Journal of Applied Physiology


Because it is not clear that the induction of orthostatic intolerance in returning astronauts always requires prolonged exposure to microgravity, we investigated orthostatic tolerance and autonomic cardiovascular function in 16 healthy subjects before and after the brief micro- and hypergravity of parabolic flight. Concomitantly, we investigated the effect of parabolic flight-induced vomiting on orthostatic tolerance, R-wave-R-wave interval and arterial pressure power spectra, and carotid-cardiac baroreflex and Valsalva responses. After parabolic flight 1) 8 of 16 subjects could not tolerate 30 min of upright tilt (compared to 2 of 16 before flight);2) 6 of 16 subjects vomited; 3) new intolerance to upright tilt was associated with exaggerated falls in total peripheral resistance, whereas vomiting was associated with increased R-wave-R-wave interval variability and carotid-cardiac baroreflex responsiveness; and 4) the proximate mode of new orthostatic failure differed in subjects who did and did not vomit, with vomiters experiencing comparatively isolated upright hypocapnia and cerebral vasoconstriction and nonvomiters experiencing signs and symptoms reminiscent of the clinical postural tachycardia syndrome. Results suggest, first, that syndromes of orthostatic intolerance resembling those developing after space flight can develop after a brief (i.e., 2-h) parabolic flight and, second, that recent vomiting can influence the results of tests of autonomic cardiovascular function commonly utilized in returning astronauts.

  • postural tachycardia syndrome
  • vomiting
  • microgravity
  • hypergravity
  • autonomic
  • space flight

orthostatic intolerance is common in returning astronauts (6,13). However, it is not clear that prolonged exposure to microgravity is always required to induce this problem. For example, some astronauts with normal preflight orthostatic tolerance have become orthostatically intolerant after space missions as brief as 2.5 days (Shuttle program) (7) and 9 h (Mercury program) (25). In addition, we have recently noted that individuals returning from even shorter (i.e., 2-h) parabolic flights commonly complain of orthostatic lightheadedness, even when motion sickness has not been present (47). Most impressively, in military aviators, mere seconds of exposure to gravitational forces in the headward direction are sufficient to increase the risk of syncope during subsequent exposure to gravitational forces in the footward direction, the so-called “push-pull effect” (1, 30). All of these phenomena suggest that autonomic cardiovascular function may be altered very rapidly after changes in ambient gravity.

In the case of parabolic flight, the gravitoinertial resultant force vector remains constantly perpendicular to the deck of the aircraft (24). Therefore, passengers who remain in the seated upright position throughout the parabolas effectively experience changes in gravity only along the headward-to-footward, or Gz, axis of the body (24, 48). A principal goal of the present study was to formally determine whether or not seated parabolic flight alone, which specifically consists of brief, repetitive exposures to both 0.01 Gz and 1.8 Gz(48), is able to induce changes in autonomic cardiovascular and orthostatic function persisting into the postlanding period. We were particularly interested in identifying changes in autonomic cardiovascular function immediately after landing because an equivalent time period has been utilized to identify the majority of autonomic abnormalities in returning astronauts (6, 7, 10, 11,13, 57).

Besides orthostatic lightheadedness, another condition that commonly affects both returning astronauts and individuals participating in parabolic flight is vomiting due to motion sickness (23, 44). In addition, Buckey et al. (6) recently described a unique form of orthostatic intolerance in two returning astronauts, possibly related to recent vomiting, that differed from the archetypal forms of post-space flight intolerance in that it was unassociated with any clear hypotensive event. To our knowledge, however, the effect that recent vomiting has on orthostatic tolerance has not been prospectively studied in any human subject group, including returning astronauts. A second, related goal of this study, therefore, was to utilize the inevitable severe motion sickness generated in some individuals during and after parabolic flights to investigate relationships between recent vomiting and concomitant changes, if any, in autonomic cardiovascular and orthostatic function.

Our specific hypotheses were that 1) autonomic cardiovascular dysfunction and orthostatic intolerance do indeed occur with an increased frequency after a seated 2-h parabolic flight but that 2) the type and/or degree of autonomic cardiovascular dysfunction and orthostatic intolerance necessarily differs in individuals who have and who have not vomited as a result of such flight.



Sixteen healthy test subjects (10 men and 6 nonpregnant women, mean age 32 yr, range 22–45 yr) participated in the study, which was approved by the Johnson Space Center Institutional Review Board. All subjects were free of cardiopulmonary, renal, or other systemic disease, and each gave written, informed consent after passing a U.S. Air Force Class III physical examination (51). In addition, all subjects were nonsmokers who had normal blood pressure (BP), hemoglobin/hematocrit, creatinine, electrolytes, liver function tests, and urinalyses (including drug screens). Caffeine, alcohol, heavy exercise, anti-motion sickness medications, and all other medications were strictly prohibited beginning 24 h before any testing, which commenced in the morning hours.

Parabolic Flights

While loosely restrained at the waist, subjects flew 4 sets of 10 parabolas in the seated position aboard NASA's KC-135 aircraft, a Boeing 707 specifically modified for parabolic flight. During their flights, subjects were instructed to avoid unnecessary head movements by looking forward at a computer monitor-sized target placed ∼1.5 m in front of them. As verified by an accelerometer mounted inside the aircraft, single parabolas consisted of the following three phases, each lasting ∼20–25 s: 1) “pull-up” with increased G-load of up to +1.8 Gz; 2) microgravity (∼0.01 Gz); and 3) “pull-out” with increased G-load of up to +1.8 Gz (see Ref.48).

Motion Sickness Scoring

The nature of the parabolic flight stimulus is such that the majority of subjects typically experience either no motion sickness symptoms or severe symptoms that include vomiting. Therefore, for statistical analyses related to our second hypothesis, subjects were separated into two principal groups: vomiters and nonvomiters. For descriptive purposes only, Graybiel et al.'s standard 16-point motion sickness scale (15) was also utilized to obtain maximum motion sickness scores for each subject during 1) seated activities on the aircraft and 2) postflight tilt testing, ∼40–70 min after exit from the aircraft.

Cardiovascular and Cerebrovascular Measurements

Cardiovascular data were collected during identical pre- and postflight sessions in the supine and tilted-upright positions at a room in NASA's hangar facility at Ellington Air Field, Pasadena, TX. The temperature and relative humidity were controlled in this testing room both pre- and postflight within the ranges, respectively, of 22–24°C and 40–50%. The preflight session occurred 1–5 days before parabolic flight, and the postflight session took place immediately after parabolic flight. Two to three hours before each session, subjects consumed the same breakfast consisting of fruit, cereal grains, and optional low-fat milk.

Before actual testing, subjects were first instrumented with1) electrocardiographic leads and electrodes (including an electrode for impedance measurements of abdominal muscle respiratory excursions; Physio-Control, Redmond, WA), 2) impedance cardiographic leads and electrodes (BoMed NCCOM-3, Irvine, CA), and 3) a finger photoplethysmographic device (Finapres 2300, Ohmeda, Englewood, CO) that was fixed to the level of subjects' hearts (via a tilt-table arm extension) for beat-to-beat estimates of BP. The continuous cardiovascular signals from all of these devices were digitally recorded and integrated by using a special software program (29,48) that automatically entrains beat-to-beat heart rate (HR), stroke volume (SV), and mean BP (MBP) to create a real-time pictorial representation for beat-to-beat cardiac output [(CO) = HR × SV] and total peripheral resistance[(TPR) = MBP/CO]. Although the BoMed bioimpedance device may not give accurate information pertaining to absolute SV, it does provide accurate, reliable, and reproducible estimations of changes in SV (36).

Throughout supine and upright testing, end-tidal CO2 was also measured via a nasal probe (Puritan-Bennett, Wilmington, MA), and a 2-MHz flat ultrasound probe (TRANSPECT, Medasonics, Mountain View, CA) was mounted over the right temporal bone to obtain transcranial Doppler (TCD ) recordings of blood flow velocities through the right middle cerebral artery. The principal TCD indexes derived for the present study were the middle cerebral artery mean flow velocity (MCA-MFV) and the estimated cerebral vascular resistance (CVRest), which is the estimated MBP (hydrostatically corrected) at the level of the circle of Willis divided by the MCA-MFV (13). Finally, manual recordings of systolic and diastolic (SBP and DBP, respectively) and pulse BP were also obtained on a minute-to-minute basis before, during, and after tilt via a sphygmomanometer attached to the arm opposite the Finapres. During tilt, these manual recordings were increased to every 30 s at the onset of new symptoms, TCD changes, or a marked decrease in BP.

Both pre- and postflight, the specific sequential activities of test subjects were as follows: 1) ambulation to the testing room;2) instrumentation (as noted above); 3) supine rest for 15 min; 4) supine controlled breathing at 0.25 Hz for 5 min or until 256 consecutive heart beats and beat-to-beat arterial pressures were recorded for subsequent spectral analyses;5) supine carotid-cardiac baroreflex testing (10 min);6) supine Valsalva maneuver testing (5 min); 7) 3–5 min of additional supine rest; and, finally, 8) upright tilt testing (30 min or less). The majority of these activities are described in greater detail below. On the aircraft itself, both immediately before and after flight, subjects also performed Valsalva tests in the seated position. These seated tests complemented our earlier investigation (in the same subjects) of seated responses to Valsalva maneuvers during the in-flight period (48). For all subjects, the time from the last parabola until the beginning of the postflight seated testing protocol was ∼20–25 min. Thereafter, transfer to the testing room in the hangar (∼50 m from the aircraft) and the beginning of the postflight supine testing protocol occurred within 10–15 min of aircraft wheel stop. Therefore, variances in the time of transfer were only on the order of 5–10 min, usually to allow motion sick subjects time to recover from an episode of emesis. In the testing room itself, maximum delays before upright tilt were also on the order of 5–10 min, usually for the repetition of a carotid-cardiac baroreflex or Valsalva test. Although we did not allow subjects to consume any food or fluid in flight, if they became hungry or thirsty, we did allow them to have sips of water and/or a cracker immediately before the postflight supine testing protocol.

Tilt Tests

After supine autonomic testing, both pre- and postflight, subjects were secured and pitched acutely (within 10–12 s) into the 80-degree head-up position by using a standard clinical autonomic tilt table (Tri W-G, Valley City, ND). Once attained, the 80-degree head-up position was sustained for 30 min or until orthostatic intolerance ensued. During minutes 1-10 of the upright position, both pre- and postflight, a majority of subjects also performed controlled breathing at 0.25 Hz for a total of 5 min. Orthostatic intolerance in this study was defined on the basis of the following tilt-test termination criteria: subject request at any time; sudden drop of SBP > 25 mmHg or of DBP > 15 mmHg; sudden and sustained drop in HR of > 15 beats/min; absolute HR < 40 beats/min for subjects whose resting absolute HRs were > 50 beats/min; absolute SBP < 90 mmHg; severe lightheadedness, severe nausea, or actual vomiting.

For tilt-related analyses, we first obtained the value of each cerebral vascular and cardiovascular parameter (except those derived from the Finapres) at the time of the manual BP measurements. Serial values for each parameter were then averaged for individual subjects according to three epochs: epoch 1, the average of the two minute-to-minute values in the supine position immediately preceding upright tilt; epoch 2, the average of all minute-to-minute values from minutes 1-10 of upright tilt (excluding any value obtained during controlled breathing or during or after the minute of orthostatic failure, if failure occurred during this epoch); and epoch 3, the average of either one or two values from the last minute of upright tilt. The epochal averages from individual subjects were then used to derive corresponding epochal averages for groups of subjects (i.e., whole group, tolerant subjects, intolerant subjects). In one individual, the averages from epoch 1 for MCA-MFV and CVRest had to be obtained from a slightly earlier portion of the supine period, both pre- and postflight, because of intermittent electrical interference in the TCD.

Derivation of Power Spectra

Spectral powers were derived from the 5-min series of consecutive R-wave-R-wave (R-R) intervals, SBPs, and DBPs collected in the supine position during metronome-controlled breathing (5) at 0.25 Hz both pre- and postflight. Before preflight testing, subjects chose a comfortable respiratory excursion (tidal volume) and practiced breathing to the metronome at that excursion. They were then asked to use this same excursion throughout all subsequent pre- and postflight tests involving controlled-frequency breathing. During data collection itself, on the basis of our observations of end-tidal CO2 levels and abdominal and nasal respiratory movements and tracings, we also provided verbal feedback to the subjects as necessary to ensure that they were maintaining gross consistency in respiration.

For spectral analyses, the Welch algorithm for averaging periodograms (56) was used in accordance with the method of Rabiner et al. (43). Specifically, the continuous series of artifact-free R-R intervals, SBPs, or DBPs was fitted to a cubic spline function, interpolated at 8 Hz to obtain equidistant time intervals, and divided into seven equal overlapping segments. Segments were then detrended, Hanning window filtered, fast-Fourier transformed, and averaged to produce the spectrum estimate. Spectral power was integrated over three defined frequency bandwidths: “low” frequencies between 0.05 and 0.15 Hz; “high” (or respiratory) frequencies between 0.20 and 0.30 Hz; and all frequencies (i.e., “total power”) below 0.50 Hz (22). We also calculated a “sympathovagal index,” defined as the ratio of the low-frequency power of SBP to the high-frequency power of R-R intervals. This index resembles (but is not identical to) the sympathovagal index recently proposed by Novak et al. (38).

Carotid-Cardiac Baroreflex Responsiveness

Both pre- and postflight, supine carotid-cardiac baroreflex responsiveness was measured in subjects via pressure changes applied to a tightly sealed Silastic neck chamber connected to a computer-controlled bellows (E-2000 Neck Baro Reflex System, Engineering Development Laboratories, Newport News, VA) (53). R-R interval responses to the neck stimuli were measured rather than HR responses because the former more accurately reflect actual changes in efferent vagal activity (40). During held expiration, neck chamber pressure was raised to +40 mmHg and then reduced to −60 mmHg in consecutive R-wave-triggered steps of −20 mmHg. This sequence was then repeated seven times, and the responses were averaged for each test subject. R-R interval responses to carotid baroreceptor stimulation, defined as carotid distending pressure (SBP minus neck pressure), were reduced to the maximum slope of the stimulus-response relation, the maximum range of R-R interval responses, and the operational point (10, 11).

Valsalva Measurements

Valsalva maneuvers were completed at an expiratory pressure of 30 mmHg for 15 s, as previously described (48). Before the strains, which were performed in triplicate, subjects first had at least 15 min of rest in the assigned postural configuration (i.e., supine or seated). Each strain was also preceded and followed by at least 1 min of controlled-frequency breathing at 0.25 Hz. To produce the strains, subjects blew into a mouthpiece connected by short plastic tube to a calibrated pressure gauge while the electrocardiogram, impedance cardiogram, and arterial and expiratory pressures were continuously recorded.

Because responses during phases I and III of Valsalva maneuvers are believed to reflect mostly mechanical changes (3, 45), we focused our analyses on variations in MBP during the “autonomic” Valsalva phases II and IV. Changes in MBP during phases II and IV were specifically calculated as follows: 1) Δ early phase II (phase IIe) was the change in MBP occurring between the maximal MBP value during phase I and the minimal MBP value during phase IIe; 2) Δ late phase II (phase IIl) was the change in MBP occurring between the minimal MBP value during phase IIe and the maximal MBP value during phase IIl; and 3) Δ phase IV was the change in MBP occurring between the minimal MBP value during phase III and the maximal MBP value during phase IV. In addition to the absolute changes in MBP, we also calculated the temporal duration of changes in the MBP response during Valsalva phases IIe and IIl(see Ref. 48). During the immediate postflight period in the aircraft, one subject was not able to perform seated Valsalva maneuvers because of severe motion sickness.


All results are reported as means ± SE. Because normality was often violated, we used nonparametric statistics for all comparisons. Specifically, we used the Wilcoxon signed-rank test for within-group comparisons (i.e., before vs. after parabolic flight) and the Mann-Whitney rank sum test for between-group comparisons (i.e., vomiters vs. nonvomiters) (14). Statistical significance was accepted at P < 0.05.


Individual Subject Characteristics

Table 1 outlines the susceptibility of individual subjects to both motion sickness and orthostatic intolerance. Six of the 16 subjects vomited after parabolic flight, whereas 10 did not. In addition, eight subjects (5 of the 6 vomiters and 3 of the 10 nonvomiters) had orthostatic intolerance postflight. However, only six of these eight subjects had new orthostatic intolerance, because two subjects (subjects 15 and16, both female vomiters) also had orthostatic intolerance preflight. These two individuals experienced typical episodes of vasovagal presyncope (33) at similar periods of upright tilt both pre- and postflight.

View this table:
Table 1.

Motion sickness and orthostatic intolerance after parabolic flight

Responses to Upright Tilt

Tilt-related hemodynamic parameters pre- and postflight for the whole group (n = 16) and for vomiters (n = 6) vs. nonvomiters (n = 10) are shown in Table 2. For the whole group, significant changes postflight compared with preflight included decreased DBP, MBP, and TPR and increased SV and CO in the supine position; decreased SBP, TPR, and MCA-MFV, and increased CO duringminutes 1-10 of the upright position; and decreased DBP, MBP, TPR, MCA-MFV, and end-tidal CO2 and increased HR and CO during the last minute of upright tilt. The decreased end-tidal CO2 during the last minute of postflight upright tilt was not related to any change in the natural respiratory rate (i.e., 3.5 ± 0.1 s/breath postflight vs. 3.5 ± 0.1 s/breath preflight; P > 0.05). In general, the directional changes that occurred postflight within the nonvomiter subgroup closely paralleled those that occurred in the whole group. On the other hand, within the vomiter subgroup, upright HR was not accentuated postflight, and pre- to postflight changes (decreases) in upright TPR were also less extreme than in the nonvomiter subgroup. Both pre- and postflight, there were also no significant differences in tilt-related parameters between vomiters and nonvomiters, with the one exception being that, postflight, vomiters had a lower SBP than nonvomiters during the last minute of upright tilt. Figure 1 shows the HR and manual BP changes that occurred during pre- and postflight upright tilt in each of the 14 subjects who were tolerant to preflight upright tilt.

View this table:
Table 2.

Responses to upright tilt: whole group and vomiters vs. nonvomiters

Fig. 1.

Individual pre- and postflight manual blood pressure and heart rate (HR) responses to upright tilt in the 14 subjects who were orthostatically tolerant before flight. Solid lines, nonvomiters; broken lines, vomiters. Note especially the overall greater range of responses postflight.

Postflight Orthostatic Failure

To reiterate, 6 of the 14 subjects who were tolerant to preflight upright tilt developed orthostatic intolerance during postflight upright tilt (Table 1). Factors that distinguished the six newly intolerant subjects from the eight persistently tolerant subjects were as follows. First, in the supine position, the newly intolerant subjects had decreased TPR compared with the persistently tolerant subjects, both postflight (9.1 ± 0.1 vs. 13.0 ± 1.1 mmHg · l−1 · min−1,P < 0.01) and preflight (10.2 ± 0.5 vs. 14.6 + 1.0 mmHg · l−1 · min−1,P < 0.01). Second, during the transition from the early portion of upright tilt to the end of upright tilt, the newly intolerant subjects also had greater increases in HR than the persistently tolerant subjects, both postflight (14 ± 4 vs. 2 ± 2 beats/min, P < 0.05) and preflight (8 ± 2 vs. 1 ± 2 beats/min, P < 0.05). The one difference that occurred between the two groups postflight but not preflight was that, during this same final upright transition, the newly intolerant subjects developed significantly greater decreases in TPR than the persistently tolerant subjects only postflight (−1.9 ± 0.6 vs. +0.6 ± 0.7 mmHg · l−1 · min−1,P = 0.02), not preflight (−0.1 ± 0.7 vs. +0.4 ± 0.8 mmHg · l−1 · min−1, P= 0.49).

Figures 2-4 illustrate in detail the representative modes of postflight orthostatic failure that developed in newly intolerant subjects, as listed in Table 1. In nonvomiters with new orthostatic intolerance, significant postural tachycardia always developed along with either absolute hypertension (Fig. 2) or hypotension (Fig.3) (see also Fig. 1). Newly intolerant nonvomiters also developed exaggerated upright hypocapnia and cerebral vasoconstriction postflight relative to preflight, their overall upright signs and symptoms resembling those found in the clinical postural tachycardia syndrome (POTS) (27, 38, 39). In contrast, in newly intolerant vomiters, postflight alterations were more localized to the cerebral circulation, with less remarkable changes affecting the systemic circulation unless and until actual vomiting recurred (Fig. 4). Although all of the newly intolerant vomiters had at least mild decreases in upright SBP postflight (relative to preflight), none of these individuals developed absolute hypotension (i.e., SBP < 90 mmHg; Fig. 1).

Fig. 2.

Preflight (gray) and postflight (black) responses to upright tilt in 1 of the 3 nonvomiters who developed postural tachycardia and orthostatic intolerance during postflight tilt testing only. The lines within each tracing represent temporal regressions from the end of upright tilt back to the second minute of the upright position. CVRest, estimated cerebral vascular resistance; MCA-MFV, mean flow velocity of blood in the right middle cerebral artery; CO2, nasal end-tidal carbon dioxide level; MBP, mean blood pressure via Finapres; SV, stroke volume; TPR, total peripheral resistance. Preflight, this subject (subject 7, Table 1) completed the maximum 30 min of upright tilt in an unremarkable fashion. Postflight, however, he developed palpitations and severe lightheadedness, requesting premature termination of the tilt test nearupright minute 21 (dark vertical line). Besides progressive postural tachycardia, this subject's postflight signs included1) a gradual failure of the upright TPR response;2) mild relative hypertension; and 3) relative hypocapnia and cerebral vasoconstriction with a reversal of the direction (compared to preflight) of the temporal regression slopes for CVRest, MCA-MFV, and CO2.

Fig. 3.

Preflight (gray) and postflight (black) responses to upright tilt in a second nonvomiter (subject 5, Table 1) who developed postural tachycardia and orthostatic intolerance during postflight tilt testing only (see Fig. 2 for an explanation of temporal regression lines). The postflight postural tachycardia of this subject occurred more immediately than that of subject 7 (Fig. 2). In addition, it was ultimately accompanied, not by persistent hypertension, but by sudden, symptomatic hypotension. Although the Finapres signal was unfortunately lost in this subject betweenminutes 0 and 3.5 of postflight upright tilt, the hypotensive episode postflight can nonetheless be seen between the dark vertical lines (near upright minute 22.5), which represent the beginning and end, respectively, of emergent downward tilt. The strong relationship between MCA-MFV and end-tidal CO2 is also illustrated by the similar detrending (i.e., acute lowering) that occurred in these two parameters during controlled breathing (stippled area, upright minutes 5–10) both pre- and postflight.

Fig. 4.

Expanded view of minutes 8–14 only of the postflight response to the upright position in a subject (subject 13, Table 1), who experienced nausea and multiple episodes of vomiting during parabolic flight. Although this subject was virtually free of motion sickness symptoms in the supine position 35–40 min after landing, she redeveloped mild nausea nearupright minute 11 of the postflight tilt test that progressed (suddenly) to retching and vomiting beginning atupright minute 12.5. The most illustrative aspect of her response to postflight tilt was that, when she redeveloped mild nausea (open bar), she increased the depth of her abdominal respiratory-muscle excursions (bottom) and became relatively hypocapnic and lightheaded. At the same time, MCA-MFV decreased and CVRestincreased (with TPR remaining relatively unchanged vs. preflight; not shown). During her subsequent upright retching and emesis, exemplified by the inordinately large, paroxysmal abdominal respiratory-muscle excursions in the shaded region, a bradycardia developed, end-tidal CO2 and MCA-MFV decreased further, and the Finapres-derived parameters (MBP, TPR, and CVRest) swung wildly, possibly (but not definitively) related to finger-cuff-related artifact. Of the other two vomiters who completed uneventful preflight (but not postflight) tilt tests, one (subject 14, Table 1) had a postflight pattern very similar to that of subject 13, whereas the other (subject 12, Table 1) did not develop postflight upright nausea. Instead, subject 12 developed severe (and isolated) lightheadedness after only 4 min of postflight upright tilt along with abrupt changes in end-tidal CO2, MCA-MFV, and CVRest resembling those shown in the open bar of Fig. 4.

Spectral Power of Supine R-R Intervals and Arterial Pressures

Pre- to postflight changes in supine R-R interval and arterial pressure spectral powers for the whole group and for vomiters vs. nonvomiters are shown in Table 3. After parabolic flight, the high-frequency and total power of R-R intervals increased in vomiters (P < 0.05 for total power only) but decreased (nonsignificantly) in nonvomiters, leading to significant between-groups differences in these parameters. In addition, vomiters had no changes in their arterial pressure spectral powers postflight, whereas nonvomiters, like the group as a whole, had increases in the total power of both SBP (P < 0.05) and DBP (P < 0.01). The nonsignificant decrease in the low-frequency power of SBP in vomiters and the nonsignificant increase in nonvomiters translated into a significant between-groups difference in this parameter postflight (P < 0.05). Sympathovagal index followed the same general pattern, with the difference between vomiters and nonvomiters reaching a similar level of significance postflight (P < 0.05).

View this table:
Table 3.

Spectral power of supine R-R intervals and arterial pressures pre- and postflight

The power spectral results might be summarized as follows: the vomiter group tended to respond to parabolic flight with enhanced R-R interval variability, whereas the nonvomiter group tended to respond to parabolic flight with enhanced arterial pressure variability, primarily in the low-frequency region. These relationships are illustrated in detail in Fig. 5.

Fig. 5.

Average R-wave-R-wave (R-R) interval and systolic pressure spectral power in vomiters and nonvomiters before and after parabolic flight. Spectral power was integrated over three defined-frequency bandwidths: low (0.05–0.15 Hz), high (0.20–0.30 Hz), and total (0–0.50 Hz). Vomiters displayed significantly greater (P < 0.05) R-R interval high-frequency and total spectral powers after parabolic flight. Nonvomiters displayed a significantly greater (P < 0.05) systolic pressure low-frequency spectral power and sympathovagal index (see text) after parabolic flight.

Carotid-Cardiac Baroreflex Responses

For the group as a whole, parabolic flight did not affect the maximum slope, range, or operational point of the carotid-cardiac baroreflex. However, postflight compared with preflight, vomiters had a significant increase in the maximum slope (3.7 ± 0.6 vs. 2.4 ± 0.7 ms/mmHg, P = 0.03), a nearly significant increase in the range (204 ± 31 vs. 153 ± 40 ms,P = 0.06), and no change in the operational percentage, whereas nonvomiters had no changes in any of these parameters (Fig.6).

Fig. 6.

Average carotid baroreflex stimulus-response relations from vomiters and nonvomiters before and after parabolic flight. Vomiters had a nearly significant increase in range (P= 0.06), a significant increase in maximum slope (P = 0.03), and no change in operational percentage of the baroreflex. Nonvomiters had no changes in these parameters. The open symbols represent the R-R intervals at 0 mmHg neck pressure.

Valsalva Responses

Table 4 shows the pre- and postflight Valsalva responses of the whole group and of vomiters vs. nonvomiters. Postflight, in the seated (but not in the supine) position, the absolute MBP responses of the whole group and of vomiters became significantly attenuated during Valsalva phases IIeand IIl (P < 0.05), whereas the absolute MBP responses of nonvomiters were unchanged. In addition, the temporal duration of seated Valsalva phase IIe increased postflight in vomiters (P < 0.05) but decreased (nonsignificantly) in nonvomiters, leading to a significant postflight between-groups difference in this parameter (P = 0.02).

View this table:
Table 4.

Effect of parabolic flight on responses to Valsalva maneuvers

Autonomic Changes in Tolerant Vs. Intolerant Nonvomiters

Finally, because autonomic cardiovascular function was independently influenced by the presence of recent vomiting (Tables3-4 and Figs. 5-6), we also analyzed postflight differences in supine autonomic cardiovascular function between nonvomiters who were tolerant vs. intolerant to postflight upright tilt (Fig.7). Compared with the seven nonvomiters who remained tolerant during postflight upright tilt, the three nonvomiters who became intolerant had 1) greater percentage increases in the low-frequency power of SBP and DBP from pre- to postflight (P < 0.05 for each), 2) greater percentage increases in the sympathovagal index from pre- to postflight (P < 0.05), 3) greater percentage increases in the absolute MBP response during supine Valsalva phase IIl from pre- to postflight (P < 0.05), and 4) a trend toward decreases (rather than increases) in the range (P = 0.07) and maximum slope (P = 0.07) of the carotid-cardiac baroreflex from pre- to postflight.

Fig. 7.

Postflight percent changes in measures of supine autonomic cardiovascular function in nonvomiters who did and who did not develop postflight orthostatic intolerance; n = 7 (tolerant nonvomiters) and 3 (intolerant nonvomiters), respectively. *P < 0.05 vs. tolerant nonvomiters, Mann-Whitney's rank sum test.


Our results indicate that seated parabolic flight alone, which consists of brief, repetitive exposures to both microgravity (0.01 Gz) and hypergravity (1.8 Gz), is sufficient to reduce orthostatic tolerance in susceptible individuals. This conclusion is supported not only by the fourfold increase in the number of subjects who developed orthostatic intolerance after (compared to before) parabolic flight (Table 1), but also by the statistically significant hemodynamic changes, i.e., decreased BP, TPR, and MCA-MFV and increased HR and CO, that developed postflight in the group as a whole (Table 2). Our data also indicate that subjects who have and who have not vomited as a result of parabolic flight develop differing syndromes of orthostatic intolerance as well as differing directional changes in autonomic cardiovascular function after flight.

Postflight Orthostatic Intolerance

The six subjects who developed new orthostatic intolerance after parabolic flight were distinguished from the eight persistently tolerant subjects by having decreased TPR levels in the supine position (both pre- and postflight), increased HR responses in the upright position (both pre- and postflight), and decreased TPR responses in the upright position (postflight only). Because increased HR (13) and decreased TPR responses (6, 13) are also key indicators for (and predictors of, Ref. 13) postspaceflight orthostatic intolerance, the mechanisms underlying orthostatic intolerance after space flight and parabolic flight might be similar. Parabolic flight, however, does not involve prolonged exposures to microgravity but only short, repetitive exposures to both micro- and hypergravity. Therefore, several factors historically invoked to explain post-spaceflight orthostatic intolerance, i.e., sustained cephalad fluid-shifting, disuse of baroreceptors, loss of cardiac mass, etc., cannot explain postparabolic flight orthostatic intolerance. Other etiologies must be sought. Moreover, in the present study, three of the six subjects who developed new intolerance after flight were resistant to motion sickness, with two of the three being completely resistant at all times (Table 1). Therefore, the postflight presyncope of these individuals also cannot be attributed to sickness-related factors such as nausea or fluid loss.

Recently, Fritsch-Yelle et al. (13) suggested that changes in central modulation of autonomic nervous system function might contribute to suboptimal TPR responses in orthostatically intolerant returning astronauts. In addition, Yates and Kerman (58) have suggested that sustained unloading of the otolith organs could contribute to central autonomic dysfunction after space flight, because signals from the otoliths are known to be involved in central control of peripheral adrenergic function (58, 59). The notion that altered labyrinthine inputs could contribute both to the push-pull effect and to suboptimal TPR responses after either space flight or parabolic flight seems to be supported by an early finding of Colehour and Graybiel (8). In the 1960s, these investigators demonstrated that, unlike healthy subjects, individuals with bilateral labyrinthine deficiency do not have increases in urinary norepinephrine excretion immediately after an acrobatic flight stress. In the present study, the hypertensive presyncope that developed in one of the subjects with POTS-like intolerance after parabolic flight (i.e.,subject 7, Fig. 2) also potentially supports a central etiology for postflight autonomic dysfunction. In the setting of clinical POTS, upright hypertension is especially suggestive of dysautonomia originating in the brain stem (19, 28).

With respect to the new orthostatic intolerance of vomiters, this was invariably accompanied, not by absolute hyper- or hypotension (i.e., SBP >140 or <90 mmHg, Fig. 1), but by a worsening of upright hypocapnia and cerebral vasoconstriction (Fig. 4). This result is potentially important because it might provide a physiological explanation for the nonhypotensive form of orthostatic intolerance recently described (but not explained) in two returning astronauts by Buckey et al. (6). The ability of hypocapnia alone to elicit lightheadedness, cerebral hypoperfusion, and early presyncope in patients prone to clinical orthostatic intolerance has recently been demonstrated conclusively by Novak et al. (39). However, in both Novak et al.'s patients and in the subjects who developed POTS-like intolerance in the present study (Figs. 2-3), exaggerated hypocapnia during upright tilt can possibly be attributed to a compensatory respiratory response to inadequate peripheral vasoconstriction (39). On the other hand, in our motion sick subjects, the corresponding hypocapnia cannot entirely reflect such compensation because it sometimes occurred before (or even outside the context of) any overt failure of systemic vasoconstriction (e.g., Fig. 4). One possibility is that, when nausea progresses during the development of motion sickness, it simply induces anxiety and, therefore, acute hyperventilation. However, inasmuch as motion sickness requires the presence of a functioning vestibular apparatus for induction (31), acute hyperventilation (and resultant cerebral vasoconstriction) in motion-sick subjects could also be partially driven by a change in supine-to-upright vestibulorespiratory regulation (59), i.e., by a brainstem-mediated (as opposed to a cerebral hemisphere-mediated) phenomenon.

The changes in end-tidal CO2 in our overall group postflight (Table 2) support a second recent finding of Novak et al. (39); namely, that individuals who are prone to orthostatic intolerance only develop significant hypocapnia (relative to tolerant subjects) after assumption of the upright position. As in Novak et al.'s patients, exaggerated upright hypocapnia in this study (i.e., postflight) was also not due to any significant increase in the average upright respiratory rate. However, it may not necessarily follow that exaggerated upright hypocapnia, when it occurs without a change in respiratory rate, is strictly attributable to an increase in the average upright tidal volume. Healthy subjects, for example, often have significantly decreased end-tidal CO2 after movement to the upright position but little or no simultaneous change in respiratory rate, tidal volume, or alveolar minute ventilation (4, 21, 52). In addition, Serrador et al. (52) recently suggested that decreased end-tidal CO2 in healthy subjects in the upright position may reflect redistribution of blood and tissue CO2 stores rather than changes in minute ventilation, alveolar ventilation, dead space (42), CO (42), or CO2 production.

Changes in Autonomic Cardiovascular Function


At least three findings from this study suggest that fluctuations in efferent vagal-cardiac nerve traffic are intrinsically heightened in the minutes after emesis. First, in the supine position postflight compared with preflight, vomiters had increases in the total spectral power of R-R intervals (Table 3). This increase was not likely due to respiratory factors (5, 26) because, as described, we actively controlled respiration rate during the collection of these data, and supine end-tidal CO2 was also not changed. Second, vomiters also had increases in the slope of the carotid-cardiac baroreflex after flight, a change that is believed to reflect increased vagal control over the sinus node (10). This second finding might help to explain why the carotid-cardiac baroreflex slope is most significantly decreased in entire groups of returning astronauts, not on landing day, but 1–2 days after landing (10, 11), when severe motion sickness in some of the crewmembers is presumably no longer a factor. Third, immediately after flight in the seated position in the aircraft, vomiters had temporally prolonged MBP responses during Valsalva phase IIe (Table4). Temporal prolongation of phase IIe also occurs when seated Valsalva maneuvers are performed during parabolic microgravity (48), i.e., when efferent cardiovagal influences on HR are presumably increased (34).

The finding of increased R-R interval variability after emesis does not contradict reports of decreased (9, 17) or unchanged (35) HR variability in previous human studies during the development of motion sickness because, in those studies, changes in HR variability were not studied in the context of actual vomiting. On the other hand, increased R-R interval variability after emesis is consistent with the increased “coefficient of variance of R-R intervals” observed in squirrel monkeys taken all the way to vomiting during a visual-vestibular stimulus (18). It may be, therefore, that the directional change in the variability of R-R intervals during motion sickness depends, in part, on the degree of motion sickness attained, with prodromal symptoms (including moderate nausea) associated with decreased R-R interval variability [or with unchanged R-R interval variability when nausea-related respiratory alterations (2) are experimentally negated (35)] and the actual emetic and postemetic periods associated with increased R-R interval variability. Taken together, these findings suggest that if a cardiac “stress response” independent of respiratory changes occurs during the development of motion sickness (17, 32) (it may not, Ref.35), it is nonetheless superseded by increased fluctuations in vagal-cardiac nerve traffic during and/or after emesis itself, a situation that might be roughly paralleled during tilt testing when the acute development of vasovagal presyncope is accompanied by increases in both the respiratory and nonrespiratory fluctuations of R-R intervals (37).


The relative increases in the low-frequency power of SBP and DBP and in the sympathovagal index in nonvomiters postflight (Table 3) were especially evident in the three nonvomiters who developed orthostatic intolerance (Fig. 7). The supine autonomic changes in these three individuals were, therefore, again, the most reminiscent of clinical POTS (38). However, unlike patients with clinical POTS, who have attenuated MBP responses during supine Valsalva phase IIl (46, 49, 50), the subjects with POTS-like intolerance in this study had, like astronauts returning from spaceflight (11), accentuated MBP responses during this same Valsalva phase (Fig. 7). Two factors might explain this apparent discrepancy. First, clinical POTS is a heterogeneous disorder, and many patients with POTS do in fact have accentuated MBP responses during supine Valsalva phase IIl (P. A. Low, personal communication). Second, NASA investigators, including ourselves, typically calculate the magnitude of Valsalva phase IIl by using the delta BP between the trough value in phase IIeand the peak value in phase IIl (11, 12, 48). On the other hand, until recently, most clinicians studying POTS have calculated the magnitude of phase IIl as the absolute or percent offset of the BP vs. the baseline BP obtained before the beginning of the maneuver (28, 46, 49, 50). It should be noted that, in situations in which both phase IIe and phase IIl are determined to be accentuated by using the NASA method, the use of the earlier clinical method may determine that phase IIl is actually attenuated, depending on the absolute increment in BP during phase I, the absolute decrement in BP during phase IIe, and the absolute increment in BP during phase IIl. Therefore, when cross-referencing the results of studies employing the Valsalva maneuver, these differing historical methods of analyses should always be kept in mind.


One limitation to this study was that we did not continuously monitor subject core temperatures. However, with respect to a potential effect of subject heat stress on our results, it should be noted that the KC-135 is air-conditioned similarly to commercial aircraft and also that the “shirt sleeves” environment of parabolic flight is such that heat stress is rarely, if ever, the major concern that it is during/after heavily-suited space flight. Another limitation was that we did not directly measure plasma volume and hormones such as catecholamines, renin-angiotensin-aldosterone, and arginine vasopressin (AVP). Regarding this limitation, however, it is interesting to note that our 16 subjects as a group actually had increased supine SV postflight compared with preflight, with no difference between vomiters and nonvomiters with respect to the average postflight increment in this parameter (Table 2). These increases in supine SV postflight provide circumstantial evidence that many of our subjects, including vomiters, may have experienced intravascular volume expansion rather than contraction after flight, perhaps because the majority of parabolic flight time is actually spent in +Gz, which would be expected to expand plasma volume, not contract it (41). In vomiters, the additional volume-expanding effects of a neurohormone such as AVP (20), which can be increased up to 60-fold after emesis during parabolic flight (23), must also be considered, along with AVP's potential baroreflex-enhancing effects (16, 55). A possible corroboration of expanded, rather than contracted, intravascular volume in vomiters might also be found in the attenuated MBP responses that they developed postflight during seated Valsalva phases IIe and IIl (Table 4). In healthy subjects, when plasma volume is expanded by saline administration, the typical result is attenuated BP responses during these same Valsalva phases (12). On the other hand, neither volume expansion nor volume contraction alone appears to influence carotid-cardiac baroreflex responsiveness as measured by neck-cuff devices (54).


We studied orthostatic tolerance and autonomic cardiovascular function in 16 healthy test subjects before and after a seated 2-h parabolic flight. After flight, the incidence of orthostatic intolerance during a 30-min tilt test increased fourfold, with new intolerance being associated especially with decreased upright TPR responses relative to preflight. Of the six newly intolerant subjects, three had vomited as a result of parabolic flight (newly intolerant vomiters) whereas three had not (newly intolerant nonvomiters). After flight, the newly intolerant nonvomiters (none of whom was significantly motion sick) developed a form of orthostatic intolerance resembling clinical POTS. This form of intolerance was characterized by an exaggerated sympathovagal index in the supine position, an exaggerated hypocapnia and cerebral vasoconstriction in the upright position, postural tachycardia, a gradual failure of the upright TPR response, and either absolute hypo- or hypertension. On the other hand, the newly intolerant vomiters developed a form of orthostatic intolerance that was not characterized by a clear hypotensive or hypertensive event but rather by a more isolated hypocapnia and cerebral vasoconstriction during nausea and/or acute lightheadedness in the upright position. During controlled breathing in the supine position postflight compared with preflight, vomiters also had autonomic changes suggestive of increased fluctuations in efferent vagal-cardiac nerve traffic. The most important conclusion from this study is that syndromes of orthostatic intolerance resembling those occurring after space flight can occur after a brief (i.e., 2-h) parabolic flight.


We thank the test subjects who enthusiastically volunteered for this challenging study; Don Barker for assistance with data collection; Dr. Mike Kassam, Jorge Serrador, Paul Dunphy, and Paul Picot for assistance with the transcranial Doppler and CO2 data; Ann Sanders for general assistance; Drs. Alan Feiveson and Dick Calkins for statistical assistance; Lynette Bryan, Noel Skinner, Linda Billica, and Bob Williams for technical assistance inside the aircraft; Dr. Suzanne Schneider for review of the manuscript; the Johnson Space Center Neurosciences and Cardiovascular Laboratories for help with physiological measurements; and the Johnson Space Center Human Test Subject Facility for the recruitment and care of test subjects.


  • This research was supported by NASA Grants 199161156 and 199161157 and a NASA Life Sciences Young Investigator Award.

  • Address for reprint requests and other correspondence: T. T. Schlegel, NASA Johnson Space Center, Mail Code SD3, Houston, TX 77058 (E-mail: tschlege{at}

  • 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.


View Abstract