Mental health of people in Australia in the first month of COVID‐19 restrictions: a national survey

Abstract Objectives To estimate the population prevalence of clinically significant symptoms of depression, generalised anxiety, thoughts of being better off dead, irritability, and high optimism about the future, and of direct experience of COVID‐19, loss of employment caused by COVID‐19 restrictions, worry about contracting COVID‐19, or major disadvantage because of the restrictions; to examine the relationship between these experiences and reporting mental symptoms. Design, setting, participants Anonymous online survey of adult Australian residents, 3 April – 2 May 2020. Main outcome measures Self‐reported psychological status during the preceding fortnight assessed with the Patient Health Questionnaire 9 (PHQ‐9; symptoms of depression) and the Generalised Anxiety Disorder Scale (GAD‐7). Optimism about the future was assessed with a 10‐point study‐specific visual analogue scale. Results 13 829 respondents contributed complete response data. The estimated prevalence of clinically significant symptoms of depression (PHQ‐9 ≥ 10) was 27.6% (95% CI, 26.1–29.1%) and of clinically significant symptoms of anxiety (GAD‐7 ≥ 10) 21.0% (95% CI, 19.6–22.4%); 14.6% of respondents (95% CI, 13.5–16.0%) reported thoughts of being better off dead or self‐harm (PHQ‐9, item 9) on at least some days and 59.2% (95% CI, 57.6–60.7%) that they were more irritable (GAD‐7, item 6). An estimated 28.3% of respondents (95% CI, 27.1–29.6%) reported great optimism about the future (score ≥ 8). People who had lost jobs, were worried about contracting COVID‐19, or for whom the restrictions had a highly adverse impact on daily life were more likely to report symptoms of depression or anxiety, and less likely to report high optimism than people without these experiences. Conclusions Mental health problems were widespread among Australians during the first month of the stage two COVID‐19 restrictions; in addition, about one‐quarter of respondents reported mild to moderate symptoms of depression or anxiety. A public mental health response that includes universal, selective and indicated clinical interventions is needed.


Research
MJA 213 (10) ▪ 16 November 2020 459 We defined scores of 8 or more as indicating great optimism, and scores of 0-7 as indicating no to moderate optimism about the future.

Experience of COVID-19 and COVID-19 restrictions
We asked whether respondents had direct experience of COVID-19 (diagnosis with or testing for COVID-19, or living with or knew someone with COVID- 19), or had lost employment because of COVID-19 restrictions; how worried they were about contracting COVID-19 (visual analogue scale, 0-10); and to what degree COVID-19 restrictions had affected daily life (visual analogue scale, 0-10).

Socio-demographic characteristics
We asked questions with fixed response options to ascertain age, sex, residential postcode, birthplace (Australia or overseas), living circumstances, and occupation. Information about state of residence, remoteness (urban or rural), and socio-economic position (Index of Relative Socio-economic Advantage and Disadvantage) were derived from respondents' postcodes using the most recent Australian Bureau of Statistics data. 4 Further details about the survey are included in the online Supporting Information.

Statistical analysis
Population prevalence rates with 95% confidence intervals (CIs) were estimated for mental health assessment parameters and experiences of COVID-19 and restrictions, adjusted for differences in selected socio-demographic characteristics (state, socio-economic position decile, sex, age) between the respondents and the Australian population at September 2019. 4 The characteristics of respondents with experiences of COVID-19 and related restrictions, and associations between mental health assessment parameters and experiences of COVID-19 and restrictions were assessed by multiple logistic regression, adjusted for selected socio-demographic characteristics.
All analyses were undertaken in Stata 16. Further details about the statistical analysis are included in the Supporting Information.

Ethics approval
Our investigation was approved by the Monash University Human Research Ethics Committee (reference, 2020-24080-42716).

Results
After excluding 80 respondents not resident in Australia and 33 under 18 years of age, there 460 were 15 121 eligible respondents; 13 829 (91.5%) contributed complete data and were included in our analyses. Respondents were drawn from all Australian states and territories, socio-economic positions, age groups, and living situations. Compared with the national population, similar proportions were born overseas, the proportion from Victoria was larger and that from New South Wales smaller, the proportion of women was higher, and the socio-economic position distribution was skewed to higher levels (Box 1).
Respondents living in Victoria, Queensland, Western Australia and the Australian Capital Territory were less likely to have had direct experience of COVID-19 than those from NSW. Respondents in areas of highest socio-economic position or born overseas were more likely to report direct experience of COVID-19; people who were at least 70 years old, retired, or caring for dependent relatives at home were less likely to have had direct experience of COVID-19 (Box 3).
Respondents in major cities were more likely to have lost jobs because of the restrictions than people in rural or regional areas; respondents aged 18-29 years were more likely than older respondents, and students more likely than those in paid positions, to have lost jobs. ACT residents were less likely than those in NSW to have lost jobs (Box 3).
High worry about contracting COVID-19 was most common among respondents in the lowest socio-economic quintile; it was more common among people who were unemployed, doing unpaid work caring for children or dependent relatives, or retired than among those in paid employment; and among those who did not identify as male or female. Respondents aged 18-29 years were less frequently worried about contracting COVID-19 than people in other age groups (Box 3).
Experiencing a high negative impact from COVID-19 restrictions was more likely for respondents in major cities than those in regional or remote areas; for people living alone; for those who were students or unemployed or doing unpaid work caring for children or dependent relatives than for respondents in paid employment; and for women than men (Box 3).

Associations between COVID-19 experiences and selfreported mental health symptoms
After adjusting for state, remoteness, socio-economic quintile of residence, sex, age group, living situation, place of birth, and employment status, people who had direct experience of COVID-19 were more likely to report clinically significant anxiety than those who had not. Respondents who had lost jobs and people who were very worried about contracting COVID-19 were more likely to report clinically significant symptoms of depression and anxiety, thoughts that they would be better off dead, and irritability. People for whom the restrictions had exerted a highly negative impact on daily life were more likely to report clinically significant signs of depression and anxiety, thoughts of self-harm, and increased irritability. Optimism was more common among people without direct experience of COVID-19, those who had not lost jobs, and people who did not find the COVID-19 restrictions too difficult (Box 5).

Discussion
We have reported the first estimates of population levels of clinically significant symptoms of depression and anxiety 462 among adults during the first month of COVID-19 restrictions in Australia. These data suggest a widespread change in the mental health of the Australian adult population. About onequarter of respondents reported mild to moderate symptoms of depression or anxiety, which is substantially higher than found by a survey of American adults during 2005-2008 (subthreshold depression, 17%) 5 or by a 2014 systematic review (median point prevalence of subthreshold anxiety, 4.4%). 6 The point prevalence of clinically significant symptoms of depression (27.6%) was much higher than reported for a randomly selected Australian adult population (aged 32-58 years; 3.7%) 7 and for other high income countries (3.3-10.8%; Supporting Information, table 3). Few estimates of the community prevalence of thoughts of being better off dead have been published; the 14.6% we found was much higher than the 1.8% for a random sample of South Australians in 2000. 8 The mental health symptoms we assessed are indicators of normal psychological adjustment to abnormal circumstances that challenge the adaptive capacity of individuals, and reduce their access to social support and opportunities for participation. Depression and thoughts of being better off dead are most likely when people experience loss and feel trapped, humiliated, and powerless. 9-11 "Disenfranchised grief" describes experiences of loss that might not be recognised by the individual affected or by others. 12,13 Everybody experienced some loss of liberty, autonomy, and agency as everyday activities were limited by the COVID-19 restrictions; privacy was affected by the close scrutiny of adherence to prescribed health behaviours that, paradoxically, required isolation. Many people missed events of lifetime significance: weddings, end-of-life support for loved ones, milestone celebrations. Occupational identity and the ability to earn an income are fundamental to individuality, sense of purpose, and autonomy in adults; their loss is a profound one, leading to demoralisation and depression. Unrecognised losses that do not attract the social support or rituals that follow bereavement can induce powerlessness rather than the problem-solving needed to manage psychological pain. Anxiety is increased by threat, danger, and uncertainty; the absence of definite knowledge about the evidence underlying specific restrictions and their duration contributed to uncertainty.
Our data indicate that some groups were especially vulnerable to mental health problems during the COVID-19 restrictions: women and people aged 18-29 years; people living in regional and rural areas or in the lowest socio-economic positions, and those not in paid employment before the pandemic; people who had lost jobs or opportunities for study; people living alone, who have fewer opportunities for daily interactions with family and friends; and people whose main occupation is 4 Respondents' self-assessment of mental health symptoms during the past two weeks Characteristic Number

Generalised Anxiety Disorder Scale (GAD-7)
Total score, mean (95% CI) 5.5 (5.3-5.7) Moderate/ severe (clinically significant) symptoms (score, ≥ 10) The consequences for occupational and social functioning are highly relevant to national recovery. People with these mental health problems are less motivated, energetic, socially engaged, and confident, and less able to concentrate, plan, organise, or trust. A public health approach has been essential to containing COVID-19, and our findings indicate that a public mental health approach is needed for recovery. 14 This would include universal interventions for the entire population, selective strategies for people with psychological problems, and indicated interventions for those with specific risks or needs.
As the mental health problems we report were associated significantly with the perceived risk of contracting COVID-19 and the consequences of the restrictions, some improvement is expected as the pandemic ebbs and restrictions are lifted. However, universal, psychologically informed mental health strategies will still be needed. Strict messages and public policies ensured that distancing and isolation restrictions were observed; political and civic leaders acknowledging the magnitude and the psychological costs of individual contributions to the public good might now be helpful in reducing social suffering. 15 Social relationships are predicated on trust, but when everyone is suspected of being able to transmit disease, trust in relationships is diminished. Activities providing engagement with other people offer essential opportunities for discussing life situations, experiencing empathy, and exploring solutions. Experiencing empathy is less likely during social media interactions than in personal encounters. 16 Clear messages about safe social engagement with others are needed to reassure people, particularly those who are living alone or are afraid of contracting the disease. Government and non-government agencies accompanied the restrictions with advice about the benefits for mental health of maintaining routines, social connections, and exercise, and the potential harms of isolation, lack of access to purposeful activities, and increased alcohol consumption. 17 Similar guidance is needed about recapturing agency and resuming healthy social and economic participation, and the need for an adjustment period. 18,19 The mental health of people who have lost jobs will benefit from empathic, courteous, and encouraging assistance that does not rely exclusively on their own initiative for finding employment.
Strengthening the psychological skills of staff and embedding mental health workers in employment agencies would be more effective than expecting people seeking work to also attend health services.
Our survey findings suggest that it would be appropriate for primary care clinicians to assess symptoms of depression and anxiety, and ideas of self-harm, in people in the more vulnerable groups we have identified. Increased access to mental health care should be provided to people whose symptoms are not ameliorated by universal or selective mental health promotion strategies. Telehealth consultations should be recommended with caution; they require internet access, an appropriate personal device, and privacy, none of which are assured for people in lower socio-economic positions. Integrating mental health care into community services can reduce barriers to access.

Strengths and limitations
Our respondents comprised a large and diverse sample of people in Australia; we weighted their responses according to the characteristics of the national population, employed standardised psychometric measures that permit comparisons with other populations, and could distinguish worry about contracting COVID-19 from the impacts of restrictions.
As participation in the survey was self-selective, the representativeness of the sample cannot be assessed, nor a response rate calculated. As for all online surveys, it was less accessible to people with lower computer proficiency or English fluency, without internet access, or in lower socio-economic positions; their experiences may have been underestimated. On the other hand, although the survey was advertised in neutral terms, people with mental health problems may have been more likely than others to complete it, leading to overestimation of symptom prevalence. A short, structured survey cannot gather nuanced information about mental health, and our data are not diagnostic; estimates of symptom prevalence based on self-reports are generally higher than those based on clinical interviews. 20 While thoughts of being better off dead were assessed with a single question, we did not assess suicide intent. Cross-sectional surveys identify associations, not causal relationships. Nevertheless, our data provide an indication of the consequences of the first month of restrictions for the mental health of Australians and could inform public health planning and clinical service responses.

Conclusion
The United Nations' policy brief, COVID-19 and the need for action on mental health (May 2020), concluded that the pandemic is leading to a "major mental health crisis", and that mental health is a priority for which each country must urgently plan a response. 21