Sociodemographic Correlates of Perceived Psychosocial Health in Times of the COVID-19 Pandemic: A Community-Based Online Study in China

The coronavirus (COVID-19) pandemic and the ensuing sociopolitical measures to control and curb its spread have been affecting people's psychosocial health and well-being through various complex pathways and in unprecedented ways. The present study aims to assess the sociodemographic correlates of psychosocial health situation of Chinese community-dwelling residents. Methods: This was a cross-sectional survey that was carried out online and using a structured questionnaire during April 2020. In total, 4788 men and women with the age range of 11-98 years were included in the analysis. Outcome variables were the change in the experience of hopelessness, loneliness and depression before and during the pandemic, and the explanatory variables included demographic and social capital related variables. Results: Respectively 34.80%, 32.50% and 44.84% of the participants expressed feeling more hopeless, lonely, and depressed during the pandemic. The percentage of all three indicators was comparatively higher among women than among men: hopelessness (50.67% vs 49.33%), loneliness (52.44% vs 47.56%), and depression (56.22% vs 43.78%). Being married was associated lower odds of loneliness among men [Odds ratio= 0.63, 95% CI=0.45,0.90]. Loneliness was negatively associated with smoking [Odds ratio= 0.67, 95% CI=0.45,0.99] and positively with drinking [Odds ratio= 1.45, 95% CI=1.04,2.02]. Compared with those in the lowest income bracket (<10K), men [Odds ratio= 0.34, 95% CI=0.21,0.55] and women [Odds ratio= 0.36, 95% CI=0.23,0.56] in the highest (>40K) had the lowest odds of reporting perceived hopelessness [Odds ratio= 0.35, 95% CI=0.25,0.48]. Smoking also showed negative association with depression only among men [Odds ratio=0.63, 95% CI=0.43,0.91]. Conclusion: More about one-third of the participants reported worsening in the experience of hopelessness and loneliness, with more than two-fth of


Study settings and sampling methods
In this study, residents in eastern, central, and western China were selected through directional strati cation and convenience sampling. According to the epidemic prevalence of COVID-19 on April 1 2020, top two provinces and lower one provinces in terms of the number of cases were selected from each region. Therefore, we selected Hubei, Hunan and Shanxi provinces from Central China, and selected Guangdong, Zhejiang and Fujian provinces from eastern China. Due to the similarities in local conditions and customs between Sichuan and Chongqing, we chose only one of those two provinces with a comparatively higher prevalence in western China. According to the comprehensive in uence of the city in each province, the provincial capital and another city are selected in each province. Sixty households from both rural and urban households in one city and all households aged over ten years were invited to participant in the survey. A total of 7,118 residents from 1,920 households in 8 provinces (16 cities) were surveyed. Due to the low response of residents in Guangdong and Zhejiang provinces, only half of the households attend in the survey, so we combined them together in the analysis.
Data collection was conducted from April 4 to April 15 of 2020, a project manager in each province recruited and coordinated provincial survey training, and six local investigators are recruiting base from each city household income to send online questionnaires and control quality. Half of them are from urban areas and most are college students. After receiving training in online data collection, each investigator was asked to send online questionnaires to 20 local families on their social network, including friends, relatives, native classmates and so on. Each eligible family member was invited to ll out an online questionnaire on an average of 15 minutes. A secret gift will be sent to encourage the participants completing the submission through the WeChat. Due to the limitations of objective factors such as age, education level, and space distance, residents may lose the ability to participate in the online survey.
It is suggested to invite the young offspring living together to answer the questions according to their choice. If there is di culty in investigating the surrounding 20 families, a supplementary survey is carried out by other investigators to complete the remaining household survey.
Meanwhile, the follow-up investigation of quality control measures were took during the data collection process. (1) Conducted a preliminary survey, group and trained the investigators. (2) Each researcher was independent, allowing the relationship between students at different learning stages. (3) Before distributed the online questionnaire, the names and eligible family Numbers of 20 family members were required to generate a unique questionnaire number. 4) Questionnaires for each family were sent out one by one and asked to convey a message: "Those who carefully complete the questionnaire will receive a secret gift."Many trap questions were set in the questionnaire to identify people who did not answer the questions carefully. (5) The project manager will be required to check the quality of each questionnaire according to the threshold value of survey time exceeding 450 seconds and the consistency of the two groups of questions set in the questionnaire.

Outcome and explanatory measurements
In this study, we adopted a tactical approach to capture three key domains of psychosocial health that are more likely to occur during a pandemic including hopelessness, loneliness and depression.
The second outcome variable is perceived hopelessness, which is a commonly used construct used in population-based studies as an indicator of psychosocial well-being such as depression and suicide [25,26] and has been studied in the context of predicts general health and social functioning among population with mood disorders, showing the wider applicability of this construct in the context of psychological well-being. For this study, it was measured by the question: would you say since the beginning of the pandemic you have been feeling hopeless: same as before, little worse than before, far worse than before. Hopelessness is associated with increases the risk of emotional maladjustment and a range of negative mood states, both in the general population and clinical settings. [27] The second outcome variable is perceived loneliness, which was measured by the question: would you say since the beginning of the pandemic you have been feeling lonely: same as before, little worse than before, far worse than before. Loneliness is widely a prevalent phenomenon globally and have been a popular topic of research across various domains including chronic health conditions, psychological stress, and anxiety. [28] Loneliness is a common human emotion that is linked to feeling of insecurity, vulnerability and isolation and is associated with overall morbidity and mortality in adult populations.
Although there is no universally agreed de nition of loneliness, it is generally understood as not just being alone, but perceived feeling of lack of an attachment gure, social network, and absence of a circle of people that allows an individual to develop a sense of belonging, of company, of being part of a community. [28] [29] The third outcome variable is perceived depression which was measured by the question: would you say since the beginning of the pandemic you have been feeling depressed: same as before, little worse than before, far worse than before.
A single-item measure of self-rated depression (SRD) is being used increasingly population-based health surveys for its ease of application and high sensitivity to objectively measured health outcomes including all-cause mortality among cognitively intact community-dwelling older adults. [31] One-item questions for measuring general health conditions are increasingly used in epidemiologic surveys [1] and measure by questions like: "In general, would you say your mental health is: Excellent, Very Good, Good, Fair or Poor?" [30] Explanatory variables included: Age

Data analysis:
Data analyses were performed using Stata version 14. The prevalence of sample population reporting hopelessness, loneliness, and depression was presented as percentages. Following that, the relationship between the three outcome and explanatory variables were measured by multivariable regression methods. Given the dichotomous nature of the outcome variables, a binary logistic regression model was used to generate the odds ratios and their 95% con dence intervals. The variance in ation factor (VIF) was used as a measure of collinearity to ensure that none of the predictor variables in the nal model was highly associated with each other. All statistical tests were two-tailed and p values below 0.05 were considered statistically signi cant.

Ethics statement
The protocol was reviewed, the ethical approval was obtained from the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology (2020S107). The oral informed consent was obtained from each participant before taking the online survey. Finally, a total of 6,253 residents over the age of 10 completed the survey, of which 4,788 were eligible. The participation ratio was 87.85%(6253/7118), and the valid participation ratio was 67.13% (4778/7118).

Results
Basic demographic characteristics and the prevalence of reporting hopelessness, loneliness, and depression were presented in Table 1. Respectively 34.80%, 32.50% and 44.84% of the participants expressed feeling more hopeless, lonely, and depressed during the pandemic. The percentage of all three indicators was comparatively higher among women than among men: hopelessness (50.67% vs 49.33%), loneliness (52.44% vs 47.56%), and depression (56.22% vs 43.78%). As shown in Fig. 1, men were less likely to report same level of hopelessness (49.3% vs 57.7%), loneliness (46.7% vs 53.3%) and depression (49.5% vs 50.5%) during the pandemic than before compared with women. More than half of the women reported having higher level of hopelessness (55.4%), loneliness (50.5%) and depression (58.0%) during the pandemic than before.

Results of regression analysis
Factors associated with hopelessness, loneliness, and depression were presented in

Discussion
The COVID-19 pandemic, through its various socio-cultural and environmental factors, is exerting adverse effects on people's psychosocial health around the world. The present study aimed to understand the state and the associated factors of three contextual mental health indicators, namely hopelessness, loneliness and depression among Chinese community-dwelling residents. Initial descriptive analysis showed that more about one-third of the participants reported worsening in the experience of hopelessness and loneliness, with more than two-fth of worsening depression during the pandemic compared with the time before. Notably, the percentage of all three indicators was comparatively higher among women than among men, implying that gendergradient in the vulnerability to mental health implications of the pandemic. There is a growing volume of literature on mental health repercussions of the pandemic, but the sex-differences in the mental health related outcomes are not very clear. However, the prevalence of psychological disorders, especially that of major depressive disorders has been found to be higher among women in previous studies. [32,33]. In the context of COVID-19, women might be at higher risk for poor mental health outcomes due to issues related to increased incidence of intimate partner violence (IPV) and loss of livelihood. In addition, women who are pregnant and experiencing di culties in receiving routine antenatal care might experience psychological challenges that are being ignored by themselves and their caregivers. While this study didn't include these potential factors such as pregnancy, quality of marriage, it is recommended that future research underscored these issues to better understand the sex-disparity in mental health outcomes from COVID-19.
Results further revealed that the worsening experience of hopelessness, loneliness and depression are correlated with a range of sociodemographic and economic factors. We found that participants in the higher age groups had relatively higher odds of reporting hopelessness, loneliness and depression, except for those in the oldest age groups (70 + years), in whom the association was reversed. In general, being currently married, living in the higher income households, being employed in white collar occupation had protective effects on the three outcomes, however inconsistently across the outcomes and the sex groups. For instance, being employed in blue-collar jobs showed higher odds for hopeless and depression, but not for loneliness. For hopelessness, the associations were signi cant only among men, and depression only among women. Expectedly, the we found a strong positive association between reporting hopelessness and household income. The current body of literature provides evidence of the physical and psychological morbidities resulting from nancial constraints, [34][35][36] with a handful of studies brie y focusing on the construct of hopelessness. [37][38][39][40] The intersection between nancial and mental well-being is a complex one and is mediated with the underlying bene ts of material advantage. Nonetheless, this result should be interpreted with caution since we had data only on raw income which may not be indicative of the actual nancial situation of the participants. It is also worthy of noting that household income didn't show any signi cant association with loneliness and depression. While the link between socioeconomic status and mental health is relatively clear, our ndings enrich the literature by showing contrastingly that income brackets are more likely to be corelated with sense of hopelessness, compared with the more commonly studied construct of depression.
Regarding health and health related behavior, we found that tobacco smoking was negatively associated with loneliness and depression, while drinking was positively associated with loneliness and only among men. Several studies have so far discussed that the use of both smoking and drinking are being triggered by the psychosocial stress resulting from the pandemic. [16,17,41,42] As the world's largest consumer of tobacco and alcoholic products, China is already having a serious mental health burdens resulting from tobacco and alcohol induced diseases. Special effort should therefore be made to control the use of these products in order to minimize their psychological consequences during the pandemic. Having NCDs was also found to be associated with higher odds of reporting loneliness and depression both among men and women. In China, NCDs represent a major contributor to mental health related morbidities and mortalities especially among elderly population, [43][44][45] and the current situation is likely be further aggravating given the higher susceptibility of elderly population to COVID-19 infection. While the healthcare system is being overstrained with the COVID-19 patients, the mental healthcare needs of people with chronic diseases should be given special priority at the same time. Lastly, we found that participants in the urban areas had higher odds of reporting hopelessness, loneliness and depression, indicating that urban population share a higher susceptibility to psychological stressors compared with their rural counterparts. The underlying reasons behind this urban-rural difference might be rooted to the factors such as population density and relative risk of cross-transmission, differences in the type of employment, and availability of essential goods and services.
Urbanization or urban residency is a well-documented contributor to psychological stress, [46][47][48] and the effect of the stressor can get intensi ed by the unique aspects brought by the pandemic. Besides urbanicity, results also showed important differences in the outcome variables across provinces, denoting the potential role of geographic factors in the mental health implications of the pandemic.

Strengths and limitations
To our knowledge, ours is the rst cross-sectional study to assess the situation of mental health status in terms of hopelessness, loneliness, and depression in the context of COVID-19 among Chinese population. sample size was relatively large and included participants with broad age range. One important aspect of the study is the contrasting measurement of the outcome factors before and during the pandemic. This method of subjective measurement of mental health status is relatively simpler and yet captures important information regarding the change in the situation speci c to the pandemic. It should be kept in mind that this method doesn't re ect whether or not people were in sound mental health status prior to the pandemic, but rather the shift which can be used effectively in other crisis settings such as natural disasters. Our results should be interpreted with caution because of several limitations. First, the data are cross-sectional and the associations cannot indicate causality. Second, the conclusions cannot be generalized to the entire population of China due to inadequate sample size. Third, data were self-reported, and therefore the chance of reporting bias cannot be ignored. We were also unable to include these potential factors such as pregnancy and spousal relationships which are likely to be associated with the outcome variables among women. Also, nancial situation was measured in terms of raw income and not as subjective assessment of solvency, which could have given a better re ection of the association between material wealth and psychological health.

Conclusions
The present study aimed to explore the state of selected mental health indicators and their associated factors among community-dwelling residents in China. Findings showed that more about one-third of the participants reported worsening in the experience of hopelessness and loneliness, with more than two-fth of worsening depression during the pandemic compared with the time before, with the percentage of all three indicators being comparatively higher among women than among men. Several socioeconomic and lifestyle factors were found to be associated with the outcome variables, most notably participants marital status, household income, smoking, alcohol drinking, existing chronic conditions, and urbanicity. Although the data are crosssectional and hence no causal inference can be made of the associations, out study makes an important contribution to the current literature regarding the mental health situation among population who are not directly affected by the pandemic, but among the healthy and community dwelling population.
These ndings will hope understand the sociodemographic groups sharing a higher susceptibility to psychosocial stress arising from the pandemic and design proper intervention strategies.

Declarations -Ethical Approval and Consent to participate
The ethical approval was obtained from the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology (2020S107). The oral informed consent was obtained from each participant before taking the online survey.
-Consent for publication Not applicable -Availability of supporting data The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

-Competing interests
The authors declare that they have no competing interests -Funding ST received funding from the Fundamental Research Funds for the Central Universities at http://www.hust.edu.cn/ (Grant NO: 2020kfyXGYJ071). The funders had no role in study design, data collection, and analysis, decision to publish, or preparation of the manuscript.
-Authors' contributions ST was responsible for collecting the data, drafted the outline of this study, performed data management, and revised the manuscript. GW performed the data analysis and drafted the rst manuscript. All authors read and approved the nal manuscript. -Acknowledgements Not applicable