Acute BDNF and cortisol response to low intensity exercise and following ramp incremental exercise to exhaustion in humans
Introduction
Research in cell and molecular biology has revealed that brain derived neurotrophic factor (BDNF) and cortisol (COR) have impact on neurogenesis in the brain of all mammalian species including humans (Jacobs et al., 2000). These factors directly change the basic structure and morphology of the brain. It is well established that chronically elevated COR levels inhibit the proliferative activity in the hippocampus (Gould et al. 1992; Sapolsky 1993). Dysfunction of neuronal plasticity, neurogenesis or remodeling was related to stress or increased levels of glucocorticoid hormones, and associated with the etiology of a variety of diseases including Alzheimer’s or mood disorders (Duman 2002; Fahnestock et al. 2002; Jacobs et al. 2000).#
On the other hand, recently BDNF was shown to promote and improve neuronal plasticity, retard neuron cell death, induce neural regeneration and stimulate neuronal survival particularly in motor and sensory neurons of the peripheral and central nervous system (Kishino et al. 2001; Molteni et al. 2004; Skup 1994). Since BDNF is readily crossing the blood–brain barrier in both directions and taking into account the presence of a high-capacity saturable transport system, it can be assumed that peripheral circulating BDNF is transported into the brain and contributes to neuroplasticity (Pan et al. 1998; Poduslo and Curran 1996). Apart from the key role as mediator of neuronal plasticity, BDNF is also known for a functional role in the periphery e.g. in repair processes at the site of traumatic injury. Serum BDNF protein is stored in platelets and is released upon agonist stimulation (Fujimura et al., 2002). Platelet BDNF is not acquired from the megakaryocyte precursor cells or pituitary gland, but is probably acquired from other sources via the blood circulation. The peripheral sources of BDNF in humans might be immune cells, endothelial cells and vascular smooth muscle cells, however, as BDNF secretion in the central nervous system contributes to the amount of circulating BDNF alterations of peripheral BDNF levels might also partly reflect the variance of secretion in the human brain (Lommatzsch et al., 2005).#
The potential of exercise to affect neural function is becoming increasingly recognized. It is well established that acute anaerobic exercise increases glucocorticoid (GC) production (Kraemer and Ratamess, 2005). It has also been implicated that excessive endurance training may impair the hypothalamo–pituitary–adrenal axis (HPAA) function in elderly athletes, as these athletes revealed elevated plasma COR concentration response to CRH stimulation despite suppression with dexamethasone (DEX) (Heuser et al. 1991; Strüder et al. 1999). It was postulated that this dysfunction of HPAA may occur in response to repeated short-term episodes of hypercortisolaemia induced by exercise stress and that frequent periods of transient excessive HPAA activation in athletes might decrease corticotrophic sensitivity to negative feedback signal based on hippocampal neuronal cell degeneration mediated via reduced corticosteroid receptors (Heuser et al. 1991; Sapolsky 1993).#
On the other hand, there is evidence that dynamic exercise has beneficial effects on brain health. Despite consensus that exercise reduces the probability to get depression, Alzheimer's or dementia, there are few studies covering the mechanisms how exercise produces these putatively beneficial effects. Recently, there is growing evidence that exercise promotes brain neuroplasticity. The beneficial effect of exercise on neuroplasticity may be related to exercise-induced increments in neuroplastic factors such as BDNF, but the mechanisms or mode and threshold required for this effects, respectively, are not yet defined. Taking into account that in the adult nervous system BDNF plays a predominant role in neuronal plasticity (Egan et al., 2003), the potential role of peripheral exercise-induced upregulation of BDNF may help to increase the brain's resistance to damage and neurodegeneration that occurs with age.#
As shown by Oliff et al. (1998), the hippocampal full-length BDNF mRNA expression is rapidly influenced by physical activity as significant increases in expression levels as soon as 6 h after voluntary wheel running were found. It has also been demonstrated in rats that BDNF concentrations are elevated by voluntary exercise in a running wheel, and this was associated with cell proliferation and neurogenesis (Van Praag et al. 1999ab). Repeated wheel running in rats increased the concentration of BDNF in the hippocampus, cerebellum, cortex and lumbar spinal cord. Within days of the exercise training BDNF mRNA alterations were found in rats, and these changes were sustained after several weeks without exercise (Cotman and Engesser-Cesar, 2002). Rhodes et al. (2003) also reported an increase of BDNF concentration and an augmentation of neurogenesis in the hippocampus of mice having access to a running wheel. In mice selectively bred for high levels of wheel running, the correlation between running distance and neurogenesis was lost as a result of a possible ceiling effect suggesting that there might be limits to exercise-induced neurogenesis in training programs. Other animal studies have revealed that BDNF response to a single exercise demand varies depending on the mode of exercise (e.g. Oliff et al., 1998). As the only available study in humans showed that like in animals exercise at moderate intensity induces an augmentation of BDNF (Gold et al., 2003), we hypothesized that in humans BDNF response to exercise also differs depending on the mode of exercise.#
Thus, the aim of the present study was to investigate BDNF concentration responses to 10 min of moderate aerobic exercise and to a following ramp incremental exercise of similar duration aiming to reach maximal oxygen uptake, and to compare these responses with exercise-induced alterations in serum COR. Furthermore, given the evidence that the release of some hormones and growth factors during exercise are triggered by changes in acid-base homeostasis (e.g. Rojas Vega et al. 2003 2006; Luger et al. 1992), we assessed the question if changes of pH, HCO3, pCO2, BE and lactate concentration are related to BDNF alterations. The research is of potential relevance for a more precise determination of the effect of exercise on neural plasticity in humans.#
Results
Power output, maximal oxygen uptake, heart rate and lactate
During the ramp test to exhaustion lasting 7.3±1.1 min, subjects reached a peak power output of 431.3±57.9 W, a relative peak power output of 5.9±0.7 W/kg body wt and a maximal oxygen uptake of 56.6±8.6 mL/kg/min. After the warm-up period with moderate exercise, the heart rate was significantly augmented to 128.4±7.0 beats per min while LA had not changed from concentration at rest (Fig. 1). Maximal heart rate (189.3±10.3 beats per min) was reached at exhaustion while maximal LA concentrations were found 3 and 6 min post exercise.#
Blood gases
Capillary HCO3−, pH, pCO2 and BE are shown in Fig. 1. There were no significant differences in any of the blood gas parameters between rest and the end of the warm-up period. Exercise to exhaustion induced a significant drop of capillary HCO3−, pH, pCO2 and BE at cessation of exercise and during the recovery period. The lowest values of these blood gas parameters were found between 3 min and 10 min of the recovery period.#
Serum BDNF and cortisol
Serum BDNF and COR concentrations (Fig. 2) did not change during 10 min of moderate exercise in the warm-up period. During the ramp test to exhaustion serum BDNF concentration significantly increased. No significant differences were found between BDNF concentrations at rest before exercise and during the recovery period. Significant differences were found between values at end of exercise and values 10 and 15 min post exercise. COR concentrations were increased from rest 10 and 15 min post incremental exercise. No significant correlations were found between BDNF or COR, respectively, and changes in blood lactate and blood gases.#
Discussion
There is accumulating evidence that exercise can promote brain health and function by protecting neurons and improving neuronal plasticity (Cotman and Berchtold, 2002). Several animal studies have demonstrated that exercise increases neurotrophic factors such as BDNF (Neeper et al. 1995 1996). For example, voluntary wheel-running for several days – which closely mimics the choices humans have when exercising – was shown to enhance BDNF production in the hippocampus and other CNS areas of rats, suggesting a mediation role of exercise-induced neurotrophin release (Cotman and Engesser-Cesar, 2002). Recently, it was also shown in humans that blood BDNF concentration is augmented after 30 min of moderate exercise at 60% of maximal oxygen uptake assessed during incremental exercise to exhaustion with 25 W steps every 2 min (Gold et al., 2003). The novel finding of the present study is that in humans immediately after moderate aerobic exercise over 10 min blood BDNF concentration is not altered whereas immediately after a following ramp incremental exercise to exhaustion of similar duration increased blood BDNF concentration is found, pointing to exercise-intensity dependent transient neurotrophic factor induction in humans. The present study also shows that the acute response to high-intensity exercise differs between BDNF and COR. Following the exercise-induced augmentation, BDNF concentration rapidly returns during the recovery phase to basal concentration while increased COR concentration was found 10 and 15 min post exercise.#
Several mechanisms and/or mediators are involved in the exercise-induced HPAA activation (Tremblay and Chu, 2000). Whereas lactate seems to play a role in the stimulation of COR secretion during exercise (Kraemer and Ratamess, 2005), the required signal for activation of BDNF is yet unknown. In the present study, the comparison between BDNF kinetic and changes in acid-base status as well as blood lactate following short time anaerobic exercise suggests that others factors are involved in the BDNF release. Despite still decreased capillary HCO3− pH, pCO2 and BE or increased capillary lactate, respectively, BDNF concentration had already returned to basal levels post exercise. As a co-release between BDNF and serotonin (5-HT) has been demonstrated (Mattson et al. 2004ab), it can be speculated that during prolonged aerobic endurance demands augmented free fatty acids concentration in the blood increases free tryptophan and thus also brain 5-HT syntheses (for review see Strüder and Weicker, 2001), thereby affecting BDNF concentration. Similarly, during anaerobic exercise, chemosensitivity-related brain 5-HT activation (Rojas Vega et al., 2006) might mediate the exercise-induced BDNF augmentation under exercise.#
It is beyond the scope of the present study to prove if the single spike in BDNF serum concentration is sufficient to produce a similar rise of BDNF in the brain. However, as BDNF crosses the blood–brain barrier (Pan et al. 1998; Kastin et al. 1999), it seems likely that the exercise-induced BDNF elevation in the blood leads to an increased entry of this neurotrophin into the CNS. Moreover, a significant increase in BDNF mRNA expression was found in rats following a single bout of 6 h or 12 h running, respectively (Oliff et al., 1998). The expression of the transcriptional forms of BDNF was also upregulated. The authors concluded that the rapid neurotrophin response to these stimuli suggests that exercise-induced BDNF might have a significant role in the brain resistance to damage and neurodegeration.#
The elevation of BDNF mRNA in rats persisted for 10 days after a single 6 h exercise demand (Oliff et al., 1998). In humans serum levels of BDNF rapidly declined after 30 min of moderate exercise (Gold et al., 2003). The present study shows that exercise to exhaustion induces an elevation of BDNF peaking at cessation of exercise, followed by a fast return to baseline level post exercise. The physiological function of the acute transient augmentation of serum BDNF concentration may contribute to promote synaptic plasticity and to improve cognitive functions (Van Praag et al. 1999b; Vaynman et al. 2004) as well as to enhance exercise performance (Rhodes et al., 2003). Neural regeneration and remyelination might also be promoted because the activation of the BDNF signal is also required for the priming effect of exercise on axonal regeneration (Ebadi et al., 1997). In the present study, the subjects sport activities were primarily cycling, basketball, soccer or swimming on average for 45 min with moderate intensity 3 days per week. It remains to be determined in future studies if the BDNF response to exercise differs between athletes and more sedentary individuals. This might also particularly be of relevance in the light of the growing health concerns associated with inactive life-styles and the possible role of exercise as a therapeutic tool for disorders which involve a decrease in BDNF.#
Concerning COR, our results are in accordance with a recent study by Ratamess et al. (2005) showing that exercise-induced COR increase depends on exercise protocols which elicit high lactate augmentations. This also corresponds with previous studies in endurance athletes, which have revealed that intensive anaerobic demands cause an increase in plasma COR concentration, while moderate aerobic exercise does not elicit any response. Prolonged elevations of COR concentrations until 30 min after anaerobic exercise have also been reported before (Kraemer and Ratamess 2005; Schwarz and Kindermann 1990).#
It has been demonstrated that training status and age affect baseline and post stimulation cortisol secretion in humans. For example, increased cortisol concentrations have been shown to occur in trained athletes, possibly as an adaptive mechanism (Mastorakos and Pavtlatou, 2005). In a study by Luger et al. (1987) investigating highly trained athletes, a mild basal hypercortisolism and attenuated cortisol response after exhaustive exercise were found in comparison to untrained and moderately trained subjects. In the present study, the training status of the subjects was not accompanied by adaptive changes of the baseline cortisol concentration which was found to be within normal ranges (10.8±3.4 μg/dL). Moreover, in our study the hormonal responses to moderate and high-intensity exercise were similar to those of other studies investigating moderately trained subjects with similar physical characteristics (i.e. Barreca et al., 1988). However, the conclusion of the present study should be restricted to the specific age group and training status of the subjects which were investigated.#
GCs have been suggested to accelerate the process of biological aging and impair cognitive function (Sapolsky, 1993). It has been hypothesized that aging per se is probably not associated with changes in the HPAA function but rather that the extent of lifelong cumulative effects of acute stress-induced elevations in GCs and free radicals determine HPAA dysregulation in the elderly. As intensive exercise training also consists of repeated transient episodes of acute stress leading to HPAA activation, it may, consequently, also promote these functional modifications. A possible effect of elevation of COR concentration on neuroplasticity is also the inhibition of the birth of new cells as had been demonstrated in in vitro studies (Gould et al., 1992). The extent of hippocampal neuron degeneration and memory deficits was shown to be positively correlated with adrenal activity augmentation in rat (Landfield et al., 1978).#
Smith et al. (1995) have demonstrated in rats that exposure to corticosterone decreases BDNF mRNA. However, the effect was small, and either confined to the dentate gyrus or more than 2 h of exposure was required. In the present study, exercise to exhaustion induced an augmentation of plasma cortisol concentration similar to the one found after mild stress. As a single acute exercise demand causes only a short-term transient elevation of cortisol, it seems rather unlikely that these exercise-induced effects of cortisol are sufficient to impair neuroplasticity in humans. However, frequently repeated high physical stress or high levels of glucocorticoids might produce deleterious effects on the neuroplasticity. Acute or chronic administration of glucocorticoids was shown to reduce BDNF expression suggesting a role for glucocorticoids in neurodegeneration (Cosi et al. 1993; Smith et al. 1995). In rats subcutaneous injection with 10 mg corticosterone once or seven consecutive days reduced BDNF mRNA in the dentade gyrus to levels which were 65–75% of those in the controls (Smith et al., 1995). The effects of glucocorticoids are anatomically selective. In contrast to the effects in hippocampus, abolition of endogenous glucocorticoid secretion by adrenalectomy resulted in severe loss of dentate granule cells. This phenomenon is reversible by the addition of exogenous glucocorticoids (Sloviter et al., 1993), suggesting that glucocorticoids are necessary for the maintenance of dentade gyrus granule neurons and basal expression of BDNF mRNA in the hippocampus and cerebral cortex (Barbany and Persson, 1993). Recently, McMillan et al. (2004) showed that chronically enhanced cortisol induces an augmentation in BDNF in primates. These conflicting results might be related to the duration of glucocorticoid exposure or differences in the used animal models.#
Future studies should further investigate whether prolonged augmentations of BDNF synthesis can be achieved in humans by acute exercise or exercise training. It was suggested by Mattson et al. (2004ab) that in humans 5-HT and BDNF might mediate the health benefit of exercise because at the molecular level they both stimulate the production of proteins involved in cellular stress adaptation, growth and repair, neurogenesis, learning and memory and cell survival. Future studies should establish a dose–response relationship between exercise and neurotrophic factors, therefore allowing to optimize the health benefit of exercise e.g. in rehabilitation programs. Interesting steps in this direction have already been done in an animal study by Berchtold et al. (2005) showing that exercise on less and alternating days is as effective as daily exercise. It was also found in this study that BDNF protein remained elevated for several days after exercise had ceased and that exercise primes a molecular memory reducing the threshold necessary for BDNF increase during repeated exercise after a resting period of 2 weeks.#
In summary, it was shown in the present study and previous research addressing in humans the acute effects of exercise on blood substrate concentrations that (1) prolonged moderate exercise increases BDNF but not COR concentration, (2) short-term moderate exercise does not immediately alter BDNF and COR concentrations and (3) short-term high-intensity exercise to exhaustion following short-term moderate exercise results in an immediate increase of BDNF concentration and a rapid decline to basal value as well a prolonged elevation of COR concentration.#
Experimental procedures
Subjects
Eight recreational athletes (age: 24.6±1.3 year, height: 178.9±6.3 cm, weight: 73.4±8.1 kg) were recruited for the study which was approved by the local Ethics Committee. They were informed about the aim of the study and received written detailed explanation about all tests, potential discomforts, risks and procedures employed in the investigation. Afterwards the subjects gave their written consent. The subjects were medically screened by a physician before participation. The examination of the subjects included medical history, blood analysis and resting electrocardiogram. All persons were in good health and free of any physical condition placing them at risk for participation in the study. During pre-examination each subject also carried out a cardiopulmonary exercise test (CPX) consisting of incremental exercise on a cycle ergometer (Ergoline 9500) with 40 W steps every 5 min. A 12-leads ECG and blood pressure were registered on a ZAN 680 Ergospirometer system (Oberthulba, Germany) throughout the test. No evidence of pathological cardiopulmonary limitations was found in any of the examined volunteers. To minimize exagerated hormonal response due to novel stress condition in the main trials, the subjects were familiarized with all test procedures and all employed equipment before their arrival on the main test day.#
Exercise test
The subjects refrained from strenuous physical exercise for 2 days before the main experimental session. At the time the exercise test was scheduled, the subjects had received a standardized vegetarian meal with water ad libitum at 9 p.m. the night before test days, had fasted overnight and had had a standardized breakfast (1 roll with 25 g cherry jam, 300 mL decaffeinated fruit tea with 0.02 g dextrose-saccarin sweetener) two or more hours before arrival. The subjects were advised to avoid coffee as well as caffeinated beverages for 24 h prior to testing. After arrival, a catheter was placed in a superficial antecubital vein of the right arm for venous blood sampling. The catheter was kept patent by 1 mL of normal saline and the subjects rested for 30 min before a baseline sample (5 mL) was obtained. Capillary samples from the hyperaimized earlobe were also collected for blood gas and lactate analysis.#
The exercise started for all subjects between 9:30 a.m. and 10:30 a.m. on a cycle ergometer. Each subject's individual seat and handlebar heights were reproduced on a SRM cycle ergometer (Jülich, Germany) during the test. After the subject was given instructions and the feet were secured in the pedal straps, a 10 min warm-up period followed at a power of 2 W·kg−1 body wt during which the pedal rate was 70 rpm. At the end of the warm-up period, the subjects proceeded directly with the ramp exercise test. This test consisted of an initial 2 min of cycling at 2 W·kg−1 body wt followed by incremental exercise to exhaustion with 25 W steps every 30 s. It was ensured that subjects had reached maximal aerobic capacity by fulfilling the following criteria: reaching of maximal heart rate (HRmax: 220 minus age in years), rating of perceived exertion (RPE) higher than 17, increase of respiratory exchange ratio (RQ) higher than 1.1, ratio of respiratory minute volume in mL and the oxygen uptake during that same minute higher than 30–35, no further increase of the quotient of oxygen uptake in mL and pulse rate during the same minute, and increase of capillary blood lactate concentration (LA) above 8 mmol/L. After exhaustion was reached, an active recovery period of 15 min succeeded during which the subjects pedaled in a free-wheel fashion solely against the intrinsic frictional resistance of the ergometer. The fixed times for the collection of capillary and venous blood samples were after the 10 min warm-up period on the cycle ergometer, at maximum exertion during the ramp exercise test and at 3, 6, 10 and 15 min post exercise. The heart rate was continuously registered with a heart rate monitor (Polar x-Trainer, Helsinki, Finland). Ventilatory measurements during exercise were obtained on breath-by-breath basis, averaged and reported for 10 s intervals.#
Blood measures
Blood bicarbonate concentration (HCO3−), base excess (BE), pH and carbon dioxide pressure (pCO2) were analyzed in arterialized blood of the earlobe using the blood gas analyzer AVL OPTI CCA (Georgia, USA). A deep puncture into the earlobe was made with a lancet so that a free flow of blood exuded from the wound without squeezing the area. A preheparinized capillary tube (200 μL) was used to collect the blood and placed deep into the drop of blood. To maintain the anaerobic conditions, the sample was immediately inserted without air bubbles into the electrode chambers for the blood gas measurement.#
Capillary lactate concentration was measured in duplicate using the lactate analyzer Biosen 5130 EKF (Magdeburg, Germany). By means of calibrated 20 μL one-time micropipette (BRAND, Hagen, GER), capillary blood samples were extracted from the earlobe, filled in 2 mL Safe-Lock vessels and blended manually with its 1000 μL system solution. The analysis of the samples was based on the enzymatic–amperometric measuring method and was administered with the help of the technical device “Biosen C line” (EKF-diagnostic, Barleben, GER). The technical device was equipped with EKF chip-sensors which transform the sample concentration into evaluable, electrical signals.#
Venous blood samples (5 mL) were drawn in pre-chilled serum venipuncture tubes. Serum was separated by centrifugation (3000 rpm for 10 min at 4°C) and stored in Eppendorf tubes at −70°C. Serum BDNF concentrations were analyzed by enzyme immunoassay using ELISA kits by Chemicon (Temecula, CA, USA) with a detection range from 7.8 pg/mL to 500 pg/mL and no cross-reactivity with others neurotrophins. The intra-assay and inter-assay variations were ±3.7% (125 pg/mL) or ±8.5% (125 pg/mL), respectively. Serum COR concentrations were analyzed using the ES 300 analyzer and ELISA kits by Boehringer (Germany). The measuring range was between 0.036 μg/dL and 63 μg/dL. The intra-assay and inter-assay variations were ±1.3% (7.31 μg/dL) or ±1.1% (46 μg/dL), respectively.#
Data analysis
Data are presented as mean±standard deviation. Factorial analysis of variance with repeated measurement was applied for statistical analysis of the data using the software program easystat 2.0 (Lüpsen, Cologne, Germany). This program automatically carries out sphericity test for homogenity of variances. In case of non-homogenity of variances, multiple comparison of means is done with Bonferroni adjustment. Significance level for all analyses was set at p<0.05.#
Figures and Tables
References
27. S.A.NeeperF.Gomez-PinillaJ.ChoiC.W.CotmanExercise and brain neurotrophinsNature3731995109
43. M.S.TremblayS.Y.ChuHormonal response to exerciseSports Endocrinology2000Humana PressNew Jersey130