Knowledge, Preventive Practices, and Depression Among Chinese University Students in Korea and China During the COVID-19 Pandemic: An Online Cross-sectional Study

Background: To investigate the knowledge, preventive practices, and depression of Chinese university students living in South Korea (International Group) and Mainland China (Mainland Group) regarding the Coronavirus Disease 2019 (COVID-19) outbreak and explore the determinants of depression among these students. Methods: An online cross-sectional questionnaire survey was conducted both in China and South Korea from March 23 to April 12, 2020. The online questionnaire included questions on knowledge and preventive practices related to COVID-19, and PHQ-9 which was used to diagnose depression in target groups. A total of 420 Chinese university students were nally included in the study (171 students from South Korea and 249 students from mainland China). Results: Majority of these students had a good level of knowledge of COVID-19. The International Group performed better than the Mainland Group regarding preventive practices; however, the percentage of students with moderate-to-severe depression was higher among the International Group. More severe depression was related to high levels of concern about family members, about getting COVID-19, and suspecting themselves of having come into contact with patients. Meanwhile, taking preventive behaviors more comprehensively decreased the depression state of both the groups. Conclusions: During the COVID-19 pandemic, the depression status of students of the International Group was signicantly more severe than that of the Mainland Group (χ 2 = 5.50, p < 0.05). The level of depression among students in both the groups was affected by their concern regarding their family members, getting COVID-19, and whether they suspected themselves of having come into contact with patients. Therefore, psychological counseling and education programs are necessary in order to support and improve the mental health of International Group students.

are becoming increasingly common among college students [22,23,24]. Potential problems such as anxiety, stress, and depression predicted a lower year-end degree commitment and negatively affected their academic performance [25,26]. The COVID-19 outbreak coincided with the Chinese lunar New Year holiday and the opening of South Korean universities; thus, similar population in both countries have been affected, directly or indirectly. To prevent the outbreak from escalating, universities in China and South Korea have postponed the beginning of the semester and canceled all campus events such as workshops, conferences, sports, and other activities [27]. The mental health of college students who are forced to stay at home for a long time with decreasing collective activities may be affected, leading to anxiety or depressive symptoms [28,29]. Thus, the psychological condition of university students cannot be overlooked and must be monitored.
Since the viral outbreak was rst reported in China, the Chinese have been targeted and blamed for the spread of COVID-19 and stigmatized internationally. An example of this would be the use of the terms "China virus" or "Wuhan virus" by the media [30]. Chinese students account for the largest proportion of foreign students studying in South Korea [31]. Most of them live alone on their own and have never known or experienced a severe outbreak in China beforehand. Previous research shows that the transition of studying abroad adds another layer of stress that can exacerbate and amplify previously existing mental problems [32]. There is a high probability that their loneliness abroad along with global discrimination and stress over the epidemic in Korea may affect their academic performance and depression level [33].
Research has been conducted in different countries to identify the risk and protective factors contributing to depression and anxiety of university students during the COVID-19 pandemic. University students in the United Arab Emirates and Jordan demonstrated that adequate knowledge, good attitudes, and low-risk practices were among the protective factors toward the prevention of COVID-19 [34,35]. Moreover, having relatives or acquaintances infected with COVID-19 is a risk factor for anxiety increase of university students [36,37]. In their study on Nigerian university students, Rakhmanov and Dane presented that an increase in the level of knowledge may help decrease anxiety levels [38]. In addition, a study on French university students showed that knowledge regarding the pandemic may be used to reduce its negative impact (like stress and anxiety) in the vulnerable population [39]. Although some university students had a good level of knowledge regarding COVID-19 and its preventive practices, comparative research in three countries indicated that health authorities should take their depressive status seriously [40]. However, there is no research that examined knowledge regarding COVID-19 and the risk factors leading to depression during the pandemic among overseas Chinese university students. Particularly, Chinese students in South Korea have become a vulnerable group. Therefore, using an online questionnaire survey, this study aimed to explore the conditions and determinants of knowledge, preventive practices, and depression among the Chinese university students in mainland China and South Korea during the COVID-19 pandemic.

Study design
We conducted the online survey from 23 March to 12 April 2020 [1] during the COVID-19 pandemic. The crosssectional survey was completed through an anonymous questionnaire. Our target populations were Chinese univeristy students studying in South Korea (hereby referred to as the International Group) and in Mainland China (hereby referred to as the Mainland Group). We used the same questionnaires for the two groups, and these were written in simpli ed Chinese characters. The questionnaire was distributed via the Naver Online Survey (Tool) for the International Group and through the Surveystar Online Survey (Tool) for the Mainland Group. Before conducting the survey, we revised and veri ed the contents of the questionnaire through an online pilot survey and ensured that the statements were appropriate and understandable.

Data collection
First, the sample size needed for the study was calculated using the G*Power 3.19 program. When calculated based on parameters of the two-sided test, χ 2 test, a residual variance of 0.83, α probability = 0.05, and power = 0.95 for F tests and linear multiple regression analysis, the minimum total sample size was estimated to be 356.
Second, during the initial screening of the online questionnaire, a statement regarding the purpose of the research and the con dentiality and privacy of individuals was written on the rst page of the survey questionnaire. Participants could only ll in the questionnaire after reading this statement and clicking "AGREE" to con rm their consent. In addition, we stipulated that the main questions in the survey were mandatory questions, which means the participants had to complete all answers before they submitted the online questionnaire. All the measures above resulted in a 100% response rate for our study.
Finally, 461 respondents were collected in this study via snowball sampling, wherein we recruited more respondents among the respondents' acquaintances. There were 180 responses from Chinese students in South Korea collected from March 23 to April 8, 2020, and 281 responses from Chinese students in mainland China collected from April 2 to 12, 2020. In line with the research aim, which was to survey university students, respondents who answered "Employed," "Unemployed," and "Others" to the occupation question were removed, leaving a total of 420 students (171 from the International Group and 249 from the Mainland Group).
All respondents expressed their willingness to participate and understood the background and purpose of the study.

Measurements
After data collection, we compared knowledge, preventive practices, and depression among Chinese university students in South Korea and China. This study used two questionnaires: 1) Questionnaire on COVID-19 and 2) Patient Health Questionnaire-9, which aimed to evaluate the target respondents' basic demographics characteristics, knowledge about COVID-19, preventive practices, and depressive symptoms. Most questions in the COVID-19 questionnaire were found to have reasonable validity and reliability in Wang's research on the Chinese general population [15]. Yonsei Global Health Center (YGHC) made changes to Part E (Precaution measures) and Part F (Additional information) according to the speci c situation in the two countries. The Patient Health Questionnaire-9 (PHQ-9) was included to the study to provide a baseline for the incidence of depression.

General Demographics
In this study, to re ect the demographic characteristics of the respondents, the basic survey asked questions related to the respondents' sex, age, education level, marital status, family size, whether they had medical insurance, whether they had chronic diseases, whether they had traveled abroad in the past 14 days, and whether they had experienced quarantine. The choices for educational level were "Undergraduate" and "Graduate," and the choices for marriage status were "Single" and "Married." In terms of family size, the choices included "1-person family," "2-person family," "3~5 persons family," and "more than 6-member family." Questions about having medical insurance, chronic illness, having traveled abroad, and having selfquarantined were answerable by either "yes" or "no." Respondents were also asked about their "self-assessed physical condition," which was answerable by either "above good" and "below fair."

Knowledge and perception regarding COVID-19
Understanding and perception of COVID-19 and other topics were evaluated through a self-enumeration questionnaire, which included questions regarding the student's knowledge on transmission pathway, information satisfaction, sources of related information, con dence about diagnose, degree of concern about this disease, perceived probability, and concern about family members.

Preventive practices of COVID-19
Nine basic preventive practices were incorporated into the questionnaire. The responses to the questions corresponded to the degree to which a measure was practiced on a daily basis (1 = "Never do this" and 5 = "Do this every day"), and the total score indicated how well the preventive practices were performed.
Cronbach's alpha coe cient of the preventive practices of the COVID-19 scale was 0.78.

Patient Health Questionnaire-9
Depressive symptoms could be diagnosed based on the nine-item criteria for depression in the Diagnostic and Statistical Manual of Mental Disorders published by the American Psychiatric Association [41]. Each question on the PHQ-9 may have been answered as follows (with their corresponding score):"not at all" (0 points), "several days" (1 point), "more than half the days" (2 points), and "almost every day" (3 points) [42]. Participants were divided into ve groups according to the total score: 0-4, 5-9, 10-14, 15-19, 20-27, which corresponded to "minimal or none," "mild," "moderate," "moderately severe," and "severe" depression, respectively [43]. The higher the score, the more severe the depression.
Depression was assessed according to the score on the PHQ-9. With a sensitivity of 88% and a speci city of 88% for detecting major depressive disorders, a score of 10 has been recommended as the cut-off score for diagnosing depression [44]. Thus, in this study, respondents who scored 10 or more points were classi ed as "moderate-to-severe," and respondents who scored less than 10 points were classi ed as "minimal-tomild." Cronbach's alpha coe cient of the scale in this study was 0.89.

Statistical Analysis
In this study, STATA 15·0 and SPSS 24·0 were used to conduct statistical analysis. The speci c analytical methods are as follows: 1. First, descriptive statistics, t-test, and Chi-square test were performed to compare each variable between the International Group and Mainland Group.
(2) Second, to explore the determinants of the different depression levels, hierarchical regression was performed.

General Characteristics
As shown in Table1, there were signi cant differences between the two groups in terms of "Age" (t = 5.427, p < 0.001), "Educational level" (χ 2 = 8.509, p = 0.004), "Marital status" (χ 2 = 4.153, p = 0.042), "Family size" (χ 2 = 32.126, p < 0.001), "Medical insurance" (χ 2 = 10.699, p = 0.001), and "Self-quarantine" (χ 2 = 42.230, p < 0.001). The mean age for the International Group was 24.08 ± 4.14 years compared with that of Mainland Group, which was 22.12 ± 2.28 years. There were higher proportions of graduate students, married respondents, and respondents with families comprising one to two members in the International Group than in the Mainland Group. In addition, there was a higher percentage of students who did not have medical insurance (16.4%) in the International group. In contrast, nearly half (47.8%) of the Mainland Group students had experienced being self-quarantined. In total, more than 90% of respondents reported good self-assessed physical condition. Table 2 shows the students' responses with regard to their knowledge about COVID-19. About 99.8% of them knew that the virus can spread through droplets, 88.8% knew that it could be transmitted via contacting contaminated objects, and 66.9% knew that it could be transmitted through air. The two groups were also updated on information regarding "Infected cases," "Death cases," and "Recovered cases." Regarding the sources of information, 92.1% of respondents got information from one to three sources such as the internet, TV, or family members. Overall, the respondents attained a good level of basic COVID-19 information satisfaction.

Knowledge about COVID-19
Between the International Group and Mainland Group, there were signi cant differences in terms of "con dence in diagnosis" (χ 2 = 15.647, p < 0.001), "concerns about the disease" (χ 2 = 4.246, p = 0.039), and "high perceived probability of getting infected" (χ 2 = 12.379, p < 0.001). Although more respondents in the Mainland Group were highly con dent in the diagnosis and were highly concerned about the disease, there was a higher number of respondents in the International Group whose perceived probability of infection was high. Among International Group students, 34.5% thought they were highly likely to be infected (19.3% in Mainland Group). However, some of them also thought they were more likely to survive after infection (91.8% in International Group VS. 86.7% in Mainland Group). Furthermore, no statistically signi cant difference was found in terms of "Concern about family members." 3.3. Differences in preventive practices between the Chinese students in the two countries during COVID-19 Table 3 illustrates the performance of preventive measures against COVID-19 in the International Group and Mainland Group. On the whole, the two groups were signi cantly different in terms of seven practices. The mean scores of International Group in the rst six practices were signi cantly higher than those of the Mainland Group, including the practices of "Covering mouth when coughing and sneezing" (t = 3.28, p < 0.001), "Washing hands with soap and water" (t = 2.93, p < 0.001), "Wash hands immediately after coughing, rubbing nose or sneezing" (t = 1.76, p < 0.05), "Washing hands after touching contaminated objects" (t = 7.11, p < 0.001) and "Avoiding public transportation" (t = 2.23, p < 0.05). However, this excluded the practice: "Wearing mask regardless of the presence or absence of symptoms." On the contrary, the Mainland Group achieved higher levels of performance than that of International Group in terms of the last three practices, including "Sitting in one row while having a meal" (t = -5.81, p < 0.001) and "Avoiding meeting more than 10 people" (t = -10.13, p < 0.001).

Depressive symptoms resulting from the analysis
A Chi-square test was performed to determine the relationship between the level of depression symptoms and if the student was part of the International Group and Mainland Group. The depressive symptoms were categorized into "minimal-to-mild" and "moderate-to-severe" (cut-off score of 10) in this study. As Table 4 shows the difference between the two groups, which was con rmed as statistically signi cant (χ 2 = 5.50, p < 0.05), the depression status of the International Group was more severe than that of the Mainland Group.
A simple linear stepwise regression analysis was conducted to explore the factors affecting the depression status of the respondents, including all the variables. The results (displayed in Table  5) showed that "information satisfaction," "patients' contact history," "concern about family members," and "self-assessed physical condition" were statistically signi cantly related to the respondents' depression. These four variables were subjected to hierarchical regression in different models in combination with variables such as demographics characteristic, knowledge score, and preventive score. Table 6 displays the results of hierarchical regression analysis on the determinants of depression on COVID-19 with four models. In Models 1 and 2, there were no statistically signi cant relationships between age, sex, education level, or marital status and the depression scores of respondents. Students who had a better assessment of their health had lower depression scores (t = -1.865, p < 0.05). As more variables were added in Models 3 and 4, the factors "Not sure contact patient's history" (p < 0.001), "Highly concerned about family members" (p < 0.01), and "Highly concerned about this disease" (p < 0.05) were associated with statistically signi cant increases in depression scores. In the four models, the preventive practice scores all had signi cantly negative relationships with depression scores. In short, the better the performance of preventive practices by the students, the lower the depression score. The Watson value was close to 2, indicating that the observed value was independent. Although R was small, the p values of F value in these four models were less than 0.01, showing a strong correlation of the interpretative power of the models.

Students
This study found that Chinese university students had a certain degree of knowledge of COIVD-19, which is in accordance with other research results [15,45]. Combined with the ndings on information satisfaction and information sources, most of them correctly understood how the virus is transmitted and have received detailed information on the cases. This shows that the publicity work and health education of schools, health institutions, and mass media are making an impact [46]. Although there were high percentages of students who perceived themselves as having high "con dence about diagnosis" and that they would "survive after infection," many students reported being highly "concern[ed] about family members" in this study. This is also consistent with the fact that as con rmed and suspected cases continue to increase, more provinces and countries are affected by the epidemic [13], which increases the public attention.
Some studies on the effects of the COVID-19 outbreak on Chinese university students' psychological state and their associated factors have been previously conducted [15,28,40]; however, these only included respondents from mainland China. After the case of "Patient No. 31," wherein a con rmed case participated in a gathering in Daegu at the Shincheonji Church of Jesus, there was a sudden outbreak in South Korea that attracted worldwide attention and suggested that upgraded quarantine and isolation were necessary [47,48]. Because most of the International Group students in South Korea live alone, their fears are exacerbated by the fact that they may have to experience a prolonged quarantine period by themselves (83.1% of them had self-quarantined). This may also explain why the students of International Group and Mainland Group experienced similar worries and concerns about this disease and family members. In addition, it should be noted that a number of students from the Mainland Group had a high perceived probability of getting infected by the disease. This is also consistent with the situation in mainland China, where COVID-19 has spread throughout almost every province since January 2020 [49]. Overall, both groups did well in performing these preventive practices. The Mainland Group only performed better than the International Group in terms of "avoiding using elevator" and "sitting in one row while having a meal." The International Group scored higher in all other preventive situations. Therefore, effective measures to prevent the virus should be more publicized through health education and publicity work by health institutions and mass media in mainland China.
Depressive disorder is one of the most common mental disorders, with a lifetime prevalence of 6.9% and a 12-month prevalence of 3.6% in the Chinese general population [50]. The average scores of PHQ-9 in these two groups were 7.20 (95% CI: 6.390-7.800) for the International Group and 6.20 (95% CI: 5.583-6.819) for the Mainland Group, and the proportion (28.7%) of the International Group, who experienced mild-tosevere symptoms, was much higher than that of the Mainland Group (18.9%), with a cut-off of 10 points. The prevalence of depression in the college students who participated in this study (12.6%) is higher than the average prevalence [51]. Factors such as performance of preventive practices, patient contact history, concern about family members, and concern about this disease and their relationships with depression scores in these respondents also showed that the COVID-19 outbreak may impact the psychological state of these university students, especially for International Group students. Thus, universities in South Korea urgently need to provide the necessary psychological interventions and health education measures for these students.
Improving university students' knowledge and prevention of the virus is conducive to their psychological health. The lack of expertise and knowledge regarding COVID-19 may cause students to be excessively worried about the damage brought on by the pandemic, resulting in a higher risk perception and more feelings of panic and anxiety [52]. As demonstrated in this study, the more comprehensive the preventive measures, the better the psychological state of university students, the lower the risk of mild depression, and the more positive their response to the epidemic. Therefore, pertinent departments and universities should make effective use of social networking platforms and social software, among other forms of media, to attract university students to receive relevant and comprehensive news, information, and education on COVID-19.
The thoughts and feelings of university students could also affect their mental health. The study showed that students who felt good about their bodies and did not suspect themselves as having come into contact with a patient, those with lower levels of stress, and those who were not as worried about their families and this disease had lower depression scores. During this outbreak, the generation of rumors and their in uence cannot be overlooked [53]. Negative and false information regarding the epidemic may result in great psychological consequences in the students because this may make them feel negative and require the companionship of family and friends at this time [54], which may not be possible for most of the students studying abroad due to the countries' ban on transportation and migration. Thus, more social and school support is necessary to help them develop and maintain a positive mindset. At the same time, government departments should release fair information in time in order to reduce discrimination, and schools should guide their students to be more caring.

Implications
Both government departments and universities should provide and ensure the dissemination of fair information. Furthermore, they should conduct a variety of effective health education and health activities tailored according to the characteristics of the two student groups in a timely manner so that university students could more comprehensively understand information regarding COVID-19 and better implement the relevant preventive measures. Moreover, it is necessary to provide psychological consultation and aid to reduce the negative effects of the outbreak on university students.

Limitation and Future Research
This study had several limitations. Due to the restrictions of various activities and limitations brought on by the COVID-19 pandemic, several adjustments had to be made. First, a web-based questionnaire was adopted in this study, which may have some shortcomings. One example is that self-assessed levels of physical condition and depression scores may not always be aligned with assessment of their real situation because participants may choose the "socially desirable response." Second, the snowball sampling method was employed; therefore, it may not be possible to make statistical inferences from the sample that are applicable to the population since the participants were not randomly selected. Third, the severity of the epidemic varies from region to region, thus the effects of the pandemic on depression among university students may also differ according to region. Therefore, it is necessary to conduct further studies on their depression status, and we are planning to conduct a prospective study on a comparable group.
Notwithstanding the above limitations, this study provides useful information on knowledge, preventive practice, and depressive symptoms among Chinese students in two countries (China and South Korea), which had suffered the largest COVID-19 outbreaks at the time this study was conducted. Our results could be used as evidence-based reference in the provision of psychological interventions for university students in different areas during the outbreak.

Conclusions
The present study investigated knowledge, preventive practices, and depression status of Chinese university students living in South Korea and Mainland China during the COVID-19 outbreak and further explored the determinants of depression among the students in both the groups. The results showed that the majority of the respondents had a satisfactory level of knowledge regarding COVID-19. The International Group students performed better than the Mainland Group students regarding preventive practices; however, the percentage of students with moderate-to-severe depression was higher in the International Group. Depression was associated with high levels of concern about family members and about getting COVID-19, and with suspecting themselves as having had contact with patients. Meanwhile, comprehensively performing preventive behaviors may be associated with a decrease in the depressive state in both groups.

Declarations
Ethics approval and consent to participate: Ethical approval for the study was obtained from Yonsei University Institutional Review Committee (IRB) before the data collection (Task No. 1041849-202005 -SB-054-01). Written informed consent was obtained from respondents for inclusion before they participated in the study. Con dentiality was maintained throughout the study by not recording participant names on questionnaires.
Availability of data and materials: The questionnaire used in the study are included in the submission. The datasets, however, are not available publicly, but can be made available upon request from the corresponding author.
Consent for publication: Not applicable.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.