Skip to main content

Table 2 Crime and built environment characteristics and their associations walking frequency/week in adults 25–65 years (n = 3,487)

From: Does walkable neighbourhood design influence the association between objective crime and walking?

Spatial variable

Mean (SD)

Relative change (CI)1

p

Individual models

   

Burglary (400 m)

5.72 (6.47)

1.086 (1.030-1.145)

0.002

Burglary (1600 m)

93.54 (80.60)

1.007 (1.003-1.011)

0.002

Personal crime in public space (400 m)

1.46 (5.75)

1.077 (1.013-1.145)

0.017

Personal crime in public space (1600 m)

32.43 (68.46)

1.010 (1.004-1.016)

0.001

Residential density (400 m)2

11.86 (27.65)

0.999 (1.000-1.001)

0.375

Residential density (1600 m)2

12.73 (8.27)

1.004 (1.000-1.009)

0.066

Street connectivity (400 m)3

61.96 (30.27)

1.000 (1.000-1.001)

0.454

Street connectivity (1600 m)3

56.85 (18.90)

1.002 (1.000-1.003)

0.078

Local destinations (400 m)4

3.94 (10.58)

1.004 (1.001-1.007)

0.015

Local destinations (1600 m)4

80.06 (104.52)

1.001 (1.000-1.001)

0.000

Hotels, pubs, clubs & restaurants (400 m)5

0.23 (0.88)

1.057 (1.016-1.100)

0.006

Hotels, pubs, clubs & restaurants (1600 m)5

1.55 (8.56)

1.008 (1.004-1.012)

0.000

  1. 1From negative binomial log-linear models and represents change in walking frequency per unit increase in the spatial variable, except for burglary/personal crime variables where they represent change per increase of 10/year. All models adjust for age, sex, marital status, education and IRSD. 2Residential density calculated as the ratio of residential dwellings to residential area in hectares. 3Street connectivity calculated as the ratio of three-way intersections (or more) to the service area. 4Local destinations calculated as the count of all retail and service destinations in the service area. 5Subset of local destinations that are likely to serve alcohol. Bold denotes p < 0.05.