We have shown previously that treadmill exercise within lower body negative pressure (LBNPex) maintains upright exercise capacity (peak oxygen consumption, V̇o2peak) in men after 5, 15, and 30 days of bed rest (BR). We hypothesized that LBNPex protects treadmill V̇o2peak and sprint speed in women during a 30-day BR. Seven sets of female monozygous twins volunteered to participate. Within each twin set, one was randomly assigned to a control group (Con) and performed no countermeasures, and the other was assigned to an exercise group (Ex) and performed a 40-min interval (40–80% pre-BR V̇o2peak) LBNPex (51 ± 5 mmHg) protocol, plus 5 min of static LBNP, 6 days per week. Before and immediately after BR, subjects completed a 30.5-m sprint test and an upright graded treadmill test to volitional fatigue. These results in women were compared with previously reported reductions in V̇o2peak and sprint speed in male twins after BR. In women, sprint speed (−8 ± 2%) and V̇o2peak (−6 ± 2%) were not different after BR in the Ex group. In contrast, both sprint speed (−24 ± 5%) and V̇o2peak (−16 ± 3%) were significantly less after BR in the Con group. The effect of BR on sprint speed and V̇o2peak after BR was not different between women and men. We conclude that treadmill exercise within LBNP protects against BR-induced reductions in V̇o2peak and sprint speed in women and should prove effective during long-duration spaceflight.
- artificial gravity
- lower body negative pressure
the majority of planned tasks during spaceflight and exploration missions are designed to require submaximal efforts, but it is inevitable that some work, including a response to an emergency situation, will include near-maximal to maximal bouts of exertion or moderate to high levels of endurance. Many of these critical tasks will require an absolute level of work that cannot be scaled to an individual's cardiopulmonary or musculoskeletal fitness. Although many of the disparities between men and women are reduced when controlling for lean tissue mass, total body mass, and total hemoglobin concentration, women generally have lower absolute levels of maximal oxygen consumption and muscular strength (22). Therefore, physical fitness may be more critical for women than for men during spaceflight to protect the crewmembers' ability to perform mission critical tasks.
Cardiovascular and musculoskeletal deconditioning is an unavoidable consequence of exposure to spaceflight when exercise countermeasures are not performed (5, 6, 16, 37, 48). Whether women and men experience similar losses in exercise capacity during spaceflight remains unclear at this time because insufficient data from female astronauts exist from which to draw firm conclusions (22). In the bed rest (BR) model of spaceflight, the reduction of participants' aerobic capacity relative to their pre-BR fitness level during cycle ergometry testing is not different in men and women (9, 10, 26). However, no published studies have investigated aerobic and anaerobic exercise performance during upright locomotion in women after BR, which is more functionally important than supine exercise capacity (29, 30, 47), and few investigations have specifically examined the effectiveness of an exercise countermeasure in women (26, 41, 44).
In this investigation we sought to determine whether the performance of a supine treadmill exercise countermeasure during lower body negative pressure (LBNPex) would protect upright exercise performance in women after BR as well as had been previously documented in men (30). Specifically, we hypothesized that 40 min of exercise against LBNP, plus 5 min of postexercise resting LBNP, performed 6 days/wk would prevent the loss of upright aerobic capacity (peak oxygen consumption, V̇o2peak), changes in submaximal exercise responses, and decreased sprint speed in women after 30 days of 6° head-down tilt BR. Utilizing a BR and countermeasure protocol that was the same as our previous BR study in men (30), we also compared the amount of deconditioning in control subjects between women and men. We hypothesized that the amount of deconditioning in women and men would not be different when expressed relative to their pre-BR fitness levels but that the absolute change in aerobic capacity and sprint speed would be less in women. This study is unique in its evaluation of an exercise countermeasure to protect aerobic and anaerobic locomotion tasks during a moderate-duration BR utilizing identical female twins and in its ability to directly compare these results to those obtained from male twins who participated in the same BR and countermeasure protocols.
Seven pairs of female twins (24 ± 3 yr, 164.5 ± 9.4 cm, 56.6 ± 9.8 kg, means ± SD) volunteered to participate. A unique aspect of this study was that by using identical twins as test subjects, we were able to control for genetic factors related to responsiveness to BR; the only factor differentiating the two twins at the end of BR was the performance of the countermeasure protocol. Monozygosity was confirmed from a DNA sample obtained using a cheek swab kit and using standard DNA polymorphism analysis (STR markers D3 S1358, vWA, D16 S539, D2 S1338, D8 S1179, D21S11, D18 S51, D19 S433, Tho1, and FGA). Protocols were reviewed and approved by the Institutional Review Board at the University of California-San Diego (UCSD) and the Committee for the Protection of Human Subjects at the National Aeronautics and Space Administration (NASA) Lyndon B. Johnson Space Center (JSC). Subjects received verbal and written explanation of all procedures before providing written informed consent before participation.
Subjects were admitted to the UCSD General Clinical Research Center (GCRC) 7 days before BR for a period of ambulatory control during which they underwent familiarization and pre-BR testing sessions. Subjects provided a detailed medical history and received a complete physical examination, including a urine pregnancy test to confirm that each subject was not pregnant before proceeding with the study. If subjects were taking prescribed oral contraceptives, these medications were continued for the duration of the study since the risk of deep vein thrombosis during BR is considered low in healthy subjects. Twin pairs were housed together as roommates.
The female twins completed a period of 30 days of strict 6° head-down BR. Except for daily 10-min showers and the countermeasure sessions, both conducted in the horizontal posture, all activities during BR were conducted in 6° head-down tilt, including eating, drinking, urination, bowel movements, and transport to and from testing and countermeasure sessions. Subjects remained in the GCRC for 3 days after the BR for post-BR testing and readaptation to normal upright activities.
Subjects completed two pre-BR test sessions that consisted of a series of tests performed in serial order: an orthostatic tolerance test to presyncope (49), sprint trials, balance testing, and a graded upright treadmill test (GXT) to volitional fatigue. All tests were conducted 2 h postprandial, and the pre-BR test sessions were separated by at least 72 h. The first test session served as familiarization, and only the data from the second pre-BR test day was used as the baseline measure. After the completion of all pre-BR tests and immediately before the start of BR, one twin within each twin pair was randomly assigned to the no-exercise control (Con) condition and the other to the exercise countermeasure group (Ex). Immediately post-BR, the testing was repeated in all subjects in the same order as pre-BR and at the same time of day. Orthostatic tolerance data have been reported elsewhere (49).
Subjects consumed a diet consisting of 55% carbohydrate, 15% protein, and 30% fat. The Dietary Department staff of the UCSD GCRC prepared the food, and diet records were maintained to insure consistency throughout the pre-, in-, and post-BR periods. Initial caloric consumption was prescribed for each subject using the Harris-Benedict equations, adjusting for self-reported activity level before GCRC admission (correction factor of 1.4–1.5) (23). Thereafter, caloric consumption was adjusted based on daily body mass measurements such that subjects maintained their body mass within ± 1.0 kg. Body mass was measured in the supine position at 0700 each morning before breakfast using a bedside scale (Century CC894; Hill-Rom, Batesville, IN). In addition, sodium consumption was maintained at 3,500 mg/day, calcium consumption was targeted at 800–1,200 mg/day, and dietary fiber was prescribed at 25 g/day. Fluid intakes and output were measured daily. Fluid intake was ad libitum, but no caffeinated or alcoholic beverages were allowed.
During BR, the Ex subjects performed 40 min of supine treadmill exercise within LBNP, followed by 5 min of resting supine LBNP without exercise (30) 6 d·wk−1. Target exercise intensities for the exercise portion of this protocol consisted of 7 min at 40%, 3 min at 60%, 2 min at 40%, 3 min at 70%, 2 min at 50%, 3 min at 80%, 2 min at 60%, 3 min at 80%, 2 min at 50%, 3 min at 70%, 2 min at 40%, 3 min at 60%, and 5 min at 40% pre-BR V̇o2peak. Target speeds to achieve these exercise intensities were prescribed based on a linear relationship between treadmill speed and oxygen consumption measured during the pre-BR GXT.
The exercise device used for this study was the same as used in previous studies in men (29, 30, 47). The device consisted of a vacuum chamber in which a supine subject ran comfortably on a vertically oriented treadmill (PaceMaster SX-Pro; Aerobics, Little Falls, NJ). Chamber pressure was reduced using a high-capacity vacuum cleaner. A small amount of leakage was allowed in the chamber such that airflow through the system minimized heat accumulation caused by the treadmill motor and the exercising subject. Interchangeable wooden plates were used so that the aperture size through which the subject's legs and lower torso projected into the chamber could be adjusted. The size of the elliptical opening was chosen to be approximately twice the subject's waist cross-sectional area (29, 30, 47) such that the LBNP required to produce 1.0 times body weight (BW) was 45–55 mmHg. The linear relationship between LBNP and footward force for each subject was used to determine the appropriate amount of LBNP to produce footward loading titrated between 1.0 and 1.1 times BW for each individual's countermeasure sessions.
The subjects wore a neoprene waist seal with shoulder straps to prevent the seal from sliding down the subject's body and to provide some amount of axial loading of the spine. A sling supported the subject's upper body with a solid back support outside the chamber, and the hips were supported with a soft sling inside the chamber. Cuffs were fitted around the knees and ankles. A bungee cord connected the cuffs at their respective positions through pulleys suspended from the ceiling of the chamber. In this way, the legs counterbalanced each other and minimized the work required to move the legs vertically against gravity. The distance between the treadmill and the elliptical chamber opening was adjusted such that the iliac crest was positioned at the level of the opening when the subject was pulled down to the treadmill by the suction of LBNP.
Graded exercise test.
Subjects performed two upright treadmill GXTs before BR and one GXT at the end of BR. The first test served as a familiarization session and consisted of one 5-min stage of level walking at 4.8 km/h [3 mile/h (mph)] followed by three 3-min stages of 6.4, 8.1, and 9.7 km/h (4, 5, and 6 mph) at 0% grade. Thereafter, treadmill speed was held constant but grade was increased in increments of 3% each minute until volitional fatigue.
Immediately pre- and post-BR, the GXT consisted of 5 min of level walking at 4.8 km/h (3 mph) followed by three 3-min stages of increasing exercise intensity (∼65, 75, and 85% of pre-BR V̇o2peak) at 0% grade. Similar to previous studies (29, 30, 47), treadmill speeds for this second GXT were individually prescribed based on the results of the familiarization test. For this set of subjects, the level running speeds were 6.8 ± 1.5, 7.9 ± 1.5, and 9.0 ± 1.6 km/h (4.2 ± 1.1, 4.9 ± 1.3, and 5.6 ± 1.5 mph; means ± SD). After the third steady-state running stage, treadmill speed was held constant but grade was increased by 3% each minute until a volitional fatigue end point was achieved.
Subjects ran on a treadmill (model ST55 or Q65; Quinton Instruments, Seattle, WA) calibrated for speed and grade. Oxygen consumption (V̇o2; l/min and ml·kg−1·min−1), carbon dioxide production (V̇co2; l/min), ventilation rate (V̇e; l/min), and respiratory exchange ratio (RER) were measured continuously with a metabolic cart (Qplex I Metabolic Cart; Quinton Instruments). Heart rhythm and heart rate (HR) were monitored continuously with a three-lead ECG system and a heart rate monitor (Polar Vantage NV; Polar Electro, Lake Success, NY).
Measures of V̇o2, V̇co2, V̇e, and RER were determined by averaging the last two 30-s samples for each submaximal exercise stage. HR was measured by averaging the last two 15-s heart rates measured in the last 30 s of each stage. Peak oxygen consumption (V̇o2peak) was determined as the average of the highest level of V̇o2 measured in two consecutive 30-s samples. V̇o2peak was accepted when at least two of the following criteria were achieved: RER was >1.1, peak HR was >85% of the age-predicted maximum, and/or a plateau of the V̇o2 curve was observed (<150 ml/min increase). Using a similar protocol in a separate group of subjects, we have observed a high degree of reliability for this measure of V̇o2peak [intraclass correlation (ICC) = 0.90, SE = 1.3 ml·kg−1·min−1].
Subjects performed a maximal sprint from a standing start over a 30.5-m (100-ft.) distance in an enclosed hallway. All sprints were conducted in the same location with the same test operators. The sprint times were measured with standard stopwatches by two different test operators, and their respective measurements were averaged. Subjects performed two sprint trials at each testing session, separated by at least 1 min, and the faster of the two trials was used for subsequent analysis and computation of average speed. Sprint time has been demonstrated to have a high degree of reliability across similar distances (ICC = 0.85–0.97) (11, 33).
Plasma volume measurement.
Plasma volume was measured at 0700 the day before the second GXT (ambulatory control) and at the same time on the day before the end of BR (day 29) using the 125I-labeled human serum albumin (RISA) dilution method (18). Subjects remained recumbent for at least 30 min before and throughout each plasma volume test. Subjects were supine during pre-BR plasma volume measurements but remained in head-down tilt during the end of BR measurements. No exercise was performed in the 24 h before the pre-BR plasma volume measurement.
Approximately 20 min before injection of 125I, three drops of a saturated potassium iodine solution (Lugol's solution) were mixed with orange juice and given orally to each subject to reduce the amount of radioactive iodide absorbed by the thyroid gland. A venous blood sample was obtained for the background measurement before the injection of 10 μCi of 125I-labeled human serum albumin through a venous catheter inserted in the antecubital space. Venous samples were drawn 10 and 20 min after injection from a venous catheter in the opposite arm.
Body composition was assessed from a whole body scan using a standard clinical dual-energy X-ray absorptiometer (model DPX-IQ; Lunar, Madison, WI). From these scans, lean tissue mass, fat mass, and percent body fat were determined. Body composition was measured once before the start of BR and 2 days after the end of BR. Using similar measurements, previous authors have reported that the coefficient of variation is 0.9% for the whole body (40) and 1.3–1.4% for the legs (13, 40).
A power analysis was performed to determine the number of subjects required to achieve statistical significance based on the primary dependent variable of interest, V̇o2peak. The mean pre-BR V̇o2peak as well as test-retest variance was considered in a group of subjects performing the same GXT protocol from previous studies in our laboratory (29, 47). Assuming that the mean decrease in aerobic capacity for female control subjects was similar to that observed in previous studies' male subjects with the same BR duration (19, 30) and that all assumptions of the test were met, only four subjects per group were required to detect a difference between the Con and Ex subjects' change in V̇o2peak at a power of 0.80, assuming that the countermeasure was 100% effective. If the countermeasure was assumed to be only 50% effective, seven subjects would have been required to detect a group difference after BR.
Pre- to post-BR peak exercise responses, plasma and blood volume, body composition, sprint performance, and body mass were compared using a two-factor repeated-measures ANOVA in which group was the nonrepeated factor and pre- to post-BR was the repeated factor. Submaximal exercise data were compared using a three-factor repeated-measures ANOVA in which group assignment was the nonrepeated factor and exercise intensity and BR were the repeated factors. Submaximal data from the third running stage were not included, since not all Con subjects were able to complete this stage after BR. Tukey's honest significant difference test was used post hoc to identify between- and within-group differences. When significant differences were observed, the effect size (Cohen's d) within groups was calculated as the mean difference from pre- to post-BR divided by the pooled standard deviation. Similarly, the effect size of the treatment was calculated as the difference between the groups' post-BR means divided by the pooled standard deviation.
To address sex effects of BR on V̇o2peak and sprint speed, we examined data previously published regarding identical male twins who participated in the same BR and countermeasure protocols (30). V̇o2peak and sprint speed were compared individually using a repeated-measures ANOVA in which sex was a nonrepeated factor and performance (V̇o2peak, sprint speed) was a repeated factor. In addition, unpaired t-tests were used to compare the percent change in V̇o2peak and sprint speed after BR between women and men. Furthermore, the relationships between pre-BR V̇o2peak and sprint speed and the BR-induced decrease in V̇o2peak and sprint speed in control (nonexercise) subjects was explored using bivariate correlations within sexes and with data pooled from both men and women. Finally, the effect of sex and bed rest on submaximal exercise (during the first running stage) RER, V̇e, and percent maximal HR (%MHR) were examined using a repeated-measures ANOVA in which sex was the nonrepeated factor and the submaximal exercise variable (RER, V̇e, and %MHR) was a repeated factor.
Statistical analyses were completed using Statistica 4.1 for Macintosh (StatSoft, Tulsa, OK) and SigmaStat v11.0 (Systat Software, Chicago, IL). Statistical significance was accepted a priori at P ≤ 0.05. Data are means ± SE, unless otherwise noted.
All but one of the Ex subjects participated in all of the prescribed exercise sessions. This subject experienced intermittent, unexplained knee pain with running such that she exercised at <1.0 BW for four exercise sessions early in the BR period and missed one session altogether. However, she exercised at 1.0 BW during the remaining countermeasure sessions.
Ex subjects were exposed to LBNP for 45.0 ± 0.4 min per session (Fig. 1). Subjects walked and ran for an average distance of 4.3 ± 0.3 km (2.7 ± 0.2 miles) per exercise session. Across all exercise sessions, the average LBNP was 51.1 ± 5.1 mmHg, which corresponded to a mean weight of 1.0 ± 0.1 BW. Most subjects completed their exercise at 1.0 BW for the first 2 wk. Four of seven subjects exercised at least three sessions at 1.05 BW. Only one subject exercised at 1.1 BW for an entire session.
Peak exercise responses.
Total test time was not different between the two groups before BR (Con: 16.2 ± 0.4 min; Ex: 15.9 ± 0.5 min). Total test time was reduced in the Con group after BR (12.8 ± 0.4 min, P < 0.001, d = −3.53) but was maintained in the Ex group (15.3 ± 0.7 min). Post-BR total test time in the Con group also was less (P = 0.001, d = −1.78) than in the Ex group. Similarly, the peak treadmill grade achieved was reduced in the Con group after BR (pre-BR: 6.9 ± 1.1%; post-BR: 0.0 ± 0.0%, P = 0.001, d = −3.50) but was not different pre- to post-BR in the Ex group (pre-BR: 6.0 ± 1.7%; post-BR: 5.6 ± 1.2%). The peak treadmill grade achieved in the Con group after BR was less (P = 0.001, d = −2.46) than in the Ex group. All of the subjects completed the three level running stages before BR, and the entire Ex group completed these stages after BR. However, only two of the seven Con subjects were able to complete the third running stage after BR.
Before BR, V̇o2peak was similar between the groups, whether expressed in absolute values (Fig. 2) or relative to body mass (Con: 44.0 ± 1.6 ml·kg−1·min−1; Ex: 44.4 ± 1.9 ml·kg−1·min−1). V̇o2peak was reduced by 16 ± 3% after BR in the Con group (36.8 ± 1.0 ml·kg−1·min−1, P < 0.001, d = −2.01) but not in the Ex group (−6 ± 5%; 41.6 ± 1.3 ml·kg−1·min−1). After BR, V̇o2peak was less in the Con group than in the Ex group (P < 0.05, d = −1.56). Peak HR, RER, V̇e, and rating of perceived exertion (RPE) were not different between groups pre-BR, and there was no change in any of these peak responses following BR.
Submaximal exercise responses.
Before BR, submaximal V̇o2 was similar between the groups during the walking and the first two running stages (15.3 ± 0.3, 25.8 ± 1.1, and 33.9 ± 0.7 ml·kg−1·min−1, mean for all subjects). Submaximal V̇o2 was not different after BR and was similar between groups (16.1 ± 0.3, 25.8 ± 1.0, and 33.6 ± 0.9 ml·kg−1·min−1).
Submaximal HR during walking and the first two running stages did not differ between the two groups before BR (Fig. 2). After BR, HR was greater during walking and running in the Con group (32 ± 4, 24 ± 5, and 12 ± 5%, respectively; P < 0.001, d = 1.45–1.57) but was elevated in the Ex group only during walking (8 ± 2%, P < 0.001, d = 1.57). After BR, submaximal HR during walking and running was higher in the Con group than in the Ex group (P < 0.05, d = 1.02–1.29).
Submaximal RER during walking and the first two running stages was not different between the two groups before BR (Fig. 3). After BR, Con group RER increased during the walking and running (P < 0.05, d =0.98–2.02). There was no effect of BR on submaximal RER in the Ex group. After BR, submaximal RER was greater in the Con group than in the Ex during walking and running (P < 0.05, d = 0.62–1.79).
Submaximal V̇e during walking and the first two running stages was not different between the two groups before BR. After BR, V̇e in the Con group was greater than pre-BR during walking and submaximal running (21 ± 6, 20 ± 4, and 22 ± 4%, respectively; P < 0.001, d = 0.55–0.87). V̇e was not different during any of the submaximal exercise stages pre- to post-BR in the Ex group. After BR, V̇e was greater in the Con group than in the Ex group during submaximal running (P < 0.001, d = 0.46–0.76) but not during walking. Pre-BR RPE was similar between the groups during walking and running and did not change in either group after BR.
All subjects successfully completed both sprints without incident during the pre-BR period, and there was no difference in sprint time between the two groups before BR. During post-BR testing, both the Con and Ex subject within the same twin pair fell during the first attempt. After BR, sprint time was not different than pre-BR in the Ex group (+9 ± 3%) but was increased in the Con group (+35 ± 9%, P < 0.05, d = 2.01; Fig. 4). The post-BR sprint time was greater in the Con group than in the Ex group (P < 0.05, d = 1.62). Similarly, average sprint speed was decreased in the Con group (−24 ± 5%, P < 0.001, d = −1.97) but not in the Ex group (−8 ± 2%, P = 0.07). The post-BR sprint speed also was slower in the Con group than in the Ex group (P < 0.01, d = −1.57).
Plasma volume, body composition, and resting HR.
Plasma volume was not different between groups before BR (Con: 2,457 ± 194 ml; Ex: 2,518 ± 157 ml) or after BR (Con: 2,379 ± 169 ml; Ex: 2,451 ± 115 ml). Plasma volume was unchanged in either group as a result of BR (Con: −2 ± 5%; Ex: −2 ± 3%).
Con subjects had a greater (P < 0.001) body mass than the Ex subjects before BR, and body mass was unchanged within either group as a result of BR (Table 1). The Con subjects had more lean tissue (P < 0.01) and fat mass (P < 0.001) but were not different from the Ex subjects with regard to percent body fat. Lean tissue mass did not change during BR in the Con subjects but increased in the Ex subjects (P < 0.05, d = 0.17). Fat tissue mass (P < 0.001, d = 0.19) and percent body fat (P < 0.01, d = 0.25) increased in the Con subjects during BR. Fat tissue mass tended (P = 0.06) to increase during BR in the Ex subjects, but percent body fat was not different after BR.
Supine HR was not different between the groups before BR (Table 2). In the Con group, resting HR was greater after BR both while subjects were supine (P < 0.05, d = 1.48) and standing (P < 0.001, d = 1.73). In the Ex group, supine HR was not different after BR, but standing HR was significantly greater (P < 0.05, d = 1.15). However, post-BR supine and standing HR were not different between groups.
Sex effects of bed rest.
There was a main effect of sex on V̇o2peak such that V̇o2peak was lower (P = 0.02) in the women than in the men [pre-BR: 51.9 ± 2.7 ml·kg−1·min−1; post-BR: 42.1 ± 1.8 ml·kg−1·min−1 (30)]. Both men and women experienced a decrease after BR (P < 0.001), but the effect of BR was not different between women and men (P = 0.23). Similarly, there was no difference between the sexes in the percent change in V̇o2peak relative to pre-BR V̇o2peak (women: −16 ± 3%, men: −18 ± 2%, P = 0.51). However, the absolute change in V̇o2peak (ml·kg−1·min−1) in the Con subjects as a result of BR was related to the pre-BR V̇o2peak in females (r = −0.79, P = 0.03), in males (r = −0.79, P = 0.02), and when both male and female subjects were considered together (r = −0.79, P < 0.001; Fig. 5).
There also was a main effect of sex on average sprint speed such that women were slower (P = 0.01) than the men [pre-BR: 22.4 ± 1.0, 18.6 ± 1.5 km/h (30)]. Both men and women experienced a decrease in sprint speed after BR (P < 0.001), but the effect of BR was not different between women and men (P = 0.62). Similarly, the percent change in sprint speed relative to pre-BR sprint speed was not different between women and men (women: −24 ± 5%; men: −18 ± 4%, P = 0.37). In contrast to V̇o2peak results, the decrease in average sprint speed (m/s) as a result of BR was not related to the pre-BR sprint speed in the women (r = 0.19, P = 0.68) or the men (r = 0.55, P = 0.16) or when the subject data were pooled (r = 0.03, P = 0.28).
Submaximal HR, expressed as a percentage of maximal HR, during the first running stage was not different between sexes before BR (P = 0.99, Table 3). Submaximal HR was increased in both women and men following BR (P < 0.001), but there was no difference between the sexes after BR (P = 0.60) and there was no significant interaction between sex and BR (P = 0.62). Although submaximal V̇e and RER were less in women than in men before and after BR, there was no interaction between BR and sex for either RER (P = 0.82) or V̇e (P = 0.81).
This was the fourth in a series of BR studies designed to determine the efficacy of LBNPex as a countermeasure to BR-induced deconditioning. This study is unique because it is the first to examine the efficacy of an exercise countermeasure to preserve V̇o2peak and sprint speed during BR in women. The principal finding of this study is that the combined LBNP and exercise protocol performed 45 min/day, 6 days/wk, maintains upright aerobic capacity and sprint speed in women equally as well as had been previously demonstrated in men (30). In addition, the percent decrease in V̇o2peak and sprint speed after BR in female subjects who performed no countermeasure were similar to changes previously reported in male twins who participated in the same testing protocol (30). Sex had no specific effect on either peak or submaximal aerobic exercise responses after BR, and the reduction in V̇o2peak was related to pre-BR fitness levels in both women and men. However, pre-BR sprint speed was not a predictor of the decrease in sprint speed as a result of BR.
Exercise capacity with BR.
Decreased exercise capacity after BR is a consistent finding (5, 6, 16, 37, 48). Although the previous models used to predict loss of V̇o2peak were developed with predominantly male subjects, the BR-induced percent reduction in aerobic capacity in our female Con subjects is similar to that which would be predicted after this duration of BR (6). Consistent with our previous studies with men, V̇o2peak was preserved in the Ex women who performed the LBNPex countermeasure.
During the first 30 days of BR, the decline in V̇o2peak has been suggested to be largely explained (∼70%) by a decrease in plasma volume in previous studies (5, 6). However, the lack of change in plasma volume in either female group suggests that factors other than plasma volume influenced the post-BR V̇o2peak. Although we observed a decrease in plasma volume in male subjects who underwent an identical study protocol, the post-BR decrease in plasma volume did not correlate with the reduction in V̇o2peak (30). Ferretti et al. (14) suggested that the largest factor explaining the decrease in V̇o2peak after a similar duration of BR is cardiovascular oxygen transport, primarily determined by decreases in cardiac output and red cell mass, and peripheral factors such as decreased mitochondrial volume density and oxidative metabolism potential (24).
A loss of plasma volume in women during BR has not been a consistent response across studies. Women were reported to have a similar decrease in plasma volume (−9.5 ± 1.4%) as men (−9.1 ± 0.9%) after 7 days of BR (12, 34) and following 14 and 17 days of BR in men (−11%) and women (−13%), respectively (10). In contrast, Fortney et al. (17) reported that the decrease in plasma volume was significantly less in women (−9.7%) than in men (−15.2%) after 13 days of BR. In ambulatory subjects, plasma volume fluctuates during the menstrual cycle, and the transient increases in plasma volume associated with increased estrogen levels also are present when subjects are subjected to BR (15). Because we did not specifically control for menstrual cycle with regard to entrance into BR, this source of variability may explain the lack of a plasma volume loss in our present results.
Elevated HR, V̇e, and RER during submaximal exercise after BR are also consistent findings in subjects who perform no countermeasures (6, 27, 29, 30, 47). These findings generally are concomitant with a decreased aerobic capacity, stroke volume, and plasma volume (6, 25, 27), although plasma volume apparently was not a contributor to these alterations in the current study. Similar to previous reports, when countermeasures were performed that maintain aerobic capacity (27, 29, 30, 47), the LBNPex countermeasure prevented the increase in HR, V̇e, and RER during submaximal exercise after BR. The countermeasure apparently preserves pre-BR levels of central (venous return and stroke volume) and peripheral factors (oxygen delivery and utilization and substrate utilization) important during upright submaximal and maximal exercise.
Maintenance of anaerobic capacity and power may be important in the event of emergency situations, such as egress from a damaged space vehicle after landing (47), but we are unaware of any other groups that have examined sprint performance after BR. In our experience, sprint time is increased consistently in men who performed no countermeasures during BR [15 days: +16% (47); 30 days: +24% (30)], and we observed similar results in the Con women in this study (+ 35%). In all cases, however, sprint time after BR was not significantly different from pre-BR sprint time when subjects performed the LBNPex countermeasure.
Sprinting performance is determined by the integration of multiple factors, including anaerobic and aerobic metabolic capacity, muscle power, muscle endurance, and neuromuscular coordination. Unpublished observations from these male and female subjects suggest isokinetic knee extensor strength (−19%) and endurance (−14%) were reduced in the Con subjects, which likely contributed to decreased sprint speed, but were maintained in Ex subjects (39). However, in both female and male identical twins (30), equal numbers of subjects in each group fell during post-BR sprint trials, suggesting that LBNPex does not completely protect posture control and neuromuscular coordination during dynamic, self-supported exercise. Difficulties with sprinting may be augmented as the duration of BR increases. Con subjects in our previous 15-day BR study appeared to run awkwardly after 15 days of BR, but no subjects fell in either group (47).
Resting responses and body composition.
Similar to our previous observations in men (30), supine (+27 ± 7%) and standing (+40 ± 9%) HR was higher after BR in women Con subjects, which is consistent with a decrease in aerobic fitness and a change in autonomic tone. Contrary to our previous report in men, however, standing HR also was greater (28 ± 8%) after BR than before BR in the female Ex subjects. The inability of the LBNPex protocol to protect standing HR responses in these female subjects is consistent with the general observation that women are more susceptible to orthostatic intolerance after BR (21) and spaceflight (50).
Subjects typically lose lean tissue (28, 40) and gain fat mass (28) during BR studies when dietary intake is maintained near pre-BR levels. Decreased lean body mass in these previous investigations is the result of reduced protein synthesis (13), but the female Con subjects in the current study did not decrease lean mass as expected. Subjects in previous studies were predominantly male and had higher pre-BR lean tissue mass. Also, Ex subjects in the current study surprisingly experienced a significant increase in lean tissue mass, even though aerobic exercise of this type is typically not expected to increase muscle mass (1).
Effect of sex and pre-BR fitness on post-BR exercise.
Having upright aerobic and anaerobic exercise data in women and men after identical BR protocols provides a unique opportunity to examine sex effects on BR deconditioning. Few studies have examined the effect of sex on the change in aerobic capacity after BR. Those that have been conducted consistently report that although the male subjects had higher pre-BR V̇o2peak values than their female counterparts, the loss of aerobic capacity expressed as a percentage of the pre-BR values was independent of sex in BR durations up to 20 days (6, 9, 10, 26). We also observed that percentage reductions in V̇o2peak following 30 days of BR in our control male (30) and female subjects were not different from each other (Fig. 6).
In addition, we observed a significant relation between pre-BR V̇o2peak and the absolute change in V̇o2peak after BR that was independent of sex. Taylor et al. (45) and Saltin et al. (38) were the first to report that subjects with a higher aerobic capacity had a greater absolute reduction in V̇o2peak than those with lower fitness. Several subsequent studies confirmed this hypothesis (6, 9, 10, 43), but this is not consistently the case (20). Interestingly, in two separate studies, Convertino et al. (9, 10) reported that there is a significant relationship between initial and post-BR V̇o2peak in men but not in women. The lack of relationship between pre-BR V̇o2peak and the BR-induced loss in women in these earlier studies might be related to the lower pre-BR V̇o2peak values measured in many of these subjects: 24 of the 33 women (73%) in those two studies (9, 10) had pre-BR V̇o2peak values <35 ml·kg−1·min−1. Female subjects in our study had pre-BR V̇o2peak values ranging from 36 to 49 ml·kg−1·min−1.
Sex also had no effect on the submaximal response to exercise after BR. Although submaximal V̇o2 was not different after BR, both women and men in the nonexercise control groups experienced similar increases in HR (women: +19 ± 4%, men: +16 ± 2%), V̇e (women: +20 ± 4%; men: +18 ± 6%), and RER (women: +6 ± 2%; men: +6 ± 2%). Similar elevations in HR and RER in men and women agrees with previous reports (9, 10), but the BR effect on submaximal V̇o2 (10) and V̇e (9) in women differs between ours and previous investigations. The disparate submaximal V̇o2 findings are not attributed to different levels of exercise intensity between our study (59 ± 3% pre-BR V̇o2peak) and a previous one [55–60% pre-BR V̇o2peak (10)], but could be related to differences in testing posture. Specifically, the slower pulmonary oxygen uptake and leg blood flow kinetics associated with supine compared with upright exercise (8, 31) might explain the discrepancies between these reports. Disparities between these studies in regard to V̇e might be related to the different exercise postures (7) or to the length of BR [30 vs. 17 days (9)], although others have reported increased ventilation after shorter BR durations (7, 27, 29, 47).
Similar to previous reports (32, 35), the women in this study had lower sprint speeds than their male counterparts in our previous study (30). Both men and women experienced similar reductions in sprint speed relative to pre-BR performance, but the absolute change in sprint speed was unrelated to pre-BR performance. Women might have been expected to experience greater decrements in sprinting performance than men, because others (46) have suggested a greater rate of muscle atrophy during long duration BR, but this was not the case. This lack of relation may be the product of the study population (none reported anaerobic exercise training before BR) or that the changes in potential contributors to sprint performance (i.e., creatine phosphate and glycogen stores, glycolytic enzymes, muscle strength and power, and neuromuscular coordination) after BR are independent of pre-BR anaerobic fitness. No studies are yet available to elucidate these relationships.
Summary and spaceflight applications.
Exercise countermeasures employed to protect crew health must be sufficient to promote and maintain high levels of function, such as aerobic and anaerobic fitness, in both female and male astronauts. As of 2001, women comprised 22% of the active astronaut corps (22). The LBNPex device and the countermeasure protocol employed in this and previous investigations (29, 30, 47) maintains upright exercise responses and sprint performance in both sexes. Upright exercise capacity after spaceflight is important operationally, because crewmembers must be prepared to perform an unaided egress from the space vehicle in the event of emergency after landing (29, 47). The results of the current study are relevant particularly to the women in the U.S. astronaut corps, because the V̇o2peak measured in our subjects is similar to the average value for female astronauts as reported by Harm et al. [36.5 ± 7.0 ml·kg−1·min−1, 2.19 ± 0.48 l/min (22)].
The LBNPex device is an exercise countermeasure that integrates protection for both the cardiovascular and musculoskeletal system. LBNPex effectively prevents spinal muscle loss and intervertebral disc deconditioning (3), reduces bone degradation (42, 51), and attenuates orthostatic intolerance (49) normally experienced with long-duration BR and spaceflight. Higher footward loading can be achieved using LBNPex than can be tolerated using the current flight treadmill/harness system [50–70% BW (4)], and LBNPex simultaneously imposes a gravity-like cardiovascular stimulus (2, 36). In our series of investigations, subjects have exercised comfortably within LBNP against loads up to 120% of their BW for as long as 40 min. In the event that this countermeasure is adopted by NASA, we expect that the same exercise prescription could be applied to both the male and female astronauts with similar levels of protection for the cardiovascular and musculoskeletal systems.
This research was supported by National Aeronautics and Space Administration Grants NCC 2-1133 and NAG9-1425 and by National Institutes of Health Grant M01 RR00827 to the UCSD GCRC.
We thank the test subject volunteers for participating; Michael Zeigler, MD, and the staff of the UCSD GCRC for efforts throughout the study; Chris Rogers, MD, Amy Langemack, Debbie O'Leary, PhD, Scott M. Smith, PhD, Kunihiko Tanaka, PhD, Jill Monroe, Eli Groppo, Maneesh Bawa, MD, Monique Keehan, Eva Brzezinski, RD, Erin O'Brien, RD, Alan Moore, PhD, and Mark Guilliams, MS, for their respective contributions to the success of this project; and Lesley Lee, Meghan Everett and Chris Miller for editorial comments.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2009 the American Physiological Society