Does Exercise Increase Your Stress Tolerance? What Happens at Work, in Burnout, and in Depression

Key Takeaways: Does Exercise Increase Your Stress Tolerance?

• Does Exercise Increase Your Stress Tolerance? The evidence suggests that exercise can improve several outcomes related to stress resilience, including blood-pressure reactivity, subjective energy, fatigue, and vitality, although the strength of evidence varies by outcome.
• A single exercise session can modestly reduce blood-pressure reactivity to a later stressor, while findings for acute cortisol responses are less consistent.
• Regular exercise training appears to improve energy, vitality, and fatigue over time, which may be clinically relevant for people dealing with sustained stress.
• Exercise is itself a physiological stressor, but unlike ongoing psychological stress, it is typically time-limited and followed by a recovery period that can support adaptation.
• In occupational burnout, the evidence for exercise as a treatment is much weaker than the evidence for general work-stress resilience. Reducing the underlying stress load often becomes the more important clinical priority.
• In depression, exercise has a much larger randomized-trial evidence base, with walking or jogging, strength training, and yoga all showing meaningful reductions in depressive symptoms.
• The practical challenge is often feasibility: patients with burnout or depression may understand that exercise could help but still struggle with fatigue, low initiative, or uncomfortable physical symptoms that make training difficult.
• Exercise is best viewed as one tool within a broader clinical picture rather than as a universal solution to stress, burnout, or depression.

Introduction: Does Exercise Increase Your Stress Tolerance

Patients with burnout and depression often come to my practice looking for practical ways to cope better with stress. Understandably, much of the discussion focuses on how to reduce the stress itself — lowering the workload, creating more room for recovery, and removing unnecessary demands. That is often the most logical place to start. But another question comes up as well: is there anything that can help a person tolerate stress better in the first place?

Exercise is part of that conversation surprisingly often. Some patients are hesitant to keep exercising when they already feel exhausted, because training can seem like just another demand on an overloaded system. I understand that concern. At the same time, when exercise is still manageable, I often try to avoid letting it disappear automatically along with other basic lifestyle habits. One reason is that patients may understandably focus on the immediate effort exercise requires and have a harder time seeing the potential benefits it may provide over time.

That tension is what prompted me to write this article. People can approach the same question from very different points on the spectrum. Someone working long hours under mounting pressure may wonder whether regular training can improve their ability to cope with stress before they reach burnout. Someone who is already depressed may wonder whether exercise is worth attempting at all, or whether it simply adds another burden.

These questions ultimately lead to the same broader issue: how exercise interacts with the systems involved in the stress response, and whether those effects translate into something clinically meaningful. This article looks at what the evidence actually supports at different points on that spectrum — from the healthy but stretched-thin worker, to the point where training itself may become an additional load, to exercise as part of the treatment landscape in major depressive disorder. Just as importantly, it looks at where the evidence is weaker than the popular narrative may suggest.

How Exercise Affects Stress Tolerance: The Underlying Mechanisms

Before looking at burnout or depression specifically, it helps to establish what “stress tolerance” actually means physiologically and what exercise does to it. The stress response runs largely through the hypothalamic-pituitary-adrenal (HPA) axis and the autonomic nervous system, with cortisol among the endocrine markers used to assess stress responses [4]. A systematic review and meta-analysis of randomized clinical trials and crossover studies found that a single bout of physical exercise attenuates systolic, diastolic, and mean blood pressure reactivity to a subsequent laboratory stress task, with pooled effect sizes in the small-to-moderate range; the authors noted the clinical relevance of an effect this size still needs more study [1]. Not every study replicates this: one randomized laboratory study using the Trier Social Stress Test found that a 30-minute high-intensity exercise bout did not significantly attenuate the subsequent blood pressure or cortisol stress response, despite prior work pointing the other way; the stress test was followed by a video lecture used to measure comprehension and mind-wandering, not by the stressor itself [2].

Regular training may go further than blunting a single stress response — the “cross-stressor adaptation hypothesis” proposes that becoming fitter through repeated exercise-induced stress also reduces the body’s reactivity to unrelated, non-exercise stressors. A 12-week randomized controlled trial of endurance training in healthy men found support for this: the trained group showed reduced heart rate, heart rate variability, and cortisol reactivity to a standardized psychosocial stress test compared with a wait-list control; the relaxation-training comparison group also showed reduced cortisol reactivity, though not reduced heart rate or heart rate variability reactivity [3]. That said, this hypothesis is not universally supported — other reviews describe the evidence for its generalization to stressors outside exercise itself as limited [4].

Cortisol responses also appear to vary by exercise modality and dose, at least in populations with elevated psychological distress. A 2025 network meta-analysis of exercise interventions in adults with psychological distress found that overall, exercise was associated with moderate cortisol reductions, with mind-body practices such as yoga producing the largest effect; high-intensity interval training tended to increase cortisol levels, although this was not statistically significant. The dose-response relationship was not linear — it followed an inverted U-shape, with the greatest cortisol reduction observed at approximately 530 MET-minutes per week, beyond which the effect plateaued [8].

Separate from cortisol reactivity, a comparatively consistent finding in this literature concerns subjective feelings of energy and fatigue rather than any hormonal measurement. A meta-analysis of 16 experiments and 678 participants found that a single bout of exercise increased self-reported energy consistently, with a standardized mean effect of 0.47 [5]. A larger, more recent meta-analysis of 81 randomized trials and over 7,000 participants examined the effects of ongoing (not single-session) exercise training and found small-to-moderate improvements in feelings of energy, vitality, and reduced fatigue compared with control groups; the size of each of these three effects was moderated somewhat differently — fatigue by exercise intensity and program duration, energy by intensity and modality, and vitality by participants’ baseline health status, exercise intensity, modality, and training location [6].

From a clinical perspective, this distinction also makes intuitive sense to me. Exercise is itself a physiological stressor, but it is a stressor that is specifically capable of driving adaptive responses. A narrative review frames exercise as a controllable, time-limited challenge that can recalibrate stress-related systems when repeated bouts are matched by sufficient recovery (and bioenergetic support); without recovery the same load may add stress burden rather than resilience [15].

A narrative review frames exercise as a controllable, time-limited challenge that can recalibrate stress-related systems when repeated bouts are matched by sufficient recovery (and bioenergetic support); without recovery the same load may add stress burden rather than resilience [15]

This is different from simply “being stressed.” Ongoing psychological stress can also lead to biological adaptation, but it does not usually produce the same kind of structured training–recovery–adaptation response that exercise does. In other words, exercise does not merely add stress to the system; it can also provide the stimulus for the body to adapt to that stress.

That distinction is clinically useful. Patients may understandably think that exercise is just another burden on an already overloaded system. In reality, exercise and persistent psychological stress do not affect the body in exactly the same way. Exercise is typically a time-limited physical challenge with a clear recovery phase, whereas psychological stress may be prolonged, recurrent, and less clearly separated from recovery. The same review describes exercise, in many human contexts, as more controllable, time-limited, and predictable than chronic psychological stress [15].

Exercise, Work Stress, and Stress Tolerance Before Burnout

For someone working hard but not yet burned out, the practical question is whether exercise measurably improves their stress tolerance for occupational demands. At the population level, a large meta-analysis of workplace physical activity intervention studies — synthesizing outcomes across roughly 38,000 subjects across all outcomes studied — found a reported job-stress effect size of 0.33 at follow-up, though this comparison was based on only three two-group studies and the paper’s own reported 95% confidence interval for it crossed zero; other job-stress comparisons within the same dataset were also not statistically significant. Fitness and work attendance improved more robustly in the same analysis [7]. A separate observational study of 2,660 Swedish healthcare workers and social insurance officers used latent profile analysis to identify six distinct occupational stress/mental-health profiles, and found that employees in the more resilient profiles — those with only moderate mental health problems despite high stress, or good mental health despite moderate stress — were more physically active than their non-resilient counterparts within the same stress level [13]. This kind of cross-sectional association cannot establish that exercise causes the resilience — people who are already coping well may simply have more capacity left over to exercise — but it’s consistent with the buffering mechanisms described above.

Modality-specific data among a stress-exposed occupational group also exists: a systematic review and network meta-analysis of physical relaxation methods for occupational stress in healthcare workers found that, compared with non-intervention controls, physical relaxation methods overall reduced occupational stress at longest follow-up, and yoga specifically was ranked highest among the modalities tested [12].

Taken together, these studies suggest modest benefits across several outcomes relevant to stress responses and perceived occupational stress — plausibly operating through the same acute and cross-stressor mechanisms described in the previous section, plus the direct energy/fatigue benefits.

Many patients would probably benefit from greater resilience to stress if that capacity could be improved. Reducing the stress itself remains important, of course, and in practice that is often where the conversation starts. But there is also a separate question: can we improve how well the body and mind tolerate stress when it cannot be removed completely?

Exercise is one potential part of that picture, but the practical reality is not always simple. In my clinical experience, people under substantial stress may find exercise unusually difficult to maintain. Some already have physical symptoms associated with stress, such as palpitations, and those sensations may sometimes become more noticeable during exercise. Even when there are no prominent physical symptoms, some patients simply feel that stress has consumed most of their available energy, leaving very little motivation or capacity for training.

That is an important clinical distinction. The theoretical value of exercise is not the same thing as a person’s ability to perform it while they are already struggling. When I discuss exercise in this setting, I think it is useful to keep both sides in view: the potential role of exercise in supporting stress tolerance, and the very real difficulty some patients have in continuing to exercise when stress itself is already producing fatigue or uncomfortable physical sensations.

Where the Stress-Tolerance Buffer Runs Out: A Note on Overreaching

The stress-tolerance framework above assumes exercise is being layered onto an already-stressed system as a recovery tool, not as an additional load competing for the same limited recovery capacity. That assumption breaks down once training volume or intensity outpaces recovery. At that point, exercise increasingly adds to the person’s total training and recovery load rather than helping offset it — the territory of functional and non-functional overreaching and, at the far end, overtraining syndrome. That progression and how it can be distinguished from burnout and depression are covered in more depth in [Functional Overreaching vs. Non-Functional Overreaching]; the role of HPA-axis dysregulation specifically is covered in [Cortisol and Overtraining], which is candid that the exact mechanism and timeline remain incompletely established. Poor sleep can add to the overall recovery burden and is covered separately in [Sleep for Athletes]. The short version for this article: more exercise is not unconditionally better for stress tolerance, and a person already running an occupational stress deficit needs a program sized to add recovery capacity, not subtract from it.

This is also the point where the evidence for exercise as a treatment for burnout specifically — as opposed to a general buffer against work stress in people who are not yet burned out — gets notably weaker, as the next distinction shows.

In my clinical experience, overreaching is something I mainly encounter in people who train seriously. They are often highly active recreational athletes or competitive athletes with a substantial and sustained training load.

It is also relatively unusual, in my experience, for someone to arrive at this point completely by accident. The overall training pattern often makes overreaching a plausible possibility before the diagnosis is even considered, and patients themselves may already suspect that their training load has exceeded their ability to recover. That clinical context matters, because overreaching is usually easier to interpret when it fits with the broader history of training volume, intensity, and recovery rather than being inferred from fatigue or poor performance alone.

Burnout specifically: a more mixed picture

It’s worth being explicit that occupational burnout is a distinct question from general work-stress resilience, and the trial evidence does not support treating them the same way. A systematic review and meta-analysis restricted to randomized controlled trials of exercise therapy in people already diagnosed with burnout identified only four RCTs with usable data (248 participants total). Three of the four individual studies reported a significant reduction in burnout or emotional exhaustion scores, but the pooled meta-analytic estimate had a wide confidence interval crossing the line of no effect (SMD 95% CI −0.41 to 0.09, I² = 0%), and the authors concluded that their results do not support the widespread assumption that exercise therapy is an effective treatment for burnout, while also cautioning that the small number of available randomized trials is itself a major limitation rather than proof of no effect [9].

This does not contradict the workplace evidence above. A program aimed at buffering everyday work stress and a program aimed at treating someone who already has established occupational burnout are different clinical situations. It is also worth being precise about what the workplace data actually show: those studies measured job stress, not burnout incidence, so they suggest that exercise may influence a related outcome but do not demonstrate that exercise prevents burnout from developing.

Once burnout has progressed far enough to cause clear functional impairment, the clinical focus often shifts away from trying to build more stress tolerance and toward reducing the stress load itself. In practice, that may mean creating real distance from the main stressor — often work — rather than simply adding more coping strategies on top of an already overloaded situation.

In my clinical experience, this is the point where temporary sick leave can become relevant, alongside occupational health support or psychological care. The central idea is that recovery may be difficult if the person remains continuously exposed to the same stressor at the same intensity.

I would not, however, automatically view exercise as something that needs to be stopped. If it remains manageable and does not clearly worsen the person’s overall condition, it may still have a place during recovery. The more important distinction, in my view, is between reducing the harmful occupational load and unnecessarily removing every other potentially supportive activity from a patient’s life.

Exercise and Clinical Depression

At the other end of the spectrum, the randomized-trial evidence base is considerably larger — though the honest caveat is that confidence in most of it is still rated low by formal grading standards, reflecting substantial risk of bias across the included trials, including the inherent difficulty of blinding participants to an exercise intervention.

A large 2024 systematic review and network meta-analysis in The BMJ pooled 218 randomized trials, 495 treatment arms, and 14,170 participants with depression, defined either by clinician diagnosis or by meeting an established clinical symptom threshold. Compared with active control conditions, moderate-to-large reductions in depressive symptoms were found across several exercise modalities: walking or jogging (Hedges’ g −0.62), yoga (g −0.55), strength training (g −0.49), mixed aerobic exercise (g −0.43), and tai chi or qigong (g −0.42); dance showed the largest effect in this dataset (g −0.96) though based on far fewer and less diverse trials. For comparison, the same analysis found cognitive behavioral therapy alone produced a moderate effect (g −0.55) and SSRIs alone a smaller effect (g −0.26) against the same type of active control condition. In the same network model, the point estimate for exercise combined with an SSRI (g −0.55) was larger than the point estimate for SSRIs alone — though this reflects the network’s modeled comparisons, not a study that directly tested exercise-plus-SSRI against SSRI alone for statistical superiority, and it was not larger than the point estimates for the best-performing solo exercise modalities [10].

A few findings from that same analysis are directly relevant to prescribing exercise for someone who is depressed rather than simply stressed:

• Effects were dose-dependent on intensity: light activity such as walking or gentle yoga still produced a clinically meaningful effect, but vigorous exercise (for example running or interval training) produced a larger one on average [10].
• Counterintuitively, studies that described giving participants more autonomy or choice over their exercise prescription tended to show weaker effects than studies coded as providing less autonomy — the authors suggest this may reflect the low self-efficacy characteristic of depression making it harder for people to self-select an adequately challenging program without guidance, though they caution against over-interpreting this. This was an exploratory, study-level association rather than a randomized test of prescription style itself, so it shouldn’t be read as proof that removing autonomy improves outcomes [10].
• Strength training and yoga had significantly lower dropout odds than active control conditions in this analysis [10].

On the mechanistic side, a small randomized controlled trial in hospitalized inpatients with major depressive disorder (n=25) specifically tested whether a six-week aerobic exercise program altered cortisol reactivity to a standardized psychosocial stress test — the same HPA-axis question raised earlier in this article, but in a clinically depressed population. The result was negative: cortisol reactivity did not change from baseline in either the exercise or the control group [11]. This doesn’t contradict the symptom-level findings above — depressive symptoms and TSST cortisol reactivity are different outcomes — but it’s a reminder that the mechanistic story is not as settled as the clinical trial evidence for symptoms.

It bears repeating that the credibility ratings attached to almost all of these individual comparisons were graded low or very low, reflecting substantial risk of bias across the included studies, including the inherent difficulty of blinding participants to an exercise intervention — this doesn’t mean the effects are illusory, but it does mean the precise size of the benefit should be held with some humility. The paper’s own framing is that these exercise modalities could be considered alongside psychotherapy and antidepressants as core treatment options, not that exercise has been shown to outperform them [10].

One of the most difficult parts of discussing exercise with patients who are depressed is that the advice can collide directly with the illness itself. A patient may reasonably ask: how am I supposed to exercise when I struggle to initiate even basic activities? In my clinical experience, that is often the central practical problem. The issue is not necessarily that the patient does not understand that exercise may help, but that they may lack the drive or initiative needed to act on that information.

This creates a frustrating gap between what may be beneficial in theory and what is realistically achievable in practice. If someone else were structuring the activity and providing external support, the situation might be different, but many patients are left having to initiate the behavior themselves. That can make this group particularly difficult to help through lifestyle advice alone.

I also think there is a risk in pushing the message too hard. Telling a depressed patient repeatedly that they “need to exercise” may not make the behavior more achievable. In some patients, it may simply add another standard they feel they are failing to meet. From a clinical perspective, that is an important distinction: knowing that something may help is not the same as being able to do it, and reminding a patient of that gap does not necessarily close it.

Same Lever, Different Starting Points: Building Stress Tolerance at Every Stage

Put side by side, a coherent picture of stress tolerance emerges even though the underlying trial bases don’t overlap much:

• The healthy, hard-working, not-yet-burned-out person gets a modest, real buffering effect from acute exercise on blood-pressure reactivity, and — probably the most dependable benefit day to day — improved subjective energy, with fatigue and vitality benefits building over weeks of regular training; evidence on cortisol reactivity specifically is more heterogeneous across study designs. This is work-stress-oriented territory, and the workplace-intervention data offer some support, though the job-stress finding specifically rests on a small number of studies.
• The person already in occupational burnout is in a different evidence category. The mechanisms plausibly still apply, but the dedicated randomized-trial evidence for exercise as a treatment, specifically, is thin and doesn’t currently support strong claims either way.
• The person with major depressive disorder has by far the largest randomized-trial evidence base among the three situations discussed here, with effect sizes for several exercise modalities comparable in magnitude to CBT and larger than the point estimate for SSRIs alone in the same network model — while still carrying real caveats about study quality and certainty, and while depression itself can make it harder to initiate the exercise that would help; studies coded as allowing less participant autonomy over their exercise prescription tended to show stronger effects in this dataset, though this was an exploratory association rather than a randomized test of prescription style.

Conclusion: Does Exercise Increase Your Stress Tolerance?

Exercise is not a universal answer to stress, burnout, or depression, but neither is it simply “more stress” added to an already overloaded system. The evidence suggests that its role depends heavily on where a person is starting from. In someone who is stressed but still functioning reasonably well, exercise may support several aspects of stress tolerance, including blood-pressure reactivity, subjective energy, and longer-term fatigue and vitality. Once burnout is established, the priority often shifts toward reducing the stress load itself, while exercise may still remain part of recovery if it is manageable. In depression, the trial evidence for exercise is considerably stronger, but the practical challenge is often not knowing what may help — it is having enough drive and initiative to actually do it.

From a clinical perspective, that is probably the most important takeaway. Exercise can be useful, but its value cannot be separated from context, recovery, and feasibility. A recommendation that makes sense physiologically may still be unrealistic for someone who is exhausted, overwhelmed, or struggling to initiate basic activities. The goal is therefore not to tell every stressed or depressed patient to simply exercise more, but to understand where exercise fits into the broader picture — as a possible tool for adaptation and resilience, not as another standard the patient is expected to fail.

References

1. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9276760/
2. https://journals.plos.org/plosone/article/file?id=10.1371%2Fjournal.pone.0318222&type=printable
3. https://link.springer.com/article/10.1007/s10865-014-9562-9
4. https://www.germanjournalsportsmedicine.com/fileadmin/content/archiv2008/heft07_08/ArtikelGerber.pdf
5. https://www.tandfonline.com/doi/abs/10.1080/21641846.2013.843266
6. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9206544/
7. https://pubmed.ncbi.nlm.nih.gov/19765506/
8. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12736704/
9. https://www.jssm.org/hf.php?id=jssm-17-475.xml
10. https://www.bmj.com/content/384/bmj-2023-075847.full
11. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7291068/
12. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8263904/
13. https://www.sciencedirect.com/science/article/abs/pii/S1469029214001071
14. https://www.nice.org.uk/guidance/ng222/chapter/Recommendations
15. https://pubmed.ncbi.nlm.nih.gov/42182059/