The purpose of this study was to test the hypothesis that repeated exposure to high acceleration (G) would be associated with enhanced functions of specific mechanisms of blood pressure regulation. We measured heart rate (HR), stroke volume (SV), cardiac output (Q̊), mean arterial blood pressure, central venous pressure, forearm and leg vascular resistance, catecholamines, and changes in leg volume (%ΔLV) during various protocols of lower body negative pressure (LBNP), carotid stimulation, and infusions of adrenoreceptor agonists in 10 males after three training sessions on different days over a period of 5–7 days using a human centrifuge (G trained). These responses were compared with the same measurements in 10 males who were matched for height, weight, and fitness but did not undergo G training (controls). Compared with the control group, G-trained subjects demonstrated greater R-R interval response to equal carotid baroreceptor stimulation (7.3 ± 1.2 vs. 3.9 ± 0.4 ms/mmHg, P = 0.02), less vasoconstriction to equal low-pressure baroreceptor stimulation (−1.4 ± 0.2 vs. −2.6 ± 0.3 U/mmHg, P = 0.01), and higher HR (−1.2 ± 0.2 vs. −0.5 ± 0.1 beats · min−1 · mmHg−1,P = 0.01) and α-adrenoreceptor response (32.8 ± 3.4 vs. 19.5 ± 4.7 U/mmHg, P = 0.04) to equal dose of phenylephrine. During graded LBNP, G-trained subjects had less decline in Q̊ and SV, %ΔLV, and elevation in thoracic impedance. G-trained subjects also had greater total blood (6,497 ± 496 vs. 5,438 ± 228 ml, P = 0.07) and erythrocyte (3,110 ± 364 vs. 2,310 ± 96 ml,P = 0.06) volumes. These results support the hypothesis that exposure to repeated high G is associated with increased capacities of mechanisms that underlie blood pressure regulation.
onset of rapid and high-level head-to-foot acceleration (G) can overwhelm cardiovascular mechanisms responsible for maintaining cerebral perfusion, vision, and consciousness in pilots who undergo aerial combat maneuvers in high-performance aircraft, often leading to catastrophic loss of consciousness (3). This G-induced loss of consciousness represents an exaggerated challenge to mechanisms involved in regulation of blood pressure. Although the emphasis of high-G acceleration research has been typically placed on description of acute hemodynamic responses during exposure, several observations suggest that repeated exposure to increased G acceleration can induce a protective “G-training” effect against development of hypotension. For instance, regular exposure at high-G acceleration enhances tolerance to high G in humans, rats, guinea pigs, and dogs (3), whereas prolonged layoff from exposure in high-G profiles (G layoff) has resulted in reduced endurance to high-G acceleration (30). In addition, the individuals least susceptible to presyncopal episodes after spaceflight were career pilots of high-performance aircraft before they became astronauts (17). Finally, high-performance aircraft pilots have demonstrated less orthostatic hypotension during the transition from supine to head-up tilt compared with nonpilot control subjects (32).
Application of protective technology (e.g., anti-G suits, anti-G straining maneuvers) was consistently utilized and therefore could not explain alterations in acceleration tolerance (3, 30). It seems likely that increased acceleration performance following intermittent repeated exposure to high G may partly explain adaptations in physiological functions associated with blood pressure regulation. Therefore, the purpose of this study was to test the hypothesis that repeated exposure to high levels of G acceleration induces significant physiological adaptations that increase the responsiveness and capacities of mechanisms that underlie blood pressure regulation.
Twenty men volunteered to participate as subjects for this investigation after all procedures and risks associated with the experiments were explained and their voluntary written informed consent to participate in the study was obtained. Reflex and hemodynamic responses and adaptation to repeated high-G training similar to those used in the present investigation are significantly attenuated in women compared with men (8, 13). Therefore, in an effort to minimize variability between groups in a cross-sectional comparison, female subjects were excluded. All procedures were approved by the Institutional Review Board. Ten G-training subjects were recruited from a panel of individuals certified for and familiarized with riding the centrifuge, and matched for height, weight, body fat, and aerobic fitness with 10 control subjects. All subjects were nonsmokers and normotensive and their selection into the study was based on results of a screening evaluation comprising a detailed medical history, physical examination, blood chemistry analysis, urinalysis, and resting and treadmill electrocardiogram (ECG) to assure absence of cardiovascular disease. Individuals taking prescription drugs were excluded, and subjects refrained from taking any medications at the time of the experiments. Because of potential effects on vascular volume and baroreflex functions, subjects were asked to refrain from exercise and stimulants such as caffeine and other nonprescription drugs 48 h before testing. During an orientation period that preceded the experiments, subjects were made familiar with the laboratory, the protocol, and procedures.
G-training subjects (n = 10) were assigned to undergo three sessions (different days) of exposure to high acceleration (G training), with each session separated by 24–48 h. Therefore, duration of the entire training period was 5–7 days. The control subjects (n = 10) had no high-G exposure. No subject was a high-performance aircraft pilot, and all subjects met the same selection criteria to participate in the study. Acceleration training was conducted at the centrifuge facilities at Brooks (Texas, 7 subjects) and Wright-Patterson (Ohio, 3 subjects) Air Force Bases. Each centrifuge was configured with an upright seat that was adjusted to either 13° (5 subjects) or 30° (5 subjects) seat-back position. Subjects wore flight suits with standard CSU-13B/P anti-G suits (David Clark, Worcester, MA) that were inflated during centrifuge training sessions. G-suit pressure profiles consisted of inflation at 1.5 psi (∼78 mmHg) for each +1 G above 2 G (i.e., 1.5 psi at +3 G, 3.0 psi at +4 G, 4.5 psi at +5 G, etc.) to a maximum of 11 psi (∼569 mmHg) at or above +9 G. Subjects also used anti-G straining maneuvers during acceleration exposure. Because the intention in the present investigation was to expose subjects to the highest acceleration stimulus possible, centrifuge training protocols were individualized based on characteristics of the centrifuge and tolerance of each subject. Because G-layoff effects on acceleration endurance occur within 14 days (30), subjects did not undergo acceleration exposure for a minimum of 30 days between their last exposure and the three centrifuge training sessions. Centrifuge training sessions consisted of a warm-up acceleration with gradual onset rate exposure (+0.1 G/s) for 50 s, ∼5 min of recovery, followed by a rapid onset rate exposure (+6.0 G/s) to a maximum of +7.0 G for 30 s (7 subjects). After the warm-up acceleration exposures, subjects performed a series of simulated air combat maneuvers (SACM). One cycle of a SACM consisted of exposure to +4.5 G for 15 s immediately followed by exposure to +7.0 G for 15 s (6 subjects) or +5.0 G for 15 s followed by exposure to +9.0 G for 15 s (3 subjects). One subject tolerated training sessions consisting of SACM exposures from +5.0 G for 15 s up to +12.0 G for 15 s. Subjects repeated SACM cycles until they met one or more of the following preset criteria:1) medical monitor decision, 2) volitional termination by the subject due to physical fatigue, 3) nausea, 4) 100% loss of peripheral vision using a multicolored light bar, 5) 50% central field loss of vision, and 6) loss of consciousness. The sum of the products of G levels and time in seconds of exposure to those levels was calculated to provide a total cumulative G-exposure index for each training session.
The experimental protocol consisted of 3 days of tests and measurements of physical and physiological functions beginning the day after the final (third) G-training session. On day 1 subjects underwent the following tests: 1) cardiopulmonary baroreflex control of peripheral vascular resistance, 2) aortic baroreflex control of heart rate (HR), 3) adrenergic receptor responsiveness, and 4) measurement of plasma volume. On day 2 subjects underwent measurements for1) R-R interval variability, 2) carotid baroreflex control of R-R interval response, 3) measurement of leg volume (LV) and compliance, and 4) hemodynamic responses to lower body negative pressure (LBNP). On a third test day, subjects underwent tests for measurement of their maximal oxygen uptake (V˙o 2 max) and estimated body composition. All tests conducted on days 1 and 2were always separated from V˙o 2 max tests by a minimum of 48 h. All measurements were conducted at the same time of day and in the same sequence.
Day 1 Tests
Measurement of cardiopulmonary baroreflex control of forearm vascular resistance.
Subjects were instrumented for beat-to-beat measurements of HR, systolic (SBP) and diastolic (DBP) blood pressures, and estimated central venous pressure (CVP). After instrumentation, subjects were placed in the LBNP chamber at a right lateral decubitus position, and a 20-gauge catheter was inserted into an antecubital vein of the dependent right arm for measurement of CVP. With the right arm being suspended, the valves in the veins become incompetent, resulting in an unimpeded column of blood. Under these conditions, the pressure in the large vein of the right arm reflect CVP when the pressure transducer is centered at heart level (18). The catheter was then cleared with isotonic saline solution and connected to a Baxter Uniflow model 43–260 pressure transducer (Baxter Healthcare, Irvine, CA) for measurement of venous pressure. Before connection to the catheter, the transducer was positioned at the level of the midsternum with a ruler and level and was calibrated with a known 20-mmHg pressure introduced from an Omega digital manometer (Omega Engineering, Stamford, CT). Forearm blood flow was measured with the left arm slightly elevated (∼10 cm) above heart level using venous occlusion plethysmography with a dual loop mercury-in-Silastic strain gauge placed around the left forearm at the point of maximal circumference. Venous outflow from the forearm was prevented by the placement of a cuff around the brachium just above the elbow using an occlusion pressure of +40 mmHg. Arterial occlusion to reduce blood flow to the hand was applied by a wrist cuff inflated at a pressure of +250 mmHg. After wrist-cuff inflation for 1 min, six 10-s occlusions were repeated during the final 2 min of each test condition, and the average of the six measurements represented the flow for that test condition. The relative change (%) in strain gauge length over 10 s was quantified as a volume of blood per unit time, i.e., flow. Mean arterial pressure (MAP) was calculated by dividing the sum of systolic pressure and twice diastolic pressure by three. An index of forearm vascular resistance (FVR) was calculated by dividing MAP by average flow during the final 2 min of each test condition and expressed as peripheral resistance units (PRU in mmHg · min · 100 ml · ml−1). After instrumentation, subjects remained quiet for 15 min. After a 2-min data collection period with the LBNP at 0 mmHg (pre-LBNP), subjects underwent continuous decompression at 5, 10, 15, and 20 mmHg every 2 min. The LBNP protocol was designed to elicit the vascular constriction response caused by unloading cardiopulmonary baroreceptors (22, 39). Forearm blood flow and CVP were measured continuously throughout the LBNP test. A stimulus-response relationship of the baroreflex was derived by plotting estimated CVP against respective FVR, and the responsiveness of the reflex was determined by calculating the slope from least-squares linear regression analysis.
Measurement of aortic-cardiac baroreflex.
Subjects remained in the LBNP chamber at a right lateral decubitus position, and a 20-gauge catheter was inserted into an antecubital vein of the left arm for drug infusion. After instrumentation, subjects rested quietly for 15 min, after which 3 min of resting data were obtained. The aortic-cardiac baroreflex was assessed using a technique described previously (35). The protocol was initiated with a steady-state infusion of phenylephrine (PE) into the left arm catheter with a goal of increasing MAP by 15 mmHg. LBNP was applied (ranging from 5 to 20 mmHg) until estimated CVP was returned to pre-PE infusion levels, and neck pressure equal to 1.4 times the increase in MAP (25) was then applied to the anterior two-thirds of the neck with the intention of returning mean carotid sinus transmural pressure to pre-PE infusion values. Clamping of CVP and carotid pressure at resting levels by application of LBNP and neck pressure, respectively, was designed to remove PE-induced loading of cardiopulmonary and carotid baroreceptors, thereby isolating the influence of the aortic baroreceptors. Responsiveness of the aortic baroreflex control of HR was calculated as the ratio of the difference in HR to MAP (ΔHR/ΔMAP) between pre-PE infusion and post-PE infusion with LBNP and neck pressure.
Measurements of adrenoreceptor responsiveness.
After 30 min of supine control, resting measurements of HR, MAP, and leg blood flow were made. Leg blood flow was measured with the leg slightly elevated (∼10 cm) at the ankle using venous occlusion plethysmography in a manner similar to the procedure used for forearm blood flow measurement, except that venous outflow from the calf was prevented by the placement of a cuff around the thigh just above the knee using an occlusion pressure of +60 mmHg. An index of leg vascular resistance (LVR) was calculated by dividing MAP by average flow during the final 2 min of each test condition and expressed as PRU. After resting measurements, three graded infusions of α- and β-adrenoreceptor agonists were performed through the left arm catheter used during the aortic baroreflex test with isotonic saline as a vehicle. Each infusion interval was 9 min in duration to establish steady state and allow adequate time for all measurements. The protocol and dosages of adrenoreceptor agonists were determined by laboratory experience to produce safe but significant physiological responses (8). The total volume infused was <50 ml. A recovery period of at least 25 min was allowed between the two agonist-infusion protocols to allow hemodynamic measurements to return to preinfusion resting levels. During both infusion protocols, constant monitoring of beat-to-beat MAP and HR was performed and leg blood flows were measured at each infusion level.
Graded infusion of the α1-adrenoreceptor agonist PE was used to assess the responsiveness of these vascular receptors. PE was infused at three graded constant rates of 0.25, 0.50, and 1.00 μg · kg−1 · min−1. An elevation of SBP of 20 mmHg above or reflex reduction in HR of 20 beats/min below resting were predetermined end points for test termination. No tests were terminated using these criteria. The response of α1-adrenoreceptors was assessed by relating the PE dose with the reduction in LVR. The relationships between PE doses and LVR were linear, and the slopes describing these relationships were used to represent an index of α1-adrenoreceptor responsiveness. The relationship between the change in MAP caused by graded infusion of PE and the reflex change in HR (ΔHR/ΔMAP) was calculated and used to provide an assessment of integrated arterial-cardiac baroreflex function.
After HR, SBP, and DBP returned to resting levels following PE infusions, infusions of isoproterenol (Iso) were used to assess the responsiveness of β1- and β2-adrenoreceptors. Iso was infused at three graded constant rates of 0.005, 0.01, and 0.02 μg · kg−1 · min−1. An elevation of HR by 35 beats/min above resting was the predetermined end point for test termination. No tests were terminated using this criterion. Linear regression relationships were then constructed relating the increase in HR and the decrease in LVR to the dose of Iso. The slopes describing the linear stimulus-response relationship between the dose of Iso and HR and LVR provided a measure of the systemic responsiveness of β1- and β2-adrenoreceptors, respectively.
Blood (10 ml) was taken without stasis from the left arm catheter before and immediately after the third infusion level of PE and Iso to determine the response of norepinephrine (NE) and epinephrine (Epi). Immediately following each withdrawal, whole blood was taken from the syringe, transferred to a chilled tube containing sodium EDTA, and centrifuged at 2,000 g for 20 min at 4°C. Immediately after centrifugation, the plasma was aliquoted for NE and Epi and stored frozen until analyses were performed. Plasma NE and Epi concentrations were measured by high-performance liquid chromatography (Waters, Milford, MA) according to standardized procedures (11). Plasma volume was determined by a modified dilution technique (11, 20) using sterile solutions of Evans blue dye contained in 10-ml ampules (New World Trading, DeBary, FL). For this technique a standard intravenous injection of 11.5 mg of dye diluted with isotonic saline solution (2.5 ml) was administered. A portion of each blood sample was placed in three microhematocrit tubes and centrifuged. Hematocrit was determined from the average of the triplicate readings. Total blood volume was calculated from the plasma volume and venous hematocrit measurement and total erythrocyte volume was calculated from the difference between total blood volume and plasma volume. Total circulating plasma NE and Epi were calculated as the product of plasma volume and plasma NE and Epi concentrations as an index of resting sympathetic activity and catecholamine release during infusion (19).
Day 2 Tests
R-R interval variability.
An ECG was recorded during 5 min while each subject took 15 breaths/min using a metronome. An index of resting cardiac vagal activity was assessed by calculating the standard deviation of R-R intervals.
Measurement of carotid-cardiac baroreflex.
A Silastic neck chamber device (Engineering Development Laboratory, Newport News, VA) covering the area of the carotid arteries was utilized to elicit carotid baroreceptor stimulus-cardiac reflex response relationships. The stimulus profile consisted of raising neck chamber to 40 mmHg for five heart beats, followed by successive 15-mmHg R-wave triggered decrements to −65 mmHg. Subjects held their breath at the point of midexpiration during the pressure stimulus profile. SBP was measured via auscultation before and after each neck chamber test session, and carotid pressure was calculated as SBP minus neck chamber pressure applied during each R-R interval. A stimulus-response relationship of the baroreflex was derived by plotting R-R intervals at each pressure step against respective carotid distending pressure. From the average of each five-trial sequence of responses, least-squares linear regression analysis was applied to every set of three consecutive points on the response relationship to determine the segment with the steepest slope (maximum slope) to provide an index of baroreflex sensitivity.
LV and compliance.
A series of five circumference measurements placed 5 cm apart on each thigh and calf was performed on each subject. The total geometric volume of each thigh and calf segment was estimated by calculating the volume of each sequential segment from its midcircumference value and length, and summing the values of all segments. This procedure assumes that each segment approximates the shape of a cylinder. Total LV was calculated as the sum of all segments of both legs. Compliance of both legs was measured during supine rest using a mercury-in-Silastic strain gauge placed at the point of greatest calf circumference. After 30 min of supine control, the left leg was slightly elevated (∼10 cm) at the ankle and an occlusion cuff placed just above the knee was inflated to 30 mmHg for 180 s. Leg compliance was calculated by dividing the volume change (ml/100 ml) at a plateau (i.e., point at which venous pressure equals cuff pressure) by the cuff pressure and expressed as Δvol%/ΔmmHg. The value for leg compliance was multiplied by 100 for convenience.
Hemodynamic responses LBNP.
The subjects were positioned in the LBNP chamber in the prone posture, and an adjustable saddle was placed between the legs in an effort to stabilize the body during exposure to negative pressure. The LBNP protocol consisted of a 2-min resting period followed by a continuous decompression to −15 mmHg for 10 min, −30 mmHg for 10 min, −40 mmHg for 3 min, and −50 mmHg for 3 min. All subjects completed all stages of the LBNP protocol. HR was obtained from an ECG, and SBP and DBP were measured noninvasively from the left arm with a Collins automated sphygmomanometer. In addition, beat-by-beat continuous measurement of MAP was obtained using a Finapres finger photoplethysmographic technique (Ohmeda, Madison, WI) with the hand held at the level of the midsternum. Blood pressure measurements obtained by auscultation were used to verify readings obtained from the Finapres. Four silver-tape electrodes, two around the neck and two around the thorax, were attached to a Minnesota model 304B Impedance Cardiograph (Surcom, Minneapolis, MN) for noninvasive rheographic determination of stroke volume (SV) during rest and LBNP (8, 10). Electrical thoracic impedance cardiography has been shown to accurately measure SV in humans during direct comparison with dye dilution, thermodilution, and Fick methods (31). Cardiac output (Q̊) was calculated as the product of HR and SV. Total systemic peripheral resistance (TPR) was calculated by dividing MAP by Q̊. Changes in LV (%ΔLV) during each LBNP stage were measured with a strain gauge placed around the point of maximum girth of the left calf. The %ΔLV (ml/100 ml) was calculated from circumference changes. Forearm blood flow and vascular resistance were measured using the same venous occlusion plethysmography procedure applied during measurement of cardiopulmonary baroreflex function.
Day 3 Tests
Estimation of body composition.
Skinfolds were taken at six sites (chest, thigh, ilium, abdomen, triceps, and scapula), and the sum of the skinfold measurements was used to estimate percentage of body fat according to the formula of Pollock et al. (33).
After resting HR and blood pressure were measured in the supine position, a graded treadmill protocol was used to elicitV˙o 2 max. The exercise protocol began with the subject walking at a speed of 0.9 m/s (2.0 miles/h) and 0% grade for 1 min followed by 1.3 m/s (3.0 miles/h) for 2 min. At constant grade (0%), the treadmill speed was increased by 0.45 m/s (1 mile/h) each 30 s until the subject indicated that he had reached a comfortable running speed. At this point in the test, speed was maintained constant and the grade of the treadmill was increased by 2% every minute until the subject reached volitional exhaustion. Subjects breathed through a low-resistance valve, and expired gas was collected and measured on a Vmax Series 229D metabolic cart (SensorMedics, Yorba Linda, CA) for volume and fractions of mixed oxygen and carbon dioxide.
Statistical analyses were performed using JMP Start Statistics software (SAS Institute, Belmont, CA). Descriptive statistics were performed for all variables, and results are presented as means ± SE. A one-way analysis of variance (ANOVA) was performed for comparisons between the two groups of all variables and slopes of responses. A two-way ANOVA with repeated measures was used for between-group comparisons of NE and Epi across PE and Iso infusion levels. No explicit criterion for statistical significance was used because exact P values are presented for each independent effect and reflect the probability of obtaining the observed effect given only randomization error (4).
The two subject groups are described in Table1. Height, weight, estimated body fat, resting blood pressures, andV˙o 2 max were not statistically distinguishable, although the G-trained group was younger and had higher resting HR.
All subjects expressed fatigue as the primary criterion for terminating the SACM training session. One training session was terminated because of 100% peripheral and central field loss of vision, and another session ended with loss of consciousness (2 different subjects). No centrifuge training sessions were terminated by medical monitor decision or nausea. The average index of maximal G exposure for the 10 G-training subjects during the 3 days of acceleration exposure is presented in Fig.1. Total Gz endurance increased [F(1,9) = 6.539, P = 0.0002] in a linear manner from 584 ± 69 G-s on day 1to 886 ± 83 G-s on day 3 of training. We obtained records of previous best endurance times for seven of our subjects during centrifuge runs before our investigation. The average maximal endurance time on day 3 of G training for these subjects (968 ± 73 G-s) was similar [t = 0.665,P = 0.265] to the average of their best previous efforts (888 ± 43 G-s).
Responses to LBNP
Resting thoracic impedance at the point of LBNP termination increased in the G-trained subjects (0.2 ± 0.3 Ω) by only one-third the elevation of 0.7 ± 0.2 Ω observed in the controls [F(1,18) = 5.0946, P = 0.037]. Hemodynamic responses to graded LBNP are summarized in Table2 and illustrated in Fig.2. The %ΔLV during graded LBNP from 0 through −50 mmHg was twice as large in controls compared with G-trained subjects. Control subjects demonstrated gradual reduction in SV with a subsequent decrease in Q̊ despite a compensatory elevation in HR. Q̊ was maintained in the G-trained subjects by lesser reduction in SV. The magnitude, i.e., slope, of reduction in SV and Q̊ was greater in controls than in G-trained subjects. The change in MAP from resting to 50 mmHg LBNP in both groups (−6 ± 2 mmHg) induced statistically indistinguishable elevations [F(1,18) = 1.764, P = 0.201] in HR between G-trained (26 ± 5 beats/min) and control (19 ± 3 beats/min) subjects. Consequently, the ΔHR/ΔMAP from resting to 50 mmHg LBNP in the G-trained group (−6.9 ± 4.9 beats · min−1 · mmHg−1) was statistically similar [F(1,18) = 0.102,P = 0.754] to that of the control group (−5.1 ± 3.3 beats · min−1 · mmHg−1). The decrease in Q̊ from resting to 50 mmHg LBNP in the control group (1,614 ± 322 ml/min) was greater [F(1,18) = 3.757, P = 0.069] than that observed in the G-trained subjects (174 ± 366 ml/min). Group differences in slopes of the TPR response were not statistically distinguishable (Table 2). However, statistical analysis revealed that the TPR response to graded LBNP in the G-trained subjects was not different from zero change in TPR (t = 0.514,P = 0.620) compared with the control group that demonstrated a typical increase in TPR (t = 3.123,P = 0.012). MAP was higher during exposure to 40 and 50 mmHg LBNP in the G-trained subjects compared with the control group (Fig. 2).
R-R Interval Variability
The average standard deviation of R-R intervals during controlled breathing was not statistically distinguishable [F(1,18) = 0.630, P = 0.438] between G-trained subjects (56 ± 6 ms) and controls (64 ± 9 ms).
Figure 3 shows that carotid-cardiac baroreflex responsiveness, i.e., maximum slope of the stimulus-response relationship, was greater [F(1,18) = 6.449, P = 0.021] in the G-trained subjects (7.3 ± 1.2 ms/mmHg) compared with that in the controls (3.9 ± 0.4 ms/mmHg). The maximum slope of the change in calculated HR response to beat-to-beat changes in carotid pressure stimuli was greater [F(1,18) = 12.95, P = 0.002] in the G-trained subjects (−0.46 ± 0.07 beats · min−1 · mmHg−1) compared with that in the controls (−0.19 ± 0.02 beats · min−1 · mmHg−1). However, aortic-cardiac baroreflex sensitivity was virtually equal [F(1,18) = 0.002, P = 0.970] in the G-trained (−0.72 ± 0.20 beats · min−1 · mmHg−1) and control subjects (−0.71 ± 0.20 beats · min−1 · mmHg−1).
Average stimulus-response relationships of the cardiopulmonary baroreflex control of FVR for G-trained and control subjects are plotted in Fig. 4. Differences in slopes (ΔFVR/ΔCVP) between the G-trained and control groups were compared by analyzing the least-squares linear estimates generated by each subject. Average ΔFVR/ΔCVP of the cardiopulmonary baroreflex response was 46% less [F(1,18) = 8.307,P = 0.010] in the G-trained subjects (−1.4 ± 0.2 PRU/mmHg) compared with the controls (−2.6 ± 0.3 PRU/mmHg).
The average slopes of the individual subject dose-response relationships between Iso and HR were statistically indistinguishable [F(1,18) = 1.474, P = 0.241] between the G-trained subjects (1,612 ± 182 beats · μg−1 · kg−1 · min−1) and the average response of 1,323 ± 153 beats · μg−1 · kg−1 · min−1in the controls. Likewise, there was no statistical difference between G-trained and control subjects in the average slope of the individual subject dose-response relationships between Iso and LVR [−534 ± 133 PRU μg · kg−1 · min−1 and −690 ± 294 PRU μg · kg−1 · min−1, respectively; F(1,18) = 0.234,P = 0.634]. Figure5 (top and middle) represents the regressions calculated from the means ± SE HR and LVR at each Iso level. In contrast, the average slope of the individual subject dose-response relationships between PE and LVR in the G-trained subjects (32.8 ± 3.4 PRU · μg−1 · kg−1 · min−1) was greater [F(1,18) = 4.858,P = 0.043] than the average slope of 19.5 ± 4.7 PRU · μg−1 · kg−1 · min−1measured in the control group. The regression calculated from the means ± SE LVR at each PE level is presented in thebottom panel of Fig. 5. During graded PE infusion, ΔHR/ΔMAP relationship was −1.2 ± 0.2 beats · min−1 · mmHg−1 in the G-trained group compared with −0.5 ± 0.1 beats · min−1 · mmHg−1 in the control subjects [F(1,18) = 8.51,P = 0.009].
LV and Compliance
Total LV of the G-trained subjects (12.5 ± 0.5 liters) and the controls (12.1 ± 0.5 liters) was similar [F(1,18) = 0.499, P = 0.489], while calf compliance of the G-trained group (4.8 ± 0.2 ml/mmHg) was less [F(1,18) = 6.497,P = 0.020] than that of the controls (6.2 ± 0.5 ml/mmHg).
Plasma Volume and Endocrine Responses
Average total circulating blood volume of G-trained subjects (6,497 ± 469 ml) was 19% greater [F(1,18) = 3.760, P = 0.069] compared with controls (5,438 ± 228 ml), primarily as a result of higher [F(1,18) = 4.061,P = 0.060] circulating erythrocyte volume (3,110 ± 364 ml for G-trained vs. 2,310 ± 96 ml for controls). Plasma volume was statistically similar [F(1,18) = 1.637, P = 0.218] between G-trained (3,387 ± 140 ml) and control subjects (3,129 ± 137 ml). Greater erythrocyte volume with similar plasma volume was consistent with the higher [F(1,18) = 5.881,P = 0.027] hematocrit measured in the G-trained group (47.0 ± 1.7%) compared with the controls (42.5 ± 0.5%).
Resting total circulating plasma NE and Epi were not statistically different [F(1,14) ≤ 1.146, P≥ 0.303] between the G-trained (975 ± 97 and 62 ± 5 ng, respectively, n = 7) and control (806 ± 101 and 53 ± 15 ng, respectively, n = 9) groups. During adrenoreceptor agonist infusions, plasma NE was decreased by PE and increased by Iso [F = 56.38, P < 0.0001, df = 2] in both groups (Fig.6). However, plasma NE concentrations at all infusion conditions could not be statistically differentiated [F = 0.949, P = 0.399, df = 2] between G-trained and control groups (Fig. 6).
When the head-to-foot G acceleration vector is chronically diminished as in bedrest or spaceflight, humans demonstrate reduction in orthostatic tolerance associated with impaired blood pressure regulation. These changes include increased compliance of the lower leg (29), alteration of vagal-cardiac nerve reflex responses (7, 9, 10, 16), contracted circulating blood volume (7, 21), and lower SV (7, 10, 16). It was therefore reasonable to suspect that individuals who do and do not train at high-G acceleration might demonstrate differences in some of these mechanisms. Therefore, the present investigation was the first to test the hypothesis in humans that repeated exposure to high-G acceleration would be associated with differences in blood volume and autonomic functions opposite to those reported in subjects who have undergone adaptation to microgravity. The major findings of this study supported this hypothesis by demonstrating that G training was associated with less venous compliance in the lower extremities, greater cardiac vagal nerve response to equal carotid baroreceptor stimulation, less vasoconstriction to equal low-pressure baroreceptor stimulation, lesser decline in Q̊ and SV, fluid accumulation in the legs, elevated thoracic impedance, and TPR induced by LBNP, higher α-adrenoreceptor and HR responsiveness to PE, and larger total blood and erythrocyte volume.
It is of interest to note that the significant differences in cardiovascular mechanisms between control and G-trained subjects were associated with only 3 days of exposure to high G over a 5- to 7-day training period. High-performance aircraft pilots have provided anectodal description that their subjective G tolerance following their return to flight training after a layoff period is established after two or three flight exposures (personal communications). The linear increase in maximal tolerable G exposure with 3 days of training exposure and best G performances on day 3 of training demonstrated that two exposures were not adequate to attain maximal G tolerance. However, attainment of best G endurance by day 3of G training supports the notion that the adaptations of cardiovascular mechanisms that underlie blood pressure regulation are highly labile and rapid.
As expected, LBNP elicited increased HR, TPR, and %ΔLV, and reduced SV with subsequent reduction in MAP in the control subjects to levels comparable to those reported in previous investigations (2, 5, 6,8, 16, 34). The capacity to maintain Q̊ and MAP during an orthostatic challenge can be influenced by impaired venous return. Among other factors, reduced venous return may result from a combination of lower total circulating blood volume and blood pooled in the lower body, which can subsequently cause subnormal filling pressures and shifting to a portion of the Frank-Starling curve where capacity to buffer orthostatic hypotension is limited (23). Less reduction in SV and MAP during higher levels of LBNP (i.e., 40 and 50 mmHg) was measured in our G-trained subjects. These results corroborate those of a previous investigation in which men who had undergone 8 consecutive days of G training demonstrated less reduction in SV and MAP during a stand test compared with before G training (13).
The present investigation extended previous observations by demonstrating that enhanced defense of SV and MAP in G-trained subjects was associated with larger blood volume and less blood distribution away from the central circulation, factors that may have contributed to greater venous return and cardiac filling during LBNP. This notion was supported by smaller increases in thoracic impedance and less elevation in LV in G-trained subjects during graded LBNP. Lower leg venous compliance was observed in our G-trained subjects. This was somewhat unexpected because increased leg compliance was previously reported after G training (13). Although the reason for this apparent discrepancy is unclear, significant differences in training duration, frequency, and intensity could have contributed. Regardless, less fluid pooling in the lower extremities during LBNP in the G-trained group may in part be explained by their 23% lower average leg compliance for equal total LV compared with control subjects.
In addition to cardiac filling, higher HR contributed to maintenance ofQ̊ throughout LBNP in the G-trained subjects. With a cross-sectional experimental design, the possibility that our G-trained subjects had higher resting and orthostatic HR responses without their G training cannot be entirely dismissed. However, higher HR responses associated with orthostatic training is not without precedent. In a longitudinal experimental design, increased blood pressure maintenance and orthostatic tolerance in subjects who underwent presyncopal-limited LBNP training was associated with increased HR during LBNP (24). Baseline HR was elevated in four of six subjects at the end of LBNP training, but not to the degree observed in G-trained subjects of the present study. The average maximal LBNP training level increased from approximately −60 mmHg on day 1 of training to −70 mmHg on the final day with exposure periods of ∼15 to 24 min, respectively (average rate = 0.05 to 0.07 mmHg/s). Comparisons of similarities and differences between LBNP and centrifuge acceleration demonstrated that a rate of LBNP onset equal to 2.0 mmHg/s is required to provide physiological stimulation similar to an onset rate of 0.01 to 0.2 +G/s of centrifuge acceleration (28). Because our G-trained subjects were repeatedly exposed to acceleration onset rates as high as 6.0 +G/s, it seems likely that elevated resting and orthostatic HR in G-trained subjects could have resulted from their extraordinarily intense acceleration exposure.
There is no clear mechanism to explain an elevated resting HR in our G-trained subjects. Although elevated HR may be influenced by decreased cardiac vagal tone, R-R interval variability was similar between our two groups. Likewise, similar circulating catecholamines suggested resting sympathetic activity may not be different. There was no statistical difference in HR response to Iso, suggesting that cardiac adrenergic responsiveness was similar between groups. The possibility that G-trained subjects had greater sympathetic activity at rest and during LBNP that may not be reflected by circulating NE cannot be dismissed without direct measures of sympathetic nerve activity.
Baroreceptor-mediated cardiac responses provide a means to buffer transient changes in arterial blood pressure. There is evidence to support the notion that regular exposure to high-gravitational vectors increased baroreceptor stimulus-cardiac reflex response relationships. Less hypotension was required to elicit similar tachycardia during the transition from supine to head-up tilt in high-performance aircraft pilots compared with nonpilots (32). Similarly, the HR response to similar arterial pressure stimuli induced by Valsalva maneuvers was greater in subjects after compared with before G training (13). Measurements conducted in the present investigation extended previous observations of integrated cardiac baroreflex responses by providing specific stimuli to aortic and carotid baroreceptors. We used rapid beat-to-beat fluctuations in carotid pressure stimuli and measured R-R intervals to characterize the pressure input-vagal nerve output relationship of the carotid-cardiac baroreflex (15). With this approach, we found that the carotid-cardiac vagal nerve baroreflex stimulus-response relationship in the G-trained subjects was shifted upward and to the left such that an 87% greater maximum slope existed in the region of hypotension. Attenuated cardiac baroreflex responsiveness has been associated with greater incidence of orthostatic incapacitation (8, 9, 14, 16,26). Accentuated efferent nerve cardiac reflex responses of G-trained subjects support the notion that sensitization of the carotid-cardiac baroreflex loop may represent an important mechanism for regulation of blood pressure during adaptation to repeated orthostatic challenge.
In addition to efferent nerve baroreflex assessed with measures of R-R interval, the ratio of change in HR to change in MAP (ΔHR/ΔMAP) represents a hemodynamically meaningful measure of cardiac baroreflex responsiveness. Although there was no difference in ΔHR/ΔMAP between experimental groups to equal aortic baroreceptor stimulation in the present study, G training was associated with a pronounced elevation in ΔHR/ΔMAP during isolated carotid stimulation and equal dose of PE. These latter observations are consistent with the measurement of increased carotid-cardiac vagal nerve reflex responses. Against expectation, we observed no difference in ΔHR/ΔMAP between the controls and G-trained subjects from resting to 50 mmHg LBNP. However, the finding that Q̊ reduced significantly more in the control subjects with the same reduction in MAP during LBNP makes the interpretation regarding higher carotid-cardiac baroreflex responses in the G-trained subjects clearer. At the systemic level, Q̊ rather than HR represents the end-organ reflex response to arterial baroreceptor stimulation (23). Therefore, an alternative method of interpreting the difference in baroreflex responsiveness between our subject groups is to calculate the change in Q̊ that would be expected to occur when a given reduction in blood pressure elicits a given reflex change in HR (8, 16, 23). In the present study, average relative reduction in SV from resting to 50 mmHg LBNP was 43% in controls compared with 32% in G-trained subjects. For a resting SV of 100 ml, G-trained subjects would have a 68-ml SV at 50 mmHg LBNP compared with 57 ml in control subjects. Therefore, for a similar change in MAP during LBNP in the present study (7 mmHg), the unit difference in Q̊ would be threefold greater in the G-trained group (3.2 beats/min × 68 ml/beat = 218 ml/min) compared with the controls (1.3 beats/min × 57 ml/beat = 74 ml/min). These calculations emphasize the shortcoming of using HR baroreflex response as the sole indicator of the total cardiac baroreflex response and indicate the importance of including both HR and SV in assessing the “effective” gain of the integrated cardiac baroreflex in maintenance of arterial pressure during a hypotensive challenge.
The capacity to increase TPR represents an important mechanism for buffering against development of orthostatic hypotension. Several observations indicate that the capacity to vasoconstrict was greater in our G-trained subjects. Similar vasoactive responses to Iso between experimental groups suggested that G training probably did not affect vasodilatory adrenoreceptor responsiveness. However, vasoconstrictive responses to PE suggested greater capacity to increase TPR in G-trained subjects under the same sympathetic stimulus. In addition, the elevation in FVR induced by cardiopulmonary baroreceptor stimulation in G-trained subjects was 46% less than that of the control subjects. Because reduced blood volume elicits greater vascular resistance for equal reductions in venous pressure during low-pressure baroreceptor stimulation (38), larger total circulating blood volume of G-trained subjects may have increased their vasoconstrictive reserve (8, 16). This concept is further supported by the observation that the typical increase in TPR during graded LBNP in the control group was attenuated in G-trained subjects (Fig. 2). With higher HR and SV during LBNP, our results may suggest that G training (and perhaps G suit inflation) adapts the cardiovascular system to reduce requirements for peripheral vascular constriction by increasing the contribution of cardiac mechanisms.
In the absence of pretraining measures, the possibility that high-G training induced erythrocyte volume expansion by 35% following a 5- to 7-day period can only be speculative. However, this observation is not without precedent. In a longitudinal experimental design, Beckman et al. (1) demonstrated a 33% greater total erythrocyte volume (P < 0.001) in a group of dogs who underwent chronic exposure to only 2.6 G compared with a control group. In contrast to the results of the present investigation, the dogs also demonstrated increased plasma volume. Increased specific red cell volume, red cell count, hematocrit, and F cells values for the 2.6 G group of dogs indicated that there was an absolute polycythemia and that a greater proportion of the red cells were in the periphery compared with the 1.0 G control dogs. Although the relative elevation in erythrocyte volume in subjects of the present study and dogs were similar (35 vs. 33%), cross-study comparison remains difficult because the dogs were exposed to dramatically less G acceleration (2.6 vs. 9.0 G) for significantly longer duration (90 vs. 7 days). Unfortunately, an earlier time course for high-G-induced polycythemia was not reported (1). In any event, repeated or chronic exposure to high-G acceleration is associated with greater total blood volume in dogs (36%) and humans (19%) compared with their nonexposed controls. Taken together, these observations support the possibility that high-G exposure may have contributed to greater erythrocyte volume in our G-trained subjects.
During the training sessions, concurrent protection was required for G-trained subjects by anti-G suit inflation and straining maneuvers because of extraordinarily high levels of G acceleration. Therefore, the possibility that these protective interventions could elicit autonomic and hemodynamic adaptations independent of or in addition to the stimulus provided by exposure to high-G acceleration must be considered as part of the G-training stimulus. In a previous investigation (12), acute HR, SV, Q̊, MAP, TPR, and maximal slope of the carotid-cardiac baroreflex were virtually identical with and without anti-G suit inflation. However, it is unclear whether long-term adaptation of cardiovascular functions to repeated exposures of anti-G suit inflation may occur. There is no evidence of a training effect of repeated anti-G straining maneuvers on cardiovascular functions, although chronic resistance exercise training that can induce similar straining did not result in hemodynamic changes to LBNP or baroreflex function (36).
Because this experimental design involved a cross-sectional comparison, numerous factors other than acceleration exposure could contribute to differences in physiological functions. Gender (8, 27), height (26, 27), andV˙o 2 max (5, 34) have been associated with alterations in blood volume, autonomic functions, and orthostatic responses. To minimize these confounding factors, subjects in the control and G-trained groups were all males and matched for height, weight, body composition, andV˙o 2 max. We cannot completely dismiss the possibility that self-selected volunteers to the G-training panel may have reflected inherent cardiovascular traits. However, all subjects passed physical examinations of cardiovascular functions that qualified them for G-panel subjects. Although our G-trained group was younger than the control group, there was little difference in sympathetic-circulatory regulation of arterial blood pressure between 26- and 64-year-old men during exposure to LBNP levels similar to those used in the present study (37). It therefore seems unlikely that factors other than G training could explain differences that were measured between the two experimental groups of the present investigation.
Improved orthostatic performance was associated with less venous compliance in the lower extremities, greater vagal response to equal carotid baroreceptor stimulation, less vasoconstriction to equal low-pressure baroreceptor stimulation, less decline in SV and Q̊, fluid accumulation in the legs, elevated thoracic impedance, and increase in TPR induced by LBNP, higher α-adrenoreceptor responsiveness, and larger total circulating blood volume. The results of this investigation support the hypothesis that repeated exposure to high-G acceleration is associated with changes in blood volume and autonomic functions opposite to those reported in subjects who experience chronic removal of head-to-foot gravity stimulation.
The observations from this study provide important perspective to understanding the contribution of gravity to the evolution and adaptation of cardiovascular function. Lower venous compliance and blood pooling in the legs, higher carotid-cardiac baroreflex sensitivity, greater SV and cardiac output, larger vasoconstriction reserve, and expanded total circulating vascular volume are associated with increased orthostatic performance. Therefore, it is possible that similar changes in these physiological functions after G training represent a partial explanation for increased performance during high-G acceleration. Similarly, results from these studies might provide a physiological explanation for G-layoff effects because development of orthostatic intolerance after exposure to reduced gravity, e.g., spaceflight or bedrest, is associated with reductions in these functions. This perspective is best supported by the observation that the majority of nonpresyncopal astronauts after space missions in the United States space program were career pilots of high-performance aircraft (17). It has been hypothesized that these individuals would not have been successful in their careers if they had not developed high tolerance to G forces. The data from the present study support this notion and suggest that greater tolerance to high-G acceleration may be in part a result of chronic adaptation of physiological mechanisms that underlie blood pressure regulation. If this relationship is valid, a rationale for application of regular G-training (i.e., artificial gravity) may be developed for countermeasure training of pilots, astronauts, and bedridden patients.
From an evolutionary standpoint, physiological reflexes required by humans to maintain adequate blood perfusion to the brain in our normal earth environment evolved around the need to counteract the upright posture in terrestrial gravity (1 G). Physiological mechanisms that underlie blood pressure regulation in humans (e.g., baroreflexes, control of vascular volume) have adapted over time to the constant day-to-day stimulation from assuming various G-factor stimuli, i.e., variation in posture. Figure 7 presents blood volume and carotid-cardiac baroreflex data from subjects in the present investigation who were exposed to high gravity compared with values from subjects in their normal ambulatory condition (present investigation, Ref. 9) and exposed to simulated low gravity (Ref. 9). Examination of adaptation of these physiological responses to changing gravity stimuli may be particularly significant because blood volume and carotid-cardiac baroreflex responsiveness are two primary factors (in addition to height) in predicting time to syncope (26). Comparisons of physiological functions that increase as gravity stimuli increase and decrease as gravity stimuli decrease provide insight to two fundamental biological relationships. First, our orthostatic capacities constantly adapt along a G-factor continuum. Second, mechanisms that underlie orthostatic tolerance are highly plastic and trainable. These important relationships can be instrumental in development of clinical techniques for enhanced rehabilitation of patients as well as design of operational countermeasures for patients, astronauts, and military personnel to limit the debilitating effects of disease- or mission-induced orthostatic hypotension.
The author thanks the following individuals for their part in making this project a success: Dr. Sheryl Wright at University of Texas Health Sciences Center at San Antonio, Drs. Sandra Oswald, James Slauson, Joseph Deering, Theodore Arevalo, Jeb Pickard, Cullen Hardy, and William Kruyer at Brooks Air Force Base for providing medical supervision; Craig Reister at Rothe Development for engineering support; Richard Owens, Russell Woods, Gary Muniz, and Jim Lutze at Rothe Development for laboratory technical support and assistance with data collection and reduction; TSgt Troy Murray and SSgt Charles Hinshaw at the Physiology Research Section of the Air Force Research Laboratory for medical support during experiments and collection of blood samples; Marion Merz at Bionetics, Kennedy Space Center for analysis of plasma volume; Dr. Marty Javors at University of Texas Health Sciences Center for analysis of plasma catecholamines; and the subjects for their cheerful cooperation.
This project was supported in part by an entrepreneurial research grant administered under the United States Air Force Office of Scientific Research and by the National Aeronautics and Space Administration administered under Grants 199–14–17–07 and W-19,515.
The views expressed herein are the private views of the author and not to be construed as representing those of the Department of Defense or Department of the Army.
Address for reprint requests and other correspondence: U. S. Army Institute of Surgical Res. Library, 3400 Rawley E. Chambers Ave., Bldg. 3611, Fort Sam Houston, TX 78234-6315 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2001 the American Physiological Society