Daily Step Counts Before and After the COVID-19 Pandemic (2024)

The COVID-19 outbreak had a global impact on physical, mental, and social health. Data collected early in the pandemic suggested a general decline in step counts worldwide,¹ but factors contributing to reduced activity have not been identified.

This study aimed to examine whether COVID-19 was associated with daily activity among All of Us (AOU) research program² participants. We hypothesized that a decline in activity across the US would persist after most social distancing recommendations were relaxed, and that socioeconomic factors and mental health status would continue to affect reduced activity.

Using AOU Research Program Controlled Tier data set (released June 2022), this cohort study included participants who wore a digital device tracking physical activity (Fitbit Inc) for at least 10 days each month (eMethods in Supplement 1). Daily steps (averaged monthly) were examined over 4 years from January 2018 through December 2021. Counterfactual analysis based on 2 years of activity data preceding COVID-19 was used to estimate postpandemic steps. Differences in observed and estimated post–COVID-19 daily steps were examined using Wilcoxon rank sum test with continuity correction.

Surveys were used to assess sociodemographic factors, mental health, location, and deprivation index at the time of enrollment.³ Linear mixed models examined associations of COVID-19 vaccine, comorbidities, sociodemographic, and mental health factors with differences in observed and estimated post–COVID-19 daily steps (eMethods in Supplement 1).

In this study, only the authorized authors who completed AOU Responsible Conduct of Research training accessed the deidentified data from the Researcher Workbench (a secured cloud-based platform). Because the authors were not directly involved with the participants, institutional review board review and the requirement for informed consent was exempted. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. Analysis was conducted using R version 4.2.2 (R Project for Statistical Computing). The threshold for significance was P < .05 determined in 2-sided tests.

We analyzed data from 5443 (3903 [71.7%] female; 257 [4.7%] Black, 4681 [86.0%] White, 505 [9.3%] other; median [IQR] age, 53 [38-64] years) participants with valid Fitbit data for at least 6 months pre- and post–COVID-19 (Table). Median (IQR) observed daily step counts pre–COVID-19 (ranged from January 1, 2018, to January 31, 2020) and post–COVID-19 (ranged from June 1, 2020, to December 31, 2021) were 7808 (5923-10 108) steps and 7089 (5101-9740) steps, respectively. The counterfactual model estimated participants walked 575 (95% CI, 521-629) fewer steps per day post–COVID-19 compared with observed daily steps (P < .001; Figure, A). The difference between observed and estimated post–COVID-19 steps was significantly explained by younger age (β = −243 per 10-year decrease; P < .001), Northeast region compared with other regions (Northeast vs others, β = −288, P < .001), and higher deprivation index (β = −477 per 0.1 increment; P < .001) (Figure, B). Post–COVID-19 step counts were also explained by COVID-19 vaccination status (vaccinated vs unvaccinated: β = 48; P < .01), depression (β = −36 per 1 score increment; P < .01), and psychological stress (β = −13 per 1 score increment; P < .01) (Figure, C). We found no association between reduced step counts and sex or comorbidities such obesity, diabetes, coronary artery disease, hypertension or cancer.

These findings suggest a consistent, widespread, and significant decline in activity following the onset of COVID-19 in the US. Vulnerable populations, including individuals at a lower socioeconomic status and those reporting worse mental health in the early COVID-19 period, were at the highest risk of reduced activity. We found a statistically significant decline in daily step counts that persisted even after most COVID-19–related restrictions were relaxed, suggesting COVID-19 affected long-term behavioral choices. Currently, it is unknown whether this reduction is steps is clinically meaningful over time. Any meaningful difference is likely dependent on baseline activity, age, and other patient-level factors. Our prior work in the AOU cohort suggests that modestly lower step counts over a long period could have a substantial contribution to long-term disease risk.⁴

While our data include some selection bias (eg, majority were White, young, and active and only included participants who owned and reported valid tracking data), this finding extends the work of prior studies showing an early reduction in activity.^1,5 We also imposed relatively stringent criteria for a valid tracking day, which could have biased our results toward more active individuals. The 2% of our cohort with prior COVID-19 infection likely underestimated the true prevalence as it did not include positive home tests. Despite these limitations, our study showed COVID-19 to be associated with decreased physical activity, which may have implications for future cardiometabolic risk with the potential to exacerbate health disparities related to socioeconomic status and mental health.

Accepted for Publication: January 31, 2023.

Published: March 20, 2023. doi:10.1001/jamanetworkopen.2023.3526

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2023 Desine S et al. JAMA Network Open.

Corresponding Author: Evan L. Brittain, MD, MSc, Division of Cardiovascular Medicine, Vanderbilt University Medical Center, 2525 West End Ave, Ste 300A, Nashville, TN 37203 (evan.brittain@vumc.org).

Author Contributions: Dr Brittain had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Master, Annis, Roden, Harris, Brittain.

Acquisition, analysis, or interpretation of data: Desine, Master, Annis, Hughes, Brittain.

Drafting of the manuscript: Desine, Master, Brittain.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Desine, Master, Annis.

Obtained funding: Harris.

Administrative, technical, or material support: Annis.

Supervision: Master, Annis, Brittain.

Conflict of Interest Disclosures: None reported.

Funding/Support: The All of Us Research Program is supported by the National Institutes of Health, Office of the Director (Regional Medical Centers: Nos. 1 OT2 OD026549, 1 OT2 OD026554, 1 OT2 OD026557, 1 OT2 OD026556, 1 OT2 OD026550, 1 OT2 OD 026552, 1 OT2 OD026553, 1 OT2 OD026548, 1 OT2 OD026551, 1 OT2 OD026555; IAA: Nos. AOD21037, AOD22003, AOD16037, AOD21041; Federally Qualified Health Centers: No. HHSN 263201600085U; Data and Research Center: No. 5 U2C OD023196, OT2 ActOD035404; Biobank: No. 1 U24 OD023121; The Participant Center: No. U24 OD023176; Participant Technology Systems Center: No. 1 U24 OD023163; Communications and Engagement: Nos. 3 OT2 OD023205, 3 OT2 OD023206; and Community Partners: Nos. 1 OT2 OD025277, 3 OT2 OD025315, 1 OT2 OD025337, 1 OT2 OD025276). Dr Brittain is supported by Vanderbilt University Medical Center (grant Nos. R01 HL146588 and 1 R61 HL158941-01A1).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 2.

Additional Contributions: The All of Us Research Program would not be possible without the partnership of its participants.