摘要：Objectives. We evaluated the influence of physical activity resources and neighborhood-level socioeconomic status (SES) on walking among community-dwelling older men. Methods. Participants reported time walked per day at baseline (2000–2002) and follow-up. Residential addresses were linked to a geographic information system database to assess proximity to parks, trails, and recreational facilities. Log-binomial regression analyses were conducted to test the hypothesis that men living near physical activity resources were more likely to increase or maintain time walked. Results. Average time walked per day declined by 6 minutes between baseline and follow-up ( P < .05). There was a significant interaction of neighborhood SES and physical activity with walking time ( P < .1). Proximity to parks and proximity to trails, respectively, were associated with a 22% (95% confidence interval [CI] = 1.01, 1.47) and 34% (95% CI = 1.16, 1.55) higher likelihood of maintaining or increasing walking time in high-SES neighborhoods, but there was no association in low-SES neighborhoods. Proximity to recreational facilities was not associated with walking. Conclusions. Uncovering reasons that proximity to parks and trails is not associated with maintenance of walking activity among men in low-SES neighborhoods could provide new insight into ways to promote physical activity. Walking is the most common form of exercise among older adults. 1 , 2 Walking can be integrated into daily routines 3 , 4 and can be a feature of neighborhood life. 5 – 7 A physically active lifestyle is linked to disease prevention and risk reduction, as well as to physiological, functional, psychological, and social benefits. 8 , 9 Despite this widespread knowledge, physical activity declines with age, and the majority of older adults do not meet physical activity recommendations to maintain good health. 10 , 11 Therefore, identifying factors that encourage elderly adults to maintain physical activities, such as walking, are vital for health promotion. The behavioral ecological model (BEM) of health promotion asserts that both the physical and social environments contribute to an individual's health behaviors. 12 In the context of physical activity, the BEM theory posits that proximate physical activity resources—such as parks, trails, and recreational facilities—serve as cues for physical activity behaviors. Observing or interacting with people engaging in activities in locations where physical activity occurs provides social and environmental reinforcement for one to do the same. The BEM underlies the hypothesis that physical activity will increase in populations that have access to nearby physical activity resources. To date, cross-sectional studies have generally shown positive associations between proximate physical activity resources and physical activity, 13 – 21 although some studies have reported no association. 22 , 23 Despite promising results, cross-sectional studies provide no information on the effects that proximate physical activity resources have on maintenance of physical activities. In addition, few studies have focused specifically on older adults, a population that may be particularly sensitive to the influence of accessibility to physical activity resources. 22 , 23 Finally, although neighborhood socioeconomic status (SES) affects the availability and quality of physical activity resources, 13 , 24 – 27 few studies have evaluated the extent to which neighborhood SES may affect the association between physical activity resources and level of activity. To determine the effects of proximate physical activity resources on maintenance of walking activity, we conducted a prospective study among community-dwelling men 65 years or older from the Portland, Oregon, metropolitan area. Using a novel combination of individual-level and neighborhood-level data, we sought to determine whether older men who live within one eighth, one quarter, or one half mile of physical activity resources, including parks, trails, and recreational facilities, are more likely than men who live further from these resources to maintain or increase the amount of time they spend walking. We also assessed whether relationships between physical activity resources and maintenance of walking activity differ according to neighborhood SES.
