Psychological Distress among University Staff before and during the COVID-19 Pandemic

(1) Background: The COVID-19 pandemic has distressed many populations worldwide, and since its beginning, many institutes have performed cross-sectional studies to assess mental health. We longitudinally examined psychological distress and depressive symptoms among university staff in Japan from 2019 to 2021, before and during the COVID-19 pandemic.; (2) Methods: Participants were teachers and hospital staff working at institutions related to Kobe University, who completed the Brief Job Stress Questionnaire (BJSQ) from 2019 to 2021. This study used the definition recommended by the guideline to identify high-stress. We analyzed the relationship between those who identified as having high-stress before versus during the COVID-19 pandemic using logistic regression analysis (adjusted for age, sex, and occupation).; (3) Results: Results showed that Stress Reaction scores increased slightly in 2020 and significantly in 2021. Time and other factors had a synergistic effect on mental health. The increase in Stress Reaction was significantly associated with females and nurses over the three years. Those with high-stress in 2019 had approximately twenty-fold odds ratios (OR) of having high-stress in 2020 and 2021.; (4) Conclusions: The long-term COVID-19 pandemic may disturb university staff’s mental health. Those who originally experienced high levels of stress were vulnerable to the negative effects of the COVID-19 pandemic.


Introduction
Since December 2019, repeated coronavirus disease-2019 (COVID-19) outbreaks have distressed populations worldwide due to lifestyle changes, social isolation, behavioral restrictions, and fear of infection. The prevalence of depression during the pandemic is seven times higher than the estimated global prevalence [1]. Mental health status is worse than before the COVID-19 pandemic in the Japanese population [2,3]. For the first time in 11 years, the number of suicide deaths in Japan has increased compared to before the pandemic [4].
Although mental health problems for HCWs have gathered attention, the COVID-19 pandemic has also negatively affected mental health in school-related populations [23][24][25]. Most studies have focused on students; few have examined teachers' mental health [6,9,12,13]. According to a systematic review of teachers' mental health, psychiatric symptoms vary between countries, highlighting the need for more research on the impact of the COVID-19 pandemic on teachers' mental health [9]. Moreover, previous research has shown that the pandemic's impact on mental health varies among teachers in elementary and junior high schools in Japan [26]; however, there is insufficient research investigating the mental health of Japanese university teachers. Wakui et al. suggested that administrative institutions should understand teachers' psychological and physical stress using surveys because their mental health influences students' education and development [26].
In a previous cross-sectional study, we investigated HCWs and students' mental health during the first wave of the COVID-19 pandemic using the patient healthcare questionnaire-9 (PHQ-9), which demonstrated that nurses, females, and younger people were likely to have depression symptoms [19], which aligned with findings of previous research [8,15,21,22]. However, it was impossible to fully assess COVID-19's impact on mental health due to the cross-sectional nature of that work. Therefore, we compare before and during the COVID-19 pandemic and investigate longitudinally to substantially uncover the pandemic's long-term mental health effects. We hypothesized that nurses, females, and younger people have mental distress at workplaces before the COVID-19 pandemic and the long-term pandemic may exacerbate our mental health.
To address these gaps, we examined psychological distress and depressive symptoms in the workplace and changes over time among university staff, including teachers, HCWs, and office clerks, before and during the COVID-19 pandemic in Japan.

Study Design
This retrospective cohort study covers the period from 2019 to 2021, one year before and two years after the onset of the COVID-19 pandemic. We included participants working at institutions related to Kobe University who had undergone a stress check for all three years and agreed to participate in our study. We used the Brief Job Stress Questionnaire (BJSQ), which is performed annually for staff at Kobe University, to assess stress [27]. The BJSQ is the standard method for evaluating occupational stress and is recommended by the Japanese Ministry of Health, Labor and Welfare [27]. Information is gathered through online questionnaires and the distribution of questionnaires. All participants were given the opportunity to refuse to enroll in our study.

Participants
One person was omitted because of refusal. A total of 10,177 participants working at institutions related to Kobe University (including a hospital, combined junior and senior high school, special needs school, elementary school, and preschool) underwent a stress check that the workplace conducts annually, based on the Industrial Safety and Health Act, from 2019 to 2021 [27]. A total of 1609 participants (823 females and 786 males; mean age: 44.58 ± 11.02 years in 2021) underwent the stress check all three years. The BJSQ is performed annually in November. We classified workers into six categories: office clerks, teachers, nurses, allied health professionals, doctors, and others, in accordance with the departments and workplace. We defined medical teachers in clinical fields as doctors.

Measures
The BJSQ was developed by Shimomitsu and authorized by the Ministry of Health, Labor and Welfare for implementing stress checks in workplaces across Japan; previous studies have demonstrated acceptable reliability and validity [27][28][29][30][31]. The BJSQ has been used in various studies [32][33][34][35][36]. It contains 57 items scored on a four-point Likert-type scale, and items are largely categorized into three components: Job Stressor (items 1−17, 17−68 points), Stress Reaction (items 18−46, 29−116 points), and Social Support (items 47−55, 9−36 points); items 56 and 57 are excluded because they are not used on the stress check [29,37] (Supplementary, BJSQ English version). Higher total scores for each category indicate higher stress and poorer social support. Job Stressor comprises nine subscales: quantity and quality of psychological job demands, physical burden, interpersonal relationships, work environment, job control, skill utilization, suitability of work, and meaningfulness of work. Stress Reaction comprises six subscales: lack of liveliness, irritability, fatigue, anxiety, depression, and somatic symptoms. Three of these categories are considered when diagnosing major depressive disorder according to the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5). We posited that the condition in which Stress Reaction scores are high and psychological distress is exacerbated may be indicative of depression. Social Support comprises three subscales: support from superiors, co-workers, and family/friends. The depression subscale in Stress Reaction comprises the scores of items 30 to 35 (6−24 points) [29,37] (Supplementary, BJSQ English version).
The Ministry of Health, Labor and Welfare guideline recommends the following criteria to define high-stress: a Stress Reaction of 77 points or higher; or the sum of Job Stressor and Social Support of 76 points or higher, in addition to a Stress Reaction of 63 points or higher [29]; this study used the same definition. High-stress is considered to have a sensitivity of 60.5% and specificity of 89.9% when screening for psychological distress, defined as a Kessler 6 scale of more than 13 points [30]. Those diagnosed with high-stress by the BJSQ were at risk of taking long-term sickness absence for 6.59 hazard ratios (HR) in males and 2.77 HR in females [31].

Statistical Analysis
Statistical analyses were performed using R version 4.0.3 (R Development Core Team, Vienna, Austria) with EZR version 1.54 (Jichi Medical University, Saitama, Japan) [38] and SPSS version 26.0 (IBM Corp, USA). The continuous variables for identical participants over three years were analyzed using repeated measures of one-or two-way ANOVA (factors: sex, occupation, or age in 2021); two-way ANOVA indicated interaction and main effect of p-values, respectively, and post-hoc analysis was performed using the Bonferroni method. Categorical variables were analyzed using Fisher's exact test. We performed multiple linear regression analysis to address confounding factors (adjusted for age, sex, occupation, and [years]). We analyzed the relationship between those identified as having high-stress in 2019 and 2020 or 2021 using logistic regression analysis (adjusted for age in 2021, sex, and occupation). We defined two-tailed p-value < 0.05 as statistical significance. Dummy variables were used, as necessary (occupation, office clerk = 0 and all other occupations = 1; sex, male = 0 and female = 1; [year, 2019 = 0 and all other years = 1]). Table 1 summarizes the demographic data and the number of high-stress populations.

Stress Reaction, Job Stressor, and Social Support
The Stress Reaction category of the BJSQ and its Depression subscale showed a similarly increasing tendency each year, and figures based on each factor (including sex, occupation, and age in 2021) mostly indicated the same tendency ( Figure 1). The scores of Stress Reaction in 2021 have increased significantly compared with 2019 and 2020; the results are similarly based on each factor (Table S1). Other categories of the BJSQ are described in Figure S1. The interaction of Stress Reaction showed a significant difference for all factors (Time/sex, p = 0.011; Time/age, p < 0.001; Time/occupation, p < 0.001) (

Factors Associated with Stress at the Workplace
Following adjustment for confounding factors (age, sex, and occupation), the increase in scores of Stress Reaction was significantly associated with females and nurses over three years, before and during the COVID-19 pandemic(females, p < 0.001 in 2019, p < 0.001 in 2020, p < 0.001 in 2021; nurses, p = 0.012 in 2019, p < 0.001 in 2020, p < 0.001 in 2021) ( Table 2). Teachers, nurses, allied health professionals, and doctors were related to elevated scores of Job Stressor during the 19 pandemic (teachers, p < 0.001 in 2020, p = 0.022 in 2021; nurses, p < 0.001 in 2020, p < 0.001 in 2021; allied health professional, p < 0.001 in 2020, p < 0.001 in 2021; doctors, p < 0.001 in 2020, p = 0.001 in 2021) ( Table 2). Higher age led to an increase in scores of Social Support, despite the pandemic (higher age, p < 0.001 in 2019, p < 0.001 in 2020, p < 0.001 in 2021) ( Table 2).   Table S4 shows the risk factors associated with the three components of the BJSQ. After adjusting for confounding factors (age, sex, occupation, and years), increased scores of Stress Reaction were associated significantly with females, nurses, and allied health professionals both before and during the COVID-19 pandemic (p < 0.001, p < 0.001, p = 0.046, respectively) (Table S4). Teachers, nurses, allied health professionals, and doctors were associated with increased scores of Job Stressor; the relation was stronger for HCWs than teachers, based on the standard partial regression coefficient (teachers [the standard partial regression coefficient = 0.0698/p-value < 0.001); nurses (0.3282/p < 0.001); allied health professionals (0.1465/p < 0.001); doctors (0.0856/p < 0.001)) ( Table S4). Social Support's standard partial regression coefficient for teachers was less than other occupations (teachers (the standard partial regression coefficient = − 0.0434/p-value = 0.008); nurses (0.0115/p = 0.480); allied health professionals (0.0370/p = 0.011); doctors (−0.0231/p = 0.115)) ( Table S4). Older age was found to be associated with decreases in scores of Stress Reaction and Job Stressor as well as an increase in scores of Social Support (Stress Reaction (the standard partial regression coefficient = − 0.0020/p-value = 0.002); Job Stressor (−0.0014/p = 0.030); Social Support (0.0072/p < 0.001)) ( Table S4).

Year
High using logistic regression analysis (adjusted for age in 2021, sex, and occupation). b We assessed the proportion between 2019, and 2020 or 2021 using Fisher's exact test. *** p-value < 0.001. The OR denoted in bold is significant at <0.05. CI, confidence interval; OR, odds ratio.

Discussion
To the best of our knowledge, this is the first study examining the long-term effects of mental health over three years before and during the COVID-19 pandemic among the same group of university staff in Japan. All staff had some increase in distress during the COVID-19 pandemic compared with the pre-pandemic period; the second year of the pandemic affected mental health more significantly than before the pandemic and the first year afterward. Time and factors including sex, occupation, and age synergistically affected mental health; in particular, increased Stress Reaction was related to female sex and nurses over the three years. Older people tended not to benefit from surroundings support compared to younger people. Those with high psychological distress before the COVID-19 pandemic were at approximately a twenty-fold risk of experiencing high psychological distress during the pandemic compared with those who did not have high psychological distress beforehand.
There was a slight increase in mental health problems soon after the outbreak of the COVID-19 pandemic, which was reduced approximately to pre-pandemic levels by the middle of 2020 [39]. Mental health problems peaked in April and May 2020, and depressive and anxiety symptoms decreased over time [40]; however, mental health problems remain high during the COVID-19 pandemic [40]. In addition, psychological distress effects varied significantly by population from the two months after the pandemic [41]. There were relatively few studies in 2021, limiting a definitive understanding of the longitudinal effects on mental health in Japan; our results showed that psychological distress increased slightly in 2020 compared to 2019 and significantly in 2021 compared to 2019-2020. The long duration of the COVID-19 pandemic may have had negative effects on the mental health of university staff.
We found that nurses and females felt more distress, including depressive symptoms, than other occupations and males. This was also the case before the COVID-19 pandemic, although it worsened afterward. A global study showed that the prevalence of major depression was higher in females than males before the COVID-19 pandemic [42]. Moreover, several studies showed that the prevalence of depressive symptoms during the COVID-19 pandemic was higher in females and nurses than among males and doctors [19,21,22]. Studies prior to the pandemic showed that the prevalence of depressive symptoms in nurses was relatively high; 38% in China and 18% in the USA [43,44], and the prevalence of depression was 10% in Canada [45]. Those employed in roles that involve interaction with the general public are at an elevated risk of psychological distress during the COVID-19 pandemic [46]. In Japan, females are frequently engaged in the hospitality sector, including caregivers, nurses, reception workers, and cashiers; these kinds of gender roles may impact mental health. Our findings also indicated that individuals with high psychological distress before the pandemic were approximately twenty-fold more vulnerable to distress during the pandemic. This suggests that the pandemic may have deteriorated the mental condition of those who already had distress. We emphasize the importance of regularly performing stress checks at work and handling the matter as necessary.
Older people did not feel the benefit of surroundings support from superiors, coworkers, family, and friends compared to young people. Emotional loneliness has been linked to depression in older adults [47]. However, this present study showed that young people were more likely to feel distressed, including experiencing depressive symptoms than older people, which is supported by previous studies [15,19]. It might be that other factors worsened the university staff's psychological condition, or that young people who became more stressed may have asked for social support. Our results showed that teachers were more related to a decrease in Social Support's points than other occupations, including nurses, based on the standard partial regression coefficient; teachers may make more efficient use of surroundings supports compared to nurses. Our findings could not conclude that higher stress was related directly to an increase in the demand for support from people around. Further studies are required to elucidate the relationship between age, Social Support, and workplace distress during the pandemic.
This study has several limitations. First, we did not consider social situations, including the waves of the pandemic, lockdown, and variants of COVID-19. The stress check was performed in 2020, the early stage of the third wave, and in 2021 after the fifth wave in Japan. Second, this study used self-report questionnaires and relied on subjectivity. The BJSQ was developed and validated for stress checks in the workplace and is not a specific questionnaire for major depressive disorder. Third, we could not examine mental health status before 2019. Thus, we used only one year of data from 2019, compared to the two years during the COVID-19 pandemic. Fourth, our results did not indicate the mental status of the general population. Our participants included teachers and HCWs, which may have led to biased populations. Finally, we did not collect detailed characteristics, including their household status, income, and medical history.

Conclusions
Our results showed that the psychological distress of university staff increased slightly in 2020 and significantly in 2021. The long duration of the COVID-19 pandemic may disturb mental health. Time and various factors such as sex, occupation, and age had synergistic effects on mental health. Females and nurses were associated with psychological distress before the pandemic. Those who originally felt distressed were vulnerable to the negative effects of the pandemic on mental health. While carefully monitoring the impact of a prolonged pandemic on mental health, each workplace needs to consider specific interventions for those with high stress during the pandemic.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph20032208/s1. Table S1: Stress Reaction among university staff before and during the COVID-19 pandemic; Table S2: Job Stressor among university staff before and during the COVID-19 pandemic; Table S3: Social Support among university staff before and during the COVID-19 pandemic; Table S4: Multiple regression analyses of university staff before and during the COVID-19 pandemic; Figure S1: Job Stressor and Social Support among university staff before and during the COVID-19 pandemic; The Brief Job Stress Questionnaire English version. Funding: This study received financial support from the foundation of Kinoshita Memorial Enterprise. The funding source was not involved in the study design, data collection, data analysis, interpretation, or publication decision.

Institutional Review Board Statement:
The study was approved by the Ethical Committee of Kobe University Graduate School of Medicine (registered on the Clinical Trials Registry: B190256).

Informed Consent Statement:
Our study design and all related procedures were performed in accordance with the Declaration of Helsinki. We had secondary use of the BJSQ conducted for their benefit and all participants were given the opportunity to refuse to enroll in our study. Informed consent was obtained from all subjects involved in the study. We considered preserving the anonymity of participants.

Data Availability Statement:
The raw data related to this article are subject to the following restrictions: data used for secondary purposes must be approved by the Ethical Committee of the Kobe University Graduate School of Medicine. Requests to access these datasets should be directed to the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.