The influence of specific behavioural techniques on walking and cycling outcomes remains unclear. The conclusions of the review are likely to be reliable and recommendations for better quality research are likely to be appropriate.

To assess what behaviour-change techniques are used in walking and cycling interventions targeted at adults and what characterises interventions that appear to be associated with changes in walking and cycling.

Eight databases – including MEDLINE and The Cochrane Library – were searched from 2007 to March 2011. These searches updated those of two previous systematic reviews that included studies published up to 2010. Search terms were reported. Searches were restricted to publications in English with adult study populations.

Randomised and non-randomised studies on the effect of individually targeted intervention materials to promote walking and cycling interventions in adults were eligible for inclusion. Interventions delivered at a population level (such as mass-media campaigns) were excluded. Interventions had to be compared to "standard care" or "no intervention".

Where reported, mean or median patient age ranged from 23 to 74 years and most participants were female. Interventions were delivered to a wide range of populations, generally in the community. Two-thirds of the studies promoted walking only; all others promoted both walking and cycling. Number, frequency and duration of interventions and behaviour-change techniques used varied widely. Interventions were delivered in various ways, notably via individualised marketing or group counselling. About one in five of the studies were conducted in England; nearly all of the studies conducted in England used individualised marketing. Most studies assessed total walking and/or cycling. Outcomes were generally gathered through self-report; a few studies used a pedometer as well or instead of self-report.

The authors did not state whether study selection was done in duplicate.

Studies were critically appraised according to a seven-item appraisal tool and classed as higher quality, medium and lower quality. The tool included randomisation, assessment of baseline comparability between groups and attrition rates.

The authors did not state how many reviewers were involved in assessing study quality.

It appeared that a single reviewer extracted outcomes data (effect sizes and statistical significance). Behaviour-change techniques used within each interventions were coded according to a classification tool by two reviewers independently, with disagreements resolved through consensus.

Studies were reported narratively and in tables, and were classed into one of three categories: studies that reported a statistically significant effect; studies that reported a non-statistically significant effect; and studies that did not report on the statistical significance of the effect. One-way between-group analyses of variance (ANOVA) were conducted to compare the frequency of behaviour-change techniques coded for each category, frequency of behaviour-change techniques coded according to study quality and a comparison of study quality with each category.

Forty-one studies were included (>29,392 participants, range 30 to 3,090). Where used, follow-up took place between three months and 10 years post intervention. Most were randomised controlled trials (RCT) or quasi-RCTs and the rest were cross-sectional studies. Study quality was considered generally good: most studies were rated as medium (56%) or higher (37%) quality overall. Twenty-two studies reported using randomisation methods and 31 had comparable arms at baseline or adjusted their analyses for baseline differences.

Twenty-one interventions found a statistically significant improvement on walking and/or cycling, 12 found no statistically significant change and 13 did not report the statistical significance of the effect. For interventions that reported statistically significant changes, “prompt self-monitoring of behaviour” and “prompt intention formation” behaviour-change techniques were used in 68% of the interventions. “Prompt intention formation” was among the most commonly coded techniques for interventions that reported a non-significant change. There was no evidence that any particular combination of behaviour-change techniques was associated with statistically significant changes in walking and cycling. Additional results of analyses of variance were reported.

The influence of specific behavioural techniques on walking and cycling outcomes remained unclear. There was a need for further exploration of the effectiveness of behaviour-change techniques on these outcomes.

The review questions and selection criteria were clear. A large number of bibliographic sources were consulted. Only publications in English were included so some studies may have been missed. The authors reported that only part of the data extraction was conducted in duplicate; the risk of reviewer error and bias in the other stages of the review process could not be excluded.

The reviewers classed study quality mostly as high or moderate but the evidence had several important limitations. Notably, only about half of the studies reported using randomisation methods and most studies provided only self-reported outcomes. Several studies (notably from England) used cross-sectional designs, which have significant limitations, and did not report on the statistical significance of their effect estimates. This made their results difficult to interpret. The decision not to conduct a meta-analysis appeared appropriate, given the differences between the studies. However, the lack of quantitative synthesis limited the strengths of the results (as acknowledged by the authors).

The conclusions of the review are likely to be reliable and recommendations for better quality research are likely to be appropriate.

Practice: The authors did not state any implications for practice.

Research: The authors stated that future studies of walking and cycling interventions should ensure that all aspects of intervention design are reported in detail, using standardized vocabulary and guidelines when possible. They stated that further exploration of the effectiveness of behaviour-change techniques used in walking and cycling interventions (particularly “prompt self-monitoring of behaviour” and “prompt intention formation”) would be desirable. They stated that the interaction between behaviour-change techniques content and study design characteristics should be further explored.

Engineering and Physical Sciences Research Council, UK.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.