Exercise Interventions as the Primary Treatment for Depression: Evidence from a Narrative Review

There is an increasing evidence supporting the efficacy of exercise interventions in the treatment of depression, which is a growing global health concern. However, data on the efficacy of exercise as the primary treatment for depression are scarce. This narrative review explored the efficacy of exercise interventions as the primary treatment for depressive disorders. A comprehensive search for English-language literature published between January 1965 and November 2019 was conducted via PubMed, Google Scholar, Scopus, Web of Science, PsycINFO, EMBASE, Cochrane database and Medline. Thirteen randomised control trials (RCTs) were included in the final analysis. Their results indicated that supervised aerobic exercise and high-intensity progressive resistance training (PRT) were effective in ameliorating depressive symptoms as the primary treatment compared with control groups, but they were not superior to other active treatments, such as antidepressants and cognitive behavioural therapy. Aerobic exercise and high-intensity PRT may be a promising primary treatment for depression as they may induce biopsychosocial effects (effects on neurotrophic factor, pro-inflammatory cytokines, monoamine, the hypothalamic-pituitary-adrenal axis, self-efficacy, mastery experience, adaptive coping and social interaction), which may ameliorate the severity of depressive symptoms. However, future RCTs with more comprehensive and well-designed methodologies are warranted to confirm our findings.


Introduction
Depression has become a major health burden; it is a disorder that currently affects 264 million people worldwide (1). The rapid rise in the incidence of depression is well documented, with the number of cases increasing from 172 million in 1990 to 258 million in 2017 (2). The mainstay of treatment for depression is pharmacotherapy (antidepressants) and psychotherapy (particularly cognitive behavioural therapy [CBT]), but these treatments are costly and associated with adverse effects; moreover, they exhibit high dropout rates (3)(4)(5).
In addition, depressive disorder is one of the most prevalent psychiatric disorders associated with suicidal behaviour. Hence, depressive disorder should be adequately treated to prevent the increasing trend of suicidal behaviour, which results in a loss of valuable life (6).
As an alternative to conventional treatments, exercise intervention is gaining increasing attention as a treatment option for depression and it is now recommended by clinical guidelines for the treatment of depression (7,8). Long-term physical exercise has been shown to increase the levels of a To link to this article: https://doi.org/10.21315/mjms2020. 27.5.2 intervention' and 'depressive disorders'. A refined search was then carried out with additional keywords, such as 'low intensity exercise', 'moderate intensity exercise', 'high intensity exercise', 'minor depression', 'major depressive disorder', 'premenstrual dysphoric disorder', 'dysthymia', 'physical activity for depressive disorder', 'exercise intervention for depressive disorder', 'aerobic exercise', 'stretching exercise' and 'home-based exercise'. The search was conducted independently by all three authors, and the search findings were compared; discrepancies were resolved before the final selection of articles to be reviewed.

Inclusion Criteria
Studies were eligible for review if they met the following criteria: i) they were published in English-language, peer-reviewed journals; articles in-press were included in the search; ii) they involved randomised control trials (RCTs) that investigated the efficacy of any exercise intervention as a stand-alone intervention for treatment of depression; if patients were on an antidepressant, it had been used for < 2 weeks; and iii) they included subjects aged 18 years and above, of any gender, diagnosed with any depressive disorder, and with any severity of depressive symptoms confirmed by screening with validated diagnostic and rating tools for depressive disorders. Studies were excluded if they had the following characteristics: i) they were published in languages other than English; ii) they took the form of systematic reviews, meta-analyses, case reports and series, theses, letters to the editor and editorials or academic conference proceedings; iii) they included subjects diagnosed with mental disorders other than depressive disorders; and iv) they did not give sufficient information about the efficacy of the exercise intervention and diagnoses of the subjects.

Search Results
The initial database search of the study titles and abstracts yielded 1700 studies, but 1550 studies were removed as they did not involve exercise intervention or were duplicates. Careful screening of the abstracts of the remaining 150 studies led to the removal of another 90 studies, which included diagnoses number of trophic factors, such as brain-derived neurotrophic factor, neurotrophin-3, epidermal growth factor, vascular endothelial growth factor and glial cell line-derived neurotrophic factor. These factors promote angiogenesis, synaptogenesis and neurogenesis, which in turn lead to neuron survival, proliferation and maturation. Hence, physical exercise may lead to reduced depressive symptoms (9,10).
Several systematic reviews, meta-analyses and literature studies have pointed out the efficacy of exercise interventions in treating various subtypes of depressive disorders, such as major and minor depression and dysthymia. Such research has mostly considered exercise as an adjunctive therapy to primary treatment with antidepressants or CBT across a variety of age groups, from young adults to the elderly (18-above 65 years old) (11)(12)(13)(14)(15)(16). However, few reviews have focussed on providing an overview of the efficacy of exercise interventions as the primary treatment for depressive disorders (i.e. in patients who are not currently being treated with antidepressants and psychotherapy). Hence, the goal of this comprehensive narrative review of the literature was to explore the types of exercise intervention that show efficacy as the primary treatment for depressive disorders and how their efficacy fares relative to other active depression treatments. In this narrative review, we applied the methodological rigour of a systematic review; to accomplish this, we carried out the following steps: we focussed our review based on well-defined objectives; established clear inclusion and exclusion criteria for the electronic literature search; established relevant criteria for literature selection based on the eligibility criteria; and critically assessed the methodology, key results, interpretation of findings and limitations of the selected studies (17).

Study Characteristics
All 13 studies selected for review conducted an RCT. Table 1 summarises the study characteristics.

Exercise Interventions
One study employed a two-armed RCT that compared the efficacy of supervised aerobic exercise with a placebo (stretching exercise in which interaction with healthcare providers was similar to that in the intervention group) (20). Another three studies employed two-armed RCTs to compare moderate and high-intensity progressive resistance training (PRT) with a control group (no intervention, health education) (23,24,26). A three-armed RCT study compared high-intensity PRT with low-intensity PRT and of mental illnesses other than depressive disorders. The full-text articles of the remaining 60 studies were screened and 47 were excluded because they did not provide a clear explanation of the efficacy of exercise intervention for the treatment of depressive disorders or they did not clarify if participants were on any treatment for depression other than exercise interventions. Hence, 13 studies were used in this review. The database search and selection of studies are summarised in Figure 1, while Table 1 provides a summary of all the selected studies.

Limitations
Singh et al.

Limitations
Singh et al.

Limitations
Brenes et al.

Quality of Study Assessment
Regarding the randomisation process, only six studies employed the allocation concealment method (19-21, 24, 25, 30), while another four studies did not (18,(22)(23)29). Two studies did not report on the randomisation method (27,28) and one did not blind the participants (26). The investigators were blinded in eight of the studies (19-21, 23-25, 27, 30), and three studies did not mention blinding (18,28,29). Unequal proportions of participants with major and minor depression were randomised to the intervention and control groups in two studies (23,25). The proportions of females and males randomised to the intervention and control groups were unequal in seven studies (21-23, 25-27, 30). One study had no dropout participants (23), while attrition rates in another nine studies ranged from 1.5% to 20.5% (18-20, 22, 24-26, 29, 30); three studies did not report the attrition rate (21,27,28). Ten studies performed data analysis with the intent-to-treat a control group (standard care from a general practitioner [GP]) (25). In another three-armed RCT, Blumenthal et al. (18) compared supervised aerobic exercise, sertraline (an antidepressant drug) treatment alone and combined supervised aerobic exercise with sertraline treatment without a control group. Brenes et al. (27) used a three-armed RCT approach to compare supervised aerobic exercise with sertraline-only treatment and a control group (usual treatment). Sadeghi et al. (28) conducted a three-armed RCT to compare supervised aerobic exercise with CBT and a control group (no treatment). In another three-armed RCT, supervised aerobic exercise was compared with placebo exercise (stretching exercise) and a control group, and all participants took an antidepressant for < 2 weeks (29). A three-armed RCT compared supervised aerobic exercise with sertraline alone and a placebo group (21). Belvederi Murri et al. (22) carried out a three-armed RCT that compared sertraline plus progressive aerobic exercise with sertraline plus non-progressive exercise and sertraline alone. Dunn et al. (30) conducted an RCT with a 2 x 2 factorial design to compare aerobic exercise at the public health dose (PHD; 17.5 kcal/kg/week, 5 times/week) with aerobic exercise at low dose (LD; 7 kcal/ kg/week, 3 times/week) and a control group (no treatment). Finally, an RCT with parallel groups was conducted to compare supervised aerobic exercise with home-based exercise, sertralineonly treatment and a control group (placebo pill titrated to be similar to sertraline) (19).

17-item Hamilton Rating Scale for Depression
Review Article | Exercise intervention as a treatment for depression principle (18-22, 24, 26, 27, 29, 30). All but one study compared exercise interventions with a control or placebo group (18).

Efficacy of Exercise Intervention Immediately Post-Intervention
Seven studies reported the efficacy of exercise intervention for the treatment of depressive disorders immediately after completion of the intervention (Table 1). Of the four studies that compared supervised aerobic exercise with a control group, three reported that supervised aerobic exercise led to a significantly higher reduction in the severity of depressive symptoms immediately post-intervention compared with that of the control group (19,29,30). However, another study reported that supervised aerobic exercise did not lead to a significantly greater reduction in severity of depressive symptoms post-intervention when compared with the control group (stretching exercise) (20). Dunn et al. (30) reported that the PHD of supervised aerobic exercise reduced the severity of depressive symptoms by 47% at the post-intervention measurement compared with the LD of supervised aerobic exercise (30%) and the control group (no exercise, 29%). Blumenthal et al. (19) showed that supervised aerobic exercise led to remission (HAMD ≤ 8) in 45% of participants post-intervention compared with home-based exercise (40% remission) and placebo (31% remission), but the sertraline-only treatment group had the highest remission, at 47% (19).
Blumenthal et al. (21) revealed that supervised aerobic exercise and sertraline reduced the severity of depressive symptoms at 4 months after the beginning of the intervention compared with the administered placebo, but there was no difference in the reduction of severity of depressive symptoms between the supervised aerobic exercise and sertraline groups. Legrand and Neff (29), who recruited all patients with antidepressant treatment of < 2 weeks, indicated that the effect size of depressive symptoms reduction at the postintervention measurement was significantly greater for supervised aerobic exercise compared with stretching exercise and control groups. In contrast, two studies compared active treatment groups. Blumenthal et al. (18) reported that no significant differences in the reduction of depressive symptoms postintervention were detected among supervised aerobic exercise, sertraline-only treatment and combined sertraline and supervised aerobic exercise treatment. Belvederi Murri et al. (22) reported a greater remission rate in sertraline plus progressive aerobic exercise and sertraline plus non-progressive exercise groups compared with sertraline-only treatment after 24 weeks of intervention. Shorter time to remission was also observed in the sertraline plus progressive aerobic exercise group compared with the sertraline-only group. Two studies compared high-intensity PRT with a control group in elderly patients immediately post-intervention (23,25). Singh et al. (23) reported that highintensity PRT resulted in a significantly greater reduction in depressive symptoms compared with the control group (health education) at the post-intervention measurement, and Singh et al. (25) found that high-intensity PRT caused a significantly greater reduction in depressive symptoms (61% of participants) compared with low-intensity PRT (29%) and the control (standard GP care; 21%) post-intervention.

Efficacy of Medium-and Long-Duration Exercise Intervention
Two studies investigated the efficacy of PRT compared with control groups in elderly patients (24,26). Sims et al. (26) reported that moderate-intensity PRT did not result in a significantly greater reduction in depressive symptoms compared with the control group in elderly patients immediately and 6 months after the intervention. Singh et al. (24) demonstrated that high-intensity PRT led to a significantly greater reduction in depressive symptoms compared with controls (health education) in elderly patients at 20 weeks of intervention and 26 months after the intervention. The antidepressant effect was maintained at 20 weeks of intervention and 33% of the participants still performed regular exercise after 26 months without supervision.
One study reported that supervised aerobic exercise and CBT led to a significant reduction of depressive symptoms compared with the control group at 8 weeks after completion of the intervention, but the exercise group did not differ from the psychotherapy group in terms of a reduction in depressive symptoms (28). Finally, another study demonstrated that supervised aerobic exercise with resistance training and sertraline-only treatment groups exhibited significantly greater reduction in the clinician-and hence, the results cannot be generalised to younger adult depressed patients.
Three studies demonstrated that, when compared to control groups, high-intensity PRT (3 days/week for 40-60 min per session) for 8 weeks-26 months was efficacious for reducing depressive symptoms in patients with major and minor depressive disorders and dysthymia (23)(24)(25). However, these studies only recruited elderly patients (> 60 years old) and had relatively small sample sizes (n = 32-60). Singh et al. (25) also had a short duration of intervention (only 8 weeks). In contrast, Sims et al. (26) reported that moderate-intensity PRT was not efficacious for reducing depressive symptoms in patients with unspecified depression compared with controls. However, Sims et al. (26) recruited only 36 participants, the duration of the intervention was short (10 weeks) and participants were not blinded. Only one study reported the efficacy of combined supervised aerobic exercise with resistance training (3 days/week, 60 min per session) for 16 weeks; they found reduced depressive symptoms in elderly patients (≥ 65 years old) with minor depression as compared to the control group. However, the sample size was small and the attrition rate was not mentioned (27).
It was difficult to compare studies because the selected studies showed heterogeneity in terms of sample size, sample characteristics, methodology, randomisation, blinding strategies, modes of intervention, duration of intervention, nature of the control groups, unequal distribution of gender between groups and different outcome measures applied. However, from our findings, we can conclude that supervised aerobic exercise and highintensity PRT may be efficacious as a primary treatment for depressive disorders, even without the prescription of antidepressants. Nonetheless, we recommend that more multicentre doubleblinded RCTs with larger sample sizes, equal gender distributions across comparison groups and longer follow-up durations with a higher number of repeated assessments be conducted to confirm the efficacy of supervised aerobic exercise as a primary treatment for depressive disorders. In addition, multicentre doubleblinded RCTs with larger sample sizes and equal gender distributions across comparison groups, as well as involving patients from a younger age group, are warranted to confirm the efficacy of high-intensity PRT as a primary treatment for depressive disorders. Furthermore, future administered rating tool (HAMD-17) but not in the self-administered tool (GDS) in elderly patients at 4 months post-intervention (27).

Efficacy of Supervised Aerobic Exercise and High-Intensity Progressive Resistance Training as Primary Treatments for Depression
This comprehensive narrative review examined the efficacy of exercise interventions as the primary treatment for depressive disorders. Four studies indicated that, compared with control groups, supervised aerobic exercise (daily or 3-5 days/week for 30-60 min per session) for 10 days-16 weeks is efficacious for reducing depressive symptoms in patients with major and minor depressive disorders (19,(28)(29)(30). Only one study reported that supervised aerobic exercise was no better than stretching exercise (control group) in reducing depressive symptoms (20). However, only Blumenthal et al. (19) recruited a relatively large sample (n > 100) compared with the other three studies (sample sizes = 35 to 80) (28)(29)(30). Other study limitations of note were as follows: i) Blumenthal et al. (19): most of the participants had only mild depression, and the placebo response rate was high; ii) Krogh et al. (20): the study stopped before achieving the pre-planned sample size; iii) Sadeghi et al. (28): randomisation concealment, blinding and attrition rates were not mentioned, and data analysis was not performed according to the intent-to-treat principle; and iv) Legrand and Neff (29): blinding and medication dosage were not mentioned, and the intervention was of relatively short duration. The efficacy of supervised aerobic exercise was not superior to that of other active treatments, such as antidepressants and CBT (18,19,21,26). However, Blumenthal et al.'s (18) study was limited by the lack of a control group and no mention of blinding. Belvederi Murri et al. (22) indicated that a combination of supervised aerobic exercise (progressive or non-progressive) with antidepressant (sertraline) had superior efficacy compared with sertraline-only treatment in ameliorating depressive symptoms and inducing remission. Yet, the main limitation of this study was that it included only elderly major depressive patients (age 65-85 years), Review Article | Exercise intervention as a treatment for depression level of BDNF in the brain, which may bring about an amelioration of depressive symptoms similar to the effects of antidepressant treatment (32). Exercise stimulates the release of various proteins from the neurons in the brain, and one of these is BDNF (33); BDNF may produce an antidepressant effect by stimulating neurogenesis and having a neuroprotective influence on the hippocampus (34), a vital brain structure that is involved in the regulation of emotional memory (35). An increase in hippocampal volume and size has been observed in subjects performing aerobic exercise for duration of 3 months-1 year (36,37). In a randomised clinical trial of patients with depression, subjects were randomised to an aerobic exercise plus treatment-as-usual group and treatment-as-usual only group, where treatment as usual included antidepressant treatment and psychotherapy; as a result, a significant increase in serum BDNF was observed in the exercise plus treatment-as-usual group compared with the treatment-as-usual only group (34).
Pro-inflammatory cytokines have been reported to induce depressive symptoms when their serum level increases. Various proinflammatory cytokines are increased in the serum of depressed patients compared with nondepressed patients, including interleukin-1β (IL-1β), IL-6 and tumour necrosis factor-α (TNF-α) (38). Again, aerobic exercise for a duration of 3 months has been shown to significantly reduce studies should evaluate the efficacy of exercise intervention conducted in individuals as all the studies reported herein conducted exercise intervention in groups, and therefore, they could not control for the confounding effect of social interactions and social support in the groups, which may themselves improve depressive symptoms.

Biopsychosocial Model of the Mechanisms Underlying the Efficacy of Supervised Aerobic
Exercise and Progressive Resistance Training as Primary

Treatments for Depression
The mechanism behind the efficacy of supervised aerobic exercise and PRT as a primary treatment for depression can be explained in terms of the biopsychosocial model. The biopsychosocial mechanism underlying the effect of exercise on depression is illustrated in Figure 2.

Biological Mechanism Underlying the Effect of Exercise on Depression
One important neurotrophic factor in the brain that exercise interacts with is brainderived neurotrophic factor (BDNF), which is involved in the regulation of neuroregeneration, neurogenesis, neuroprotection and synaptic plasticity (31). While BDNF decreases in depression, exercise tends to increase the According to the monoamine hypothesis of depression, an increase in serum serotonin, dopamine and noradrenaline will ameliorate depressive symptoms in patients.

Psychological Mechanism Underlying the Effect of Exercise on Depression
There are several psychological mechanisms underlying the effect of exercise on depression. Exercise can alleviate stress and tends to improve self-efficacy, self-confidence, selfperception and body image, while it prevents negative thoughts. Exercise also improves satisfaction with life and well-being. Regular exercise improves self-efficacy by increasing one's belief in one's ability to accomplish tasks and reach a certain level of achievement. This is accomplished by allowing oneself to gain mastery experiences via performing the physical activity or watching others, and in turn, allowing oneself to experience a positive affective state during exercise. Improved self-efficacy leads to the formation of a positive self-identity and selfperception as someone who could cope with a negative mood state by accomplishing tasks and reaching certain goals. Eventually, a positive selfschema is formed where one allows oneself to express positive ideas about oneself, generating improved self-esteem. These changes will help to alleviate stress and depressive symptoms (52, 53).
To improve the sense of well-being, it is necessary to ensure the need for selfcompetence, self-autonomy and improved relatedness. Regular exercise allows one to fulfil the need for self-competence by practicing and improving in the sport or physical activity of choice, improving self-autonomy through mastery experience gained in practicing physical activity and improving relatedness through sharing experience with others by joining group activities. Hence, improving the sense of wellbeing by fulfilling these psychological needs will reduce depressive symptoms (54).
Regular exercise could also be viewed as a coping mechanism to overpower maladaptive coping in depression, including social isolation and inactivity. Regular exercise contributes to an increase in pleasure and self-achievement, serving as an activity that may distract depressed persons from maladaptive coping, allow behavioural activation and help regulate the affective state (53, 55). Hence, with regular serum IL-1β, IL-6 and TNF-α, which is also positively correlated with an improvement in depressive symptoms in depressed subjects (39,40). When aerobic exercise intervention is compared with other exercise, such as strength exercise, there is a significantly greater decrease in pro-inflammatory cytokines, including TNF-α and C-reactive protein (41).
Dysregulation in the hypothalamicpituitary-adrenal (HPA) axis is thought to be a risk factor for developing depression, in which the serum cortisol level is increased and there is blunting of the dexamethasone suppression test in depressive patients (42). The effect of regular exercise on the HPA axis is still inconclusive. The basal level of serum cortisol has been reported to increase in response to regular and chronic exercise (43), but one study indicated that regular exercise significantly reduces urinary cortisol levels, serum adrenocorticotropic hormone (ACTH) and depressive symptoms (44). Other hormonal systems may also play a role in the mechanism underlying the effect of exercise on depression, such as the involvement of growth hormone and insulin-like growth factor-1 (IGF-1), which are associated with neuroprotective effects and regulate mood and sleep in the central nervous system (45,46). Growth hormone is found to increase in major depressive disorder patients after acute episodes of aerobic exercise and PRT (47). Although IGF-1 has been reported to exert an antidepressant effect in animal studies (48), regular aerobic exercise in depressed human subjects does not result in an increased serum IGF-1 level (49). Hence, the role of IGF-1 in the mechanism underlying the therapeutic effect of exercise on depression is inconclusive.
Regular exercise may alter monoamine levels in the brain. The effect of aerobic exercise on serotonin differs between brain regions and is time-course dependent. Upon acute aerobic exercise, the serotonin level increases in the mid-brain, striatum, hippocampus, cerebral cortex and brainstem. The cortical changes last for a week after discontinuation of exercise, while serotonin metabolism in the hippocampus reduces 1 day after acute exercise but increases again after 1 week of exercise cessation. In terms of dopamine, upon aerobic exercise, the dopamine level is increased in the hippocampus, striatum, mid-brain and brainstem, modulating the mood and memory (50). Similarly, chronic treadmill and wheel exercise in animal studies are shown to increase noradrenaline levels in Review Article | Exercise intervention as a treatment for depression duration of 40-60 min per session, three times a week for 10-16 weeks is recommended for the elderly. Early morning is suggested for exercise, as it is easier to commit in the morning, before the daily routine intrudes. Furthermore, since exercise can help distract from maladaptive coping, promote behavioural activation and help regulate the affective state, it would be ideal to perform exercise in the morning, especially for depressed patients with morning worsening of depressive symptoms. However, the exercise interventions reviewed were not superior to antidepressants and CBT in the treatment of depression. Future RCTs with more comprehensive and well-designed methodologies are warranted to confirm our findings. Studies that investigate the potential mechanism for exercise as the primary treatment of depression are also warranted. Our review provides useful data for clinicians about the need to include exercise interventions for the treatment of depressive patients because such interventions improve both physical outcomes and patients' mental health.

Psychosocial Mechanism Underlying the Effect of Exercise on Depression
By engaging in regular exercise in a group, one can experience a sense of meaning in life when performing the physical activity and sense of closeness with group members and facilitators by sharing experiences that increase the sense of connectedness, sense of belonging and purpose in life. These factors will allow the person to eliminate maladaptive coping like social isolation, and hence, reduce depressive symptoms. Therefore, the social setting of the group physical activity could bring about mental wellness (56)(57)(58)(59).

Limitations
A few limitations of this review should be noted. First, the review only included articles published in English and there may be articles published in other languages that would have fulfilled the eligibility criteria. However, we did not have funding to include experts who could help interpret findings from non-English articles. Studies in published theses were also excluded, yet they could be a valuable source of data. This omission was due to the time constraints that we faced. Nevertheless, we think this review covered most of the literature evaluating the efficacy of exercise interventions as a primary treatment for depression. Second, due to the heterogeneity of the selected studies, a meta-analysis would have been a more ideal way to assess the literature focussed on the efficacy of exercise interventions for treating depression.

Conclusion
This narrative review included a comprehensive review of the literature that investigated the efficacy of exercise intervention as the primary treatment of depressive disorders, filling the research gap associated with the lack of such data. We found that supervised aerobic exercise and high-intensity PRT may be efficacious in treating depressive disorders, even without the use of antidepressants. Hence, supervised aerobic exercise at a duration of 30-60 min per session, three to five times a week for 12-16 weeks is recommended for adult depressed patients, while high-intensity PRT at a