Public Park System Characteristics Associated with Physical Activity

Thesis (Ph.D.)--University of Washington, 2017-03Physical activity (PA) is protective of many chronic diseases, but most Americans are not sufficiently active. Public parks are places where PA occurs, and thus expanding park land or the facilities in parks could result in greater population levels of PA. There is a lack of robust evidence, however, linking greater park proximity or more or better park facilities with higher PA levels at the individual level. In this dissertation, we identified and characterized park visits and corresponding park-based PA using timestamp-linked travel diary, GPS, and accelerometer data collected from a population-based sample of urban King County adults observed for one to three one-week periods over 4 years. Using these data, we sought to advance the understanding of how park proximity and facilities in parks are associated with PA. In Aim 1 we assessed how proximity to parks is associated with PA. We divided total PA bout time into three mutually exclusive categories: PA that occurred during visits to home neighborhood parks, PA that occurred during visits to non-home neighborhood parks, and all other PA. We found that home neighborhood park proximity (count and area of parks within a 10-minute walk) was positively associated with home neighborhood park PA. But since home neighborhood park PA accounted for an average of only 3% of total PA, home neighborhood park proximity was not associated with total PA. In Aims 2A and 2B we tested the association between the variety of facilities in parks and PA. In Aim 2A we treated individual participants as their own controls to compare the variety of PA facilities in different parks that an individual visited while active versus sedentary. We found that each additional different type of PA facility in a park was associated with a 7% increased probability of an individual being active during a visit. In Aim 2B we assessed if the variety of PA facilities at a park was associated with duration of PA during a park visit. We observed that each additional different type of PA facility in a park was independently associated with a 7.3% greater duration of PA during the park visit. Our findings based on comprehensive and objective park visit data provide strong evidence for an association between parks and individual-level PA, but also place the association in the context of overall PA. Because park-based PA is a small proportion of total PA, investing in parks should be viewed as one of a portfolio of strategies to create an overall environment more conducive to physical activity