Lifestyle Habits and Mental Health in Light of the Two COVID-19 Pandemic Waves in Sweden, 2020

The COVID-19 pandemic has become a public health emergency of international concern, which may have affected lifestyle habits and mental health. Based on national health profile assessments, this study investigated perceived changes of lifestyle habits in response to the COVID-19 pandemic and associations between perceived lifestyle changes and mental health in Swedish working adults. Among 5599 individuals (50% women, 46.3 years), the majority reported no change (sitting 77%, daily physical activity 71%, exercise 69%, diet 87%, alcohol 90%, and smoking 97%) due to the pandemic. Changes were more pronounced during the first wave (April–June) compared to the second (October–December). Women, individuals <60 years, those with a university degree, white-collar workers, and those with unhealthy lifestyle habits at baseline had higher odds of changing lifestyle habits compared to their counterparts. Negative changes in lifestyle habits and more time in a mentally passive state sitting at home were associated with higher odds of mental ill-health (including health anxiety regarding one’s own and relatives’ health, generalized anxiety and depression symptoms, and concerns regarding employment and economy). The results emphasize the need to support healthy lifestyle habits to strengthen the resilience in vulnerable groups of individuals to future viral pandemics and prevent health inequalities in society.


Introduction
The pandemic caused by the coronavirus disease 2019 (COVID-19) has become a global public health emergency. To stop the virus, confinement, social distancing, and even full lockdowns have been implemented. Under such circumstances, there is a risk for radical changes of lifestyle habits such as physical activity (PA), sedentary behavior, smoking, diet, and alcohol consumption, which have all been previously linked to morbidity and pre-mortality [1][2][3][4]. For example, both short and long bouts of regular PA have been shown to improve physical and mental health in both children and adults [1,5,6].
During the first wave of the pandemic, several lifestyle habits seem to have changed, but with mixed reports from different countries. For example, studies from Belgium, France, and Switzerland have reported a general increase in both exercise frequency and sedentary behavior [7,8]. Conversely, in Italy, total PA decreased significantly during the first COVID-19 wave as compared to before, in all age groups and especially in men [9]. Moreover, several studies have shown small changes in dietary habits [10][11][12], while others have reported an increase in unhealthy food intake, overeating, and snacking between meals [10,[13][14][15]. Similarly, studies have indicated that alcohol consumption has not changed during home confinement [13,16], while others have reported increased alcohol consumption [15,17,18]. Smoking has been reported to both have increased [17,19] and decreased [16,20] during the first wave of COVID-19.
Negative changes in lifestyle habits and an increased risk of depression, anxiety, and stress symptoms during the COVID-19 pandemic have been reported [17,21], while a positive association between more time spent in moderate-to-vigorous PA and better mental health has also been found [9,22,23]. However, previous studies have investigated changes in lifestyle during COVID-19 in a relatively short timeframe during the spring and summer of 2020. As the pandemic has continued, we need to examine longer-term effects on lifestyle and mental health, including comparing differences between the different waves of the pandemic. Also, with different governments employing varying countermeasures and social restrictions, it is important to study the effects on lifestyle habits and health experiences in the context of different countries. Sweden is one of the countries that has caught attention worldwide as the government chose to implement mainly recommended restrictions without any full-scale lockdown. Any comparative results from such a strategy on lifestyle habits and mental health is highly relevant for future decision making in similar situations.
The main aim of the present study was therefore to investigate perceived changes in time spent sitting, daily PA, exercise, diet, alcohol, and smoking in response to the COVID-19 pandemic in Swedish working adults, and to study potential differences across age, sex, education, occupational groups, and different waves of the pandemic. An additional aim was to study the odds ratio of perceived mental ill-health in relation to perceived lifestyle changes.

Study Population
Data originated from the Health Profile Assessment (HPA) database (http://www. hpihealth.se (assessed on 3 December 2020)) which contains data from HPAs carried out in health services all around Sweden since the middle of the 1970s. An HPA includes a questionnaire about lifestyle and health experiences, measurements of anthropometrics and blood pressure, estimations of maximal oxygen consumption from a submaximal cycle ergometer test, and a person-centered dialogue with an HPA coach. An HPA is offered to all employees working for a company or an organization connected to occupational or health-related services, and is voluntary and free of charge for the employee. All data are subsequently recorded in the Health Profile Institute database. In the light of the COVID-19 pandemic emerging in March 2020, additional questions regarding working and commuting habits, perceived change in lifestyle habits, and mental health experiences in relation to the COVID-19 pandemic were added to the HPA in the middle/end of April. It was optional for the participants to answer the additional questions. This study included and compared data from three periods: April to June, July to September, and October to December, 2020. From the 21 April 2020 to 2 December, a total of 5599 men and women answered the additional COVID questions, and were included in the present analyses (Table 1). For comparative purposes, an additional 6232 men and women who performed a HPA during the same time period without answering the additional COVID-questions, as well as 20,864 men and women performing a HPA during the same time period in 2019 (Appendix B, Table A1), were included. The study was approved by the ethics board at the Stockholm Ethics Review Board (Dnr 2020-02727). Informed consent was obtained from the participants prior to participation.

Measures
The additional questions in relation to the COVID-19 pandemic are presented in Appendix A. They included questions regarding current working situation, commuting habits, and perceived change in commuting habits, as well as perceived change in sitting time, daily activity, exercise, diet, alcohol intake, and smoking due to the COVID-19 pandemic. Moreover, open questions regarding hours and minutes spent in (a) mentally passive sitting (i.e., tv-viewing, using you phone/iPad/computer to browse the internet) (b) a mentally active sitting (i.e., working, reading, solving cross-words or Sudoku), and (c) socialization (i.e., having a meal, talking with friends or family) were included, as previous studies have indicated different relationships between these different types of sedentary behavior and mental well-being [24]. Finally, questions regarding health anxiety (SHAI) [25], in terms of both one's own health and that of relatives (modified from SHAI); employment [26] and economic [27] concerns; generalized anxiety [28]; and depression [29] were included.
From the HPA, data on BMI and estimated VO 2 max [30] were derived, as well as self-reported baseline daily PA, exercise habits, sedentary behavior, diet, alcohol abuse by AUDIT-C [31], smoking habits, overall stress, perceived health, and perceived symptoms of anxiety and depression (see Appendix B). Highest educational attainment at the time for the HPA was obtained from Statistics Sweden by linking of the participants' personal identity numbers. Occupation was reported by the participants and coded according to the Swedish Standard Classification of Occupation [32], and further dichotomized into blueor white-collar workers.

Statistical Analyses
Chi-square test (percentages) or t-test (mean values) results were used to compare participants with HPA + COVID data and participants with only HPA data during the study period (21 April and 2 December 2020), as well as all participants with HPA data during the study period and participants with HPA data between the same dates in 2019. Differences in working situation, commuting habits, mental health and sitting time between subgroups ( Table 2) were tested using a Chi-square test (percentages), or t-test (mean values). Wave 1 of the COVID-19 pandemic was defined as 21st of April to 30th of June, and wave 2 as 1st of October to 2nd of December, which corresponds to the two clear wave-shapes of hospitalization due to COVID-19 in Sweden according to the Public Health Agency of Sweden [33]. From 1 July to 30 September was defined as months between the two waves, with significantly lower incidence of COVID-19. Multinomial regression modelling was used to calculate odds ratios (ORs) with 95% confidence intervals (CIs) for self-reported perceived change in six different lifestyle habits due to the COVID-19 pandemic in association to sex (women vs. men), age group (18-59 years vs. 60-78 years), educational level (University vs. non-university), occupation group (white collar vs. blue collar), baseline level of each habit, and wave of COVID-19 compared to the summer months (April-June vs. July-Septemberand October-December vs. July-September) ( Table 2). Clustering of negative and positive perceived changes in lifestyle habits, respectively, were defined as negative or positive change in two or more lifestyle habits compared to less. Daily activity was not included in the clustered variable, as change in time spent sitting and daily activity are interchangeably occurring (sitting less leads to more daily activity and vice versa). Moreover, odds ratio (OR) and 95% CI was calculated using logistic regression modelling to study the association of dichotomized mental ill-health variables in relation to sex, age group, educational level, occupation group, wave of COVID-19 pandemic, type of sitting, and perceived change in lifestyle habits. The mental health variables were dichotomized to describe mental ill-health according to the following: "Frequent health anxiety, own" (Question 7A in Appendix A, answer of reply 3 or 4 vs. 1), "Frequent health anxiety, relatives" (Question 7B, reply 3 or 4 vs. 1), "Frequent anxiety symptoms" (Question 10A, reply 3 or 4 vs. 1), "Frequent depression symptoms" (Question 10B, reply 3 or 4 vs. 1), "High concerns employment" (Question 8, reply 4 or 5 vs. 1), and "High concerns economy" (Question 9, reply 4 or 5 vs. 1). Significance level was set as α < 0.05. Data were analyzed using SPSS (version 26), R 4.0.3 (R Core Team, 2020) with the Tidyverse library [34]. Almost half of the participants answering the additional COVID-19 questions reported that their occupation required that they stay at work ( Table 2). The majority reported that they did not change their commuting habits due to the pandemic, whereas 10% reported that they had changed. Of those who changed, the greatest shift was from public transport to car (54%) and to active commuting (26%). Mean reported time spent in mentally active sitting was slightly higher compared to mentally passive sitting, with less time spent sitting while socializing (131, 119, and 82 min/day). Men and blue-collar workers spent more time in mentally passive sitting and less time in mentally active sitting compared to women and white-collar workers. Participants <60 years spent more time in mentally active sitting than those ≥60 years.

Perceived Changes in Lifestyle Habits
Most individuals stated that they had not changed their lifestyle habits due to the COVID-19 pandemic. For time spent sitting, in daily activity, and exercise, respectively, only 5%, 9%, and 10% of the participants reported a positive change, while 18%, 20%, and 20% reported a negative change. Similarly, for diet, smoking, and alcohol intake, 7%, 3%, and 8% perceived a positive change in these lifestyle habits, while 5%, 1%, and 3% perceived a negative change. Figure 1, show changes during the first and second wave. For clustering of perceived change in lifestyle habits, 13% reported a negative change in two or more lifestyle factors, whereas 8% reported a positive change in two or more lifestyle habits.
Comparing the two waves, the odds for lifestyle changes, both negative and positive, were higher during the first wave compared to the second ( Figure 1 and Table 3). For example, the odds of both a perceived positive and negative change in sitting time, daily PA, and exercise were higher during the first wave compared to the second wave. Also, the odds were higher for a perceived negative change in diet and alcohol intake during the first wave compared to the second. Demographic factors were significantly associated with changes in lifestyle habits (Table 3). Women, younger participants (<60 years), participants with a university degree, white-collar workers, and those with more adverse lifestyle habits had higher odds of changing their lifestyle due to COVID-19 pandemic.  Table A2). * HPA question regarding sitting in leisure, coded as Low/moderate = "Almost no time", "25% of time", "50% of time" and High = "75% of time", "All the time". # HPA question regarding diet, coded as Good = "Very good" or "Good" and Poor = "Neither good or bad", "Poor", "Very poor". § Including change in time spent sitting, exercise, diet, alcohol, and smoking.
only 5%, 9%, and 10% of the participants reported a positive change, while 18%, 20%, and 20% reported a negative change. Similarly, for diet, smoking, and alcohol intake, 7%, 3%, and 8% perceived a positive change in these lifestyle habits, while 5%, 1%, and 3% perceived a negative change. Figure 1, show changes during the first and second wave. For clustering of perceived change in lifestyle habits, 13% reported a negative change in two or more lifestyle factors, whereas 8% reported a positive change in two or more lifestyle habits. Comparing the two waves, the odds for lifestyle changes, both negative and positive, were higher during the first wave compared to the second (Figure 1 and Table 3). For example, the odds of both a perceived positive and negative change in sitting time, daily PA, and exercise were higher during the first wave compared to the second wave. Also, the odds were higher for a perceived negative change in diet and alcohol intake during the first wave compared to the second. Demographic factors were significantly associated with changes in lifestyle habits (Table 3). Women, younger participants (<60 years), participants with a university degree, white-collar workers, and those with more adverse lifestyle habits had higher odds of changing their lifestyle due to COVID-19 pandemic.

Mental Health Experiences
The majority of participants had low personal health anxiety, generalized anxiety and depression symptoms, as well as concerns regarding their employment and economy, with a higher proportion experiencing health anxiety for relatives (Table 4). Six percent had clustering of two or more variables of mental ill-health (Table 5). In general, women and participants <60 years had higher odds of mental ill-health compared to men and participants ≥60 years (Table 5), while participants with a university degree and white-collar workers had significantly lower odds of having concerns regarding employment or economy (only university degree participants) compared to their counterparts. As for perceived change in lifestyle habits, the odds of mental ill-health were higher during the first wave compared to the second.

Type of Sitting and Change in Lifestyle Habits in Relation to Mental Ill-Health
A negative perceived change in each lifestyle habit, compared to no or positive change, was associated with higher odds for clustered mental ill-health ( Figure 2). This was seen for all separate mental ill-health variables, except that it was not observed for perceived change in smoking.

Discussion
In Sweden, a country with relatively few social restrictions during the pandemic, we noted small changes in the lifestyle variables overall in a large cohort of workers during the first and second wave of the COVID-19 pandemic in 2020. When changes were present, they were more pronounced during the first wave compared to the second. We also noted that the pandemic impacted some segments of the population more than others; women, individuals <60 years, those with a university degree, white-collar workers, and those with unhealthy lifestyle habits at baseline had higher odds of changing their lifestyle habits compared to their counterparts. Negative changes in lifestyle habits, as well as more time spent in mentally passive sitting at home, were associated with higher odds of mental illhealth.

Changes in Lifestyle Habits in Sweden Compared to Other Countries
The present results with small changes in lifestyle habits are in line with a report in May 2020 from the Swedish National Board of Public Health, where a majority reported no change compared to before the COVID-19 pandemic, (total PA 60%, diet 71%, alcohol 79%, and smoking 77%) [35]. The decrease in daily PA (26% first wave and 20% second wave) and exercise (28% and 21%) in the present sample is noticeably lower than in a large Australian study where approximately 50% reported a decrease in PA [17]. Another study exploring the number of daily steps worldwide during the first wave (March to June 2020), concluded that Swedish citizens maintained their number of daily steps to a higher degree in comparison to other countries. For example, while the maximal decrease of average step counts was 49% in Italy, Sweden had experienced a decrease of only 7% [36]. These differences might be partly explained by differences in lockdown regulations, where Sweden implemented less harsh social restrictions with no lockdown.
The increase in sitting time (26% first wave and 17% second wave) is in line with other studies [7,8,11], and may be due to similar restrictions regarding work situations in these countries. We also investigated the previously proposed difference between mentally passive and mentally active sitting behaviors on mental well-being. For example, More time spent in mentally passive sitting (Tertile 3; ≥120 min/day vs. Tertile 1; 0 to 90 min/day) was associated with higher odds for all variables and clustering of mental ill-health (Table 5). No similar associations were seen for more time spent in mentally active sitting or time in sitting socializing. Note: All analyses mutually adjusted for sex, age group, educational level, occupational group, and wave of COVID-19. Clustered risk and frequent personal health anxiety were additionally adjusted for baseline of perceived health. Time in mentally passive and active sitting, as well as when socializing, were additionally adjusted for baseline level of total sedentary behavior.

Discussion
In Sweden, a country with relatively few social restrictions during the pandemic, we noted small changes in the lifestyle variables overall in a large cohort of workers during the first and second wave of the COVID-19 pandemic in 2020. When changes were present, they were more pronounced during the first wave compared to the second. We also noted that the pandemic impacted some segments of the population more than others; women, individuals <60 years, those with a university degree, white-collar workers, and those with unhealthy lifestyle habits at baseline had higher odds of changing their lifestyle habits compared to their counterparts. Negative changes in lifestyle habits, as well as more time spent in mentally passive sitting at home, were associated with higher odds of mental ill-health.

Changes in Lifestyle Habits in Sweden Compared to Other Countries
The present results with small changes in lifestyle habits are in line with a report in May 2020 from the Swedish National Board of Public Health, where a majority reported no change compared to before the COVID-19 pandemic, (total PA 60%, diet 71%, alcohol 79%, and smoking 77%) [35]. The decrease in daily PA (26% first wave and 20% second wave) and exercise (28% and 21%) in the present sample is noticeably lower than in a large Australian study where approximately 50% reported a decrease in PA [17]. Another study exploring the number of daily steps worldwide during the first wave (March to June 2020), concluded that Swedish citizens maintained their number of daily steps to a higher degree in comparison to other countries. For example, while the maximal decrease of average step counts was 49% in Italy, Sweden had experienced a decrease of only 7% [36]. These differences might be partly explained by differences in lockdown regulations, where Sweden implemented less harsh social restrictions with no lockdown.
The increase in sitting time (26% first wave and 17% second wave) is in line with other studies [7,8,11], and may be due to similar restrictions regarding work situations in these countries. We also investigated the previously proposed difference between mentally passive and mentally active sitting behaviors on mental well-being. For example, Hallgren et al. showed that mentally active sitting was associated with a 29% lower risk for major depressive disorders after a 13-year follow-up in middle-aged men and women, while mentally passive sitting was associated with a 26% higher risk [24]. A study comparing sitting at work (presumably predominantly mentally active sitting) and in leisure time (presumably predominantly mentally passive sitting) showed weak associations of sitting at work and frequent symptoms of anxiety and depression, while more time sitting during leisure time was associated with three to four times higher OR compared to less leisure time sitting [37]. This is comparable to the results in the present study, where more self-reported time in mentally passive sitting (<120 min/day) compared to less (0 to 90 min/day) was associated with~60% to 100% higher risk (OR) for different mental ill-health outcomes. No similar associations were found for mentally active sitting or time sitting while socializing. Although the directions of the observed associations are not clear, possible variations between different types of sitting and mental health outcomes should be considered in future studies examining the impact of pandemic restrictions, as well as in interventions targeting sitting for mental health outcomes.
For changes in sitting time, daily PA, and exercise, it was more evident that individuals with low PA levels at baseline had higher odds of a negatively perceived change in PA due to the pandemic. This is similar to a previous study by Lesser et al., which concluded that mainly inactive individuals had become less physically active during the pandemic [23]. On the contrary, a Canadian study showed that previously active adults decreased their PA, while previously inactive adults did not change their PA due to the pandemic [38]. In contrast to other studies, women in the present study had a 36-38% higher risk of decreasing their daily PA and exercise level compared to men. There were also differences between occupational groups, where white-collar workers had higher odds of increasing daily PA and exercise, while decreasing sedentary time compared to blue-collar workers.
Differences in lifestyle changes due to COVID-19 in relation to occupation groups have not been addressed in previous studies.
The small changes in diet, alcohol, and smoking habits in the present study are in line with studies from other countries [10][11][12][13]15,17,19]. However, perceived changes varied between and within subgroups. For diet, individuals with healthy diets had an approximately 80% lower risk of dietary habits deteriorating compared to individuals with poor habits. Moreover, white-collar workers were more prone to changing their diet in either direction, and had approximately 90% higher probability of worsening as well as improving their diet compared to blue-collar workers. This might be due to blue-collar workers having to be at their workplaces to a higher extent, which probably contributed to fewer possibilities to change their diet behavior compared to white-collar workers, who were able to work more from home. The large differences between blue-and white-collar workers working from home or not in this study are similar to a report from Swedish statistics. The report concluded that while 56% of individuals with a university degree or equivalent reported that they did not work from home at all, the corresponding number among individuals with occupations requiring shorter education was 97% [39].
For alcohol, young individuals in this study had higher odds of both an increase and decrease in alcohol intake, with women having a lower probability of decreasing their alcohol intake. A Canadian study concluded that younger individuals and individuals with higher educational levels had higher risks of increasing their alcohol intake compared to older individuals and those with a lower education level [15]. For smoking, our results indicated that daily smokers had a 53% higher risk of increasing their smoking compared to occasional smokers, which is in line with a small Italian study [19]. However, a study of >20,000 men and women over 16 years of age found that smokers in England were more likely to report trying to quit smoking, and rates of smoking cessation were higher than before the COVID-19 pandemic [40].
There were more pronounced changes for all lifestyle habits during the first compared to the second COVID-19 wave. As recommended restrictions in Sweden were similar during both waves, possible explanations for this might be temporal effects and change saturation. This includes perceived changes in lifestyle habits during the first wave becoming the "new normal" and that people experienced more resistance to change during the second wave [41].
As healthy lifestyle habits are important in preventing noncommunicable diseases [1,42], the need to support individuals in improving or maintaining healthy lifestyle habits during the COVID-19 pandemic in order to prevent health inequalities in society and promote national public health is emphasized.

Changes in Mental Health in Sweden Compared to Other Countries
We found a relatively low prevalence of mental ill-health, with 4% to 6% scoring high on health anxiety regarding their own health, generalized anxiety and depression symptoms, as well as concerns regarding employment and economy. Only health anxiety for relatives was more prevalent (12%). The findings regarding health anxiety for one's own health are similar to the report from the Swedish National Board of Public Health in May 2020, where 5% were very worried about their own health, whereas a higher frequency for health anxiety was noted for relatives (25%) [35]. This is lower in comparison to reports from the UK, where 37% experienced poor mental well-being [43]. We found that women, participants <60 years, and those with a perceived negative change in daily PA, sitting time, exercise, diet, and alcohol consumption, were more vulnerable to mental ill-health. The higher odds of mental ill-health in women and younger age groups has been reported in previous studies [21,43], as has the association between mental health and PA [7][8][9]38], alcohol consumption [17,18], diet [17], and smoking habits [17]. Interestingly in this study, the higher odds for women and younger individuals were also seen for health concerns for their relatives, which has not been reported previously.

Strengths and Limitations
A strength of this study is the reasonably large cohort of women and men of different ages, with a variation in educational level and occupation. The extended period of data collection (from April to December) enabled unique comparative analyses between the two waves of the COVID-19 pandemic in the total study population, as well as in subgroups. Another strength is that the study explored different components of the PA pattern, including both sitting, daily PA, and exercise, as well as different aspects of mental health (clustered mental ill-health, anxiety concern, generalized anxiety, and depression). A limitation is that the study is that it did not have data on baseline depression or anxiety. However, the analyses adjusted for self-reported general health. A limitation is the cross-sectional design, which decreased the ability to draw conclusions of causality and temporal order. Also, we examined self-reported perceived changes in lifestyle, which are not the same as within-person change based on multiple measurements. The study population consisted of individuals who accepted answering the extra covid-19-related questions, which poses a risk of selection bias. Another limitation is that data regarding lifestyle habits and changes in lifestyle habits were based on questionnaires not validated in previous work, thus risking recall bias [44]. However, questionnaires with categorical answer modes, as used in the present study, provide better validity compared to open answers for levels of PA [45].

Conclusions
Our findings suggest only small perceived changes in lifestyle habits, including time sitting, daily PA, exercise, diet, alcohol, and smoking in men and women from the Swedish working population in relation to the first two COVID-19 pandemic waves. Both negative and positive changes were more pronounced during the first wave compared to the second. Women, individuals <60 years, those with a university degree, and white-collar workers had higher odds of changing lifestyle habits compared to their counterparts. Individuals with an unhealthy lifestyle at baseline were more likely to change their lifestyle habits negatively. Thus, changes varied between sociodemographic subgroups, suggesting a clear divergence in how the pandemic waves might have impacted individuals and society. Furthermore, negative changes in lifestyle habits tended to be associated with higher levels of mental ill-health. The perceived negative changes in health-related lifestyles is a considerable public health concern, with possible implications for further increases in health inequality and mental health challenges in the light of the COVID-19 pandemic. To strengthen the resilience of both society and individuals to future viral pandemics, there is a clear need to focus on the promotion of healthy lifestyle habits, especially in socially vulnerable groups and individuals who already have an unhealthy lifestyle.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The datasets generated and/or analyzed during the current study are not publicly available due being property of HPI Health Profile Institute, but are available from the corresponding author or the HPI Health Profile Institute on support@hpihealth.se.  Table A1. Characteristics of participants with COVID-data and/or Health profile assessment-data between 21 April and 2 December in year 2020 (n = 11,844) and participants with Health profile assessment-data during the same time period in year 2019 (n = 20,864).