang et al., 2020) were consistently 236 associated with higher risk for different mental health outcomes. If confirmed in other populations 237 worldwide, these findings could be of great importance for subsequent intervention strategy for 238 global mental health related to COVID-19. 239
Relevance 240
Monitoring populations’ mental health is critical during a pandemic, as generalized fear and fear-241 induced over-reactive behaviour among the public could impede infection control (Dong and 242 Bouey, 2020). Further, the current strict lockdown measures and the home confinement of 243 unknown duration represent an unprecedented stressful event potentially leading to significant 244 long-term health costs. Epidemiological monitoring and targeted intervention should be therefore 245 timely implemented to prevent further mental health problems. Indeed, once the outbreak will be 246 over, its negative socio-economic consequences may have a detrimental effect on the population’s 247 mental health, as suggested by our finding of an heightened risk of mental health issues due to 248 COVID-19 related working difficulties and by earlier studies related to the last economic crisis 249 (Wahlbeck et al., 2011). 250
251
Limitations and future directions 252
This study has some important limitations due to the sampling technique. Relying on social 253 networks voluntary recruitment and re-sharing could have introduced an important selection bias, 254 firstly excluding people not on social networks, and secondly introducing a self-selection bias, as 255 suggested by the highly unbalanced gender ratio observed. This latter bias could have affected 256 also two other large web-based surveys in China, that reported on samples with a 64.7% and 257 67.3% proportion of woman (Qiu et al., 2020; Wang et al., 2020). For these reasons, rates of 258 mental health outcomes should be interpreted with caution. Secondly, this survey was based on 259 self-report instruments that could introduce a systematic bias and return different rates compared 260 to interview-based measures. 261
This study has also several strengths, including a very large sample size and the sampling 262 timeframe that corresponded to the pandemic peak in Italy. 263
Future studies will need to monitor the trajectory of mental health outcomes, in order to define 264 mental health interventions at a population level. 265
266
Conclusions 267
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

We found high rates of negative mental health outcomes in the Italian general population three to 268 four weeks into the COVID-19 pandemic and lockdown measures. COVID-19 related factors 269 were associated with these outcomes independently from previous mental illness or childhood 270 trauma. These findings warrant further monitoring on the Italian population’s mental health and 271 could serve to inform structured interventions in order to mitigate the impact on mental health of 272 the outbreak. 273
274
Authors contribution 275
Conceptualization: RR, VS, FP, GDL; Methodology: RR; Formal Analysis: RR; Data Curation: 276 RR, SM, GDL; Writing - Original Draft: RR, VS; Writing - Review & Editing: RR, VS, DT, 277 ADM, FP, SM, CN, AR, AS, GDL. 278
Funding 279
No specific funding was granted for this study. 280
Acknowledgments 281
This work is supported by Territori Aperti, a project founded by “Fondo Territori Lavoro e 282 Conoscenza CGIL CISL UIL”. 283
Conflicts of interests 284
The authors have no conflict of interest to disclose. 285
Contribution to the field 286
The COronaVIrus Disease 2019 (COVID-19) pandemic is a global health emergency that could 287 potentially have a serious impact on public health, including mental health. The psychological 288 impact of the COVID-19 outbreak and related lockdown measures among the Italian population 289 are unknown. In this web-based study, we report for the first time on the psychological impact of 290 COVID-19 outbreak on the general population in Italy. This study shows high rates of post-291 traumatic symptoms, Depression, Anxiety, Insomnia, Perceived stress and Adjustment Disorder 292 associated with a number of COVID-19-related risk factors. This study represents the first 293 European report on mental health in the time of the COVID-19, and it could have a strong impact 294 on subsequent research and clinical intervention strategy for global mental health related to 295 COVID-19. 296
297
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

298
BIBLIOGRAPHY 299
Bai, Y. M., Lin, C. C., Lin, C. Y., Chen, J. Y., Chue, C. M., and Chou, P. (2004). Survey of stress reactions among 300 health care workers involved with the SARS outbreak. Psychiatr. Serv. 55, 1055–1057. 301 doi:10.1176/appi.ps.55.9.1055. 302
Brooks, S. K., Webster, R. K., Smith, L. E., Woodland, L., Wessely, S., Greenberg, N., et al. (2020). The 303 psychological impact of quarantine and how to reduce it: rapid review of the evidence. Lancet 395, 912–920. 304 doi:10.1016/s0140-6736(20)30460-8. 305
Cao, W., Fang, Z., Hou, G., Han, M., Xu, X., Dong, J., et al. (2020). The psychological impact of the COVID-19 306 epidemic on college students in China. Psychiatry Res., 112934. doi:10.1016/j.psychres.2020.112934. 307
Cohen, S., and Hoberman, H. M. (1983). Positive Events and Social Supports as Buffers of Life Change Stress1. J. 308 Appl. Soc. Psychol. 13, 99–125. doi:10.1111/j.1559-1816.1983.tb02325.x. 309
DiGiovanni, C., Conley, J., Chiu, D., and Zaborski, J. (2004). Factors influencing compliance with quarantine in 310 Toronto during the 2003 SARS outbreak. Biosecur. Bioterror. 2, 265–272. doi:10.1089/bsp.2004.2.265. 311
Dong, L., and Bouey, J. (2020). Public Mental Health Crisis during COVID-19 Pandemic, China. Emerg. Infect. Dis. 312 26. doi:10.3201/eid2607.202407. 313
Hawryluck, L., Gold, W. L., Robinson, S., Pogorski, S., Galea, S., and Styra, R. (2004). SARS control and 314 psychological effects of quarantine, Toronto, Canada. Emerg. Infect. Dis. 10, 1206–1212. 315 doi:10.3201/eid1007.030703. 316
Lai, J., Ma, S., Wang, Y., Cai, Z., Hu, J., Wei, N., et al. (2020). Factors Associated With Mental Health Outcomes 317 Among Health Care Workers Exposed to Coronavirus Disease 2019. JAMA Netw. open 3, e203976. 318 doi:10.1001/jamanetworkopen.2020.3976. 319
Lee, S., Chan, L. Y. Y., Chau, A. M. Y., Kwok, K. P. S., and Kleinman, A. (2005). The experience of SARS-related 320 stigma at Amoy Gardens. Soc. Sci. Med. 61, 2038–2046. doi:10.1016/j.socscimed.2005.04.010. 321
Liu, N., Zhang, F., Wei, C., Jia, Y., Shang, Z., Sun, L., et al. (2020). Prevalence and predictors of PTSS during 322 COVID-19 Outbreak in China Hardest-hit Areas: Gender differences matter. Psychiatry Res., 112921. 323 doi:10.1016/j.psychres.2020.112921. 324
Morin, C. M., Belleville, G., Bélanger, L., and Ivers, H. (2011). The Insomnia Severity Index: Psychometric 325 Indicators to Detect Insomnia Cases and Evaluate Treatment Response. Sleep 34, 601–608. 326 doi:10.1093/sleep/34.5.601. 327
Olff, M., Aakvaag, H. F., Brewer, D., Elmore Borbon, D. L., Hyland, P., Kassam-Adams, N., et al. Screening for 328 consequences of trauma – an update on the Global Collaboration on Traumatic Stress. In press. Eur. J. 329 Psychotraumatol. 11. 330
Qiu, J., Shen, B., Zhao, M., Wang, Z., Xie, B., and Xu, Y. (2020). A nationwide survey of psychological distress 331 among Chinese people in the COVID-19 epidemic: Implications and policy recommendations. Gen. Psychiatry 332 33, 19–21. doi:10.1136/gpsych-2020-100213. 333
Shevlin, M., Hyland, P., Ben-Ezra, M., Karatzias, T., Cloitre, M., Vallières, F., et al. (2020). Measuring ICD-11 334 adjustment disorder: the development and initial validation of the International Adjustment Disorder 335 Questionnaire. Acta Psychiatr. Scand. doi:10.1111/acps.13126. 336
Spitzer, R. L., Kroenke, K., Williams, J. B. W., and Löwe, B. (2006). A Brief Measure for Assessing Generalized 337 Anxiety Disorder. Arch. Intern. Med. 166, 1092. doi:10.1001/archinte.166.10.1092. 338
Spitzer, R. L., Kroken, K., and Williams, J. B. (1999). Validation and Utility of a Self-Report Version of PRIME-MD: 339 The PHQ Primary Care Study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. 340 JAMA 282, 1737. doi:10.1001/jama.282.18.1737. 341
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

Tucci, V., Moukaddam, N., Meadows, J., Shah, S., Galwankar, S. C., and Bobby Kapur, G. (2017). The forgotten 342 plague: Psychiatric manifestations of ebola, zika, and emerging infectious diseases. J. Glob. Infect. Dis. 9, 151–343 156. doi:10.4103/jgid.jgid_66_17. 344
Wahlbeck, K., Anderson, P., Basu, S., McDaid, D., and Stuckler, D. (2011). Impact of economic crises on mental 345 health. World Health. 346
Wang, C., Pan, R., Wan, X., Tan, Y., Xu, L., Ho, C. S., et al. (2020). Immediate Psychological Responses and 347 Associated Factors during the Initial Stage of the 2019 Coronavirus Disease (COVID-19) Epidemic among the 348 General Population in China. Int J Env. Res Public Heal. 17. doi:10.3390/ijerph17051729. 349
World Health Organization (2020a). Mental Health and Psychosocial Considerations During COVID-19 Outbreak. 350
World Health Organization (2020b). World Health Organization - COVID-19. Available at: https://who.sprinklr.com/. 351
Wu, P., Fang, Y., Guan, Z., Fan, B., Kong, J., Yao, Z., et al. (2009). The psychological impact of the SARS epidemic 352 on hospital employees in China: Exposure, risk perception, and altruistic acceptance of risk. Can. J. Psychiatry 353 54, 302–311. doi:10.1177/070674370905400504. 354
Xiang, Y.-T., Yang, Y., Li, W., Zhang, L., Zhang, Q., Cheung, T., et al. (2020). Timely mental health care for the 355 2019 novel coronavirus outbreak is urgently needed. The Lancet Psychiatry 7, 228–229. doi:10.1016/S2215-356 0366(20)30046-8. 357
358
359
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

360
Table 1. Demographic characteristics and rates of mental health outcomes in the sample 361
Total North Centre South
No. / Median (% / IQR) No. / Median (% / IQR) No. / Median (% / IQR) No. / Median (% / IQR)
Age 38 (23) 38 (23) 38 (24) 38 (31)
Gender
Women 14207 (79.5) 6310 (79) 3729 (79.4) 4168 (80.6)
Men 3653 (20.5) 1681 (21) 966 (20.6) 1006 (19.4)
Education
Undergraduate 8538 (47.8) 3770 (47.2) 2243 (47.8) 2525 (48.8)
Postgraduate 7674 (43) 3411 (42.7) 2112 (45) 2151 (41.6)
Lower education 1649 (9.2) 810 (10.1) 340 (7.2) 499 (9.6)
Occupation
Housewife 1139 (6.4) 367 (4.6) 244 (5.2) 528 (10.2)
Unemployed 2094 (11.7) 793 (9.9) 484 (10.3) 817 (15.8)
Employed 10881 (60.9) 5349 (66.9) 2867 (61.1) 2665 (51.5)
Retired 291 (1.6) 124 (1.6) 77 (1.6) 90 (1.7)
Student 3456 (19.3) 1358 (17) 1023 (21.8) 1075 (20.8)
Currently on Quarantine 141 (0.8) 101 (1.3) 21 (0.5) 19 (0.4)
Working activity change
As usual 2320 (13.5) 977 (12.6) 633 (14) 710 (14.5)
Smart-working 6688 (38.9) 3088 (39.9) 1847 (40.9) 1753 (35.7)
Discontinued 7500 (43.7) 3347 (43.2) 1870 (41.4) 2283 (46.5)
More than usual 665 (3.9) 335 (4.3) 168 (3.7) 162 (3.3)
Loved one’s status
None 16312 (91.8) 6987 (87.6) 4431 (94.7) 4894 (95.5)
Infected 789 (4.4) 519 (6.5) 139 (3) 131 (2.6)
Deceased 253 (1.4) 183 (2.3) 30 (0.6) 40 (0.8)
Hospitalized 424 (2.4) 284 (3.6) 80 (1.7) 60 (1.2)
GPS PTSS 3 6604 (37) 2876 (36) 1560 (33.2) 2168 (41.9)
PHQ 15 3084 (17.3) 1349 (16.9) 703 (15) 1032 (20)
GAD 15 3700 (20.8) 1613 (20.2) 854 (18.3) 1233 (23.9)
ISI 22 1301 (7.3) 542 (6.8) 280 (6) 479 (9.3)
PSS 75th percentile 3895 (21.8) 1720 (21.5) 918 (19.6) 1257 (24.3)
ADS 4092 (22.9) 1900 (23.8) 1032 (22) 1160 (22.4)
GPS: Global Psychotrauma Screen; PHQ: Patient Health Questionnaire; GAD: Generalized Anxiety Disorder scale; ISI: Insomnia severity Index; PSS: 362 Perceived Stress Scale; ADS: Adjustment Disorder Symptom; IQR: Interquartile range. 363
364
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint

Table 2: Seemingly Unrelated Logistic Regression
PTSS Depression Anxiety Perceived Stress Insomnia ADS
OR [95%CI] OR [95%CI] OR [95%CI] OR [95%CI] OR [95%CI] OR [95%CI]
Age§ 1.49*** [1.39,1.60] 1.55*** [1.42,1.69] 1.72*** [1.59,1.87] 1.76*** [1.62,1.90] 1.01 [0.97,1.05] 1.05 [0.75,1.47]
Gender
Men 1.00 (ref)
Women 2.12*** [1.94,2.31] 1.39*** [1.24,1.56] 1.77*** [1.59,1.97] 2.06*** [1.85,2.30] 1.50*** [1.26,1.78] 1.64*** [1.45,1.84]
Region
North 1.00 (ref)
Centre 0.93 [0.86,1.01] 0.87* [0.78,0.97] 0.90* [0.82,1.00] 0.90* [0.82,0.99] 0.9 [0.77,1.05] 0.91 [0.81,1.02]
South 1.36*** [1.26,1.47] 1.25*** [1.13,1.37] 1.29*** [1.18,1.41] 1.20*** [1.10,1.32] 1.41*** [1.24,1.62] 0.95 [0.85,1.06]
COVID-19-Related Stressful Event 1.46*** [1.37,1.56] 1.58*** [1.45,1.72] 1.64*** [1.51,1.78] 1.82*** [1.68,1.97] 1.58*** [1.40,1.79] n.a. n.a.
Currently On Quarantine 1.74** [1.21,2.49] 1.49 [0.98,2.26] 1.52* [1.05,2.22] 1.42 [0.97,2.07] 1.23 [0.69,2.18] 2.28*** [1.44,3.61]
Working Activity Change
As Usual 1.00 (ref)
Smart-Working 1.01 [0.91,1.12] 0.99 [0.86,1.14] 0.97 [0.85,1.10] 1.02 [0.90,1.15] 0.9 [0.74,1.10] 1.07 [0.91,1.25]
Discontinued 1.15** [1.05,1.27] 1.40*** [1.23,1.59] 1.16* [1.03,1.31] 1.19** [1.06,1.34] 1.22* [1.03,1.46] 1.1 [0.95,1.28]
More Than Usual 1.42*** [1.18,1.71] 1.26 [0.98,1.63] 1.25 [1.00,1.57] 1.71*** [1.38,2.12] 1.29 [0.93,1.80] 1.39* [1.04,1.87]
Loved One’s Condition
None 1.00 (ref)
Infected 1.22* [1.05,1.42] 1.05 [0.87,1.28] 0.91 [0.75,1.10] 0.88 [0.73,1.05] 1.02 [0.77,1.35] 0.96 [0.79,1.17]
Deceased 1.68*** [1.30,2.16] 1.41* [1.03,1.93] 1.22 [0.91,1.65] 1.34* [1.01,1.78] 1.74** [1.18,2.54] 1.21 [0.87,1.68]
Hospitalized 1.22 [1.00,1.48] 1.09 [0.84,1.41] 1.25 [0.99,1.57] 1.1 [0.87,1.39] 1.1 [0.76,1.60] 1.16 [0.91,1.49]
In A Relationship 1.14*** [1.06,1.22] 0.92 [0.84,1.00] 1.11* [1.02,1.22] 1.11* [1.02,1.21] 1.08 [0.94,1.23] 1.07 [0.97,1.19]
Education
Postgraduate 1.00 (ref)
Undergraduate 1.12** [1.04,1.20] 1.30*** [1.19,1.43] 1.28*** [1.18,1.39] 1.25*** [1.15,1.36] 1.31*** [1.15,1.50] 1.05 [0.95,1.16]
Lower Education 1.25*** [1.11,1.41] 1.62*** [1.40,1.87] 1.51*** [1.32,1.74] 1.47*** [1.28,1.69] 1.76*** [1.46,2.13] 1.21* [1.01,1.44]
Occupation
Employed 1.00 (ref)
Housewife 1.28*** [1.11,1.47] 1.35** [1. 12,1.63] 1.31** [1.11,1.55] 1.21* [1.03,1.44] 1.39** [1.11,1.74] 1.05 [0.83,1.32]
Unemployed 1.05 [0.94,1.17] 1.59*** [1.40,1.80] 1.39*** [1.23,1.57] 1.22** [1.08,1.37] 1.33** [1.12,1.58] 1.09 [0.93,1.27]
Retired 0.9 [0.66,1.22] 1.17 [0.79,1.75] 1.02 [0.69,1.51] 1.39 [0.96,2.01] 0.88 [0.52,1.48] 0.46* [0.22,0.97]
Student 0.79*** [0.71,0.88] 1.60*** [1.41,1.83] 1.02 [0.90,1.16] 1.28*** [1.13,1.44] 1.02 [0.86,1.22] 1.16 [0.84,1.62]
Childhood Trauma 1.06 [0.99,1.13] 1.41*** [1.30,1.54] 1.29*** [1.19,1.39] 1.01 [0.93,1.09] 1.50*** [1.33,1.70] 1.10* [1.01,1.21]
Prior Psychiatric Diagnosis 1.59*** [1.48,1.71] 2.19*** [2.01,2.39] 2.10*** [1.94,2.28] 1.73*** [1.59,1.87] 1.76*** [1.56,1.98] 1.25*** [1.13,1.39]
*p<0.05; **p<0.005; ***p<0.001; PTSS: Post-Traumatic Stress Symptoms; ADS: Adjustment Disorder Symptom; § Age is standardized and reversed, younger age has an OR>1 if associated with heightened risk.
. CC-BY-NC-ND 4.0 International licenseIt is made available under a
author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the(which was not peer-reviewed) The copyright holder for this preprint .https://doi.org/10.1101/2020.04.09.20057802doi: medRxiv preprint