- Top of page
During and after maximal exercise there is a 15–30 % decrease in the metabolic uptake ratio (O2/[glucose +1/2lactate]) and a net lactate uptake by the human brain. This study evaluated if this cerebral metabolic uptake ratio is influenced by the intent to exercise, and whether a change could be explained by substrates other than glucose and lactate. The arterial-internal jugular venous differences (a-v difference) for O2, glucose and lactate as well as for glutamate, glutamine, alanine, glycerol and free fatty acids were evaluated in 10 healthy human subjects in response to cycling. However, the a-v difference for the amino acids and glycerol did not change significantly, and there was only a minimal increase in the a-v difference for free fatty acids after maximal exercise. After maximal exercise the metabolic uptake ratio of the brain decreased from 6.1 ± 0.5 (mean ±s.e.m.) at rest to 3.7 ± 0.2 in the first minutes of the recovery (P < 0.01). Submaximal exercise did not change the uptake ratio significantly. Yet, in a second experiment, when submaximal exercise required a maximal effort due to partial neuromuscular blockade, the ratio decreased and remained low (4.9 ± 0.2) in the early recovery (n= 10; P < 0.05). The results indicate that glucose and lactate uptake by the brain are increased out of proportion to O2 when the brain is activated by exhaustive exercise, and that such metabolic changes are influenced by the will to exercise. We speculate that the uptake ratio for the brain may serve as a metabolic indicator of ‘central fatigue’.
The brain takes up O2 and glucose in a ratio close to 6:1, but the ratio becomes reduced during physiological activation as demonstrated with visual (Fox et al. 1988) and mental stimulation in man (Madsen et al. 1995b) and by sensory stimulation in the rat (Madsen et al. 1995a, 1999). Furthermore, this ratio between the brain uptake of O2 and carbohydrate decreases in response to exhaustive exercise and reaches its lowest value in the first minutes of the recovery (Ide et al. 2000). During activation the brain also takes up significant amounts of lactate provided that plasma lactate is also increased as seen during intense exercise (Ide et al. 2000). Thus, the disproportionate higher glucose and lactate to O2 uptake - reflected by a reduced O2/(glucose + 1/2lactate) ratio - becomes even more pronounced, especially in the early recovery period. Such a persisting higher uptake of carbohydrates, is also observed for several minutes following mental activity in man (Madsen et al. 1995b). As cerebral activation is associated with a reduced glycogen content in the brain of the rat (Swanson et al. 1992; Madsen et al. 1995a, 1999), a reduction in the metabolic ratio supports the idea that, especially in the recovery, some of the glucose taken up by the brain is used to replenish brain glycogen stores.
Williamson et al. (1997, 1999) and Nowak (2001) provide evidence that physiological activation of the brain increases with exercise intensity and intense neuronal activity can make energy demand exceed energy production (Sappey-Marinier et al. 1992). Such transient imbalance between energy demand and production may occur in brain regions engaged during exhaustive exercise and, in turn, cause glycogen depletion and termination of exercise. In this way, the metabolic ratio may be a unique metabolic equivalent to ‘central fatigue’ which so far has been difficult to explain.
Although Ide et al. (2000) reported a drop in the metabolic ratio following exhaustive exercise, it is unknown if non-fatiguing exercise also induces an excess carbohydrate uptake in the immediate recovery. The hypothesis of this study was that the O2/(glucose + 1/2lactate) ratio for the brain would be reduced after exhaustive exercise rather than following non-fatiguing exercise. We further considered that in animals, neuronal tissue has the capacity to oxidise amino acids (Larrabee, 1984; Sonnewald et al. 1997) and that such metabolism during exercise could affect the metabolic ratio. Hence, in addition to glucose and lactate, we determined the arterio-internal jugular venous differences (a-v difference) for glutamate, glutamine, alanine and also for glycerol and free fatty acids. Cerebral perfusion was evaluated by the transcranial Doppler determined middle cerebral artery (MCA) mean flow velocity (Vmean).
Exhaustive exercise requires cerebral activation to produce the ‘command’ to exercise and supposedly to perform the integration of the sensory input from working muscle. Yet, their individual contribution to the increased metabolism of the brain is unknown. Central command and sensory input from exercising skeletal muscle may be partially separated by administering a non-depolarising neuromuscular blocking agent that, in a dose-dependent manner, reduces muscle strength. Thus, under the influence of such agents, maintaining a given work rate requires increased discharge frequency or recruitment of additional motor units and consequently a larger central drive (Asmussen et al. 1964; Leonard et al. 1985; Galbo et al. 1987). Such effect is evaluated in an additional experiment.
- Top of page
The experiments were performed according to the Declaration of Helsinki as approved by the Ethics Committee of Copenhagen (KF 01–369/97). Every participant volunteered and informed written consent was obtained. In two studies separate subjects exercised in a semi-supine position on a modified Krogh cycle ergometer at 60 r.p.m. (Galbo et al. 1987), while the a-v difference across the brain for O2 and substrates were determined. Except for incremental exercise, each bout lasted 10 min and was followed by ∼1 h of recovery.
In the first experiment we included twelve subjects of both sexes. However, two females were excluded from the study due to difficulties of drawing from the catheters, leaving one female and nine male subjects to be studied (age 25 ± 4 years; height 179 ± 8 cm; weight 75 ± 9 kg; means ±s.e.m.). The protocol included exercise at light and moderate intensity (heart rate (HR) 90 and 120 beats min−1, respectively) and exercise to exhaustion that was reached by increasing the work rate by 30 W every second minute.
In the second study with partial neuromuscular blockade a further four females and seven males (age 25 ± 1 years; height 174 ± 3 cm; weight 71 ± 3 kg) attended the laboratory twice. First, they performed incremental exercise until exhaustion to determine the maximal O2 uptake (VO2,max, 3.2 ± 0.2 l min−1). On the second day they initially exercised at a work rate of about 40 % of VO2,max (control) followed by a resting period and then exercised at the same work rate but with partial neuromuscular blockade. Subjects were curarised by cisatracurium (Nimbex, Glaxo-Wellcome, Denmark) administered intravenously to a total dose of 1.2 (0.4–2.2) mg. Data for one subject was excluded because of apparatus failure (blood analysis). When muscular relaxation was induced an Ambu-E resuscitator apparatus, neostigmine and atropine were available, but never needed.
The a-v difference across the brain was obtained by means of a 2.2 mm (14 gauge) catheter introduced percutaneously into the right internal jugular vein and advanced to the venous bulb, and a 1.1 mm (20 gauge) catheter placed in the brachial artery of the non-dominant arm. Mean arterial pressure (MAP) was obtained from the artery and integrated by a monitor (Dialogue 2000, Danica Electronic, Copenhagen, Denmark) that also calculated HR from a three lead electrocardiogram. Catherisation was well tolerated by the subjects and was not reported to affect their exercise performance.
Arterial and venous blood samples were drawn simultaneously three times at rest, twice during submaximal exercise, every second minute during maximal exercise, and several times in the recovery period. Samples were drawn anaerobically for determination of glucose, lactate and blood gas variables and were placed immediately on ice for later analysis (ABL 625, Radiometer, Denmark). Samples drawn for the determination of alanine (Graßl & Supp, 1985), glutamate and glutamine (Lund, 1985), glycerol (Eggstein & Kuhlmann, 1970) and free fatty acids (FFA; NEFA C, ACS-ACOD Method, Wako Chemicals, Richmond, VA, USA) were placed on ice and centrifuged within 5 min. Plasma was frozen in liquid nitrogen and kept at −80°C until analysed. The assays applied to analyse for amino acids, glycerol and FFA have an accuracy that is better than 1%. Measurement by way of the ABL at the relevant concentrations has an inaccuracy (mean difference between the measured value on a group of test instruments and the estimated true value as assayed by the reference method) of: glucose and lactate 0.1 mm, PO2 0.08 kPa, haemoglobin 0.2 g dl−1 and O2 saturation 0.2 %.
The cerebral blood flow was evaluated by transcranial ultrasound Doppler. The proximal segment of the MCA was insonated (Multidop X, DWL, Sipplingen, Germany) through the right temporal window. After determination of the optimal signal-to-noise ratio, the probe was secured by adhesive ultrasonic gel (Tensive, Parker Laboratories, Fairfield, NJ, USA) and a headband. The MCA Vmean was calculated as the one minute average of continuously sampled maximal frequency shifts for each heart beat. Both at rest and during maximal cycling, determination of flow velocity in the MCA has a coefficient of variation of ∼5 % (Pott et al. 1996).
During the study with patial curarisation, the VO2 was measured breath by breath (MedGraphics 2001, St. Paul, MN, USA). The level of partial neuromuscular blockade was assessed both before and during exercise as the handgrip strength for the dominant arm by means of a strain gauge dynamometer connected to a measuring bridge (Caspersen & Nielsen, Copenhagen, Denmark). Two different questions were posed to assess the exercise intensity using a scale with units from ‘6′ to ‘20′ (Borg, 1970). The first question determined how hard the exercise felt by rating the perceived exertion (RPE), which is the regular application of the Borg scale. The second question quantified the ‘will to exercise’ by asking how hard the subject tried to exercise. Such differentiation is important when the neuromuscular transmission is impaired as the subjects may be trying mentally to exercise as hard as they can while at the same time experiencing only moderate intensity.
Variables are expressed as means ±s.e.m. or as median with range. The Friedman test was used to determine whether significant changes occurred between rest, exercise and the recovery which was divided into the first 5 min and the subsequent 25 min as well as, for maximal exercise, for an additional 30 min (Ide et al. 2000). Also, incremental exercise was divided into low and high intensity periods by an arbitrary ‘lactate threshold’, which in this context means the time point where the increase in plasma lactate becomes significant. Changes were identified with the Wilcoxon matched pairs test by rank. P < 0.05 was considered to be statistically significant.