Skip to main content

Table 4 Logistic regression analyses exploring the odds of remaining an active traveller

From: Individual, socio-cultural and environmental predictors of uptake and maintenance of active commuting in children: longitudinal results from the SPEEDY study

  

Minimally adjusted model OR

Maximally adjusted model OR

(95% CI)

p

(95% CI)

p

Child characteristics

Gender

    

  Female (reference: male)

0.96 (0.53, 1.72)

0.887

0.88 (0.45, 1.72)

0.716

Weight status

    

  Overweight/Obese (reference: normal)

0.62 (0.33, 1.19)

0.152

0.60 (0.29, 1.24)

0.165

Parent characteristics

Age parent/carer left full-time education

    

  < 16 years (reference: >16 years)

1.09 (0.61, 1.94)

0.765

n.e

 

Social environment for active travel

Children walk or cycle to non-school destination at least once a week (reference: no)

    

  Yes

1.44 (0.78, 2.68)

0.246

1.42 (0.69, 2.91)

0.338

Composite rules score (reference: no rules)*

    

  Mid (score = 1)

0.90 (0.10, 8.01)

0.481

n.e

 

  High (score = 2)

0.68 (0.09, 5.40)

   

Convenient to take the car (reference: yes)

    

  No

10.74 (5.03, 22.92)

0.001

5.43 (1.95, 15.13)

0.001

Around to take child to school (reference: no)

    

  Yes

1.08 (0.56, 2.07)

0.823

n.e

 

Parental and peer encouragement (reference: neither)

     

  Parental or peer encouragement for AT

0.87 (0.41, 1.85)

    

  Parental and peer encouragement for AT

1.14 (0.50, 2.59)

0.754

n.e

 

Physical environment for active commuting

Perceptions of the environment

    

Route safety (reference: low)

    

  High

2.05 (1.06, 3.98)

0.032

1.49 (0.66, 3.37)

0.334

Neighbourhood social cohesion score (reference: low)

    

  High

1.41 (0.80, 2.51)

0.238

1.73 (0.90, 3.33)

0.102

Neighbourhood environment score (reference: low)

    

  High

1.32 (0.74, 2.36)

0.345

n.e

 

Safe to walk/play alone in my neighbourhood during the day (reference: no)

    

Yes

1.14 (0.60, 2.17)

0.669

n.e

 
 

Objective neighbourhood environment

    

Urban rural status (reference: village and hamlet)

    

  Town and fringe

1.45 (0.69, 3.05)

   

  Urban

1.29 (0.63, 2.62)

0.486

n.e

 

Deprivation (reference: Least deprived)

    

  2nd quartile

0.50 (0.22, 1.18)

   

  3rd quartile

0.61 (0.24, 1.57)

   

  Most deprived

0.60 (0.23, 1.57)

0.473

n.e

 

Road density (reference: low)

    

  High

1.59 (0.89, 2.82)

0.116

1.15 (0.57, 2.33)

0.693

Street light density (reference: low)

    

  High

1.12 (0.55, 2.27)

0.762

n.e

 

Junction density (reference: low)

    

  High

1.00 (0.56, 1.78)

0.996

n.e

 

Effective walkable area (reference: low)

    

  High

0.99 (0.56, 1.76)

0.977

n.e

 

Land use mix (reference: low mix)

    

  High mix

1.26 (0.70, 2.26)

0.449

n.e

 

Objective route environment

    

Route length between home and school (reference: >2km)

    

  1-2km

2.46 (1.12, 5.43)

 

1.24 (0.46, 3.39)

 

  <1km

6.32 (2.89, 13.85)

0.001

2.80 (0.98, 7.96)

0.020

Main road en route (reference: yes)

    

  No

0.92 (0.46, 1.83)

0.814

n.e

 

Route to school completely within an urban area (reference: no)

    

  Yes

2.71 (1.48, 4.95)

0.001

1.12 (0.49, 2.55)

0.790

Streetlight density on route (reference: low <4%)

    

  High (<4%)

0.80 (0.45, 1.42)

0.439

n.e

 

Route length ratio (reference: low)

    

  High

0.85 (0.46, 1.56)

0.602

n.e

 

Objective school environment

    

Walk to school initiative (reference: no)

    

  Yes

1.23 (0.69, 2.21)

0.470

n.e

 

Walking provision (reference: low)

    

  High

1.41 (0.77, 2.58)

0.258

n.e

 

Cycling provision (reference: low)

    

  High

1.03 (0.57, 1.87)

0.911

n.e

 
  1. Reference group is those children who switched from active to passive modes of travel. Number of children included in the final model is 419, due to missing values in some explanatory variables. Bold font indicates variable significant at p < 0.05. Where one p-value is reported for several categories, it refers to a test for trend across the groups. n.e = not entered into model, n.i = not included in final model due to collinearity with other variables.