Skip to main content

Table 1 Characteristics of the included interventions (n = 30)

From: Effectiveness of active school transport interventions: a systematic review and update

Author, Country Intervention and strategies Methods Effect on ASTa Effects on other outcomes Moderators and mediators
Buckley et al. 2013 [29] [fall event]
Idaho, USA
Designated AST day, aimed to encourage students and their families to practice AST on a specific day.
Strategies: encouragement only.
Design: Observational case study | pre-post (during, 1 day after) Duration: 1 day.
Sample: 2 primary schools.
AST measure: observation counts (trips to school only, walking and cycling combined).
Relative increase in AST (101%) on the day; remained high one day following. None examined. None examined.
Buckley et al. 2013 [29] [spring event]
Idaho, USA
Designated AST day, aimed to encourage students and their families to practice AST on a specific day.
Strategies: encouragement only.
Design: Quasi-experimental | pre-post (during, 1 day after, 2 weeks after) Duration: 1 day.
Sample: 3 primary schools (2 experimental, 1 control).
AST measure: observation counts (trips to school only, walking and cycling combined).
Increase in AST sustained at 2-week follow-up relative to the control school (χ2 = 11.6; p = 0.009). Parent escort increased by 333% on AST day (p < 0.001). Parent interviews suggested that the school journey is an opportunistic time to spend with their child. None examined.
Buliung et al. 2011 [30]
4 Canadian provinces: Ontario, Alberta, British Columbia, Nova Scotia
School Travel Planning, a school-specific intervention aimed to increase AST through a collaborative stakeholder approach.
Strategies engineering, education, enforcement and encouragement (varied between schools).
Design: Observational | pre-post (1 year).
Sample: 12 primary schools, 1489 parent questionnaires.
Duration: 1-year.
AST measure: student classroom survey + parent questionnaire (trips to/from school, all active modes combined).
Student reported data indicated a modest increase in AST at follow-up (from 43.8 to 45.9%). d = 0.05. P value is unavailable (G. Faulkner, personal communication). 13% of parents reported driving less as a result of the intervention.
According to parents the 3 most effective strategies were education, special events, and infrastructure improvements.
None examined.
Bungum et al. 2014 [31]
Las Vegas, USA
Designated AST day, aimed to encourage students and their families to practice AST on a specific day.
Strategies: encouragement only.
Design: Quasi-experiment | pre-2 post (during, 1-week) assessments Duration: 1-day event.
Sample: 2 schools (1 experimental and 1 control) | 1336 students ages 5-11.
AST measures: observation counts (trips to school only, all active modes combined).
Increase in the mode share of AST by 7.4 percentage points on the day of the event. AST was then significantly higher than in the control school (χ2 = 27.2; p < .001; d = 0.29). AST dropped to baseline rates at 1-week assessment. No effect on the number of motor vehicles observed around the schools. The increase in AST was larger in girls (χ2 = 13.5; p < .05) than boys (χ2 = 1.79; p = .056), but formal moderation analysis was not reported.
Christiansen et al. 2014 [32]
Denmark
A comprehensive school-based intervention to improve non-curricular PA through changes of the physical and school environment supported by educational activities. Intervention schools were asked to have a policy targeting AST, to offer cycling safety education.
Strategies: engineering, education, enforcement and policy.
Design: RCT | pre-post (2-year).
Sample: 14 schools (7 experimental and 7 control) | 1014 students | ages 11-13.
Duration: 2 years.
AST measure: travel diary (trips to/from school, all active modes combined).
The prevalence of AST increased from 87.8% to 88.8% in the experimental group and from 84.3% to 85.3% in the control group with no difference between groups (p = 0.30; d = 0.13). The intervention had no effect on perceived safety of the school route, parental encouragement of cycling and attitudes toward cycling. Note: improved cycling infrastructure was not implemented as planned due to limited funding. Students reporting an unsafe route to school at baseline were more likely to use AST at follow-up in the intervention group compared to students with an unsafe route in the control group (OR = 2.69; 95% CI = 1.20–6.07). No interactions for gender, parent encouragement, distance, walkability and baseline AST.
Coombes et al. 2016
United Kingdom
A technology-based intervention (Beat the Street) aimed to increase AST via incentive-motivated approaches.
Strategies: encouragement only.
Design: quasi-experimental | pre-2 post (7-weeks, 20-weeks) assessments. Duration: 9-week intervention.
Sample: 2 schools (1 intervention and 1 control) | 80 students | ages 8-10.
AST measure: travel diary (trips to/from school, all active modes combined).
At 7-week follow-up: no difference in AST between groups.
At 20-week follow-up: 10% increase in AST in the intervention group and 7% decrease in control group (p = 0.056).
No difference in accelerometer counts per minute, but there was a smaller decline in MVPA in the experimental group (−15.1 vs. -23.3 min/day; p = 0.020). Children who touched a Beat the Street box more often were more active (+3.5 min/day of MVPA for children who engaged in the intervention on the mean number of days, that is 14.5 days).
Crawford & Garrard, 2013 [34]
Victoria, Australia
[pilot schools]
The Ride2School Program, which consisted mostly of promotional activities with some infrastructure changes.
Strategies: education, encouragement, and engineering.
Design: Quasi-experimental mixed methods study | pre-post (~1 year).
Sample: 4 primary schools (2 control and 2 intervention) | participants were ages 10-13, but younger children were also counted in the observations.
Duration: 9-12 months.
AST measure: observation counts + student hands-up survey (trips to school only, all active modes combined).
Observation results show that AST increased significantly in the inner suburban pilot school and the outer suburban control school. Hands-up surveys show that AST increased in the inner suburban pilot school and no changes were found in the other schools. Qualitative data suggest that the program was easier to implement within a school that was smaller, more established, with a culture that was accepting and enthusiastic about AST, in an area of higher density and lower car use, with more infrastructure improvements and a more “hands-on” approach from the Coordinator. Results differed by level of urbanization (see “effects on AST” column).
Crawford & Garrard, 2013 [34]
Victoria, Australia
[program schools]
The Ride2School Program, which consisted mostly of promotional activities, without infrastructure changes.
Strategies: education and encouragement.
Design: Observational mixed methods study | pre-post (6 months).
Sample: 13 primary schools.
Duration: 6 months.
AST measure: student hands-up survey and parent survey (trips to school only).
Parent surveys: increase in cycling (from 13.9 to 15.9%) and decrease in scooter or skateboard use (from 6.2 to 5.0%). The proportion of parents reporting ≥1 active trips increased (adjusted OR = 1.67; 95% CI = 1.04-2.68). Child surveys: decrease in scooter or skateboard use (from 7.2 to 4.9%) and a decrease in AST (from 51.1 to 48.7%). The latter was no longer significant after adjustment. d = 0.04 for parent report and −0.06 for child report. Qualitative data suggest that program implementation varied between schools and that the program was more effective when school communities were highly motivated, when secure bike storage facilities were offered, when all active modes were promoted equally by dynamic school staff. None examined.
Ducheyne et al. 2014 [35]
Belgium
Cycling training, aimed to increase cycling skills and encourage uptake of cycling.
Strategies: education only.
Design: RCT | pre-post (1-week, 5-month) post assessments.
Sample: 3 primary schools (2 distinct experimental groups and 1 control group) | 94 students | age 9.3 ± 0.5.
Duration: 4 sessions (45 min each).
AST measure: parent reported (to/from school, time spent cycling only).
Changes in weekly time spent cycling did not differ between intervention and control group (F = 1.9; p > 0.05) Effect sizes: intervention vs. control group: d = 0.46; intervention + parent vs. control: d = 0.03. Children’s cycling skill score increased significantly more in the intervention group from pre to post (F = 16.9; p < 0.001) and from pre to 5-months follow-up (F = 16.8; p < 0.001) compared to the control group. No intervention effects for parental attitudes. None examined.
Goodman et al. 2016 [36]
United Kingdom
Bikeability, a national cycle training program for children and adults.
Strategies: education only.
Design: retrospective natural experiment.
Sample: 3336 children whose school either had offered (n = 2563) or had not yet offered Bikeability (n = 773) | age 10-11.
Duration N/A.
AST measure: parent-reported frequency of cycling (at least once a week vs. less), and whether cycling was children’s usual travel mode to school.
Children attending schools that had offered Bikeability were not more likely to cycle at least once a week (OR = 0.99; 95% CI = 0.89-1.10) and to cycle to school (OR = 0.73; 95% CI = 0.41-1.29).
Children who received Bikeability training were more likely to cycle at least once a week (OR = 1.26; 95% CI = 1.16-1.37).
Children attending schools that had offered Bikeability were much more likely to have completed the program (68% vs. 28%; p < 0.001). Children’s participation in Bikeability was identified as a potential mediator of the relationship between school exposure to Bikeability and cycling frequency; however, no main effect was observed.
Gutierrez et al. 2014 [37]
Miami, USA
Implementation of crossing guards & AST awareness campaign.
Strategies: education, equity.
Design: Quasi-experimental | pre-post.
Sample: 58 intersections near elementary schools (24 where a new crossing guard was present and 34 control).
AST measure: observation counts (to school only, walking and cycling examine separately and combined).
The number of pedestrian and cyclists did not change following the addition of crossing guards (p > 0.05; d = 0.03). Safety: increase in students’ use of supervised routes with a moderate effect size (partial η2 = 0.008). No changes in parental attitudes regarding AST safety. None examined.
Henderson et al. 2013
Atlanta, USA
Safe Routes to School, a comprehensive, federally-funded program in the US designed to increase AST through non- infrastructure and infrastructure strategies.
Strategies: education, encouragement, and engineering.
Design: Observational | pre-post.
Sample: one primary school (658 students).
Duration: 2 years.
AST measure: parent-reported (to/from school); all active modes combined.
The prevalence of AST increased from 18% to 42% in the morning trip (p < 0.0001; d = 0.66), and from 18% to 23% in the afternoon (NS; d = 0.17). Parental perception about the health benefits, perception that the school strongly encouraged AST, and enjoyment of walking/biking to school increased significantly (all p < 0.01). None examined.
Hinckson et al. 2011a
New Zealand
School Travel Planning, a school-specific intervention aimed to increase AST through a collaborative stakeholder and multi-strategic approach.
Strategies: engineering, education, enforcement, encouragement, and policy interventions.
Design: Observational | pre-post (1 to 2 years).
Sample: 33 primary schools | 13,631 students.
Duration: 1-2 years.
AST measure: student reported (to school only, all active modes combined).
AST increased by 5.9 ± 6.8%. d = 0.45. None examined. Larger increase in AST with longer follow-up period. Longer follow-up period, smaller school roll and higher pre-intervention rate of AST predicted higher rates of AST at follow-up.
Hinckson et al. 2011b [18]
New Zealand
School Travel Planning, a school-specific intervention aimed to increase AST through a collaborative stakeholder and multi-strategic approach.
Strategies: engineering, education, enforcement, encouragement, and policy interventions.
Design: Observational | pre-post (1-, 2- and 3-year).
Sample: 56 primary schools | 57,096 students | grades K-5.
Duration: 3 years.
AST measure: student reported (to school only, all active modes combined).
There was an increase in AST by the 3rd year of implementation (from 40.5 to 42.2%; OR = 2.65; 95% CI = 1.75-4.02). d after 1, 2, and 3 years = 0.17, 0.51 and 0.54 respectively. None examined. Larger increase in AST with longer follow-up period. The program was more effective in older students, in smaller schools and in the city of Auckland, but it was less effective in low SES schools (all p < 0.05).
Hoelscher et al. 2016 [39]
Texas, USA
Safe Routes to School, a comprehensive, federally-funded program in the US designed to increase AST through infrastructure and non- infrastructure strategies.
Strategies: education and encouragement; some schools received infrastructure improvements (engineering).
Design: Quasi-experimental | pre-post (3 year).
Sample: grade 4 students from 78 elementary schools. Schools were allocated to either of 3 conditions: “infrastructure” (n = 23), “non-infrastructure” (n = 21) or control group (n = 34).
Duration: 3 years.
AST measure: student reported (to/from school, all active modes combined).
Infrastructure and non-infrastructure schools had significantly higher rates of AST in the morning (p = 0.024 and 0.013 respectively) and non-infrastructure schools had significantly higher overall AST relative to control schools (p = 0.036). However, differences between groups attenuated over time. Students from both infrastructure and non-infrastructure schools had higher AST-related self-efficacy, and a similar finding was noted in infrastructure schools for parents. Students in non-infrastructure schools reported engaging in PA on more days than students from comparison schools. Parents from all types of schools perceived worse walkability and bikeability in their neighborhoods and schools over time. None examined.
Hunter et al. 2015 [40]
London, England
Reading, England
Vancouver, Canada
International school competition, aimed to increase AST via incentive-motivated approaches.
Strategies: encouragement only.
Design: Observational Mixed-Methods (4-week).
Sample: 12 primary and secondary schools | 3817 Students | 9-13 year olds.
Duration: 4 weeks.
AST measure: Objective swipe card technology and child reports (to/from school, walking only).
The percentage of walking trips measured by the swipe card decreased over the 4-week measurement period from 29 to 12%. However, at baseline 77% of children stated that they walked to school at least once in the past week and this proportion was 86% at follow-up. d = −0.61 with swipecard methodology and 0.34 with self-report. Children’s attitudes: perceived the intervention to help physical and mental health. Adult attitudes: 91% of parents and 72% of teachers surveyed stated that they thought the competition had encouraged children to spend more time walking with their friends. This was corroborated with focus groups data. None examined.
Johnson et al. 2016 [41]
England
[Bikeability school travel survey]
Bikeability, a national cycle training program for children and adults.
Strategies: education only.
Design: retrospective case-control analysis.
Sample: 1345 year 5 and 6 students.
Duration: N/A.
AST measure: frequency of cycling to school as well as cycling in general (child report).
Students who received Bikeability were more likely to cycle to school (OR = 2.25; 95% CI = 1.83-3.52). d = 0.45. Students who received Bikeability did not report more cycling in general (OR = 1.01; 95% CI = 0.75-1.38). Year 6 students who received Bikeability expressed greater confidence (OR = 1.81; 95% CI = 1.26-2.59). None examined
Johnson et al. 2016 [41]
England
[CensusAtSchool]
Bikeability, a national cycle training program for children and adults.
Strategies: education only.
Design: retrospective case-control analysis.
1745 year 7-9 students. Duration: N/A.
AST measure: frequency of cycling to school as well as cycling in general (child report).
Students who received Bikeability were more likely to cycle to school (OR = 1.60; 95% CI = 1.17-2.21). d = 0.26. Students who received Bikeability were more likely to report cycling ≥30 min in the past week (OR = 1.27; 95% CI = 1.07-1.51). None examined.
Mammen et al. 2014a [19]
Canada
School Travel Planning, a school-specific intervention aimed to increase AST through a collaborative stakeholder and multi-strategic approach.
Strategies (varied across schools): education, encouragement, engineering, enforcement.
Design: retrospective analysis (1-year following implementation).
Sample: 53 primary schools, 7827 questionnaires.
Duration: 1 year.
AST measure: parent questionnaire (to/from school, all active modes combined).
17% of the parents reported driving less as a result of the intervention. Of these, about 83% reported switching from driving to AST. No baseline data available and no hypothesis test performed. None examined. Parents of older students, those living <3 km away from school, attending urban and suburban schools, and attending medium-SES schools were more likely to report less driving.
Mammen et al. 2014b [64]
Canada
School Travel Planning, a school-specific intervention aimed to increase AST through a collaborative stakeholder and multi-strategic approach.
Strategies (varied across schools): education, encouragement, engineering, enforcement.
Design: Observational | pre-post (1-year).
Sample: 53 primary schools.
Duration: 1 year.
AST measure: Student reported hands-up survey (to/from school, all active modes combined).
Baseline and follow-up data showed that 27% and 31% of children engaged in AST to and from school, with no intervention effect. d = −0.02 for morning trip and 0.01 for afternoon trip. Changes in AST ranged from a 26% decrease and a 23% increase across schools. None examined. Schools that collected baseline data in the Fall (i.e., September) and follow-up data in Winter (i.e., February) observed a 5% decrease in AST (B = −5.36, p < .05).
McDonald et al. 2013 [42]
Oregon, USA
Safe Routes to School, a comprehensive, federally-funded program in the US designed to increase AST through non- infrastructure and infrastructure strategies.
Strategies: engineering, education, encouragement, and enforcement (varied between schools).
Design: Quasi-experimental | pre-post.
Sample: 9 primary schools and 5 middle schools (includes 9 experimental and 5 control schools).
Duration: up to 4 years.
AST measure: student reported (to/from school, walking and biking combined).
Education + encouragement were associated with increases in walking and biking by 2 and 5 percentage points respectively. Augmenting education programs with engineering improvements was associated with increases in walking and biking of 5-20 percentage points. None examined. More comprehensive programs were associated with greater increases in AST (see “effect on AST” column).
McDonald et al. 2014 [43]
Florida, Oregon, Texas, District of Columbia, USA
Safe Routes to School, a comprehensive, federally-funded program in the US designed to increase AST through non- infrastructure and infrastructure strategies.
Strategies: engineering, education, encouragement, and enforcement (varied between schools).
Design: Quasi-experimental | pre-post (5-year).
Sample: 801 schools of which 83% were elementary schools (includes 378 experimental and 423 control schools) | 65,289 students.
Duration: 5 years.
AST measure: student reported (to/from school, walking and biking combined).
Relative to control schools, each year of participation in SRTS was associated with a 1.1% increase in AST (p = 0.002; d = 0.019). Engineering improvements led to a 3.3 percentage point increase in walking and biking (p = 0.031; d = 0.12), while education + encouragement interventions led to a 0.9 percentage point increase per year (p = 0.025; d = 0.15). None examined. More comprehensive programs were associated with greater increases in AST (see “effect on AST” column).
McMinn et al. 2012 [44]
Glasgow, Scotland
Travelling Green, a 6-week school based intervention aimed to increase AST via teacher lesson plans and student packs (e.g., material to set walking goals, record behavior).
Strategies: education and encouragement.
Design: Quasi-experimental | pre-post.
Sample: 5 primary schools (2 experimental and 1 control) | 166 students | ages 8-9.
Duration: 6 weeks.
AST Measure: Step counts and MVPA measured by Actigraph accelerometers during the trips to and from school.
Intervention group had smaller decreases in mean steps (−47 vs. -205) and seconds of MVPA (−33 vs. -85) during the morning trip. Opposite results on the afternoon trip for steps (−222 vs. -120) and MVPA (−125 vs. -59). d < 0.1 for changes in steps and MVPA during the school trip. Children who received the intervention showed a smaller decline in daily step counts (−901 vs. -2528; d = 0.52) and time spent engaging in MVPA (−429 vs. -1171 s; d = 0.46). None examined.
Mendoza et al. 2011 [45]
Texas, USA
Walking School Bus, aimed to increase AST by having children walking in adult-supervised groups.
Strategies: encouragement only.
Design: RCT | data collected before and in 4th & 5th week of intervention.
Sample: 8 primary schools (4 experimental and 4 control) | 149 students | average age = 10 years.
Duration: 5 weeks.
AST measure: Student reported (to/from school, all active modes combined).
In the intention-to-treat analyses, intervention children increased their weekly percent AST from 23.8% ± 9.2% at baseline to 54.0% ± 9.2% at follow-up, whereas control children decreased their weekly percent AST from 40.2% ± 8.9% to 32.6% ± 8.9% (p < .0001; d = 0.40). Intervention children increased their MVPA from 46.6 ± 4.5 to 48.8 ± 4.5 min/day while controls decreased theirs from 46.1 ± 4.3 to 41.3 ± 4.3 min/day (p = .029; d = 0.18). Acculturation (p = .014) and parent outcome expectations (p = .025) were both associated with increased AST. Parent self-efficacy was positively associated with AST (r = 0.182; p = .032).
Østergaard et al. 2015 [46]
Denmark
Safe and Secure Cycling to school, aimed to increase cycling behaviors through a multicomponent cycling promotion program.
Strategies: encouragement, enforcement, education and engineering.
Design: Quasi-experiment | pre-post (1-year) assessments.
Sample: 25 primary schools (13 experimental and 12 control) | 2401 students | 4th & 5th grade students.
Duration: 1 year.
AST Measure: Student reported number of cycling trips to/from school in the past week (range = 0-10).
Change in the number of cycling trips to/from school were not significant (B = 0.15 trips; 95% CI = −0.25; 0.54). d = 0.02. Cardiorespiratory fitness decreased in the intervention group relative to the control group (B = −1.45 ml O2·kg−1·min−1; p < 0.0001). No change in recreational cycling, overall PA, BMI and obesity. None examined.
Sayers et al. 2012 [47]
Columbia, USA
Walking School Bus, aimed to increase AST by having children walking in adult-supervised groups.
Strategies: encouragement only.
Design: Case control analysis where the researchers compared accelerometry-measured PA between WSB participants and non-participants.
Sample: 3 primary schools | 77 students | ages 8-9.
Duration: 1 week.
None. Percentage of time spent in MVPA did not differ between WSB participants and controls (all p ≥ 0.17). d = −0.32. The age-related gradient in MVPA was attenuated in WSB participants. None examined.
Stewart et al. 2014 [48]
Florida, Mississippi, Washington, Wisconsin, USA
Safe Routes to School, a comprehensive, federally-funded program in the US designed to increase AST through non- infrastructure and infrastructure strategies.
Strategies: engineering, education, encouragement, and enforcement (varied between schools).
Design: Observational | pre-post.
Sample: 53 primary schools with SRTS projects.
Duration: 5 years.
AST measure: student hands-up survey or observation counts (travel to school only). Assessed walking and biking separately and combined.
At the school level, AST increased from 12.8% to 19.8% (p < .001; d = 0.50); walking from 8.8% to 13.3% (p < .001; d = 0.46); cycling from 2.0% to 3.2% (p = 0.085; d = 0.32). None examined. Smaller changes in cycling in schools that had higher levels of cycling at baseline (r = −0.40; p = 0.009). Were not associated with changes in AST: funding, number of students/schools per project, project type, intervention strategies, school level, enrollment, % of students eligible for free/reduced cost meals, and characteristics of the school neighborhood.
Vanwolleghem et al. 2014 [49]
West-Flanders, Belgium
A drop-off spot (500-800 m distance from school) was organized that included teacher supervision on the walk to/from the designated area.
Strategies: encouragement only.
Design: Observational | data collected before and during intervention.
Sample: 2 primary schools | 58 students | ages 6-12 (mean = 9.7 ± 1.6 years).
Length: 1 week.
AST measure: child reported the number of active trips using the drop-off spot in a diary.
The number of reported active trips per week increased from 1 to 3 (χ2 = 52.9; p < 0.001; d = 1.00). Pedometer-determined step counts before/after school hours increased significantly (+732 step counts/day; χ2 = 12.2; p < 0.001), but not daily step counts (p = 0.16). Positive perception of the intervention by principals and parents, but teachers expressed doubts about future implementation. None examined.
Villa-González et al. 2016 [50]
Spain
Intervention aimed in increasing AST through changing children safety perceptions and attitudes.
Strategies: education and encouragement.
Design: Quasi-experimental | pre-post (6 months).
Sample: 5 primary schools (3 experimental and 2 controls) | 206 students | ages 8-11.
Length: 6 months.
AST measure: student-reported number of walking and cycling trips in the previous week (range = 0-10 trips).
Increase in the frequency of active trips in intervention schools (0.6 ± 0.2) relative to control schools (−0.4 ± 0.3) [p = 0.019; d = 0.40]. When examining travel modes separately, significant changes were only observed for walking. None examined. None examined.
Xu et al. 2015 [51]
China
Click Obesity Study, a multicomponent lifestyle childhood obesity prevention program aimed to enhance lifestyle behaviors.
Strategies (for lifestyle behavior change in general): education, encouragement.
Design: RCT | pre-post (1 year).
Sample: 8 primary schools (4 intervention and 4 control) | 1182 students | 4th grade.
Length: 1 year.
AST measure: Student reported school travel mode (to/from school, walking and cycling combined).
Participants in intervention schools were more likely to change their travel mode to walking or cycling to school (OR = 2.24, 95% CI = 1.47-3.40; d = 0.45) relative to those in the control schools. Intervention participants were more likely to show a ≥ 0.5 kg/m2 decrease in BMI (OR = 1.44, 95% CI = 1.10-1.87), to increase the frequency of jogging or running (OR = 1.55, 95% CI = 1.18-2.02), and to decrease TV/computer use (OR = 1.41, 95% CI = 1.09-1.84) and red meat consumption (OR = 1.50, 95% CI = 1.15-1.95). None examined.
  1. Characteristics are reported at the intervention level because some papers reported the findings of two interventions
  2. AST active school transportation, MVPA moderate-to-vigorous physical activity, NS non-significant, OR odds ratio, PA physical activity, SES socio-economic status, SRTS Safe Routs to School, STP school travel planning, WSB walking school bus
  3. aDetails on the calculation of standardized effect sizes (Cohen’s d) are provided in Additional file 2