An international study was conducted, which aimed to examine the global impact of the COVID-19 pandemic on lifestyle behaviors, fear, depression, and perceived needs of communities [26,27]. The study was conducted in 30 countries across the globe. It is a cross-sectional web-based survey design. Moreover, a web-based questionnaire platform was developed for data collection [28].

For this analysis, only data collected in mainland China and Hong Kong between July 2020 and January 2021 were used. Study eligibility criteria included (1) aged ≥18 years; (2) being able to read and understand Chinese; and (3) having an internet access. To recruit more people with diverse sociodemographic backgrounds, multiple recruitment strategies were used to recruit study participants. The study participants were recruited by survey service providers, social media platforms such as Facebook, WeChat, and Twitter, and snowball sampling, in which the existing study participants helped to recruit additional participants to join this study. To encourage more people to complete the survey, for each completed questionnaire, HK $1 (US $0.13) would be donated to the Red Cross in the respondent’s region.

The study protocol has been published elsewhere [26]. The study was approved by the institutional review board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (reference UW 20-272). All the procedures involving human participants in this study were conducted in accordance with the ethical standards of the institutional review board and the 1964 Declaration of Helsinki and its later amendments. Electronic informed consent was obtained from each study participant.

The primary outcome of the study was the fear of COVID-19. To measure the fear of COVID-19, we adapted the 8-item Breast Cancer Fear Scale developed by Champion et al [21] for this study. The study instrument was originally developed to measure women’s emotional responses to breast cancer. In the scale developed by Champion et al [21], a 5-point Likert scale is used (1=strongly disagree, 2=disagree, 3=neutral, 4=agree, and 5=strongly agree); a higher score indicates a higher level of fear. The total score of the instrument is the sum of each item. In this paper, we changed the words “breast cancer” to “COVID-19” in all of the following 8 items: (1) the thought of COVID-19 scares me; (2) when I think about COVID-19, I feel nervous; (3) when I think about COVID-19, I get upset; (4) when I think about COVID-19, I get depressed; (5) when I think about COVID-19, I get jittery; (6) when I think about COVID-19, my heart beats faster; (7) when I think about COVID-19, I feel uneasy; and (8) when I think about COVID-19, I feel anxious.

The face validity of the adapted instrument was evaluated by an expert panel of this study.

The 4-item Patient Health Questionnaire (PHQ-4), which measures anxiety and depressive symptoms, was administered to evaluate the convergent validity of the Fear Scale. The PHQ-4 includes the 2-item Generalized Anxiety Disorder scale and the 2-item Patient Health Questionnaire (PHQ-2). A 4-point Likert scale is used (0=not at all; 1=several days; 2=more than half the days; and 3=nearly every day). The summary score of the PHQ-4 ranges from 0 to 12, with a higher score indicating greater anxiety and depressive symptoms. The PHQ-4 was validated in Chinese adults [29]. The study supported its 2-factor model and reliability [29]. Cronbach α coefficient was .87 for the PHQ-4, .80 for the 2-item Generalized Anxiety Disorder, and .80 for the PHQ-2 in this study.

A structured questionnaire was used to collect sociodemographic factors such as age, gender, and comorbidities.

The internal construct validity, convergent validity, known group validity, and reliability of the Fear Scale were assessed. The internal construct validity was evaluated using the corrected item-total correlation; a correlation coefficient of ≥0.4 indicated adequate internal construct validity. The convergent validity of the Fear Scale was determined by calculating the Pearson correlation coefficient between the total score of the Fear Scale and the total score of the PHQ-4. It was hypothesized that an absolute value Pearson correlation coefficient of at least 0.3 was required [30]. To evaluate the known group validity, independent t tests were used to compare the mean score of the Fear Scale between (1) people recruited from mainland China and people recruited from Hong Kong [31]; (2) people aged 18-59 years and people aged 60 years or older [15]; and (3) male and female participants [32]. A study among Chinese university students reported that students in mainland China had lower fear of instability related to the COVID-19 pandemic when compared with students in Hong Kong [31]. Another study among pregnant women and new mothers reported that compared with the study participants in mainland China, the level of fear related to the COVID-19 pandemic was significantly higher among study participants in Hong Kong [33]. A study in Singapore found that older age was associated with greater fear of COVID-19 [15]. Another study in the Spanish population found that fear was higher among women than among men [32]. Besides, a study in Turkey reported that the COVID-19 fear scores were higher among people with a chronic disease [34]. Therefore, we also compared the mean score of the Fear Scale between people with and without the following chronic diseases, which were highly prevalent in Chinese populations: (1) hypertension; (2) diabetes; (3) liver disease; (4) heart disease; (5) stroke; (6) chronic obstructive pulmonary disease; (7) cancer; (8) depression; (9) anxiety; and (10) insomnia. Cohen d effect sizes were also calculated. The interpretation of the effect sizes was as follows: trivial (<0.2), small (≥0.2 to <0.5), moderate (≥0.5 to <0.8) and large (≥0.8). Finally, descriptive statistics were used to describe the fear levels of the study participants. Furthermore, multiple linear regression analysis was used to explore the known associations between sociodemographic and clinical factors, on the one hand, and the Fear Scale, on the other.

A total of 2822 study participants aged 18 years or older were included in the analysis, of whom 61% (n=1721) were female and 38.8% (n=1096) were male. Three-quarters, 75.6% (n=2133), were recruited from Hong Kong while 24.4% (n=689) were recruited from mainland China. Almost half (1450, 51.4%) of the participants were married. Hypertension (294, 10.4%) was the most prevalent reported chronic disease, followed by diabetes (155, 5.5%), insomnia (90, 3.2%) and heart disease (62, 2.2%). Table 1 shows the participants’ characteristics of the study.

The mean score of the Fear Scale was 23.60 (SD 6.64), and the Cronbach α coefficient was .93. The corrected item-total correlations were >0.7 for all items. Table 2 shows the results of the internal consistency and internal construct validity. The Pearson correlation coefficient between the Fear Scale and PHQ-4 scores was 0.23 (P<.001). Table 3 shows the results of the convergent validity.

With respect to the known group comparisons, the results of the independent t tests showed that the study participants who were recruited from Hong Kong had a higher level of fear compared with the study participants from mainland China (Cohen d effect size 0.24). Furthermore, older adults (60 years or above) had a higher level of fear than younger adults (Cohen d effect size 0.39). Study participants with cancer (Cohen d effect size 0.58), heart disease (Cohen d effect size 0.44), hypertension (Cohen d effect size 0.36), liver disease (Cohen d effect size 0.33), insomnia (Cohen d effect size 0.33), and anxiety (Cohen d effect size 0.28) had a higher level of fear than those without such conditions. Table 4 and Table 5 show the results of the known group comparisons by independent t test.

The results of multiple linear regression are shown in the Multimedia Appendix 1.

In total, 47.1% (n=1330) of the participants reported that the thought of COVID-19 scared them. Moreover, 36.6% (n=1032) of the study participants reported that they felt nervous when they thought about COVID-19. About one-third of the participants reported that they felt uneasy (1003, 35.5%) and became depressed (923, 32.7%) when they thought about COVID-19. The descriptive statistics of the Fear Scale are shown in Table 2.

We also separated the analysis between data collected in Hong Kong and those collected in China. Those results are shown in Multimedia Appendix 1.

In the first part of this study, we assessed the psychometric properties in terms of internal construct validity, convergent validity, known group validity, and reliability of the Fear Scale. The Cronbach α coefficient was .93, which is far larger than the recommended cut-off value of .7. This finding supports the general agreement between the 8 items that make up the composite score of the scale to measure the fear related to COVID-19. Moreover, we found that all the coefficients of the item-total correlation, corrected for overlaps, were larger than 0.4, supporting the internal construct validity of the modified scale. These results supported the suggestion that all individual items measured the same construct as that measured by the other items. With regards to the study’s convergent validity, we found a small-to-moderate correlation between the total score of the Fear Scale and the total score of the PHQ-4. Another important finding of this study was that participants with a chronic disease had a higher fear level than those without a chronic disease. Particularly, we found that hypertension, liver disease, heart disease, cancer, anxiety, and insomnia were associated with a higher fear level.

There were some limitations in this study. First, the study was conducted in mainland China and Hong Kong. Therefore, the study findings may not be transferable to other geographic areas in which the severity of COVID-19, case fatality rate, and infection control measures are different. We expect that the fear level would be even higher in areas where the severity and case fatality rate of COVID-19 were more severe. Second, we could not explore the trajectory of the fear levels over time due to the cross-sectional nature of the study. Third, we adapted the Breast Cancer Fear Scale in this study; thus, some of the constructs related to COVID-19 could not be measured. However, as previously mentioned, there was no validated fear scale specific to COVID-19 when we planned our study. Fourth, regarding the reliability of the scale, we only evaluated its internal consistency. We were not able to evaluate the test-retest reliability of the scale due to the cross-sectional design of the study. Fifth, regarding the known group comparison, the sample size of some subgroups was small such as that of patients with diabetes and depression. There might be insufficient statistical power to detect the differences between groups. Finally, we used a web-based questionnaire platform to collect the data. People with low computer literacy would probably be excluded from the study. Accordingly, the potential sampling bias should be noted.

We found that the total score of the Fear Scale had a higher correlation with the PHQ-4 anxiety subscale than with the PHQ-4 depression subscale. In fact, there are distinct differences in psychological features between fear and depression. According to Witte [35], fear is conceptualized as negatively toned emotion accompanied by a high level of physiologic arousal stimulated by a threat. Fear can be expressed as a physiological arousal, such as feeling “jittery” and “heart beating faster,” through verbal self-reports of fear (eg, “I feel scared”) and overt acts that exhibit fear, such as facial expression [21]. These emotional and physiological reactions to perceived threats are fundamentally different from those of depression, which is manifested through the following 4 symptom clusters: (1) emotional symptoms such as feeling sad and worthless; (2) cognitive symptoms such as a negative view of the self and hopelessness; (3) motivational symptoms such as a lack of incentive; and (4) somatic symptoms such as a loss of appetite and sleep disturbances [36,37]. Additionally, it was suggested that fear and anxiety are largely distinct emotions. A meta-analysis reported only a moderate (r=0.32) relationship between measures of trait fear and anxiety [38]. Fear is an aversive psychological state during which an individual is motivated to escape a specific and imminent threat. The characteristics of fear include short-lived arousal that quickly dissipates after the threat is avoided. By contrast, anxiety is an aversive psychological state that occurs while an individual approaches an ambiguous and uncertain threat. Hypervigilance and hyperarousal are the typical behaviors characteristic of anxiety [38]. The small-to-moderate correlation between the Fear Scale and the PHQ-4 further supported the need for this study, which adapted and validated the Fear Scale to measure the fear of COVID-19. Besides, compared with study participants recruited in Hong Kong, those recruited from mainland China had a higher PHQ-4 score but lower Fear Scale total score. This finding further suggested that the constructs of fear and anxiety are different.

Participants with a chronic disease had a higher fear level than those without one. This finding was consistent with that of a matched case-control study, which found that the prevalence of anxiety symptoms and depressive symptoms and the level of stress were significantly higher among those with preexisting chronic health conditions (59%, 71.6%, and 73.7%, respectively) compared with controls (25.6%, 31.1%, and 43.3%, respectively) [39]. Evidence has suggested that the presence of comorbid chronic conditions would increase the risk of death from COVID-19 [40-42]. Moreover, one major concern with the COVID-19 pandemic was its impacts on the routine use of health care services especially for individuals with comorbidities [43]. Service disruptions due to cancellations of elective care and lockdowns hindering access to health care facilities, in addition to the diffidence of patients with a chronic disease in seeking assistance for fear of risking iatrogenic exposure, altogether increased the psychological burden of patients with a chronic disease. Thus, it was not surprising that people with a chronic disease had a higher fear level than those without.

In this study, more than 40% of the study participants reported that the thought of COVID-19 scared them. About one-third of the study participants reported that when they thought about COVID-19, they felt nervous, uneasy, and depressed. No doubt, the COVID-19 pandemic was very stressful for people and the communities in general [7]; the fear of infection was very common during the pandemic. Furthermore, people were worried that the health care system could not cope with the COVID-19 pandemic, that there were not enough hospital beds and ventilators to handle the rising number of COVID-19 cases. Another concern weighing on people’s minds was the COVID-19 recession. Fear of the COVID-19 pandemic can be overwhelming and cause strong emotions [7]. It was also noteworthy that the COVID-19 pandemic rekindled fears of the 2003 severe acute respiratory syndrome epidemic in mainland China and Hong Kong.

First, based on the psychometric evaluation, we found that the adapted scale was a valid and reliable measure to assess the level of fear related to COVID-19. Further studies can use this scale to longitudinally monitor the fear level in different communities. Second, given the high fear levels found in the study sample, it is required to provide psychosocial care for the general public to diminish the psychological burden of the pandemic. Third, the findings call for the need to provide more psychosocial care for chronic disease patients and older adults.

This study found that the psychometric properties of the Fear Scale were acceptable to evaluate the fear level of the general Chinese population. Our descriptive analysis found that more than 40% of the study participants reported nervousness when they thought about COVID-19. About one-third of the study participants reported that when they thought about COVID-19, they felt nervous, uneasy, and depressed. Additionally, we found that people with a chronic disease reported a higher fear level than those without. The findings call for the need to provide more psychosocial care for chronic disease patients and older adults.