A Multidimensional Analysis for Understanding Walking Habits in Older Adults Post-Pandemic

Term Start:

June 1, 2024

Term End:

May 31, 2025

Budget:

$150,000

Keywords:

Older Adults, Post-Pandemic Era, Walking Behavior

Thrust Area(s):

Data Modeling and Analytic Tools, Equity and Understanding User Needs

University Lead:

The University of Texas at Austin

Researcher(s):

Chandra Bhat

This study addresses a critical gap in the literature by offering a novel analytical lens to understanding walking behaviors among older adults in the post-pandemic era. The walking survey of older adults in the US population to be used in the proposed research was undertaken through the Foresight 50+ Consumer Omnibus panel survey, which constitutes a probability-based panel designed to be representative of the US household population age 50 or older. The specific walking survey instrument used in the current paper was funded by the American Association of Retired Persons (AARP) and undertaken in July 2022. Using the sample of about 1700 individuals, we undertake an analysis that has several salient aspects.

First, by utilizing data collected in July 2022, the study uniquely focuses on walking habits after the peak of the pandemic. While existing research, such as the study by Hwang et al. (2023), has explored changes in walking frequency post-pandemic, these studies often relied on respondents’ expectations of walking rather than their observed behavior. Our study overcomes this limitation by using data grounded in real-world experiences of individual walking patterns (admittedly though, as self-reported by individuals).

Second, the focus on the demographic aged 50 and above addresses a group frequently overlooked in walking research, particularly concerning the long-term effects of the pandemic. Pre-pandemic data highlights a decline in adherence to recommended physical activity levels with age. For instance, in 2018, only 45.1% of those aged over 64 met the recommended activity standards, compared to 59.4% among adults aged 18 to 64 (National Center for Health Statistics, 2019). The pandemic has exacerbated this decline in walking frequency, potentially due to diminished activity and muscle atrophy resulting from extended periods of contact restrictions. As a result, our findings can inform equitable and effective policy interventions focused on the 50+ age group, where walking can significantly improve quality of life and where the pandemic may have caused irreversible changes.

Third, the multivariate perspective integrates frequency, duration, and companionship into a unified analytical framework. The walking frequency dimension reflects the usage of walking as a mode of transportation. The duration measures the total volume and intensity of physical activity, capturing the direct effects of walking on physical health. Additionally, the companionship dimension captures the role of social factors, such as the presence of walking companions, as vital motivators for sustaining an active lifestyle, especially pertinent for the older demographic. This framework includes an ordered response probit model to evaluate weekly walking frequency (categorized as 1-2 days, 3-4 days, 5-6 days, and 7 days), an ordered response model for assessing average weekly duration (segmented into intervals of 10 minutes, 11-29 minutes, 30-59 minutes, and ≥60 minutes), and another ordered response model for the frequency of walking with someone else (ranging from Never, Rarely, Sometimes, Often, to Always). This approach allows for a comprehensive understanding of the interplay between sociodemographic, built-environment, and perceptual factors on various aspects of walking behavior among those aged 50 and above. Also, it permits controlling for unobserved factors that may lead to associations among the three outcome variables. For example, an individual’s underlying physical fitness level or motivation for an active lifestyle could simultaneously impact their walking frequency, typical walking duration, and preferred companionship. By jointly modeling these outcomes, we can account for such unobserved confounders and obtain more accurate estimates of the determinants specific to each walking dimension.

Scroll to Top