Moving to a Highly Walkable Neighborhood and Incidence of Hypertension: A Propensity-Score Matched Cohort Study

Abstract

Background: The impact of moving to a neighborhood more conducive to utilitarian walking on the risk of incident hypertension is uncertain.

Objective: Our study aimed to examine the effect of moving to a highly walkable neighborhood on the risk of incident hypertension.

Methods: A population-based propensity-score matched cohort study design was used based on the Ontario population from the Canadian Community Health Survey (2001-2010). Participants were adults ≥ 20 years of age who moved from a low-walkability neighborhood (defined as any neighborhood with a Walk Score < 90) to either a high- (Walk Score ≥ 90) or another low-walkability neighborhood. The incidence of hypertension was assessed by linking the cohort to administrative health databases using a validated algorithm. Propensity-score matched Cox proportional hazard models were used. Annual health examination was used as a control event.

Results: Among the 1,057 propensity-score matched pairs there was a significantly lower risk of incident hypertension in the low to high vs. the low to low-walkability groups [hazard ratio = 0.46; 95% CI, 0.26, 0.81, p < 0.01]. The crude hypertension incidence rates were 18.0 per 1,000 person-years (95% CI: 11.6, 24.8) among the low- to low-walkability movers compared with 8.6 per 1,000 person-years (95% CI: 5.3, 12.7) among the low- to high-walkability movers (p < 0.001). There were no significant differences in the hazard of annual health examination between the two mover groups.

Conclusions: Moving to a highly walkable neighborhood was associated with a significantly lower risk of incident hypertension. Future research should assess whether specific attributes of walkable neighborhoods (e.g., amenities, density, land-use mix) may be driving this relationship.

Citation: Chiu M, Rezai MR, Maclagan LC, Austin PC, Shah BR, Redelmeier DA, Tu JV. 2016. Moving to a highly walkable neighborhood and incidence of hypertension: a propensity-score matched cohort study. Environ Health Perspect 124:754-760; http://dx.doi.org/10.1289/ehp.1510425.

Figure 1

Study flow diagram. Low- and high-walkability areas were defined as Walk Score of < 90 and ≥ 90, respectively.

Figure 2

Event-free survival for incident hypertension in a propensity score–matched cohort of participants who moved from low- to high-walkability areas vs. from low- to low-walkability areas. Low- and high-walkability areas were defined as Walk Score of < 90 and ≥ 90, respectively. The p-value tests the difference between the Kaplan–Meier survival curves using the log-rank test. All estimates were weighted by the survey sample weights and bootstrap methods were applied. The hazard ratios, 95% confidence intervals, and p-values were derived from a Cox proportional hazards model performed on the propensity score–matched study sample of 1,057 pairs of participants balanced on age, sex, income, education, marital status, urban/rural residence, immigrant status, race/ethnicity, smoking, diabetes, BMI, stress, leisure physical activity, alcohol consumption, fruit and vegetable consumption, and index year.

Figure 3

Kaplan–Meier survival curves for annual health examination in a propensity score–matched cohort of movers from low- to high-walkability areas vs. from low- to low-walkability areas. Low- and high-walkability areas were defined as Walk Score of < 90 and ≥ 90, respectively. Kaplan–Meier survival curves were weighted using survey weights. The p-values test the differences between the Kaplan–Meier survival curves using the log-rank test. The hazard ratios, 95% confidence intervals, and p-values were derived from Cox proportional hazards models performed on the propensity score–matched study sample, which was balanced on age, sex, income, education, marital status, urban residence, immigrant status, race/ethnicity, smoking, diabetes, BMI, stress, leisure physical activity, alcohol consumption, fruit and vegetable consumption, and index year.