Perceived Stress and Its Impact on Health Behavior of Chinese Residents during the Epidemic of COVID-19: An Internet Survey

Background An outbreak of 2019 novel coronavirus disease(COVID-19) could cause a severe respiratory illness, which may come from Spray or ordinary contact, then pose a great threat to life. Meanwhile, epidemic may trigger excessive stress and have a terrible impact on mental health even health behaviors, especially some special individuals. So this study paved the way for psychological interventions focused on speci�c groups during the COVID-19 epidemic. Methods An internet survey was conducted on 2,449 residents in 20 provinces including Chongqing, Sichuan and Gansu et al, in order to collect the data of Chinese residents' perceived stress, COVID-19 awareness and health behaviors. SAS9.4 was used to analyze the relationship between health behaviors and perceived stress, and logistic regression was used to explore the in�uencing factors of health risk stress. Results The residents' perceived stress score was 22.25±7.2(total score was 56), and the incidence of health risk stress was 39.89% (977/2449). As perceived stress increased, the frequency of health behaviors such as washing hands and wearing a mask decreased in turn. Age, cognition of susceptibility to COVID-19, life-threaten level, cognition of the importance of home isolation, and cognition of the difference between common cold and COVID-19 were related to the occurrence of health risk stress. Conclusions There was a negative correlation between residents' health behaviors and perceived stress during the COVID-19 outbreak. It is of great signi�cance to improve the public's awareness of the COVID-19, to provide good social support and psychological intervention for people with health risk stress and to promote their health behaviors.

a: compared with the rst layer, P < 0.05; b: compared with the second layer, P < 0.05; c: compared with the third layer, P < 0.05; d: compared with the fourth layer, P < 0.05; e: compared with the fth layer, P < 0.05; f: compared with the sixth layer, P < 0.05.

Chinese perceived stress scale (CPSS)
The perceived stress scale (PSS) was developed by American psychologist Dr. Cohen s [7] to assess the degree of stress caused by unpredictable, uncontrollable or overloaded life.Currently, there are 3 versions of 14 items (PSS-14), 10 items (PSS-10) and 4 items (PSS-4), PSS-14 is considered as a brief and easy scale to administer and complete [8].Considering the national conditions and the local environment, in this study, we used the version of Chinese perceived stress scale (CPSS).The scale(PSS-14) was rst translated by two domestic scholars, then translated back by two Chinese-American scholars, and lastly reviewed by one Chinese-American and one domestic scholar.The overall structure and speci c items were examined and necessary modi cations were made according to China's cultural background to form a preliminary scale, which had good reliability and validity after investigation [9], so CPSS was used to investigate the perceived stress of residents during the epidemic of COVID-19, including 14 items, each item had 5 options: never, almost, sometimes, often, always, corresponding to 0 to 4 points respectively.The highest possible score was 56, because positive questions were scored from 4 to 0, while negative questions were scored in reverse.Positive items assessed the ability to respond to perceived stress, while negative element focused on assessing lack of control, negative emotions, and reactions.Take 25 points as the critical value, low stress state (0-score), medium stress state (15-score) and health risk stress state (above 25 score).That is, the higher the score, the greater perceived stress of the residents during the epidemic.After the reliability test, the Cronbach coe cient of the scale in this survey was 0.829, which had a high internal consistency.

COVID-19 Cognition and Health Behavior Questionnaires
We referred to the relevant literatures on the prevention and treatment of infectious diseases and World Health Organization(WHO) and the National Health Committee of the People's Republic of China on the prevention of COVID-19 [10][11][12][13], The rationality and scienti city of questionnaire item design were communicated online by research team.Secondly, we conducted a small-scale test of the online questionnaire to understand the response time and the feelings of the respondents, deleted misleading or excessive items, and increased the feasibility of the survey.Finally, 13 cognitive items and 10 health behaviors items for the COVID-19 were drawn up.The Cronbach coe cient of two questionnaires were 0.667 and 0.827, respectively.

Statistical analysis
The measurement data of normal distribution were described by the mean plus or minus standard deviation.The comparison between groups adopted two independent samples T test and variance analysis, and the afterwards two-two comparison adopted SNK method.The measurement data were described by the rate of adoption, the chi-square analysis was used for the comparison between groups, and Bonferroni method was used for the subsequent two-totwo comparison.The linear trend test was used to examine the relationship between health behaviors and perceived stress, and the Gamma coe cient described the correlation between health behaviors and perceived stress, Variables with P < 0.05 in univariate logistic regression were included in multivariate logistic regression, and the independent risk factors of health risk stress were examined by stepwise regression (exclusion standard 0.05).P < 0.05 was considered statistically signi cant.All analyses were analyzed in SAS9.4.(Copyright 2016 SAS Institute Inc. Cary, NC, USA) P < 0.05 was considered statistically signi cant.

Research quality control
We adopted a series of measures for strict quality control.1).Each item was set as a required answer, otherwise it could not be submitted; 2).Avoid repeated submission and lling, each user could only answer once; 3).Eliminate invalid questionnaires (logical errors, answer time less than 1 minute or more than 30 minutes)

3.results
In this study, 2,533 internet questionnaires were collected, of which 2,449 were valid and been input into database for data analysis, with an effective rate of 96.7%.There was a negative correlation between health behavior and perceived stress, 39.89% of the residents were in a state of health risk stress.

Demographic characteristics & perceived stress
Among the valid samples, 823 (33.60%) were male and 1626 (66.39%) were female.90.4% of the residents are aged between 18 and 50.There were 1,783 urban residents, accounting for 72.8%, 666 township and rural residents, accounting for 27.2%.374 (15.3%) residents with junior high school education or below, and 2,075 (84.7%) residents with senior high school education or above were recruited.In sociodemographic data, gender, age, marital status, education level, and the analysis of perceived stress are statistically signi cant.P < 0.05.Totally, through this research, we have found at least three types of residents with extremely high stress.The results showed that the female residents' perceived stress score (22.6 ± 6.99) was higher than that of the male residents, but there was no statistical difference in the incidence of health risk stress between the two groups.The residents aged 18 to 25 years old had higher perceived stress (23.78 ± 6.24) and health risk stress incidence (48.86%) than those aged over 30 years old.Compared with other occupations, students had the strongest perceived stress (23.87 ± 6.18), and 48.66% of students are in a state of health risk stress.Table 1.Besides, residents who were in the medical isolation or the home isolation observation period had higher perceived stress (23.65 ± 6.59) than those who were not in isolation.Further, this study showed that 45.19% of medical workers were in a state of health risk stress.

health behaviors & perceived stress
There was a negative correlation between health behavior and perceived stress, the Gamma correlation coe cient is between − 0.212 and − 0.379.With the increased of perceived stress, the frequency of residents' health behaviors decreased in order.such as washing hands, wearing a mask et al.The data showed that the residents who never or occasionally paid attention to the real-time information of COVID-19 had the lowest incidence of health risk stress (0.1%), and the residents who always wore a mask when going out had the highest incidence of health risk stress (74.41%).Table 2.

Multivariate Analysis of Health Risk Stress
We that of the residents were in a state of health risk stress.Demographic characteristics and cognitive items were included in the univariate logistic regression analysis, and 13 variables (gender, marriage, occupation, monthly income, etc.) were included in the multivariate logistic regression analysis.The data showed that age, cognition of susceptibility to COVID-19, life-threaten levels, cognition of the importance of home isolation, and cognition of the difference between common cold and COVID-19 were signi cantly related to the occurrence of health risk stress among residents, P < 0.05.Table 4.

4.discussion
We conducted an internet survey on 2449 Chinese residents of perceived stress and health behaviors during the COVID-19 outbreak.The 45th Statistical Report on Internet Development in China showed that China's internet population was 904 million, with the internet popularizing rate reaching 64.5% [14].
Previously, some studies had also used internet survey and con rmed that it was time-saving and convenient Compared with the traditional paper questionnaire survey [15,16].
High risk groups of health risk stress Students had higher stress scores, previously, the prevalence of perceived stress among students was high [17].At the current stage, universities delayed the start of the semester, the ministry of education advocated the suspension of classes, and stress from unexpected events, that was COVID-19, all of them had adverse effects on students.Outbreaks can put signi cant psychological stress that may lead to unfavorable effects on learning and the overall psychological health of students [18], While distressing their overall psychological health, this situation could adversely affect learning, by the increased avoidance of learning activities and reducing their concentrations.all of these results highlight the need to establish psychological support programs for students during an COVID-19 outbreak.Besides, there is no doubt that the mental health of the isolated population deserved attention, these individuals needed more social support and professional Psychological crisis interventions.
Further, Medical personnel were also the focus of attention.Initially, the health care workers were placed in a stressful condition due to the uncertainty about the mode of transmission of the disease, tremendous fear, and implementation of rigorous infection control protocols, It had been found that 30% of con rmed cases were health care workers (HCWs)during Middle East Respiratory Syndrome-Corona Virus outbreaks, During the SARS epidemic in 2003 and the in uenza A/Hemagglutinin1 Neuraminidase1(H1N1) pandemic in 2009, researchers evaluated psychological stress on HCWs using different modalities, and stress was found constantly high [18].This study showed that 45.19% of medical workers were in a state of health risk stress, a large burden in the clinical treatment and public prevention efforts in hospitals and community settings.such psychological distress may affect HCWs, who would be in a high demand and shortage during outbreaks.Challenges and stress they experienced could trigger common mental disorders, including anxiety and depressive disorders, and posttraumatic stress disorder, which in turn could result in hazards that exceed the consequences of the epidemic itself [19,20].For on-thejob medical personnel, they should further strengthen their awareness of self-protection, emphasize strict implementation of infectious disease protection measures, standardize operations, maintain a peaceful mindset, and relieve stress in a timely manner.The hospital should provide timely management and technical support, allocate and guarantee supplies, and provide online and o ine psychological counseling services.
Totally, there were at least three types of residents with extremely high stress.In order to improve e ciency and make good use of limited medical resources, it is important to establish key target groups at the initial stage and set priorities accordingly.Nowadays, The guidelines divided the population affected by COVID-19 into four levels, the rst level included medical staff, especially in the front line, the second level included suspected people, and the forth level contained susceptible groups and general public [21].It was essential to incorporate stress buster or psychological therapists into the overall deployment of COVID-19 epidemic prevention and control, the main premise of stress buster or psychological therapists is to minimize psychological harm and provide timely assistance for epidemic prevention and control, that is to say, to provide mental health support for high risk stress individuals.The work of Psychological intervention was initiated by the Chinese Society of Psychiatry, with a team of experts established to organize intervention efforts and provide technical guidance [22].In the current epidemic status, face-to-face psychological counseling requires high standards of on-site isolation to minimize risks.This service is only available to front-line medical staff who have not been infected, so it is not suitable for ordinary residents who need psychological intervention.In order to ensure the continuous provision of mental health services and reduce the risk of cross-infection, the Chinese government is developing and implementing a remote consultation network to conduct telephone or internet consultations in a safe environment [21].

Negative Relationship between health Behavior and perceived Stress
Individual's perceived stress exert signi cant effects on behavior habit and experience [23], Data analysis of residents' health behaviors and different stress states were statistically signi cant (p < 0.05).Which is consistent with the results of the impact of perceived stress on health behaviors investigated by the work during the SARS epidemic [5].At present, it is generally believed that stress is a process and the product of the interaction between human and environment, which mainly includes the mediating variable of stress source and the physical and mental response.It has been suggested that there are three types of response that can help the body respond to stressors, namely direct stress responses.This kind of response can make the body respond quickly to ght or escape when it responds to stressors.Adaptability of secondary signals of the body and brain; and an assessment of stress responses based on cortical structures and pathways involved in cognitive processes through which the subject responds to the stressor [8].For the domestic material reserve status, this survey found that 71.37% of the residents had little or no masks when the epidemic broke out, while only 6.16% had more or su cient disinfectant.The shortage of supplies and the sudden disruption of work and life may all become main sources of stress for residents during the epidemic of COVID-19.We also found residents with monthly income more than 10,000 RMB have lower rates of perceived stress and health risk stress, so we had to admit that the economic situation was also an important factor that should not be ignored.
The link between perceived cognitive de cits and depressive symptoms is partially explained by quality of sleep and perceptions of stress [24], anxiety, as well.Anxiety levels closely mirrored the daily number of new cases and were strongly associated with the intensity of the outbreak [25].Psychobehavioral responses during the SARS or H1N1 in uenza outbreaks showed a positive association between anxiety level and personal protective behaviors [25][26][27], however, if the anxiety level continues to rise, it will have adverse effects, the previous study also showed the same results [28].Logistic regression models were also used to examine the prevalence of high levels of perceived stress and poor sleep quality, increased in proportion to the heavy smartphone use and insu cient physical activity [29].This work showed that only 53.33% of residents often or always take appropriate exercise even when they are isolated in home.Relevant study [30] have shown that staying at home for a long period of time may lead to increased sedentary behaviors, thereby reducing regular physical activity, eventually leading to increased health risks or causing a cycle of anxiety, depression, and vice.Therefore, it is suggested to strengthen residents' health behaviors, pay attention to a balanced diet, maintain regular exercise and improve immunity in a safe family environment.

Residents' cognition of disease not optimistic stress state
As for knowledge, approximately 63% answered two or more questions correctly regarding knowledge of SARS [28].Nowadays, the overall disease awareness rate was 62.04% of COVID-19.This disease hasn't been fully understood, there is currently no speci c drug or vaccine, the need for disease information and health behaviors by o cial authorities is also high.This survey showed that 73.71% of the residents were very concerned about the epidemic information, which is accurate, clear, su cient, timely, and trustworthy.Public health authorities should keep monitoring the situation, as the more we learn about this novel virus and its associated outbreaks, the better we can respond [31].The media and governments at all levels should enrich the channels and methods of information disclosure, strengthen the education of disease-related knowledge, and release epidemic information in real time to meet the needs of different groups.Residents should take a rational view of COVID-19, abide by the special regulations during the epidemic, increase

Table 1
Comparison of Perceived Stress and Incidence of Health Risk Stress among Participants with Different Demographic Characteristics a: compared with the rst layer, P < 0.05; b: compared with the second layer, P < 0.05; c: compared with the third layer, P < 0.05; d: compared with the fourth layer, P < 0.05; e: compared with the fth layer, P < 0.05; f: compared with the sixth layer, P < 0.05.

Table 2
Comparison of participants' health behaviors under different stress states Nowadays, the con rmed transmission routes of the virus include droplet transmission and contact transmission of COVID-19.However, aspect of COVID-19 cognition, 74.27%-80.93% of residents could correctly answer whether taking antibiotics / Shuanghuanglian oral solution could prevent COVID-19, whether vinegar could kill the virus, and whether hot water at 56 degrees Celsius could kill the virus for 30 minutes, It can be seen that residents had a certain ability to identify rumors.But only 15.3% of the residents knew the route of transmission of the virus, 13.1% of the residents knew the masks that could prevent the virus from spreading, and the overall disease awareness rate was 62.04%.At the same time, this survey have shown that different cognitive levels of COVID-19 could affect perceived stress score and the incidence of health risk stress.Table3.

Table 3
The status quo of disease cognition and its comparison with the incidence of perceived stress and health risk stress a: compared with the rst layer, P < 0.05; b: compared with the second layer, P < 0.05; c: compared with the third layer, P < 0.05; d: compared with the fourth layer, P < 0.05.a: compared with the rst layer, P < 0.05; b: compared with the second layer, P < 0.05; c: compared with the third layer, P < 0.05; d: compared with the fourth layer, P < 0.05.

Table 4
Logistic Regression Model Results of Health Risk Stress Incidence among participants