Abstract
Background and aims:
The COVID-19 pandemic affected life for a large proportion of the global population. In this study, we examined changes in well-being among employees in Denmark before, during and after the COVID-19 pandemic.
Methods:
We used secondary survey data collected by a private company that had developed and provided a smartphone app to monitor employees’ well-being using the WHO-5 questionnaire. We included 6758 employees from 77 workplaces, yielding 111,705 observations of well-being scores. We investigated time trends in well-being by comparing well-being scores in the years before (2018 and 2019), during (2020 and 2021) and after (2022) the COVID-19 pandemic. We further compared well-being scores during the months of three distinct COVID-19 waves with the corresponding months before the COVID-19 pandemic.
Results:
We observed a clear seasonal variation, with higher well-being scores during the summer months (July and August) in each year, regardless of the COVID-19 situation. In both men and women, annual well-being scores were lowest during the COVID-19 years (68.6, and 67.3, respectively) compared with annual well-being scores before (68.9 and 69.3) and after (70.1 and 68.7) the COVID-19 pandemic. During three distinct COVID-19 waves in Denmark, women, but not men, showed statistically significant lower well-being scores compared with the corresponding months before the COVID-19 pandemic (all p<0.01).
Conclusions:
Keywords
Background
In March 2020, the World Health Organization (WHO) declared the global outbreak of COVID-19 a pandemic [1]. The COVID-19 pandemic affected large parts of populations worldwide, as societal lockdowns, requirements to work from home and other restrictions implemented to limit the spread of infections changed everyday life [2,3].
Numerous studies have examined the effect of the COVID-19 pandemic on population mental health and well-being. Systematic reviews and umbrella reviews have summarized these results [4 –9]. Overall, the reviews suggest that mental health and well-being may have decreased during the pandemic; however, there is major heterogeneity and uncertainty in the data. Furthermore, there are important methodological limitations in the literature, as a substantial number of studies were either cross-sectional or commenced after the pandemic already was underway.
Studies examining changes in mental health and well-being in occupational cohorts during the COVID-19 pandemic are surprisingly rare [10] and often limited to specific job groups, for example, health care workers [11]. Studies on mental health and well-being among employees working from home during the pandemic showed mixed results [12 –14]. A Canadian study reported that working from home during the pandemic was associated with lower burnout and stress levels but also with a decline in self-reported mental health [12]. A Danish study reported that workers working from home experienced a greater decline in mental health than those not working from home [13]. A Finnish study reported that different dimensions of occupational well-being (work engagement, job boredom, burnout) changed only slightly during the pandemic [14].
To our knowledge, no occupational cohort study had yet compared changes in well-being before, during and after the COVID-19 pandemic with multiple well-being measures in each phase of the pandemic. To address this research gap, we conducted a secondary data analysis of well-being measures, collected every other week, from 2018 to 2022 by a smartphone app.
Aims
We aimed to investigate monthly changes in well-being among Danish employees from 2018 to 2022, and whether there was an influence of the COVID-19 pandemic on these changes. We compare well-being scores before (2018 and 2019), during (2020 and 2021) and after (2022) the COVID-19 pandemic. We further examine well-being during three distinct COVID-19 waves that were characterized by high infection levels and several governmental measures to contain infections, such as mandatory work from home and partial lockdowns (see Supplemental material online for details).
Methods
Study design and setting
This study is a secondary data analysis of data collected by a private company that had provided a smartphone app to numerous workplaces in Denmark to measure employees’ well-being. Through the app, employees had the opportunity to complete a short well-being questionnaire every other week as part of a surveillance and intervention service. In cases of decreasing well-being, an external team of psychologists offered telephone consultations as tailored interventions to improve well-being at both the individual and workplace levels.
From 2015 onwards, the app company collected anonymized well-being data from users. Our research team was not involved in the data collection or intervention; instead, we used post-hoc the opportunity of analysing well-being data that was collected before, during and after the COVID-19 pandemic, that is, from 2018 to 2022.
Study participants
In the 59 months from January 2018 to November 2022, 21,806 employees at 143 workplaces completed the questionnaire at least once and were included in the study. We excluded 1872 employees that worked outside of Denmark, and a further 13,934 employees with missing information on gender, yielding a final sample of 6758 participants from 77 workplaces. During the study period, each participant completed the questionnaire between one and 111 times, resulting in 111,705 observations. See Table S-I in the Supplemental material for details.
Data collection
Well-being was assessed by the five-item World Health Organization Well-Being Index (WHO-5) [15]. WHO-5 is a generic questionnaire, measuring self-reported well-being by five positively worded items (see Supplemental material). Respondents were asked to consider the past two weeks, and rate each item on a six-point Likert scale. We calculated the WHO-5 score by summing up the scores of the items and multiplying by four, yielding a total score ranging from 0 (absence of well-being) to 100 (maximal well-being). Previous studies have shown that the WHO-5 is psychometrically valid, and the instrument has been used to assess well-being over time or to compare well-being between groups [15], including a randomized general population study in Sweden [16], and specific occupations during the COVID-19 pandemic [17].
In addition to well-being, information on employees’ gender and industry code of their workplace was available in the smartphone app.
All employees using the smartphone app signed a written consent allowing researchers to perform scientific analyses on anonymized data.
COVID-19 waves
We defined three major COVID-19 waves in Denmark based on high infection levels and governmental measures to contain infections [18]. The first wave was from March 2020 to May 2020, the second wave from September 2020 to April 2021, and the third wave from November 2021 to January 2022. During each wave, the Danish Government responded with several measures (e.g. restrictions on public gatherings, requirements to wear facemasks) aimed to mitigate the number of COVID-19 infections. The measures were more severe in the first and second wave and less severe in the third wave. For a detailed description, see Supplemental material.
Statistical analyses
We calculated mean well-being scores for each of the 59 months of the study (January 2018 to November 2022). To investigate whether the annual mean well-being scores in the years before (2018 and 2019), during (2020 and 2021) and after (2022) the COVID-19 pandemic differed, we used two-sample t-tests. We repeated this analysis for a subsample of 376 employees who had at least five responses in each of the three periods. Post-hoc, we examined the statistical difference between men and women using two-way analysis of variance. Next, we examined for each year whether the monthly mean well-being scores differed from the annual mean scores by using one-sample t-tests. Finally, we compared the mean well-being scores during the months of the three distinct COVID-19 waves with the corresponding months in the years before the COVID-19 pandemic, by using two-sample t-tests.
Ethics
The project was approved by the Danish Data Protection Agency through the joint notification of the National Research Centre for the Working Environment (no. 2015-57-0074).
Results
Gender and industry distribution of the study participants
Table I shows the gender and industry distribution of the study participants. In total, 6758 employees responded to the WHO-5 questionnaire at least once in the study period: 3826 men (56.6%) and 2932 women (43.4%). The largest proportion of the participants was employed in electricity, gas and district heating supply (30.6%), followed by wholesale and retail trade, repair of motor vehicles and motorcycles (13.9%), manufacturing (11.6%), and public administration, defence and social security (10.5%).
Gender and industry distribution of the study sample.
Other industries include ‘Water supply; sewage system, waste sorting and cleaning of soil and groundwater’, ‘Building and construction business’, ‘Accommodation facilities and restaurant business’, ‘Real estate’, ‘Administrative services and support services’, ‘Education’. Each of these industries has <40 n in total.
Annual mean well-being scores from 2018 to 2022
Figure 1 shows the annual mean well-being scores from 2018 to 2022, stratified by gender. In both men and women, we observed the lowest well-being score in 2021 (68.2 and 66.7, respectively). In men, well-being was highest in 2022 (70.1), whereas in women it was highest in 2018 (70.6). In men, well-being decreased from 2018 to 2019 (–0.5 points), increased slightly in 2020 (+0.3 points), decreased in 2021 (–0.8 points) and increased strongly in 2022 (+1.9 points). In women, well-being continually decreased from 2018 to 2021 (–2.3, –0.2 and –1.4 points, respectively) before increasing again in 2022 (+2.0 points). Further details are provided in Table S-II in the Supplemental material.

Annual mean well-being scores from 2018 to 2022, men and women.
In both men and women, the annual mean well-being score during the pandemic years (2020 and 2021) was statistically significantly lower than the annual mean well-being score after the pandemic (2022) (Supplemental Table S-II). In men, the mean well-being score before the pandemic was 68.9, decreased to 68.6 during the pandemic (–0.3, t= –1.88, p=0.06) and increased to 70.1 (+1.5, t= –8.56, p<0.001) after the pandemic. In women, the mean well-being score before the pandemic was 69.3, decreased to 67.3 during the pandemic (–2.0, t= –8.99, p<0.001) and increased to 68.7 (+1.4, t= –5.92, p<0.001) after the pandemic (Supplemental Table S-II). When comparing the differences in mean well-being scores in the three pandemic periods (before, during and after), we found that the differences from before to during the pandemic and from before to after the pandemic differed statistically significantly between men and women (F=29.78, p<0.001 and F=29.83, p<0.001, respectively). The differences in the mean well-being scores from during to after the pandemic did not differ statistically significant between men and women (F=0.57, p=0.45).
When we restricted the analyses to a subsample of participants who had at least five responses in each period (before, during, after the pandemic) all results among women were directionally similar in the main sample and the subsample. The change from during to after the pandemic was statistically significant in the main sample only (Supplemental Table S-III). Among men, the change in well-being from before to during the pandemic was also directionally similar in both samples (decline in well-being in both samples). The change from during to after the pandemic (increase in the main sample, decrease in the subsample) and from before to after the pandemic (increase in the main sample, decrease in the subsample) were directionally different in the two samples. See Supplemental Table S-III for details.
Monthly mean well-being scores from 2018 to 2022
Figures 2 and 3 show the monthly mean well-being scores from 2018 to 2022 in men and women, respectively. We found a clear seasonal variation with a sharp increase in well-being in July and August, followed by a sharp decrease in September in both genders. This pattern was similar across all five years, regardless of whether the year was before, during or after the pandemic (Figures 2 and 3).

Monthly mean well-being scores from 2018 to 2022, men.

Monthly mean well-being scores from 2018 to 2022, women.
In both men and women, the monthly mean well-being scores were statistically significantly higher in July/August than the annual mean well-being scores in all five years from 2018 to 2022 (p=<0.001 in each year). See Tables S-IV and S-V in the Supplemental material for details.
Mean well-being scores during three distinct COVID-19 waves
Figure 4 shows the monthly mean well-being scores from 2020 to 2022 in men and women with time indication of the three COVID-19 waves in Denmark: March 2020 to May 2020, October 2020 to April 2021, and November 2021 to January 2022.

Monthly mean well-being scores during three distinct COVID-19 waves (pale grey indications), men and women.
In men, the monthly mean well-being scores during the three COVID-19 waves were not statistically significant different from the corresponding months in the years before the pandemic (68.5 vs. 68.7, t=0.46, p=0.64; 67.7 vs. 67.7, t=0.03, p=0.98; and 68.5 vs. 67.6, t= –1.89, p=0.06, respectively) (Supplemental Table S-VI). In women, the monthly mean well-being scores during the three COVID-19 waves were statistically significantly lower than the corresponding months in the years before the pandemic (67.3 vs. 69.1, t=3.25, p=<0.01; 66.4 vs. 68.6, t=5.66, p=<0.001; and 66.0 vs. 68.7, t=4.56, p=<0.001, respectively) (Supplemental Table S-VII). Further details in both men and women are provided in Tables S-VI and S-VII in the Supplemental material.
Discussion
Summary of results
Well-being decreased from before (2018 and 2019) to during (2020 and 2021) the COVID-19 pandemic. We observed this decrease in well-being among both men and women and in both the main sample and the subsample. From during to after the pandemic, we observed an increase in well-being among both women and men in the main sample. In the subsample, we observed a small and statistically not significant increase among women and a decrease among men. Further, we found a clear seasonal variation in well-being, with well-being peaking in the summer months of July and August and sharply declining in September. We observed this pattern in both men and women and in all study years, regardless of whether the years were before, during or after the pandemic. However, the annual mean well-being differed statistically significantly when comparing years during the pandemic with years before and after the pandemic, indicating an association of the COVID-19 pandemic with lower well-being. This association was more pronounced in women than in men. When comparing three distinct COVID-19 waves with the corresponding months in the years before the pandemic, we found in women that well-being was statistically significantly lower during all three COVID-19 waves. In men, we did not find statistically significant differences for any of the three COVID-19 waves compared with the corresponding months in the years before the pandemic. Although our findings showed statistically significant changes related to the COVID-19 pandemic, the absolute changes in the scores were small. Topp et al. reviewed the literature on the WHO-5 and suggested that for clinical trials, a change of 10 points on the WHO-5 scale can be considered as clinically significant [15]. The changes in WHO-5 score that we observed in the different phases were considerably smaller (ranging from –2.0 (before to during the pandemic, women) to +1.5 (during to after the pandemic, men)). Thus, our showed study statistically significant changes but not clinically significant changes in well-being in relation to the COVID-19 pandemic.
Comparison with previous studies
To the best of our knowledge, this is the first study that examined repeatedly-assessed well-being before, during and after the COVID-19 pandemic, thus we cannot directly compare our results with results from previous studies. However, reviews have suggested that the pandemic adversely affected population mental health and well-being, though study results showed high heterogeneity [4 –9]. One review reported that women’s mental health and well-being was more affected by the COVID-19 pandemic than men’s [4], which is congruent with our findings. Furthermore, when we compared well-being in the months of three distinct COVID-19 waves with the corresponding months before the COVID-19 pandemic, we found a stronger decline in women than in men. One explanation for lower well-being among women during the COVID-19 pandemic could be a potential disproportionate impact on women, such as the double burden of caring for children while working under restrictions during lockdowns [19,20].
The vast majority of the previous reviews were based on cross-sectional studies, thus examining mental health and well-being during the COVID-19 pandemic without comparison with measures before the pandemic [4 –9]. An exception is a review by Robinson et al. on longitudinal studies that compared mental health before and during the COVID-19 pandemic in 2020 [8]. The authors reported a small increase in mental health symptoms in the beginning of the pandemic (March and April 2020) but that symptoms were back to pre-pandemic levels in May to July 2020. In our study, we observed a strong decline in well-being in March 2020 in both men (–2.4 points) and women (–2.9 points). However, when we compared the mean well-being during the first COVID-19 wave in Denmark (March to May 2020) with the corresponding months before the pandemic (March to May 2018 and 2019), we observed statistically significant lower well-being during the pandemic only in women but not in men. The year with the lowest mean well-being was in both men and women not the first but the second COVID-19 year (2021), a period that was not covered in the review by Robinson et al. [8].
Two studies by Thygesen et al. on mental well-being during the COVID-19 pandemic in Denmark found that mental well-being in both men and women decreased in the autumn months from 2019 to 2020 and then increased from 2020 to 2021 [21,22]. We also observed in both men and women that well-being was lower in October 2020 compared with October 2019. In contrast, well-being was lower in October 2021 compared with October 2020 in women, whereas well-being was the same in October 2020 and October 2021 in men.
Our findings on clear seasonality of well-being are in agreement with the results of a systematic review that reported higher prevalence and symptoms of depression in the winter months compared with the summer months [23]. However, the review also reported a high heterogeneity in the findings [23]. Lower levels of depressive symptoms and higher levels of well-being during the summer months could be explained by sleep patterns, levels of physical activity, and neurological factors [23]. In addition, social factors may play a role, as summer months coincide with the main annual vacation period in many countries, including Denmark.
Strengths and limitations
A major strength of this study is the data availability of frequently collected well-being data from 2018 to 2022 from the same data source, allowing us to analyse monthly changes before, during and after the COVID-19 pandemic. Most other studies on well-being during the pandemic either started data collection immediately after the beginning of the pandemic, or had to rely on limited available data [4 –9]. The use of a validated instrument to measure well-being is also a strength of this study [15]. A recent review concluded that although smartphone apps to monitor well-being are increasingly used, research studies analysing the results are rare [24]. Our study is one of the few so far that took advantage of the rapidly growing amount of app data.
Several limitations have to be considered. First, this is a secondary analysis of data that was not collected for research purpose, but with the purpose of monitoring and intervening on well-being. The data was collected by a private, non-research company and could preferably have contained more information on participant’s background (e.g. sociodemographic characteristics, and history of well-being or mental health problems), as this could strengthen the understanding of the differences and changes in well-being. It is particular regrettable that information on gender was missing for a large proportion of the respondents, which substantially reduced our analytic sample. Additionally, variables related to the COVID-19 pandemic, such as participants having contracted COVID-19 or COVID-19 policies of the workplace (e.g. on allowing or mandating employees to work from home during the pandemic) would have been of interest but were not available. A Danish study further called for more mixed-methods studies, as they found significant variations in people’s individual experiences of the first lockdown [25]. These studies would benefit from information on participants’ perceptions and reactions to the different phases of the pandemic, such as the phases of lockdowns, restrictions and vaccination campaigns. Second, although our study sample represented employees from different industries and workplaces in Denmark, we cannot assume that our study sample was representative for the Danish labour market. The study sample was self-selected both with regard to the workplaces that decided to use the smartphone app, and with regard to the participants deciding to respond to the questionnaire in the smartphone app. Third, the number of participants that responded to the questionnaire varied from month to month, thus not always the same participants responded, which might have caused bias. However, when we restricted the analysis to individuals who had at least five responses in each of the three COVID-19 periods (before, during, after the pandemic), results were similar to the main analysis, indicating that selective non-response had not substantially affected the results.
Conclusions
The results of this study suggest that the COVID-19 pandemic was associated with a slight decrease in employees’ well-being in Denmark. The results were more pronounced in women than in men. Independent of the COVID-19 situation, well-being showed a clear seasonal variation.
Supplemental Material
sj-docx-1-sjp-10.1177_14034948251352659 – Supplemental material for Monthly changes in well-being among Danish employees from 2018 to 2022. Was there an influence of the COVID-19 pandemic?
Supplemental material, sj-docx-1-sjp-10.1177_14034948251352659 for Monthly changes in well-being among Danish employees from 2018 to 2022. Was there an influence of the COVID-19 pandemic? by Sandra Nielsen, Iben L. Karlsen, Marie L. Bak, Johan S. Abildgaard, Reiner Rugulies and Jeppe K. Sørensen in Scandinavian Journal of Public Health
Footnotes
Availability of data and materials
Due to the Danish Data Protection Act, data cannot be made publicly available. Researchers interested to know more about the data are encouraged to contact the corresponding author.
Declaration of conflicting interests
The authors have no conflicts of interest to declare.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Danish Working Environment Research Fund (grant number: 03-2019-09).
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
