This investigation was designed to measure aerobic capacity (V̇o2peak) during and after long-duration International Space Station (ISS) missions. Astronauts (9 males, 5 females: 49 ± 5 yr, 77.2 ± 15.1 kg, 40.6 ± 6.4 ml·kg−1·min−1 [mean ± SD]) performed peak cycle tests ∼90 days before flight, 15 days after launch, every ∼30 days in-flight, and on recovery days 1 (R + 1), R + 10, and R + 30. Expired metabolic gas fractions, ventilation, and heart rate (HR) were measured. Data were analyzed using mixed-model linear regression. The main findings of this study were that V̇o2peak decreased early in-flight (∼17%) then gradually increased during flight but never returned to preflight levels. V̇o2peak was lower on R + 1 and R + 10 than preflight but recovered by R + 30. Peak HR was not different from preflight at any time during or following flight. A sustained decrease in V̇o2peak during and/or early postflight was not a universal finding in this study, since seven astronauts were able to attain their preflight V̇o2peak levels either at some time during flight or on R + 1. Four of these astronauts performed in-flight exercise at higher intensities compared with those who experienced a decline in V̇o2peak, and three had low aerobic capacities before flight. These data indicate that, while V̇o2peak may be difficult to maintain during long-duration ISS missions, aerobic deconditioning is not an inevitable consequence of long-duration spaceflight.
- aerobic capacity
- International Space Station
understanding the effects of prolonged microgravity exposure and concurrently developing methods to preserve astronauts' health and performance during long-duration spaceflight are key objectives of the National Aeronautics and Space Administration (NASA) Human Research Program. Preserving aerobic capacity, or at least mitigating losses, is desirable to protect the ability of astronauts to perform strenuous extravehicular tasks or emergency actions during a survival-type scenario in which the vehicle lands in a remote and unplanned location. In addition, although the precise plans regarding space exploration destinations are not well defined, future exploration activities may require a well-maintained aerobic capacity. Astronauts participating in International Space Station (ISS) missions perform exercise countermeasures using a treadmill, cycle ergometer, and a resistance exercise device in an attempt to defend bone, cardiovascular, and musculoskeletal health during spaceflight (31, 42, 44, 47); however, the effectiveness of these exercises to preserve aerobic fitness is not known because maximum aerobic capacity (V̇o2peak) during and after ISS missions has not been measured.
Two separate reports regarding peak cycle tests performed after short-duration Space Shuttle missions (8–14 days) showed a reduction in V̇o2peak and peak power by 22 and 14%, respectively (25, 30). To date, only Levine et al. measured V̇o2peak during short-duration spaceflight and reported no effect of spaceflight on V̇o2peak or peak power (25). In the absence of maximal exercise testing during or after long-duration spaceflight missions, changes in V̇o2peak have been inferred from examination of the heart rate (HR) response to submaximal cycle exercise (31). During the Skylab program flown in the early 1970s, astronauts did not experience a change in HR response to submaximal exercise during their 28- to 84-day flights (28, 29, 37) compared with preflight. This was assumed to reflect no change in V̇o2peak during flight. Conversely, most ISS astronauts experience an elevated HR response to submaximal graded exercise testing during the first 1–2 mo of flight, followed by a less exaggerated response later in the mission. These data suggest that, early in ISS flights, aerobic capacity is reduced compared with preflight and improves throughout flight (31, 34). The reason for the discrepancy in the HR response to the submaximal tests between Skylab and ISS astronauts is not known but may be because of differences in exercise training intensities, durations of the exercise countermeasures sessions, and exercise equipment available for use during the missions. Comparisons between ISS and Skylab results also may be confounded by differences in the atmospheric conditions existing in the stations. ISS is maintained at essentially earth-like atmospheric conditions, but the Skylab atmosphere was ∼70% O2, ∼30% N2, and ∼0.5% CO2 at a pressure of 34 × 103 Pa (Po2 ∼179 mmHg, Pn2 ∼76.5 mmHg, Pco2 ∼1.3 mmHg at a total pressure of ∼255 mmHg). Following both Skylab and ISS flights, crewmembers exhibited elevated HR responses to submaximal exercise initially, with a return to preflight levels occurring 1 mo after return to Earth (6, 31).
Examination of the submaximal HR response to graded exercise may provide an indication of change in aerobic fitness; however, there is significant variation in the estimates of V̇o2peak derived from using submaximal HR paired with either submaximal V̇o2 and/or cycle power data (17, 32). Metabolic gas analysis during tests to maximum effort must be performed to provide an accurate measure of V̇o2peak during and after long-duration spaceflight. The primary purpose of the current investigation was to document, for the first time, V̇o2peak during and after long-duration ISS missions. Based on previous literature demonstrating an elevated HR response to exercise during ISS missions and Space Shuttle data showing reduced V̇o2peak immediately after spaceflight, our hypotheses were that: 1) V̇o2peak would decline in the first month of flight and slowly increase toward preflight levels during the remainder of the mission, and 2) V̇o2peak would be lower than preflight immediately after landing and return to preflight levels within 30 days. This study provides the first time course measurements of V̇o2peak during and after long-duration spaceflight.
Subjects, test schedule, and facilities.
Testing for this study was performed during ISS Increments 19–33 (April 2009-November 2012). Fourteen astronauts [9 males, 5 females; 49 ± 5 yr, 175 ± 7 cm, 77.2 ± 15.1 kg (mean ± SD)] assigned to missions ranging from 91 to 192 days (164 ± 26 days) participated. The study was approved by the Institutional Review Board at the National Aeronautics and Space Administration Johnson Space Center (NASA JSC, Houston, TX), and the astronauts provided written informed consent before participation in the study. Subjects were scheduled to perform peak cycle tests at ∼9 mo and 90 days before launch, on flight day 15 (FD15), every subsequent 30 flight days, and on recovery days 1 (R + 1), R + 10, and R + 30 (Fig. 1); however, actual testing did not always occur when scheduled because of logistical or other mission constraints. All preflight and R + 30 tests were performed in the Exercise Physiology Laboratory at NASA JSC. The R + 1 and R + 10 tests were conducted in laboratory facilities either at NASA JSC, the Gagarin Cosmonaut Training Center (GCTC) in Star City, Russia, or at the NASA Kennedy Space Center in Florida. Room temperature at each of these facilities was controlled between 20 and 23°C. Testing during flight was conducted in the U.S. Laboratory Module of the ISS. The cabin conditions of ISS were similar to those on the ground and are maintained at ambient pressure of ∼760 mmHg, with a Po2 of ∼160 mmHg, Pco2 of ∼4 mmHg, and a Pn2 and other trace gases of ∼596 mmHg. The cabin temperatures ranged from 20 to 22°C and the relative humidity from 30 to 40%.
A peak cycle test was performed ∼9 mo before flight to satisfy a NASA medical requirement. This protocol consisted of three continuous 3-min stages (either at 50, 75, 100 W or 50, 100, 150 W, depending on the astronauts' body weight and self-report of activity levels), followed by 1-min stepwise increments of 25 W to volitional fatigue. The results of this test were used to establish the submaximal workloads for subsequent experiment-specific testing individualized for each subject. The experiment-specific protocol consisted of a 2-min resting period followed by three continuous 5-min power levels prescribed to elicit 25, 50, and 75% of the individual's preflight V̇o2peak. These were immediately followed by 1-min stepwise increments of 25 W until subjects reached their symptom-limited maximum. This protocol allowed for measurement of steady-state cardiovascular and metabolic variables during different submaximal exercise intensities (data not reported in this manuscript), followed by measurements of V̇o2peak. All ground-based exercise testing was performed in the upright seated posture on an electronically braked cycle ergometer (Lode Excalibur Sport, Groningen, NL), and in-flight tests were performed using the cycle ergometer with vibration isolation system (CEVIS), manufactured by the Danish Aerospace Company (DAC, Odense, DK). During in-flight CEVIS exercise, a waist belt was used to stabilize the subject's hips, cleated cycling shoes were clipped into the pedals, and side-mounted hand grips at approximately the waist level were used to stabilize the upper body and allow efficient transmission of force into the pedals. The cycle ergometers (both flight and ground) were calibrated via dynamometry before the experiment.
The dependent variables were peak power (W), HR (via ECG), V̇o2, carbon dioxide production, respiratory exchange ratio (RER), and ventilation (V̇E). For all experiment-specific testing, both on the ground and on board the ISS, metabolic gas analysis data were obtained using the portable pulmonary function system (PPFS, commissioned by the European Space Agency and manufactured by DAC). The PPFS uses two types of technology for gas analysis. CO2 is measured using a photoacoustic method of gas analysis (18). An Oxigraf model X2004 sensor (Oxigraf, Mountain View, CA), a laser diode absorption spectroscopy sensor (measurement range 0–100% O2 to a resolution of 0.01%), is integrated into the PPFS to measure O2. During exercise testing, the subject inspires through a custom-designed two-way nonrebreathing valve, and the expired gases are sampled in a 15-l capacity anesthesia bag, which serves as a mixing reservoir. Ventilation is measured on the inspired side of the nonrebreathing valve using a custom-designed differential pressure flowmeter (pneumotach). The PPFS gas analyzer module was calibrated before each testing session using reference gases, and the pneumotach was calibrated with a 3-l volumetric syringe. Temperature, humidity, and gas concentrations of O2 and CO2 were measured before each testing session and were used as the inspired values for metabolic gas analysis calculations. Before experimental testing, each PPFS underwent stringent acceptance tests to validate instrument performance. These tests included verifying that the gases were measured within the required accuracy for the specified concentration range (%volume) (O2: ±0.3%, concentration range: 0–100%; CO2: ±1%, concentration range: 0–12%). The differential pressure flow meters were required to operate in the range of −15 to 15 l/s with a required accuracy of 2% or 20 ml/s. The PPFS obtains raw data contributing to metabolic calculations at a frequency of 100 Hz. Further information regarding the technology used for metabolic gas analysis within the PPFS is contained in a paper by Clemensen and colleagues (7).
A proprietary software package [Agile Data Analyzer and Monitor (ADAM); DAC] was used to compute metabolic gas analysis variables after testing. Within ADAM, the user specifies the time range for metabolic calculations. V̇E is calculated as the sum of the inspiratory flow within the specified time range divided by the range interval, which is defined as the elapsed time between initiation of an inspiration closest to the beginning of the specified time range to the initiation of an inspiration nearest the end of the specified time range. The resulting value is converted to a “per minute” value. The mixed-expired gas fraction data (FeO2, FeCO2) are the averages that occur in the mixing bag over the specified range. Within the metabolic calculations, the ADAM software takes into account the fixed delay time for the sensors to detect expired gas (6 s) and a variable time shift based upon the ventilation and breathing frequency of the subject to match the gas fraction data to ventilation. The formulas for the time-shift calculation are found in the paper by Jensen and coworkers (22). As inspired ventilation is measured, V̇o2 is calculated using the Haldane transformation (48).
All pre- and postflight testing was conducted by the same three test operators. Testing during flight was conducted in the U.S. Laboratory Module of the ISS. The astronauts participating in this study performed three preflight training sessions on setting up and operating the PPFS so they could run the in-flight tests by themselves. The in-flight tests were supported on the ground by a member of the science team (one of the test operators who conducted the preflight and postflight testing) and by a DAC team member. Live video feed was required for all in-flight tests for the investigator team to confirm proper equipment set-up, calibration, and data collection.
Data reduction and calculations.
The peak metabolic gas analysis data and peak HR data were obtained from the final 60 and 10 s of exercise, respectively. Peak power was calculated using a time-weighted average over the final 60 s of exercise before cool down. For example, if the subject achieved 30 s at 350 W, the value recorded for peak power attained was 337.5 W (one-half of the final 60 s of the test was spent at 325 W, one-half was at 350 W).
Exercise countermeasures performed during ISS flight.
Astronauts performed cycle ergometry, treadmill running, and resistance exercise during flight as prescribed by Astronaut Strength, Conditioning, and Rehabilitation (ASCR) specialists (31, 44). Exercise prescriptions generally consisted of 6 days/wk of upper and lower body resistance exercise and 4–6 days/wk of moderate- to high-intensity aerobic exercise. Exercise time, power, speed, and HR were recorded during cycle ergometry and treadmill exercise.
The test sessions conducted during flight did not consistently follow the planned schedule due to logistic constraints and the need to accommodate other mission activities. As a result, the unbalanced nature of our repeated-measures longitudinal data set led to the use of mixed-model linear regression analyses (8) with V̇o2peak, peak power, peak HR, peak RER, and peak V̇E as outcomes and preflight body weight as a covariate. Besides allowing for random differences between subjects, the mixed models contained fixed coefficients representing 1) the change early in-flight due to the microgravity environment, 2) the mean change per flight day, and 3) mean values of these peak responses during recovery from flight, all adjusted for differences in preflight body weight. Based on the regression model, 95% confidence intervals (95% CI) for the peak responses preflight, in-flight, and postflight were calculated for a subject of average weight (77.2 kg for our subjects). Comparisons of in-flight and postflight peak exercise values to preflight were performed as a set of preplanned comparisons using significance tests following the mixed-model linear regression analyses. Significance thresholds for P values were calculated to allow for the testing of multiple null hypotheses per variable using the Hochberg (20) procedure for controlling the family-wise error rate to 0.05.
Peak exercise response.
All preflight and in-flight tests were performed normally with no adverse events. On R + 1, one test was terminated early due to subject symptoms (dizziness). On R + 30, one subject could not be tested due to a medical condition that arose unrelated to the study. Preflight body weight was slightly higher than postflight body weight (preflight = 77.2 ± 15.2 kg; R + 1 = 75.3 ± 14.4 kg; R + 10 = 75.7 ± 15.6 kg; R + 30 = 76.2 ± 14.8 kg, mean ± SD).
Mean peak values for V̇o2, power, HR, RER, and V̇E are presented in Table 1. The mixed-model linear regression showed mean V̇o2peak and peak power were substantially decreased at the first in-flight test (∼FD15) by 17% [95% confidence = (−22%,−13%)] and 24% [95% confidence = (−29%,−20%)], respectively (Fig. 1). Subsequent to FD15, V̇o2peak and peak power demonstrated modest linear upward trends during flight (V̇o2peak: 0.0013 l·min−1·day−1, P = 0.02; peak power: 0.177 W/day, P < 0.001) (Fig. 2). On R + 1 and R + 10, mean V̇o2peak and peak power values were significantly lower than preflight but had recovered to levels not distinguishable from preflight by R + 30. The V̇o2peak results for individual subjects and time points are displayed in Fig. 3. Although many of the subjects experienced a decline in V̇o2peak during and/or early following flight, this was not a universal finding. During the first one-half of flight, the mixed-model estimate of mean peak V̇E was significantly lower than preflight levels (Δ = ∼10 l/min) but did not differ from preflight levels by FD180 (Δ = −4.0 l/min) (Table 1). Postflight V̇E was not different from preflight except on R + 10 where a significant elevation was found (∼11.5 l/min).
Aerobic exercise training performed during ISS flight.
HR data were available for 82% of the daily exercise sessions. HR data from one astronaut were not reported because of poor data quality throughout the entire mission. Subjects completed an average of five to six aerobic exercise sessions per week [161 ± 39 min/wk (mean ± SD)] at an average intensity of 73.8 ± 13.2% of peak HR. The subjects averaged 58 ± 15% of their exercise sessions on the treadmill and 42 ± 15% of the sessions on the cycle ergometer. There were eight astronauts who had a preflight V̇o2peak of >40.0 ml·kg−1·min−1. Of these, four maintained their V̇o2peak from preflight to postflight (<3% decrease from preflight). Although the subgroupings are too small to statistically analyze, the four with a preflight V̇o2peak of >40.0 ml·kg−1·min−1 and who maintained their V̇o2peak exercised on the CEVIS at a higher percentage of peak HR [79 ± 6 vs. 68 ± 20% (mean ± SD)] and spent a greater percentage of their time exercising above 70% of peak HR [76 ± 30 vs. 63 ± 32% (mean ± SD)] compared with those who experienced a decrease in V̇o2peak.
This study is the first to report V̇o2peak from exercise tests performed during and after long-duration ISS missions. Compared with previous reports of changes in V̇o2peak during and following short-duration spaceflight (25, 30), this study included a much larger sample size to describe changes in astronauts' V̇o2peak. The primary findings of the study are: 1) mean V̇o2peak declined (∼17%) early and improved only slightly throughout the remainder of the missions, 2) mean V̇o2peak was reduced on R + 1 from preflight by ∼15% but recovered by R + 30, and 3) not all of the astronauts experienced a decline in V̇o2peak during or following their missions.
Change in V̇o2peak during spaceflight.
Previous studies that have measured V̇o2peak during spaceflight were from short-duration flights and report conflicting results. Different from our ISS results, Levine et al. (25) and Moore et al. (30) concluded that neither V̇o2peak nor peak power was changed while in microgravity during Space Shuttle missions. In contrast, V̇o2 data obtained at 85% of preflight maximal power during a 17-day mission (n = 4) showed a decline in V̇o2 of 6% on FD2 to 11% on FD11 (45). During flight, three of the crewmembers attained their preflight peak HR at the 85% stage, suggesting the in-flight tests were near maximal. These data provided the first evidence that aerobic deconditioning likely occurred during spaceflight. The discrepancy in reported change in V̇o2peak between studies is likely due to a variety of factors, including: 1) durations of the flights in the current study are much longer than those of previous studies, 2) higher mean V̇o2peak of the astronauts in the current investigation and in the Trappe et al. (45) study [3.16 ± 0.10 and 3.59 ± 0.13 (SE) l/min, respectively] compared with those in the Levine et al. (25) study (preflight V̇o2peak: 2.76 ± 0.33 l/min), 3) substantial differences in the metabolic gas analysis hardware used to measure V̇o2peak, and 4) differences in the in-flight exercise countermeasures.
Examination of individual data (Fig. 3) provides the impression that those with a higher initial V̇o2peak experience greater decreases from preflight to the first in-flight test. This relationship was evaluated using an instrumental variable regression with preflight body weight as the exogenous instrument. The Wu-Hausman test for endogeneity of the predictor (preflight V̇o2peak) indicated that there was some effect of regression to the mean (P = 0.04), but, even after correcting for that effect, there was still a significant negative correlation (corrected R2 = 0.59; P = 0.006) (18). Thus, the data support the contention that those astronauts who have higher initial aerobic capacities are more prone to loss of V̇o2peak early during flight. However, this finding should not be interpreted that a high preflight aerobic capacity is undesirable. Although the astronauts with high capacities tended to lose more, they typically remained at higher levels than crew who started at lower levels (see Fig. 3, subjects F vs. N as a good example).
For the ISS astronauts studied here, the early in-flight decline of V̇o2peak is likely due to several factors, including time without exercise in the several days, up to a week, following docking with the ISS, space motion sickness, and cephalad fluid shifts that contribute to a decrease in blood volume (BV), primarily the plasma component, potentially resulting in decreased muscle perfusion pressure and initiating some degree of central cardiovascular deconditioning (47).
With regard to the relationship between BV and aerobic deconditioning, Coyle and coinvestigators reported a 6% (P < 0.01) reduction in V̇o2peak, accompanied by a 9% (P < 0.01) decline in BV following a 2- to 4-wk detraining period of eight endurance-trained subjects (13). A recent paper documented a 7% (P < 0.05) increase in BV accompanied with a 10% (P < 0.05) increased V̇o2peak following 6 wk of endurance training (5). The investigators then removed, via phlebotomy, the training-induced BV and observed V̇o2peak returned to a level not differing from pretraining values. Thus, acute and rapid changes in BV, such as those induced by spaceflight, likely contribute to the decline in V̇o2peak. Although speculative, differing changes in BV may partially explain why the subjects who possessed higher V̇o2peak levels before flight had more difficulties in maintaining during flight than those with lower aerobic capacities. Decreased red blood cell mass has been documented to occur early during spaceflight; however, the concomitant decline in plasma volume likely results in little to no net change in the oxygen-carrying capacity of the blood (2, 3, 41).
The reduction in muscle perfusion pressure that occurs in microgravity can be simulated with supine exercise performed in 1 G. A well-documented 10–15% reduction in V̇o2peak is observed in supine compared with upright exercise, which is close to the early decrease in V̇o2peak observed during flight in the current study (35). In opposition, it should also be noted that this “postural effect” was not observed in the short-duration Space Shuttle studies that measured in-flight V̇o2peak and peak power (25, 30).
Once established in microgravity, and after the initial decline in V̇o2peak, the subject group of the current study experienced a modest recovery in aerobic capacity and peak power attained (Fig. 2). The trend for improved V̇o2peak observed during flight may be due to physiological normalization to the spaceflight environment, as well as the cardiac, BV, and peripheral muscle training adaptations expected in individuals who perform regular aerobic exercise. It is possible that some device or “postural”-specific training is also involved. Exercise on the ISS is more similar to exercise conducted in the supine posture, rather than in the upright posture, on Earth. Ray et al. reported significantly greater improvements in supine V̇o2peak after training in the supine compared with the upright posture (36).
Change in V̇o2peak following spaceflight.
The postflight result showing a significant decrease in V̇o2peak (∼15%) immediately upon landing with a return to preflight levels by R + 30 is consistent with previous spaceflight investigations (25, 30). Following short-duration missions, V̇o2peak was similarly decreased in the first few days after landing but recovered during the 2 wk following flight (21, 27, 43). Levine et al. (25) showed an approximate 16% reduction in V̇o2peak measured 24–48 h after landing and a full recovery to preflight levels by R + 7. Moore et al. (30) reported similar reductions on the day of landing. The current study showed the astronauts' V̇o2peak recovered by R + 30 but remained below preflight at R + 10. Additionally, our data showed a surprising spike in peak V̇E at the R + 10 time point only. This may indicate a compensatory mechanism for diminished cardiovascular function, peripheral deconditioning, or overshoot in the recovery of plasma volume; however, at this point there are no data to clearly explain this result.
Reductions in V̇o2peak early following flight may be attributed to a combination of factors, including decreased plasma volume, reduced red cell mass, orthostatic intolerance, and deconditioning (central cardiac and peripheral muscle) (2, 3, 10, 26). Although the relative contributions of these factors toward reductions in V̇o2peak are difficult to determine, peak cycle tests performed in the upright and supine posture provide some insight in evaluating the specific effects of orthostasis vs. deconditioning on V̇o2peak. Reductions in upright V̇o2peak after long-duration bed rest are ∼10% greater than when measured in the supine position (9, 11, 21, 43).
The longer recovery observed in the current study may be due to increased deconditioning during ISS compared with Space Shuttle flights or the higher preflight mean V̇o2peak of the astronauts in this study. Bed rest studies indicate that V̇o2peak is fully recovered within a week of bed rest lasting ∼2 wk (45), whereas recovery can take up to 2–4 wk after longer-duration bed rest (43). Conversely, long-duration bed rest studies that included a resistance and aerobic exercise countermeasure or a high-intensity aerobic exercise routine successfully protected V̇o2peak immediately after bed rest (23, 38, 46).
Effects of deconditioning and reconditioning on V̇o2peak.
Deconditioning of the cardiac and peripheral muscle that occurs during spaceflight likely affects in-flight and postflight V̇o2peak. Data regarding cardiac atrophy and reduced cardiac function after spaceflight are mixed. Four astronauts participating in a 10-day mission demonstrated an average decrease in left ventricular (LV) mass (12 ± 7%) (33). Four astronauts who participated in flights lasting from 129 to 144 days demonstrated greater decreases in LV ejection fraction and percent fractional shortening as well as an increase in LV end-diastolic volume (27). However, preliminary reports from the NASA Integrated Cardiovascular Study, involving eight male and five female astronauts, some of whom were also subjects of the present study, indicate that no deleterious effects of prolonged spaceflight were observed in cardiac structure or function (1). In addition, cardiac atrophy during long-duration spaceflight showed much variation between individual astronauts but as a group did not change over flight (+1%) (39). Bed rest data suggest that LV impairment occurs due to changes in the distensibility of cardiac tissue (15, 16, 33) and a reduction in cardiac diastolic function (26). There is evidence that the addition of an exercise countermeasure during bed rest prevents or mitigates losses in cardiac function (19, 40). Impaired muscle strength and oxidative capacity due to leg muscle atrophy and fiber type shifting have also been reported (41, 44). An investigation studying skeletal muscle changes of nine ISS astronauts on flights of ∼6 mo documented atrophy of both the soleus and gastrocnemius (average −15% and −10% change in volume, respectively) and a shift from the slower oxidative fibers toward the faster glycolytic phenotypes (45). These findings suggest that aerobic metabolism of working muscles may be impaired during long-duration flight.
The recovery time of V̇o2peak after long-duration bed rest and the effects of reconditioning programs have not been specifically studied; however, similar to the astronauts in the current study, most individuals in bed rest recover their aerobic capacity within 2–4 wk with the resumption of normal daily activities (12, 43). All astronauts participate in a reconditioning program following flight beginning on R + 0 and continuing daily through R + 45. The program focuses on activities to improve aerobic conditioning, weight training, stretching, balance, and proprioception. The progression of the reconditioning program is individualized to each astronaut.
Exercise countermeasure effectiveness.
In-flight exercise intensity and volume are important considerations in interpreting the study results. Notably, examination of the individual test results (Fig. 3) shows that seven of the subjects were able to attain V̇o2peak values at or above preflight levels during their missions, providing evidence that aerobic detraining is not an inevitable consequence of long-duration flight. Of these astronauts, three (Fig. 3, subjects A, J, and K) were participants in an ongoing exercise study that focuses on high-intensity, low-volume exercise training as a method to efficiently maintain aerobic fitness and muscle strength, and one (subject H) exercised at a relatively high intensity (averaging >80% of maximal HR on both the CEVIS and treadmill). The remaining three astronauts not experiencing a loss in V̇o2peak possessed low aerobic capacities before flight (Fig. 3, subjects B, F, and M). Subject B's absolute V̇o2peak does not look particularly low compared with the others in the study; however, the preflight relative V̇o2peak of this subject was 31 ml·kg−1·min−1, much lower than the mean of all subjects (41 ml·kg−1·min−1). The data of this study tend to support the notion that higher-intensity exercise can be used to mitigate losses in V̇o2peak during spaceflight as evidenced by the observation that individuals with maintained V̇o2peak spent ∼80% of their in-flight aerobic training time at ∼80% of maximum HR. Subjects who showed declines in V̇o2peak spent, on average, less time exercising at high in-flight HR. An important finding of this study was that, when prescribed appropriately, exercise using the current ISS hardware can protect V̇o2peak at preflight levels.
Studies involving spaceflight present challenges not encountered in ground-based laboratory investigations. Due to launch delays, the PPFS did not arrive until approximately half-way into the mission of the first three astronaut subjects. One test session of another subject was cancelled in favor of mission priorities during a period when only two astronauts were aboard the ISS. Several of the test dates were shifted during flight, primarily due to conflicts with either extravehicular activities or arrival of provisioning vehicles. A potential limitation caused by the microgravity environment was the influence on filling of the mixing bag of the PPFS. On the ground, the bag fills and the mixing volume stabilizes at ∼4 l; however, without the effect of gravity we observed, via video downlink, the bag inflated in proportion to ventilation, peaking at ∼⅔ full at heavy exercise. To examine the “worst case” influence of this change, we calculated the difference in V̇o2peak for the astronaut that attained the highest value during flight using 10 l as the mixing volume instead of 4 l. The difference in V̇o2peak was negligible (3.817 l/min for 4 l vs. 3.812 l/min for 10 l).
The last preflight measurement and the first in-flight data collection session were separated by ∼100 days; therefore, the early in-flight changes cannot automatically be attributed to spaceflight alone. However, the astronauts enrolled in this study were specifically asked and all reported that they continued their normal exercise habits during the ground portion of this time span. The ergometers used for ground testing and for testing on ISS were not identical; therefore, the preflight vs. postflight comparisons and analysis of data taken solely during flight may be more accurate than the ground vs. flight comparisons. The limited availability of astronaut time did not allow for a secondary testing session using a supramaximal stage to verify that a true V̇o2 max (plateau in V̇o2 with an increase in power) was achieved; however, several studies indicate that V̇o2peak determined during incremental tests does not differ from V̇o2 max in well-motivated and healthy subjects (14, 24). Additionally, ∼50% of subjects do not show a plateau in exercise V̇o2 with maximum effort (4). The astronaut participants were well-motivated volunteers, were briefed on the importance of giving an all-out effort for each test, and consistently reported that they gave their best effort. The investigators are confident the results accurately reflect the astronauts' capacities to perform aerobic exercise before, during, and following flight.
In conclusion, most astronauts demonstrated a rapid decline in V̇o2peak in the early weeks of the mission that improved slightly throughout flight but did not fully recover to preflight levels. V̇o2peak again decreased abruptly on R + 1 and recovered to preflight levels 30 days after landing. There was a significant amount of variability in the in-flight and postflight change in V̇o2peak between astronauts that appears to be related to in-flight aerobic exercise intensity and the aerobic capacity of the individual astronauts before flight. These results provide evidence that, although many astronauts experience a decline in V̇o2peak during ISS missions, use of the aerobic exercise hardware aboard the ISS combined with exercise prescriptions of sufficient exercise intensity can be used to effectively prevent decline in aerobic capacity.
This research was supported by the NASA Human Research Program.
No conflicts of interest, financial or otherwise, are declared by the authors.
Author contributions: A.D.M. and S.M.C.L. conception and design of research; A.D.M. and M.E.D. performed experiments; A.D.M., M.E.D., A.H.F., and P.K. analyzed data; A.D.M., M.E.D., A.H.F., and L.L.P.-S. interpreted results of experiments; A.D.M. drafted manuscript; A.D.M., M.E.D., S.M.C.L., A.H.F., P.K., and L.L.P.-S. edited and revised manuscript; A.D.M., M.E.D., S.M.C.L., A.H.F., P.K., and L.L.P.-S. approved final version of manuscript; M.E.D., S.M.C.L., and A.H.F. prepared figures.
Successful completion of this study is due to the immense support of members of the National Aeronautics and Space Administration (NASA) family and the larger “spaceflight community” of the European Space Agency (ESA), Japanese Aerospace Exploration Agency, Canadian Space Agency, and Russian Agencies. We extend our thanks to the following: 1) the astronauts who enthusiastically contributed their time and efforts to the research, 2) the members of the NASA Johnson Space Center (JSC) Exercise Countermeasures Laboratory who helped in the conduct of the study, 3) the personnel of the Danish Aerospace Company in Odense, Denmark, who provided excellent hardware and sustaining engineering support, 4) the ESA Life Sciences and Medical Operations Offices, 5) the International Space Station Medical Project support group, 6) physicians and support personnel of NASA JSC Medical Operations, and 7) the NASA Human Research Program support team. We thank Erik Hougland and Jennifer Wilson for many hours of dedicated training and in-flight monitoring support and Jamie Guined for assistance during data collection. Also, we thank Benjamin Levine for counsel during the preparation of this manuscript.
- Copyright © 2014 the American Physiological Society