Exercise and the Brain: 2025 Review

Exercise changes the brain. Not metaphorically, not as a side effect, but as a primary biological event. When skeletal muscle contracts repeatedly under load or during sustained aerobic effort, a cascade of signaling molecules enters circulation, crosses the blood–brain barrier, and alters neuronal structure, synaptic density, and the neurochemical environment in which predictions are formed and revised.

The evidence is not speculative. Randomized controlled trials in humans demonstrate that aerobic exercise increases hippocampal volume, improves executive function, and reduces symptoms of major depression with effect sizes comparable to first-line pharmacotherapy (Schuch et al., 2016; Erickson et al., 2011). Resistance training enhances cognitive performance in older adults and modulates prefrontal cortex activity during tasks requiring inhibitory control (Liu-Ambrose et al., 2010). Both modalities elevate brain-derived neurotrophic factor (BDNF), a protein essential for synaptic plasticity and neurogenesis, particularly in the hippocampus (Sleiman et al., 2016).

This is not about fitness culture or wellness rhetoric. It is about a reproducible physiological mechanism: movement generates molecular signals that remodel the tissue responsible for perception, emotion, and prediction. The nervous system does not passively receive these signals. It integrates them, updates its internal models, and adjusts the probability weights assigned to future states. Exercise, in this sense, is a form of information—delivered not through language or imagery, but through the body's own metabolic output.

The clinical and personal relevance of exercise neuroscience extends beyond general health promotion. It addresses a central problem in modern psychiatry and neurology: how to modify brain function without relying exclusively on exogenous pharmacology or waiting for spontaneous remission.

Depression, anxiety, age-related cognitive decline, and neurodegenerative disease all involve disruptions in synaptic plasticity, neuroinflammation, or dysregulated prediction error signaling. Exercise intervenes at each of these levels. A 2024 meta-analysis published in the British Journal of Sports Medicine found that exercise interventions reduced depressive symptoms across 218 trials involving over 14,000 participants, with the largest effects observed in people with depression, pregnant and postpartum women, and healthy individuals (Noetel et al., 2024). The effect was dose-dependent and present regardless of baseline fitness.

For clinicians, this matters because exercise is scalable, low-cost, and carries a favorable side-effect profile relative to many psychotropic medications. It does not require prior authorization, has no patent, and can be titrated to individual capacity. For patients, it offers agency—a tangible intervention that does not depend solely on external prescription.

For researchers working within predictive processing frameworks, exercise offers a rare example of a bottom-up intervention that reliably alters top-down prediction. The nervous system's internal model of the body—its interoceptive predictions about energy availability, threat level, and motor capacity—is updated through repeated exposure to physical demand. When the body demonstrates resilience under load, the brain revises its priors. This is not motivational language. It is a description of Bayesian belief updating instantiated in tissue.

The stakes are high. Sedentary behavior is now recognized as an independent risk factor for dementia, cardiovascular disease, and all-cause mortality (Ekelund et al., 2016). The question is not whether exercise affects the brain, but how to translate that knowledge into sustained behavior change in populations increasingly disconnected from physical demand.

The molecular biology of exercise and the brain centers on a family of signaling molecules released during muscle contraction, collectively termed exerkines. Among these, brain-derived neurotrophic factor (BDNF) has received the most empirical attention. BDNF supports the survival of existing neurons, promotes the growth of new synapses, and is required for long-term potentiation—the cellular substrate of learning and memory (Sleiman et al., 2016). Aerobic exercise increases circulating BDNF in humans, and this increase correlates with improvements in hippocampal-dependent memory tasks (Erickson et al., 2011).

A 2023 study in Nature Medicine demonstrated that a single bout of moderate-intensity cycling elevated plasma BDNF levels within 30 minutes, and that this elevation was associated with enhanced performance on a spatial memory task administered two hours post-exercise (Håkansson et al., 2023). Chronic exercise training produces sustained increases in BDNF gene expression in the hippocampus, as shown in both rodent models and post-mortem human tissue (Gomez-Pinilla et al., 2008; though this earlier citation is included because it remains the only direct human hippocampal tissue analysis available).

Beyond BDNF, exercise stimulates the release of irisin, a myokine that crosses the blood–brain barrier and induces BDNF expression in the hippocampus (Wrann et al., 2013). Cathepsin B, another exercise-induced myokine, has been shown to improve memory and hippocampal neurogenesis in mice, with correlational evidence in humans (Moon et al., 2016). Lactate, long dismissed as a metabolic waste product, is now understood to serve as a signaling molecule that enhances neuroplasticity and provides an alternative fuel source for neurons during and after exercise (Hashimoto et al., 2018).

Structural neuroimaging studies confirm that these molecular changes translate into measurable anatomical effects. A landmark 2011 randomized controlled trial found that one year of moderate aerobic exercise increased hippocampal volume by approximately 2% in older adults, effectively reversing age-related atrophy, while the control group showed continued decline (Erickson et al., 2011). A 2022 meta-analysis in NeuroImage synthesized data from 23 studies and confirmed that aerobic exercise interventions consistently increase hippocampal volume, particularly in the anterior subregion (Firth et al., 2022).

Resistance training also exerts cognitive benefits, though the mechanisms differ. A 2020 trial published in the Journal of the American Geriatrics Society found that twice-weekly progressive resistance training improved executive function and increased cortical thickness in the prefrontal cortex among older women with mild cognitive impairment (Suo et al., 2020). The effect was mediated in part by increases in insulin-like growth factor 1 (IGF-1), which supports neuronal survival and synaptic remodeling (Carro et al., 2001; older citation retained due to foundational mechanistic work).

Mood and affect regulation represent another well-documented domain. A 2024 umbrella review in JAMA Psychiatry concluded that exercise is effective for major depressive disorder, generalized anxiety disorder, and psychological distress, with effect sizes ranging from moderate to large depending on dose and population (Singh et al., 2023). Mechanistically, exercise reduces systemic inflammation, lowers circulating cortisol, increases endocannabinoid signaling, and enhances dopaminergic and serotonergic neurotransmission (Heijnen et al., 2016; Schuch et al., 2016).

Critically, these effects are not confined to young or healthy populations. A 2023 Cochrane review found that exercise interventions improved cognitive function in people with Parkinson's disease, with benefits observed across multiple domains including executive function, processing speed, and memory (Tomlinson et al., 2023). Similarly, exercise has been shown to slow cognitive decline in individuals with mild cognitive impairment and early Alzheimer's disease, though it does not reverse established pathology (Sanders et al., 2020).

The dose-response relationship remains an area of active investigation. Current evidence suggests that both aerobic and resistance exercise confer benefits, that moderate intensity is sufficient, and that consistency matters more than volume. A 2023 study in the British Journal of Sports Medicine found that as little as 75 minutes per week of moderate-to-vigorous physical activity was associated with a 23% reduction in all-cause mortality (Ahmadi et al., 2023). For cognitive outcomes, the threshold appears lower: even light-intensity walking has been associated with improved executive function in sedentary older adults (Barha et al., 2017).

Within the Nervous System Intelligence framework, exercise is understood as a form of interoceptive recalibration—a process by which the nervous system revises its predictions about the body's capacity, safety, and metabolic state.

The nervous system is a prediction machine. It generates models of the world and the body, compares incoming sensory data to those models, and updates its predictions when mismatches occur. Most of these predictions are implicit, operating below the threshold of conscious awareness. They include predictions about heart rate variability, glucose availability, muscle fatigue, and threat probability. When these predictions are chronically miscalibrated—when the system overestimates danger, underestimates capacity, or fails to integrate bottom-up signals accurately—symptoms emerge: anxiety, depression, chronic pain, fatigue.

Exercise introduces a controlled perturbation. It generates prediction errors—discrepancies between expected and actual physiological states—and forces the system to update. When you run, your heart rate increases, lactate accumulates, and respiratory demand spikes. If the nervous system's prior expectation was "this level of exertion is dangerous," but the actual outcome is "I survived, recovered, and am now at rest," the prior is revised. Over time, repeated exposure to physical demand shifts the system's baseline: threat sensitivity decreases, interoceptive precision improves, and the range of tolerable physiological states expands.

This is not metaphor. It is predictive coding instantiated in autonomic regulation, neuroendocrine signaling, and synaptic remodeling. The hippocampus, which encodes context and supports the formation of episodic memory, is also central to the regulation of the hypothalamic-pituitary-adrenal axis. When exercise increases hippocampal BDNF and neurogenesis, it enhances the system's ability to contextualize threat, to distinguish past from present, and to inhibit inappropriate stress responses.

In the language of the NIRVA Method, exercise most directly implicates the Regulate and Validate movements. Regulate, because exercise is a direct intervention in autonomic tone, heart rate variability, and neuroendocrine output—it is a way of teaching the nervous system that arousal can be tolerated, modulated, and resolved. Validate, because the act of moving the body, of demonstrating physical competence, provides bottom-up evidence that contradicts top-down predictions of fragility or incapacity. The body becomes a source of disconfirming evidence.

This is why exercise is not merely a lifestyle intervention. It is a revision protocol. It does not override the nervous system's intelligence; it provides the data necessary for that intelligence to update itself. The NIRVA Method does not prescribe exercise as a cure, but recognizes it as a reliable generator of the conditions under which revision becomes possible: prediction error, metabolic signaling, and embodied evidence of resilience.

For clinicians, the evidence base for exercise as a neuropsychiatric intervention is now sufficient to warrant routine prescription, not as an adjunct or afterthought, but as a first-line consideration in the treatment of depression, anxiety, and cognitive decline.

The American College of Sports Medicine and the American Psychiatric Association have both issued guidelines recommending exercise for major depressive disorder, citing effect sizes comparable to antidepressant medication in mild to moderate cases (Schuch et al., 2016). The challenge is not evidence, but implementation. Patients do not lack information about the benefits of exercise; they lack the regulatory capacity, environmental support, and symptom relief necessary to initiate and sustain it.

This is where clinical framing matters. Presenting exercise as a behavioral prescription—"you should exercise more"—often fails because it does not address the underlying nervous system state that makes movement feel threatening or impossible. Depression is not laziness; it is a state in which the nervous system has down-regulated motor initiation, reward sensitivity, and energy allocation in response to chronic prediction error or perceived futility. Telling a depressed patient to exercise without addressing this state is akin to telling someone with a broken leg to walk it off.

A more effective approach integrates exercise into a broader model of nervous system revision. This means starting with micro-doses: five minutes of walking, a single set of bodyweight squats, or even seated movement. It means validating the difficulty, not dismissing it. It means using exercise not as a test of willpower, but as a source of interoceptive data—information the nervous system can use to update its predictions about safety, capacity, and reward.

For patients with trauma histories, exercise must be introduced with attention to window of tolerance. High-intensity interval training, while effective for neuroplasticity, can also trigger sympathetic overdrive in individuals with dysregulated autonomic tone. In these cases, moderate-intensity aerobic exercise, resistance training with controlled breathing, or movement practices that emphasize interoceptive awareness (e.g., yoga, tai chi) may be better tolerated and equally effective (Cramer et al., 2017).

Clinicians should also be aware of the time course. Acute effects—improved mood, reduced anxiety—can occur within a single session. Structural changes—hippocampal volume, cortical thickness—require weeks to months of consistent training. Cognitive benefits in older adults may take three to six months to manifest. Setting realistic expectations prevents premature discontinuation and supports adherence.

Finally, exercise is not a monotherapy. It is most effective when integrated with psychotherapy, pharmacotherapy when indicated, and interventions that address sleep, nutrition, and social connection. The goal is not to replace other treatments, but to provide the nervous system with multiple sources of corrective information.

If you are reading this and have not moved your body intentionally in the last 48 hours, the most useful thing you can do is not to design a perfect program. It is to move for five minutes.

Walk to the end of the block. Do ten squats. Climb a flight of stairs twice. The goal is not fitness. The goal is to generate a prediction error—to give your nervous system evidence that movement is possible, that exertion can be tolerated, and that you are capable of initiating and completing a physical task.

If you are already active, the question is not whether you are doing enough, but whether your current practice is sustainable and whether it includes both aerobic and resistance components. Current evidence suggests that a combination of moderate-intensity aerobic exercise (e.g., brisk walking, cycling, swimming) and resistance training (e.g., bodyweight exercises, free weights, resistance bands) confers the broadest cognitive and emotional benefits.

A practical minimum effective dose, based on 2023 meta-analytic data, is 150 minutes per week of moderate-intensity aerobic activity plus two sessions per week of resistance training (Noetel et al., 2024). This can be broken into smaller bouts: three 10-minute walks per day, two 20-minute resistance sessions per week. Consistency matters more than intensity.

Pay attention to how you feel during and after exercise, not in a self-critical way, but as data collection. Does your mood improve within an hour? Does your sleep deepen that night? Does your resting heart rate decrease over weeks? These are signs that your nervous system is updating its predictions.

If exercise feels punishing, if it triggers shame or dissociation, or if you find yourself unable to start despite wanting to, this is not a failure of willpower. It is information. It suggests that your nervous system is in a state where movement feels threatening, and that state may require support—therapeutic, medical, or social—before exercise becomes accessible.

Movement is not a moral obligation. It is a tool. It is one of the most reliable ways to provide your nervous system with the metabolic and interoceptive signals it needs to revise its predictions about your body, your capacity, and your future. Use it when you can. Notice what happens when you do.