Increased Psychological Distress, Loneliness, and Unemployment in the Spread of COVID-19 over 6 Months in Germany

Background: The COVID-19 pandemic poses a challenge to global mental health. Loneliness and isolation may put people at higher risk for increased psychological distress. However, there is a lack of research investigating the development of COVID-19-related distress over time. Materials and Methods: We undertook an online survey among general population (N = 1903) in Germany throughout 6 months from the peak transmission period in April to the off-peak period by September 2020. Results: We found that the average prevalence of psychological distress caused by the COVID-19 pandemic significantly rose from 24% to 66% between the peak and off-peak transmission period, respectively. Unemployment rate and loneliness increased negative mental health outcomes, although the number of active COVID-19 cases decreased from April to September. Psychological distress scores increased mostly in female, young, and lonely people. Conclusions: Our results underline the importance of considering innovative alternatives to facilitate employment opportunities, distant contacts, and self-help over the course of the pandemic. Our study highlights the urgent need to pay attention to mental health services specifically targeting female, young, unemployed, and lonely people.


Introduction
The current COVID-19 pandemic has a massive impact on global mental health [1][2][3], causing sudden lifestyle changes through social distancing and isolation at home, with severe social and economic consequences [4,5]. Scientists urge mental health research to be central to the social contexts that affect the assessment and treatment of pandemics within and across countries [6,7].
A first wave of pioneering mental health surveys regarding COVID-19 were launched in January and February 2020 in China [8][9][10]. Among those studies, Shanghai Mental Health Center designed a COVID-19 Peritraumatic Distress Index (CPDI) questionnaire to assess psychological distress among the general population in China [10]. As the worldwide COVID-19 cases dramatically increased, several countries initiated their own nationwide mental health surveys between the end of March and the middle of May 2020 [10][11][12][13][14][15][16]. They used the same CPDI design, employed previously in China [10], to assess mental status of the general population during this period of peak transmission of COVID-19 [10][11][12][13][14][15][16]. The results show that the general population across countries experienced psychological distress and that the prevalence of psychological distress varied from low (11.5% (N = 410) in Nepal [11]) to mild and moderate (24.1% (N = 1007) in Germany [12] and 28.6% (N = 1035)

Self-Reported Psychological Distress
We used the COVID-19 Peritraumatic Distress Index (CPDI) questionnaire to capture peritraumatic psychological distress during the COVID-19 pandemic [10]. The CPDI was developed by psychiatrists from the Shanghai Mental Health Center, and we used a German version [12]. The CPDI has 24-items that are rated on a 5-point scale ranging from 0 ("strongly disagree") to 4 ("strongly agree"). A score below 28 indicates no distress, a score between 28 and 51 indicates mild to moderate distress, and a score above 52 indicates severe distress (see Supplementary Materials Table S1).

Measures of Loneliness
We used the German version [38] of the short-form UCLA Loneliness Scale  to measure an individual's subjective perception of loneliness or social isolation [39] (see Supplementary Materials Table S2).

Data Analysis
Statistical tests of the CPDI and ULS-8 questionnaires were performed using SPSS Statistics Version 26 (SPSS Inc., Chicago, IL, USA). Differences were considered statistically significant at p < 0.05 and highly statistically significant at p < 0.001. We used a one-way ANOVA to assess psychological responses over time for each of the six time periods. In the ANOVA, we used "month" (April, May, June, July, August, and September) as an independent variable and the respective "CPDI score" as a dependent variable. We also tested sociodemographic differences between the six-month groups in order to control its potential confounding effect. To test whether change in loneliness can account for the change in distress across the months, we performed an analysis of covariance and used "month" as a fixed factor and "loneliness score" as a covariate and we also explored their interactions. To further investigate if such changes are associated with the sociodemographic variables (i.e., gender, age, and years of education), we conducted a multiple linear regression analysis in R version 4.0.3. (www.r-project.org). We built a first model with "ULS-8 loneliness scores", "month", "gender", "age", and "years of education" as predictors and "CPDI distress scores" as the outcome and a second model with "month", "gender", "age", and "years of education" as predictors and "ULS-8 loneliness scores" as the outcome. The assumptions for both models were visually checked. An independent t-test was used for pairwise comparisons of the respective peak (in April) and off-peak (from May to September) periods, high versus low levels of loneliness groups, males versus females, elder versus younger groups, and high versus low levels of education groups (two-tailed p values were assumed).
We tested the hypotheses that (1) there is an increase in self-reported psychological distress over the course of the pandemic, (2) the increase in self-reported distress is significantly associated with joblessness as an indicator of an economic downturn, and (3) lonely persons feel more distressed. We explored whether the number of active COVID-19 cases as an indicator of the risk to become infected was correlated with distress and loneliness, and we explored the effects of age and gender.

Group Description
From April to September 2020, 1903 respondents in Germany (1437 females; age range: 18-81, mean = 38.32, SD = 13.40) participated in our survey. Each participant completed the survey only once. The participants' sociodemographic variables and group comparisons are presented in Table 1. With regard to sociodemographic variables between the six-month groups, we did not find a significant difference in "gender" (F (5, 1897) = 0.98, p = 0.431) but we found significant differences in "age" (F (5, 1889) = 43.55, p < 0.0001) and "education" (F (5, 1828) = 30.00, p < 0.0001) among the six groups. In general, there were gender, age, and education differences in CPDI scores. Female respondents experienced relatively higher levels of psychological distress compared to males, t (1901) = 3.01, p = 0.003. Younger respondents experienced relatively higher levels of psychological distress compared to elders, t (1836.830) = −8.59, p < 0.001. Respondents with fewer years of education experienced relatively higher levels of psychological distress compared to those with more years of education, t (1662.836) = −9.66, p < 0.001, as shown in Table 1. Over the last six months (183 days), in Germany, there have been an average of 19,630 active cases per day, which reached its peak with an average of 54,715 active cases in April (with a peak number on 6 April 2020 at 72,865 cases on this day) and an off-peak average of 6739 cases in July (down to 5882 total cases in a single day on 19 July 2020).

Increased Psychological Distress over the Course of the Pandemic in Germany
Self-reported psychological distress increased over the course of the pandemic in Germany, F (5, 1897) = 112.17, p < 0.001. There was a significant difference in psychological outcomes between the peak versus off-peak transmission period of COVID-19 in Germany, t (1485.474) = −22.09, p < 0.001. The average COVID-19 Peritraumatic Distress Index (CPDI) score for the off-peak transmission period from May to September (mean = 38.94, SE = 19.84) was 17.08 higher than the average CPDI score for the peak transmission period in April (mean = 21.85, SE = 12.64), shown in Figure 1.
outcomes between the peak versus off-peak transmission period of COVID-19 in Germany, t (1485.474) = −22.09, p < 0.001. The average COVID-19 Peritraumatic Distress Index (CPDI) score for the off-peak transmission period from May to September (mean = 38.94, SE = 19.84) was 17.08 higher than the average CPDI score for the peak transmission period in April (mean = 21.85, SE = 12.64), shown in Figure 1.
During the period of off-peak transmission of COVID-19 pandemic, 66% of respondents in Germany experienced psychological distress varying from 40% mild to 26% severe. On the other hand, during the period of peak transmission of COVID-19, only 24% of populations in Germany reported distress with 21% mild and 4% severe stress.
Interestingly, the daily average of active cases per month did not statistically significantly correlate with perceived distress (Pearson r = −0.658, p = 0.155, two-tailed). Instead, increased perceived distress over time was significantly associated with increased unemployment rate in Germany (Pearson r = 0.874, p = 0.023, two-tailed). An increase in the unemployment rate was statistically significantly associated with the prevalence of active COVID-19 cases (Pearson r = −0.886, p = 0.019, two-tailed), indicating that people could be left without work due to COVID-19, which in turn affected perceived stress levels. Data of the unemployment rate in Germany were retrieved from the Federal Statistical Office of Germany (https://www.destatis.de/EN/Themes/Economy/Short-Term-Indicators/Labour-Market/arb210a.html). Date of the active cases of COVID-19 in Germany (i.e., by removing deaths and recoveries from total cases) were retrieved from Worldometer's COVID-19 data (https://www.worldometers.info/coronavirus/country/germany/). The monthly average of active cases was calculated by the sum of daily active cases divided by the number of days in that month.

Loneliness Predicts Psychological Distress
We found statistically significant differences in psychological distress scores between the groups by month when adjusted for ULS-8 loneliness score, F (4, 890) = 11.36, partial η 2 = 0.049, p < 0.0001, and there was no significant interaction between independent variable "month" and the covariate "ULS-8 loneliness score", F (4, 886) = 1.65, partial η 2 = 0.007, p = 0.161, indicating that perceived loneliness predicted change in psychological in Germany (i.e., by removing deaths and recoveries from total cases) were retrieved from Worldometer's COVID-19 data (https://www.worldometers.info/coronavirus/country/germany/). The monthly average of active cases was calculated by the sum of daily active cases divided by the number of days in that month.
During the period of off-peak transmission of COVID-19 pandemic, 66% of respondents in Germany experienced psychological distress varying from 40% mild to 26% severe. On the other hand, during the period of peak transmission of COVID-19, only 24% of populations in Germany reported distress with 21% mild and 4% severe stress.
Interestingly, the daily average of active cases per month did not statistically significantly correlate with perceived distress (Pearson r = −0.658, p = 0.155, two-tailed). Instead, increased perceived distress over time was significantly associated with increased unemployment rate in Germany (Pearson r = 0.874, p = 0.023, two-tailed). An increase in the unemployment rate was statistically significantly associated with the prevalence of active COVID-19 cases (Pearson r = −0.886, p = 0.019, two-tailed), indicating that people could be left without work due to COVID-19, which in turn affected perceived stress levels.
There was a significant difference in self-reported psychological distress between individuals who reported high versus low levels of loneliness, t (850.392) = 20.03, p < 0.001. The average CPDI score for individuals who reported higher levels of loneliness (mean = 49.86, SE = 17.52) was 22.68 higher than the average CPDI score for those who reported lower levels of loneliness (mean = 27.18, SE = 15.56), shown in Figure 2.  Due to a high proportion of female participants, we also conducted additional analysis of males and females separately. We also found that the same change in perceived loneliness over time predicted change in psychological distress in males and females. In male participants, the predictors "ULS-8 loneliness scores" (b = 2.22, t (197) = 12.23, p < 0.0001) and "month" (b = 3.20, t (197) = 2.60, p < 0.01) statistically significantly increased psychological distress, and in female participants, the predictors "ULS-8 loneliness scores" (b = 2.16, t (656) = 21.90, p < 0.0001) and "month" (b = 2.72, t (656) = 4.79, p < 0.0001) statistically significantly increased psychological distress. Due to a high proportion of female participants, we also conducted additional analysis of males and females separately. We also found that the same change in perceived loneliness over time predicted change in psychological distress in males and females. In male participants, the predictors "ULS-8 loneliness scores" (b = 2.22, t (197) = 12.23, p < 0.0001) and "month" (b = 3.20, t (197) = 2.60, p < 0.01) statistically significantly increased psychological distress, and in female participants, the predictors "ULS-8 loneliness scores" (b = 2.16, t (656) = 21.90, p < 0.0001) and "month" (b = 2.72, t (656) = 4.79, p < 0.0001) statistically significantly increased psychological distress.

General Discussion
This study is among the first national-sample studies to track temporal changes in population mental health from the lockdown period to the subsequent period of rather low active COVID-19 cases per months. Consistent with our hypothesis, we found an overall increase in psychological distress over 6 months' time within a general population in Germany. Additionally, we found that this increase in psychological distress was associated with the persistent and substantial rise of unemployment rates in Germany. Moreover, our data suggests that years of education is a protective factor for psychological distress. This might be due to higher income, higher job security, or home office opportunities associated with increased level of education. Finally, we found that loneliness predicted COVID-19-related distress.
According to the Federal Statistical Office of Germany (Destatis), the unemployment rate in Germany remained stable at 5.1% until March 2020 when restrictions were in place due to COVID-19. Since then, the unemployment rate has risen rapidly from 5.8% in April to 6.4% in August by a 0.1% monthly rise (https://www.destatis.de/EN/Themes/ Economy/Short-Term-Indicators/Labour-Market/arb210a.html). Cross-sectional studies showed that unemployment was associated with increased mental health problems, depression and anxiety disorders especially among young people [18]. Unemployment insurance (i.e., government payments to eligible individuals to replace part of pre-job loss income during their job search, "Arbeitslosengeld" in Germany) may act as a "safety net" not only to protect individuals from hardship but also to enhance mental health resilience [19]. Wanberg et al. (2020) found that higher perceived unemployment insurance generosity was associated with better mental health via reduced time pressure and financial strain [19].
Beyond creating employment opportunities and enhancing unemployment insurance, other strategies that enable individuals to remain active in their job search included productive use of time on meaningful activities, conserving financial resources, social supports, and cognitive techniques (e.g., holding a positive outlook, reevaluating expectations, reappraisal, determination, and agency) [40]. In the era of COVID-19, researchers suggest that it is high time to design evidence-based interventions for unemployed individuals and to examine work-family balance and unemployment issues among youth [24].
Our data suggest that loneliness is a predictor for COVID-19-related distress. In Germany, a sharp spike in people seeking for mental health help appears to be mainly driven by anxiety, depression, and loneliness [41], and the federal states of Germany, where stricter measures were in place, such as "stay-at-home" orders in Bavaria, Saarland, and Sachsen-Anhalt, had an even stronger increase in helpline contacts [41]. This may be particularly challenging for females and young people [17,42,43]. To deal with the immediate and long-term consequences of COVID-19, it is essential to tap in the power of social community participation [44,45]. Qi et al. (2020) found that lower levels of social support were associated with a higher prevalence of mental health problems among adolescents during the COVID-19 outbreak [44]. Thus, it appears to be necessary to provide incentives and to support vulnerable individuals in response to COVID-19 [45]. For example, online communities are a way to promote social participation and tangible support by sharing online workshops, resources, and guides [24]. Internet-and mobilebased psychological testing and assessment can also capture a mental health profile of the community and can provide evidence to innovate and strengthen practice and policy amid a global pandemic [46,47]. For example, a study on the Germany's largest free mental health helpline "TelefonSeelsorge" (https://www.telefonseelsorge.de/) showed that the demand for mental health services has been increased by approximately 20% from 1800 to 2400 contacts per day in the first week of the lockdown [41].

Limitations
Our current results must remain tentative due to the lack of daily and momentary state variations in psychological distress to detect more subtle effects of COVID-19-related restrictions and economic downturn on mental health. Our survey neither elicited an individual's employment nor disentangled the causal interplay between unemployment and poor mental health. An increase in fear of unemployment may also worsen mental health [21]. Our study is also limited by a lack of validated and comparable, pre-COVID-19 baseline data against which to measure change either within individuals or across the population as a whole. In addition, pre-pandemic life circumstances may also remain important determinants of people's mental health during the pandemic. The generalizability of results is limited to populations that share characteristics with our sample (e.g., the population has access to the Internet to complete our survey). At the time of our study, the COVID-19 crisis is still unfolding, which further limits the generalizability of our findings. We hope that the reported descriptive associations encourage replications of the research design in different settings and further theoretical elaboration.

Outlook and Future Perspectives
Intermittent social distancing may remain in place until 2022, and a resurgence of COVID-19 is possible until 2024 [36]. An inclusive response to and recovery from COVID-19 requires an integrated longitudinal approach to anticipate the impact across different target groups. The use of mobile digital health interventions, such as mobile health (mHealth) tools, is one possible solution to deliver the objective to mitigate negative psychosocial consequences of the COVID-19 pandemic [33]. Virtual reality exercise has also been used in mental health and wellness promotion [48]. Education, self-care, and family support should form part of mental health prevention strategies, which can involve multiagency collaboration among housing, education, and employment services, with support from the voluntary and mental health sectors [49]. These interventions could include digital forms of study groups, peer group sessions, mentoring, and psychological counseling [50]. Friendship, interaction, social support, and studying with others have been argued to impact their well-being and academic success, but they often require meeting opportunities and informal settings to develop [51]. In a long-term perspective, it is important to examine whether the rules, knowledge, attitudes, and coping styles developed during the pandemic are maintained post-pandemic. It might be a long and unclear road until the COVID-19 pandemic ends. Future studies are needed to investigate stress changes during the course of the pandemic, particularly among vulnerable groups including females and young people [17,42,43], people with chronic diseases [52], and medical professionals [53]. Integration of long-term and short-term research in response to COVID-19, recovery strategies, and resilience may help us understand what we need to do to better manage future epidemics and pandemics.

Data Availability Statement:
The data presented in this study are available on request from the authors.