Impact of risk for severe COVID-19 illness on physical activity during the pandemic

Abstract

Background

Precautions to mitigate spread of COVID-19 such as the closing of exercise facilities impacted physical activity behaviors. Varied risks for severe COVID-19 may have influenced participation in regular physical activity to maintain precautions.

Objective

Describe differences in the amount and intensity of physical activity between adults at high versus low risk for severe COVID-19 illness during the pandemic. We hypothesized that over 13 months, 1) high-risk adults would have greater odds of inactivity than low-risk adults, and 2) when active, high-risk adults would have lower metabolic equivalent of task minutes (MET-min) than low-risk adults.

Methods

This longitudinal observational cohort study surveyed U.S. adults’ demographics, health history, and physical activity beginning March 2020 using REDCap. Using self-report, health history was assessed with a modified Charlson Comorbidity Index and physical activity with the International Physical Activity Questionnaire. Repeated physical activity measurements were conducted in June, July, October, and December of 2020, and in April of 2021. Two models, a logistic model evaluating physical inactivity (hypothesis 1) and a gamma model evaluating total MET-min for physically active individuals (hypothesis 2), were used. Models were controlled for age, gender, and race.

Results

The final sample consisted of 640 participants (mean age 42.7 ± 15.7, 78% women, 90% white), with n = 175 categorized as high-risk and n = 465 as low-risk. The odds of inactivity for the high-risk adults were 2.8 to 4.1 times as high than for low-risk adults at baseline and 13 months. Active high-risk adults had lower MET-min levels than low-risk adults in March (28%, p = 0.001), June (29%, p = 0.002), and July of 2020 (30%, p = 0.005) only.

Conclusions

Adults at high risk of severe COVID-19 illness were disproportionately more likely to be physically inactive and exhibit lower MET-min levels than adults at low risk during the early months of the COVID-19 pandemic.

Introduction

The impact of the COVID-19 pandemic extends beyond direct infections to include health behaviors of people globally. Beginning in April 2020 policy makers in the United States began implementing recommendations; altering everyday life through lockdowns, physical distancing, mandated mask-wearing, continual testing, and vaccination policies.¹ Guidelines and policies governing COVID-19-related behaviors were not always clear or implemented universally, leaving individuals to interpret a deluge of conflicting information. The Centers for Disease Control and Prevention (CDC) encouraged everyone regardless of health status to adopt COVID-19 precautions (e.g., physical distancing, vaccination, wearing a mask, and improving ventilation).¹ Because of the severity of COVID-19 and the lack of proven treatments, social and structural changes rapidly occurred to reduce people's risk of being exposed. Gyms, restaurants, and other areas where people gathered were closed, and people were encouraged to practice physical distancing, a new form of self-management for many. With reduced access to public gyms or exercise groups, the amount and intensity of people's physical activity decreased, and their sedentary lifestyles increased.3, 4, 5 During the pandemic, 50% of U.S. adults reported practicing more sedentary activities, and spending less time being physically active than prior to the pandemic.^{6 , 7}

Physical activity is associated with decreased risk of type 2 diabetes, depression, various cancers, stroke, cardiovascular disease including heart failure, and overall mortality.⁸ Physically active individuals also tend to sleep better, feel better, and function better overall.⁸ Little or no moderate-to-vigorous physical activity combined with high levels of sedentary behavior increases individuals’ risk of all-cause mortality.⁹ There are population-based differences in physical activity that are important to consider. Non-Hispanic White individuals have reported meeting physical activity recommendations at higher rates than non-Hispanic Black or Hispanic individuals.¹⁰ Additionally, physical activity peaks in middle age, with some gender-based differences.¹⁰ Regular physical activity is particularly important for older adults to lower their risk of cognitive decline,¹¹ and injuries due to falls.¹² With the number of health benefits physical activity provides, it is not surprising that numerous health organizations, including the American Heart Association, encourage adults to engage in 150 min per week of physical activity at an intensity level of moderate or higher.^{8 , 10}

The nationwide reduction in physical activity levels during the COVID-19 pandemic3, 4, 5, 6 may have impacted some populations more than others. It is unclear how physical activity behaviors may have differed between people with conditions that put them at high risk for severe COVID-19 illness (as defined by the CDC) and people at low risk for severe COVID-19 illness. The socioecological model,¹³ the guiding framework for this study, suggests that the individual factors of individual risk are just one level of factors potentially contributing to decreased physical activity. As lockdown policies were lifted, gyms and other public physical activity locations began re-opening—a community-level factor that impacts activity choices for individuals. Despite the increased availability of publicly accessible activity venues, people at high risk for severe COVID-19 illness may have continued to practice social distancing due to heightened health-related concerns. This was problematic for individuals with multiple existing chronic conditions because maintaining or improving cardiovascular health through physical activity is imperative to mitigate worsening illness. For instance, physical activity reduces the risk for developing complications among people with obesity, arthritis, depression, cardiovascular disease, and mental health disorders.^{14 , 15}

The CDC emphasized the elevated risk of severe COVID-19 illness or death for people over the age of 65 and those with certain chronic conditions (e.g., cancer, chronic kidney disease, diabetes, heart conditions, etc.).² Information is needed to delineate when and to what degree adults at high risk for severe COVID-19 illness experienced reductions in physical activity in order to design potential targeted interventions. For instance, it is important to know if the availability of vaccinations mitigated impacts on physical activity differently for people at low and at high risk of severe COVID-19 illness. Phase 1 of COVID-19 vaccinations began on December 14, 2020 with healthcare personnel and long-term care residents vaccinated during Phase 1a.¹⁶ Over the following weeks and months the vaccine became more widely available. During Phase 1b, eligibility was extended to adults aged 75 and older as well as essential workers.¹⁶ In Phase 1c, eligibility was further extended to adults aged 65 and older, as well as individuals aged 16–64 with medical conditions that increase their risk of severe COVID-19 illness.¹⁶ It is unknown if the availability of vaccines prompted people at high risk for severe COVID-19 illness to resume pre-pandemic amounts or intensities of physical activity or if they had sustained reductions in physical activity. It is critical to determine the trends in people's physical activity during the pandemic to guide targeted intervention development for those with sustained reductions.

In this manuscript, findings are presented from a longitudinal study in which differences in the amount and intensity of physical activity are described between adults in the U.S. at high and low risk for severe COVID-19 illness as defined by CDC. The two main aims of this study are: 1) to describe differences in physical inactivity between people with high-risk conditions and those without high-risk conditions and 2) to determine the differences in the amounts of physical activity between individuals at low and high risk for severe COVID-19 illness over 13 months. The authors hypothesized that over 13 months, 1) high-risk adults would have greater odds of inactivity than low-risk adults, and 2) when active, high-risk adults would have lower MET-min than low-risk adults.

Methods

Study design and procedures

This longitudinal observational cohort study included participants from the Behavioral Outcomes During Social Distancing Study.¹⁷ Participants were initially enrolled using social media and internet-based recruitment methods. Several versions of recruitment advertisements featuring images of individuals representing various underrepresented population groups were created with the help of a graphic designer. These advertisements were promoted through paid (Facebook) and unpaid (Twitter, Facebook, Instagram, and Reddit) advertising on social media platforms. Both open and targeted postings were used in groups or channels to increase underrepresented population recruitment. The posting was shareable, and emails were sent to professional networks with a request to post the advertisements on their social media profiles. Interested individuals received information about the study and were asked a single question about the risks and benefits of the study to complete informed consent. Consented participants completed baseline data collection between March 23 to June 20, 2020. Questionnaires with demographic items were administered last due to the personal nature of these items.¹⁸ To reduce the potential for fraudulent or automated-bot responses, several reading-check items were included. Responses to reading-check items were reviewed to identify potential threats to data quality. At the completion of baseline questionnaires, participants indicating interest and providing email address contact information were enrolled in the longitudinal cohort study and invited to subsequent surveys sent out periodically during the pandemic. Rather than timing data collection on enrollment, all participants received invitations to follow-up surveys at the same time. Unfolding events of the pandemic prompted subsequent data collection including changes in state or national policies and the availability of vaccinations (see Fig. 1 ). This team of researchers met weekly to determine when and if pandemic events may alter perceptions of the pandemic or health behaviors warranting additional data collection. Follow-up surveys were sent to all longitudinal participants after recruitment closed in June, July, October, and December of 2020, and in April of 2021. All surveys were delivered to participants using the secure REDCap platform with individual links sent from within REDCap for longitudinal surveys.¹⁹ As an incentive for participating in the study, at baseline participants indicated their interest in a raffle to win one of 25 electronic gift cards valued at $25. All participants received a $5 electronic gift card at each of the five follow-up measurements, except for those who indicated that they did not want one. This study was designated as exempt by the [masked university] Institutional Review Board (STUDY2003910440).

Fig. 1.

COVID-19 timeline with sample responses.

Fig. 1 legend: Timeline of COVID-19 deaths and cases by month with selected federal and state policies and vaccine availability during the pandemic. Deaths and cases are reported in K+ (thousands) or M+ (millions). Study timepoints with sample responses are noted along the bottom. This figure is printed with permission from the authors.

Sample

Inclusion criteria for this longitudinal cohort sample were 1) English-speaking, 2) U.S. adults, 3) aged 18 years and older. For the purposes of this analysis, participants were excluded if they had not completed the physical activity and comorbidity baseline questionnaires, and at least one physical activity follow-up questionnaire. The sample for this analysis included 640 participants. Full sampling descriptions from baseline are reported elsewhere.¹⁷

Measures

Comorbidities

Participants self-reported the presence of underlying medical conditions using a modified Charlson Comorbidity Index, with increased representation of conditions associated with increased risk of life-threatening COVID-19.^{21 , 22} The original wording of the conditions was altered to improve readability for laypeople and to align with CDC wording of at-risk conditions.² Conditions assessed included heart attack, heart failure, heart disease or stroke, dementia, chronic pulmonary disease, liver disease, diabetes, kidney disease, leukemia, lymphoma, human immunodeficiency virus/acquired immunodeficiency syndrome, and hypertension.

Physical activity

To measure physical activity, the International Physical Activity Questionnaire (IPAQ) short form, a seven-item measure of physical activity levels (e.g., frequency and duration) was used.²⁰ Though the IPAQ is not as robust as measuring physical activity with actigraphy, it demonstrates good test-retest reliability (Spearman correlation coefficient = 0.80).²⁰ The IPAQ includes both leisure and work-related activity which is appropriate for our sample of both working and non-working adults. Activity was calculated as metabolic equivalent minutes per week (MET-min), which is the amount of energy expended during physical activity over the week. A ceiling limit of 180 min for any bout was implemented, and a maximum of 21 h of activity was permitted each week. Higher total MET-min indicated greater physical activity. Physical activity levels were categorized as walking (3.3 METs), moderate (4.0 METs), or vigorous (8.0 METs). Physical inactivity was defined as having no activity bouts greater than 10 min.

Sociodemographic factors

Age, gender, race, employment status, industry, education, and income were assessed using self-report measures. Race, ethnicity, and gender were assessed with choose all that apply questions.

Statistical method

After the longitudinal surveys were closed, a de-identified dataset was downloaded and directly uploaded into SPSS 27.²³ Statistical analyses were conducted using SAS/STAT ® version 9.4.²⁴ To account for lower variability in response distribution, item responses were collapsed for gender (man, woman, other, not-reported) and race (non-white, white, not-reported). Individuals were categorized as high-risk if they reported having any of the following conditions: heart attack, heart failure, heart disease or stroke, dementia, chronic pulmonary disease, liver disease, diabetes, kidney disease, leukemia, lymphoma, human immunodeficiency virus, and/or hypertension (as described by the CDC).² Individuals with none of these conditions were considered low-risk. Demographic characteristics were compared between high- and low-risk participants at baseline (March 2020). Chi-square or Fisher's exact tests were used for the categorical variables and Wilcoxon rank sum was used for the continuous variables, as they are not normally distributed.

Analysis of study aims

To examine the differences in physical activity between high-risk adults and low-risk adults, a two-part model was created, which simultaneously assessed the association of physical inactivity with high/low risk (hypothesis 1) and, for those who were physically active, the association of total MET-mins with high/low risk (hypothesis 2). A logistic model was fit to determine the probability of no physical activity while a gamma model (generalized linear model with the gamma distribution and log link) was utilized for total MET-mins which better suits right-skewed data. A random subject intercept was included in each of the two parts of the model to incorporate the correlation of data from repeated measurements. Models were first fit with covariates of risk group, survey timepoint, and the interaction between risk group and survey timepoint. Models were then fit with additional demographic variables of age, gender, and race. Parameter estimates and 95% confidence intervals from the logistic models were calculated as were odds ratios of no physical activity for high-risk vs. low-risk. Estimated means of total MET-mins and rates of estimated means for high-risk vs. low-risk. Statistical significance was set at an alpha level of 0.05.

Results

Sample characteristics

Table 1 shows the descriptive statistics for the 640 participants meeting inclusion criteria for this analysis. Of the 640 participants between 322 and 512 completed each of the follow-up surveys (see Fig. 1). Respondents primarily reported being non-Hispanic white (90%) and women (78%). Approximately half of this sample reported working in an industry determined to be essential by the CDC.²⁵ Participants ranged in age from 18 to 81 years of age, with a mean age of 42.7 ± 15.7, and more older adults were categorized as high risk than low risk.

Table 1.

Sample Sociodemographics.

Variable	Overall	High Risk Group	Low Risk Group	p-value
	n=640	n = 175	n = 465
Age in years, mean ± SD	42.7 ± 15.7	50.8 ± 16.1	39.6 ± 14.4	< 0.001
Gender, n (%)				0.623
Woman	497 (78.4)	136 (78.6)	361 (78.3)
Man	103 (16.2)	30 (17.3)	73 (15.8)
Othera	34 (5.4)	7 (4.0)	27 (5.9)
Race, n (%)
Non-Hispanic White	572 (89.7)	157 (89.7)	415 (89.6)	0.976
Hispanic or Non-White	66 (10.3)	18 (10.3)	48 (10.4)
Employment Status, n (%)
Full-time employment	329 (61.3)	69 (47.3)	260 (66.5)	<0.0001c
Part-time employment	56 (10.4)	14 (9.6)	42 (10.7)
Student	46 (8.6)	9 (6.2)	37 (9.5)
Retired	53 (9.9)	32 (21.9)	21 (5.4)
Unpaid Volunteer or Caregiver	24 (4.5)	7 (4.8)	17 (4.3)
Disabled	15 (2.8)	9 (6.2)	6 (1.5)
Unemployed	14 (2.6)	6 (4.1)	8 (2.0)
Industryb, n (%)
Essential	325 (50.8)	88 (50.3)	237 (51.0)	0.878
Non-essential	315 (49.2)	87 (49.7)	240 (51.7)
Income Level, n (%)
$0 to $54,999	100 (28.8)	30 (29.4)	70 (28.6)	0.163
$55,000 to $99,999	98 (28.2)	30 (29.4)	68 (27.8)
$100,000 to $149,999	75 (21.6)	27 (26.5)	48 (19.6)
$150,000 to $199,999	36 (10.4)	10 (9.8)	26 (10.6)
$200,000 and above	38 (11.0)	5 (4.9)	33 (13.5)
missing	293	73	220

Note:.

^a“Other” includes questioning, non-binary, agender, gender queer/gender nonconforming, choose to not identify, and gender not listed.

^b“Essential” included healthcare and public health, food and agriculture (including grocers and other food/beverage retail), financial services, energy and utilities, emergency services, defense services, information technologies, manufacturing, communications, transportation (including public and private transport of people or goods), commercial facilities (including apartments, condos, and mixed-use), and restaurants/bars/food service industries.

^cEmployment comparison is of full-time employment versus not.

Study variable statistics

Observation of mean physical activity reported suggest MET-mins were generally stable during the study period. Slight changes in the total reported mean MET-mins for the sample and the high and low risk groups can be seen in Table 2 .

Table 2.

Sample physical activity minutes over time.

Timepoint	Overall		High Risk Group	Low Risk Group
	n	mean ± SD	mean ± SD	mean ± SD
Mar. 2020	640	2242.6 ± 2608.1	1965.7 ± 2586.9	2346.8 ± 2611.2
June 2020	512	2669.1 ± 3184.2	2177.1 ± 2755.2	2856.1 ± 3317.1
July 2020	322	2279.7 ± 2589.3	1843.6 ± 2167.3	2438.6 ± 2713.6
Oct. 2020	416	2219.8 ± 2914.8	2106.2 ± 2732.4	2266.9 ± 2990.4
Dec. 2020	392	1984.7 ± 2614.6	1831.8 ± 2416.8	2040.6 ± 2685.2
Apr. 2021	377	2305.8 ± 2805.5	2167.6 ± 2503.1	2362.8 ± 2923.6

Note: MET-min is the weekly mean metabolic equivalent minutes reported.

The baseline comorbidities placing individuals in the high-risk group reported by participants included heart attack (n = 4, 0.6%), heart failure (n = 4, 0.6%), heart disease or stroke (n = 16, 2.5%), chronic pulmonary disease (n = 72, 11.3%), liver disease (n = 8, 1.3%), diabetes (n = 42, 6.6%), kidney disease (n = 6, 0.9%), leukemia (n = 2, 0.3%), lymphoma (n = 2, 0.3%), and hypertension (n = 75, 11.7%%). Dementia and human immunodeficiency virus were not reported by any of the participants. Participants who did not report any of these conditions were categorized as low-risk for severe COVID-19. Of the 640 participants, 27% were in the high-risk group and 73% low-risk group (see Table 1). There were no statistically significant differences between groups in gender, race, essential vs. non-essential industry, or income levels. As compared to the high-risk group, the low-risk group was significantly younger and more likely to be employed full-time.

Inactivity

Fig. 2 shows the percentage of participants reporting inactivity in each the low- and high-risk groups at each study timepoint. High-risk adults demonstrated a higher percentage of inactive individuals at each timepoint as compared to low-risk adults (see Fig. 3 ). The odds of high-risk adults reporting inactivity compared to low-risk adults is significantly greater in both March 2020 (p = 0.004) and in April 2021 (p = 0.009; see Table 3 ).

Fig. 2.

Percentage of inactive participants at each time point by risk group.

Fig. 2 legend: Percentage of inactive participants in each the high- and low-risk groups at each survey timepoint.

Fig. 3.

Log odds of inactivity for high-risk adults.

Fig. 3 legend: The odds of inactivity for adults at high risk of COVID-19 illness as compared to adults at low risk of COVID-19 illness at each survey timepoint. All data reported on the natural log scale with 95% confidence limits in the shaded area.

Table 3.

Odds of high-risk adults reporting inactivity.

Time	Odds Ratio	95% CI	p-value
Mar. 2020	2.81	1.40, 5.64	0.004
June 2020	2.85	0.98, 8.26	0.054
July 2020	3.48	0.95, 12.73	0.059
Oct. 2020	2.60	0.79, 8.62	0.118
Dec. 2020	2.04	0.70, 5.93	0.190
Apr. 2021	4.12	1.42, 11.93	0.009

Note: CI, confidence interval.

Physical activity

The mean weekly number MET-min activity reported for both groups fluctuated over time as shown in Fig. 4 . For individuals that were active (reporting at least a single bout of activity more than 10 min), the rate of MET-min activity was lower in the high-risk group than the low-risk group (see Fig. 5 ). Significant differences were determined in the rate of MET-min activity. Rates were 28% lower in the high-risk group as compared to the low-risk group in March 2020, 29% lower in June of 2020, and 30% lower in July of 2020 (see Table 4 ). Subsequent timepoints were not significant for differences in MET-mins by group.

Fig. 4.

Weekly MET-min reported at each time point by risk group.

Fig. 4 legend: The mean weekly number of metabolic equivalent minutes (MET-min) for individuals at high-risk and at low-risk for severe COVID-19 over time.

Fig. 5.

MET-min of total physical activity by risk group.

Fig. 5 Legend: Weekly mean metabolic equivalent minutes (MET-min) of physically active adults at high risk of COVID-19 illness are shown in a red line and adults at low risk of COVID-19 illness in a blue line. All data reported on the natural log scale with 95% confidence limits represented as shaded areas.

Table 4.

Rate of total physical activity in high-risk group as compared to low-risk group among physically active participants.

Time	Rate	95% CI	p-value
Mar. 2020	0.72	0.58, 0.88	0.002
June 2020	0.71	0.57, 0.89	0.003
July 2020	0.70	0.54, 0.90	0.005
Oct. 2020	0.82	0.65, 1.04	0.100
Dec. 2020	0.81	0.81, 1.03	0.088
Apr. 2021	0.90	0.70, 1.14	0.381

Note: CI, confidence interval.

Discussion

Early in the pandemic, evidence showed that individuals with a variety of chronic illnesses were at a greater risk for life-threatening COVID-19 illness. This life-threatening risk, according to the CDC, includes the need for hospitalization or intensive care, mechanical ventilation and an increased mortality risk.² Warnings about the risk of severe COVID-19 illness were intended to encourage health behaviors that would reduce potential exposure to the SARS-CoV-2 virus including social distancing and mask-wearing. As the pandemic continued, it became clear that of the individuals who developed COVID-19, people with hypertension, diabetes, and cardiovascular and respiratory system diseases were the most vulnerable groups.²⁶

During the COVID-19 pandemic changes to everyday life included increased sedentary behaviors and decreased physical activity.3, 4, 5 These results can be seen in the outcomes of this study as well. This study expands the knowledge of physical activity in regard to differences between adults considered to be at low- as compared to high-risk for severe COVID-19 illness. During the first month of the pandemic, participants in this study who were at high risk were more than twice as likely to be inactive as those at low risk. High-risk adults reported more inactivity at all timepoints and fewer MET-mins early in the pandemic than low-risk adults in this study. This study also identified difference in ages by group, representative of increases in proportions of high-risk conditions seen in older adults.²⁷

Unfortunately, social distancing has drawbacks, particularly for people aiming to improve their health. To prevent worsening of chronic conditions, self-management behaviors, including sufficient physical activity, are a necessity.  With 6.7 billion gym visits in the U.S. in 2019, the closing of gyms may have played a role in the reported 50% of U.S. adults that reported becoming less active during the pandemic.⁶ The need for viral mitigation and disruption of established self-management behaviors associated with physical activity were complicated by confusion about mask-wearing, social distancing, and the closing of gyms and group exercise.

A variety of potential changes during the 13 months of this study may have contributed to changes in inactivity or the MET-mins reported. Other studies have suggested confinement,²⁸ changes in active commuting,²⁹ advanced age, relationship status, living alone, and depression³⁰ may be contributors. Interestingly, in this study, when the vaccine became available in December 2020 to high-risk individuals, physical activity levels did not improve. It may be that behavior changes related to the pandemic were firmly in place after 10 months without vaccines available. Unfortunately, both the high- and low-risk adults demonstrated increases in inactivity percentages at the time of vaccine availability. One potential opportunity for further inquiry is examining how COVID-19 infection or vaccination at an individual level may have impacted physical activity and inactivity.

Another potential contributor to physical activity changes is the fluctuating information, policies, and guidelines available over time. Although the results of this study do not control for the changes in policies and guidelines, it is notable that there were fluctuations in physical activity in both groups that may be reflective of the frequently changing policies, guidelines, and available information. These fluctuations could be seen not just in the U.S., but globally as well. Although difficult to disentangle, environments individuals lived in during the pandemic impacted physical activity, which includes policies and built environments.³¹

This shift to non-significant differences later in the pandemic may be explained by sustained changes in the spaces people use for physical activity or by weather-related changes. In 2019, the U.S. had over 40,000 fitness facilities, but by July 2021 22% of these facilities had permanently closed.⁶ As of January 2022, 25% of all fitness facilities across the United States have closed due to the pandemic.³² This time period also saw an increase in at-home physical activity, including the use of live stream exercise and physical activity sessions.³³ Additionally, the non-significant differences occurred during the coolest months of the year. A systematic review prior to the pandemic suggests that globally physical activity declines in winter months, which may have only been exacerbated during the pandemic.³⁴ More research is needed to explain the rationale for differences versus no differences in high-risk as compared to low-risk adults.

There are limitations of this study that should be considered. First, individuals were categorized as low-risk vs. high-risk without attention to disease severity. Although this is consistent with CDC recommendations,² it may not sufficient in the prediction of physical activity. Second, the participants in this study are primarily composed of middle-aged, non-Hispanic White women, limiting the generalizability of the study findings. This convenience sampling online may have contributed to the lack of diversity. Finally, self-reported physical activity is not as robust as other objective measurements.²⁰

Although this study is limited in the interpretation of potential explanatory variables for the declines in physical activity, the findings show that insufficient physical activity clearly exists. Interventions at every level (individual, community, etc.) to increase physical activity could be beneficial. A recently published review of physical activity trends during the pandemic suggests intervening at the community level by improving outdoor spaces.³¹ This includes increasing the accessibility and the number of green spaces for people to freely use. In lower-income urban neighborhoods, the accessibility and quality of outdoor spaces are associated with better health in the communities.³⁵ Improving access to outdoor spaces is particularly beneficial to high-risk adults because they may have less concern about being exposed to COVID-19 or other potentially severe infections in these spaces compared to fitness facilities. Developing community-level interventions by improving outdoor spaces is one means of addressing the issues of inactivity and reduced MET-min for individuals at both high and low risk for severe COVID-19 illness. It should be noted that as many areas of the US experience weather related changes that can impact the safety of outdoor spaces. Weather as well as other neighborhood safety concerns may also need to be addressed to improve safe and accessible spaces. Much work still needs to be done at all levels as the evidence continues to reveal declining physical activity and the general worsening of cardiovascular health.

Conclusion

Physical activity behaviors fluctuated over the pandemic. Adults at high risk of severe COVID-19 illness were disproportionately more likely to be physically inactive as compared to adults at low risk. During the early months of the COVID-19 pandemic, adults at high risk also exhibit lower MET-min levels than adults at low risk.

Declaration of Competing Interest

There are no conflicts of interest to disclose.