Exercise and Depression: A Review of Reviews
- 10k Downloads
There has been considerable research interest in the effects of exercise upon depression outcomes. Recently, health agencies in the United Kingdom (UK) and beyond have made several guidance statements on this issue. Therefore, this review seeks to provide a synthesis of evidence regarding the effectiveness of exercise in the management of depression (including postnatal depression) in adults. Studies were identified by searching PubMed, Medline, Cochrane Library (CENTRAL) and PsychINFO using relevant search terms. The article describes how meta-analyses from peer reviewed journals have reported exercise as treatment for depression is more effective than no treatment, as effective as traditional interventions in some instances, possibly a promising approach to postnatal depression and has equivalent adherence rates to medication. However, reviews have also raised concerns about the methodological quality of trials, possible overestimation of treatment effects and lack of data regarding long term benefits. Based on the available evidence it is concluded that while awaiting further high quality trial evidence it would seem appropriate for exercise to be recommended in combination with other treatments.
KeywordsExercise Depression Postnatal depression
Chief Medical Officers’ recommendations for physical activity to achieve health benefits
• For general health, adults should achieve a total of at least 30 min a day, of at least moderate intensity physical activity on five or more days of the week.
• For many people it is likely that 45–60 min of moderate intensity physical activity a day is necessary to prevent obesity.
• Children and young people should achieve a total of at least 60 min of a least moderate intensity physical activity each day. At least twice a week this should include activities to improve bone health, muscle strength and flexibility.
• The recommended levels of physical activity can be achieved either by doing all the daily activity in one session, or through several shorter bouts of activity of 10 min or more. The activity can be lifestyle activity performed as part of everyday life (e.g. climbing stairs or brisk walking) or structured exercise or sport, or a combination of these.
Search Strategy for Identification of Studies
Systematic reviews and meta-analyses of observational studies and randomized controlled trials published between January 1990 and November 2007 were eligible for inclusion. Only English language reports that involved adult human populations were included. Relevant studies were identified by searching PubMed, Medline, Cochrane Library (CENTRAL) and PsychINFO using relevant search terms. Terms used in the search were depression, clinical depression, depressive symptoms, exercise, physical activity, meta-analyses and systematic review. Recent guidance from health agencies were also consulted in the development of this report.
Review of the Literature
Available Evidence Regarding Depression and Exercise in General Populations
Early meta-analyses (McDonald & Hodgson, 1991; North, McCullagh & Tran, 1990) reported moderate to large effects for exercise on depression. However, these meta-analyses included non-RCTs and studies where non-depressed populations were recruited and consequently these reviews should be interpreted with caution. Craft and Landers (1998) attempted to provide more precise answers regarding the relationship between depression and exercise by only including studies where individuals had been diagnosed with clinical depression and co-morbid depression with another mental illness (e.g. schizophrenia, anxiety and paranoia) in their meta-analysis. Thirty studies were included, many of which were unpublished dissertations. Analyses showed that exercise was associated with reduced depression (Effect Size = −0.72) but the inclusion of observational studies and non-RCTs again limits the conclusions from this review. Additional analyses showed that neither exercise program characteristics (e.g. duration, intensity, frequency and mode of exercise) nor participants’ characteristics (e.g. age, gender and severity of depression) moderated the effects. Specific details regarding the findings of these earlier meta-analyses can be found in each respective report.
Other reviews and meta-analyses have included only RCTs. Specifically Lawlor and Hopker (2001) included only RCTs where participants had been diagnosed with depression. Based on 14 included studies, an overall mean effect size of −1.1 for exercise interventions relative to no treatment was reported. Despite this large effect, the authors concluded that the effectiveness of exercise could not be determined because most trials were of poor quality with inadequate follow-up. Another recent meta-analysis (Stathopoulou, Powers, Berry, Smiths, & Otto, 2006) included four RCTs that were not available at the time of previous meta-analyses. Studies that did not target clinical levels of depression, were not published in peer-reviewed journals or did not include a non-active comparison group were excluded from the review. The subsequent analyses involving 11 RCTs yielded a very large treatment effect in favor of exercise compared to control conditions (Effect Size = 1.42). There was no significant relationship between sample size and effect, or publication year and sample size. There was, however, a significant association between publication year and effect size, with more recent studies associated with larger effects.
A recent systematic review of RCTs (Sjosten & Kivela, 2006) which focused specifically on older people (defined as over 60 years or defined as ‘older’ or ‘elderly’ within the included studies) concluded that exercise interventions may be efficient in reducing depression or high levels of depressive symptoms. The authors of this report commented that exercise might be effective in the short-term but that effects tended to reduce over time in studies reporting follow-ups beyond the length of the intervention. Furthermore, the non-depressed older adults benefited less than older depressed adults. Similar to previous authors, concerns were raised in this review about the methodological quality and heterogeneity of included trials. The main methodological concerns raised were related to the lack of allocation concealment in trials and the omission of intention to treat analyses.
Additional and Ongoing Trials
Searches also located four RCTs that have examined the effects of exercise on depression, which post date all previous meta-analyses and systematic reviews. Knubben et al. (2007) randomized patients with a moderate to severe depression undergoing standard clinical antidepressant drug treatment to walking or placebo low intensity stretching and relaxation exercise groups. After 10 days, reduction in depression scores in the walking group was significantly larger than in the placebo group. In addition, the proportion of patients with a clinical response was greater for the walking group. Brenes et al. (2007) conducted a pilot RCT and found exercise significantly reduced minor depression relative to sertraline in older adults. Blumenthal et al. (2007) randomized patients who had been diagnosed with major depression to one of four conditions; supervised exercise in a group setting, home based exercise, standard antidepressant treatment (sertraline) or placebo pill for 16 weeks. At 4-month follow-up patients receiving active treatments tended to have higher remission rates than the placebo controls. All treatment groups had lower Hamilton Depression Rating Scale (Hamilton, 1960) scores after treatment, but scores were not significantly different from the placebo group, suggesting that patient’s beliefs and expectations about treatment can be substantial. Legrand and Heuze (2007) assessed the effect of an eight week aerobic exercise program where participants with elevated levels of depression (n = 23) were randomized to low frequency exercise (one aerobic session per week and deemed the comparator), individualized high frequency exercise (3–5 aerobic exercise sessions per week) or group based high frequency exercise interventions. Participants randomized to the high frequency exercise interventions reported lower depression scores than those assigned to the low frequency exercise intervention at 8-week follow-up, highlighting that the exercise and depression relationship may be dose dependent. Furthermore, there were no differences in depression scores between participants in the individual and group based high frequency exercise groups, perhaps suggesting that the additional effects of exercising in a group are not important in the relationship between exercise and depression.
Registered ongoing RCTs of exercise in the management of depression and postnatal depression
Title: A pragmatic randomised controlled trial to evaluate exercise as a treatment for depression (TREAD) (UK) (National Research Register publication ID: N0632188872)
Principal research questions: Does exercise, in addition to usual care in primary health care, change the outcome for depression and alter subsequent use of antidepressant medication
Methodology and sample: Patients (N = 780) randomized to exercise plus usual care or usual care: Patients aged 16–69 years who have just started on antidepressant medication for depression and depressed patients who are not currently on antidepressant medication but who wish to pursue further treatment
Title: Treatment with exercise augmentation for depression (TREAD) (USA) (ClinicalTrials.gov Identifier: NCT00076258)
Principal research questions: Assess the relative efficacy of two doses of aerobic exercise (high or low frequency) to augment (adjunct) selective serotonin reuptake inhibitors treatment of major depressive disorder
Methodology and sample: Patients (N = 122) who remain symptomatic following an adequate acute phase trial of antidepressant therapy are randomized to exercise or antidepressant treatment
Title: The feasibility of an exercise intervention in depressed postpartum women: A pilot randomised controlled trial. (National Research Register publication ID: N0222178615)
Principal research questions: The study aims to evaluate the feasibility and acceptability of a pram pushing exercise intervention in postpartum women reporting depressive symptomatology
Methodology and sample: Patients (N = 44) randomized to exercise plus usual care or usual care
Title: Decreasing health care utilization with alternative approaches for the treatment of depression: A randomised trial of exercise for postpartum depression (ClinicalTrials.gov Identifier: NCT00384943)
Principal research questions: Examine change in depressed mood and other health outcomes scores immediately following a 3-month intervention and at 3 and 6 months post-treatment
Methodology and sample: Patients (N = 90) randomized to exercise or no exercise groups
Clinical Significance of Treatment Effects
Craft and Perna (2004) attempted to translate the effect sizes from three meta-analyses (Craft & Landers, 1998; Lawlor & Hopker, 2001; North et al., 1990) of exercise and depression to a binominal effect size to allow interpretation in terms of clinical significance. According to the authors values from these meta-analyses (−0.72, −0.94 and −1.1) reflect an increase in success rate due to treatment (exercise) of between 67 and 74% respectively. If a 50% reduction in symptoms is taken as the cut-off for a therapeutic response, these scores are very encouraging.
Evidence Available Regarding Postnatal Depression
Searches did not reveal any published meta-analyses of the effects of exercise upon postnatal depression (PND). However, a recent comprehensive review (Daley, MacArthur, & Winter, 2007) identified two small RCTs (Armstrong & Edwards, 2003, 2004) of the effects of exercise (group pram walking) interventions upon PND. Both trials reported substantial reductions in depression as measured by the Edinburgh Postnatal Depression Scale (Cox, Holden & Sagovsky, 1987) in those participants randomized to receive an intervention involving exercise. Although the results from these trials are promising, the sample sizes were very small and they were conducted in the Gold Cost resort area in Australia, where the persistent warm weather conditions are likely to have positively influenced mothers’ motivation to exercise by pushing their child in a pram. Some caution is required therefore when applying these findings to other environments. Nevertheless, compliance to exercise interventions has been reported to be good (66–75%) in this population (Daley et al., 2007) and observational studies have demonstrated that postnatal women are likely to be receptive to the suggestion from health professionals to exercise as part of treatment (Currie & Develin, 2002; Wilkinson, Philips, Jackson, & Walker, 2003).
Given that PND can have adverse consequences for the mother and infant (Department of Health, 2004; Hay et al., 2001), some women are reluctant to take antidepressants after giving birth (Whitton, Warner, & Appleby, 1996), and after returning home new mothers may have limited access to their own psychological and health care interventions due to financial and time constraints, further evaluation of novel adjunctive interventions in the management of PND would appear worthwhile. Based on this limited evidence in women specifically with PND and evidence from depressed general populations, NICE (2007) recommended in their recent guidance on the management of antenatal and postnatal mental health that exercise should be considered as a treatment for women who develop mild or moderate depression during the postnatal period. It is now imperative that we investigate whether this clinical advice is justified. More trials are also needed because the demands of motherhood are sufficiently different to preclude direct imputation from studies involving general populations of depressed individuals.
Exercise Versus Conventional Treatment
Several studies have evaluated exercise against alternative treatments for depression, most notably psychotherapy and antidepressants; a summary of their findings is provided here. Exercise was not significantly different from psychotherapy, or other types of behavioral and pharmacological interventions in one meta-analysis (Craft & Landers, 1998). In another meta-analysis exercise was as effective as cognitive therapy (Lawlor & Hopker, 2001). Dunn, Trivedi, Kampert, and Clark (2005) also reported response and remission rates comparable to other depression treatments in participants randomized to receive an exercise intervention at the public health recommended dose. Very recently Blumenthal et al. (2007) found remission rates for patients randomized to receive a home based exercise intervention or medication (sertraline) were similar. These findings are clearly relevant because many patients would prefer not to take antidepressant medication and the high prevalence of depression (Ormel et al., 1994) means there is often limited availability of psychological interventions. Exercise is relatively side effect free, a comparatively cheap alternative, and can be performed at the convenience of the individual, in contrast to antidepressants or psychotherapy, which is reliant on a visit to a physician or therapist. Antidepressants are known to have a latency of several weeks before taking effect; exercise in contrast, has potential to provide immediate psychological benefits.
Additional Benefits of Exercise
Depressive illness is strongly associated with physical disease (Peveler, Carson, & Rodin, 2002) and research has reported that depressed patients often demonstrate reduced physical health status (Martinsen & Medhus, 1989). Thus, the rationale for exercise as an intervention may extend beyond the potential benefits for depression since exercise can also provide a range of physiological health benefits (e.g. cardiovascular risk reduction and weight management) that traditional interventions do not (Department of Health, 2004). Furthermore, some symptoms of depression remain despite antidepressant treatment, for example, fatigue and reduced cognitive function and evidence has shown that exercise can improve these symptoms (Eriksen & Bruusgaard, 2004; Etnier et al., 1997). Collectively, this may make exercise an ideal ‘all round’ treatment for depression.
Regarding postnatal depression specifically, studies have shown that childbearing can contribute to the development of physical diseases such as obesity (Heliovaara & Aromaa, 1981; Williamson, Kahn, & Byers, 1991). Related to this, maternal body weight and postnatal weight have been shown to predict psychological health following birth (Jenkin & Tiggermann, 1997; LaCoursiere, Baksh, Bloebaum, & Varner, 2006) and therefore exercise may have an additional role to play in the health and well-being of postnatal women.
Mechanisms of Action
Exercise as a treatment option for depression may not be immediately obvious to patients and they may want to know the process or processes by which exercise might improve their symptoms. Despite extensive research regarding the exercise and depression relationship the mechanism(s) of action are not well understood. However, a number of hypotheses can be offered. A brief overview of some of the more prominent explanations is included here and readers are directed to other publications for more comprehensive comment (Daley, 2002; Ernst, Olson, Pinel, Lam, & Christie, 2006).
Arguably the most well known explanation is the endorphin hypothesis (Steinberg & Sykes, 1985), which proposes that exercise is associated with release of endogenous opiates such as beta-endorphins and consequently improved mood and feelings of well-being result. Several studies have shown increases in plasma endorphins after exercise (Bortz, Angwin, & Mefford, 1981; Farrell, Gates, & Maksud, 1982) but it remains unclear whether such changes reflect brain chemistry and are directly linked to reductions in depression. Similarly, the monoamine hypothesis (Pierce, Kuppart, & Harry, 1979) has suggested that exercise results in an increase in the availability of brain neurotransmitters (dopamine, serotonin and noradrenaline) that typically are found to be reduced with depression. While studies (Dunn, Reigle, Youngstedt, Armstrong, & Dishman, 1996) involving animals have indicated that exercise increases levels of serotonin and norepinephrine in the brain, this relationship remains unclear in humans.
Some researchers (Bahrke & Morgan, 1978; Gleser & Mendelberg, 1990) have suggested that exercise can serve as a source of distraction or a ‘time out’ strategy from daily worries and depressing thoughts. Studies have documented that exercise is associated with self-esteem and self-efficacy enhancement in a range of populations (Fox, 2000; Taylor & Fox, 2005). Thus, it is possible that exercise influences depression through enhancement of self-esteem. People who are depressed often feel they lack control and feelings of mastery in their lives and do not feel they have skills to bring about change. Promoted in the appropriate way, exercise may provide patients with a sense of achievement, efficacy and self-determination in their recovery process. Exercise may be a form of behavioral activation, which is typically an important component of effective psychotherapy interventions for depression (Brosse, Sheets, Lett, & Blumenthal, 2002).
Promoting Exercise for Depression Management: Possible Problems and Implications
While the available evidence favors a positive relationship between exercise and depression and patients’ views about the usefulness of exercise as a treatment for mental ill-health have been positive (Martinsen & Medhus, 1989; Pelham & Campagna, 1991), exercise is not without its problems. We must therefore, remain pragmatic about the complexities surrounding exercise promotion in depressed patients. For exercise to be a successful treatment for depression, a great deal of energy, commitment and motivation will be required from patients, many of whom will lack the motivation to complete even routine everyday tasks. Dropout from treatment is a critical factor in determining treatment success and patients need to adhere to the exercise intervention to experience therapeutic benefit. Continued involvement in exercise after remission may also be an important prophylaxis in preventing relapse. While drop-out from exercise programs has been identified as a concern in depressed populations (Blumenthal et al., 1999; Doyne et al., 1987; Sing, Clements, & Fiatarone, 1997) at approximately 20% (Stathopoulou et al., 2006), encouragingly this rate is similar to, and in some cases, better than antidepressant medication to treat depression (MacGillivray et al., 2003). As with any type of treatment for depression, patients will require regular monitoring to ensure compliance and clinicians and health professionals will need to plan for maintenance. GP referral schemes and exercise on prescription programs can play an important role in this regard, and there is some evidence of their efficacy in facilitating exercise behavior change and improving health outcomes (Isaacs et al., 2007; Taylor, Doust, & Webborn, 1998).
• Department of Health. At least five a week: Evidence on the impact of physical activity and its relationship to health, 2004 (www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4080994)
• NICE. CG23 Depression: Management of depression in primary and secondary care (http://guidance.nice.org.uk/page.aspx?o=397933)
• NICE. CG45 Antenatal and postnatal mental health (http://guidance.nice.org.uk/CG45/niceguidance/pdf/English)
• Mental Health Foundation. Exercise referral and the treatment of mild-moderate depression: Information for GPs and health care practitioners, 2005 (www.mentalhealth.org)
• Mental Health Foundation. Up and running: Exercise therapy and the treatment of mild to moderate depression in primary care, 2005 (www.mentalhealth.org)
• Grant T (Ed). Physical activity and mental health. National consensus statements and guidelines for practice, London, HEA, 2000
• Department of Health. National Framework for Mental Health, The Stationary office, HMSO, 1999 (www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/Browsable/DH_4096400)
For women specifically with postnatal depression, complications such as childcare responsibilities, fatigue and breastfeeding routines may reduce their opportunities and enthusiasm for exercise (Daley et al., 2007). Therefore, any program that promotes exercise in this population will need to take into account these factors and provide alternative strategies and methods by which regular exercise participation can be achieved.
While the promotion of exercise in the primary care setting has evidence of being effective (Elley, Kerse, Arroll, & Robinson, 2003) this inevitably relies on health professionals being both convinced of its merits and having the correct information about effective dosage. Studies (Daley et al., in press; Faulkner & Biddle, 2001; McEntee & Halgin, 1996) have shown that many primary care health professionals are not knowledgeable about current recommendations for physical activity to achieve health benefits and mental health professionals do not routinely promote exercise as a treatment option with their patients.
Several meta-analyses of the effects of exercise on depression have been published over the last 16 years. Early reviews reported very positive results regarding the effects of exercise for depression management but these conclusions were based predominately on observational studies and poor quality controlled trials. Recent reviews that have utilized more stringent study quality inclusion criteria have reported similarly encouraging findings. However, the authors of reviews and meta-analyses have warned against taking their findings at face value because the methodological quality of trials is still not adequate to make any conclusive statement about the effectiveness of exercise as a treatment for clinical depression. Caution also needs to be used in interpreting findings because there is often considerable overlap between meta-analyses in terms of included studies. Despite these methodological concerns, reviews have reported that exercise as treatment for depression is more effective than no treatment and is as effective as traditional interventions, at least in the short term. Further, clarification of the exercise and depression relationship may be obtained when recently completed and ongoing trials are added to future meta-analyses.
It is also important to note that adherence to exercise interventions in depressed populations is comparable to medication trials. Additionally, initial findings from trials specifically involving women experiencing PND have been promising, but they have been very small, raising concerns about the potential for bias. While NICE has advised that health professionals should consider promoting exercise in the management of PND, a large trial is clearly required before women can be advised that we know exercise is an effective treatment for postnatal depression.
Given there are minimal side effects associated with participation, and the potential for other health gains exists, clinicians and health professionals should consider promoting exercise as a treatment option for depression, but be mindful of the methodological concerns raised by reviews. Whether exercise is more or less effective as a stand-alone or as an adjunctive treatment remains unclear. However, while awaiting further high quality trial evidence it would appear appropriate for exercise to be recommended in combination with other treatments.
- Craft, L. L., & Landers, D. M. (1998). The effects of exercise on clinical depression and depression resulting from mental illness: A meta-analysis. Journal of Sport and Exercise Psychology, 20, 339–357.Google Scholar
- Daley, A.J., Bassi, S., Haththotuwa, H., Hussain, T., Kalhan, M., & Rishi, S. (in press). “Doctor, how much physical activity should I be doing”? How knowledgeable are general practitioners about the Chief Medical Officer’s (2004) recommendations for active living to achieve health benefits? Public Health.Google Scholar
- Department of Health. (2004). Confidential enquiry into maternal and child health: Why mothers die 2000–2002. Royal College of Obstetricians and Gynaecologists. London: RCOG Press. www.cemach.org.uk/publications/WMD2000_2002/Content.htm. Accessed 11 Jan 2006.
- Department of Health. (2004). At least five a week: Evidence on the impact of physical activity and its relationship to health. A report from the Chief Medical Officer. http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4080994). Accessed 11 Jan 2006.
- Dunn, A. L., Reigle, T. G., Youngstedt, S. D., Armstrong, R. B., & Dishman, R. K. (1996). Brain norepinephrine and metabolites after treadmill training and wheel running in rats. Medicine and Science in Sport and Exercise, 28, 204–209.Google Scholar
- Etnier, J., Salazar, W., Landers, D., Petruzzello, S. J., Han, M., & Nowell, P. (1997). The influence of physical fitness and exercise upon cognitive functioning: A meta-analysis. Journal of Sport and Exercise Psychology, 19, 249–277.Google Scholar
- Fox, K. R. (2000). The effects of exercise on self-perceptions and self-esteem. In S. J. H. Biddle, K. R. Fox, & S. H. Boutcher (Eds.), Physical activity and psychological well-being (pp. 88–117). London: Routledge.Google Scholar
- Gleser, J., & Mendelberg, H. (1990). Exercise and sport in mental health: A review of the literature. Israel Journal of Psychiatry and Related Science, 27, 99–112.Google Scholar
- Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry, 23, 56–62.Google Scholar
- Isaacs, A. J., Critchley, J. A., Tai, S., Buckingham, K., Westley, D., Harridge, S. D., et al. (2007). Exercise Evaluation Randomised Trial (EXERT): A randomised trial comparing GP referral for leisure centre-based exercise, community-based walking and advice only. Health Technology Assessment, 11, 1–184.PubMedGoogle Scholar
- MacGillivray, S., Arnoll, B., Hatcher, S., Ogston S., Ried, I., Sullivan, F., et al. (2003). Efficacy and tolerability of selective serotonin reuptake inhibitors compared with tricyclic antidepressants in depression treated in primary care: Systematic review and meta-analysis. British Medical Journal, 326, 1014.PubMedCrossRefGoogle Scholar
- McDonald, D. G., & Hodgdon, J. A. (1991). Psychological effects of aerobic fitness training: Research and theory. New York: Springer-Verlag.Google Scholar
- National Institute for Clinical Excellence. (2004). CG23 Depression: Management of depression in primary and secondary care. London: National Health Service.Google Scholar
- National Institute for Health, Clinical Excellence. (2007). CG45 Antenatal and postnatal mental health. London: National Health Service.Google Scholar
- North, T. C., McCullagh, P., & Tran, Z. V. (1990). Effects of exercise on depression. Exercise and Sports Science Reviews, 18, 379–415.Google Scholar
- Ormel, J., von Korff, M., Ustun, T., Pini, S., Kortsen, A., & Oldehinkel, T. (1994). Common mental disorders and disability across cultures: Results from the WHO collaborative study on psychological problems in general health care. Journal of the American Medical Association, 272, 1741–1748.PubMedCrossRefGoogle Scholar
- Pelham, T., & Campagna, P. (1991). Benefits of exercise in psychiatric rehabilitation of persons with schizophrenia. Canadian Journal of Rehabilitation, 4, 159–168.Google Scholar
- Peveler, R., Carson, A., & Rodin, G. (2002). ABC of psychological medicine: Depression in medical patients. British Medical Journal, 149–152, 325.Google Scholar
- Sing, N. A., Clements, K. M., & Fiatarone, M. A. (1997). A randomized controlled trial of progressive resistance training in depressed elders. Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 52A, M27–M35.Google Scholar
- Taylor, A. H., Doust, J., & Webborn, N. (1998). GP referral randomised controlled trial to examine the effects of a GP exercise referral programme in Hailsham, East Sussex, on modifiable coronary heart disease risk factors. Journal of Epidemiology and Community Health, 52, 595–601.PubMedCrossRefGoogle Scholar