Physical inactivity has many known deleterious effects on the body. Heart disease, stroke, and metabolic diseases such as type 2 diabetes are just a few examples of the consequences of a sedentary lifestyle (2). As such, exercise interventions could prove beneficial in combating these and other conditions. It is becoming increasingly evident that exercise also is beneficial to the brain and nervous system. An endogenous substance that plays a central role in the health of neurons is brain-derived neurotrophic factor (BDNF). BDNF is a 27.0-kDa homodimer of two 13.5-kDa subunits linked by noncovalent interactions (26). BDNF acts via its tyrosine kinase receptor, TrkB, to promote neuronal differentiation and survival (5). Exercise studies with rodents have shown that treadmill running increased hippocampal BDNF (21) and have implicated BDNF as one of the central influences on brain plasticity (20).
In the rodent spinal cord injury (SCI) model, Hutchinson et al. (9) have shown that animals that underwent SCI and remained sedentary experienced significant drops in BDNF mRNA in the spinal cord and soleus muscle relative to a control group. SCI rats that underwent treadmill training recovered from allodynia, an associated condition of sensory dysfunction, and BDNF mRNA in the spinal cord and soleus muscle were not significantly different than control group levels. Using human subjects, Colcombe and colleagues (4) reported that among an older adult population, there was a reduction in the age-based decline in tissue density in the frontal, parietal, and temporal cortices as a function of cardiovascular fitness. These exercise-associated outcomes would be consistent with the neurotrophic (5) and/or neurogenic (31) actions of BDNF. Assays for BDNF or other markers of brain health, however, were not performed.
Indeed, comparatively fewer studies have examined the response of BDNF to exercise in humans. A study by Gold et al. (6) has reported increases in serum BDNF in healthy controls and multiple sclerosis patients in response to acute exercise, but only one exercise intensity was selected. To date, the present study is the only study to examine systemic BDNF responses to acute exercise in humans while incorporating varying exercise intensities.
Physical activity also has been shown to effect cognitive function. Vaynman et al. (32) report that animals that exercised for a short period outperformed their sedentary counterparts on a measure of learning and recall, the Morris water maze test. BDNF mRNA levels were also examined (as were the levels of mRNA of cAMP response-element binding protein, CREB, an associated molecular effector). The animals that were the fastest learners and that had the best recall also had the highest expression of BDNF and CREB mRNA. Human-based studies also have shown that exercise has a positive effect on measures of cognitive performance. Khatri et al. (14) examined the effects of exercise training on cognitive function in a group of depressed older individuals. They report that individuals who participated in aerobic exercises demonstrated greater increases in certain cognitive processes (memory, executive functioning) compared with participants taking antidepressant medication. Relative to rodent-based investigations, however, studies involving human participants have largely not examined possible biochemical effectors involved in the exercise-induced improvements in cognitive function.
The potential impact of exercise on the nervous system is significant; if exercise in humans can lead to BDNF-mediated maintenance or improvement of neural tissue, then exercise may become a potent weapon in combating neuromuscular disorders and the diseases of affect and cognition that correlate with a sedentary lifestyle. The purpose of this study was to determine the effects of acute exercise on serum BDNF levels in humans and the relationship between exercise intensity and BDNF responses. Additionally, the relationship between the change in serum BDNF and the change in cognitive function scores after an acute exercise bout was examined. We hypothesized that an acute exercise bout would result in an increase in serum BDNF and that the increase would be exercise intensity dependent. We further hypothesized that the change in BDNF would directly correlate with the change in cognitive function scores.
The study population consisted of 15 subjects (11 male, 4 female; age 25.4 ± 1.0 yr; height 174.7 ± 1.9 cm; weight 71.0 ± 3.1 kg; BMI 23.1 ± 0.6 kg·m−2; mean ± SE) who were physically active (i.e., involved in collegiate intramural activities) and, therefore, of adequate health to participate in the exercise protocols, but not engaged in competitive sports. According to a detailed medical history questionnaire, all subjects were free of cardiopulmonary, metabolic, or musculoskeletal disease, mental disorders, and were nonsmokers. All subjects signed an informed consent form, and the institutional review board approved the study.
General testing procedures.
Participation in the study required three visits to the exercise physiology laboratory. During the first visit, the subjects performed a graded exercise test (GXT) on a stationary cycle. The second and third visits consisted of 30-min endurance rides at fixed intensity levels performed in a randomized fashion. For each endurance ride, data from the individual's GXT was used to determine the work rate that corresponded to the target intensity. Participants were instructed to forgo strenuous exercise for 24 h before each exercise bout, and food, caffeine, and alcohol intake were prohibited for 3 h before each ride. All females who participated were in the follicular phase of their menstrual cycle, to control for time-dependent variations in estrogen levels. Additionally, all testing was performed at the same time of day to control for circadian influences.
During the first visit, the subjects became familiarized with the exercise testing equipment and then performed a GXT on a cycle ergometer (Lode, Corival) to volitional fatigue. Cardiopulmonary and metabolic parameters (maximal work rate, maximal oxygen consumption (V˙O2max), maximal heart rate, and respiratory exchange ratio (RER)) were determined on a breath-by-breath basis with samples averaged for 30-s intervals (MedGraphics, CPX/D). Blood pressure was determined before, during, and after the GXT via auscultation. Heart rate and rhythm were monitored during the GXT via electrocardiography (Quinton Instruments, Model 4000). The ventilatory threshold (VTh) was determined by the V-slope method (33).
The subjects performed two 30-min exercise bouts in a randomized fashion on the cycle ergometer at a power output that corresponded to the VTh plus 10% (high intensity) or minus 20% (low intensity) as determined for each individual from the GXT. Metabolic, ventilatory, and heart rate parameters were collected at 10-min intervals during the rides and were averaged as described above. There was a minimal rest period of 48 h between the GXT and the first endurance ride, and there was a similar rest period between the first and second endurance rides.
Blood sampling and analysis.
A 10-mL blood withdrawal via aseptic technique from the antecubital vein was taken before and after the GXT and endurance rides for the determination of blood lactate and serum BDNF levels. Blood lactate was measured with an Accusport portable lactate analyzer (Mannheim Boehringer) that was calibrated to the manufacturer's specifications before each test. The blood was allowed to clot (BD Vacutainer Plus SST) and was then centrifuged (Eppendorf Centrifuge 5804) for 12 min at 1300 relative centrifugal force. The supernatant was decanted and stored in a −80°C freezer until analysis. Serum BDNF was assayed using a Chemikine BDNF sandwich enzyme-linked immunosorbent assay (ELISA) kit. The sensitivity as reported in the ELISA kit literature was 7.8 pg·mL−1. The reported intraassay and interassay variations were ± 3.7 and ± 8.5%, respectively.
Cognitive function tests.
A cognitive function assessment, the Stroop color and word test (7,27), was administered pre- and postexercise for the GXT and the two endurance rides. The Stroop color and word test consists of three sections, all of which are variations of the same theme; subjects are required to read out loud, as quickly and accurately as possible, from a list of items for 45 s. Test 1, the word test, requires that subjects read from a page containing three printed words: red, green, or blue. The words are printed in black ink and are listed in random fashion in five columns of 20 words each. No word is allowed to follow itself in a column. Test 2, the color test, similarly consists of 100 items that are all written as XXXX and printed either in red, green, or blue ink. The subject states the color of the ink for each of the series of four X's. Test 3, the color-word test, consists of the same 100 words (red, green, or blue) as the first test, but printed in colored ink, and the color of the ink and the printed word are never the same (for example, the word "blue" could be printed in red or green ink, but never blue ink). The subject is required to say the color of the ink that the word is printed in, not the printed word itself. For the three tests, a higher score reflects a better performance. Subjects performed a practice session to become familiar with the Stroop test before it was initially administered pre-GXT.
Postexercise performance parameters from the endurance rides and serum BDNF and Stroop scores from the GXT were analyzed with paired t-tests. The serum BDNF and Stroop scores from the endurance rides were analyzed with a two-way repeated-measures ANOVA with main effects examined for condition (VTh − 20 and VTh + 10) and time (before and after exercise). Significant main effects were further analyzed with Student-Newman post hoc tests. Results are presented as means (± SE). The relationships between the pre- to postexercise changes in serum BDNF and blood lactate, and the changes in serum BDNF and Stroop scores, were examined with the Pearson product-moment correlation. A P value of < 0.05 was considered significant. Statistical analyses were conducted using SigmaStat for Windows (Jandel Scientific Software, SPSS Inc., Chicago, IL).
Graded exercise and endurance ride responses.
Table 1 lists maximal values from the GXT for six measured variables. The values for RER and lactate meet accepted criteria for a true maximal effort (RER > 1.1 and lactate >8 mM) (1). The mean VTh occurred at 57.6 ± 2.3% V˙O2max. Table 2 compares indicators of exercise intensity between the two endurance ride conditions. All variables were significantly greater at the VTh + 10 condition relative to the VTh − 20 condition (P < 0.001; P value smaller than that required from Bonferroni correction).
Serum BDNF responses.
The intraassay variation between duplicate samples was 1.653%. The GXT yielded a significant elevation in the postride serum BDNF (Fig. 1) relative to baseline (30% increase) (P < 0.001). Additionally, there was a significant correlation between the change in blood lactate and the change in serum BDNF responses from the GXT (Fig. 2). For the endurance rides, an ANOVA for repeated measures on BDNF levels revealed a significant effect of time (F1,14 = 13.09; P < 0.01), whereas neither condition nor the interaction of the factors was significant. The VTh + 10 endurance ride resulted in a significant increase in serum BDNF (Fig. 3) relative to baseline (13% increase) (P < 0.05, Student-Newman-Keuls post hoc test), whereas the VTh − 20 ride did not yield a significant postexercise increase. There were no significant correlations between the pre-post changes of serum BDNF and blood lactate for either the VTh − 20 or VTh + 10.
The results from the Stroop word and color tests, but not the color-word test, were significantly greater post-GXT compared with their respective pre-GXT scores (P < 0.05) (Fig. 4). For the endurance rides, an ANOVA for repeated measures on the Stroop word and color tests revealed a significant effect of time (F1,14 = 9.76; P < 0.01 and F1,14 = 33.72; P < 0.001, respectively), whereas neither condition nor the interaction of factors was significant for either test. An ANOVA for repeated measures on the Stroop color-word test revealed a significant effect of time (F1,14 = 7.16; P < 0.05). The VTh + 10 resulted in a significant increase in the color-word score (P < 0.05, Student-Newman-Keuls post hoc test) (Fig. 5), whereas the VTh − 20 ride did not yield a significant postexercise increase. Additionally, the color-word test from the endurance rides yielded the only significant interaction between condition and time (F1,14 = 6.45; P < 0.05). There were no significant correlations between change in the Stroop word, color, or color-word scores and change in BDNF for the GXT, VTh − 20, or VTh + 10.
The present study examined how acute exercise bouts at different intensities affected serum levels of BDNF and assessments of cognitive function. The intensity at the end of the GXT is rarely, if ever, encountered by an individual during recreational activities and fitness training. However, assessments of BDNF and cognitive function were made before and after the GXT, yielding, along with measurements from the endurance rides, a collection of data from exercise bouts of disparate intensities. Target work rates of VTh − 20% and VTh + 10% were selected for the fixed-intensity endurance rides because the VTh corresponds to the work rate at which exercise becomes difficult (25). The VTh + 10 was characterized as being between somewhat hard and hard according to the Borg scale of perceived exertion (14.2 ± 0.4; mean Borg rating of perceived exertion (RPE) ± SE), whereas the VTh − 20 was described as light (10.8 ± 0.4; mean RPE ± SE). Given the disparate intensities between the two fixed-intensity exercise conditions, as determined by both subjective responses and physiological measurements, we feel that this protocol provided valuable insight into an intensity-dependent response of a 30-min exercise bout to yield an increase in BDNF.
Reported average serum BDNF values for healthy subjects seem to vary between studies and may reflect methodological differences. The preexercise average serum BDNF value of this protocol (18,168 ± 1193 pg·mL−1 (mean ± SE)) falls approximately halfway between the reported averages (of healthy controls) from two published psychiatric studies (12,29) and is comparable with the reported median value of healthy adults (22,600 pg·mL−1) from a recent study notable for its large sample size (N = 140) (19). Thus, we feel that the average preexercise serum BDNF values presented in this study are consistent with resting serum BDNF levels in the established literature and provide a valid resting baseline for the examination of the effects of acute exercise on the serum levels of this neurotrophin.
The GXT and the VTh + 10 yielded significant increases in serum BDNF levels relative to baseline. The relative magnitude of the pre- to postexercise percent increases were as follows: VTh − 20 (10%) < VTh + 10 (13%) < GXT (30%), thus indicating an intensity-dependent induction of serum BDNF. To the best of our knowledge, we are the first to report an exercise intensity-dependent induction of serum BDNF in human participants. The change in blood lactate did correlate with the change in BDNF, but only for the GXT.
The intensities of the VTh − 20 and the VTh + 10 endurance rides employed in this study corresponded to 55.9 and 75.2% V˙O2max, respectively. The VTh − 20 did not yield a significant pre- to postexercise increase in the serum BDNF concentration, whereas the higher-intensity VTh + 10 did result in significantly elevated BDNF relative to baseline. These results are consistent with the results of Gold et al. (6), who reported significant increases in serum BDNF for multiple sclerosis patients and healthy controls after a 30-min cycle ride at 60% V˙O2max, which, according to their pre- and postexercise lactate measurements, exceeded the anaerobic threshold. Exercise of sufficient intensity and/or duration may increase the serum BDNF concentration via onset of acidosis relative to the resting state.
BDNF is a neurotrophin; it is integral in the maintenance of the healthy neuronal phenotype. Additionally, BDNF has been shown to acutely modulate presynaptic neurotransmitter release (10) and evoked excitatory postsynaptic currents via TrkB receptors (17) and to directly induce neuronal depolarization (11). It is becoming widely recognized that exercise is directly beneficial to brain health and function, probably via a BDNF-mediated mechanism. BDNF is found and made in many locations throughout the body (18,34) in addition to its namesake source of production, the brain. In this study, we assayed for peripheral serum BDNF levels via a blood withdrawal from the antecubital vein. We view it as likely that, in accordance with rodent studies, exercise in humans also results in elevated BDNF levels in the brain, because in rats there is a strong correlation between serum and cortical BDNF levels (r = 0.81) (13), and it has been reported that BDNF undergoes bidirectional transport across the blood-brain barrier (22,23). Although this study did not directly assess brain levels of BDNF, it seems plausible that BDNF in the brain increased because of onsite production, transport from the periphery, or both. Given this speculation, it is especially heartening that the subjects were able to bring about a serum BDNF increase after the VTh + 10, which was a submaximal exercise bout characterized as falling between somewhat hard and hard according to the Borg scale (3), and that all participants were able to perform for 30 min. Exercise prescriptions for the purpose of neurological health could, conceivably, be of an intensity that would discourage few from participating because of the level of difficulty.
Adequate BDNF levels are essential for cognitive function (30). Previous studies have reported post-acute exercise bout Stroop score improvements (8,15). Therefore, pre- and postexercise Stroop tests were administered, and the relationship between those results and BDNF changes was examined. The GXT word and color scores increased significantly relative to baseline, whereas the color-word score did not. The post-endurance ride word and color test scores increased relative to baseline for the VTh - 20 and VTh + 10, whereas an increase in the color-word score occurred only after the VTh + 10. The process of cognitive interference is associated specifically with the Stroop color-word task, and the anterior cingulate has been cited as a pivotal brain structure involved in its execution (24). Although rodent BDNF exercise studies have focused to a large extent on hippocampal BDNF responses, BDNF is found in many brain regions, including the anterior cingulate (28). Therefore, we felt that the color-word results also warranted analysis as we examined possible associations between pre- to postexercise changes in BDNF and the Stroop color and word test. It is worthwhile to note that the VTh + 10, as opposed to the VTh − 20, yielded increases in both BDNF and the Stroop color-word score. However, the increase in the VTh + 10 Stroop color-word performance could be attributed to an increase in activation and attention after the higher-intensity exercise (8); it is not necessarily related to the exercise-induced BDNF increase. Indeed, in all five instances where there was a significant BDNF and Stroop increase (GXT: word, color; VTh + 10: word, color, color-word), examination of the coefficient of correlation when comparing the pre- to postexercise BDNF changes with the Stroop changes did not yield significant results. The pre- to postexercise increases in BDNF, where significant, occurred in most participants (15/15 for GXT, 14/15 for VTh + 10), whereas the Stroop scores, even where significant increases occurred, were not as uniform (i.e., for the Stroop word test from the VTh + 10, 11/15 increased relative to baseline, two individuals' scores did not change, and two scores decreased). Likely, this greater variability of change in Stroop scores precluded significant associations with the changes in BDNF values. Future investigations with a larger sample size may result in significant correlations between BDNF and Stroop responses.
This study examined the response of serum BDNF and cognitive function to acute bouts of exercise at varied intensities. Chronic physical activity in humans has been shown to benefit cognitive processes (16) and brain health (4). It is tempting to speculate that repeated pulses of exercise-induced BDNF are key phenomena in the neurological and cognitive improvements that occur as a function of regular chronic exercise. Future studies should examine the effect of chronic exercise training on the interplay between cardiovascular fitness, resting levels of BDNF, and cognitive function in human subjects. Further, investigations into the response of BDNF to an acute exercise bout before and after chronic training might shed additional light on the relationship between cardiovascular fitness, circulating BDNF levels, and cognitive and neurological function.
We gratefully acknowledge Daniel Lindley for his assistance in the Pathology Laboratory. The authors thank the subjects for their enthusiastic participation in this study. This work was supported by a Research Enhancement Fund Grant, Texas Tech University, and a Faculty Incentive Grant, Texas Tech University Health Sciences Center.
1. American College of Sports Medicine. Interpretation of clinical exercise test data. In: ACSM's Guidelines for Exercise Testing and Prescription
, R. M. Otto (Ed.). Philadelphia, PA: Lippincott Williams and Wilkins, pp. 115-129, 2006.
2. Booth, F. W., M. V. Chakravarthy, S. E. Gordon, and E. E. Spangenburg. Waging war on physical inactivity: using modern molecular ammunition against an ancient enemy. J. Appl. Physiol.
3. Borg, G. A. Psychophysical bases of perceived exertion. Med. Sci. Sports Exerc.
4. Colcombe, S. J., K. I. Erickson, N. Raz, et al. Aerobic fitness reduces brain tissue loss in aging humans. J. Gerontol. A. Biol. Sci. Med. Sci.
5. Cotman, C. W., and C. Engesser-Cesar. Exercise enhances and protects brain function. Exerc. Sport Sci. Rev.
6. Gold, S. M., K. H. Schulz, S. Hartmann, et al. Basal serum levels and reactivity of nerve growth factor and brain-derived neurotrophic factor to standardized acute exercise in multiple sclerosis and controls. J. Neuroimmunol.
7. Golden, C., and S. Freshwater. A Manual for the Adult Stroop Color and Word Test
. Chicago, IL: Stoelting, pp. 1-11, 2002.
8. Hogervorst, E., W. Riedel, A. Jeukendrup, and J. Jolles. Cognitive performance after strenuous physical exercise. Percept. Mot. Skills.
9. Hutchinson, K. J., F. Gomez-Pinilla, M. J. Crowe, Z. Ying, and D. M. Basso. Three exercise paradigms differentially improve sensory recovery after spinal cord contusion in rats. Brain
10. Jovanovic, J. N., A. J. Czernik, A. A. Fienberg, P. Greengard, and T. S. Sihra. Synapsins as mediators of BDNF-enhanced neurotransmitter release. Nat. Neurosci.
11. Kafitz, K. W., C. R. Rose, H. Thoenen, and A. Konnerth. Neurotrophin
-evoked rapid excitation through TrkB receptors. Nature
12. Karege, F., G. Perret, G. Bondolfi, M. Schwald, G. Bertschy, and J. M. Aubry. Decreased serum brain-derived neurotrophic factor levels in major depressed patients. Psychiatry Res.
13. Karege, F., M. Schwald, and M. Cisse. Postnatal developmental profile of brain-derived neurotrophic factor in rat brain and platelets. Neurosci. Lett.
14. Khatri, P., J. A. Blumenthal, M. A. Babyak, et al. Effects of exercise training on cognitive functioning among depressed older men and women. J. Aging Phys. Act.
15. LaManca, J. J., S. A. Sisto, J. DeLuca, et al. Influence of exhaustive treadmill exercise on cognitive functioning in chronic fatigue syndrome. Am. J. Med.
16. Laurin, D., R. Verreault, J. Lindsay, K. MacPherson, and K. Rockwood. Physical activity and risk of cognitive impairment and dementia in elderly persons. Arch. Neurol.
17. Levine, E. S., C. F. Dreyfus, I. B. Black, and M. R. Plummer. Brain-derived neurotrophic factor rapidly enhances synaptic transmission in hippocampal neurons via postsynaptic tyrosine kinase receptors. Proc. Natl. Acad. Sci. U. S. A.
18. Lommatzsch, M., A. Braun, A. Mannsfeldt, et al. Abundant production of brain-derived neurotrophic factor by adult visceral epithelia. Implications for paracrine and target-derived neurotrophic functions. Am. J. Pathol.
19. Lommatzsch, M., D. Zingler, K. Schuhbaeck, et al. The impact of age, weight and gender on BDNF levels in human platelets and plasma. Neurobiol. Aging
20. Molteni, R., Z. Ying, and F. Gomez-Pinilla. Differential effects of acute and chronic exercise on plasticity-related genes in the rat hippocampus revealed by microarray. Eur. J. Neurosci.
21. Neeper, S. A., F. Gomez-Pinilla, J. Choi, and C. W. Cotman. Physical activity increases mRNA for brain-derived neurotrophic factor and nerve growth factor in rat brain. Brain Res.
22. Pan, W., W. A. Banks, M. B. Fasold, J. Bluth, and A. J. Kastin. Transport of brain-derived neurotrophic factor across the blood-brain barrier. Neuropharmacology
23. Pan, W., and A. J. Kastin. Polypeptide delivery across the blood-brain barrier. Curr. Drug Targets CNS Neurol. Disord.
24. Pardo, J. V., P. J. Pardo, K. W. Janer, and M. E. Raichle. The anterior cingulate cortex mediates processing selection in the Stroop attentional conflict paradigm. Proc. Natl. Acad. Sci. U. S. A.
25. Persinger, R., C. Foster, M. Gibson, D. C. Fater, and J. P. Porcari. Consistency of the talk test for exercise prescription. Med. Sci. Sports Exerc.
26. Rosenfeld, R. D., L. Zeni, M. Haniu, et al. Purification and identification of brain-derived neurotrophic factor from human serum. Protein Expr. Purif.
27. Stroop, J. R. Studies of interference in serial verbal reactions. J. Exp. Psychol.
28. Takahashi, M., O. Shirakawa, K. Toyooka, et al. Abnormal expression of brain-derived neurotrophic factor and its receptor in the corticolimbic system of schizophrenic patients. Mol. Psychiatry
29. Toyooka, K., K. Asama, Y. Watanabe, et al. Decreased levels of brain-derived neurotrophic factor in serum of chronic schizophrenic patients. Psychiatry Res.
30. Tsai, S. J. Brain-derived neurotrophic factor: a bridge between major depression and Alzheimer's disease? Med. Hypotheses
31. van Praag, H., B. R. Christie, T. J. Sejnowski, and F. H. Gage. Running enhances neurogenesis, learning, and long-term potentiation in mice. Proc. Natl. Acad. Sci. U. S. A.
32. Vaynman, S., Z. Ying, and F. Gomez-Pinilla. Hippocampal BDNF mediates the efficacy of exercise on synaptic plasticity and cognition. Eur. J. Neurosci.
33. Wasserman, K., W. L. Beaver, and B. J. Whipp. Gas exchange theory and the lactic acidosis (anaerobic) threshold. Circulation
34. Yamamoto, H., and M. E. Gurney. Human platelets contain brain-derived neurotrophic factor. J. Neurosci.