Skip to main content

Table 4 Logistic regression and Generalized Linear Model estimates for the association between neighborhood type, self-reported walkability and participation in and minutes of neighborhood-based recreational walking in the last 7-days

From: The associations between objectively-determined and self-reported urban form characteristics and neighborhood-based walking in adults

  Participation in walking for recreation Minutes of walking for recreation
(n = 1875) (among those walking ≥1 times/week; n = 1065)
  Model 1a Model 2a Model 1b Model 2b
OR (95% CI) OR (95% CI) Estimate (95% CI)* Estimate (95% CI)*
Objective neighborhood type     
Low walkable (LW) Ref. Ref. 158.55 (142.61, 174.50) 161.68 (145.07, 178.29)
Medium walkable (MW) 0.96 (0.77, 1.20) 0.90 (0.71, 1.13) 155.55 (138.03, 173.06) 157.31 (139.50, 175.11)
High walkable (HW) 0.88 (0.58, 1.32) 0.82 (0.54, 1.26) 149.11 (116.36, 181.87) 153.04 (118.19, 187.89)
Self-reported neighborhood characteristics     
Access to services   0.99 (0.89, 1.10)   -9.13 (-17.90, -0.36)
Physical barriers   1.03 (0.93, 1.14)   6.45 (-2.03, 14.93)
Street connectivity   1.00 (0.90, 1.12)   0.39 (-8.42, 9.19)
Pedestrian infrastructure   0.94 (0.85, 1.04)   -2.34 (-10.54, 5.86)
Neighborhood aesthetics   1.18 (1.06, 1.32)   2.57 (-6.09, 11.24)
Motor vehicle traffic safety   0.96 (0.86, 1.07)   -5.32 (-14.28, 3.65)
Safety from crime   0.96 (0.86, 1.08)   2.59 (-6.06, 11.24)
Recreational destination mix   1.09 (0.98, 1.21)   4.77 (-3.02, 12.56)
  1. Model 1a and 1b: adjusted for age, gender, education, home ownership, dependents, years lived in neighborhood, attitude towards walking and reasons for neighborhood choice (access to places supporting physical activity, access to services, sense of community and ease of driving).
  2. Model 2a and 2b: adjusted for model 1 and self-reported neighborhood walkability.
  3. No interaction terms were statistically significant.
  4. = p < .05.
  5. * = estimated marginal means are reported for objective neighborhood type; regression coefficients (β) are reported for all other variables.