Skip to main content

Table 1 Previous studies on the associations between GIS-derived walkability and daily steps in adults

From: Associations between neighbourhood walkability and daily steps in adults: a systematic review and meta-analysis

       Overall Walking Utilitarian Walking
1st Author, Publication Date N Age Location Sampling Design Neighbourhood Walkability Measurement (cut-off for high vs. low walkability) Measurement Findings Associationa Difference in mean steps per day for people living in high versus low walkability neighbourhoods (95 % confidence interval) b Measurement Findings Associationa Difference in mean time walking for utilitarian purposes for people living in high versus low walkability neighbourhoods (95 % confidence interval) b
Kondo, 2009 112 30 to 69 Hagi City, Japan Sampling from high and low walkable neighbourhoods using a stratified random sampling method based on sex and 5-year age strata. GIS-derived walkability based on street connectivity, residential density, land use mix (not specified) Accelerometer High walkability: 9364 steps/day; SE 567 INC 1071 steps/day (95 % CI −399 to 2540) Min/day (IPAQ)c High walkability: 3.3 min/day; SE = 2.1 0 −5 min/day (95 % CI −10 to 1)
Low walkability: 8294 steps per day; SE 491   Low walkability: 8.0 min/day; SE = 2.0
Van Dyck, 2011 350 42.4 ± 13.2 Flanders, Belgium Sampling from high and low walkable neighbourhoods based on address list provided by the local government. Urban vs. rural neighbourhoods based on GIS-derived walkability based on street connectivity and population density (not specified) Pedometer High walkability: 9323 steps/day; SD 3473 INC 548 steps/day (95 % CI −230 to 1326) Min/week (NPAQ)c High walkability: 97.5 min/week; SD = 96.4 + 76 min/week (95 % CI 58 to 94)
Low walkability: 8775 steps per day; SD 3942   Low walkability: 21.9 min/week; SD = 72.3
Dygryn, 2010 70 20 to 64 Olomouc, Czech Republic Random selection of participants in city. Walkability was determined after inclusion into the study. GIS-derived walkability based on street connectivity, residential density, floor area ratio, land use mix (upper versus lower 5 deciles) Pedometer High walkability: 11318 steps/day; SD 4091 + 2088 steps/day (95 % CI 440 to 3736) n/a n/a n/a n/a
Low walkability: 9230 steps per day; SD 2554  
Van Dyck, 2009 120 20 to 65 Sint-Niklaas, Flanders (Belgium) Sampling from high and low walkable neighbourhoods. Letters of invitation sent to randomly selected people. Letters were followed up with house visits to recruit people. Two neighbourhoods with greatest contrast in GIS-derived walkability based on street connectivity and residential density (not specified) Pedometer High walkability: 9318 steps/day, SD 3055 + 1222 step/day (95 % CI 131 to 2313) Min/week (NPAQ)c High walkability: 104.33 min/week; SD = 95.1 + 82 min/week (95 % CI 53 to 110)
Low walkability: 8096 steps per day; SD 3044   Low walkability: 22.83 min/week; SD = 61.0
Robertson, 2012 76 27 to 66 Glasgow, Scotland Sampling of people from Glasgow who were low active and part of low socioeconomic groups. Advertisement for participation was made in public locations (e.g., shops). Walkability was determined after inclusion into the study. GIS-derived commercial and residential land use mix Pedometer A one-unit increase in land use mix (from no mix to a perfect mix) was associated with:    n/a n/a n/a n/a
1896 more steps/day (SE = 583) at 6-months post community intervention + 1896 steps/day(95 % CI 754 to 3038)
1260 more steps/day (SE = 622) at 12-months post community intervention + 1260 steps /day(95 % CI 40 to 2479)
Zhang, 2014 1,100 46 to 80 Shanghai, China Stratified random samples based on even distribution of community types. Selected households were sent letters of invitation. Walkability was determined after inclusion into the study. GIS-derived street connectivity Pedometer Living in a neighbourhood one-SD above the mean street connectivity was associated with accumulating 21 more steps/day (no variance estimates reported) Unknown based on reported information Confidence intervals around the linear regression estimate could not be calculated based on the information reported in the text. n/a n/a n/a n/a
  1. aPositive relationship (+); negative relationship (−); INC (inconclusive; more research is needed to better estimate this effect); 0 (no effect)
  2. b95 % confidence intervals were recalculated based on information reported in the original manuscripts (i.e., group sample sizes, standard deviations/standard errors, and/or p-values)
  3. cInternational Physical Activity Questionnaire (IPAQ); Dutch Version of the Neighbourhood Physical Activity Questionnaire (NPAQ)