Home > DARE Reviews > Meta-analysis of internet-delivered...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet]. York (UK): Centre for Reviews and Dissemination (UK); 1995-.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet].

Meta-analysis of internet-delivered interventions to increase physical activity levels

CA Davies, JC Spence, C Vandelanotte, CM Caperchione, and WK Mummery.

Review published: 2012.

CRD summary

The review concluded that internet-delivered interventions were effective in producing small but significant increases in physical activity. Limitations of the review cast considerable doubt on the reliability of the authors' conclusions, which appear to be over-optimistic.

Authors' objectives

To evaluate the effectiveness of internet-delivered interventions to increase physical activity.


PubMed, ProQuest, Scopus, PsycINFO, CINAHL and SPORTDiscus were searched from January 1990 to June 2011 for studies published in English; search terms were reported. Reference lists of all relevant review articles were manually searched.

Study selection

Eligible studies were those with an internet-delivered (web page or email) intervention for adults (≥18 years) that targeted and measured physical activity as an outcome. Studies also had to include a comparison group that did not receive internet-delivered materials, and had to provide data to allow the calculation of effect sizes.

Most studies used interventions that incorporated both internet and email communication; the remainder used only one or the other. There was variation in comparison group treatments; some studies used alternate or minimal intervention and most studies used some form of standard care. Duration of intervention ranged from two to 52 weeks (mean 12.6 weeks). Most studies were of a general population, but some were of specific populations (overweight or with diabetes). The mean participant age was 43 years and 65% were female. Over half the studies were conducted in the USA; three were conducted in the UK.

The authors did not state how many reviewers selected studies.

Assessment of study quality

Study quality was evaluated based on the following criteria: study design, selection and specification of the study sample, specification of illness/conditions, reproducibility of the study, and outcomes specification and measurement.

Studies could receive a score up to 18 points. Each study's score was divided by 18 and multiplied by 100 to provide a Study Quality Percentage. Study Quality Percentages were then classified as good (66.7% or higher), fair (between 50 to 66.6%) and poor (less than 50%).

The authors did not state how many reviewers performed the assessments.

Data extraction

Data were extracted to calculate standardised mean differences with standard errors. Two reviewers independently extracted data with disagreements resolved by discussion.

Methods of synthesis

Meta-analyses were performed to calculate pooled standardised mean differences with standard errors, using a fixed-effect model. Heterogeneity was assessed using the Cochran Q test. Publication bias was assessed using Egger's test. A number of subgroups analyses were conducted.

Results of the review

Thirty-four randomised trials were included (9,638 participants were relevant for the comparisons of interest in this review). Twenty-four studies were rated as being of good quality and 10 were rated as fair. Attrition was reported in 28 studies with the overall rate being 23% in the intervention group. Fourteen studies conducted intention-to-treat analyses.

Internet-delivered interventions had a small but significantly greater effect on physical activity change than comparisons (SMD 0.14, SE 0.02; 34 studies). Statistically significant publication bias and heterogeneity were present. The overall mean effect for sustained physical activity at least six months post- intervention resulted in a small but significant benefit favouring internet-delivered interventions (SMD 0.11, p<0.01; 11 studies).

Post-intervention sample size, initial physical activity level, and use of educational components were found to be significant moderators of physical activity change.

Authors' conclusions

Internet-delivered interventions were effective in producing small but significant increases in physical activity.

CRD commentary

The review addressed a clear question and was supported by reproducible eligibility criteria. The restriction to searching only for studies published in English meant that some relevant studies may have been missed, and the review results appear to have been subject to publication bias. Duplicate processes were employed to reduce the risks of reviewer error and bias during data extraction, but the authors did not report on whether such methods were used to select studies or assess study quality.

Adequate primary study details were provided. Study quality was assessed, but it was unclear how well the criteria translated into an evaluation of whether individual trial results were reliable, as minimal method and result details were presented. It was therefore unclear how reliable the results of 'good' studies were.

The value of the pooled estimates appears questionable since a fixed-effect model was used, despite the presence of considerable heterogeneity. Appropriate methods were used to assess and investigate heterogeneity. However, statistically significant heterogeneity was present not only in the main analysis, but also in many of the subgroup analyses; there was also evidence that significant publication bias had affected the results. Furthermore, the authors noted uncertainty regarding whether the treatment effects found were clinically-relevant effects. Although some of these issues were discussed, they were not incorporated into the overall conclusions.

Limitations of the review cast considerable doubt on the reliability of the authors' conclusions, which appear over-optimistic.

Implications of the review for practice and research

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

Research: The authors stated a need for well-designed, adequately powered randomised studies which should ideally also evaluate cost-effectiveness and longer term outcomes. The studies should also aim to isolate the impact of specific intervention features on physical activity change, and focus on aspects which increase engagement and retention of the target population.


National Health and Medical Research Council of Australia; National Heart Foundation of Australia.

Bibliographic details

Davies CA, Spence JC, Vandelanotte C, Caperchione CM, Mummery WK. Meta-analysis of internet-delivered interventions to increase physical activity levels. International Journal of Behavioral Nutrition and Physical Activity 2012; 9:52. [PMC free article: PMC3464872] [PubMed: 22546283]

Indexing Status

Subject indexing assigned by NLM


Databases, Factual; Health Knowledge, Attitudes, Practice; Humans; Internet; Motor Activity; Randomized Controlled Trials as Topic; Sports



Database entry date


Record Status

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.

CRD has determined that this article meets the DARE scientific quality criteria for a systematic review.

Copyright © 2014 University of York.

PMID: 22546283