Journal of Applied Physiology


Early evidence from long-duration flights indicates general cardiovascular deconditioning, including reduced arterial baroreflex gain. The current study investigated the spontaneous baroreflex and markers of cardiovascular control in six male astronauts living for 2–6 mo on the International Space Station. Measurements were made from the finger arterial pressure waves during spontaneous breathing (SB) in the supine posture pre- and postflight and during SB and paced breathing (PB, 0.1 Hz) in a seated posture pre- and postflight, as well as early and late in the missions. There were no changes in preflight measurements of heart rate (HR), blood pressure (BP), or spontaneous baroreflex compared with in-flight measurements. There were, however, increases in the estimate of left ventricular ejection time index and a late in-flight increase in cardiac output (CO). The high-frequency component of RR interval spectral power, arterial pulse pressure, and stroke volume were reduced in-flight. Postflight there was a small increase compared with preflight in HR (60.0 ± 9.4 vs. 54.9 ± 9.6 beats/min in the seated posture, P < 0.05) and CO (5.6 ± 0.8 vs. 5.0 ± 1.0 l/min, P < 0.01). Arterial baroreflex response slope was not changed during spaceflight, while a 34% reduction from preflight in baroreflex slope during postflight PB was significant (7.1 ± 2.4 vs. 13.4 ± 6.8 ms/mmHg), but a smaller average reduction (25%) during SB (8.0 ± 2.1 vs. 13.6 ± 7.4 ms/mmHg) was not significant. Overall, these data show no change in markers of cardiovascular stability during long-duration spaceflight and only relatively small changes postflight at rest in the seated position. The current program routine of countermeasures on the International Space Station provided sufficient stimulus to maintain cardiovascular stability under resting conditions during long-duration spaceflight.

  • baroreflex
  • heart rate variability
  • left ventricular ejection time
  • cardiac output

astronauts participating in spaceflights of <2-wk duration might have an initial increase in vagally mediated carotid baroreceptor (arterial) baroreflex responses during spaceflight (7), but this is followed by reductions in-flight (7, 10) and on return to Earth (12, 13, 16, 43). This change in baroreflex response might contribute to poor cardiovascular responses to upright posture after spaceflight (12). Overall, the cardiovascular system adapts rapidly to the microgravity environment. In addition to observations of arterial baroreflex, there are many reports of generally small increases or no changes in heart rate (HR) for short-duration missions (7, 10, 31), but there are reports of reduced HR during in-flight experimental sessions (29) and during 24-h recordings (14). There are fewer reports of arterial blood pressure (BP) during short-duration spaceflight. One study reports a transient increase in systolic BP (SBP) on day 1 of spaceflight followed by a reduction to preflight at rest in the seated upright position (7). Other studies indicate no change in mean arterial pressure (MAP) (31) or a reduction in diastolic BP (DBP) with no change in SBP (14). Stroke volume (SV) and cardiac output (CO) were found to increase with short-duration flights (31, 37), suggesting that systemic vascular resistance (SVR) was reduced to maintain MAP.

There is little information on the cardiovascular responses to long-duration spaceflight, especially in the era of the International Space Station (ISS). Early data from the Russian Mir space station suggest that the overall cardiovascular system adapted well to the microgravity environment, with minimal changes in HR and MAP (6, 20). However, limited observations suggest that the arterial baroreflex response was reduced in-flight (6). Two recent publications provide the first information on cardiovascular responses on the ISS. Baevsky et al. (2) observed significant reductions in SBP and DBP from preflight to month 5 on the ISS but found that resting HR was not different. Verheyden et al. (44) reported no changes in BP or HR in-flight. These latter investigators also examined the baroreflex response and observed, in contrast to several investigations of short-duration spaceflight (10, 12, 13, 43), no change in baroreflex response slope during 6 mo in-flight. With the exception of the marked reduction in baroreflex response slope following 9 mo on Mir (6), there are no reports of the baroreflex on return to Earth after long-duration spaceflight.

The current study investigated cardiovascular regulation before, during, and after long-duration spaceflights. Finger arterial BP was measured during spontaneous and paced breathing to examine the arterial baroreflex, and the pulse waveform was further analyzed to derive indicators of cardiac function and arterial vascular responses. On the basis of the available literature when the study was initiated in 2001 (6, 13, 28), it was hypothesized that the arterial baroreflex would be reduced during and after long-duration spaceflight. Furthermore, it was hypothesized that postflight HR would be elevated and that indicators of vascular function derived from the finger pulse pressure (PP) waveform would reflect reductions in SV and impaired vasoconstrictor responses.



Six male astronauts (41–55 yr old, 175.7 ± 5.0 cm height, 81.0 ± 9.6 kg body wt) volunteered to participate in the study after receiving full verbal and written details of the experiment. Each astronaut signed a consent form approved by the Office of Research Ethics at the University of the Waterloo and the Committee for the Protection of Human Subjects at Johnson Space Center. The experiment conformed to the guidelines in the Declaration of Helsinki.

Three of the astronauts launched and landed on the space shuttle. The durations of their missions to the ISS were 153, 120, and 52 days. The other three astronauts travelled to and from the ISS on the Russian Soyuz. Their mission durations were 175, 199, and 180 days. Preflight tests were conducted ∼30 days before flight; in-flight tests were conducted ∼2–3 wk after launch (In-flight-1), which was close to the time of data collection in some previous short-duration flights, and ∼2–3 wk before returning to Earth (In-flight-2); postflight tests were conducted on the day after landing (R+1) for four of the six astronauts, on R+2 for one, and on R+3 for the other, according to their schedule for return to the test site after landing. All testing was completed between May 2007 and December 2009.

Experimental Protocol

The three shuttle-launched astronauts participated in preflight testing at Johnson Space Center (Houston, TX). Two of the three completed postflight testing at Kennedy Space Center (Cape Canaveral, FL) on R+1 and the other on R+3 at Johnson Space Center after landing at the Dryden facility in California. Two of three Soyuz-launched astronauts completed pre- and postflight testing at the Gagarin Cosmonaut Training Center (Star City, Russia), the other completed preflight testing in Houston, and one astronaut completed postflight testing on R+2, rather than R+1, because of weather constraints. All astronauts used a fluid-loading protocol immediately prior to return to Earth; one astronaut experienced nausea on landing day. The slight delay in testing two astronauts could have influenced the results, but previous investigations showed that autonomic changes persist for several days after short-duration spaceflights (13), and we noted no differences.

Pre- and postflight data were collected in the supine posture during a 5-min period of spontaneous breathing, as well as during spontaneous and paced breathing in the seated posture. The seated posture reflects nonstressed upright rest on Earth (31). Recent investigations of astronauts from the ISS used semisupine (2), as well as supine and standing, reference postures (44). In the seated posture, 5 min of spontaneous breathing was followed by 5 min of paced breathing, with 5 s of inspiration followed by 5 s of expiration guided by watching the second hand of a clock. In-flight measurements were identical for each astronaut and required the astronaut to set up the BP device and then complete the spontaneous and paced breathing protocols according to procedures and computer guidance.


Ground-based studies for all shuttle astronauts incorporated the finger arterial pressure measurement (Finometer, Finapres Medical, Amsterdam, The Netherlands), with data sampled at 1,000 Hz and stored on a personal computer via an analog-to-digital conversion (PowerLab, ADInstruments, Castle Hill, NSW, Australia). Ground-based studies in Russia used a different finger BP device (Finapres, Finapres Medical) with the same instrumentation for data collection at 1,000 Hz. In-flight measurements were made with the National Aeronautics and Space Administration continuous BP device, which was part of the Human Research Facility on the ISS. The in-flight raw data were sampled at 200 Hz and downlinked for analysis. For all ground and in-flight tests, the BP cuff was allowed to stabilize and complete its self-calibration process before the Physiocal was disabled to permit uninterrupted 5-min data collection periods.

Data Analysis

All data were analyzed with an offline version of algorithms in the Nexfin monitor (BMEYE, Amsterdam, The Netherlands) (5). The data from ground collections were downsampled to 200 Hz to match the in-flight sampling rate; then the data were processed for each beat to obtain an estimate of RR interval, SBP, DBP, and MAP, PP, SV, CO, SVR, rate of rise of arterial pressure (dP/dt), and left ventricular ejection time (LVET), which was corrected for changes in HR to determine LVET index (LVETi = 1.7·HR + LVET) (27).

For each variable, a beat-by-beat time series was constructed from a ≥4-min table portion within the 5-min data collection window for spontaneous or paced breathing. Mean values of each variable were calculated from all beats in this window. Spectral analyses were performed on the unevenly spaced time series as follows. Data were resampled at 4 Hz by cubic spline interpolation. Any linear trends were removed, and the power spectra were obtained with a fast Fourier transform-based approach (Welch's periodogram: 256 points, 50% overlap with preceding segment, Hanning window). Two frequency bands were set: 0.05–0.15 Hz [low frequency (LF)] and 0.15–0.5 Hz [high frequency (HF)], and auto-spectral power densities for each were calculated for all variables. The HF lower-limit cutoff was modified in three individual spontaneous breathing tests, because visual inspection confirmed that the respiratory peak overlapped the 0.15-Hz boundary. In these three tests (1 from In-flight-1 and 2 from In-flight-2, all different individuals), the boundaries were set at >0.125 Hz for two tests and at >0.117 Hz for the third test. During paced breathing, only LF spectral powers were computed, as the breathing pattern forced variability into this frequency range. The dynamic interrelationship defining estimates of the arterial baroreflex responses between SBP and the RR interval was determined from the cross-spectra within the LF and HF bands for spontaneous breathing and in the LF band only for paced breathing. Only values for which coherence was ≥0.5 were included in computation of transfer function gain and phase.

Spontaneous baroreflex slope was determined by the beat sequence method, which has been shown to be highly correlated with gain computed by the cross-spectral method and with the baroreflex response slope obtained during administration of drugs to raise or lower BP (24, 33). Sequences of three or more beats in which the RR interval and SBP increased or decreased and the absolute change in SBP between beats was >0.5 mmHg were included. The beat lag for matching RR interval and SBP to achieve the highest correlation was conducted as described previously (3). Regression slopes were determined for each beat sequence, and the mean value was computed for each subject.


The main analysis examined the potential differences within the primary baroreflex study for the main effect of test times (seated preflight, In-flight-1, In-flight-2, and seated postflight) with a one-way repeated-measures ANOVA followed by Student-Newman-Keuls post hoc test. Spontaneous and paced breathing were examined separately. Given the specific hypothesis that postflight would differ from preflight cardiovascular variables, paired t-tests were used to compare the preflight seated values with the postflight seated values during spontaneous and paced breathing. Specific planned comparisons were made by paired t-tests for the supine tests to determine if they differed from seated data at the corresponding time period (pre- and postflight).

Statistical analyses were performed with SigmaPlot 11 software. Significance was set at P < 0.05, but given the nature of the study with a limited number of astronauts, P < 0.1 is reported to reveal possible trends. All data are presented as means ± SD.


Cardiovascular Variables

Pre- and postflight supine and seated.

During the spontaneous breathing component of the protocol preflight, the RR interval was longer in the supine than seated position, although the absolute difference was small (Tables 1 and 2). HR tended to be lower and the standard deviation (SD) of the RR interval tended to be larger in the supine than seated position. SV was greater in the supine position, and CO tended to be higher. The estimate of LVETi from the finger pulse wave was significantly longer in the supine than seated posture. Arterial baroreflex slope during spontaneous breathing was 14.8 ± 6.5 and 13.6 ± 7.4 ms/mmHg in supine and seated positions, respectively; this difference was not significant (Fig. 1). During paced breathing, the slope was 13.4 ± 6.8 ms/mmHg.

View this table:
Table 1.

Cardiovascular variables for seated postures pre- and postflight compared with in-flight

View this table:
Table 2.

Cardiovascular variables measured during spontaneous breathing in supine posture with statistical comparisons to spontaneous breathing in seated posture during the baroreflex protocol pre- and postflight

Fig. 1.

A: baroreflex slope obtained during spontaneous breathing (black bars) and paced breathing at 6 breaths/min (gray bars). In the supine posture, only spontaneous breathing sessions were collected. Preflight (∼30 days before launch) and postflight (within 1–3 days of landing) data were collected in the seated position. In-flight-1, ∼14–21 days in-flight; In-flight-2, ∼14–21 days before the end of the flight. NS, not significantly different between supine and seated postures (by paired t-test). Values are means ± SD. *Significantly reduced compared with each other paced breathing measurement, P < 0.05. B: individual subject values at each time point during spontaneous breathing tests only. Open symbols, shuttle landing; filled symbols, Soyuz landing.

After spaceflight, there was a trend for HR to be lower in the supine than seated position, and again estimated LVETi was longer in the supine than seated position (Tables 1 and 2). Arterial baroreflex slopes, 10.6 ± 4.5 and 8.0 ± 2.1 ms/mmHg while supine and seated, respectively, were not different from each other, but the postflight supine value was less than the preflight supine value (Fig. 1; P < 0.05).

On-ground and in-flight.

From preflight to in-flight, there were no differences in any cardiovascular variable measured during spontaneous paced breathing, with the exceptions of the estimates of LVETi, which were increased in-flight (Table 1), and SV (Fig. 2) and CO, which was significantly increased only In-flight-2, and SVR, which was decreased only In-flight-2 during paced breathing (Table 1). Preflight arterial baroreflex slope was not different from In-flight-1 (11.0 ± 3.7 and 12.3 ± 5.3 ms/mmHg during spontaneous and paced breathing, respectively, respectively) or In-flight-2 (11.8 ± 5.3 and 14.4 ± 6.4 ms/mmHg, respectively; Fig. 1).

Fig. 2.

Relationship between estimated left ventricular ejection time index (LVETi) and estimated stroke volume (SV) values obtained from the finger arterial pressure waveform for all test conditions (supine, seated, and paced and spontaneous breathing) on the ground (open symbols) and in microgravity (paced and spontaneous breathing, filled symbols). Each astronaut is represented by a different symbol. Pooled regression lines for on-ground (dashed line, r2 = 0.35) and in-flight (solid line, r2 = 0.32) are shown.

Postflight measurements revealed several differences with respect to preflight or in-flight (Table 1). The postflight estimate of LVETi was reduced compared with the two in-flight measurements. With the repeated-measures ANOVA, only CO was increased postflight compared with preflight during spontaneous and paced breathing, and total SVR was significantly lower postflight during paced breathing (Table 1). Arterial baroreflex slope tended to be reduced (P = 0.058) postflight compared with preflight (8.0 ± 2.1 ms/mmHg, 75.1 ± 40.1% of preflight) during spontaneous breathing and was significantly reduced during paced breathing postflight (7.1 ± 2.4 ms/mmHg, 66.1 ± 33.1% of preflight) compared with preflight, In-flight-1, and In-flight-2 (Fig. 1).

Postflight comparisons were also made directly with preflight by paired t-tests. HR was ∼5–6 beats/min higher postflight, the RR interval tended to be smaller, and the SD of the RR interval was reduced postflight (Table 1).

Spectral Analysis

Pre- and postflight supine and seated.

In the preflight comparisons of RR interval spectral power in the seated (Table 3) and supine (Table 4) postures, only HF and total spectral powers tended to be different (P < 0.1). There were no significant differences in any of the other spectral power indicators between supine and seated preflight measurements.

View this table:
Table 3.

Comparison of natural logarithm-transformed spectral powers during the baroreflex protocol in the seated posture pre- and postflight with in-flight during spontaneous and paced breathing

View this table:
Table 4.

Natural logarithm-transformed values of spectral powers measured during spontaneous breathing in the supine posture with statistical comparisons to spontaneous breathing in the seated posture during the 10-min baroreflex protocol pre- and postflight

In the postflight comparisons of supine and seated postures, HF and total RR interval spectral powers were greater in the supine posture. In the postflight comparisons, the HF spectral power of SBP and the HF spectral power of PP were significantly greater in the supine than seated posture.

On-ground and in-flight.

HF spectral power for the RR interval and for arterial PP was significantly reduced in-flight compared with preflight in the seated posture during spontaneous breathing (Table 3). The spontaneous breathing frequency determined from the peak frequency in the HF band shifted to lower frequencies in-flight but was not significantly changed from preflight (0.237 ± 0.037 Hz) to In-flight-1 (0.193 ± 0.026 Hz) and In-flight-2 (0.188 ± 0.054 Hz). During paced breathing, PP LF power was less than preflight.

LF, HF, and total RR interval spectral power were reduced postflight during spontaneous breathing compared with preflight (Table 3), but breathing frequency was unchanged postflight (0.223 ± 0.032 Hz). LF and total RR interval power were also less postflight than In-flight-1 and In-flight-2, while HF RR interval power was less In-flight-2 (Table 3). During spontaneous breathing, the HF power of arterial PP postflight was greater than in-flight but was not different from preflight, and SV HF power was greater than In-flight-1. Under the paced breathing conditions of postflight testing, there were no differences in spectral power compared with in-flight or preflight (Table 3).

Changes in Spontaneous Baroreflex Slope

The data for each individual astronaut relating the slope of the spontaneous baroreflex with the RR interval recorded at that specific time point are shown in Fig. 3. In general, the RR interval decreased the supine to the seated position preflight and to postflight, and simultaneously the slope of the spontaneous baroreflex was reduced. Three of the six subjects portrayed in Fig. 3 had significant correlation coefficients for the reduction in baroreflex slope relative to the reduction in RR interval.

Fig. 3.

All 6 data points obtained during spontaneous breathing components of the test (supine and seated pre- and postflight plus the 2 in-flight sessions) for the relationship between baroreflex slope and mean RR interval during that data collection. Correlation coefficients and corresponding P values are shown for each astronaut. Open symbols, shuttle landing; filled symbols, Soyuz landing.

The estimates of the baroreflex gain calculated from the spontaneous beat sequence method and from the cross-correlation between systolic arterial pressure and RR interval were highly correlated during spontaneous breathing in the HF (r = 0.87) and LF (r = 0.78) ranges and during paced breathing in the LF range (r = 0.97). Furthermore, the mean values obtained by the two methods did not differ, so only data from the sequence analysis are included in Figs. 1 and 3.


This study provides the first data on cardiovascular stability and control from the arterial baroreflex response slope and indicators of cardiac function derived from the finger pulse wave during and after long-duration missions to the ISS. In-flight cardiovascular function was similar to that observed previously in short-duration flight. There was only modest evidence of cardiovascular deconditioning postflight, suggesting that the in-flight countermeasures were reasonably effective in these astronauts.

Reduced baroreflex response slopes during the paced breathing phase of postflight testing were consistent with the hypothesis; however, the magnitude of change was less than anticipated. Similarly, the 5–6 beat/min increase in HR during the postflight testing was significant but of very small magnitude. Despite the small change in HR postflight, there was a marked reduction in LF, HF, and total RR interval spectral power during spontaneous breathing compared with preflight and in-flight. Previous data from short-duration spaceflights (7, 10, 12, 13, 16, 43) combined with early research of long-duration astronauts on the Russian Mir station (6, 28) provided the rationale for expecting a marked change in arterial baroreflex of astronauts living up to 6 mo on the ISS. Our in-flight data showing essentially unchanged baroreflex responses contrast with these early observations, but they are in line with the recent observations of Verheyden et al. (44) during 6-mo sojourns to the ISS. Our data, which are the first reported after long-duration flight, reflect considerably less cardiovascular deconditioning than in two cosmonauts who spent 9 mo on Mir (6). For some astronauts, the individual changes in baroreflex response slope were correlated with changes in RR interval, suggesting a mechanism related to the overall reduction in parasympathetic activity to the heart.

The current study also provided new insight into cardiovascular stability from analysis of the finger pulse waveform. In contrast to the hypothesis, SV was maintained during and after spaceflight when referenced to the preflight upright seated position. The LVETi was longer in-flight without a change in SV, in contrast with simultaneous changes on Earth (38). A consequence of the late spaceflight and postflight time points of maintained or slightly increased SV, along with small increases in HR, was that the estimate of CO was significantly elevated compared with preflight baseline, while SVR was reduced to maintain MAP. The reductions in HF RR interval power in-flight and postflight in HF, LF, and total RR interval power might reflect reduced parasympathetic modulatory effects on HR (32, 39). However, when the in-flight results are viewed along with significant reductions in HF power of PP and SV, it might suggest that the within-breath modulatory influence of the respiratory pump on venous return and cardiac filling was reduced in-flight, even though overall SV and MAP were maintained.

Long-Duration Spaceflight

The ISS astronauts in the current study represent the first six-person crew, signifying the transition to greater possibilities to conduct science on this major international laboratory. Opportunities for investigations of physiological responses to spaceflight have been limited by the demands placed on astronauts for construction activities and by restricted access to space following the tragedy of the Columbia accident on 1 February 2003. Time available for research in-flight and immediately on return to Earth is limited, and problems related to the small sample size have been considered previously (34, 44). In the current study, the primary investigation was designed around 5-min periods of spontaneous and paced breathing. This consistent procedure was employed in the seated posture pre- and postflight, as well as In-flight-1 and In-flight-2. No position on Earth, even supine, replicates exactly the conditions of microgravity, and uncertainty exists over the best reference point (30). In this study, the seated posture on Earth was compared with the supine position, and there were few differences, probably reflecting the fact that the relaxed seated posture that we used was a relatively minor orthostatic challenge. Additional factors that can affect science on the ISS include the work cycle and daily routines, including diet, exercise, sleep patterns, and sleep-shifting related to activities including arrival of other crew and supplies. Constraints were placed on all astronauts to avoid sleep shifts, to restrict food and caffeine intake during the previous 2 h, and to avoid heavy exercise (heart rate >80% maximum) on the day before testing or any exercise on the day of testing.

Data collection within the first 2–3 wk in space and the last 2–3 wk of flight allowed astronauts to accommodate to the routine of spaceflight and avoided any space motion sickness that might have occurred in the first few days of flight. While the timing of our first sample might be similar to some shorter-duration flights, the shorter flights are inherently more stressful with tasks to complete before the end of mission. Our postflight measurements were complicated by the landing site, but there were no obvious differences related to this issue (Figs. 1 and 3).

Baroreflex Response Slope

During long-duration spaceflight, there was no change in baroreflex response slope in the spontaneous or paced breathing protocols. These results are consistent with the recent observations by Verheyden et al. (44) of six astronauts who also spent up to 6 mo on the ISS. However, the results contrast with some (7, 10, 12, 13, 16), but not all (44), measurements made during short-duration spaceflights and with limited information from three cosmonauts on previous long-term missions on the Mir station (6). The current data showing no change in baroreflex response slope during up to 6 mo on the ISS confirm the stability of the cardiovascular adaptation to spaceflight under the current conditions on board the ISS. The potential role of exercise countermeasures in maintaining cardiovascular stability is considered in Exercise Countermeasures on the ISS.

The 25–34% reductions in postflight spontaneous baroreflex response slope during spontaneous and paced breathing are the first reports of baroreflex function on return from the ISS. Previous data from three cosmonauts after 9 mo on Mir indicated >50% reductions (6). These data compare with the ∼65% lower baroreflex responses in patients with autonomic failure with respect to age-matched controls (15). The differences between spaceflight studies could have been due to individual differences, the longer Mir missions, or recent upgrades to exercise facilities and time available for exercise on the ISS; however, the Russian countermeasure system might have included lower body negative pressure (26), which could improve the response, although this was not reported (6).

Baroreflex response has been measured frequently after short-duration missions. Fritsch et al. (12) first reported a 32% reduction in baroreflex following 4- to 5-day missions and then reported a 15% reduction after 8- to 14-day missions (13). Gisolf et al. (16) found a breathing frequency-dependent reduction in the supine baroreflex response slope after 10- to 11-day missions to the ISS that was 33% at 6 breaths/min but 0% at 15 breaths/min. After the same-duration missions, Verheyden et al. (43) observed a 40% reduction in the baroreflex response that appeared to be independent of breathing frequency. These latter results are more consistent with our own observations, with average 25–34% reductions during spontaneous and paced breathing.

Large individual differences were observed in the absolute value of the baroreflex slope, as well as in the change of the slope with spaceflight (Figs. 1 and 3). Unfortunately, orthostatic tolerance tests were not conducted on these astronauts to determine if those with the largest change in baroreflex slope were intolerant. The results do emphasize the relatively small change in baroreflex response in some individuals, even after long-duration spaceflight. The individual with the largest change in the baroreflex response slope also had the longest RR interval. While it is unknown if changes in intrinsic HR might contribute to the change in RR interval, reductions in resting parasympathetic activity could account for the faster RR interval and the reductions in the baroreflex response slope (23, 44).

Cardiovascular Responses

The unchanged HR during spaceflight, which is consistent with most results from the ISS (2, 44) and short-duration spaceflights (7, 10, 45), can be taken as an index of cardiovascular stability. In the current study, CO was significantly elevated In-flight-2 during spontaneous and paced breathing as a consequence of small increases in HR and SV. Previous reports of SV and CO during long-duration spaceflights suggest that values were maintained relative to preflight (20), although left ventricular volume might be reduced (1). During shorter flights, CO might be maintained (37), reduced (35), or even increased (31), depending at least in part on the preflight reference point. In our postflight testing, CO was significantly elevated with respect to preflight values primarily because of an increase in HR. We found no difference between postflight seated and postflight supine CO, again reflecting the relatively low orthostatic stress of our seated posture. Elevation in CO near the end of flight and especially postflight was unexpected when referenced to short-duration spaceflights (37), but astronauts returning from 6 mo on the ISS had small but not significant increases in SV and CO (36). The mechanisms responsible for elevated, rather than reduced, CO after long-duration spaceflights on the ISS remain to be determined. Changes in blood volume could influence CO, but there are no blood volume data pre- vs. postflight for these astronauts. All astronauts followed a fluid-loading protocol (47) prior to reentry, and all astronauts had been on Earth for at least one overnight period, so their blood volumes might have been close to preflight levels. Early research from short-duration flights indicates marked loss of cardiac muscle mass (35), which, if it also occurred with long-duration flights, would be expected to reduce SV and CO. It is possible that exercise countermeasures preserved or enhanced cardiac mass, as seen in bed rest (8).

Estimates of LVETi and dP/dt were derived from analysis of the finger pulse waveform. dP/dt was not different at any time during or after spaceflight, including comparisons between supine and seated postures. On the other hand, LVETi increased during long-duration spaceflight compared with pre- and postflight in the seated posture. Normally, increases in LVETi are associated with increased SV (38), but SV did not differ from preflight in the seated position. Di Rienzo et al. (7) reported longer LVET (although it was not corrected for HR) during up to 6–14 days of spaceflight; however, they did not measure SV. Thus we know of no data to confirm or contradict our observations of longer LVETi in-flight without a change in SV. Collectively, these data suggest that there might be alterations in cardiac contractility during spaceflight, as shown by the relationship between LVETi and SV (Fig. 2); however, available echocardiography data do not support this notion (1, 20). Nonetheless, ground-based observations of a longer preejection period in women participating in 60 days of head-down bed rest (22) and slower cardiac relaxation in men after bed rest (9) provide rationale for future investigations.

Spectral Power Indicators

The reduction in HF spectral power for the RR interval in-flight compared with preflight in the seated position contrasts with recent data obtained on the ISS, where RR interval variability was unchanged relative to supine (44) and semirecumbent (2) positions. However, our finding of no change in LF and total powers was consistent with these other studies. The cause of these differences is unknown. It is unlikely that breathing frequency played a role, as the small reduction in the current study might have increased HF power, especially as we reduced the lower bound for the HF band in three cases to avoid losing some respiratory frequency-associated power into the LF band. Lower HF power is often attributed to reduced parasympathetic activity (32, 39), but in the current case, a strong case for this link cannot be made, as the RR interval and the SD of RR interval were not reduced. Each of the LF, HF, and total spectral power indicators was reduced postflight relative to preflight, a finding in general agreement with other studies (2, 7, 17, 29, 44). Postflight changes in spectral power were associated with indicators from changes in RR interval, SD of RR interval, and HR that parasympathetic activity might have been reduced compared with preflight. There were no significant changes in the LF-to-HF ratio, which is often taken as a marker of sympathetic activity or sympathovagal balance (32), indicating only minor changes in overall autonomic balance during or after spaceflight.

Another factor that could contribute to the magnitude of RR interval variability is the variation in arterial pressure. The HF PP spectral power during spontaneous breathing and the LF PP spectral power during paced breathing in-flight were reduced relative to pre- and postflight. This was associated during spontaneous breathing with smaller HF SV power at least for In-flight-1. We speculate that the reduction in HF RR interval power was related to the reduced magnitude of HF PP without a change in coupling, as indicated by the unchanged baroreflex slope. It is possible that removal of the impact of gravity on blood flow through the lungs caused smaller variations in venous return and cardiac filling during the respiratory cycle (42).

Exercise Countermeasures on the ISS

The potential negative physiological consequences related to the reduced physical demands of living in the microgravity environment have been well recognized (4, 20, 45). Exercise countermeasures have been developed during the Russian long-duration spaceflights (20, 26) and during the American space program (19) with the intent of bringing the astronauts back to Earth in good health and capable of meeting any demands associated with emergencies that could arise on landing. However, many studies during the early phases of long-duration spaceflight reveal poor orthostatic tolerance (28), considerable loss of bone mass (40), and impaired muscle function (11, 41). Activities performed by the astronauts in recent investigations of exercise performance and muscle function (11, 41) were similar to those performed by the astronauts in the current study. Our subjects were allocated up to 2.5 h/day during which they could set up for the exercise, complete their workout, and clean up after the session. On average, over the entire missions, each astronaut in this study completed 1.3 ± 0.3 sessions of exercise per day. The types of activity on the cycle and treadmill and resistive exercise varied considerably between individuals but were similar to those described by Trappe et al. (41).

The impacts of these in-flight exercise-training sessions have not been evaluated systematically with respect to the postflight cardiovascular stability. The current study provides some insight into this issue, but the challenge of seated upright rest was not severe and should not be taken as evidence for or against cardiovascular tolerance of a tilt table test or the effects of lower body negative pressure. Bed rest simulations of spaceflight have generally not shown benefits of exercise on resting HR, RR interval variability, blood pressure regulation including the baroreflex response, or orthostatic tolerance (18, 23, 25), unless the exercise is combined with a form of artificial gravity induced by lower body negative pressure (21, 46). Yet the results for resting HR and arterial baroreflex in the current investigation and the recent studies of astronauts on the ISS (2, 44) suggest that in-flight cardiovascular control was well maintained by the daily exercise regimen, and very limited observations for postflight responses suggest that stability has been maintained, at least under conditions of minimal stress.


The primary limitation of this study was the small sample size, which is characteristic of the limited opportunity to conduct research on astronauts. However, this study would have had sufficient statistical power to detect a difference in baroreflex slope if the magnitude of change reported in early investigations of spaceflight (6) (i.e., available when the study protocol was approved in 2001) had been replicated. We did have sufficient statistical power to detect changes in spectral power indicators in-flight and postflight, but the interpretation of these changes remains speculative. No data were available for the orthostatic tolerance of these astronauts that could have been compared with the changes in baroreflex response; the tolerance test was deleted as a standard medical operations requirement. The study was confined to male astronauts, so the results might not be applicable to women, who have greater orthostatic intolerance after spaceflight (47). Measurements of cardiovascular function were obtained by analysis of the blood pressure waveform, with the assumption that the changes in vascular properties that could occur with spaceflight had no impact on the algorithms used to derive arterial pressure, SV, and other indicators. Future studies could incorporate alternative, independent methods to compare estimates of SV on Earth with those in space to verify the application of the algorithm. Time with the astronauts constrained the experimental design. In the supine position, we were able to collect data only during spontaneous breathing; in the seated posture, we were able to measure cardiovascular responses at two different respiratory frequencies, i.e., during spontaneous breathing and at 6 breaths/min. The slow 6 breaths/min paced breathing frequency has been used in other investigations and permitted comparisons. Spontaneous respiration is subject to large between-individual differences, and changes in frequency can occur under various conditions. An experimental design that includes HF and LF paced breathing with regulation of arterial Pco2, in addition to spontaneous breathing, might provide the best experimental design but would require a greater time commitment.


The current results show that resting RR interval and baroreflex responses were well maintained in astronauts living up to 6 mo on the ISS. The changes that we observed in-flight were the prolongation of LVETi, an increase in estimated CO late in-flight, and reductions in HF RR interval and PP spectral powers, none of which is an indication of cardiovascular deconditioning. The change in LVETi suggests that cardiac contractile function might be altered during spaceflight, but if so, it recovered quickly after landing. Very small increases in resting HR and a 25–34% reduction in the arterial baroreflex response postflight reflected relatively modest levels of cardiovascular deconditioning. These postflight changes were somewhat less than expected based on short-duration flights and early reports of long-duration missions and suggest that the current countermeasures on the ISS, which include exercise training, are keeping cardiovascular control mechanisms well prepared for return to Earth, but this should be confirmed with specific orthostatic tolerance testing.


This research was supported by CSA Grant 9F007-02-0213. P. P. P. P. Pereira-Junior is a recipient of fellowships from the Canadian Bureau for International Education and the Brazilian Council for Scientific and Technological Development. D. Xu is a recipient of a fellowship from the Ontario Ministry of Research and Innovation.


No conflicts of interest, financial or otherwise, are declared by the authors.


We thank the astronauts for their enthusiasm and dedication to the success of the project (Cardiovascular and Cerebrovascular Control on Return From ISS). The assistance and support of personnel at the Canadian Space Agency (CSA) and National Aeronautics and Space Administration were essential. In particular, we thank the support teams at CSA, the experiment support team, and the cardiovascular laboratory at the Johnson Space Center, Kennedy Space Center, Dryden Flight Research Center, and the Gagarin Cosmonaut Training Center.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
View Abstract