INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

HEAVY RESISTANCE EXERCISE is associated with a pronounced elevation in mean arterial pressure (MAP) that occasionally may result in intracerebral bleeding (9, 12, 30). Although this suggests an increase in cerebral blood flow, the weight lifter's "blackout" has been linked to a decrease in cerebral blood flow provoked by the intense expiratory strain (a Valsalva-like maneuver) often performed to stabilize the trunk (4). A marked rise in the intrathoracic pressure increases central venous pressure (CVP) and may, in turn, lead to a critical reduction in the cerebral perfusion pressure (27, 32).

With application of a constant force during leg extension, the elevation of MAP depends on the mode of ventilation (23), and it is largest with a concomitantly performed Valsalva maneuver (19). During a Valsalva maneuver, a precipitous rise in MAP leads to parallel changes in the transcranial Doppler (TCD)-determined middle cerebral artery (MCA) blood velocity (V_mean; 27, 32). We hypothesized that during intense exercise the associated straining, i.e., a Valsalva maneuver, may be of deterministic importance both for central and cerebral hemodynamics. We considered that when continued ventilation is advocated during heavy resistance exercise (23), it is to avoid hypothetical harmful consequences for the brain. However, the separate effects of the developed muscle force and the strain on cerebral blood flow and oxygenation are unknown. The purpose of this study was to delineate the separate contributions of heavy static exercise and the associated rise in intrathoracic pressure on the cerebral and central hemodynamic responses.

In contrast to studies in which the TCD-determined MCA V_mean or cerebral blood flow was determined intermittently during static exercise (16, 29) or focused on the depression established during a maximal lift (7), we took advantage of the ability of the Doppler technique to record beat-by-beat MCA V_mean. We hypothesized that MCA V_mean would be influenced by the steep increase in MAP at the onset of static exercise and that its time course would be affected by changes in both CVP and the arterial carbon dioxide tension (Pa_CO2). Abrupt changes in MAP associated with intense exercise may increase transmural pressure of the insonated cerebral artery, and the diameter of the MCA may change, violating the assumption of a linear relationship between changes in MCA blood velocity and cerebral blood flow (18). To surpass the uncertainty of changes in MCA diameter, we combined TCD and near-infrared spectroscopy (NIRS) as methods based on different physical principles, assuming that concordant changes would indicate a change in cerebral blood flow (35).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Ten healthy subjects [4 women; age 28 (21-38) yr (median with range), weight 76 (50-85) kg, and height 179 (162-191) cm] participated in the study after informed consent to the protocol as approved by the Ethics Committee of Copenhagen [KF 01-120/96] was given.

After 30 min of rest, the subjects were asked to perform static two-legged exercise while sitting with a lower back support adjusted to keep the ankle and knee joints bent 90°. Extension strength was determined with a steel rod mounted in bar bearings on an iron frame that also carried the seat. Deformation of the rod was recorded by four strain gauges connected to a Wheatstone bridge, and a potentiometer allowed for visual feedback (Caspersen & Nielsen, Copenhagen, Denmark). The protocol included two series each with three 15-s bouts of static exercise followed by 5 min of recovery, aiming at a force that was determined before the protocol as the maximal voluntary contraction force for 15 s of static two-legged exercise with continued ventilation. In random order, leg extension was performed with continued ventilation or with a concomitantly performed Valsalva maneuver (n = 10). Ventilatory variables were obtained with an oxyscreen apparatus (MedGraphics, St. Paul, MN). The influence of straining on the exercise response was assessed in another session: six subjects [3 women; age 29 (21-44) yr, weight 75 (55-97) kg, and height 176 (158-190) cm] performed three 15-s Valsalva maneuvers without exercise and three 15-s bouts of static two-legged exercise together with a Valsalva maneuver aiming at a similar intrathoracic strain as guided by visual feedback from the tracing of CVP.

The proximal segment of the right MCA was insonated (Multidop X, DWL, Sipplingen, Germany) through the posterior temporal "window." Once the optimal signal-to-noise ratio was obtained, the probe was covered with an adhesive ultrasonic gel (Tensive, Parker Laboratories, Orange, NJ) and secured with a headband. The V_dia and V_sys were the diastolic and systolic velocities, respectively. V_mean was computed as the integral of the maximal frequency shifts over one beat divided by the corresponding beat interval.

In six subjects, both TCD cerebral blood velocity and local cerebral tissue oxygenation were measured. Cerebral oxygenation was assessed by NIRS (NIRO 500, Hamamatsu Photonics, Hamamatsu, Japan). The optodes on the forehead were placed 4.5 cm apart, covered with black rubber for light shielding, and fixed against the skin by adhesive tape. The NIRO 500 uses wavelengths of 775, 825, 850, and 905 nm to calculate the concentration (c) changes () of oxyhemoglobin (cHbO₂) and deoxyhemoglobin (cHb) by applying the modified Lambert-Beer law: c = OD/( · L · B), where OD is the attenuation of light expressed as changes in optical density,  is the extinction coefficient of the chromophore (mM/cm), L is the distance between the optodes (cm), and B is a pathlength factor that takes into account the scattering of light in the tissue (5.98). Measurements were sampled at 2 Hz.

Arterial pressure was measured with a Finapres model 5 (n = 10; Netherlands Organization for Applied Scientific Research, Biomedical Instrumentation; TNO-BMI, Amsterdam, The Netherlands). The Finapres cuff was applied to the midphalanx of the middle finger of the dominant arm and placed at heart level. Beat-to-beat systolic and diastolic pressures as well as MAP were computed after analog-to-digital conversion at a sampling rate of 100 Hz. MAP was the integral of pressure over one beat divided by the corresponding beat interval, and heart rate (HR) was the inverse of the interbeat interval.

A catheter (1.7 mm ID, 16 gauge) was placed in the superior caval vein through the basilic vein for CVP and venous O₂ saturation. CVP was recorded from a transducer (Bentley, Uden, The Netherlands) fastened to the subject in the midaxillary line at the level of the right atrium and connected to a monitor (8041, Simonsen & Weel, Copenhagen, Denmark). Stroke volume (SV) was obtained from the arterial pressure pulse wave by model flow analysis (FAST-mf/-cZ system, TNO, Amsterdam, The Netherlands). This method computes a continuous aortic flow waveform by simulating a nonlinear, time-varying model of the aortic input impedance with an accuracy to a thermodilution estimate of ~5% (11). To illustrate the feasibility of this method during a Valsalva maneuver in one subject, model flow  SV was compared with a transthoracic Doppler ultrasound determination of SV (Fig. 1). Cardiac output (CO) was the product of SV and HR and was expressed relative to control.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Comparison of beat-by-beat stroke volume (SV; thin line) during a Valsalva maneuver determined by model flow from continuous Finapres arterial blood pressure (ABP) and transthoracic Doppler ultrasound (bold line). Interruptions in pressure tracings (top) represent determination of Finapres volume clamp set point.

A 1.0-mm internal diameter (19-gauge) catheter was placed in the brachial artery of the nondominant arm for Pa_CO2 and O₂ saturation. For one run under each condition, blood samples were obtained anaerobically in heparinized syringes (QS50, Radiometer, Copenhagen, Denmark) at rest, during the last seconds of leg extension, and ~10 s into the recovery. Samples were analyzed immediately (OSM-3, ABL, Radiometer).

Tracings of the recorded variables were checked for artifacts and transformed to equidistantly resampled data at 2 Hz by polynomial interpolation.

Critical closing pressure (CCP) is hypothesized to integrate the intracranial pressure and the arteriolar tone (2, 5). CCP is estimated as the pressure axis intercept of the linearly extrapolated line between the systolic and diastolic pressure-flow (velocity) values for single cardiac cycles (22). Thus this concept assumes that intracranial capillaries "collapse" instantaneously when, during one cardiac cycle, blood pressure declines below CCP.

The Friedman test was used to determine whether significant changes were established over time, and only if such a difference was indeed present Wilcoxon's matched pair signed-rank test was applied for comparisons with the baseline value. Linear regression analysis was applied to test for interdependencies, and paired observations were tested by a two-tailed t-test. A P value of <0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Leg extension with a Valsalva-like maneuver resulted in an instantaneous elevation in CVP to a maximum of 56 ± 11 mmHg and remained elevated until the end of straining (on average 40 ± 7 mmHg; Fig. 2). MAP followed a three-phasic course with a rapid increase from 98 ± 2 to 141 ± 6 mmHg (within 2 ± 1 s) followed by a reduction (after 8 ± 1 s) to 113 ± 7 mmHg and a maximum at 151 ± 7 mmHg just before the end of the strain (Fig. 2). CO decreased by 24 ± 4% reflecting a lowered SV (35 ± 4%) not fully compensated for by the increase in HR (by 20 ± 2 beats/min). Within 1 ± 1 s, MCA V_mean increased from 59 ± 5 to 77 ± 7 cm/s, and after 3 ± 1 s MCA V_mean reached the baseline level and decreased further to 46 ± 5 cm/s after 7 ± 1 s followed by a second increase toward the end of the contraction (to 69 ± 10 cm/s).

View larger version (47K): [in this window] [in a new window]   Fig. 2.   Systolic, mean, and diastolic middle cerebral artery mean blood velocities (VMCA) and ABP, central venous pressure (CVP), critical closing pressure (CCP), heart rate (HR), and changes in model flow stroke volume (SV) and the difference between mean arterial pressure and CVP (MAP-CVP) during static two-legged exercise (shaded bars) with a Valsalva maneuver (n = 7) or with continued ventilation (n = 10). Values are means ± SE.

After the contraction, CVP reached the baseline level and MAP dropped after 2 ± 1 s to 90 ± 7 mmHg followed by a transient increase to 107 ± 5 mmHg. SV and CO increased 34 ± 7 and 72 ± 12%, respectively. MCA V_mean dropped sharply to below the baseline level (to 46 ± 3 cm/s after 1 ± 1 s) with a subsequent overshoot (to 73 ± 8 cm/s).

Leg extension with continued ventilation increased CVP by 9 ± 1 mmHg within 3 ± 1 s, but, averaged over the whole contraction the increase was not significant (1 ± 1 mmHg; Fig. 2). Within 3 ± 1 s from the onset of the contraction, MAP increased to 120 ± 7 mmHg. During the leg extension, CO increased 11 ± 5% as did HR (by 22 ± 2 beats/min), corresponding to a reduction in SV by 12 ± 3%. After an initial maximum of 70 ± 6 cm/s (within 3 ± 1 s), V_mean decreased to the baseline level within 5 ± 1 s, whereas MAP remained elevated. After static exercise, MAP decreased to below the resting level to reach 79 ± 3 mmHg (after 4 ± 1 s), whereas SV (by 35 ± 6%) and CO increased (by 80 ± 10%). MCA V_mean was lowest 3 ± 1 s after the contraction at 40 ± 5 cm/s.

Compared with leg extension with a Valsalva maneuver, MAP and CVP were lower when ventilation was maintained during exercise, as was the initial increase in MCA V_mean. At the end of exercise, Pa_CO2 was diminished and O₂ saturation was elevated both with and without a Valsalva maneuver (Table 1). Venous O₂ saturation remained stable during the contraction and decreased to a lowest value ~10 s into the recovery. Respiratory variables are presented in Fig. 3, demonstrating hyperventilation during leg extension with continued ventilation and an increase in pulmonary oxygen uptake after exercise especially with a Valsalva maneuver.

View larger version (50K): [in this window] [in a new window]   Fig. 3.   Ventilatory frequency (f), minute ventilation (E), O2 uptake (O2), and end-tidal carbon dioxide tension (PETCO2) during leg extension with a Valsalva maneuver (circles; n = 7) or with continued ventilation (n = 10). Values are means ± SE. , Significantly different from baseline, P < 0.05.

CCP. The average CCP is presented in Fig. 2 and follows a time course that is similar to that of CVP during exercise with a concomitantly performed Valsalva maneuver. In contrast, during two-legged exercise with continued ventilation, CCP tended to increase only toward the end of the contraction, and it declined slowly after the exercise.

MCA V_mean vs. MAP and the difference between MAP and CVP. Both with and without a Valsalva maneuver, from the onset of leg extension to the maximum MCA V_mean, MCA V_mean paralleled changes in MAP (r² = 0.95 ± 0.02 and 0.93 ± 0.02, respectively). The diverging time courses of the difference between MAP and CVP (MAP-CVP) and MCA V_mean are shown in Fig. 2.

Influence of leg extension on the Valsalva response. The central and cerebral hemodynamic variables from a Valsalva maneuver with and without accompanying two-legged static leg extension are presented in Table 2 with a representative recording in Fig. 4. Cerebral hemodynamic variables were not different between the two conditions, but after the maneuver, HR and in turn CO were higher with static exercise.

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Recording during two-legged static exercise with a concomitantly performed Valsalva maneuver vs. a resting Valsalva maneuver. SV and CO, changes in model flow SV and cardiac output (bold lines), respectively; cHbO2, changes in cerebral tissue oxygenated hemoglobin concentration; CCP, changes in CCP. Roman numerals indicate phases of the Valsalva maneuver.

NIRS vs. MCA V_mean. A representative time course of the NIRS-determined cHbO₂ and cHb and the TCD-determined MCA V_mean is presented in Fig. 5; both cHbO₂ and MCA V_mean demonstrated the distinct phases of the Valsalva maneuver, whereas cHb remained statistically unchanged (Table 2). Peak changes in cHbO₂ were delayed compared with those of MCA V_mean by 1.5 ± 1.0 s, and both the nadir during phase II of the maneuver and the maximum after the maneuver were less pronounced in cHbO₂ compared with MCA V_mean. Correlation analysis between the peak changes in cHbO₂ and MCA V_mean revealed a r² = 0.77 (n = 6).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Average time course of changes in near-infrared spectroscopy-determined oxygenated (cHbO2, thin line) and deoxygenated (cHb, dotted line) hemoglobin concentration from the frontal cortex and changes in the transcranial Doppler-determined middle cerebral artery mean blood velocity (Vmean, bold line) during a Valsalva maneuver and a Valsalva maneuver with concomitantly performed 2-legged static exercise (n = 6).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study evaluated the MCA V_mean response to intense static two-legged exercise and focused on its modification by the respiratory pattern and the associated changes in intrathoracic pressure employed during exercise. The onset of intense static leg extension was associated with a rapid rise in MAP and the MCA V_mean, and this was particularly pronounced with a concomitantly performed Valsalva maneuver. In contrast, a reduction in MCA V_mean during and after intense static exercise appeared to be induced both by expiratory straining and by hyperventilation.

TCD monitors blood velocity rather than volume flow and changes in the diameter of the insonated vessel could modulate velocity independently of flow (18). During craniotomy in anesthetized patients, Giller et al. (8) found the diameter of the large cerebral vessels unchanged with large changes in arterial pressure. With MRI, the MCA diameter was found to be stable in healthy individuals during hypocapnia (34), hypercapnia, and lower body negative pressure (31), suggesting that the MCA is not involved in the regulation of cerebral vascular resistance. However, even small changes in the diameter could lead to a large change in blood velocity, and the spatial resolution of the MRI technique (~0.5 mm) is not sufficient to exclude the possibility that changes in vascular diameter may effect MCA blood velocity. The assumption of a stable MCA diameter may be invalid during the conditions of the present study with dynamic changes in both blood pressure and intrathoracic pressure, changes in Pa_CO2, and presumably also changes in sympathetic tone. Thus, especially at the onset of static leg extension with a Valsalva maneuver, the quick and large rise in MAP could increase transmural pressure and distend the MCA, leading to underestimation of an increase in cerebral blood flow. To relate the MCA V_mean response to another independent indicator of cerebral blood flow, we assessed NIRS-determined cHbO₂ and cHb in the frontal cerebral tissue. Because in arterial blood hemoglobin is almost fully saturated with O₂, cHbO₂ may increase with both active and passive arteriolar dilation and recruitment of capillaries (13). Accordingly, cerebral activation increases regional cerebral blood flow and cHbO₂ (35). During the Valsalva maneuver, and irrespective of concomitantly performed two-legged static exercise, we found peak changes in cHbO₂ closely related to those of MCA V_mean, and the cHb remained constant throughout the maneuver, suggesting that during a Valsalva maneuver cortical arteriolar volume follows changes in the arterial inflow or that an increased O₂ flow-to-O₂ uptake ratio and elevated velocity will reduce fractional O₂ extraction and more of the vascular bed will therefore contain blood with an increased HbO₂.

In our untrained subjects, heavy static exercise with a Valsalva maneuver exerted similar cerebral and central hemodynamic responses to a Valsalva maneuver without accompanying exercise. In patients with a subarachnoid aneurysmal hemorrhage, activities associated with straining precede the bleeding in a large number of cases (30). When our subjects avoided performing a Valsalva maneuver during exercise, the elevation in MAP was both slower and less pronounced (by ~20 vs. 40 mmHg). At the onset of the contraction and irrespective of the mode of ventilation, MCA V_mean increased in parallel with MAP (R²  1), even though the elevation in MAP was within the upper limit of what is referred to as the cerebral autoregulatory range. This finding reflects the observation that cerebral autoregulation onsets with a delay (~2 s) and establishes blood velocity (flow) homeostasis within ~10 s (1, 26). The question whether the subsequent decline to or below the baseline level in spite of a still-elevated blood pressure in fact reflects an autoregulatory response is not addressed in the present study, and such effort would require multivariate analysis methods taking the changing Pa_CO2, CCP, and eventually cerebral metabolic demand into account (24). The results indicate that the onset of resistance exercise, especially when performed with a Valsalva-like maneuver, increases blood flow to the brain (23). Whether an occasional association with intracerebral bleeding (9, 12) is due to an elevation in flow per se or due to an increase in transmural pressure cannot be answered from the data of the present study. It may be speculated that continued ventilation during lifting of a heavy weight may be "safer" because the contraction is then associated with a lesser increase in MAP and, presumably, cerebral blood flow (23).

During weight-lifting exercise in the elite athlete, an increase in the intrathoracic pressure up to 260 mmHg has been suggested to protect the brain by counteracting the pronounced influence of arterial pressure (up to 480/350 mmHg; Ref. 19) on the cerebral perfusion pressure, i.e., the difference between MAP and the intracranial pressure. During straining, intracranial pressure is modified by an elevation in both spinal fluid pressure and central and/or jugular venous pressure (10). However, the diverging time course of MCA V_mean and MAP-CVP suggests that CVP has a delayed or attenuated influence on intracranial pressure and, in turn, the cerebral perfusion pressure. We propose that the delayed influence of the straining-associated rise in CVP on MCA V_mean is due to the compliance of both the cervical and cerebral venous vascular beds. In contrast to the upright position, such a delay or attenuation is not noted when subjects perform a supine Valsalva maneuver, i.e., in that position MCA V_mean parallels MAP-CVP (27). We consider that, with straining in the upright position, some time and pressure are needed to fill the collapsed outflow veins. This notion is supported in an animal preparation by Kongstad and Grände (17), who demonstrated an increase in venous pressure with no influence on cerebral tissue pressure for as long as the venous pressure was below the tissue pressure. Only when the venous pressure equals, or exceeds, the tissue pressure does collapse of the outflow vein disappear and the two pressures increase in parallel. Also, in the dog, jugular venous collapse appears to constitute a resistance in the transmission of central to cerebral venous pressure in the head-elevated position (33).

To shed further light on the mechanism counteracting the increase in MAP during exercise, we calculated the CCP that reflects the theoretical positive pressure at which flow becomes zero, and this is considered to integrate vascular tone and intracranial pressure (2). During the leg contraction with straining, CCP mirrored CVP in both magnitude and time course. During exercise with continued ventilation, CCP increased slightly toward the end of the exercise bout when the subjects hyperventilated. For obvious reasons intracranial pressure was not monitored in our subjects. Nevertheless, during static exercise with a Valsalva maneuver, the elevated CCP may indicate a pronounced increase in ICP (6), whereas the changes in CCP during exercise with continued ventilation more likely are attributable to an elevation in cerebrovascular tone related to the lower Pa_CO2 (3).

Pa_CO2 is considered an important factor for the reduction in MCA V_mean during and after the contraction. With continued ventilation during leg extension, the drop in Pa_CO2 reflects hyperventilation, whereas exercise with a Valsalva maneuver is associated with a pronounced reduction in CO and, in turn, a reduced washout of CO₂ from the tissue (21). In resting humans, MCA V_mean changes ~2.6% for 1 Torr change in Pa_CO2 (20), and for a rapid change in Pa_CO2, the MCA V_mean response is established after ~15 s (25, 28). During leg extension with continued ventilation, Pa_CO2 decreased ~5 Torr, and MCA V_mean reached the baseline level ~5 s after an initial ~25% elevation, corresponding to a ~5% MCA V_mean change for 1 Torr change in Pa_CO2. Such "CO₂ reactivity" would be even larger (~15%/Torr) during the leg extension with a concomitant Valsalva maneuver when Pa_CO2 decreased only ~2 Torr. Although it is unknown whether static exercise may alter the normal cerebrovascular CO₂ reactivity, the magnitude and time course of the changes in MCA V_mean indicate that factors other than Pa_CO2 contribute to the reduction in MCA V_mean. We consider that the rise in CVP and CCP during exercise with a Valsalva maneuver considerably taxes cerebral perfusion pressure and together with the cerebrovascular consequences of hyperventilation may explain occurrence of a weight lifter's blackout (4).

We considered the potential influence of changes in CO for MCA V_mean. The elevation in MCA V_mean is reduced during dynamic exercise with pharmacological cardioselective -blockade that blunts the exercise induced increase in CO (15). Sympathetic pathways probably mediate such an effect of a reduction in CO on cerebral vascular resistance as the attenuated MCA V_mean is reversed by a unilateral stellate blockade (14). In this study, CO increased ~10% when ventilation was maintained during exercise, whereas MCA V_mean decreased throughout. With a concomitantly performed Valsalva maneuver, CO continuously decreased to as low as 50% of the baseline value, whereas MCA V_mean increased in the second half of the exercise. These results render a simple relationship between dynamic changes in CO and those of MCA V_mean unlikely under the circumstances of the present study.

Taken together, in untrained subjects performing intense static two-legged extension, the largest increase (~30%) in MCA V_mean is observed at the onset of exercise performed with a concomitant Valsalva-like maneuver, although CVP also increases. Marked reductions (20-30%) are noted during the sustained exercise with a Valsalva-like maneuver and after static two-legged extension with continued ventilation. We furthermore present that changes in cerebral blood velocity as an index of cerebral blood flow as gauged in a large cerebral artery are followed by NIRS-determined concentration changes in oxygenated hemoglobin in frontal cerebral tissue. The results suggest that intense exercise with a concomitant Valsalva maneuver raises cerebral blood flow, whereas a reduction in cerebral blood flow is expected during or after exercise related to both pronounced straining and hyperventilation. A possible limitation may be that we could not measure the effects of intense static exercise concomitant to straining on arterial and transmural pressures from the MCA toward the small cerebral blood vessels. We acknowledge that when relating changes in measured variables that are supposed to reflect cerebral blood flow, uncertainties remain regarding the causes of the weight lifter's blackout and exercise-associated cerebral hemorrhage.