Relationships between Physical and Social Behavioural Changes and the Mental Status of Homebound Residents in Hong Kong during the COVID-19 Pandemic

In Hong Kong, social distancing has been adopted in order to minimise the spread of COVID-19. This study aims to examine the changes in physical health, mental health, and social well-being experienced by local residents who were homebound during the pandemic. An online questionnaire in both Chinese and English versions was completed by 590 eligible participants from 24 April to 13 May 2020. The questionnaire found that individuals aged 18 to 25 years spent more time resting and relaxing but experienced more physical strain. Working status was associated with social contact, with participants working full-time jobs scoring higher in “maintaining social communication via electronic means” and “avoiding social activities outside the home”. Additionally, approximately one third of the participants (29.7%) had moderate to severe depression, and participants aged 18 to 25 were found to have higher scores in PHQ-9. Changes in physical health and social contact were significantly associated with developing depressive symptoms. From the results, it is clear that the COVID-19 pandemic has the potential to exert a negative impact on the mental health status of individuals.


Introduction
The outbreak of Coronavirus Disease 2019 (COVID- 19) was first reported in Wuhan, China, as a pneumonia of unknown cause. The World Health Organization (WHO) named the new virus COVID-19 on 11 February 2020 and declared the outbreak a world pandemic on 11 March 2020 [1]. COVID-19 causes respiratory infections with symptoms ranging from common cold-like to more severe diseases like Severe Acute Respiratory Syndrome (SARS) [2]. It is mainly transmitted through contact with respiratory droplets. However, many people only experience mild symptoms in the early stage, and the estimated incubation period ranges from 1 to 14 days [3]. By 31 August 2020, health-related information with informed consent. Only those who were Hong Kong residents and had been homebound for at least four days a week due to COVID-19 at the time of study were recruited. Those who are cognitively impaired were excluded in this study. Data was collected from 24 April to 13 May 2020.

Survey Instrument
A questionnaire comprising four parts, in both English and Chinese versions, was posted online. The first part collected demographic data, covering age, gender, average daily studying or working hours at home, and changes in daily routines. The second part-"physical health changes"-consisted of five questions about changes in physical health, rated in terms of "Much decreased", "Decreased", "No change", "Increased', and "Much increased". The third part was comprised of seven items about the aspects of social contact and social support. The fourth part on mental status was made up of 10 items, nine of which were adopted from the validated Patient Health Questionnaire-9 (PHQ-9). The additional question enquired about the increase of stress due to working from home. Participants were required to answer nine questions related to the frequency of depressive symptoms by selecting one of the following options: "Not at all", "Several days", "More than half the days" or "Nearly every day" [22]. The severity of depressive symptoms was determined by the total score ranging from 0 to 27. A higher score represents a greater level of depressive symptoms. Participants were classified into five different groups based on the severity of depressive symptoms: (i) having minimal or no depressive symptoms (0-4); (ii) mild (5-9); (iii) moderate (10)(11)(12)(13)(14); (iv) moderately severe (15)(16)(17)(18)(19); and (v) severe depressive symptoms (20)(21)(22)(23)(24)(25)(26)(27). Typically, participants with scores of 10 and above would be advised to seek medical attention.

Statistical Analysis
Descriptive statistics for demographic variables including household members, education level, working status, and changes in activities of daily living were reported in frequencies and proportions. Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) Statistics (Version 25, IBM Crop., Armonk, NY, USA).
Chi-square tests of independence were used to test for statistical significance between categorical variables. An independent t-test was used to determine whether there was a difference in gender and score regarding physical, social, and mental aspects. One-way Analysis of Variance (ANOVA) was employed to assess the difference in working status and total score on physical, social, and mental status. Pearson's correlation coefficient was used to investigate the relationship between social change, physical change, and mental status. Multivariate regression models were used to evaluate the covariates significantly associated with the four outcome variables (i.e., changes in physical health, social contact, social support and depressive severity) separately. Ten dummy demographic variables were coded to be included in the regression model. All p-values < 0.01 were considered to be statistically significant. To evaluate the reliability of this study, the internal consistency was examined and the Cronbach's alpha values of the multi-item scales were considered acceptable when the value was at least 0.5 [23].

Participant Characteristics
Seven hundred and eighty seven (787) responses were received, with 590 eligible participants in total. The demographic factors are presented in Table 1. A plurality of participants were 18 to 25 years old (46.9%) and had attained a tertiary education level or above (86.5%). 95.9% of participants stayed at home more often and 71.7% paid more attention to personal hygiene.

Physical Health Changes
The Cronbach's alpha of this construct was 0.554, which was above the cut-off value of 0.5 [24]. A t-test of independence showed no significant differences between gender and total score in physical health change (t = 0.358, df = 588, p = 0.720), but individuals who were studying scored higher in physical health changes than those who were not studying (t = 2.727, df = 588, p < 0.05). However, chi-square tests showed a significant difference between age group and "time to rest", χ 2 (24, n = 590) = 48.713, p = 0.002 (Table 2)  In multivariate regression models, three variables were significantly associated with a change in physical health ( Table 3). The most important factor in predicting a positive change in physical health was being younger than 18 years old with a coefficient of 1.435, which was significantly higher than the other age groups.

Social Health Changes
The Cronbach's alpha of items regarding social contact and social support were 0.507 and 0.785 respectively, both of which were above the recommended value [24]. There was a significant difference between gender and total score in social contact, but the effect was small, with females (M = 6.02) scoring higher than males (M = 5.60), (t = 3.147, df = 588, p < 0.005, d = 0.3). Since the data was not normally distributed and the variance was heterogeneous, a Robust Welch ANOVA was conducted to examine the effect of participants' working status on their total score in "maintaining social communication via electronic means" and "avoiding social activities outside the home". This showed a significant effect of presentation condition: F(2235.1) = 33.42, p < 0.001. From post-hoc Games-Howell analysis, the mean of social health changes in participants who were studying was significantly higher than that of the working and underemployed groups.
Nearly half of the participants increased their usage of electronic tools to communicate with others (48.6%), while just over half increased their avoidance of social activities outside their homes (50.8%). However, 45.8% of participants experienced no change in sharing feelings with family members or addressing their emotional health. Overall, participants aged 18 to 25 years increased their use of electronic means to maintain communication (f o = 119 and f e = 82.6) and increased their avoidance of social activities (f o = 154 and f e = 140.8) when compared with other age groups.
Individuals aged below 55, females, and degree-and higher degree-holders were found to be positively and significantly associated with an increase in the usage of electronic tools and the avoidance of outside activities. However, the association between demographic variables and social support was not significant.

Mental Health Status
The Cronbach's alpha of ten items was 0.926, which demonstrated good internal consistency [24]. The assumptions of normality and homogeneity of variance were violated and therefore a Robust Welch ANOVA was conducted to examine the effect of working status on total PHQ-9 score. This revealed a significant effect of presentation condition: F(2587) = 51.934, p < 0.001. By comparing means in post-hoc Games-Howell, individuals who were studying scored higher than those in the working and underemployed groups.
A chi-square test showed that there was a significant relationship between age group and depressive severity: χ 2 (4, n = 590) = 123.801, p < 0.001, as well as between educational level and depressive severity: χ 2 (4, n = 590) = 25.904, p < 0.001 (Table 5). 65.3% of participants experienced increased stress due to staying at home and 29.7% experienced moderate to severe levels of depressive symptoms. Overall, a significantly higher number of participants aged 18 to 25 years old than expected From the result of multivariate regression, social contact was found to be significantly associated with the total score of PHQ-9, indicating that the increase in the use of electronic devices and the decrease in outside activities was positively associated with a higher level of depressive severity. However, physical health changes were negatively associated with depressive severity, meaning that individuals with more time to relax and rest were at a lower risk of experiencing depressive symptoms. The 18 to 25 age group scored significantly higher in the total score of PHQ-9 than the below-18 age group.  Increased stress due to staying at home 0.000 0.006

Relationship between Changes in Physical and Social Health and Mental Health Status
The relationship between changes in physical and social health, and mental health status is shown in Table 6. There was a significant positive correlation between total physical score and social contact score (r = 0.146, n = 590, p < 0.001), and total physical score and social support score (r = 0.194, n = 590, p < 0.001). Total social contact score was also positively correlated with total PHQ-9 score (r = 0.211, n = 590, p < 0.001) and social support score (r = 0.220, n = 590, p < 0.001). Also, there was a significant negative correlation between the total physical score and the PHQ-9 score (r = −0.097, n = 590, p < 0.05). Table 6. Pearson correlations between changes in physical and social health and mental health status.

Discussion
Since COVID-19 became a pandemic, people have been urged to maintain social distance and increase personal and environmental hygienic habits to mitigate the transmission of the coronavirus. These measures enable individuals to avoid the inhalation of droplets expelled by infected persons as well as contact with contaminated surfaces, objects, and body parts [25]. The Hong Kong Government has implemented special work and online teaching arrangements for government departments and schools respectively to reduce social contact and, thus, the risk of spreading the novel coronavirus in the community. Employees are allowed to work from home, and all face-to-face teaching in educational institutions has been suspended. This study indicated that 95.9% of participants stayed at home more often to maintain social distancing, and over 70% paid more attention to their personal hygiene during the pandemic, which is similar to the health behaviour changes noted during the outbreak of SARS in Hong Kong in 2003 [25]. After experiencing the SARS epidemic, over 98% and 70% of the participants in Hong Kong reported the adoption of good personal hygiene and the wearing of face masks, respectively [26]. This practice has likely influenced the everyday routines of most Hong Kong residents, and in turn resulted in good personal hygiene maintenance during the current pandemic. However, this study showed that being homebound did not change participants' behaviours regarding diet or physical exercise. It is possible that they believed they were already in tune with their health, or that it takes a longer period of time to change diet and exercise habits. Additionally, participants who studied or worked from home needed to rely more on electronic devices to fulfil their work commitments or attend classes, resulting in additional use of the Internet and digital screens.
This study found that 29.7% of participants reported moderate to severe depressive symptoms and this result was similar to studies conducted in China and Italy, which found that 25% and 32.8% of the participants, respectively had moderate to severe depressive symptoms during the COVID-19 pandemic [27,28]. The level of depressive symptoms can be affected by the risk perception of health crisis. Risk perception is subjective judgement of possible consequences of an event, influenced by cognitive, emotional, social and cultural variation between different individuals and countries [29,30]. Personal knowledge, trust in government, healthcare professionals and science, and information of COVID-19 collected from friends and family are significant predictors of risk perception [30]. Additionally, affective risk perception, which is strongly subjective, is positively associated with the depressive symptoms during the COVID-19 pandemic, while negative association is found between cognitive risk perception, meaning individual judgement is based on scientific support, and depressive symptoms [27]. Risk perception may result in changes in daily routines such as paying more attention to personal hygiene and browsing more prevention information during the pandemic, hence, affecting depressive symptoms.
Owing to school and university closures, students aged between 18 to 25 stayed at home to study and attended online lessons and courses. Unsurprisingly, this study found that younger participants tended to use electronic devices more often to maintain their social life without physical contact with classmates and friends. They used electronic devices to communicate with their friends and lecturers and spent more time using online learning platforms [31]. Such learning modes facilitated the interaction between students and lecturers in receiving more feedback related to their assignments and course progress [32]. Moreover, younger people tended to have higher usage rates of electronic products than older people. They were more likely to visit different social media platforms [33,34] and, therefore, may have preferred staying at home and relying on electronic devices to obtain information and seek entertainment. This group also experienced increased physical strain, which could be explained by their more frequent usage of electronic devices; often a cause of prolonged physical discomfort such as neck, shoulder, and upper extremity pain [35].
This study also found a significant difference in social contact changes according to gender. Females tended to have higher scores in social contact, meaning that they communicated with friends and family members by using electronic devices and avoided outdoor social activities. This result is in line with that of another study which found that females were more frequent users of social media, texting, and online video calls for social connectivity [36]. This may be due to the gender differences in displaying emotions, as females tend to express their emotions of happiness, sadness, and fear to others [37]. In view of the COVID-19 pandemic, social networking and online communication tools, such as Facebook and WhatsApp, could be useful platforms to enhance communication and social connection among friends and family members.
Participants demonstrated no significant increase in sharing feelings and caring for family members. This is not in line with previous studies which showed that people shared feelings more often with family and friends during a health disease crisis [13,38]. In particular, individuals aged 18 to 25 indicated no change in such social support. People might avoid discussing COVID-19 to minimise the trigger of negative emotions such as anxiety regarding self-protection [39]. It may also be due to easy access to health information and news, and hence people feel unnecessary to share related feelings with others.
More importantly, the 18 to 25 age group felt increased levels of stress due to studying at home for more than half of each day, and almost every day. Although the participants who were studying, and thus using electronic devices, were maintaining social contact, they still showed moderate and moderately severe levels of depressive symptoms. A similar trend among young people was found by another local study which discovered that people who had not experienced the SARS outbreak in 2003 were associated with a poorer mental health status [11]. Scores of participants who were studying were higher than those who were working and underemployed. These participants were also predicted to have higher scores in PHQ-9 than other age groups. In contrast, participants aged 56 or above indicated no increase of stress arising from working from home. This may be because those aged 18 to 25 years had not experienced such a learning arrangement before and were more likely to become worried about their academic progress, learning support, and the teaching materials [15,40], as well as about when normal classes would resume. Younger people also tend to use social media more, which might trigger more stress. Further support provided by educational institutions, teachers, and peers may help to optimize students' learning progress, and thus relieve the stress and depressive symptoms induced by studying at home [40]. Using telemedicine to provide psychological counselling can also be enhanced to reduce psychological distress [14].
Those aged 56 and above experienced much lower levels of depressive symptoms, which could indicate that they prefer working from home instead of going about their typical working day. In addition, this study could not find a significant difference between participants who were working and underemployed with respect to changes in mental health. This finding contrasts with another study which found that staying at home due to the COVID-19 pandemic was associated with greater financial concern and health anxiety [41]. Considering the uncertain economic environment caused by the pandemic, it would be expected that working individuals and underemployed participants would be more depressed and have higher anxiety levels than those who were studying. One possible reason for this unexpected outcome is that the participants had relatively stable jobs or were approaching retirement, therefore they showed less significant concern.
A negative correlation between physical health change and total PHQ-9 score was found in this study. This is similar to another study which showed that participants who made positive changes to their physical health, including increasing physical activity and maintaining healthier diets, were associated with not only better physical health but also better mental health [42]. An increase in physical activity was associated with a decrease in depressive symptoms which implied that physical activity at home could potentially relieve stress and depressive symptoms [43]. Furthermore, a positive association and correlation between social contact and total PHQ-9 score was found in this study, meaning that participants relying more on social contact via online platforms and fewer outdoor social activities might have higher levels of depressive symptoms. In fact, participants who had inadequate physical social interaction and a reduction of face-to-face communication during the pandemic might be more prone to depressive symptoms. Spending more time using screens was significantly associated with developing depressive symptoms [44]. The reduction of time spent having social interactions was correlated to the depressive symptoms [45], whilst infrequent social interaction could also predict higher depressive symptoms [46]. Therefore, participants who were homebound during the pandemic with insufficient social interactions might be more likely to develop depressive symptoms.
This study has a number of limitations. First, the behavioural changes and mental status of participants were only measured during the pandemic, as the situation before the pandemic could not be identified. Future studies could aim to investigate and compare the difference of behavioural changes and mental status of participants after the pandemic. Second, socioeconomic variables including income and social class could be important variables through which the findings of this study could be interpreted. Third, 46.9% of the participants were aged 18 to 25 indicating selection biases. This could be due to the fact that this study was only accessible via an online platform; therefore, younger subjects were easier to recruit. The overrepresentation of younger subjects in this study is not necessarily representative of all residents in Hong Kong.

Conclusions
This is the first time that Hong Kong residents have experienced such quarantine and social distancing arrangements for more than two months. To our knowledge, this is the first study to examine the physical, social, and mental health changes of Hong Kong residents who stayed at home to study or work during the COVID-19 pandemic. Local residents experienced disruptions to their daily routines and changes in behaviour such as an increasing amount of time spent relaxing and resting. Stress level among younger age (18)(19)(20)(21)(22)(23)(24)(25) people are higher than that of the other age bands due to worry about academic progress and learning support. Likewise, more than three quarters of the participants chose electronic means to maintain their social contact as well as avoiding outdoor social activities. It is noteworthy that nearly a third of the participants in this study experienced moderate to severe depressive symptoms. Such high prevalence of depressive symptoms should provide the Government and health professionals with some insight into the mental status of the local residents, particularly students. Further studies to investigate depression among Hong Kong students during and after major social events like the pandemic are warranted. Moreover, the findings of this study have implications on practices in community health and in the Government's preparedness and response plan for the outbreak [47], particularly in the early identification and management of depression among at-risk individuals.