Physical Activity and Mental Health Declined during the Time of the COVID-19 Pandemic: A Narrative Literature Review

(1) Introduction: Mental health (MH) and physical activity (PA) share a bi-directional relationship, but most studies report MH as the outcome. With diminishing pandemic-related MH, this review examines the impact of diminished MH on PA. (2) Methods: This narrative literature review included 19 empirical studies published since the COVID-19 pandemic. Electronic databases such as MEDLINE, PsycINFO, and CINAHL were searched for English language articles in peer-reviewed journals using equivalent index terms: “anxiety”, “depression”, “stress”, “mental health”, “exercise”, “activity”, “COVID-19”, “coronavirus”, and “2019 pandemic”. The search reviewed 187 articles with double-rater reliability using Covidence. A total of 19 articles met the inclusion criteria. (3) Results: MH themes that impacted PA were depression and/or anxiety (n = 17), one of which identified inadequate coping and excessive pandemic stress (n = 2). In addition, women are more likely to suffer diminished MH and reduced PA throughout the pandemic. (4) Conclusion: Current research suggests that individuals with pre-pandemic MH episodes are correlated with more effective coping skills and fewer adverse effects from COVID-19 than expected. As we emerge from this pandemic, equipping all individuals, especially women, with positive coping strategies may accelerate a seamless return to PA.

The World Health Organization (WHO) and the Center for Disease Control and Prevention (CDC) recommend that early to midlife adults (i.e., aged 18-64 years) participate in 150-300 min of moderate-intensity (or 75 min of high-intensity) aerobic PA and two muscle-strengthening sessions at moderate intensity or greater per week [19]. Individuals who do not meet the minimum PA standards are 20% to 30% more at risk for mortality than individuals who meet the criteria [19].
Beyond the minimum PA requirements, it is recommended that individuals limit their sedentary time [20] (ST). Excessive ST is associated with all-cause mortality, cardiovascular disease, cancer, and type-2 diabetes [19]. The term "Sedentary Lifestyle Syndrome" (SLS), coined by Charansonney to describe long-term sedentary behavior, can be triggered by stress (i.e., a global pandemic) [21]. Prolonged "stay-at-home" orders increased sedentary not able to isolate experience a heightened concern for common symptoms (such as cough and fever) that are now indicative of a COVID-19 infection, further exacerbating fear and anxiety for those actively limiting their own exposure [43]. Research has supported that these psychological factors, and others such as stress and social isolation, correlate with adverse health behaviors [44][45][46][47][48]. Psychological well-being promotes many healthy behaviors, while the lack thereof deters them [48].
As demonstrated, the complexity of stressful events can affect physical, social, mental, and general health [42]. Aptly, the COVID-19 pandemic has provided an opportunity to examine if diminished MH can actually be a precursor for decreased PA and potentially a contradiction for PA prescriptions [42]. Stanton et al. cross-sectionally surveyed 1492 adults in Australia and discovered that COVID-19's social distancing measures negatively impacted PA (48.9%), sleep (40.7%), alcohol consumption (26.6%), and smoking (6.9%) since the onset of the COVID-19 pandemic [40]. They illustrated that these negative health behaviors in PA, sleep, smoking, and alcohol consumption were associated with increased depression, anxiety, and stress [40]. They further ascertained that long-lasting MH effects might arise from fear of infection, confusion, anger, post-traumatic stress symptoms, separation, frustration, boredom, lack of resources and information, financial loss, and stigma [40]. Likewise, other current literature identifies the most vulnerable psychological factors as isolation, fear, emotional stress, anxiety, and depression [38]. According to historically similar pandemics, any of these four acute MH illnesses may previse reduced PA [38]. The current pandemic allows us to analyze the MH and PA relationship in real-time compared to historical analysis from similar pandemics.
In summary, the bleak reality is that many people are neglecting their mental and physical well-being. This observation leads us to consider alternative barriers to PA during these events. The pandemic's contribution to diminished MH promotes prolonged ST, decreased PA, and other adverse health behaviors [22,23,38]. In response to this recurrent co-occurrence and the known bi-directional relationship between PA and MH, we pose this research question: during excessively stressful events (i.e., the COVID-19 pandemic), does diminished MH serve as a barrier to PA? To answer this question, this review explores whether the pandemic's MH impact is a significant barrier to PA. Therefore, the study aims to review existing literature for articles that present diminished MH as an indicator of decreased PA to evaluate this alternate direction. We hypothesize that countless studies measure both MH and PA, but only a few will infer that acutely diminished MH may negatively impact PA.

Design
A systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist [49]. A comprehensive literature search with narrative methods was performed for this study to identify peer-reviewed articles addressing the role of mental health on physical activity during the COVID-19 pandemic.

Eligibility Criteria
Each study meeting the eligibility criteria was included in the systematic review: articles published in peer-reviewed journals in English, published after 16 March 2020, included adult subjects 18 years and older, included both MH and PA factors, and used various research designs and methodologies. The exclusion criteria were articles published before 16 March 2020, articles without any connection to PA, articles that targeted pregnant women or actively COVID-positive patients as participants, and articles using PA as the predictor and MH as the outcome.

Study Selection
The study selection consisted of three steps. First, each author independently screened all titles and abstracts (n = 187) of relevance for this systematic review [47]. In addition, the reference lists of all the included studies were scanned for relevant papers. Removal of three duplicate publications resulted in selecting 184 published papers. Second, the abstracts of all relevant articles were screened independently for eligibility by each author. Inter-rater reliability was 79%. The 63 articles deemed irrelevant were excluded. Third, the full papers of the included publications were obtained and screened for inclusion and exclusion criteria. The full text of 121 articles was reviewed. Furthermore, 102 articles were excluded, and 19 studies were selected to be analyzed. Figure 1 outlines the search process of the literature [48].

Eligibility Criteria
Each study meeting the eligibility criteria was included in the systematic review: articles published in peer-reviewed journals in English, published after 16 March 2020, included adult subjects 18 years and older, included both MH and PA factors, and used various research designs and methodologies. The exclusion criteria were articles published before 16 March 2020, articles without any connection to PA, articles that targeted pregnant women or actively COVID-positive patients as participants, and articles using PA as the predictor and MH as the outcome.

Study Selection
The study selection consisted of three steps. First, each author independently screened all titles and abstracts (n = 187) of relevance for this systematic review [47]. In addition, the reference lists of all the included studies were scanned for relevant papers. Removal of three duplicate publications resulted in selecting 184 published papers. Second, the abstracts of all relevant articles were screened independently for eligibility by each author. Inter-rater reliability was 79%. The 63 articles deemed irrelevant were excluded. Third, the full papers of the included publications were obtained and screened for inclusion and exclusion criteria. The full text of 121 articles was reviewed. Furthermore, 102 articles were excluded, and 19 studies were selected to be analyzed. Figure 1 outlines the search process of the literature [48].

Data Extraction
Each author independently selected data from the 19 studies using the data extraction tool. The authors met to compare the data and resolve inconsistencies by referring to the full-text article and thorough discussion. The following data were extracted: citation, country of study, the aim of the study, population demographics (e.g., age, gender), study designs, data source(s) (survey), and key observation(s) of the study (See Appendix A).

Quality Assessment
The assessment of each article's quality was determined using the Joanna Briggs Institute (JBI) critical appraisal tools (See Appendix B) [50]. The criteria used JBI guidelines

Data Extraction
Each author independently selected data from the 19 studies using the data extraction tool. The authors met to compare the data and resolve inconsistencies by referring to the full-text article and thorough discussion. The following data were extracted: citation, country of study, the aim of the study, population demographics (e.g., age, gender), study designs, data source(s) (survey), and key observation(s) of the study (See Appendix A).

Quality Assessment
The assessment of each article's quality was determined using the Joanna Briggs Institute (JBI) critical appraisal tools (See Appendix B) [50]. The criteria used JBI guidelines to evaluate whether each study is good quality and has minimal risk of bias. The study used a cross-sectional appraisal including eight criteria [51]. There are four answer choices in the JBI, namely "yes", "no", "unclear", and "not applicable". Conclusions were based on the results of the review. The more "yes" answers in the JBI critical appraisal column, the better and more valid the publication will be. The researchers independently evaluated the quality of each study, and disagreements were resolved by discussion within the review team. Identifying confounding factors and strategies to deal with confounding factors were not mentioned. This is of little concern for the current study. None of the studies were excluded based on their quality appraisal (see Appendix B).

Data Analysis
Due to the variability in the MH and PA measures, results could not be combined by meta-analysis. A narrative synthesis of the study was conducted. Tables and narrative summaries are used to present the study participant characteristics and the findings of the studies.

Study Characteristics
All the studies were published in or after 2020. Out of the 19 studies, three of the studies were conducted in the United States [52][53][54], six of the studies were conducted in Europe [55][56][57][58][59][60], two of the studies were conducted in Asia [60,61], two of the studies were conducted in Australia [40,62], one of the studies were conducted in the Middle East [63], three studies were conducted in South America [64][65][66], one of the studies were conducted in Canada [67], and one of the studies was conducted in multinational region [68].

Mental Health Assessment
Two studies used the 21-item Depression, Anxiety and Stress Scale (DASS-21) [40,62]. One study used the 16-item Quick Inventory Depressive Symptomatology tool [52]. One study used the Illness Attitude Scale [69]. One study used the Diagnostic and Statistical Manual of Mental Disorders, Version Four criteria (DSM-IV) [53]. Three studies used the Generalized Anxiety Disorder (GAD-7) [55,65,67]. One study used the Patient-Reported Outcomes Measurement Information System (PROMIS) [68]. One study used the Beck Anxiety Inventory tool [63] and one study used the Beck Depression Inventory tool [62]. Two studies used the Center for Epidemiological Studies-Depression Scale (CES-D) [56,61]. One study used Zung's Self-reported Anxiety Scale (SAS) [57]. One study used the 14-item Hospital Anxiety Depression Scale [64]. Two studies used the Perceived Stress Scale (PSS) [54,60]. One study used the Yesavage Geriatric Depression Scale [43]. One study used a Mental Stress Indicator Score [59]. One study used a previous depression diagnosis [66].

MH and PA Interaction
Seventeen articles suggested that increased anxiety and depression reduced PA levels [40,43,[52][53][54][55][56][60][61][62][63][64][65][66][67][68][69]. Stanton et al. found that community-dwelling adults reported increased anxiety, depression, and stress symptoms and decreased PA levels [40]. Coughenour et al. and Moriarty et al. found that college students reported higher levels of depression engaged in fewer minutes of PA [53,54]. Marashi et al. also reported that those whose MH symptoms increased significantly also reduced their PA during theCOVID-19 pandemic [67]. Those with more anxiety and depression reported a more significant decrease in PA level [67]. Stanton et al. found that females reported higher psychological distress scores than males [40]. Two studies reported that anxiety and PA were determined based on the participants' behavior before the pandemic. Kaygisi reported that female participants who were less anxious during the pandemic were more likely to engage in sports before the pandemic [63]. However, Choi and Bum found that those who were anxious about being infected with COVID-19 were less likely to participate in sports activity during the COVID-19 pandemic [69].
Patients receiving medical care reported increased anxiety and depression reported decreased PA. Almandoz et al. reported bariatric patients who reported increased anxiety and depression reduced their PA levels by 47.9% [52]. Van Der Heide et al. reported that individuals diagnosed with Parkinson's disease had an increase in poor MH outcomes and a decrease in PA due to worsened Parkinson's disease symptoms [60].

Discussion
Quarantine, isolation, and other social distancing measures are critical to reducing exposure to this novel virus. Unfortunately, the COVID-19 pandemic has had detrimental health consequences in both physical and MH. Some of the most concerning consequences discussed here include significant increases in emotional stress, depression, and/or anxiety paired with a continuous decline of PA. This mirrors the outcomes of historical pandemics that seem to repeat with extraordinary precision. This study provides evidence that acute MH illnesses such as anxiety, depression, and emotional stress can, in fact, hinder PA participation. Because diminished MH may impede PA, this study suggests that MH care should take precedence over PA reinstatement as we recover from the pandemic. These findings are unprecedented and greatly contribute to the existing literature by uniquely invoking the reverse directionality of the bi-directional relationship between MH and PA.

Research Implications
To our knowledge, there are no studies that identify the degree of or duration at which MH may impede PA. Longitudinal studies are needed to better understand the directionality of the bi-directional relationship between MH and PA. Until this is evaluated, we cannot accurately discern the significance of this impedance. Additionally, the degree of diminished MH that interferes with PA is important to accurately define and to guide PA recommendations. Currently, there is no existing acute MH assessment to indicate if physical abilities are impacted. Furthermore, evaluating populations with existing MH diagnoses would contribute toward even greater validity of this barrier. Lastly, it may be interesting to introduce MH practices such as meditation, relaxation, deep social connection, coping skills, and self-reflection into PA regimes to boost physical skills beyond current limits. In any manner, MH practices should be incorporated into all levels of healthy lifestyle interventions.

Study Limitations
Because this is a novel event with global attention, the demographics in these studies were reported in various ways, making generalizability difficult. Similarly, it should be noted that many of the included studies used predominantly female samples. Again, this makes generalizability less reliable. Because of the vastness in geographic locations and population cultural variations, MH and PA measurement tools varied. While most of the measurement tools were previously validated, they differed between each study. Due to the swiftness of related publications, it is possible that articles could have been incidentally omitted from this narrative review, especially those published in a language other than English and those that failed to explicitly compare the impact of diminished MH on PA. Lastly, and most notably, some cross-sectional studies assumed a potentially misleading causal relationship between MH and PA.

Conclusions
Obviously, more desirable pandemic-era PA behaviors would lessen post-pandemic physical health consequences. However, acute MH episodes may be a contraindication for concurrent PA participation in the current pandemic. Assessing one's current stress levels may be a helpful tool for evaluating one's PA readiness. Incorporating MH assessments into existing PA pre-screening protocols would help identify if an individual should prioritize MH practices before initiating a PA regimen. For better PA outcomes in future pandemics, preventative MH and coping practices appear to be necessary. To learn from previous shortcomings, it would be wise to exit this pandemic with MH strategies in mind for immediate and long-term utilization for all adults experiencing diminished MH, especially those most affected: women and older adults. Furthermore, providing standardized MH care as a long-term health prevention strategy may alleviate many other adverse health behaviors during future pandemics, including excessive substance use, inadequate sleeping patterns, poor nutrition, and sedentary and screen times. Even more grand, achieving optimal MH may induce the intended objectives of Healthy People 2030.

Conflicts of Interest:
The authors declare no conflict of interest. College students reported higher depression scores (p < 0.01) and reduced PA (p = 0.01) after "stay-at-home" orders were issued.

Appendix A. Characteristics of Included Studies
There was a small but significant (p = 0.04) correlation between changes in total minutes of PA and depression scores. Seniors (p = 0.05) and Hispanic students (p = 0.03) were less likely to report worsening depression scores than first year and non-Hispanic white students. Asian students were significantly more likely to report decreased PA than non-Hispanic white students. This study suggests that COVID-19 and its consequences may contribute to reduced PA and greater depressive symptoms in college students and that sub-groups of college students have been affected differently.

IPAQ-Short Form-Turkish version
The results showed that the post-menopausal women who exercised before the pandemic had higher PA levels during the pandemic. Post-menopausal women with more grandchildren engaged in less PA and reported higher anxiety levels. The levels of anxiety and PA were negatively associated with each other. Zung's Self-Reported Anxiety Scale (SAS), Spanish version PA levels were self-reported before and during the pandemic weekly.
The study found PA as a coping resource. The study found that the prevalence of a sedentary lifestyle was four times higher during the pandemic than in the pre-pandemic stage (21% pre-pandemic, 87% during the pandemic). The negative change in PA was due to the restrictions to prevent the spread of COVID-19. The factors for decreased PA were the lack of home-based programming.

Marashi, 2021 [67] Canada
Crosssectional online survey The study aimed to examine the relationship between PA and sedentary behavior and how it impacts perceived barriers and motivators to PA during the COVID-19 pandemic. Participants reported higher psychological stress and moderate levels of anxiety and depression during the pandemic. Participants with the highest reported mental health deterioration were the least likely to be active. Most participants were unmotivated to exercise because they were anxious. The findings highlight the paradox between mental health and PA. People who wanted to be active to improve their mental health but found it challenging to be active due to their poor mental health. Likewise, participants who were more depressed were less motivated to engage in PA.   Pre-COVID, individuals diagnosed with depression were more likely to have a higher prevalence rate of physical inactivity. During COVID, the physical inactivity incidence rate did not differ among people diagnosed with depression compared with the general population.