Secondary Logo

Journal Logo


Exercise Enhances and Protects Brain Function

Cotman, Carl W.; Engesser-Cesar, Christie

Author Information
Exercise and Sport Sciences Reviews: April 2002 - Volume 30 - Issue 2 - p 75-79
  • Free



Exercise is known to impact nearly every system in the body. Benefits of regular exercise include improved cardiovascular health, greater bone mineral density (BMD), and decreased risk for cancer, stroke, and diabetes, yet many people still choose not to participate. The extra incentive many sedentary individuals need to begin exercising might be found in new research showing that exercise is a means of enhancing and protecting brain function. Unlike erythrocytes (red blood cells), which have a lifespan of 120 d, the millions of neurons in the brain must last the 80+ yrs of a lifetime. Age-related neuron dysfunction and degeneration cause cognitive decline and personality changes in some cases. Many people turn to supplements, such as gingko biloba, as a solution to cognitive impairments. Unfortunately, no proof exists that any one supplement can maintain brain function at an optimal level during the aging process. One could imagine, though, that the brain can produce molecules that nourish neurons and ensure overall brain health. The ideal candidate molecule would have the ability to protect neurons, to increase neuronal plasticity, to enhance learning, and to assist in the overall maintenance of the brain. Such molecules do exist and fall into a class of proteins called growth factors. Several specific growth factors, called neurotrophic factors, target neurons. (Neurons are the key cells in the brain and trophin is the Greek word for nutrition, hence the term neurotrophin.)

One neurotrophic factor, brain-derived neurotrophic factor (BDNF), is capable of mediating the beneficial effects of exercise on brain plasticity. BDNF is a member of a neurotrophic family that supports the health and functioning of the primary neuronal type in the brain, glutamatergic neurons. These neurons are the major projection neurons connecting cognitive, sensory, and motor brain regions. BDNF is synthesized in the cell body and transported back and forth to synapses, where neuronal communication occurs. In addition to mediating synaptic efficacy, BDNF also promotes differentiation, neurite (process) extension, and the survival of a variety of neuronal populations in culture and in vivo. BDNF is regulated by neuronal activity, and mechanisms that induce BDNF, such as exercise, enhance learning.


We developed an animal model to examine the effects of exercise on the brain, specifically to see if exercise regulated BDNF. We chose voluntary wheel running as it closely mimics the choices humans have when exercising because animals dictate the time, speed, and distance of running throughout the experiment. Some rats voluntarily run up to an astounding 20 km in one night! At the beginning of the experiment, we predicted that with physical activity the major sites of gene induction in the brain would be in motor and sensory regions. However, the brain region showing the earliest and most sustained neurotrophic upregulation in response to exercise was the hippocampus. The hippocampus is a brain region important in learning and memory and is a target for early deterioration in neurodegenerative diseases like Alzheimer’s disease (AD).

Initial experiments were designed to measure the effects of acute exercise bouts on BDNF activity. In situ hybridization measures of BDNF messenger RNA (mRNA) found a 20% increased abundance of BDNF mRNA from control levels in the hippocampus after 2–7 nights of running (Fig. 1) (6,7). Increases in BDNF mRNA levels occur when running activity is above a minimum threshold level of approximately 500 m·d−1. BDNF mRNA increases with greater distances run up to a ceiling at several km·d−1. Experiments looking at 3 wk of exercise demonstrated that the effect of exercise on BDNF mRNA levels was not merely a short-term, transient effect. Indeed others have shown that increases in mRNA for BDNF and its receptor trkB are observed in the hippocampus after 6 wk of voluntary wheel running (15). Recently, experiments showed that the increase in abundance of BDNF mRNA is paralleled by increases in BDNF protein levels (Fig. 1).

Figure 1
Figure 1:
Voluntary running is an easily monitored and quantifiable method of exercise. A. Animals have access to running wheels for a defined length of time. During this time computers record the distance animals run. B. Enzyme-linked immunosorbent assay quantification of BDNF protein levels (pg·mL−1) in hippocampal lysate after 5 d of running shows an increase in protein in the running (EX) animals compared with sedentary (SED) animals. Error bars represent SEM (*P = 0.0244). C and D. Coronal sections depicting in situ hybridization of BDNF mRNA in the rat hippocampus after 6 d of voluntary running wheel activity (C) and sedentary animals (D). Note the increase in mRNA in the dentate gyrus (DG) and CA1-CA3 regions of the hippocampus and in the cortex (ctx) after exercise.

Maintenance of cerebral BDNF levels is important for effective neural function and longevity (Table 1) (10). There is strong evidence that BDNF is important in the ability of synapses to encode change, as measured by long-term potentiation (LTP). In LTP, patterns of synaptic stimulation lead to a long-term potentiation of the synaptic response; therefore, LTP is regarded as synaptic analog of learning and memory. LTP is impaired in BDNF-deficient mice, but can be restored with BDNF infusion. This suggests that LTP formation is BDNF dependent. In addition to its effects on plasticity (LTP), BDNF demonstrates its neuroprotective role by promoting survival of hippocampal, striatal, and septal neurons in culture and in vivo as well as by protecting the brain against insults such as ischemia and axotomy. Indeed, as predicted, animals that exercised showed improved learning (14) and increased resistance to stroke (2).

Table 1
Table 1:
The neurotrophic factor, BDNF, participates in numerous functions in the brain and is best known for its roles in neuroplasticity and neuroprotection


In addition to BDNF, other genes in the brain are likely to be affected by exercise. It is also probable that BDNF interacts with pathways that trigger changes in genes causing synaptic change. High-density oligonucleotide microarrays allow the simultaneous analysis of the activity of ∼5000 genes. Tong et al. (13) discovered that in addition to BDNF, exercise induces changes in other genes known to be associated with neuronal activity, synaptic structure, and neuronal plasticity (Fig. 2). Thus, gene expression profiles suggest that exercise has the ability to enhance brain plasticity and encoding in a manner that may translate directly into structural change of neurons or synapses. Neuronal connections can be strengthened and made more efficient through enhanced synaptic capacity and the addition of new neurons.

Figure 2
Figure 2:
Exercise-induced gene expression changes in rat hippocampi. After 3 wk of running, gene expression changes were detected using Affymetrix high-density oligonucleotide microarray chips (Affymetrix, Inc., Santa Clara, CA). Various significant changes in gene expression can be broadly classified into categories regulating plasticity, metabolism, immune function, and degeneration processes. For details see (13).

Typically, healthy neurons have synaptic connections with thousands of cells. These connections are essential to the successful coordination of all brain activity, including cognition. To promote neural health, synapses need to maintain synaptic protein levels. Decreases in the presynaptic vesicle proteins synaptobrevin, synaptophysin, and synaptotagmin correlate with cognitive decline in the brains of those with AD (12). Given the loss of function likely to be found with decreased synaptic proteins, any means of upregulating synaptic proteins would be a significant finding. We suggest, and literature supports our prediction, that BDNF interacts with a pathway that triggers changes in synaptic proteins. BDNF acts to increase levels of these important synaptic proteins. BDNF mutants have low levels of the synaptic vesicle proteins, synaptophysin and synaptobrevin, and mice with a mutation in the BDNF receptor, trkB, have decreased levels of synaptophysin and synaptotagmin. In addition, mice with a mutation in BDNF have a decreased number of docked synaptic vesicles. Interestingly, infusion of BDNF to the brain of normal mice results in an increase in the number of docked vesicles (10). This suggests that low BDNF levels may result in impaired neuronal communication, whereas increased BDNF levels may improve neuronal communication.

As mentioned earlier, microarray analysis shows that exercise induces changes in a number of genes that regulate synapses (Fig. 2) (13). Genes whose expression was upregulated include those associated with membrane trafficking, vesicle recycling, and synaptic plasticity. For example, synaptotagmin, the synaptic vesicle protein believed to be the calcium sensor necessary to trigger release of the neurotransmitter, was upregulated. In addition, VESL/homer, genes that have been implicated in synaptic growth, and COX-2, a gene believed to be involved in activity-dependent synaptic plasticity, were both upregulated after exercise. These data and others are important because they indicate voluntary exercise enhances synaptic function. Additional research is needed to link these changes directly to BDNF.

Neuronal connections can also be selectively strengthened by the recruitment of new neurons formed from neurogenesis, the process whereby neural stem cells proliferate and differentiate into neurons and glial cells. Until recently, stimuli to induce neurogenesis were unknown. However, an increased number of new neurons were recently discovered in the hippocampus of mice living in an enriched environment where animals are exposed to stimulation such as social interaction, learning, and exercise in running wheels. Further experiments then looked at which component of the enriched environment—exercise or learning—was more important for stimulating neurogenesis. Mice were placed in conditions of swimming, swimming with learning, voluntary wheel running, or an enriched environment. Increased neurogenesis was seen only in the voluntary wheel running group and the enriched environment group (14), suggesting that running alone is sufficient to increase neurogenesis, allowing for the possibility of strengthening neuronal connections.


For various reasons, many women agonize over decisions regarding estrogen replacement. On one hand, estrogen reduces bone loss and protects against the onset of dementia, whereas on the other hand, it may increase the risk of breast cancer. In the brain, estrogen deficits are known to affect neuronal function, survival, and synaptogenesis negatively, as well as to decrease the availability of BDNF in the hippocampus. We sought to determine whether estrogen deprivation and replacement interacted in any way with exercise to influence BDNF gene and protein regulation. Although a few days of running can offset BDNF losses from short-term estrogen deprivation, it cannot offset BDNF mRNA and protein decreases after long-term estrogen deprivation (1). However, long-term estrogen replacement combined with exercise increases BDNF protein levels even more than estrogen replacement alone, providing more evidence that exercise initiates pathways that protect against functional losses (Fig. 3). Interestingly, the amount of voluntary running is dependent on estrogen, such that those without estrogen were significantly less active than those with estrogen. Restoring estrogen returned activity to normal levels. Controlled clinical studies are still needed, but it appears that the benefits of hormone replacement in women may include increased exercise participation and induction of BDNF.

Figure 3
Figure 3:
Estrogen and exercise cooperatively interact to upregulate BDNF levels in the brain. A. Hippocampal BDNF protein is increased after estrogen manipulation and exercise treatment; however, exercise potentiation of BDNF protein is dependent upon estrogen. OVX, ovariectomized; OVX/ER, estrogen-replacement; ex, exercise; sed, sedentary. Error bars represent SEM (*,P < 0.05; **, P < 0.0001). B. Estrogen increases BDNF directly, through enhancing availability of BDNF, and indirectly, by stimulating voluntary running. Both effects raise important considerations for women deciding upon hormone replacement.


The finding that BDNF levels are upregulated with exercise has potential significance for depression, an important clinical field of research. Exercise is often recommended to individuals suffering from depression as a means of improving their psychological well-being. In general, research supports a correlation between improvement in depression and participation in either aerobic or anaerobic exercise. Patients who remain active after clinical interventions report less depression than those who return to sedentary lifestyles. Overall exercise was found to have positive effects on depression for both clinical and healthy participants. Experiments showing that antidepressants increase BDNF mRNA in animals link BDNF to depression. Interestingly, in animal models of depression, BDNF itself has antidepressant effects (11). Our finding that exercise increases BDNF mRNA levels suggests a possible physiological reason for the benefits of exercise in depressed persons. We investigated the combined effect of exercise and antidepressants in our voluntary running model.

Hippocampal BDNF mRNA level was increased with either exercise or the antidepressant tranylcypromine. The combination of exercise with antidepressant administration produced additive increases in BDNF mRNA abundance (Fig. 4) (9). Antidepressant treatment in humans may take up to 2 wk to reach maximal effectiveness. A similar period is required for antidepressant-triggered increases in BDNF mRNA levels. However, when combined with exercise, antidepressants may be effective more quickly. Our research showed that when antidepressant treatment was combined with exercise, BDNF mRNA was increased in as little as 2 d. Thus, exercise may decrease the time required for antidepressants to be effective, a finding that is potentially clinically significant (9).

Figure 4
Figure 4:
Exercise and select antidepressants influence BDNF levels. A. Total BDNF mRNA levels in the dentate gyrus of the hippocampus after 1 wk of treatment with antidepressant, exercise, or the combination of both (expressed as percentage of control values). The combined intervention (exercise with antidepressant) potentiates the effect of either treatment alone. Error bars represent SEM (*P < 0.05). CTR, control; TC, antidepressant tranylcypromine; EX, exercise. B. Exercise and select antidepressants converge upon a shared molecular pathway to influence BDNF levels. Combined treatment potentiates the effect and reduces the time required for a response to treatment.

Another implication of research on BDNF and exercise comes from the finding that if exercise is terminated abruptly, BDNF levels initially fall below normal before returning to control levels by 30 d (15). The decrease in BDNF after a sudden withdrawal of access to wheel running may correlate with the depression often reported by athletes whose training has been unexpectedly interrupted by injury.

Exercise and Antidepressants Together Improve Stress Response

Exposure to stress causes a decrease in BDNF mRNA levels in the hippocampus, which may be linked to depression. We investigated whether exercise and antidepressants could counteract stress-induced decreases in BDNF mRNA (8). Animals underwent a forced swim test (a stressful event) after 1 wk of treatment with tranylcypromine, 1 wk of voluntary exercise, or a combination of these two interventions. Antidepressant administration or exercise alone prevented the BDNF decrease caused by the acute stress. The combination of exercise with antidepressant led to significantly greater increases in hippocampal BDNF mRNA levels than did either treatment alone. It is possible that select antidepressants and exercise converge at a cellular level to promote brain health. The impact of exercise on reducing brain stress responses is an exciting area for future research.


Exercise increases BDNF, aids in treatment of depression, is affected by estrogen levels, and may induce neurogenesis, but why should this motivate the general populace to exercise? The answer is really twofold. First, gross characterizations of diseased brains include cortical atrophy and decreasing synaptic connections, both of which are correlated with cognitive impairments. Secondly, as previously mentioned, BDNF is a neuroprotective factor capable of sustaining neuron viability. The conclusion then, on a systems level, is amazingly simple; exercise results in upregulated BDNF levels, which should lead to less cortical atrophy and hence improved cognitive function, as supported by animal models.

Our argument that exercise can mediate brain function is supported by research from trials with people aged 60–75 yrs. A group of people who underwent aerobic exercise (walking) was compared with a group who underwent anaerobic exercise (stretching) (4). The aerobic group showed improvements in neuropsychological measures of executive control, linking the enhanced ability to increased physical fitness. The argument that improvements might be due to decreased depression levels leading to increased information-processing efficiency is countered by the lack of effect in anaerobic exercisers. If depression levels were a factor, improvements would have been expected in both groups of exercisers.

Exercise may also be a physiological stimulus capable of initiating protective brain mechanisms. Two studies examining activity levels of adults indicate that physical activity and enriched environments may help decrease the risk of cognitive impairments. Comparisons of clinical evaluations, spaced by 5 yrs, of 4615 adults aged 65 yrs and older revealed that high levels of physical activity corresponded with decreased incidence of cognitive impairments (5). This result was particularly significant in women. A similar study examining retrospective surveys from 193 possible or probable participants with AD and 358 healthy controls aged 20–60 yrs concluded that those with AD participated in fewer activities during middle adulthood than did healthy controls (3). Activities examined were categorized as passive, intellectual, and physical, suggesting that participants with more active lifestyles were living in more enriched environments and were less likely to have AD. Additional clinical studies are needed, because nearly all of the research to date has been retrospective.


Exercise is a universal activity capable of inducing change throughout the body. Recently, several exciting findings in the basic sciences have extended this concept to include the brain as an organ that directly benefits from exercise. Changes in gene expression in the hippocampus, a brain region critical for learning and memory, suggest that physical activity initiates modifications in molecular mechanisms supporting the health and enhancing the plasticity of the brain. BDNF availability is critical for many of these mechanisms. Animal models have shown that exercise reduces the extent of damage after brain injury; furthermore, human studies suggest that exercise can delay the onset of AD. No other behavioral or pharmaceutical intervention can make such claims. Although some may exert neuroprotective or neuroplasticity effects, no other intervention has yet been shown to successfully influence both of these endpoints. Understanding the fundamental mechanisms by which exercise can impact brain molecular mechanisms and function will allow for the optimization of human cognitive health and improve exercise compliance.


We thank Cheryl A. Cotman for her artwork. This work was supported by the National Institute on Aging Grant AG13411 and the Christopher Reeve Paralysis Foundation.


1. Berchtold, N. C., Kesslak, J. P. Pike, C. J. Adlard, P. A. Cotman. C. W. Estrogen and exercise interact to regulate brain-derived neurotrophic factor mRNA and protein expression in the hippocampus. Eur. J. Neurosci. 14: 1992–2002, 2001.
2. Carro, E., Trejo, J. L. Busiguina, S. Torres-Aleman. I. Circulating insulin-like growth factor I mediates the protective effects of physical exercise against brain insults of different etiology and anatomy. J. Neurosci. 21: 5678–5684, 2001.
3. Friedland, R. P., Fritsch, T. Smyth, K. A. Koss, E. Lerner, A. J. Chen, C. H. Petot, G. J. Debanne. S. M. Patients with Alzheimer’s disease have reduced activities in midlife compared with healthy control-group members. Proc. Natl. Acad. Sci. U.S.A. 98: 3440–3445, 2001.
4. Kramer, A. F., Hahn, S. Cohen, N. J. Banich, M. T. McAuley, E. Harrison, C. R. Chason, J. Vakil, E. Bardell, L. Boileau, R. A. Colcombe. A. Ageing, fitness and neurocognitive function (Letter). Nature. 400: 418–419, 1999.
5. Laurin, D., Verreault, R. Lindsay, J. MacPherson, K. Rockwood. K. Physical activity and risk of cognitive impairment and dementia in elderly persons. Arch. Neurol. 58: 498–504, 2001.
6. Neeper, S. A., Gómez-Pinilla, F. Choi, J. Cotman. C. Exercise and brain neurotrophins (Letter). Nature. 373: 109, 1995.
7. Neeper, S. A., Gómez-Pinilla, F. Choi, J. Cotman. C. W. Physical activity increases mRNA for brain-derived neurotrophic factor and nerve growth factor in rat brain. Brain Res. 726: 49–56, 1996.
8. Russo-Neustadt, A., Ha, T. Ramirez, R. Kesslak. J. P. Physical activity-antidepressant treatment combination: impact on brain-derived neurotrophic factor and behavior in an animal model. Behav. Brain Res. 120: 87–95, 2001.
9. Russo-Neustadt, A. A., Beard, R. C. Huang, Y. M. Cotman. C. W. Physical activity and antidepressant treatment potentiate the expression of specific brain-derived neurotrophic factor transcripts in the rat hippocampus. Neuroscience. 101: 305–312, 2000.
10. Schinder, A. F., Poo. M. The neurotrophin hypothesis for synaptic plasticity. Trends Neurosci. 23: 639–645, 2000.
11. Siuciak, J. A., Lewis, D. R. Wiegand, S. J. Lindsay. R. M. Antidepressant-like effect of brain-derived neurotrophic factor (BDNF). Pharmacol. Biochem. Behav. 56: 131–137, 1997.
12. Sze, C. I., Bi, H. Kleinschmidt-DeMasters, B. K. Filley, C. M. Martin. L. J. Selective regional loss of exocytotic presynaptic vesicle proteins in Alzheimer’s disease brains. J. Neurol. Sci. 175: 81–90, 2000.
13. Tong, L., Shen, H. Perreau, V. Balazs, R. Cotman. C. W. Effect of exercise on gene-expression profile in the rat hippocampus. Neurobiol. Dis. 8: 1046–1056, 2001.
14. van Praag, H., Christie, B. R. Sejnowski, T. J. Gage. F. H. Running enhances neurogenesis, learning, and long-term potentiation in mice. Proc. Natl. Acad. Sci. U.S.A. 96: 13427–13431, 1999.
15. Widenfalk, J., Olson, L. Thoren. P. Deprived of habitual running, rats downregulate BDNF and TrkB messages in the brain. Neurosci. Res. 34: 125–132, 1999.

running; BDNF; depression; estrogen; neuroplasticity; neuroprotection

©2002 The American College of Sports Medicine