Association Of Perceived And Objective Neighborhood Walkability With Accelerometer Measured Physical Activity And Sedentary Time

PURPOSE: We aim to investigate the association of neighborhood walkability with accelerometer-measured physical activity (PA) and sedentary behavior (SB) levels and examine whether objective and subjective measures of walkability result in similar findings. METHODS: PA and SB from the first 7,689 Maastricht Study participants were measured using accelerometers for a period of seven days. Mean daily step count, light intensity PA, moderate to vigorous PA (MVPA), and SB were calculated. Objective walkability was measured by the 7-component Dutch Walkability Index within 500 m Euclidian buffers around residential addresses of participants. Subjective walkability was obtained from the Abbreviated Neighborhood Environment Walkability Scale. Linear regression models analyzed the associations of walkability with PA and SB, controlling for potential confounders. RESULTS: Objective walkability was negatively associated with light intensity physical activity in the most walkable quartiles (Q3: b = -15.05, 95% CI = -21.38, -8.73; Q4: b = -14.58, 95% CI = -20.94, -8.23). Compared to participants living in the objective, least walkable neighborhoods, those in the most walkable quartiles had statistically significant higher SB levels (Q3: b = 13.18, 95% CI = 6.53, 19.83; Q4: b = 11.64, 95% CI = 4.95, 18.32). For subjective walkability, mean daily step count was significantly higher in the most walkable quartiles (Q3: b = 378.02, 95% CI = 126.76, 629.29; Q4: b = 509.60, 95% CI = 243.38, 775.81). Higher subjective walkability was positively associated with MVPA (Q3: b = 3.56, 95% CI = 1.82, 5.29; Q4: b = 4.40, 95% CI = 2.56, 6.23). CONCLUSIONS: Living in a neighborhood with higher objective walkability was associated with lower levels of PA and higher SB levels while higher subjective walkability was associated with higher levels of PA. These results show contrasting findings and thus, the effect of walkability on participant physical activity and sedentary behavior within our sample is to be determined. Supported by the European Regional Development Fund via OP-Zuid, the Province of Limburg, the Dutch Ministry of Economic Affairs (grant 31O.041).