Acute mental health responses during the COVID-19 pandemic in Australia

The acute and long-term mental health impacts of the COVID-19 pandemic are unknown. The current study examined the acute mental health responses to the COVID-19 pandemic in 5070 adult participants in Australia, using an online survey administered during the peak of the outbreak in Australia (27th March to 7th April 2020). Self-report questionnaires examined COVID-19 fears and behavioural responses to COVID-19, as well as the severity of psychological distress (depression, anxiety and stress), health anxiety, contamination fears, alcohol use, and physical activity. 78% of respondents reported that their mental health had worsened since the outbreak, one quarter (25.9%) were very or extremely worried about contracting COVID-19, and half (52.7%) were worried about family and friends contracting COVID-19. Uncertainty, loneliness and financial worries (50%) were common. Rates of elevated psychological distress were higher than expected, with 62%, 50%, and 64% of respondents reporting elevated depression, anxiety and stress levels respectively, and one in four reporting elevated health anxiety in the past week. Participants with self-reported history of a mental health diagnosis had significantly higher distress, health anxiety, and COVID-19 fears than those without a prior mental health diagnosis. Demographic (e.g., non-binary or different gender identity; Aboriginal and Torres Strait Islander status), occupational (e.g., being a carer or stay at home parent), and psychological (e.g., perceived risk of contracting COVID-19) factors were associated with distress. Results revealed that precautionary behaviours (e.g., washing hands, using hand sanitiser, avoiding social events) were common, although in contrast to previous research, higher engagement in hygiene behaviours was associated with higher stress and anxiety levels. These results highlight the serious acute impact of COVID-19 on the mental health of respondents, and the need for proactive, accessible digital mental health services to address these mental health needs, particularly for those most vulnerable, including people with prior history of mental health problems. Longitudinal research is needed to explore long-term predictors of poor mental health from the COVID-19 pandemic.

If the data are held or will be held in a public repository, include URLs, accession numbers or DOIs. If this information will only be available after acceptance, indicate this by ticking the box below. For example: All XXX files are available from the XXX database (accession number(s) XXX, XXX.).
• If the data are all contained within the manuscript and/or Supporting Information files, enter the following: All relevant data are within the manuscript and its Supporting Information files. The acute and long-term mental health impacts of the COVID-19 pandemic are unknown. The 20 current study examined the acute mental health responses to the COVID-19 pandemic in 5070 adult 21 participants in Australia, using an online survey administered during the peak of the outbreak in Australia 22 (27 th March to 7 th April 2020). Self-report questionnaires examined COVID-19 fears and behavioural 23 responses to COVID-19, as well as the severity of psychological distress (depression, anxiety and stress), 24 health anxiety, contamination fears, alcohol use, and physical activity. 78% of respondents reported that 25 their mental health had worsened since the outbreak, one quarter (25.9%) were very or extremely worried 26 about contracting COVID-19, and half (52.7%) were worried about family and friends contracting COVID-27 19. Uncertainty, loneliness and financial worries (50%) were common. Rates of elevated psychological 28 distress were higher than expected, with 62%, 50%, and 64% of respondents reporting elevated depression, 29 anxiety and stress levels respectively, and one in four reporting elevated health anxiety in the past week. 30 Participants with self-reported history of a mental health diagnosis had significantly higher distress, health 31 anxiety, and COVID-19 fears than those without a prior mental health diagnosis. Demographic (e.g., non-32 binary or different gender identity; Aboriginal and Torres Strait Islander status), occupational (e.g., being a 33 carer or stay at home parent), and psychological (e.g., perceived risk of contracting COVID-19) factors were 34 associated with distress. Results revealed that precautionary behaviours (e.g., washing hands, using hand 35 sanitiser, avoiding social events) were common, although in contrast to previous research, higher 36 engagement in hygiene behaviours was associated with higher stress and anxiety levels. These results 37 highlight the serious acute impact of COVID-19 on the mental health of respondents, and the need for 38 proactive, accessible digital mental health services to address these mental health needs, particularly for 39 those most vulnerable, including people with prior history of mental health problems. Longitudinal research 40 is needed to explore long-term predictors of poor mental health from the COVID-19 pandemic. 41 ramifications (e.g., unemployment, social isolation), along with fears of COVID-19 are likely to have 48 significant and long-term impacts on the mental health of the community. Research into past pandemics, 49 such as the 2003 outbreak of Severe Acute Respiratory Syndrome (SARS), has shown higher rates of illness 50 fears, psychological distress (e.g., depression, anxiety, stress), insomnia and other mental health problems 51 (e.g., posttraumatic stress) in people with pre-existing mental illness, front-line health care workers (2), and 52 survivors of severe and life-threatening cases of the disease (3-6). 53 High quality research into the mental health impacts of COVID-19 is urgently needed (7) to inform 54 evidence-based policy decisions, prevention efforts, treatment programs and community support systems, 55 particularly for those who are most vulnerable and those who are at risk of experiencing poor mental health 56 outcomes during and after this pandemic. In marked contrast to the rapidly growing literature into the 57 physical health consequences of COVID-19, there is currently limited information about the mental health 58 impacts of the COVID-19 outbreak in the general population. However, some recent research has emerged 59 from China with community participants (8-10), and health care worker samples (11). In a cross-sectional 60 survey of 52,730 participants in China conducted between the 31 st January to the 10 th February 2020 (10), 61 29.3% of respondents experienced mild to moderate psychological distress, and 5.1% experienced severe 62 distress. In another survey of 1210 members of the general public (half of whom were students) conducted 63 between 31 st January to 2 nd February 2020, Wang et al. (8) found that over half (53.8%) of participants rated 64 the psychological impact of the COVID-19 outbreak as moderate to severe, three quarters were worried 65 about their family members contracting COVID-19, and rates of moderate to severe depression, anxiety and 66 stress were 16.5%, 28.8%, and 8.1% respectively. In a follow-up survey four weeks later, rates of 67 depression, anxiety and stress remained unchanged (12). In another survey of 7236 self-selected volunteers 68 from 3 rd to 17 th February 2020, Huang & Zhao (13) found that 20.1%, 35.1%, and 18.2% of respondents 69 reported symptoms of depression, generalised anxiety disorder (GAD), and insomnia on self-report 70

measures. 71
Together these studies demonstrate the elevated psychological distress in the general community 72 during the initial COVID-19 outbreak in China. These studies also give some early insights into factors that 73 may increase a person's vulnerability to experiencing poor mental health during the pandemic. Preliminary 74 evidence suggests that i) demographic factors (younger participants, females, college students, and those 75 with low educational attainment), ii) occupational factors (migrant workers, nurses), iii) health-related 76 factors (history of chronic illness, poor self-rated health (8)), and iv) greater exposure to COVID-19 and the 77 worst affected regions of the outbreak (10), are associated with higher distress levels. In contrast, engaging 78 in precautionary behaviours (e.g., hand hygiene, wearing a mask) have been associated with lower distress 79 (8, 12). As COVID-19 has spread to communities outside of China, more research is urgently needed to 80 explore the mental health impacts of the outbreak, and to identify groups who are vulnerable to poorer 81 mental health in other countries. 82 To our knowledge there are no published findings on the mental health of the general community 83 during the COVID-19 pandemic in Australia. However, we conducted a previous online survey of the 84 knowledge, attitudes, behaviours and risk perceptions of 2174 people from the general community, shortly 85 after the first death occurred from COVID-19 and when confirmed COVID-19 cases were low in Australia 86 (March 2 nd -9 th 2020) (14). In that study, we found one in three participants were very or extremely 87 concerned about an outbreak, and that participants perceived their risk of personally contracting COVID-19 88 as relatively high (rated as 70% likelihood of contracting the virus). Moreover, most participants (61%) 89 expected that they would experience moderate to severe symptoms of COVID-19 if they contracted the 90 virus. We did not measure mental health outcomes, or how afraid individuals were of personally contracting 91 COVID-19. Therefore, the current study extended our previous survey and investigated the mental health of 92 Australian residents during a 12-day period from 27 th March to 7 th April 2020, which is now considered to 93 be the time of the peak in new cases, and the steady decline in new cases. Three days prior to recruitment, an 94 international travel ban had been implemented in Australia, and from 28 th March 2020, all travellers arriving 95 in Australia from overseas were required to undergo a mandatory 14-day quarantine in designated 96 accommodation. On the first day (27 th March) of the study recruitment period, there was a total of 3378 97 confirmed cases and 13 deaths related to COVID-19 in Australia, with 328 new cases diagnosed on the 27 th 98 March. Over the next two days, there was an increase of 785 new cases in Australia. Finally, over the 99 remaining days of the study, the number of daily new cases steadily declined, with 93 new cases reported on 100 the last day of recruitment (7 th April 2020). There was a total of 5988 confirmed cases (including 3392 101 active cases) and 49 deaths at the end of the survey period. 102 Drawing from past research (8, 10, 12) we assessed demographic characteristics, fears of COVID-19, 103 risk perceptions and behavioural responses to the outbreak, psychological distress (depression, anxiety, 104 stress), and alcohol use. We included measures of health anxiety and contamination fears due to their 105 potential role in influencing behaviour, health service use, and anxious reactions to viral outbreaks (15-18), 106 as well as physical activity levels, and loneliness, due to the expected negative impacts of social distancing 107 measures on these variables, and due to their important role in mental and physical health (19, 20). Finally, 108 we assessed financial worries, as we expected unemployment, and financial insecurity, which have already 109 resulted from this outbreak, to have significant, negative impacts on mental health (7, 21). Our primary aim 110 was to provide the first snapshot of the mental health of the general community during the initial COVID-19 111 outbreak (and enforcement of social distancing laws) in Australia. The second aim was to explore the 112 relationship between specific demographic and sample characteristics with depression, anxiety and stress, to 113 identify factors that are associated with increased vulnerability for poorer mental health during the COVID-114 19 pandemic. While we acknowledge that the data from an online survey may not be representative of the 115 entire population, they provide an important opportunity to (i) identify vulnerable groups who are risk of 116 poorer mental health during COVID-19, (ii) determine the socio-demographic and psychological factors that 117 predict psychological distress, and (iii) examine whether the findings from past pandemics, and from China, 118 apply to the Australian context during the COVID-19 pandemic. Based on research from past pandemics, 119 and Chinese research, we expected that between 20-35% would worry about contracting COVID-19 and 120 experience elevated psychological distress, and that specific demographic variables including younger age, 121 being a student, unemployed, female, or with lower educational attainment would predict higher distress 122 levels in the current cohort. We also expected people with lived experience of prior mental health diagnoses 123 would have higher rates of distress and would be vulnerable to poorer mental health during the current 124 pandemic. Finally, we predicted that engaging in precautionary hygiene behaviours would be associated 125 with lower distress. Participants were asked to complete single item measures of i) how lonely they were feeling, ii) how 159 worried they were about their financial situation, and iii) how uncertain they were feeling about the future, the past week they engaged in moderate to strenuous activity, and ii) the average number of minutes they 172 exercised at this level, and screened for hazardous alcohol use using the Modified Alcohol Use Disorders 173 Identification Test (AUDIT-C; 28). All questionnaire responses were anchored to the past week, except for 174 the AUDIT-C (past month), and the Padua contamination subscale (general). The mental health and lifestyle 175 questionnaires were administered in randomised in order to minimise responding biases. 176

COVID-19 Variables, Fears and Perceived Risk 177
Participants were asked about their own COVID- participants were asked about how much information they had seen, read or heard about coronavirus 196 (nothing at all, a little, a moderate amount, a lot). 197 Health-Protective Behaviours 198 To assess social distancing, hygiene and buying behaviours, participants were asked whether they 199 had engaged in a total of 16 behaviours during the previous week (see Table 2). Health-Protective Behaviours 242 The percentage of respondents who reported having engaged in a range of distancing and hygiene 243 behaviours during the past week is presented in Table 2. During the previous week, handwashing and social 244 distancing (avoiding social events and gatherings) were the most common behaviours. 245 Note. Numbers represent n and proportion (%) in brackets.

Mental Health 246
More than three quarters of participants reported that their mental health had been worse since the 247 outbreak, with 55.1% selecting 'a little worse', and 22.9% selecting 'a lot worse'. A small proportion 248 reported improvements in their mental health since the outbreak (5.5%) (see Figure 1). A chi square analysis 249 revealed that there was a significant difference in the impact of COVID-19 on mental health for participants 250 with and without a prior mental health diagnosis (ꭓ 2 (4) = 141.44, p <.001), with 26.6% of those with a 251 prior mental health diagnosis saying their mental health had been 'a lot worse', relative to 13.4% in the 252 group without a mental health diagnosis.   Hazardous drinking levels were defined as an AUDIT-C score of 3 or more for women and other genders, 265 and 4 or more for men (28, 29). 266 Comparison between people with and without prior mental health diagnosis 267 People with and without a self-reported history of mental health diagnosis were compared in their severity of 268 COVID-19 fears, mental health, distress, health anxiety, alcohol use, contamination fears, and physical 269 activity. People with a previous self-reported mental health diagnosis reported higher uncertainty, loneliness, 270 financial worries, COVID-19 fears (self and others), believed they were more likely to contract COVID-19, 271 had lower perceived behavioural control, had higher rates of psychological distress, health anxiety and 272 contamination fears, and lower physical activity than those without a self-reported mental health diagnosis 273 history. There were no differences in alcohol use between these groups (see Table 4). 274 Impact of self-isolation: Compared to people who were not in self isolation, people who self-reported being 275 in self-isolation reported higher uncertainty, loneliness, financial worries, and COVID-19 fears (self and 276 others), rated the symptoms of COVID-19 as more serious, but believed they were less likely to contract 277 COVID-19, and perceived more behavioural control over COVID-19. They also had higher rates of 278 psychological distress, health anxiety and contamination fears, and lower alcohol use than those not in 279 isolation. There were no differences in physical activity between these groups (see Table 5). 280

Predictors of Depression, Anxiety and Stress 281
Separate linear regression analyses were conducted to explore the demographic, occupational, and 282 psychological predictors of DASS-21 depression, anxiety and stress severity (see final model in Table 8). 283 We entered demographic predictor variables (gender, age, occupational status, education, Aboriginal and/or 284 Torres Strait Islander and carer status) in the first step. In the second step, we entered general health 285 variables including chronic illness, mental health diagnosis history, and self-rated health. In the third step, 286 we entered uncertainty about the future, loneliness, worry about finances. In the final step, we added 287 COVID-19 variables (whether they were in self-isolation, hygiene behaviours, exposure to COVID-19 288 information, risk perceptions including perceived likelihood, perceived control, and severity of illness, 289 concern/worry about contracting COVID-19, and concern/worry about loved ones contracting COVID-19. 290 Depression. Demographic Table 8 and accounted for 48.5% of the variance in depression scores. 297 Controlling for the other variables in the model, being female, more well educated, older, and having better 298 self-rated health were all associated with lower depression, whereas being unemployed, a student, retired, 299 carer or stay at home parent were associated with higher depression. Mental health and chronic illness 300 diagnoses were associated with higher depression, as were increased uncertainty about the future, loneliness, 301 and financial worries. Of the COVID-19 variables, higher worry about COVID-19 and perceived 302 behavioural control over COVID-19 infection were associated with lower depression, whereas perceiving 303 higher illness severity was associated with higher depression. 304 Anxiety. Controlling for other variables in the model, being female, non-binary or different gender identity, and being 313 Aboriginal and/or Torres Strait Islander were predictors of higher anxiety. Older age, and more well 314 educated (certificate, degree or higher) were predictors of lower anxiety. In contrast to depression, only 315 being a student predicted worse anxiety. Having a chronic illness, and prior history of mental health 316 diagnosis were associated with higher anxiety, whereas better self-rated health was a predictor of lower 317 anxiety. Similar to depression, increased uncertainty about the future, loneliness, and financial worries were 318 also associated with higher anxiety. Of the COVID-19 variables, more hygiene behaviours, worry about 319 COVID-19, worry about loved ones contracting COVID-19, and higher perceived illness severity were 320 predictors of higher anxiety, whereas increased exposure to COVID-19 information, and perceived control 321 over COVID-19 predicted lower anxiety. 322 Stress. In the first step, demographic variables accounted for 10.8% of the variance in anxiety scores (R 2  323  Controlling for other variables in the model, identifying as non-binary or different gender identity, 331 Aboriginal and/or Torres Strait Islander, predicted higher stress. Being more well-educated with a trade 332 certificate, and older age, were predictors of lower stress. Being a stay at home parent was a predictor of 333 higher stress. Having a chronic illness, and prior history of mental health diagnosis were associated with 334 higher stress, whereas better self-rated health was a predictor of lower stress. Increased uncertainty about the 335 future, loneliness, and financial worries were also associated with higher stress. Of the COVID-19 variables, 336 more hygiene behaviours, worry about loved ones contracting COVID-19, and higher perceived likelihood 337 of contacting COVID 19 were predictors of higher stress. Higher perceived control over COVID-19 338 predicted lower stress. 339 To rapidly respond to the evolving COVID-19 situation, we administered online validated self-report 355 questionnaires rather than diagnostic interviews. It is important to note that these questionnaires assessed 356 symptoms of distress during the past week and should not be taken as indicative of a 'diagnosis' of a 357 depressive or anxiety disorder. We found higher than expected levels of acute distress based on research in 358 China during the COVID-19 pandemic (8), and compared to normative data (22, 31). Between 20.3-24.1% 359 of the current sample were experiencing severe or extremely severe levels of depression, anxiety and stress, 360 and a further 18-22% moderate symptoms. Only 38% of the current sample had normal depression, 50% had 361 normal anxiety, and 46% had normal stress levels, whereas in the Chinese sample reported by Wang et al. 362 (8) 64-69% had normal anxiety, stress and depression on the DASS-21. These differences may be due to the 363 high proportion of people with pre-existing mental health diagnoses (70%) in our sample, which have been 364 shown to be a vulnerable group (8, 10), or because of the significant proportion with a self-reported chronic 365 illness (38%), who may be more susceptible to more severe COVID-19 disease, and therefore more 366 distressed. Having a personal history of chronic illness was a consistent predictor of higher depression, 367 anxiety and stress, whereas better self-rated health was associated with better mental health. Compared to 368 the Australian population, this sample appeared to have poorer health, with 30% reported being in fair or 369 poor health (compared to 15% in the Australian population), and 30% reporting being in very good or 370 excellent health (compared to 56% of Australians) (32). 371 Our data gave some insights into other demographic variables which predict higher psychological 372 distress. Specific occupational factors predicted higher distress levels: student status (depression and  373 anxiety), being an at home parent (depression and stress), a carer or retired (predicted higher depression), 374 whereas education was associated with lower psychological distress. In contrast to past research, identifying 375 as female predicted lower depression, however identifying as non-binary or a different gender identity was 376 associated with higher self-reported anxiety and stress. Identifying as Aboriginal or Torres Strait Islander 377 also predicted worse anxiety and stress levels. These groups may be particularly vulnerable during the 378 current pandemic, and longitudinal research is needed to explore the longer term predictors of poorer mental 379 health over time.

380
Our results confirm fears about the potential impact of the COVID-19 pandemic on people with lived 381 experience of mental illness (7). Participants with a self-reported history of mental health problems were 382 more afraid of COVID-19 and more worried about their loved ones contracting COVID-19, had higher 383 distress, depression, anxiety, health anxiety and contamination fears, and higher rates of elevated health 384 anxiety (26% versus 11%) than those without pre-existing mental health diagnoses. Relative to those 385 without mental health issues, a greater proportion of people with self-reported mental health problems had 386 elevated health anxiety (26% versus 11%), and said their mental health had been 'a lot worse' since the 387 outbreak (26% versus 13%). Having a history of mental health issues was a consistent predictor of higher 388 depression, anxiety and stress. 389 Because we did not collect any information about the history and nature of these mental health 390 diagnoses, we cannot determine whether these individuals had higher distress prior to the pandemic, or 391 whether distress increased as a result of the pandemic, due to inability to access usual supports, social 392 isolation or loneliness (7). However, our findings highlight the need for proactive mental health 393 interventions for those who are experiencing elevated symptoms of depression, anxiety and stress during the 394 current COVID-19 pandemic, regardless of whether the distress is an exacerbation or recurrence of pre-395 existing mental health concerns, or new onset. Digital interventions, which have been shown to be highly 396 effective and cost-effective for depression and anxiety treatment (33) will be crucial to respond to these 397 ongoing mental health concerns, as they have capacity to deliver high quality interventions for distress at 398 scale, and to those in social isolation who are unable to attend face-to-face services (7, 34). 399 This study provides new knowledge about the rates of health anxiety during the COVID-19 400 pandemic. Over one in four (26%) of people with a prior history of mental health issues, and 11% of those 401 without pre-existing mental health issues reported elevated health anxiety in the past week, which is higher 402 than rates of health anxiety in the general Australian population (3.4% (35)), and closer to the rates of health 403 anxiety observed in general practice (10%) and outpatient medical clinic settings (20-25%) (36 Participants rated on average that there was a 50% likelihood of contracting the virus personally, and higher 415 perceived risk was associate with higher depression and stress levels. In the current cohort approximately 416 one third of participants expected COVID-19 to lead to severe symptoms (32.1%), and in some cases death 417 (4%), which is higher than in our previous study, where we found only 25% expected severe symptoms. The 418 expected severity of the COVID-19 illness differs markedly to the reality for most people, as studies show 419 that 80% of people will experience no or mild symptoms (40). These findings reinforce the need for 420 education campaigns to address these misperceptions, especially as research has shown that these beliefs are 421 associated with engagement with distress. These risk perceptions explained a relatively small amount of 422 variance in the regression analyses, with perceived control over COVID-19 a consistent predictor of better 423 mental health and higher perceived severity of illness associated with higher depression and anxiety. 424 However, it is important to note that other predictors, including loneliness, financial stress, uncertainty, 425 demographic factors, and prior history of mental and chronic illness were stronger predictors of distress. 426 427 Similar to Wang et al. (8), some of the most common precautionary behaviours were avoiding 428 touching objects that had been touched by others, washing hands, and using hand sanitiser. Participants also 429 commonly reported staying at home and avoiding social events and socialising with others outside of the 430 household. In contrast to media portrayals of panic buying, excessive purchasing behaviour was not 431 common. In previous research, higher engagement in hygiene behaviours, such as handwashing have been 432 associated with lower distress and anxiety, suggesting behavioural control may be protective for mental 433 health. However, in the current cohort we found some inconsistent results, with engagement in more hygiene 434 behaviours associated with higher anxiety and stress levels (they were not associated with depression). 435 These where the use of precautionary measures, such as avoiding sharing utensils, hand hygiene and wearing 437 masks were associated with lower stress, anxiety and depression. However, the current findings are 438 consistent with some research from the SARS epidemic, in which moderate levels of anxiety were 439 associated with higher uptake of precautionary behaviours (41). It is possible that the association we found 440 was due to people who were higher in anxiety or stress using these behaviours in an attempt to control 441 anxiety. 442 Finally, concerns have been raised about the potential impact of social isolation and quarantine on 443 physical inactivity, as well as increased alcohol use and abuse. On the AUDIT-C brief screener for alcohol 444 use, approximately 52.7% met criteria for hazardous drinking levels, which is higher than the 42% found in 445 primary care samples in Australia (42) and higher than USA-based population samples (35 %-45%) (43). 446 However it is important to note that participants with a prior experience of mental health problems had 447 lower rates of hazardous drinking, and lower rates of inactivity. In the current sample, 42.7% met the 448 national physical activity recommendations of 150 minutes or more of moderate to vigorous activity over 449 the past week, which are similar to the population based normative data from the Australian National Health 450 survey (43-44%) (32). We will be following up these participants longitudinally to explore whether activity 451 levels decrease further as isolation restrictions proceed. Given the importance of exercise and physical 452 activity in maintaining mental health and promoting overall health and wellbeing, interventions could be 453 used to assist increasing activity levels for those sedentary at home. 454 Limitations 455 The results are based on a convenience sample recruited online, who were mostly women (85%) and 456 well educated, and a significant proportion reported having lived experience of a mental health diagnosis 457 (70%). This may overestimate the symptom severity and impact of COVID-19, especially given past studies 458 have shown worse impact of pandemics on those with pre-existing mental illness, and in females. It may 459 also mean that the results cannot generalise to the broader Australian population. Results are also based 460 solely on validated self-report measures, due to their ease and speed of assessment, and administration. 461 Conducting diagnostic interviews to assess mental health diagnoses with more than 5000 participants in 10 462 days would not have been feasible. Future studies need to explore the impact of COVID-19 on mental health 463 of COVID-19 patients, given evidence of increased rates of Post -Traumatic Stress Disorder, sleep 464 disturbance and depression in SARS patients (5, 44). Finally, the study was cross-sectional; the next step in 465 our research is to track this cohort over time, to explore how their mental health changes as the pandemic 466 evolves in Australia. 467 468