Physical activity on anxiety symptoms during the COVID-19 pandemic: a systematic review

Abstract This systematic review aimed to investigate the effect of physical activity on anxiety symptoms during the COVID-19 pandemic. Three databases (PubMed; Scopus and Web of Science) were searched in the period from 2020 to 2022 the following base terms were used: “Physical Activity”, Exercise, Anxiety and COVID-19. Peer-reviewed, primary studies published in English, Portuguese and Spanish using valid and reliable measures were included. Eighteen studies met the eligibility criteria, of which 17 were cross-sectional, 2 were cohort studies. The number of participants ranged from 43 to 2,301, aged between 18 and 65 years or older. A decrease in PA practice or an increase in sitting time were associated with higher levels of anxiety symptoms. Additionally, participants who did not meet PA recommendations were more likely to experience moderate to severe anxiety symptoms. The results showed that physical activity is associated with the alleviation of anxiety symptoms during confinement in the COVID-19 pandemic.


INTRODUCTION
The COVID-19 pandemic has been the target of research regarding mental health, due to the impact it has on the global health of the population 1,2 .The restrictive measures of isolation and social interaction, used to reduce the spread of the virus 3,4 were important factors in modifying the habits of the entire world population, resulting in weight gain 5 , restriction of leisure activities and reduction of physical activity (PA) 6,7 promoting deleterious effects on mental health 8,9 .
Recent evidence suggests that people who are kept in isolation and quarantine experience significant levels of anxiety, anger, confusion, and stress [10][11][12][13][14] , as it was also seen in the influenza epidemic in 2008 15 , and H1N1 in 2009 16 .The first systematic review and meta-analysis on the prevalence of stress, anxiety, and depression in the general population resulting from the COVID-19 pandemic showed that the prevalence of stress, anxiety, and depression are 29.6, 31.9, and 33.7%, respectively 9 , revealing that a significant portion of people is becoming mentally ill, not directly from the SARS-CoV-2 virus infection, but due to the nuisances of the pandemic in general.In a recent systematic review, it was found that the prevalence of anxiety and depression in 2020 were 32.6% and 27.6%, respectively 17 .
One of the problems related to mental disorders during the COVID-19 pandemic was physical inactivity.Puccinelli et al. 18 showed that 30% and 23.3% of the subjects evaluated in their study (n=1853) had moderate/severe symptoms of depression and anxiety, respectively, associated with a reduction in PA.The health benefits of regular exercise are well established in the scientific literature 17,[19][20][21][22][23] , which can be an important intervention in combating the negative effects of the COVID-19 pandemic on mental health 22,24,25 .
The recommendations for regular PA during this pandemic period follow the standards established by the World Health Organization (WHO) and the American College of Sports Medicine (ACSM) [26][27][28] , as there are no studies that can guide the real need for changes in current recommendations.Thus, it is not known whether social distancing, lockdown, and other physical restrictions due to the COVID-19 pandemic allowed studies on exercise to be performed following the traditional parameters recommended by international institutions.The real effects of these training programs on anxiety disorder during the pandemic period are still unclear since there is still no compilation of studies performed to date.
Therefore, the present systematic review aimed to evaluate the effects of physical activity specifically on anxiety symptoms in studies conducted during the COVID-19 pandemic.Further, the training protocols were analyzed and compared to traditional recommendations provided by the WHO and ACSM.

Study design
This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Checklist 29 .Studies were included according to the Participants, Intervention, Comparison, Outcomes, and Study design (PICOS) criteria: 1): men and women over 18 years old, physically active or sedentary, with no history of joint or musculoskeletal damage, and without In July 2022, the searches were updated to include studies that were published between January and July.We selected ten studies for full reading.For the analysis, only one study met the eligibility criteria 30 .As five other studies [31][32][33][34][35] did not meet eligibility criteria, we could only use them for our discussion.

Study selection criteria
Studies were included in this review after adhering to the criteria: (i) studies performed during the COVID-19 pandemic period, published in English, Portuguese or Spanish, in peer-reviewed journals; (ii) intervention studies with exercise or observational studies (screening of physical activity or exercise); (iii) sample composed of people with an assessment of "anxiety" through validated instruments.

Data screening
Each study was selected based on "Titles and Abstract."Then, the full text was evaluated to confirm inclusion or exclusion.Title, abstract and full-text screening was performed by one reviewer and verified by three other independent

Data extraction
The following information was extracted: author, year, type of study, objective, sample, instruments used, intervention, and results.Three independent reviewers extracted data.A fourth reviewer gathered the extracted information and organized the results.

Risk of bias
The Joanna Briggs Institute manual ( JBI) 36 was used to evaluate the observational/cross-sectional studies.It contains eight analytical criteria for a cross-sectional study construction.As for the evaluation of cohort studies, the Joanna Briggs Institute for cohort studies was used, which contains eleven analytical criteria for this type of study.Responses consist of "yes", "no", "confused", and "not applicable" for both questionnaires.The aim of this evaluation is to estimate the methodological quality of a study and to determine the extent to which the possibility of bias has been addressed in its design, conduction, and analysis.

Study selection
The search resulted in 442 possible eligible studies (Figure 1).After removing 231 duplicates, 201 studies were selected for the title and abstract reading phase, resulting in forty-two studies for detailed analysis (Chart 1).A total of eighteen studies were eligible, which included seventeen cross-sectional studies and two cohort studies (Chart 2).

GAD-2
The association between the different domains of PA (ie, work domain, transport domain, domestic domain, and garden and leisure) did not find significant differences related to anxiety.Sitting time and PA levels were also verified, no significant difference was found for anxiety levels.
Antunes et al. 42  GAD-7 During confinement, less PA was directly associated with less sleep regularity and more severe insomnia symptoms.The effects of exercise on anxiety symptoms did not exist before the blockade.In addition, there was a significant indirect effect between PA and anxiety symptoms during the block, so that less PA was associated with more severe insomnia symptoms and, therefore, greater anxiety symptoms.
Haider et al. 45  .47.Anxiety levels were similar in physically active and inactive students.Physically active and inactive male college students did not differ in their anxiety levels.The effect of the interaction between gender and PA was also not statistically significant.

Characteristics of the studies
The included studies were published between 2020 and 2022.A total of sixteen studies collected data between February and June 2020.In the study by Duncan et al., data were collected from April to October, while in the study by Coakley et al., data were collected from October to November 2020, with both studies collecting data in the second half of the year.In only one study, the date of data collection was not specified 39,52 .
The number of participants ranged from 43 to 2,301, aged between 18 and 65 years or older.Several instruments were used to analyze the practice of PA: i) International Physical Activity Questionnaire (IPAQ), ii) Physical Activity and Sedentary Behavior Questionnaire (PASB-Q), iii) Telephone risk factor surveillance survey (Surveillance of Risk and Protective Factors for Chronic Diseases by Telephone Survey -VIGITEL) -physical activity section containing 5 items, iv) Physical Activity Vital Sign (PAVS), v) Godin Leisure Questionnaire (GLQ).Three studies did not use a questionnaire, but a model of questions about physical activity during the period of social isolation.

Risk of bias
Most observational/cross-sectional studies met the main methodological quality criteria.Some studies have demonstrated confusion in inclusion and exclusion criteria 37,38,40,47,51 .The cohort studies met all JBI criteria 52,53 .More details in Supplementary file 1.No studies were excluded after individual assessments, more information can be seen in the Supplementary file 1.

Observational studies
Marashi et al. 38 and Coakley et al. 39 in their respective studies, found that people who reported decreased PA practice or increased sitting time experienced higher levels of anxiety symptoms 38,39 .PA has an association on reducing state and trait anxiety 42,43 .Indeed, it was found that vigorous PA 48 had a negative and significant correlation with trait anxiety 43 , whereas low-intensity PA was not associated to reducing anxiety symptoms.
In addition, some studies have shown that the duration and frequency of PA practice are important for reducing anxiety levels 27,30,46,50 .The minimum duration of activity and weekly frequency with some effectiveness were 30 min and 3 times a week, respectively 27,45 .Schuch et al. 46 found that practicing 15 minutes or more of vigorous PA/day provides a lower chance of presenting anxiety symptoms when compared to those who perform 30 minutes or more of moderate PA/day 46 .Reduced PA practice may have indirect effects on anxiety, as demonstrated in the study by Lewis et al, in which the decrease in PA increased the symptoms of insomnia and, consequently, anxiety 44 .Haider et al. 45  individuals who spend more than 60 min/day outdoors and who managed to maintain a moderate to vigorous level of PA were more likely to have high mental well-being, without depression and anxiety symptoms 45 .

Recommendations for Physical Activity
When the practice of PA was analyzed according to the WHO and ACSM recommendations as parameters, we found that most of the populations studied did not meet the minimum recommendations for health benefits.Still, for those who were active, with the beginning of the pandemic and social restrictions, there was a decrease in PA levels.Al-Ajlouni et al. 30 found that participants who did not meet the WHO guidelines had a higher prevalence of moderate to severe anxiety symptoms.
Among the studies analyzed, only the study by Duncan et al, did not meet any of the minimum recommendations of physical Activity 52 .There were variations in compliance with the recommendations in the studies before and during social isolation.The studies by Haider et al. 45 , Meira et al. 27 and Wood et al. 37 obtained the highest percentages of the sample meeting the recommendations before social isolation, being 89%, 86.9% and 82.58%, respectively.These percentages were reduced to 76.3%, 65.6% and 51.6% during social isolation, respectively.Even with the efforts of the WHO and ACSM with illustrative publications for people to remain physically active, many individuals decreased the level of physical activity over the period of social isolation and other restrictions, as can be seen in more detail in the Chart 3. Al-Ajlouni et al. 30 31.5% of participants met the PA guidelines Duncan et al. 52 Did not meet the recommendations.
Rees-Punia et al. 53 38.21% of the participants met the recommendations.
Wood et al. 37 82.58% of the participants met the recommendations.However, only 51.6% met the recommendations during the lockdown.
Marashi et al. 38 The participants met the recommendations.However, the data is unclear.
Coakley et al. 39 51% of university students met the recommendations.
Kua et al. 40 56.5% of participants met the recommendations before the lockdown.After the lockdown, only 39.8% met the recommendations.
Pears et al. 41 56% of the participants met the recommendations.
Antunes et al. 42 68.1% of the participants met the recommendations.
Lewis et al. 44 Participants met the recommendations before the lockdown.Lesser & Nienhuis 49 36,6% of the participants met the recommendations.
Jacob et al. 503,41% of the participants met the recommendations.

DISCUSSION
The present systematic review investigated the effect of exercise and PA on anxiety symptoms during the COVID-19 pandemic.Main findings revealed that exercise and PA are important factors for reducing anxiety levels, and that greater intensity, duration, and frequency were related to greater effects.Vigorous-intensity activities 46 or longer duration 50 were more effective for this result than moderate-intensity activities.
Disarrangements resulting from the COVID-19 pandemic have had a negative impact on individuals mental health, such as, increasing the incidence of symptoms of anxiety and depression 12,55 .Because of locomotion restrictions, an increase in sitting time and sedentary behavior 56 were major reasons for the reduction of exercise and physical Activity 57 in a large portion of the population, even in those who were already physically active.Time spent using screens and general sedentary behavior 58,59 are risk factors associated with increased anxiety 53 .
Physical activity is considered a positive mediator of anxiety 60,61 and even an individual with generalized anxiety disorder can sustain high levels of PA, for example by activating compensatory mechanisms 62 and regulatory systems of the sympathetic and parasympathetic systems, regulating the stress response through the hypothalamic-pituitary-adrenal (HPA) axis 63 , in addition stimulating the brain-derived neurotrophic factor (BDNF), which can balance the neurobiological responses to stress 64 in cortical and limbic regions of the brain, such as in the hippocampus, where in the BDNF Val66Met polymorphism (which decreased BDNF expression and signaling) has been associated with reduced hippocampal volume and executive function, increased susceptibility to anxiety, and depressive behaviors 65 .In addition, the positive effect of physical activity includes increased blood flow to the hippocampus and an increase in size, as well as decreased neuro-inflammation 66 , an increase in neuroprotection, and overpowering the negative effects of anxiety 67,68   The lowest engagement in PA was seen in the study by Lesser and Nienhuis, with 36.6% 49 of participants meeting the recommendations proposed by the WHO or ACSM, while the highest engagement was observed in the study by Haider et al. 45 , with 76.3%, both during the lockdown 45 .Such a discrepancy shows how much people diverged in the ways to keep themselves active according to the location.The non-commitment of all participants to the minimum PA recommendations may be one of the factors related to some divergent results, in studies that did not show differences in the anxiety profile in individuals who performed PA 19,51,69,70 .Overall, participants who meet recommended PA guidelines are less likely to experience greater anxiety than those who do not 53,71 .
Evidence from a systematic review conducted by Wolf et al. 72 suggests that people who regularly perform PA with greater volume and frequency and maintain a stable PA routine have fewer symptoms of depression and anxiety.For example, those who reported the most total time spent in moderate to vigorous PA were 12 to 32% less likely to experience depressive symptoms and 15 to 34% less likely to experience anxiety 72 .
The study conducted by Kim et al. 73 found that compared to the sedentary group (0-600 METs-min/week), individuals who achieved 600-6000 METsmin/week had a significantly lower risk of anxiety.After stratifying the data by sex, optimal PA ranges were 600 to 9,000 METs-min/week for men, but 1,200 to 3,000 METs-min/week for women.Furthermore, engagement in more than 6,000 METs-min/week was found to be not associated with higher risk of anxiety symptoms, suggesting a specific PA dose-response regardless of the type of activity practiced.In this sense, a cross-sectional study, which reported that people who were not practicing PA during COVID-19 confinement presented a higher level of stress, anxiety, and depression, highlighting the importance of home physical training to reduce the impacts of physical inactivity due to confinement measures, due to the pandemic situation on mental health 74 .
As identified by the self-reports of participants in the studies addressed, the specifications of PA in terms of frequency and duration were crucial to reduce anxiety during the COVID-19 pandemic.In fact, other studies and institutions advocate the practice of frequent and prolonged PA.Both WHO and ACSM have recommended at least 150 minutes per week of moderate to vigorous PA and, during the COVID-19 pandemic, have suggested people to maintain their usual levels of PA.As observed in Chart 3, there was a reduction in the practice of PA due to social restrictions.Evidence supports the importance of regular PA practice in the prevention and treatment of chronic diseases, cognitive decline, improvement of the immune system and increased longevity 75 .The COVID-19 pandemic reinforced the relevance of PA for physical and mental health, mainly because of PA reduction and increased sedentary behavior, especially due to social isolation protocols 76 .Physical inactivity and sedentary behavior were related to the most severe outcomes of COVID-19 77 , especially for the elderly who presented themselves as a risk group 78 .The low-cost and low-risk nature of PA shows that it should be implemented in public policies during pandemics.
According to a study conducted with Australians who underwent three data collection periods (April, July, and August 2020), people participating in at least 150 minutes of moderate or vigorous physical activity per week had significantly higher resilience scores than those not participating in PA.As resilience is a response to persistent stressors, like the COVID-19 pandemic 79 , it can vary based on what individuals are capable of coping with 80 .Interventions that aim to minimize psychological distress can be designed with an understanding of resilience and its changes over time 81,82 .
Mental health and well-being are associated with the behaviors we adopt throughout life, so PA has been a positive way to maintain mental health during periods of social isolation 33 .A study comparing the effects of isolation in New Zealand and the United Kingdom found that the as more PA practiced and the less sedentary behavior, the better the perception of well-being and the lower the DASS-9 score 32,33 .
It is worth mentioning that the prevalence of anxiety and depression may vary in these surveys, due to the methodological heterogeneity between the studies.Wu et al. 83 found that the prevalence of anxiety ranged from 6.3% to 87.5% in 53 studies with seven populations (I 2 = 99.6%), and the point prevalence of depression ranged from 3.1% to 87.3% among 48 studies and seven populations (I 2 = 99.6%) 83.The I 2 values presented in both studies, above 99%, show the degree of heterogeneity that exists.
These are the main limitations of this systematic review which, in addition to the disparity in the data of the analyzed studies, could not provide sufficient quantitative information to conduct a meta-analysis.Finally, the absence of experimental studies limits us to cautious statements about the level of PA and anxiety.If, on the one hand, vigorous and moderate intensity as well as continuous frequency of at least three times weekly can mitigate anxiety, on the other hand, light exercise may not have the same effect.Nevertheless, the observational nature of the analyzed studies, which were the majority in this systematic review, makes it impossible to establish cause-and-effect relationship.

CONCLUSION
Despite the positive consequence of physical activity for the reduction of anxiety symptoms during the COVID-19 pandemic, the results of the present systematic review are still inconclusive.However, it is worth mentioning the importance of intensity, session duration and weekly frequency, which, in some studies, showed better effects in more active individuals, suggesting a proportional dose-response relationship of physical activity on anxiety.The higher the intensity and frequency of PA, the lower the risk of developing anxiety symptoms.Taking part in vigorous physical activity for at least 15 minutes per day may relieve anxiety symptoms.

COMPLIANCE WITH ETHICAL STANDARDS Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.This study was funded by the authors.

Ethical approval
This research is in accordance with the standards set by the Declaration of Helsinki

Figure 1 .
Figure 1.Flowchart of search strategy results.

Chart 3 .
Changes in Physical Activity during the COVID-19 pandemic STUDIES Physical Activity Recommendations -Adults and Older Adults WHO ACSM David et al.Physical activity on anxiety symptoms LFFF; EHES; EOB).Any discrepancies in the inclusion or exclusion of scientific articles were resolved in consultation with a fifth reviewer (RSMJ).The retrieved articles were cataloged, and the duplicates were removed. 30 demonstrated that 45ider et al.4589.0% of participants met the recommendations before and 76.3% during the lockdown.Note.World Health Organization (WHO) -Adults and older adults should perform 150 to 300 minutes/week of moderate and 75 to 150 minutes/week of vigorous-intensity PA.Strength training for at least two days for large muscle groups.For older people, training in multicomponent activities at least 3 days a week is recommended26.American College of Sports Medicine (ACSM) -Adults should perform 150 minutes/week of moderate and 60 minutes/week of vigorous-intensity aerobic PA.For strength and endurance improvement, train large muscle groups at least two days a week.Older adults should perform 150 minutes/week of moderate-intensity and 75 minutes/week of vigorous-intensity PA, or a combination of both.Perform strength training at least two days a week and other types of training such as balance 54 .
Adults and older adults should perform 150 to 300 minutes/week of moderate and 75 to 150 minutes/week of vigorous-intensity PA.Strength training for at least two days for large muscle groups.For older people, training in multicomponent activities at least 3 days a week is recommended26.American College of Sports Medicine (ACSM) -Adults should perform 150 minutes/week of moderate and 60 minutes/week of vigorous-intensity aerobic PA.For strength and endurance improvement, train large muscle groups at least two days a week.Older adults should perform 150 minutes/week of moderate-intensity and 75 minutes/week of vigorous-intensity PA, or a combination of both.Perform strength training at least two days a week and other types of training such as balance 54 .