COVID-19 started in Wuhan, China in December 2019 [1] and was declared a pandemic on March 11, 2020 [2]. As of January 10, 2021, there were 87,364 deaths and over 89.4 million confirmed cases in China and overseas [3]. In China, the initial phase of the COVID-19 outbreak induced numerous stressors, as it impacted almost every aspect of daily life, from work and entertainment to service use and social interactions [4]. For instance, entry to and exit from Wuhan and many cities and regions have been prohibited since the Chinese Lunar New Year (CNY). Other personal and public control measures include closure of offices and public areas (eg, shopping areas, offices, and restaurants), massive quarantines, staying at home for a long period of time, suspension of school classes, and cancellation of events [4]. It is important to study the negative psychological responses potentially caused by the COVID-19 epidemic, as previous studies have reported high prevalence of depression and other mental health problems in various populations during the 2003 severe acute respiratory syndrome (SARS) epidemic, the 2009 novel influenza A (H1N1) pandemic, and the 2014-2016 Ebola outbreak [5-8]. Mental health problems have significant public health implications, as they affect the public’s use of measures for prevention [9]. The study of negative psychological responses and associated factors related to the COVID-19 outbreak in China allows global public health and mental health workers to assess related service demands and design effective interventions.

Although a number of studies have looked at the factors of depression during the COVID-19 period, fewer studies were conducted in the initial phase of the outbreak, which started after the Chinese government announced the disease’s person-to-person transmission property on January 20, 2020, and started the first controversial Wuhan lockdown 3 days after [10]. To our knowledge, no similar studies were conducted outside China around that period of time, as COVID-19 had not become a pandemic until March and gained less attention outside China in early February 2020. Psychological responses are context specific. The initial phase of the COVID-19 outbreak (in terms of the first few weeks) in the world meant uncertainties and a lack of information. It is imperative to document the community responses at the initial stage of new emerging infectious diseases (especially those that eventually become a pandemic) to inform preparations for future outbreaks.

Furthermore, the first COVID-19 outbreak occurred in China during the CNY, which involved high mobility, as billions of people were returning to their hometowns and, hence, created a high risk of spreading the virus to the entire country. Mobile populations have special relevance in this specific time and country context, one of which was university students, as the majority of students were studying in cities away from their hometowns. This study investigates psychological responses of depression and associated factors among university students in China from 10 to 20 days since COVID-19 was believed to involve person-to-person transmission. Our literature search found only four similar university student studies conducted during the same time period [11-14]. Three of them covered only one or two cities and provinces, and the fourth one claimed to involve 29 provinces but had a sample size of only 2216, while this study covers 16 cities in 13 provinces (n=23,863). Moreover, this study is population-based, while no sampling frame was mentioned in the four other studies.

Although factors of depression during the COVID-19 period have been widely reported. There are still substantial research gaps that are being filled by this study besides the aforementioned fact that such studies were scarce during the initial weeks. Cognitive, behavioral, and psychosocial factors related to COVID-19 were all found to be potential determinants of depression [15-17]. Although a lot of studies have investigated associations between COVID-19–related cognitive factors and preventive behaviors [18,19], only a few studies have looked at COVID-19–related cognitive factors of depression. For instance, the four Chinese university student studies conducted during the initial outbreak mainly mentioned lifestyle factors such as sleep and psychological attributes such as resilience but not COVID-19 cognitive factors [11-14]. Several studies of the population in China and overseas looked at cognitive factors such as perceived susceptibility and severity, which were positively associated with depression [15,20-23].

According to the fear appeal theory, perceived threat comprises perceived susceptibility and perceived severity [24]. The theory provides a framework to include cognitive factors in this study. Perceived chances for oneself or one’s significant others in contracting a disease (perceived susceptibility) [9] was significantly associated with psychological problems such as depression during the SARS, H1N1, and Ebola outbreaks [5,7,8]. In general, perceived severity of a disease was significantly associated with negative psychological responses [7,8]. The public’s belief that H1N1 could cause severe irreversible bodily damage was associated with mental distress [7]. Anticipation about the likelihood and scale of a potential outbreak reflects perceived severity of the epidemic at the community level. For instance, perceived chances of having large H1N1 and Ebola outbreaks were associated with mental health problems [5,7]. We contended that perceived chance of controlling the COVID-19 epidemic would be positively associated with depression, as such a perception may reduce perceived severity due to potential negative impacts (eg, finance, work, and social relationship).

Misconceptions that H1N1 could be transmitted via some unconfirmed modes of transmission such as waterborne transmissions increased perceived susceptibility, which was associated with mental distress [8]. The association between misconceived mode of COVID-19 transmission and depression was less clear. One study conducted in the general population during the initial phase of COVID-19 in China found that the perceived mode of transmission via droplets increased risk of depression, while perceived transmission via contaminated objects or airborne transmission was nonsignificant. As people were forming their perceptions based on the risk, severity, and mode of transmission during the initial outbreak phase amid uncertainties and a lack of knowledge, their associations with depression might be different from those obtained from subsequent studies conducted during the later phases of the pandemic.

COVID-19–related behavioral factors of depression are also important, as behavioral responses would occur during initial outbreaks of new emerging infectious diseases. An important and unprecedented response was staying at home during the CNY in China, keeping in mind that there were then no clear strict social distancing policies in China, and there was no penalty for going out in most Chinese cities. It is important to understand the level of staying at home during the critical initial outbreak phase, as it had contributed to the control of COVID-19 in China by reducing social contacts. Research has found that quarantine was positively associated with depression [25], but staying at home during the holidays is different from quarantine and working from home. Other studies found that social distancing, or more precisely compliance to social distancing policies, was positively associated with depression [26,27], while nonsignificant results have also been reported (eg, [28]). No study has looked at the association between voluntarily staying at home and depression during the initial COVID-19 period in China. Another important unique behavioral factor was close contacts with people from Wuhan, which was the first and most important epicenter of COVID-19. At the time of the survey, a majority of the COVID-19 cases in China were detected from Wuhan, while many cases detected outside Wuhan were related to visitors of Wuhan [10]. Close contacts with such people was a unique stressor that was investigated only in this study. Choosing to stay at home as a prevention strategy may reduce perceived susceptibility, as it lowers the likelihood of contracting COVID-19. Having closely contacted high-risk people such as those who had travelled to Wuhan may increase perceived susceptibility. Thus, it is essential to look at whether they have associations with depression after adjusting for perceived severity and susceptibility.

Potential psychosocial factors related to COVID-19 may be associated with mental distress at the community level. For instance, perceived discrimination is a risk factor of depression [29] and has been associated with mental distress in SARS research [30]. Health care workers and patients who recovered from SARS were discriminated against by the public [31]. As of January 10, 2021, there were 50,340 confirmed COVID-19 cases in Wuhan and 17,809 in the rest of the Hubei Province (where Wuhan is located) [32]. Five million people had traveled out of Wuhan during the CNY [33]. Those who had visited Wuhan or Hubei might have encountered discrimination. In addition, people at risk of contracting the virus were prone to encountering discrimination (eg, hospital workers and their close contacts, and family members, coworkers, and neighbors of infected cases). No study has looked at COVID-19–related discrimination and its potential effect on depression during the COVID-19 outbreak in China.

Mental distress due to COVID-19 (eg, panic, anxiety, and emotional agitation) is potentially associated with depression. It is understood that prevalence of depression among university students prior to the COVID-19 outbreak was not low; not all depressive symptoms were caused by COVID-19 although the pandemic could have inflated the risk of depression. This study thus has the novelty of measuring both general depressive symptoms and the level of self-reported mental distress directly attributed to COVID-19, based on a scale used in understanding mental distress due to SARS and H1N1 [7,34,35]. It is imperative to understand the associations between cognitive, behavioral, and psychosocial factors, and both mental distress due to COVID-19 and depressive symptoms, which were tested in this study.

Another research gap is that few studies have looked at the mechanisms between COVID-19–related factors and depression during the pandemic period. According to the common sense model of illness representation, cognitive perceptions, how a person feels about a disease (ie, emotional representation), and their coping responses such as behavioral responses would determine health outcomes, including mental health status [36,37]. Thus, it is contended that cognitive, behavioral, and psychosocial factors, and mental distress due to COVID-19 would be positively associated with depression. Furthermore, this study tested the mediation hypothesis that emotional responses (ie, mental distress due to COVID-19) mediated between the studied cognitive factors (eg, perceived bodily damages and perceived infection risk of COVID-19), behavioral factors (eg, staying at home and having close contacts with people who visited Wuhan), and psychosocial factors (eg, perceived discrimination related to COVID-19), and depressive symptoms. No study has looked at such mediations, and thus, this study contributes to the literature of mental distress during the initial COVID-19 period.

This study investigates the level of depression among 23,863 university students of 26 universities located at 16 cities in 13 provinces of China 10 days after the official recognition of person-to-person transmission by the Chinese government and during the 8th to 17th days of the CNY in China, which was the initial rising phase of COVID-19. Based on the literature search, besides background and contextual factors, this study investigates the associations between the following factors and both mental distress due to COVID-19 and depression: cognitive factors (ie, misconceptions about modes of transmission; perceived risks of contracting COVID-19 for self, family members, and classmates; perceived permanent bodily damages of COVID-19; and perceived chance in controlling the epidemic in China in the coming 6 months), behavioral factors (ie, staying at home behavior during the CNY and close contacts with people who had visited Wuhan before the CNY), and psychosocial factors (ie, perceived discrimination) related to COVID-19. We further tested the hypothesis that mental distress due to COVID-19 would mediate between the aforementioned cognitive, behavioral, and psychosocial factors, and depression. The literature has not reported similar studies.

This population-based study, which covers 26 universities of 16 cities in 13 provinces of China, revealed a relatively high level of depressive symptoms among university students during the initial phase of the COVID-19 outbreak in China (10-20 days since the official recognition of the person-to-person transmission property of COVID-19). This study identifies a number of significant COVID-19–related cognitive (ie, perceived infection risk and perceived controllability of the epidemic), behavioral (ie, staying at home behavior and close contacts with people from Wuhan), and psychosocial (ie, perceived discrimination) factors of both mental distress due to COVID-19 and depressive symptoms; nonsignificant factors included perceived airborne transmission and perceived permanent bodily damage. As expected, mental distress due to COVID-19 was significantly and positively associated with depressive symptoms; it further mediated the associations between some of the cognitive/behavioral/psychosocial factors and depressive symptoms. The findings present a rough snapshot of what happened in the country and shed light on how people may react emotionally to new emerging infectious diseases and how various types of responses (eg, cognitive and behavioral responses) would be associated with such emotional responses.

The COVID-19 pandemic has affected people’s mental health. The findings of this study reveal prevalent mild to severe depression in 40% (9326/23,863) of university students in China, according to the PHQ-9. The prevalence obtained from studies among university students in China prior to the COVID-19 period seemed lower (eg, 23.8% among Chinese university students reported in a meta-analysis [40] and 29.5% reported during the SARS outbreak among Chinese university students [41]), but such prevalence was not exactly comparable, as different tools and sampling methods were used. Three of the four studies on Chinese university students conducted during a similar time period also used the PHQ-9. Two of the studies, which were conducted in Guangzhou [11] and 29 Chinese provinces [14], presented prevalence of moderate to severe depression (PHQ-9≥10) of 7.7% and 23.3%, respectively, compared to the 14.5% (3464/23,863) of this study, denoting geographical variations in addition to differences in sampling methods. This study has the strength that it was population-based and had a class-based sampling frame, while the others were distributed conveniently online.

Depressive symptoms were prevalent among university students at the initial phase of the COVID-19 outbreak in China; this can be seen from the high levels of self-reported mental distress directly attributed to COVID-19, which is understandable, as uncertainties and the Wuhan lockdown were alarming and worrisome. Furthermore, mental distress due to COVID-19 was positively associated with depressive symptoms. Thus, the mental distress directly attributed to COVID-19 might have increased the already high level of depression among university students during the initial phase of the epidemic in China. As mental distress and depression affect preventive behaviors [9] and individual well-being [42], and acute stress may turn into chronic depression [43,44], health care workers need to integrate mental health promotion with prevention of COVID-19 at the initial stages of outbreaks for new emerging infectious diseases.

The context of this study was unique and relevant, as it was conducted during the CNY and soon after the first outbreak and lockdown in Wuhan. It can be seen that students who stayed in the university’s city or with family members were less likely than others to be depressed. They might have received better support from their significant others, and social support and coping resources are protective against mental health problems [45,46]. The perceived number of confirmed cases in the province they were located in was positively associated with depression; it is plausible that the perception might increase perceived susceptibility and perceived severity of the epidemic and thus depression. It is interesting that travel restrictions on entering and exiting from the city of stay, which was then a new preventive measure, was not significantly associated with depression. The findings suggest that such drastic restriction, if implemented orderly and with good support, does not necessarily cause substantial panic or negative impacts on mental distress. A number of countries implemented even more severe lockdowns such as bans on going out soon after the completion of this study (eg, Italy), which was followed by many other countries (eg, parts of the United Kingdom, France, and Australia). In some countries, the strict social distancing measures were associated with depression [27,47]. Future studies should review this new measure and identify ways to minimize its adverse mental health effects. This study hence documents the initial responses to preliminary lockdown measures in the COVID-19 pandemic.

This study has interesting findings that involve interpretations in the context of the initial outbreak and in comparison with other studies. Spearman analysis and the SEM showed that some COVID-19–related cognitions (perceived chance of controlling COVID-19 and perceived risks of infection) were significantly associated with both mental distress due to COVID-19 and depressive symptoms; such findings corroborate with other studies [15,20-23]. However, it is unexpected that perceived bodily damages (a reflection of perceived severity) and perceived airborne transmission (possibly a misconception) were not significantly correlated with mental distress due to COVID-19 and depression, while such correlations were significant in similar H1N1 research [7,8]. It is plausible that this study was conducted in the early prepandemic phase of COVID-19 when no clear information was given about long-term harms and modes of transmission.

The behavioral factors of self-reported mental distress due to COVID-19 and depressive symptoms also illustrated the uniqueness of this study. The duration of staying at home during the CNY was protective against depression, which has not been reported in the global literature (except one general population study that found a nonsignificant result [15]). However, other studies have reported positive associations between social distancing and isolation and depression [27,47]. It has some special implications, as there were then no clear and strict social distancing policies in China (except in Wuhan), and people could leave home without facing penalties. The government, however, pledged for national support to contain COVID-19; staying at home during the CNY hence might have involved altruism and social responsibility, which were negatively associated with depression [48]. People may also feel safer at home. Thus, a short period of staying at home (for the 7-day holiday) during the initial outbreak period of a new emerging infectious disease may not cause mental distress but was instead protective. It seems that social distancing policies need to be exercised as early as possible during new outbreaks of important emerging infectious diseases to increase effectiveness and minimize distress. In addition, we found that close contacts with people coming from Wuhan was a risk factor of depression. To our knowledge, no study has looked at this variable, although other studies looked at visits to Wuhan (eg, [13]). This finding has important implications, as Wuhan was the epicenter where the first outbreak occurred, and the virus was spread to other regions. The variable became nonsignificant in the SEM, possibly because of controlling for a potential confounder of perceived discrimination.

This study also investigated the psychosocial factor of perceived discrimination. Over half (13,874/23,863, 58.1%) of the participants perceived discrimination related to COVID-19, possibly because of their traveler status. Associations between perceived discrimination and depression were similarly reported in previous studies related to SARS and H1N1 [7,8,49]. In fact, the association with depression was the strongest one among all the factors of this study. Thus, university students may feel more depressed during the COVID-19 epidemic, not only because of related perceived susceptibility and severity but also the way they are treated by others. To our knowledge, only one Canadian general population study had looked at such an association but found a nonsignificant finding [16]. The situation in China was unique. The country faced strong international pressure during the study period; Wuhan was accused of spreading the disease to other regions. Indeed, COVID-19 was initially labelled as the Wuhan virus. Travelers might be regarded as potential carriers of the virus; discrimination found a fertile ground to grow and might have a powerful negative effect on mental distress. It is imperative to investigate whether perceived discrimination related to epicenters has caused mental distress in the later phase of the pandemic. For instance, there were over 22 million detected cases in the United States as of January 10, 2021, and a new viral strain of higher infectivity was found in the United Kingdom where the incidence of COVID-19 is soaring. The level of perceived discrimination and its association with depression might be country specific due to politicization. When facing outbreaks of new emerging infectious diseases in the future, stigma needs to be removed from the location of the outbreak. The effects of the generalized perceived discrimination need to be investigated.

This is one of the few studies that looked at the mechanisms behind the associations between COVID-19–related factors and depression during the COVID-19 period. The findings suggest that COVID-19–related cognitive, behavioral, and psychosocial factors, and mental distress directly attributed to COVID-19 were all associated with depression, while the relationships between some of the COVID-19 cognitive, behavioral, and psychosocial factors, and depression may be partially mediated and suppressed through mental distress due to COVID-19. Specifically, perceived risk and perceived discrimination may have an indirect effect on depression via mental distress due to COVID-19; such risk factors might have increased mental distress due to COVID-19, which would in turn increase risk of depression. In addition, confidence in controlling the epidemic and staying at home could potentially be protective against depression via reduction of mental distress due to COVID-19, which was in turn positively associated with depression. Interventions to improve these cognitions or preventive behaviors may thus reduce depression directly or via reduction of mental distress due to COVID-19. Moreover, the mediation and suppression model presented in Figure 1 are supported by the common sense model [36], which suggests that diseases (COVID-19 in this case) as stimuli trigger cognitive representations (perceptions related to COVID-19) and emotional representations (negative psychological responses to COVID-19). The two types of responses would in parallel determine the coping process and health outcomes (depression in this case). In addition, the theory postulates that the cognitive responses would have an effect on the emotional responses. The findings and the model suggest that both cognitive and emotional outcomes are important in jointly determining mental distress during the COVID-19 period. No study has tested this contention. Future longitudinal studies are warranted to test the full common sense model in the context of the COVID-19 epidemic.

This study has the strength of covering a large number of university students who were staying in most of the provinces in China. The data thus presents a crude national scenario. This study has some limitations. First, it did not have national coverage. Selection bias may exist, as classes and departments were not randomly selected. Second, we did not cover important interpersonal factors (eg, subjective norms and social support), which were associated with many health-related behaviors [50,51]. Third, the relatively mild magnitudes of some mediation and suppression effects of mental distress due to COVID-19 imply existence of other unstudied mechanisms. Last, the cross-sectional study design does not allow for causal inferences, as depression may also change perceptions. Longitudinal studies are needed to confirm these contentions.

The findings suggest that mental distress due to COVID-19 and depression were prevalent among university students in China during the initial COVID-19 outbreak period. The former may have further increased the prevalence of the latter. Various cognitive, behavioral, and psychosocial responses to COVID-19 showed both direct and indirect effects (via mental distress due to COVID-19) on depression. Thus, interventions to improve such multidimensional factors might reduce mental distress during the initial COVID-19 outbreak period. The associations between some of the studied factors and depression may change over time as more information and experiences were obtained by the public, signifying early investigation of community responses to avoid mental distress, which would carry over to later phases and affect prevention behaviors. Some of the findings may shed light on handling new emerging infectious diseases that occur in the future. It is important to validate the findings in general and specific populations in China and in other countries.