Depression is one of the most common comorbidities in people with neurological injury. Following a stroke, traumatic brain injury or spinal cord injury, the prevalence of depressive symptoms can reach 30–50% of patients, according to different published series. This has direct clinical implications: comorbid depression is associated with poorer functional outcomes, lower adherence to rehabilitation programmes and a significant reduction in quality of life.
Over the past two decades, scientific evidence on the role of physical exercise in managing depressive symptoms has grown considerably. Several high-quality meta-analyses and systematic reviews have evaluated the efficacy of different exercise modalities, the most appropriate doses, and the populations most likely to benefit. This article reviews that evidence and contextualises it in the field of neurorehabilitation.
The meta-analysis by Schuch et al. (2016), published in the Journal of Psychiatric Research, analysed 25 randomised controlled trials with a total of 1,487 participants. Results showed a large and statistically significant antidepressant effect of exercise (effect size: SMD = -1.11; 95% CI: -1.54 to -0.68). Even when the analysis was restricted to studies with the lowest risk of bias, the effect remained significant (SMD = -0.53), reinforcing the robustness of the findings.
More recently, the network meta-analysis by Singh et al. (2023), published in the British Medical Journal, represented an important methodological advance. This study included 218 randomised controlled trials with 14,170 participants and directly compared multiple exercise modalities against each other and against active controls. The authors concluded that physical exercise is an effective treatment for depression, with effects comparable to those of psychotherapy and pharmacotherapy in certain subgroups.
The Cochrane review by Cooney et al. (2013), which included 39 trials with 2,326 participants, also found a moderate effect of exercise on depressive symptoms compared with no treatment or control interventions. The authors noted that, although the overall quality of evidence was moderate, results were consistent across different subgroups and study designs.
The accumulated evidence suggests that physical exercise can contribute significantly to reducing depressive symptoms, with effect sizes that, in some analyses, approach those of established pharmacological and psychotherapeutic interventions.
One of the most relevant aspects of the Singh et al. (2023) meta-analysis was the direct comparison between different exercise modalities. Results allowed a preliminary efficacy hierarchy to be established, although the authors emphasise that differences between modalities were, in many cases, modest.
Walking, running, swimming or cycling on a stationary bike are the most studied forms of aerobic exercise. The pioneering work by Blumenthal et al. (1999), published in Archives of Internal Medicine, compared a supervised aerobic exercise programme (walking or running 30 minutes, 3 times per week for 16 weeks) with sertraline and the combination of both in patients with major depressive disorder. All three groups showed similar reductions in depressive symptoms, suggesting that aerobic exercise could have comparable efficacy to pharmacotherapy in mild to moderate depression.
A subsequent follow-up study by the same group (Babyak et al., 2000) showed that participants assigned to the exercise group had lower relapse rates at 10 months than those treated exclusively with medication, suggesting a possible long-term protective effect.
The meta-analysis by Gordon et al. (2018), published in JAMA Psychiatry, focused specifically on the effect of resistance training on depressive symptoms. With 33 trials and 1,877 participants, the authors found a significant reduction in symptoms (SMD = -0.66; 95% CI: -0.86 to -0.46). A relevant finding was that the antidepressant effect was independent of physical health status, total training volume and strength gains, suggesting that mechanisms do not depend exclusively on muscular adaptation.
In the context of neurorehabilitation, adapted strength training is particularly interesting, as many patients with neurological injury present significant muscle weakness as part of their clinical picture.
Combining aerobic exercise and strength training in the same programme has shown promising results. Singh et al. (2023) observed that multicomponent programmes were among the interventions with the largest effect sizes. From a clinical perspective, this combination is logical, as it simultaneously addresses cardiorespiratory capacity and neuromuscular function — two areas frequently compromised in people with neurological injury.
Yoga appeared in the Singh et al. (2023) meta-analysis as one of the modalities with significant effect on depressive symptoms. A review by Cramer et al. (2017), published in the Journal of Affective Disorders, analysed 23 trials and reported a moderate effect of yoga versus passive controls. The authors noted that the combination of physical, breathing and meditative components may offer additional mood benefits beyond the purely mechanical effects of movement.
Most studies demonstrating significant effects used programmes of 3 to 5 sessions per week. The Schuch et al. (2016) meta-analysis found that programmes with greater weekly frequency tended to show more pronounced effects. NICE guidelines for depression recommend structured exercise programmes of at least 3 sessions per week.
Singh et al. (2023) observed that vigorous intensity exercise tended to produce greater reductions in depressive symptoms than light or moderate intensity. This should be qualified in the neurorehabilitation context: many patients with neurological injury have functional limitations that affect their ability to reach vigorous intensities, at least in the early stages of the programme.
A pragmatic approach is to prescribe moderate intensity (perceived effort 5–6 out of 10 on the modified Borg scale) as a starting point, with gradual progression towards higher intensities according to patient tolerance and clinical evolution.
Programmes evaluated in the main meta-analyses lasted between 8 and 16 weeks, with 30 to 60 minute sessions. Evidence suggests that longer programmes can consolidate benefits and reduce the risk of relapse. The follow-up study by Babyak et al. (2000) suggested that continued exercise practice was associated with lower rates of depressive recurrence in the medium term.
There is preliminary evidence of a dose-response relationship between exercise volume and reduction of depressive symptoms. Dunn et al. (2005), in a trial published in the American Journal of Preventive Medicine, randomly assigned participants with mild to moderate depression to different doses of aerobic exercise and found that the dose consistent with public health recommendations (17.5 kcal/kg/week) produced greater reductions than a low dose (7 kcal/kg/week).
It is essential to emphasise that current evidence positions exercise as a complement to conventional depression treatment, not a substitute. NICE guidelines (2022) recommend structured exercise as one option within a stepped approach that may include psychotherapy (cognitive behavioural therapy, behavioural activation) and pharmacotherapy.
In clinical practice, the decision to incorporate exercise into the therapeutic plan should be made individually, considering:
Prescribing exercise for people with neurological injury and comorbid depression has specific features that differentiate it from prescription for the general population. The principles of individualisation, gradual progression and safety are of critical importance here.
In people with significant functional limitations, robotic neurorehabilitation technology can facilitate access to exercise programmes that would otherwise be unfeasible. Robotic systems allow:
At GNeuro, the approach to comorbid depression in people with neurological injury is integrated within the individualised neurorehabilitation programme. The clinic, located in Ourense, has state-of-the-art robotic equipment (robotic gait training system, upper limb rehabilitation robot, functional electrical stimulation cycling device, etc.) that allows exercise programmes to be designed adapted to the specific needs of each patient.
What type of exercise is most effective for depressive symptoms?
The evidence suggests that both aerobic exercise and strength training can contribute to reducing depressive symptoms. The meta-analysis by Singh et al. (2023) published in the BMJ indicates that walking, running, yoga and strength training showed significant effects. The choice of exercise type should be individualised according to the patient's functional capacities, preferences and clinical situation.
Can exercise replace antidepressant medication?
No. Physical exercise should be considered as a complement to conventional treatment, not a substitute. Clinical guidelines, including those of NICE, recommend exercise as part of a multimodal approach that may include pharmacotherapy and psychotherapy. Any changes to medication must be supervised by the responsible physician.
How much exercise is needed to observe mood benefits?
According to available evidence, exercise programmes of 3 to 5 sessions per week, with a duration of 30 to 60 minutes per session and moderate to vigorous intensity, can contribute to improving depressive symptoms. Benefits are usually observed progressively over several weeks, and continued adherence is a key factor.
Is exercise safe for people with neurological injury and comorbid depression?
Therapeutic exercise supervised by qualified professionals is generally safe for people with neurological injury, provided it is adapted to the patient's functional capacities. At GNeuro, exercise programmes are designed on an individual basis and carried out with robotic technology (robotic gait training system, upper limb rehabilitation robot) that allows precise control of load and assistance, contributing to programme safety.
At GNeuro we design individualised therapeutic exercise programmes for people with neurological injury, integrating state-of-the-art robotic technology with an evidence-based clinical approach.
Request assessment