Factors associated with meeting physical activity guidelines during the COVID-19 pandemic

The COVID-19 pandemic caused global changes in the way people conduct their daily lives. For instance, shelter in place/stay at home mandates (heretofore Stay-at-Home) enacted to help reduce community transmission restricted the movement of individuals [1]. Restricted movement, in turn, may have contributed to the reduction in physical activity across many demographic groups [2‐7]. However, the reasons for the decreases in physical activity during COVID-19, or more generally, the factors associated with physical activity during this time, are still not well understood. Understanding these factors is critical to informing future infectious disease mitigation policies that promote, rather than hinder physical activity [8]. This is especially important given evidence of the positive impact of physical activity on mental health during the pandemic [9], immune response to vaccination [10], COVID-19 outcomes [11], and critically, that physical inactivity will continue to be an important risk factor in future pandemics [12‐14].

Less is known about factors associated with physical activity during the COVID-19 pandemic among U.S. adults. In the U.S., COVID-related state and municipal ordinances included the wide-spread closure of gyms and open spaces, such as parks, for prolonged periods of time [15‐17], and the closing of some public open spaces, such as parks [18, 19]. If indoor recreational facilities were closed, but outdoor green spaces remained open, individuals living in areas with fewer of these public outdoor spaces or in unsafe neighborhoods may have experienced reduced opportunities for physical activity [20‐22]. Prior to the pandemic, studies had demonstrated that the neighborhood environment, including perceived safety, were associated with physical activity [23], but it is unknown if and how these factors would impact physical activity during the COVID 19 pandemic. However, under Stay-at-Home order, higher income individuals who tend to have more access to yard space or exercise equipment in the home [24‐26] may not have been as affected as low-income individuals. Considerable gaps remain in understanding how Stay-at-Home orders and neighborhood safety impacted physical activity during the COVID-19 pandemic.

Similarly, knowledge regarding how sociodemographic characteristics are associated with physical activity during the pandemic could help inform future promotional strategies for specific individuals and/or communities. For instance, while pet ownership has been found to be associated with physical activity in other countries during COVID-19 [27], this relationship has yet to be investigated with a U.S. sample during the COVID-19 pandemic. In this same context, even with restricted movement, pet owners may still need to go outside the home to tend to their pets’ needs.. Moreover, the physical activity of adults may have been negatively impacted in households with school-age children due to remote learning [28, 29]. Lastly, while socioeconomic factors are known to be associated with physical activity, including income and education [30], it is important to confirm this relationship held during the COVID-19 pandemic.

The purpose of this study is to assess the association of Stay-at-Home orders and length, neighborhood safety, and demographics variables with meeting physical activity guidelines in a predominately Hispanic and female group of survey respondents during the initial phase of the COVID-19 pandemic (April to June, 2020).

The average age of respondents was 42.7 years old (Table 1). The majority of respondents were female, Hispanic, had a bachelor’s degree or higher, made more than $70,000 per year, were employed, had a pet and no children. Most respondents reported there was a Stay-at-Home order where they lived at the time they completed the questionnaire (78%), which is not surprising given 56% of the sample came from two neighboring counties on the Texas-Mexico border. At the time the survey was administered, the average amount of time respondents reported being under a Stay-at-Home order was 49.5 days, with a minimum of 31 days and a maximum of 110 days (data not shown). As a result of the pandemic, 46% of participants said they increased all types of intensities of activity (mild, moderate and vigorous) with all others reporting some variation in their activity levels.

We explored the association between neighborhood safety and physical activity. In the adjusted association of neighborhood safety tertiles with meeting physical activity guidelines, we see, on average, each subsequent tertile (progressing from most to least safe) is associated with lower odds of meeting physical activity guidelines (Fig. 2A). In the adjusted association of the continuous neighborhood safety scale with meeting physical activity guidelines, we see an inverse association between the two variables, which then levels off at values higher than 15, indicating that higher levels on the neighborhood safety scale (i.e. the highest perception of safety) no longer impact the probably of meeting physical activity guidelines (Fig. 2B).

We found significant associations between race/ethnicity, education, length of Stay-at-Home order, and neighborhood safety tertiles with physical activity level before adjustment (Table 2). We tested the interaction of both income and sex separately with physical activity, and found a statistically significant interaction with income. Therefore, we stratified all results by income. Among individuals in the lowest income group (i.e. < $40,000), those with some college as compared to a college degree or higher had 74% lower odds of meeting physical activity guidelines. Among individuals in the < $40,000 income group, living in the most unsafe neighborhood (i.e. tertile 3) was associated with an 85% lower odds of meeting physical activity guidelines as compared to those in the 1^st tertile (safest neighborhood). We found no statistically significant association between neighborhood safety and physical activity in either the $40,000—$69,999 or ≥ $70,000 income groups. Lastly, we found a strong, statistically significant association between pet ownership and physical activity, but only in the highest income group (i.e. ≥ $70,000).

In this assessment of U.S. adults, representing a predominately Hispanic, Texas-based sample, we found lower levels of perceived neighborhood safety (as compared to the highest level of perceived neighborhood safety) were associated with lower odds of meeting physical activity guidelines, but only in adults reporting a household income of less than $40,000 a year. These findings mirror those found by colleagues, who identified that the COVID-19 pandemic decreased use of outdoor spaces in low socioeconomic status (SES) areas, but actually increased use in high SES areas [25]. Our findings could be because of the confluence of unsafe neighborhoods with low income or low SES neighborhoods [34, 35], or could be because above a certain income, individuals have sufficient disposable income to pay for equipment in their homes or have enough private outdoor area on their property to allow them to maintain their physical activity [24‐26]. Importantly, irrespective of income bracket, we saw an inverse association between neighborhood safety and physical activity, but this leveled off at values on the neighborhood safety scale higher than 15 (i.e. the least safe neighborhoods). This leveling off at 15 would seem to indicate that higher levels on the neighborhood safety scale (i.e. the highest perception of safety) no longer impact the probability of meeting physical activity guidelines. Individuals only need a specific threshold of safety met for them to feel comfortable in outdoor spaces.

As compared to those with a college education or higher, holding an Associate's degree, completing technical school, or some college was associated with lower odds of meeting physical activity guidelines, but only in adults reporting a household income of less than $40,000 a year. Previous data has shown an association of low income and lower education with lower levels of physical activity [36, 37]. These findings may reflect the overall patterns of health inequities or the demand on many essential workers during the pandemic [38‐41], many of whom have jobs requiring less education and tend to receive lower salaries. It also may reflect that those with lower incomes had to continue working in-person despite the COVID-19 risks and therefore had less time to be physically active. Our study contributes to the literature on the association of lower education and lower income with lower levels of physical activity during the early part of the pandemic.

Moreover, we found pet ownership in the highest income bracket (≥ $70,000) was associated with increased odds of meeting physical activity guidelines, though this statistically significant association was not found in the two lower income brackets. A similar association was found in a study conducted in Singapore in an affluent sample [27]. Stay-at-Home orders and other business-specific decisions to continue working at home during the first four months of the pandemic favored office workers and those with higher income [41, 42], and thus, may have provided more time for these at-home workers with pets, particularly those with dogs, to get outside and engage in more physical activity. It is important to note that regardless of the COVID-19 pandemic, dog-ownership specifically is associated with walking [43, 44], so these findings may reflect a persistent underlying association.

There are several limitations to this study. Foremost, this was a cross-sectional study, which limits our ability to infer causation and observe how these same individuals were impacted by social and demographic factors before and well after the first few months of the COVID-19 pandemic. Our sample was a convenience sample of individuals willing to complete an online survey, thus, our results are not representative of any specific demographic group and it should be assumed that individuals with less access to reliable internet may have been more likely to be excluded from this sample. More specifically, we had a limited number of men included in this study (21%). We also did not assess the actual neighborhood environment, so it is possible that those reporting “unsafe” neighborhoods may not truly live in one. However, research shows perceived, rather than objective, neighborhood environment is more critical for physical activity [23].

We found perceived neighborhood safety, educational levels and pet ownership were associated with meeting physical activity guidelines during the early months of the COVID-19 pandemic, though associations varied by annual household income. These findings can inform approaches to promoting physical activity during subsequent waves of COVID-19 or future pandemics, potentially targeting different SES groups with messaging most appropriate to their needs.