Journal of Applied Physiology
Vol. 83, No. 3, pp. 810-816, September 1997
PULMONARY CIRCULATION AND LUNG FLUID BALANCE

Pulmonary tissue volume, cardiac output, and diffusing capacity in sustained microgravity

Sylvia Verbanck¹, Hans Larsson², Dag Linnarsson², G. Kim Prisk³, John B. West³, and Manuel Paiva⁴

¹ Akademisch Ziekenhuis, Vrije Universiteit Brussel, 1090 Brussels, Belgium; ² Karolinska Institute, S-17177 Stockholm, Sweden; ³ Department of Medicine, University of California San Diego, La Jolla, California 92093-0931; ⁴ Biomedical Physics Laboratory, Université Libre de Bruxelles, 1070 Brussels, Belgium

ABSTRACT

INTRODUCTION

MATERIALS AND METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Verbanck, Sylvia, Hans Larsson, Dag Linnarsson, G. Kim Prisk, John B. West, and Manuel Paiva. Pulmonary tissue volume, cardiac output and diffusing capacity in sustained microgravity. J. Appl. Physiol. 83(3): 810-816, 1997.In microgravity (µG) humans have marked changes in body fluids, with a combination of an overall fluid loss and a redistribution of fluids in the cranial direction. We investigated whether interstitial pulmonary edema develops as a result of a headward fluid shift or whether pulmonary tissue fluid volume is reduced as a result of the overall loss of body fluid. We measured pulmonary tissue volume (Vti), capillary blood flow, and diffusing capacity in four subjects before, during, and after 10 days of exposure to µG during spaceflight. Measurements were made by rebreathing a gas mixture containing small amounts of acetylene, carbon monoxide, and argon. Measurements made early in flight in two subjects showed no change in Vti despite large increases in stroke volume (40%) and diffusing capacity (13%) consistent with increased pulmonary capillary blood volume. Late in-flight measurements in four subjects showed a 25% reduction in Vti compared with preflight controls (P < 0.001). There was a concomittant reduction in stroke volume, to the extent that it was no longer significantly different from preflight control. Diffusing capacity remained elevated (11%; P < 0.05) late in flight. These findings suggest that, despite increased pulmonary perfusion and pulmonary capillary blood volume, interstitial pulmonary edema does not result from exposure to µG.

spaceflight; zero gravity; acetylene rebreathing experiments

INTRODUCTION

TWO OF THE BEST DOCUMENTED physiological responses to weightlessness ever since the early manned spaceflights are the cephalad shift of bodily fluids and the net loss of body weight. There are variable results concerning the amplitude of these changes and their dependence on time spent in flight and recovery postflight [see Norsk and Epstein (12) for review]. On the basis of the cephalad fluid shift, together with a postulated increase in central venous pressure in microgravity (µG), predictions for manned spaceflight included a risk for interstitial pulmonary edema (14). The loss of body weight has been associated with loss in bodily fluids, including interstitial fluid and plasma volume, as well as loss of a so-called lean body mass and fat components (10). The contribution of different body compartments to the overall body fluid deficit has been estimated by an electrical impedance method that showed a loss of fluid in lower extremities (4). Simultaneous ultrasonic measurements of soft tissue thickness on the tibia and on the forehead in the same subjects showed similar results (9).

On the one hand, one might expect to see a marked increase in pulmonary tissue volume (Vti) early in spaceflight caused by a headward shift in body fluids (12), by an increase in pulmonary capillary blood volume (15), and by any interstitial pulmonary edema that may have occurred (14). On the other hand, the overall loss of body fluids (12), and the suggestion that lung interstitial tissue has a lower compliance for fluid accumulation than do some other body tissues (e.g., muscle; see Ref. 11), may result in a decrease in Vti.

In the present study, we report the first measurement of Vti in weightlessness, obtained by means of the inert soluble gas-rebreathing method. As well as Vti, which incorporates pulmonary capillary blood volume, the acetylene (C₂H₂)-rebreathing tests performed by four astronauts during the 10-day Spacelab-D2 mission, also allowed computation of cardiac output (c) and stroke volume (SV). The latter parameters have been obtained in space before, both from using nitrous oxide rebreathing tests performed during the SLS-1 mission (15) and from echocardiographic (ECG) measurements during the STS51-D mission (6). From the carbon monoxide (CO)-rebreathing traces, we also calculated CO-diffusing capacity (DL_CO). c, SV, and DL_CO showed marked increases after only ~24 h in µG and trends that are consistent with previous results obtained during the SLS-1 mission covering a comparable period of sustained µG. In contrast, we observed no significant change in Vti after ~24 h in µG and a 25% decrease with respect to ground (1-G) control by the end of the 10-day exposure to µG, with a slow recovery postflight.

MATERIALS AND METHODS

Table  1.   No. of rebreathing tests performed with low-O2 and high-O2 rebreathing gas mixtures Subject No. Preflight In-flight Postflight Early Late R + 2  R + 4  R + 9  Low O2 High O2 Low O2 High O2 Low O2 High O2 Low O2 High O2 Low O2 High O2 Low O2 High O2 1 4 4 2 2 1 1 2 2 2 2 2 2 2 5 5 2 2 1 2 2 2 2 2 2 3 4 4 3 3 2 2 2 2 2 2 4 4 4 1 1 2 2 2 2 2 2 Low O2, 45% O2; High O2, 86.1% O2; see MATERIALS AND METHODS for details. In-flight, early, and late refer to measurements on in-flight mission days 1 or 2 and days 9 or 10, respectively (see MATERIALS AND METHODS). R + 2, R + 4, and R + 9 are postflight sessions 2, 4, and 9 days, respectively, after return to Earth.

RESULTS

Table 2 lists the average preflight values of all measured parameters on the four subjects. All parameters except DL_CO and K_CO, where a change was expected, were not significantly different when computed from the rebreathing tests with low-O₂ or high-O₂ rebreathing gas and are obtained from four low-O₂ and four high-O₂ tests, i.e., eight data points (see Table 1). DL_CO and K_CO values in Table 2 are subject averages of the four preflight low-O₂ rebreathing tests. In general, preflight SE values were of the order of 2-3%, except for Vti which had an average preflight SE of 4.5%.

Table  2.   Preflight data for each subject Subject 1  Subject 2  Subject 3  Subject 4  FRC, ml 3,316 ± 102  2,977 ± 106  3,076 ± 89  4,473 ± 104   c, ml/min 5,818 ± 177  5,571 ± 150  7,111 ± 237  5,903 ± 192  Vti, ml 725 ± 38  461 ± 13  619 ± 29  721 ± 33  DLCO, ml · min1 · mmHg1 STPD 25.4 ± 0.8  27.4 ± 0.4  24.6 ± 0.5  28.4 ± 0.6  KCO, ml · min1 · mmHg1 · l1 BTPS 6.2 ± 0.3  7.1 ± 0.4  6.4 ± 0.2  5.3 ± 0.2  HR, beats/min 76.8 ± 2.1  83.9 ± 2.4  90.4 ± 2.7  76.4 ± 3.0  SV, ml 76.3 ± 0.9  66.6 ± 1.4  75.5 ± 1.4  77.4 ± 1.3 Values are means ± SE from all rebreathing data collected over 2 preflight experiment sessions. FRC, functional residual capacity; Vti, lung tissue volume; c, cardiac output; HR, heart rate; and SV, stroke volume are averages of low-O2 and high-O2 rebreathing tests. DLCO (lung CO-diffusing capacity) and KCO, (DLCO per unit volume) are average values from low-O2 rebreathing tests.

Figures 1-4 show the changes of normalized averages (± SE) of all parameters in Table 2 as a function of time during and after a 10-day period of µG. The preflight data point ( at 100%) results from pooling all data from the two preflight experiment sessions and has an arbitrary time tag for clarity of representation. For in-flight and postflight sessions, intervals on the time axis are representative of the number of days separating the experiment sessions. All  symbols are the average results from the four subjects whenever rebreathing data were collected on all four subjects within 24 h of each other. This was the case for the late in-flight session and during all postflight experiment sessions. All  symbols represent the averages of rebreathing data from subjects 1 and 2 only. For the early in-flight sessions, these are the only data available. For late in flight, these are a subset of the rebreathing data available on all four subjects (Table 1). We display the data in this manner to avoid statistically skewing results from differing subject populations.

FRC (Fig. 1) was reduced in flight compared with preflight, although this was no longer significant in four subjects measured late in flight. Based on data from subjects 1 and 2, FRC did not change in flight between the early and the late data collection sessions. Vti (Fig. 2) was unchanged early in flight compared with preflight (n = 2), but by late in flight Vti showed a significant reduction of ~25% (n = 4; P < 0.001). Postflight, Vti remained depressed through R+9 but showed a trend toward a return to normal. c (Fig. 3A) was elevated early in flight (n = 2) and returned to preflight control by late in flight. This was confounded by significant changes in HR (Fig. 3B), which was depressed early in flight and significantly elevated on R+2. SV (Fig. 3C) provides a clearer picture and shows a large (~40%) increase early in flight, but SV is reduced by late in flight to a level not significantly different from preflight controls. Both DL_CO (Fig. 4A) and K_CO (Fig. 4B) were significantly elevated in both the early and the late in-flight sessions and returned to baseline postflight. There was no apparent temporal change in either DL_CO or K_CO postflight.

DISCUSSION

This paper reports a set of cardiopulmonary parameters obtained during a sustained 10-day period of weightlessness aboard the Spacelab-D2 mission. The major difference between our experiments and previous measurements during spaceflight (15) is that, by using C₂H₂ and CO in the rebreathing gas mixture, we were able to compute Vti. A Vti value of, say, 600 ml measured from soluble gas rebreathing (17, 20) comprises ~100 ml dry tissue, 100 ml capillary blood volume, and 400 ml extravascular fluid (2). As such, its value in weightlessness could be increased by a cephalad fluid shift and/or interstitial pulmonary edema or could be decreased by a net loss of body fluid.

Although one might expect a Vti increase due to the fluid shift from the lower body compartments, this was clearly not the case for the two subjects studied after ~24 h of µG (Fig. 2). A possible explanation is that the loss of body fluid counteracts, and therefore masks, the Vti increase that the cephalad fluid shift might have caused. A summary of data from Apollo and Skylab missions reported in a review by Norsk and Epstein (12) shows that decreases in total body mass start within the first day of spaceflight. In flight, the decreasing Vti trend did not reach significance for the subgroup of subjects 1 and 2 (dotted line between , Fig. 2). However, the decrease of the average Vti value for all four subjects late in flight with respect to the preflight baseline was found to be highly significant (P < 0.001) and considerably below baseline (a 25% decrease). Together with the late in-flight Vti data, the steady and consistent slow recovery toward preflight baseline Vti values during the 10-day period after the mission also supports a correlation between the decrease of Vti and the net loss of body fluids in µG. The temporal changes of Vti indeed mimic the body weight profiles summarized in Norsk and Epstein (12). Unfortunately, we have too few data points to actually correlate Vti with body mass, because body mass measurements were performed neither onboard Spacelab-D2 nor during the R+9 session. Nevertheless, in support of the relation between a Vti decrease and body fluid loss, we note the average body weight decrease of 2.5 kg between preflight and R+2 and the average increase of 0.6 kg between R+2 and R+15 (i.e., the session where body weight was determined closest to and after R+9).

The Vti derived from the C₂H₂ rebreathing test has been used to assess clinically mild and even inapparent interstitial pulmonary edema in humans, with Vti values ranging from 100 to 200 ml/l of TLC (the corresponding Vti value for a group of normal subjects was 80 ml/l of TLC) (13). Permutt (14) hypothesized that subclinical pulmonary edema could develop on entry into the µG environment, and, if this were the case, it would be expected to increase Vti. The lack of an increase in Vti early in flight suggests that interstitial pulmonary edema does not occur as a result of exposure to µG. Nevertheless, we cannot totally exclude the possibility that a net fluid loss could have masked any potential factors contributing to Vti increase, such as the postulated interstitial pulmonary edema. Even if edema were to have occurred early in flight, the marked Vti decrease late in flight suggests that it must have resorbed later in flight. However, the hypothesis of time dependence in the occurrence and resorption of interstitial pulmonary edema is contradicted by the membrane diffusing-capacity data obtained in µG by Prisk et al. (15). In particular, the increase in membrane diffusing capacity in µG and the absence of any dependence on time spent in flight over a 10-day flight, counters the suggestion that the in-flight time dependence of Vti could have been due to early edema and absence thereof late in flight. Taken together, these data suggest that interstitial pulmonary edema as a result of exposure to µG did not occur.

The changes that we observed suggest that, despite documented cephalad fluid shifts from the lower extremities (4, 9, 12), this fluid does not move to the gas-exchanging region of the lung. Such a conclusion is consistent with recent findings of the absence of a rise in central venous pressure (5) and with the finding that the pulmonary interstitium is up to 30 times less compliant than other body tissues with respect to fluid accumulation (11). The transendothelial filtration coefficient and the Starling pressures (hydrostatic and osmotic pressure differences between capillary and interstitum) determine the rate and direction of fluid transfer. Increased capillary recruitment must be assumed to have increased the transendothelial filtration coefficient (11), but this effect appears to be reversed by the combined influences of a loss of plasma water (12) and the low tissue compliance in the lung, the latter tending to increase the hydrostatic tissue pressure markedly for any given increase of tissue fluid volume. Thus, despite clear evidence of excessive fluid in the tissues of the upper body, as evidenced by puffy faces, increased tissue thickness (9), and increased pulmonary capillary blood volume and SV (15), the lungs do not develop interstitial pulmonary edema in µG.

The other cardiopulmonary parameters computed here were FRC, c, DL_CO, and K_CO, which can all be compared with recently reported data obtained during the Spacelab SLS-1 mission (7, 15). For instance, the FRC decreases in µG shown in Fig. 1 are in good agreement with the 15% decrease reported by Elliott et al. (7): an 8% FRC decrease for the four subjects () and an average 16% FRC decrease for the subset of two subjects (). Also consistent with the findings of Elliott et al. (7) is the fact that FRC did not show any adaptation effects between early and late in flight.

c increases early in flight and returns to 12% below baseline toward the end of the mission. Postflight c is not significantly different from 100% from R+2 onward. The general pattern is compatible with the SLS-1 data (15), except that the magnitude of the initial (~24 h) increase was <50% the increase in c observed in SLS-1 (14% in D-2 vs. 35% in SLS-1). In both studies, c was no longer significantly different from preflight baseline value by the end of the mission and postflight. The impact of the different methods used for computation of c [namely, the methods of Sackner et al. (17) in D-2, the method of Ayotte et al. (3) in SLS-1, and the use of C₂H₂ instead of N₂O] are expected to be small, and the differences more likely reflect differences between the subject populations, especially with respect to the HR response. As in SLS-1, the SV pattern amplifies the behavior of c, with a more marked increase early in flight due to a combination of the 14% rise in c and a 19% decrease in HR. The initial change we observed in SV was 40%, which was also smaller than that seen during SLS-1 (15). The overall pattern of change was also similar, with a SV that is not significantly different from baseline by the end of the mission.

Finally, the diffusing capacity results (DL_CO and K_CO) coincide with the SLS-1 observations, namely that diffusing capacity increases in µG to a new steady-state level, supporting the concept of uniform capillary filling. Of particular interest is the fact that the techniques used in D-2 and SLS-1 were different, namely rebreathing vs. single-breath breath-hold test. The single-breath test was a vital capacity maneuver, with the measurement of CO uptake being essentially performed at TLC, where significantly negative intrathoracic pressures may have enhanced pulmonary capillary filling in µG. In contrast, the rebreathing maneuver used here (i.e., a breathing pattern with a baseline end-expiratory lung volume corresponding to FRC and a tidal volume of ~2 liters) effectively determines DL_CO for a mean lung volume of approximately FRC + 1 liter, providing more normal physiological conditions.

Direct comparison of DL_CO values between SLS-1 and D-2 is not possible. Two comparative studies (1, 16) exist where DL_CO values have been measured in the same subjects at rest by using both the single-breath and the rebreathing technique. When comparing measurements performed at similar lung volumes, Adaro et al. (1) found that DL_CO was 30% larger when measured by rebreathing compared with the single-breath method. However, Rose et al. (16) found no DL_CO differences by using either technique at two selected lung volumes. The difference in volume at which DL_CO is measured in the D-2 and SLS-1 study is a good reason to compare K_CO instead, even though K_CO itself is also known to depend on alveolar volume. In fact, Stam et al. (18) found K_CO to increase by ~30% when performing the single-breath maneuver at 50% instead of 100% TLC. When deriving K_CO from the DL_CO values and alveolar volumes reported in Rose et al. (16), K_CO is found to be only 10% larger when performing the single-breath test or rebreathing test at approximately FRC + 1.5 liters rather than near TLC. The results of Adaro et al. (1) and Stam et al. (18) both contribute to explain the 50% larger preflight K_CO values in D-2 than in SLS-1. Nevertheless, the findings of Rose et al. (16) would have predicted smaller differences between SLS-1 and D-2 on purely technical grounds. Also, the slightly higher alveolar O₂ concentration (typically 20-25%) in D-2 with respect to SLS-1 (16-18%) is expected to have a small but counteracting effect on the quantitative discrepancy between K_CO results obtained from both missions. Despite these methodological differences and any supposed physiological effects that might be induced, the results obtained in D-2 and SLS-1 are similar. Figure 4B shows that the 32% increase of K_CO in response to µG was actually in quantitative agreement with Prisk et al. (15), who found K_CO increases ~25% without any significant difference between measurements performed at different stages of the mission.

Because we have not been able to subdivide DL_CO into its components, capillary volume and membrane diffusing capacity (because of too few data points for each level of O₂ concentration), we cannot confirm the most important finding of the SLS-1 study, namely the increase of both these components in µG and, derived from it, the µG model suggesting more uniform capillary filling in µG. Nevertheless, by obtaining an increase of the low-O₂ DL_CO with respect to preflight (with a good agreement when normalizing for alveolar volume, i.e., K_CO), at least one of the conditions leading up to this model was met. We might have expected to see an increase in Vti in µG due to the likely increase in pulmonary capillary volume, since Vti incorporates capillary blood volume. However, it is most probable that the Vti parameter obtained by rebreathing is not sufficiently sensitive to pick up a capillary blood volume increase on the order of only 20 ml (15).

ACKNOWLEDGEMENTS

We acknowledge the dedication of the crew of the Spacelab-D2 mission and of all members of the Microgravity User Support Center of Cologne and the support of the European Space Agency.

FOOTNOTES

Address for reprint requests: S. Verbanck, Dienst Pneumologie (CPNE), AZ-VUB, Laarbeeklaan 101, 1090 Brussels, Belgium (E-mail: pnevks{at}az.vub.ac.be).

Received 23 December 1996; accepted in final form 7 May 1997.

REFERENCES