The impact of COVID-19 pandemic on physical and mental health of Asians: A study of seven middle-income countries in Asia

The coronavirus disease (COVID-19) pandemic has impacted the economy, livelihood, and physical and mental well-being of people worldwide. This study aimed to compare the mental health status during the pandemic in the general population of seven middle income countries (MICs) in Asia (China, Iran, Malaysia, Pakistan, Philippines, Thailand, and Vietnam). All the countries used the Impact of Event Scale–Revised (IES-R) and Depression, Anxiety and Stress Scale (DASS-21) to measure mental health. There were 4479 Asians completed the questionnaire with demographic characteristics, physical symptoms and health service utilization, contact history, knowledge and concern, precautionary measure, and rated their mental health with the IES-R and DASS-21. Descriptive statistics, One-Way analysis of variance (ANOVA), and linear regression were used to identify protective and risk factors associated with mental health parameters. There were significant differences in IES-R and DASS-21 scores between 7 MICs (p<0.05). Thailand had all the highest scores of IES-R, DASS-21 stress, anxiety, and depression scores whereas Vietnam had all the lowest scores. The risk factors for adverse mental health during the COVID-19 pandemic include age <30 years, high education background, single and separated status, discrimination by other countries and contact with people with COVID-19 (p<0.05). The protective factors for mental health include male gender, staying with children or more than 6 people in the same household, employment, confidence in doctors, high perceived likelihood of survival, and spending less time on health information (p<0.05). This comparative study among 7 MICs enhanced the understanding of metal health in the general population during the COVID-19 pandemic.


Introduction
Emerging psychiatric conditions and mental well-being were identified as the tenth most frequent research topic during the COVID-19 pandemic [1]. A recent systematic review found that relatively high rates of symptoms of anxiety, depression, post-traumatic stress disorder and stress were reported in the general population and health care professionals during the COVID-19 pandemic globally [2,3]. Asia has a number of middle income countries (MICs) that face tremendous economic challenges and limited medical resources to maintain physical and mental well-being during the pandemic [4]. This extended to North America as well, with the sudden change in economic security during COVID-19 projected to increase suicide rates [5]. During the pandemic, the Asia Pacific Disaster Mental Health Network recommended to establish a mental health agenda for Asia [6]. It is therefore important to conduct research to assess psychiatric status of Asians living in MICs to develop capacity of various health systems to respond to COVID-19. Previous studies mainly focused on mental health of individual Asian countries during the pandemic without cross comparison [7][8][9].
With no prior comparative study found on physical and mental health of Asians living in MICs during the COVID-19 pandemic, this study aimed to investigate the impact of the pandemic on physical and mental health in 7 Asian MICs (China, Iran, Malaysia, Pakistan, Philippines, Thailand and Vietnam), identify differences among countries, understand their concerns and precautions toward COVID-19, as well as to identify protective and risk factors associated with mental health outcomes.

Study design and study population
This was a cross-sectional study that involved seven countries. The recruitment was conducted after COVID-19 became an epidemic in each country. To minimize risks of COVID-19 infection, a respondent-driven sampling strategy on recruiting the general public was utilized where new participants were electronically invited by existing study respondents rather than face-to-face interaction. The respondents completed the questionnaires through an online survey platform ('SurveyStar', Changsha Ranxing Science and Technology in China, SurveyMonkey in Philippines, and Google Forms in other countries).
Thailand (Chulalongkorn University, COA No. 147/2563), and Vietnam (Hanoi Medical University, QD 75/QD-YHDP&YHDP). All IRBs allowed participants aged 12 years to 17 years to participate in this study and provide their own consent because the online survey did not pose any risk to research participants. All respondents provided informed consent. Confidentiality was maintained because no personally identifiable information was collected.

Measures and instruments
The COVID-19 online questionnaire designed by the National University of Singapore [10] had five sections: demographic, physical symptoms related to COVID-19 in the past 14 days, knowledge and concerns about COVID-19, precautionary measures against COVID, and views of health information required. Psychometric properties of the questionnaire were established in the initial phase and peak of the COVID-19 epidemic [8,9].
The psychological impact of COVID-19 was measured using the well-validated Impact of Event Scale-Revised (IES-R) in the Asians for determining the extent of psychological impact after exposure to a traumatic event (i.e., the COVID-19 pandemic) within one week of exposure [11][12][13][14]. In this study, the Cronbach's alpha for different versions of IES-R is very high in all countries and ranges from 0.912-0.950. Cronbach's alpha of 0.70 or higher in measuring the internal consistency is considered "acceptable" in most social science research [15].
The mental health status of respondents was measured using the Depression, Anxiety and Stress Scale (DASS-21) [16], which has been used to assess mental health in Asians [17,18]. Furthermore, DASS-21 assessed three domains (i.e. anxiety, depression and stress) and its psychometric properties was validated across clinical and non-clinical samples in different cultures and languages during the COVID-19 pandemic [19]. In this study, the Cronbach's alpha (internal consistency) for different versions of DASS-21 is as follows: stress scale ranges from 0.839-0.934, anxiety scale ranges from 0.784-0.914, and depression scale ranges from 0.878-0.943. The IES-R and DASS-21 scales were previously used in research related to the COVID-19 epidemic [8,12,20,21]. The DASS and IES-R questionnaires are available in the public domain, and so permission is not required to use these two questionnaires [22,23].

Statistical analysis
Descriptive statistics were calculated to compare demographic characteristics, physical symptoms and health service utilization, contact history, knowledge and concern, precautionary measure and additional health information variables among 7 MICs. One-Way analysis of variance (ANOVA) was calculated to compare the mean IES-R and DASS-21 scores between 7 MICs in order to determine whether the associated population mean IES-R or DASS-21 scores were significantly different. If there were significant differences among 7 MICs, the Least Significant Difference (LSD) would calculate the smallest significant between mean scores of two countries with different combinations. Any difference larger than the LSD is considered a significant result. We used linear regressions to calculate the univariate associations between independent and dependent variables including the IES-S score and DASS-21 stress, anxiety and depression subscale scores for all respondents separately. All tests were two-tailed, with a significance level of p <0.05. Statistical analysis was performed on IBM SPSS Statistics version 21.0.

Results
A total of 4479 participants from 7 MICs in Asia completed the survey. The distribution of the number of participants by country is listed as follows: China (27%), Philippines (19%), Malaysia (16.2%), Iran (12.3%), Thailand (11.6%), Pakistan (11.3%), and Vietnam (2.7%). Fig  1 compares  More than half of participants stayed in a household with more than 3-5 people across all countries except Pakistan (49%). Majority of respondents from Philippines, Pakistan, Vietnam and Malaysia were employed when the study was conducted. Table 1 shows the association between demographic characteristics of all participants and mental health parameters. Demographic characteristics associated with lower psychological impact were male gender whereas age younger than 30 years and students were associated with higher psychological impact. Participants who have children were associated with lower stress, anxiety and depression whereas participants with higher education, single and separated status were associated with higher stress, anxiety and depression. Staying with 6 or more people and those who were employed were associated with lower anxiety and depression.      Table shows the frequency of physical symptoms that resemble COVID-19 infection and there were significant differences among all countries. During the COVID-19 pandemic, the most common physical symptoms reported by general population in the 7 countries were headache (23.13%), cough (21.86%) and sore throat (19.29%). About 8.13% of respondents consulted General Practitioner (GP); 2.69% were hospitalized; 3.89% were tested positive for COVID-19 and 57.1% had a health insurance. Pakistani had the significantly highest proportion of respondents consulted GP (27.5%), hospitalized (16.4%), receiving COVID-19 test (17.2%) and being isolated (17.8%). Table 2 shows the association between physical symptoms and mental health outcomes. The physical symptoms that were significantly associated with higher scores in all mental health outcomes (IES-R and DASS-21 subscales) including rhinitis and persistent fever with cough or breathing difficulties. Chills or rigors, headache and nausea or vomiting were associated with higher DASS-21 stress and anxiety scores. Myalgia, cough, dizziness and sore throat were associated with higher score of IES-R. Usage of medical services such as seeing a doctor, hospitalization, recent COVID-19 testing, quarantine, poor rating of health status that were significantly associated with higher scores in all mental health outcomes (IES-R and DASS-21 subscales). History of chronic illness were significantly associated with higher DASS-21 subscale scores. Having medical insurance coverage was associated with higher IES-R scores.
S3 Table shows the belief of route of transmission among participants in 7 MICs and there were significant differences among all countries. Out of all participants, there were a small number of participants who did not agree with transmission of COVID-19 being via droplets (10.34%) and contaminated objects (17.21%). It is interesting to note that China (60.5%) and Vietnam (59.8%) demonstrated significantly higher percentage of participants who believed in airborne transmission compared to .64.76% of participants from the other five countries who did not agree that COVID-19 was airborne transmitted.
Participants expressing confident and very confident in their doctors diagnosing COVID-19 were very high in Malaysia (93.8%) and China (92.9%); level of confidence was much lower in Iran (65.5%) and Pakistan (62.6%). About 50.26% of participants reported that they were likely and very likely to contract COVID-19, with Malaysian participants demonstrating the highest perceived risk of COVID-19 (72.8%) whilst the Filipino demonstrated the highest proportion of participants believing that they would not contract COVID-19 (53.2%). About 89.8% of Thai participants believed that they would survive if contracted with COVID-19

Direct contact with people infected by COVID-19
Yes <0.001 <0.001 0.14 (-0. 10  information on COVID-19 with Filipino participants having the highest proportion for spending more than 2 hours per day to view information (47.2%). Table 3 shows the association between knowledge and concerns related to COVID-19 and mental health parameters. Agreement with airborne, contact with contaminated objects and droplet transmission was associated with higher DASS-21 in all subscales. Likelihood of contracting COVID-19, discrimination against by other countries and contact with people infected with COVID-19 were associated with higher IES-R or DASS-21 scores. Confidence in one's own doctor diagnosing COVID-19, high likelihood of survival if infected with COVID-19 and spent less than two hours per day to monitor information relating to COVID-19 were associated with lower level of IES-R or DASS-21 scores.
S4 Table shows the prevalence of precautionary measures and there were significant differences among 7 MICs (p<0.001). High percentages were reported by participants covering their mouth and nose after sneezing (98.0%), avoided sharing utensils (90.8%), practised hand hygiene (98.9%), washed hand after touching contaminated objects (96.2%), and wear face masks (93.5%). All Vietnamese participants (100%) responded wearing a face mask. About 68% of respondents felt that people were too worried about COVID-19 with Malaysia (90.5%), Thailand (90.5%) and Pakistan (86.6%) as the top three countries. Approximately 53% of respondents spent 20-24 hours per day at home; with China (84.7%), Iran (73.5%) and Philippines (55%) as the top three countries. Table 4 shows the association between precautionary measures related to COVID-19 and mental health parameters. Avoidance of sharing cutlery dealing meals was associated with higher anxiety and depression. In contrast, hand hygiene practice was associated with lower IES-R and DASS-21 in all subscales. Wearing a face mask was associated with lower levels of stress and depression. Worries about COVID-19 was associated with significantly higher levels of DASS-21 in all subscales. Shorter duration of homestay was associated with higher levels of anxiety, depression and stress as compared to those who stayed at home for 20-24 hours per day.
S5 Table compares the health information needs of participants from 7 MICs and there were significant differences among 7 MICs. The Chinese had the highest proportion who wanted to understand the symptoms of COVID-19 (91.6%), the prevention method (93.7%), effectiveness of drugs and vaccines (94.1%), number of infected cases and location (95.9%), travel advice (96.9%), mode of transmission (94.5%), required regular information update (92.7%) and personalized information (96.8%). The Iranians had the highest proportion who sought advices regarding treatment methods (90.4%) and Malaysians had the highest proportion who wanted to understand local outbreaks (94.2%). Table 5 shows the association between health information needs about COVID-19 and mental health parameters. Most additional information including information on COVID-19 symptoms, prevention, treatment advice, needs for regular updates, knowledge on local transmission, effectiveness on drugs and vaccines, number of infected people based on geographical locations, travel advice and transmission mode of COVID were associated with higher IES-R scores. In contrast, the need for more personalized information, information on the effectiveness of drugs and vaccines, travel advices, transmission mode were associated with significantly lower level of depression.

Discussion
The main findings of this first multinational population-based study in MICs in Asia during the COVID-19 pandemic are summarized as follows. First, Thai respondents reported the highest levels of IES-R and DASS-21 scores. Second, Pakistani respondents reported the second highest levels of DASS-21 scores. Comparatively, Vietnamese respondents reported the lowest levels in DASS-21 scores. Third, Iranian respondents demonstrated the lowest confidence in their doctors whilst Pakistani respondents had the highest proportion who believed they would not survive COVID-19 and reported discrimination.
Assessing COVID-19's association with respondents' mental health, the three most common physical symptoms associated with adverse mental health were headache, cough and sore throat. Risk factors associated with adverse mental health during the COVID-19 pandemic include age <30 years old, high education background, single and separated status, discrimination by other countries, contact with people with COVID-19 and worries about COVID-19. Protective factors for mental health during the COVID-19 pandemic include male gender, staying with children, staying with 6 or more people, employment, confidence in own's doctors diagnosing COVID-19, high perceived likelihood of surviving COVID-19, spending less time on health information, hand hygiene practice and wearing a face mask. Importantly, these findings will be significantly helpful for healthcare administrators in Asia at the national and local community levels [24] when preparing for the next wave of COVID-19 outbreak and future pandemics [25]. Iran had the highest total reported COVID cases (386,658) and number of COVID cases per 1 million people (4,593), as well as the highest number of deaths from COVID (22,293) and deaths per 1 million people (265) [26]. Pakistan had the second highest number of cases (298,509) and deaths (6,342) [26]. Of the 7 MICs, Vietnam had the lowest total numbers and rates across all seven countries, with 1,049 reported cases, 35 deaths and rates of just 11 cases and 0.4 deaths per 1 million [26]. As a result, Vietnamese respondents reported the lowest IES-R and DASS-21 scores. Vietnam has adopted several strategies to combat COVID-19 including development of the action plan and response strategies to optimize the utilization of human resources and equipment [24]; address the health information needs based on the diverse socioeconomic, demographic, and ethnic factors [27]; re-design communication activities for a more effective dissemination of information related to the epidemic [28]; safeguarding the health of workforce [29] to ensure minimal impact on economy and involvement of the grassroot system and village health collaborators to combat pandemics [30,31].
Thailand recorded the second lowest number of total cases (3,444) and deaths (58), and similarly the second lowest case rates (49) and death rates (0.8) per 1 million [26]. Surprisingly, we found that Thailand was the country with the highest IES-R and DASS-21 depression scores. This could be due to the impact of COVID-19 on the economy in Thailand. Among all MICs in Asia, the disruption on COVID-19 pandemic is the most severe on Thailand economy, due to its reliance on tourism as compared to other MICs. For 2020, the International Monetary Fund has predicted Thailand's GDP to be reduced by 6.7 percent which is highest among Asian countries [32]. Pakistan ranked second in terms of DASS-21 scores and number of COVID cases and deaths. The congruence between psychological parameters and epidemiology of COVID-19 in Pakistan was due to poor sanitation, lack of basic preventive measures, lack of proper testing and medical facilities. Pakistani health professionals started protesting and threatened to quit work due to lack of Personal Protective Equipment (PPE) [33]. Currently, the vaccination coverage in rural Pakistan remains unsatisfactory amid various barriers including price, hesitancy, and low level of awareness [34]. Eid-ul-Adha is an annual religious festival that could not be cancelled due to religious obligations and led to a sharp spike in COVID-19 cases [35]. The unpreparedness and contradictory policies resulted in an alarming high rate of COVID-19 spread and worsening mental health and discrimination faced by Pakistani people. Iranian respondents demonstrated lowest confidence in their doctors. The economic sanctions that prevented medical supplies, equipment and drugs from arriving in Iran could lead to low confidence among Iranians [36].
This study highlighted unique protective factors for mental health in MICs of Asia. In this study, more than 90% of respondents agreed to wear masks to prevent COVID-19. During the initial stage of COVID-19 pandemic, medical and public health experts from the US and some European countries believed that there was no direct evidence of airborne transmission of COVID-19 [37]. In contrast, respiratory clinicians and public health experts from Asia argued that lack of evidence does not equate to evidence of ineffectiveness of face masks [38]. The use of face masks by Asians have played an important role in controlling the spread of COVID-19 [39]. This study showed the association between the use of face mask and lower DASS-21 anxiety and depression scores. This finding might support the postulation that wearing face mask could offer psychological benefits, such as feeling less vulnerable to infection via perceived control [37]. Staying with children and more than 6 people in the same household were protective factors due to the values of family support among Asians. Compared with western countries, family support has a greater influence on reducing the risk of adverse mental health in Asia [10]. The findings of this first multinational study have several implications for health and government policies. Firstly, the health authorities should offer psychological interventions to the general population who are at higher risk of developing adverse mental health including women, people younger than 30 years and single and separated status. High education background is a risk factor and online psychological interventions such as cognitive behaviour therapy (CBT) and mindfulness-based therapy could improve mental health for highly educated individual [40]. For countries with high IES-R scores (Thailand, China and Iran), online trauma-focused CBT that promotes trauma narration, problem solving related to problems associated with COVID-19 and home based relaxation could be helpful in reducing psychological impact [9]. Second, as physical symptoms resembling COVID-19 infection (e.g., rhinitis, persistent fever with cough, breathing difficulties) were associated with high IES-R and DASS-21 scores groups. There is an urgent need to develop accurate, rapid diagnostic tests in general practitioners' clinics, community and rural settings [31]. A negative COVID-19 test result may alleviate anxiety, depression, stress and psychological impact. Enhancing the capacity of health system to combat COVID-19 may increase the confidence of public and improve mental health. Third, based on our findings, the WHO, governments and health authorities should provide regular updates on the effectiveness of vaccines and treatment methods. Mis-information related to the cause of COVID-19 [41], rumours [42] and inconsistent information [43] on COVID-19 symptoms, prevention, treatment and transmission mode were associated with negative psychological impact. Local governments, news agencies, professional and advocacy organisations should all provide health information and advices related to COVID-19 that are consistent with national guidelines and avoid mis-information [44]. It is important to identify group-specific demands would be helpful to provide proper information related to COVID-19 to fulfil the need of different population groups [27]. Various governments should offer relief packages to safeguard employment and economy to protect mental health. Additionally, the level of policy stringency in response to COVID-19 or pandemics, as measured by the Oxford Stringency Index, may influence mental health and should be moderated accordingly by respective governments [45]. This study has several limitations. First, the findings of this study were based on seven MICs in Asia and could not be generated to other countries. The study population had different sociodemographic characteristics as compared to the general population in the world due to sampling bias because only participants with Internet access could participate in this online survey. The respondent sampling method also compromised the representativeness of samples. The study population was female predominant (proportion of female in the study population: 67.76%; world population 49.58%) [46] and a high proportion of the study population possessed a university degree (85.6%). Thus, there is a potential risk of sampling bias because we could not reach out to potential respondents without Internet access. The second limitation was the cross-sectional nature of this study and inability to demonstrate cause and effect relationship. The third limitation was that we did not record demographic data regarding pre-existing mental illness of the study participants. The fourth limitation is that selfreported levels of psychological impact, anxiety, depression and stress may not always be aligned with objective assessment by mental health professionals. Nevertheless, psychological impact, anxiety, depression and stress are based on personal feelings, and self-reporting was paramount during the COVID-19 pandemic. The fifth limitation is that we did not study other aspects of the pandemic such as the potential threat of self-medication of hydroxychloroquine and cholorquine [47] and precautionary measures of walkthrough sanitization gates [48]. Lastly, we were unable to calculate the response rate. For potential respondents who were not keen to participate in the online survey, no response was recorded, and we could not collect any information from them.

Conclusions
In conclusion, this multi-national study across 7 MICs in Asia showed that Thai reported the highest mean IES-R and DASS-21 anxiety, depression and stress scores. In contrast, Vietnamese reported the lowest mean scores in IES-R and DASS-21 anxiety, depression and stress scales. The risk factors for adverse mental health include age < 30 years, high education background, single and separated status, discrimination by other countries, contact with people with COVID-19 and worries about COVID-19. The protective factors for mental health include male gender, staying with children, staying with 6 or more people, employment, confidence in own's doctors diagnosing COVID-19, high perceived likelihood of surviving COVID-19, spending less time on health information, hand hygiene practice and wearing a face mask.
Supporting information S1