Exercise and Physical Activity for Mental Health Prevention: Dose-Response Relationships, Neurobiological Mechanisms, and Population-Level Evidence

Physical activity has emerged as one of the most robust modifiable risk factors for mental disorders, with a strength of evidence that now rivals — and in some contexts exceeds — many pharmacological and psychotherapeutic interventions for prevention. Unlike most psychiatric treatments, which are deployed after symptom onset, exercise operates across the full spectrum from universal prevention (reducing incidence in healthy populations) to indicated prevention (attenuating progression in at-risk individuals) to adjunctive treatment (improving outcomes in diagnosed disorders).

The epidemiological evidence is striking. A landmark prospective study by Harvey and colleagues (2018), drawing on the HUNT cohort of over 33,000 adults followed for 11 years, estimated that 12% of future cases of depression could have been prevented if all participants had engaged in at least one hour of physical activity per week — a population-attributable fraction (PAF) larger than many established risk factors. Meta-analytic data from Schuch et al. (2018), synthesizing 49 prospective studies with 266,939 person-years of follow-up, found that individuals with high physical activity levels had 17% lower odds of developing depression (adjusted OR = 0.83, 95% CI: 0.79–0.88) compared to those with low activity, with a clear dose-response gradient.

What has changed in recent years is not merely the accumulation of more data, but a deeper understanding of the specific neurobiological mechanisms through which exercise exerts its effects, the dose-response parameters that inform clinical prescriptions, and the comparative effectiveness data that allow clinicians to position exercise relative to established treatments. This article provides a comprehensive, research-grounded review of these domains, drawing on meta-analytic evidence, landmark cohort studies, and randomized controlled trials (RCTs) to articulate the current state of the science.

Neurotransmitter and Neuroendocrine Systems

Exercise modulates multiple neurotransmitter systems implicated in psychiatric disorders. The most extensively studied include:

• Serotonergic system: Aerobic exercise increases tryptophan availability in the brain by promoting the peripheral uptake of branched-chain amino acids (BCAAs) into skeletal muscle, thereby reducing competition for transport across the blood-brain barrier. This results in increased central serotonin (5-HT) synthesis, particularly in the dorsal raphe nuclei. Animal models demonstrate upregulation of 5-HT_1A autoreceptor sensitivity and increased serotonin transporter (SERT) expression following chronic exercise, paralleling mechanisms targeted by SSRIs.
• Dopaminergic system: Exercise activates mesolimbic and mesocortical dopamine pathways, increasing D2 receptor availability in the striatum. Positron emission tomography (PET) studies in humans (e.g., Boecker et al., 2008) have demonstrated exercise-induced dopamine release in the ventral striatum. This mechanism is particularly relevant for anhedonia, motivational deficits, and the reward-processing abnormalities central to both depression and substance use disorders.
• Noradrenergic system: Acute exercise robustly activates the locus coeruleus–norepinephrine (LC-NE) system, increasing arousal, attention, and stress responsivity. Chronic exercise training modulates the sensitivity of this system, leading to attenuated noradrenergic responses to psychosocial stressors — a form of cross-stressor adaptation.
• Endocannabinoid system: The phenomenon historically attributed to "runner's high" via beta-endorphins is now more accurately attributed to exercise-induced increases in circulating endocannabinoids, particularly anandamide (AEA). Sparling et al. (2003) demonstrated significant elevations in plasma AEA following moderate-intensity running, and subsequent work has confirmed CB1 receptor-mediated anxiolytic and analgesic effects.
• GABAergic system: Magnetic resonance spectroscopy (MRS) studies by Streeter and colleagues (2010, 2012) demonstrated that yoga and aerobic exercise increase cortical GABA levels, particularly in the thalamus and anterior cingulate cortex, offering a plausible mechanism for anxiolytic effects.

Hypothalamic-Pituitary-Adrenal (HPA) Axis Regulation

HPA axis dysregulation — characterized by elevated basal cortisol, blunted cortisol awakening response, and impaired glucocorticoid receptor (GR) sensitivity — is a core biological feature of major depression, PTSD, and chronic stress states. Regular exercise normalizes HPA axis function through multiple pathways: it restores glucocorticoid receptor expression in the hippocampus and prefrontal cortex, reduces adrenal sensitivity to ACTH, and enhances negative feedback inhibition. The net effect is a more resilient cortisol response pattern — lower baseline levels with preserved acute reactivity.

Brain-Derived Neurotrophic Factor (BDNF) and Neuroplasticity

Perhaps the most compelling neurobiological mechanism involves the neurotrophic hypothesis. Exercise potently upregulates brain-derived neurotrophic factor (BDNF), a key mediator of hippocampal neurogenesis, synaptic plasticity, and dendritic arborization. A meta-analysis by Szuhany, Bugatti, and Otto (2015) across 29 studies confirmed that a single bout of exercise increases peripheral BDNF levels and that regular exercise training elevates resting BDNF concentrations, with larger effects observed in individuals with depression. The landmark BDNF Val66Met polymorphism (rs6265) moderates this relationship: Met allele carriers show attenuated BDNF secretion in response to exercise, potentially explaining some of the individual variability in exercise response.

Neuroinflammatory Pathways

Chronic low-grade systemic inflammation — elevated IL-6, TNF-α, and C-reactive protein (CRP) — has been identified as a transdiagnostic mechanism linking depression, anxiety, psychosis risk, and cognitive decline. Exercise exerts anti-inflammatory effects through the release of myokines (particularly IL-6 from contracting muscle, which paradoxically acts as an anti-inflammatory cytokine in the exercise context), downregulation of adipose tissue-derived pro-inflammatory mediators, and enhancement of regulatory T-cell function. A meta-analysis by Kandola et al. (2019) found that physical activity's protective effect against depression was partially mediated by reductions in inflammatory biomarkers, with CRP reductions of approximately 20-30% in regular exercisers.

Structural and Functional Brain Changes

The Erickson et al. (2011) RCT demonstrated that one year of moderate aerobic exercise increased hippocampal volume by approximately 2%, effectively reversing 1-2 years of age-related atrophy. This structural change correlated with increased serum BDNF and improved spatial memory performance. Functional neuroimaging studies have shown that exercise enhances default mode network (DMN) connectivity, strengthens prefrontal-amygdala regulatory circuits (relevant to emotion regulation), and increases white matter integrity in frontolimbic tracts.

One of the most clinically actionable advances has been the characterization of dose-response relationships, which enable evidence-based "prescriptions" for exercise as a preventive intervention.

Depression Prevention

The Choi et al. (2019) Mendelian randomization study, using genetic instruments for objectively measured physical activity in over 611,000 individuals, provided strong causal evidence that increasing physical activity reduces depression risk, with an estimated 26% reduction in odds per standard deviation increase in accelerometer-measured activity. Critically, this study controlled for confounding through genetic design, substantially strengthening causal inference.

The dose-response curve is nonlinear and characterized by diminishing returns. The Pearce et al. (2022) analysis of 15 pooled prospective studies (191,130 participants) found that the steepest risk reduction occurs at low volumes of activity: moving from sedentary to approximately 150 minutes/week of moderate-intensity activity (equivalent to ~8.8 MET-hours/week) was associated with a 25% reduction in depression risk. At half the recommended dose — approximately 75 minutes/week — a 18% risk reduction was still observed. Beyond 300 minutes/week, additional protective effects were minimal, suggesting a plateau.

Anxiety Prevention

The evidence for anxiety disorders, while somewhat less extensive, follows a similar pattern. Schuch et al. (2019) meta-analyzed 13 prospective studies and found that individuals with high physical activity had 26% lower odds of developing anxiety disorders (OR = 0.74, 95% CI: 0.62–0.88). The anxiolytic threshold appears to be somewhat lower than for depression, with benefits emerging at as little as 30 minutes of moderate activity, three times weekly.

Exercise Intensity and Modality

Both aerobic exercise and resistance training demonstrate mental health benefits, though through partially distinct mechanisms. A landmark RCT by Gordon et al. (2018) found that resistance exercise training (RET) significantly reduced depressive symptoms with a pooled effect size of SMD = -0.66 across 33 trials — a moderate-to-large effect comparable to pharmacotherapy. Notably, the antidepressant effect of RET was present regardless of baseline health status, total prescribed volume, or significant improvements in strength, suggesting that the psychological and neurobiological effects of resistance training are not entirely dependent on physiological fitness gains.

For aerobic exercise, moderate intensity (defined as 50-70% of maximum heart rate, or a perceived exertion level permitting conversation) appears to be the minimum effective dose for neurobiological adaptations. High-intensity interval training (HIIT) has shown particularly robust effects on BDNF elevation and cardiovascular fitness, with some evidence of superior antidepressant effects compared to moderate continuous training, though head-to-head data remain limited.

Current Guideline Recommendations

The World Health Organization (2020) recommends 150-300 minutes of moderate-intensity or 75-150 minutes of vigorous-intensity aerobic physical activity per week for adults, with additional muscle-strengthening activities on 2 or more days per week. The American Psychiatric Association (2024 update) and NICE guidelines now include physical activity as an evidence-based recommendation for depression management, though specific dose prescriptions remain less standardized than pharmacological dosing.

The critical clinical question is not whether exercise "works," but how its effects compare to established pharmacological and psychotherapeutic interventions. Several landmark studies and meta-analyses address this directly.

Exercise vs. Antidepressants

The SMILE (Standard Medical Intervention and Long-term Exercise) study by Blumenthal et al. (2007) randomized 202 adults with major depressive disorder (MDD) to supervised aerobic exercise, sertraline (50-200 mg), or combined treatment for 16 weeks. Remission rates were comparable across groups: 46% for exercise alone, 47% for sertraline, and 45% for combined treatment (no significant differences). At 10-month follow-up, the exercise-alone group had significantly lower relapse rates (30% vs. 52% for medication, p = 0.028), suggesting durable benefits beyond discontinuation.

The large-scale meta-analysis by Noetel et al. (2024), published in the BMJ and incorporating 218 RCTs with 14,170 participants, found that walking/jogging, yoga, strength training, and dance all showed moderate-to-large effects on depression symptoms, with effect sizes comparable to cognitive-behavioral therapy (CBT) and pharmacotherapy. Walking and jogging demonstrated the strongest evidence base, with a pooled effect size of approximately SMD = -0.62 to -0.67 relative to active controls.

Exercise vs. Psychotherapy

Head-to-head comparisons with CBT are fewer but informative. The Knapen et al. (2015) systematic review found that structured exercise programs and CBT produced comparable symptom reductions for mild-to-moderate depression, with both superior to waitlist controls. Exercise may offer advantages in domains not well-captured by symptom scales alone — including cardiovascular fitness, metabolic health, and the prevention of antidepressant-associated weight gain.

Number Needed to Treat (NNT)

Calculating NNT for exercise as a preventive intervention is challenging due to heterogeneity in study designs and populations, but available estimates are informative. Based on the Schuch et al. (2018) meta-analytic data, the NNT for exercise to prevent one case of incident depression is approximately 9-12 in general adult populations — comparing favorably to many pharmacological interventions in primary prevention (e.g., statins for cardiovascular events, NNT ≈ 20-40 over 5 years). For treatment of existing depression, the NNT has been estimated at 4-6 based on response rate differences in RCTs, comparable to SSRI NNTs in the STAR*D trial context.

Adjunctive Effects

When combined with antidepressants, exercise appears to produce additive benefits. The TREAD (Treatment with Exercise Augmentation for Depression) study by Trivedi et al. (2011) demonstrated that augmenting SSRI treatment with a public health dose of exercise (16 kilocalories per kilogram per week) yielded remission rates of 28.3% versus 15.5% for a lower exercise dose — a clinically meaningful difference in a treatment-resistant population.

Major Depressive Disorder (MDD)

As outlined above, exercise has the strongest evidence base in MDD. Effect sizes for exercise interventions in diagnosed MDD range from SMD = -0.50 to -0.80 across meta-analyses — moderate to large effects. The benefits are observed across age groups, genders, and depression severity, though supervised and structured programs consistently outperform unsupervised recommendations.

Anxiety Disorders

A Cochrane review by Stubbs et al. (2017) found that exercise significantly reduced anxiety symptoms compared to control conditions, with moderate effect sizes (SMD ≈ -0.40 to -0.58). Generalized anxiety disorder (GAD) and panic disorder show the strongest responses. For panic disorder specifically, aerobic exercise has been shown to reduce anxiety sensitivity — the fear of anxiety-related bodily sensations — through interoceptive exposure mechanisms. Patients repeatedly experience elevated heart rate and somatic activation in a safe context, which parallels and may enhance exposure-based psychotherapies.

Post-Traumatic Stress Disorder (PTSD)

Emerging evidence supports exercise as an adjunctive intervention for PTSD, with a meta-analysis by Rosenbaum et al. (2015) finding significant reductions in PTSD symptom severity (SMD = -0.35). Proposed mechanisms include enhanced extinction learning (exercise-induced BDNF facilitates fear extinction consolidation), reduced hyperarousal through autonomic regulation, and improved sleep architecture.

Psychotic Disorders

Aerobic exercise programs in schizophrenia and related disorders have demonstrated improvements in negative symptoms, cognitive functioning, and hippocampal volume. Firth et al. (2017) meta-analyzed 10 RCTs and found that exercise significantly improved total symptom severity, negative symptoms, and social functioning. Aerobic exercise specifically improved cognitive domains of working memory, attention, and social cognition — areas poorly addressed by antipsychotic medication. These findings are particularly significant given the metabolic side-effect burden of second-generation antipsychotics and the reduced life expectancy (15-20 years) observed in schizophrenia.

Substance Use Disorders

Exercise activates reward circuitry through dopaminergic mechanisms, providing a natural reinforcer that may partially substitute for substance-related reward. RCTs have demonstrated that exercise reduces cravings, improves abstinence rates, and attenuates withdrawal symptoms across alcohol, nicotine, and stimulant use disorders. Lynch et al. (2013) found that aerobic exercise as an adjunct to standard treatment produced abstinence rates approximately 10-15 percentage points higher than standard treatment alone at 12-week follow-up in alcohol use disorder.

Cognitive Decline and Dementia Prevention

Physical activity represents one of the strongest modifiable risk factors for Alzheimer's disease and all-cause dementia. The Livingston Commission (2020) estimated that physical inactivity accounts for approximately 1.6% of the population-attributable fraction for dementia globally. The Erickson et al. (2011) hippocampal volume findings provide a direct neuroanatomical substrate for these protective effects.

Understanding moderators and predictors of exercise response is critical for clinical decision-making. The evidence identifies several key factors:

Factors Associated with Greater Benefit

• Higher baseline symptom severity: Counterintuitively, individuals with more severe depression show larger absolute symptom reductions from exercise, though adherence is more challenging. This mirrors pharmacotherapy findings (higher baseline severity predicts larger drug-placebo differences).
• Supervised and structured programs: Meta-analyses consistently show larger effects for supervised exercise (SMD ≈ -0.72) compared to unsupervised exercise recommendations (SMD ≈ -0.33). The DOSE (Depression Outcomes Study of Exercise) trial found that supervised group sessions outperformed home-based programs.
• Aerobic fitness improvements: Participants who achieve measurable improvements in cardiorespiratory fitness (VO2max) tend to show greater symptom reductions, suggesting a dose-response relationship between physiological adaptation and clinical benefit.
• Social context: Group-based exercise programs show additive effects beyond the physiological effects of exercise alone, likely mediated through social connectedness, behavioral activation, and accountability.
• Comorbid physical conditions: Individuals with comorbid metabolic syndrome, cardiovascular disease, or obesity often show particularly large mental health benefits, likely reflecting the shared inflammatory and metabolic pathways linking these conditions to depression.

Factors Associated with Reduced Benefit or Poor Adherence

• BDNF Val66Met polymorphism (Met carriers): Individuals carrying the Met allele at the BDNF Val66Met locus show attenuated neurotrophic responses to exercise, potentially reducing antidepressant efficacy. While this remains a research finding rather than a clinical tool, it illustrates the genetic basis for individual response variability.
• Anhedonia and psychomotor retardation: These core features of severe melancholic depression create a motivational paradox: the patients most likely to benefit from exercise are often least able to initiate and maintain it. This necessitates graded approaches and integration with behavioral activation frameworks.
• Chronic pain conditions: While low-intensity exercise is beneficial for most chronic pain states, acute pain exacerbation during exercise can create negative reinforcement cycles that undermine adherence.
• Unsupported self-directed programs: Simply advising patients to "exercise more" without structured support is largely ineffective. Dropout rates in unsupervised exercise trials for depression range from 30-50%, compared to 10-20% in supervised programs.
• Personality factors: Low self-efficacy, external locus of control, and high neuroticism predict poorer exercise adherence in psychiatric populations.

Long-Term Outcome Data

The most critical question for prevention science is durability. Blumenthal et al. (2007) demonstrated lower relapse rates at 10-month follow-up for exercise compared to sertraline. A 2-year follow-up of the HUNT cohort data showed that the protective effects of regular physical activity persisted across the follow-up period, with consistent dose-response relationships. However, the preventive effects appear to be contingent on maintained activity — when individuals cease exercising, protective effects attenuate over 6-12 months, suggesting that exercise functions more as an ongoing protective behavior than a time-limited treatment with enduring remission effects.

In clinical practice, exercise as a mental health intervention rarely occurs in diagnostic isolation. The prevalence of psychiatric comorbidity and co-occurring medical conditions shapes both the rationale for and implementation of exercise interventions.

Depression-Anxiety Comorbidity

Approximately 50-60% of individuals with MDD have comorbid anxiety disorders (DSM-5-TR). Exercise is uniquely positioned for this comorbidity because it simultaneously addresses both conditions through overlapping but also distinct mechanisms — serotonergic and BDNF-mediated pathways for depression, plus interoceptive exposure and autonomic regulation for anxiety. The NNT for exercise in comorbid anxiety-depression populations has not been specifically calculated but clinical trials suggest comparable or larger effect sizes in comorbid versus pure diagnostic groups.

Metabolic Syndrome and Depression

The bidirectional relationship between metabolic syndrome and depression is well-established, with shared inflammatory, HPA axis, and lifestyle pathways. Approximately 25-30% of individuals with type 2 diabetes have clinically significant depression (Anderson et al., 2001). Exercise simultaneously addresses both conditions, improving insulin sensitivity, reducing visceral adiposity, lowering inflammatory biomarkers, and alleviating depressive symptoms. For this population, exercise may be the single most efficient intervention available.

Antipsychotic-Induced Metabolic Effects

Second-generation antipsychotics (particularly olanzapine and clozapine) are associated with weight gain of 4-10 kg in the first year and substantially elevated rates of metabolic syndrome. The CATIE trial demonstrated that metabolic side effects were a leading cause of medication discontinuation in schizophrenia. Structured exercise programs can partially attenuate these metabolic effects while simultaneously improving the negative and cognitive symptoms least responsive to pharmacotherapy.

Insomnia Comorbidity

Sleep disturbance is present in approximately 75% of depressive episodes and is both a symptom and a risk factor for recurrence. A meta-analysis by Kredlow et al. (2015) found that both acute bouts and chronic exercise training improved sleep quality, with moderate effect sizes (d = 0.47 for chronic exercise). The mechanisms include body temperature regulation (thermoregulatory cascade hypothesis), circadian rhythm entrainment, and reduction of pre-sleep cognitive arousal.

Chronic Pain and Depression

The comorbidity of chronic pain and depression ranges from 30-50% depending on the pain condition and clinical setting. Exercise-induced hypoalgesia, mediated through endocannabinoid and opioidergic systems, provides dual benefit. However, clinicians must carefully titrate exercise intensity to avoid pain flares that could worsen both pain and depression through negative reinforcement.

The individual-level clinical evidence translates into substantial population-level implications. The burden of mental disorders is enormous: the Global Burden of Disease study estimates that depression and anxiety disorders are the leading and sixth leading causes of disability worldwide, respectively, accounting for over 80 million disability-adjusted life years (DALYs) annually.

Population-Attributable Fractions

The population-attributable fraction (PAF) — the proportion of cases that would be prevented if the risk factor were eliminated — provides a measure of public health impact. Harvey et al. (2018) estimated a PAF of 12% for depression attributable to physical inactivity, meaning roughly 1 in 8 cases of depression could theoretically be prevented through population-wide increases in physical activity. For context, this is comparable to the PAFs for other major modifiable risk factors: childhood adversity (PAF ≈ 20-30%), social isolation (PAF ≈ 5-15%), and problematic alcohol use (PAF ≈ 5-10%).

Economic Analysis

The economic case for exercise-based prevention is compelling. Depression alone costs an estimated $326 billion annually in the United States (Greenberg et al., 2021), including direct treatment costs, workplace productivity losses, and suicide-related costs. Exercise-based interventions are relatively inexpensive to implement at scale — walking programs require minimal infrastructure, and group-based programs leverage economies of scale. Cost-effectiveness analyses have estimated that exercise referral schemes produce incremental cost-effectiveness ratios (ICERs) of $2,000-$10,000 per QALY gained, well below conventional willingness-to-pay thresholds ($50,000-$100,000 per QALY).

Implementation Science Challenges

Despite strong evidence, the translation from efficacy (controlled trials) to effectiveness (real-world implementation) faces substantial barriers:

• Adherence: Population-level physical activity guidelines are met by only approximately 25-30% of adults in high-income countries (WHO, 2022). In psychiatric populations, adherence rates are even lower — estimated at 15-20% for sustained moderate-intensity activity.
• Health equity: Physical inactivity disproportionately affects populations already at elevated psychiatric risk: low socioeconomic status, racial/ethnic minorities, individuals with disabilities, and residents of low-resource neighborhoods. Exercise interventions that fail to address structural barriers (access to safe spaces, time constraints, cultural factors) risk widening health disparities.
• Clinical integration: Exercise referral schemes (e.g., the UK's Exercise on Referral programs) have shown mixed results in systematic reviews, partly due to insufficient dose prescription, lack of ongoing support, and poor integration with mental health services. The evidence suggests that co-located exercise professionals within mental health settings produce better outcomes than external referral.

The FITT-VP framework (Frequency, Intensity, Time, Type, Volume, Progression) provides a structure for exercise prescription in psychiatric contexts:

• Frequency: 3-5 sessions per week; 3 sessions appears to be the minimum frequency for sustained neurobiological adaptations
• Intensity: Moderate intensity (50-70% HRmax, RPE 12-14 on the Borg scale) as the evidence-based minimum; vigorous intensity permissible for those with adequate cardiovascular fitness
• Time: 30-60 minutes per session; sessions as brief as 20 minutes show significant acute mood effects, but durations of 45+ minutes are associated with larger chronic adaptations
• Type: Aerobic exercise (walking, jogging, cycling, swimming) has the strongest evidence base; resistance training is independently effective; yoga and mind-body practices show additional benefits for anxiety and trauma-related disorders
• Volume: The minimal effective dose for depression prevention is approximately 8.8 MET-hours/week (equivalent to 150 minutes of brisk walking); optimal dose may be higher (17.5 MET-hours/week or ~300 minutes/week of moderate activity)
• Progression: Begin at 50% of the target dose for sedentary individuals and increase by no more than 10-20% per week; behavioral activation principles should guide initial engagement

Safety Considerations

Exercise is generally safe for psychiatric populations, but several considerations merit attention. Individuals on lithium must maintain adequate hydration due to the risk of lithium toxicity during dehydration. Beta-blockers limit heart rate response and require perceived exertion-based monitoring rather than heart rate targets. Individuals with eating disorders require careful exercise monitoring to prevent compulsive exercise. Acute psychotic episodes may impair judgment regarding safety during exercise. Pre-exercise cardiovascular screening (using the PAR-Q+ or equivalent) is recommended for previously sedentary individuals over 45 years of age or those with cardiovascular risk factors.

Motivational Strategies

Given the motivational deficits inherent to depression, successful exercise programs for psychiatric populations typically incorporate: motivational interviewing techniques; behavioral activation principles that schedule exercise as a valued activity rather than relying on mood-contingent motivation; self-determination theory-informed approaches that support autonomy, competence, and relatedness; and gradual dose escalation to build self-efficacy before targeting optimal therapeutic doses.

Precision Exercise Medicine

Analogous to precision pharmacotherapy, emerging research aims to match exercise prescriptions to individual biological profiles. The BDNF Val66Met genotype, inflammatory biomarker status (high vs. low CRP), and HPA axis phenotype may eventually inform personalized exercise prescriptions. For example, individuals with elevated CRP (>3 mg/L) — a subgroup that tends to respond poorly to SSRIs — may be particularly responsive to anti-inflammatory effects of exercise. This "immunometabolic" subtyping of depression represents a promising avenue for targeted exercise intervention.

Acute vs. Chronic Effects

A methodological distinction that remains underexplored is the separation of acute mood-enhancing effects (single-bout improvements lasting 2-6 hours, mediated by endocannabinoids and monoamine release) from chronic adaptations (structural brain changes, sustained neurotrophic upregulation, HPA axis recalibration). Clinical prescription should leverage both: frequent sessions to maximize cumulative acute effects while sustained training drives chronic neuroplastic adaptations.

Exercise and the Gut-Brain Axis

An emerging frontier involves exercise-induced changes in gut microbiome composition. Preliminary evidence suggests that exercise increases microbial diversity and the abundance of short-chain fatty acid (SCFA)-producing bacteria, which modulate neuroinflammation and blood-brain barrier integrity through vagal and humoral pathways. This represents a novel mechanism linking exercise to mental health, though human evidence remains in early stages.

Digital and Technology-Assisted Delivery

App-based exercise programs, wearable activity trackers, and virtual group exercise platforms offer scalability advantages, particularly post-COVID-19. Early RCTs of app-delivered exercise interventions for depression show modest but significant effects (SMD ≈ -0.30), with engagement and retention remaining key challenges.

Limitations of Current Evidence

Despite the strength of the overall evidence base, important limitations must be acknowledged:

• Blinding: Exercise RCTs cannot be double-blinded, introducing expectancy and attention biases that likely inflate effect sizes relative to pill-placebo controlled pharmacotherapy trials.
• Publication bias: Funnel plot analyses in exercise meta-analyses suggest moderate publication bias, with Duval and Tweedie trim-and-fill analyses reducing pooled effect sizes by approximately 10-15%.
• Active control selection: Many exercise RCTs use waitlist or treatment-as-usual controls, which inflate effect sizes. Studies using active attention controls (e.g., stretching, social groups) show smaller but still significant effects.
• Adherence in real-world settings: The gap between trial conditions (supervised, incentivized, selected populations) and clinical reality remains substantial.
• Causal inference: While Mendelian randomization studies strengthen causal claims, residual confounding remains possible in observational studies. Reverse causation — depression causing inactivity rather than inactivity causing depression — is partially addressed by prospective designs but not entirely eliminated.

While the evidence for exercise is robust, responsible clinical practice requires clear delineation of its limitations as a standalone intervention:

• Severe MDD with suicidal ideation: Exercise should not be used as a primary treatment for individuals with active suicidal ideation or severe functional impairment requiring urgent pharmacological or psychotherapeutic intervention. It is appropriate as an adjunct.
• Bipolar disorder: Exercise in manic or hypomanic phases may exacerbate sleep deprivation and arousal. Exercise prescriptions for bipolar disorder should be mood-state contingent, with monitoring for signs of overactivation.
• Treatment-resistant depression (TRD): While the TREAD study demonstrated benefit for exercise augmentation in partial SSRI responders, TRD typically requires multimodal approaches including medication optimization, evidence-based psychotherapy, and potentially neuromodulation (TMS, ketamine). Exercise augmentation in this context is evidence-based but insufficient alone.
• Eating disorders: Exercise in anorexia nervosa can be contraindicated due to compulsive exercise patterns and the risks of excessive energy expenditure in an already malnourished state. Careful clinical judgment is required.

The diagnostic pitfall most relevant to exercise-as-intervention is the assumption that presenting symptoms represent a disorder that exercise alone can address. A patient presenting with fatigue, low mood, and psychomotor slowing may have MDD amenable to exercise — or may have hypothyroidism, anemia, sleep apnea, or an emerging neurodegenerative condition requiring differential medical workup. Exercise should complement, not replace, thorough diagnostic assessment.

The evidence base for exercise as a mental health prevention strategy has reached a level of maturity that demands clinical integration rather than merely academic acknowledgment. Key takeaways for clinical practice include:

• Exercise is causally linked to reduced depression incidence, with Mendelian randomization and large prospective studies converging on a risk reduction of 17-26% for physically active individuals.
• The dose-response curve shows greatest marginal benefit at low doses: even 75 minutes/week of moderate activity provides meaningful protection, with optimal effects at 150-300 minutes/week.
• Effect sizes for exercise in treating existing depression (SMD = -0.50 to -0.80) are comparable to SSRIs and CBT, with NNTs of 4-6 for treatment and 9-12 for prevention.
• Neurobiological mechanisms are specific and multi-system: serotonergic, dopaminergic, neurotrophic (BDNF), anti-inflammatory, and HPA axis pathways all contribute.
• Supervised, structured programs consistently outperform recommendations alone, and integration within mental health services is associated with better outcomes than external referral.
• Benefits extend across diagnostic categories — depression, anxiety, PTSD, psychotic disorders, substance use, and cognitive decline — making exercise a rare transdiagnostic preventive intervention.
• Individual variability exists and is influenced by genetic factors (BDNF Val66Met), symptom profiles (anhedonia reduces adherence), and social-contextual factors.

Physical activity should be considered a first-line preventive intervention for common mental disorders and an evidence-based adjunctive treatment across the severity spectrum. The clinical imperative now is not generating more efficacy data but solving implementation challenges — ensuring equitable access, integrating exercise into mental health care pathways, and developing strategies to sustain engagement in the populations that stand to benefit most.