|Walking for depression or depressive symptoms: a systematic review and meta-analysis
|Robertson R, Robertson A, Jepson R, Maxwell M
The review concluded that walking significantly reduced depressive symptoms in some populations and was a promising treatment. The authors recommended that their conclusions should be treated with caution, which was appropriate due to the limited evidence and wide variation in participants, controls, interventions and outcome measurement.
To evaluate the effectiveness of walking for depression or depressive symptoms.
Eleven databases were searched, including MEDLINE, EMBASE, PsycINFO, Physical Education Index, PsycARTICLES, SPORTDiscus, Cochrane Central Register of Controlled Trials (CENTRAL) from inception to January 2012 for publications in any language; search terms were reported. And internet search was conducted. Bibliographies of relevant articles and the trials themselves were searched. Past and ongoing trials investigators were contacted for further trials.
Randomised controlled trials (RCTs) of a walking intervention as treatment in any setting for depression (as defined by the trials authors) in adults (aged 18 years or more) were eligible for inclusion. All categories of depression were included (including postpartum depression) except bipolar disorder. Trials of participants with other mental health disorders in addition to depression and trials of patients selected with specific medical conditions (such as cancer) were excluded.
Eligible interventions were any structured or semi-structured walking treatment, indoors or outdoors, either as a single treatment or in combination with other depression treatments. Trials where some participants in the intervention group carried out other activities such as higher-intensity jogging and trials where the intervention included activities other than walking (such as cycling or weight training) were excluded. Trials where interventions had minor additions to walking (such as warm up, cool down or other short periods of stretching) were included. A validated depression assessment had to be used for outcome measurement.
The included RCTs were in developed countries and globally well spread. A large study was added that included overweight (64%) or moderately depressed patients where not all the patients were depressed. The amount of supervision and whether walking was in a group varied. Most interventions were aimed at a specific exercise intensity. Average intervention length was 3.5 sessions over 6.2 days to six months; individual session length ranged from 20 to 50 minutes and the number of sessions ranged from daily to five per week. The most common controls were social contact or support, also usual care and (in individual studies) stretching and cognitive therapy. Participants and settings varied and included postpartum women, students, women only, patients in different age groups (range 20 to 70 years) and elderly patients (mean 72.5 years). Depression diagnosis details were reported for each study. Six different psychometric instruments were used to measure depression outcomes. Six trials objectively measured fitness level outcomes. Further details were reported.
Two independent reviewers performed the study selection. Differences in opinion were resolved by discussion with the other two reviewers.
Assessment of study quality
Two reviewers performed the quality assessment and each checked the other's input. Criteria included allocation concealment, intention to treat (ITT) analysis, sample size calculation and blinding of assessors.
Means with standard deviations (SD) were extracted for the intervention and control groups and used to calculate mean differences with 95% confidence intervals (CI). Two reviewers performed the data extraction and checked each other's input. Authors were contacted for further data. Some study data were extracted from an earlier review (Mead et al 2008, see Other Publications of Related Interest).
Methods of synthesis
Results were pooled to give standardised mean differences (SMD) with 95% CIs using a fixed-effect model where between-study heterogeneity was low and a random-effects model where there was significant heterogeneity. Between-study heterogeneity was determined using the Χ² and Ι² statistics. Subgroup analyses were performed for outdoor walking, indoor walking and walking in a group. Sensitivity analyses excluded the largest study (as it included patients with high body mass index but not necessarily depression) and lower quality studies.
Results of the review
Eight RCTs were identified (341 participants, range 11 to 127). One trial met all four quality criteria, one trial met three criteria, two trials met two criteria and four trials met only one criterion. A forest plot indicated no relationship between effect size and study quality. Completion rates varied from 75% to more than 90%.
Walking versus controls had a significant positive effect on symptoms of depression (SMD -0.86, 95% CI -1.12 to -0.61; Ι²=86%; fixed-effect model; eight RCTs). The effect remained significant but smaller when the lower quality studies were excluded (SMD -0.69, 95% CI -0.99 to -0.39; four RCTs) and increased when the largest study (which included some overweight and some depressed participants) was excluded (SMD -1.01, 95% CI -1.92 to -0.11; random-effects model; seven RCTs).
Subgroup analyses found the effect was also significant for walking outdoors (SMD -0.60, 95% CI -0.91 to -0.28; four RCTs), walking indoors (SMD -1.35, 95% CI -1.84 to -0.86; three RCTs) and walking as a group activity (SMD -0.60, 95% CI -0.96 to -0.24; two RCTs) versus controls.
Results were provided for fitness measures in individual studies.
Walking had a statistically significant large effect on the symptoms of depression in some populations but the evidence base from randomised controlled trials was limited. Further investigation of effective interventions was needed, especially in primary care.
The review addressed a well-defined question in terms of study design, participants, interventions and outcomes. The search performed was extensive. Study quality was assessed and suitable criteria were used. Study quality was generally not high. Efforts were made to reduce error and bias throughout the review process. Relevant study details were reported.
A fixed-effect model was used for the main meta-analysis, which seemed inappropriate in view of the high statistical heterogeneity. No statistical heterogeneity results were reported for the subgroup analyses. The number of participants in the included studies was low and an extra study was added which did not quite fit the inclusion criteria in order to increase the overall numbers of participants.
The authors recommended that their conclusions should be treated with some caution since the included trials were diverse for participants, controls, interventions and outcome measurement. Given this, and the other limitations of the evidence, caution seems appropriate as the evidence may not be reliable.
Implications of the review for practice and research
Practice: The authors suggested that walking might be a useful adjunct to other interventions for treating depression. Health professionals could refer patients to suitable exercise referral schemes where walking was an option. They pointed out that the available evidence was inadequate to support NICE recommendations for group exercise.
Research: The authors recommended that further investigation should establish the frequency, intensity, duration and types of effective walking interventions, particularly in primary care populations, in order to provide recommendations for clinical practice. Future research should clarify the optimum role for healthcare professionals in delivering walking interventions and which factors encouraged primary care patients to adhere to walking programmes. Research should determine whether recruitment factors affected success.
Robertson R, Robertson A, Jepson R, Maxwell M. Walking for depression or depressive symptoms: a systematic review and meta-analysis. Mental Health and Physical Activity 2012; 5(1): 66-75
Other publications of related interest
Mead GE, Morley W, Campbell P, Greig CA, McMurdo M, Lawlor DA. Exercise for depression. Cochrane Database of Systematic Reviews 2008
Subject indexing assigned by CRD
Depression; Depressive Disorder; Humans; Walking
Date bibliographic record published
Date abstract record published
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.