Foreword
Introduction
Results
References
List of Acronyms
Abstract
Methods
Discussion
Acknowledgements
List Figures / Tables

Discussion

Main Findings

We investigated the physiological responses to 16 weeks of high resistance training on the development of strength, muscle hypertrophy and increases in bone mass, using two types of resistive devices; 1) a dynamic constant external resistance device (free weights) and 2) a dynamic variable resistance device, using elastomer. We hypothesized that the interim Resistive Exercise Device (iRED) would be a less effective training-modality to elicit increases in muscle strength and hypertrophy and bone mass compared to similar training with free weights. We further hypothesized that by increasing the training volume while using the iRED, the effectiveness of the training program would become more similar to the effects of free weight training.

The major findings of this study were that strength increased for all exercising groups. Training intensity progressively increased throughout the study for each group. Fig 17 shows a total iRED-training period overview for one subject, which clearly shows the increase in training intensity. Depicted are the average peak forces for each training session and the peak forces used in the statistical analysis, respectively represented as solid and open circles.


Figure 17. Example of the recorded peak forces for the squat exercise for all training sessions for one iRED3 subject (subject 18). There is a reduction in intensity after 7 weeks (sessions 22-24) of training and after 15 weeks of training (sessions 46-48), in accordance with the protocol. The figure further shows the three representative high-intensity days from which the statistical analyses were performed.

Furthermore, the lifted weight during the one repetition maximum (1RM) tests were increased for all exercises (with the exception for the iRED6 group for the squat exercise) at the end of the study. As shown by the DEXA data, all exercising groups exhibited signs of muscle hypertrophy in the legs after training, as the FW group and both iRED groups had increased lean body mass in the legs, consistent with increased thigh and calf muscle volumes, observed by the MRI data. No consistent changes in bone mineral density were observed during this study, with the exception of the increase in bone mineral density of the lumbar vertebrae L4 in the FW group.

This data confirms that the iRED is capable of producing increases in muscle mass and muscle strength, but the results do not confirm that the effect of FW training is superior to the effect of iRED training, since there were no differences between the groups. Furthermore, doubling the training volume did not change the training effects of the iRED. Furthermore, our data does not confirm that strength training increases bone mass, with the exception for bone mineral density in the FW group. We believe that this was due to the short duration of the training program. Changes in bone BMC and BMD are typically observed after longer durations of resistive training (Tsuzuku et al., 1998; Chilibeck et al., 1995).

Background

An interim Resistance Exercise Device (iRED) has been developed to provide resistance for strength training to combat the deleterious effects of weightlessness on the musculoskeletal system of the American astronauts aboard the International Space Station. This elastomer-equipped device was chosen as an interim resistive exercise device during the assembly phase of ISS, since severe restrictions limit the type of device that can be flown.

The unique environment inside the ISS imposes technical requirements on the exercise equipment. For obvious reasons, free weight cannot be used on board a space vehicle and the device has to rely on another source of resistance. Electrical braking to create resistance cannot be used for the first increments on board the ISS, because of the overall shortage of power on the station (Berg and Tesch, 1992). Another criterion that should be met is that the training apparatus should be lightweight, and small (Berg and Tesch, 1992), and the apparatus or its use should provide no vibrations to the ISS (Cavanagh et al., 1992).

Special Characteristics of Elastomer Training

The iRED consists of polymer strings that, when stretched, generate resistance. A number of exercise devices, particularly those designed for home use, also have elastic components, such as springs or bands, as their source of resistance. The resistance is provided by a standard elastic component, which is proportional to the distance it is stretched: F R = k* x, where F R is the resistive force, k is a constant that reflects the physical characteristics of the elastic component, and x is the distance that the elastic component is stretched beyond its resting length. The most obvious characteristic of elastic resistance is that the more elastic component that is stretched, the greater the resistance (Hughs et al, 1999). Further more, the forces necessary to stretch the rubber progressively increase when the rubber is stretched further.

As observed during the training sessions with the iRED and during the calibration procedures, the problem that arises is that every movement begins with low resistance and ends with high resistance. This is contrary to the force capability patterns of virtually all human muscle groups. When working against free weights, a set level of inertia must be overcome to move the center of mass (weight combined with the lifted body) over a range of motion requiring the greatest force capabilities to beat the beginning of the range of motion, with a substantial drop-off in force capability toward the end of the range of motion (Beachle and Earle, 1994).

In fact, maximal ground forces (as measured with a force-plate during the squat exercise) are generated at the bottom of the ROM, with a knee-joint angle of approximately 90-110 degrees. Peak forces with free weight training occur at the bottom of the range of motion (ROM). This peak force is caused by decelerating the moving object and overcoming inertia to change directions of the weight. This means that the greatest motor recruitment is required at the bottom range of the motion. The basic concept behind this is, that skeletal bones are actually poor biomechanical levers. Great forces are required to overcome the opposing inertia. While standing up from the lowest part of the range of motion, i.e. when the posterior borders of the thighs are parallel to the floor, great forces are required to overcome the opposing inertia. While the range of motion increases, the knee-joint angle also increases. This leads to a decrease in the moment arm. A biomechanical advantage occurs since both the moment arm decreases, and the moved weight now has a momentum. This reduces the necessity for high muscular forces at the end of the range of motion, i.e. while almost standing tall.

With the iRED device, however, the force exerted in the parallel position in the squat and deadlift is less than that of the peak force, which progressively increases as the movement continues upwards. This means that the peak force at the parallel position, while standing up, is low compared to the peak loads seen with lifting free weights. In this exercise device, the force curve is not linear over the whole range of movement. There is no need to overcome the inertia as seen with free weight, nor is there the necessity to decelerate the weight in the descending phase, because of the drop-off in weight, when using polymer. This means that the initial loads with the iRED are low compared to the loads seen with free weights.

However, the peak force during the iRED exercises was found not to be solely dependent on deflection only. Peak force was also affected by the acceleration with which the motion was performed, generating a peak force where the interaction of the deflection and acceleration was maximal. Depending on the rate of ascent peak forces could occur anywhere throughout the range of motion. Unpublished data by our lab suggests that peak force as seen with free weight lifting is best mimicked by iRED lifting when subjects perform the exercises at a fast pace in the ascending phase. For slow movements, peak force would occur somewhere in the upper parts of the range of motion.

Unpublished data from our lab has shown that the difference in strength curves between iRED and free weights, results in different concentric/eccentric ratios per training repetition. The ration is approximately 0.95 for FW lifting and approximately 0.75 for iRED lifting. This means that even if the same protocols (e. g. sets and repetitions/set) are used, FW lifting is superior in providing an eccentric component. Based on this general idea, a combination of concentric and eccentric training will lead to superior gains in muscle strength (Dudley et al., 1991).

Training Protocol

Our protocol was designed to elicit increases in strength, muscle and bone. A "circuit" approach was used that included training days for strength (6-8 repetition days) as well as muscle mass (10-12 repetition days).

A training program designed to increase body mass (bodybuilding) involves using moderate, typically ranging from 8 RM to 12 RM that allow the athletes to perform more repetitions that is typical of a strength training program, but heavy enough loads to elicit concentric and eccentric contraction failure (inability of the muscle to shorten or lengthen under control) within 6-12 repetitions. The higher overall training volume, coupled with a moderate relative intensity (expressed as a percentage of one-repetition maximum or 1RM), appears to be optimal for increasing muscle girth (Fleck and Kraemer, 197) and could possibly combat the muscle atrophy seen in space flight. The training protocol used in this study is based on the recommendations made by a team of on-site and off-site exercise science experts (the medical operations Bone mineral, Exercise Integrated Product Team, or BME IPT)

The training intensity for each subject was primarily based on 1RM strength. The 1RM value for the FW group was directly applicable for determining training intensity since load, or intensity and number of reps are inversely related (Beachle, 1994); the heavier the load, the fewer the number of repetitions that can be preformed.

The training intensity for each subject was primarily based on the 1RM strength value. The 1RM value for the FW group was directly applicable for determining training intensity since load, or intensity, and number of reps are inversely related (Beachle, 1994); for example, the heavier the load, the fewer the number of repetitions that can be preformed.

The Exercises

The choice of exercises was based on multiple-joint exercises that involve many muscle groups in one exercise, direct force vectors through the spine and hip, and exercises that allow greater absolute loads to be used in training. Based on Wolff’s law that states that stress or mechanical loading applied to the bone via the muscle and tendons has a direct effect on bone formation and remodeling (Chamay and Tschantz, 1972, Layne and Nelson, 1999), both muscles and bones would benefit from the same exercise.

We selected the squat, heel raise and deadlift as the training exercises since they form the cornerstones of the American inflight strength-training countermeasure program. They also primarily target the antigravity muscles (with their connective tissues) and the load-bearing bones to which they attach.

Squat

An integral part of most strength programs is the squat, which is known as a multi-joint, closed kinetic chain exercise (CKC) (Wilk et al., 1996). CKC's are believed to be superior to open kinetic chain exercise (OCK. CKC knee exercises are considered safer and more effective since they place less strain on the anterior cruciate ligament (ACL) and elicit hamstring co-contraction together with the quadriceps (Wilk et al., 1996). Furthermore, it is advantageous to strengthen the antigravity muscles in a similar manner to real (terrestrial) movements. The squats and heel raises were performed using a Smith Machine for the free weight group.

Deadlift

The deadlift exercise is also known as a multi-joint, closed kinetic chain exercise (CKC). As a compound exercise, the movement spans three joints with extension occurring at the hip, knee, and ankle joints, being a true whole body exercise and thus utilizing several large muscle groups. The exercise was used to develop the muscles of the lower back, back of the thighs and hips. Heavier weights would also have an effect on the forearm muscles, which were involved in gripping the bar, as well as muscles that were involved in trunk stabilization.

Heel Raise

The heel raise primarily targets the Triceps Surae muscles (e.g. the Gastrocnemius and Soleus muscle). The primary function of the Triceps Surae muscles is to extend the foot at the ankle joint, a movement known as a plantar flexion, to raise the body and flex the toes. Calf muscles are important antigravity and postural muscles.

Free weight Results

Changes in Strength

One-repetition max (1RM) refers to the heaviest weight that can be lifted with proper technique for only one repetition. This is the most common measure of weight-lifting strength. Although repetitions in weightlifting involve alternating concentric and eccentric contraction phases, a weightlifting test in this sense is primary a test of concentric contraction strength. Since one of the primary goals of in-flight training regimes is to maintain full functional muscular capacity throughout the full range of motion, the 1RM measure of muscle strength was used during this study. 1RM testing, however, requires an adequate lifting technique. We performed four 1RM sessions prior to the start of the study, so that the subjects could familiarize with the equipment and familiarize with lifting heavier loads than they were accustomed to (Baechle, 1994). Increasing the number of 1RM sessions to four would likely lead to a more accurate 1RM value. The rational for separating the trials by one week served a dual purpose. First, we wanted to enable the subjects to completely recover from the prior session and second, we wanted to avoid training the subjects in the baseline period.

1RM values significantly increased for the FW group in the first half of the training study. During the second half, fewer increases were observed. This corresponds well to the literature (Häkkinen et al., 1985). Initial gains are the highest due to neuro-motor learning, and then level off, when further strength increases rely more on muscle hypertrophy (Moritani and deVries, 1979; Chilibeck, 1998). It has been suggested that the contributions of either factor may be related to the complexity of the exercise (Chilibeck, 1998). A more neural adaptation might be necessary for a more complex exercise. This would explain why the 1RM for the deadlift also increased during the second phase of the study, whereas the easier exercises (squat and heel raise) did not.

We believe this effect to be partly due to training with the Smith Machine. When squatting and performing heel raises on the Smith Machine, the movement is along a guided line. Basically, that means the subjects could lean against the weight and had to only focus on lifting the weight. In this modality, it would be easier to increase the absolute amount of weight, early in the study.

The deadlift exercise was the most complex exercise, since it allowed movements in the x- z- and to a lesser degree in the -axis. In essence, the deadlift exercise is basically quite similar to the squat, but the weight is now held in front of the subject, instead of being placed on the upper back. The lower absolute loads for the deadlift exercise when comparing to the squat are believed to be the result of basically two factors. The first one, is that the deadlift didn’t make use of the Smith Machine, thereby increasing its complexity. The second reason is that, as a result of this, the subjects would break proper form more easily. In order to assure subject’s safety, the 1RM trials were halted when proper form was lost. Experience in performing the deadlift exercise caused a steady increase in the lifted loads.

Looking at the training data, a similar observation was made. The high-intensity (83%) training sessions were chosen, to represent increases in training intensity, because they would correlate the best with increases in strength and hence the 1RM values (Baechle and Earl, 1994). Training intensities increased similar to the 1RM values. The range of motion was kept constant during the training sessions, but the number of training repetitions declined during the second half of the training study. We believe this to be a consequence of the training protocol itself. After the mid-1RM, true maximal strengths were obtained and training intensities were changed accordingly. However, 83% of the 1RM value appeared to be too heavy to lift 8 times.

Changes in Muscle

Lean tissue (muscle) and total mass increased in the legs, which indicates that the free weight training program was capable of eliciting muscle hypertrophy at the end of the study. It is a well known fact that strength training increases muscle tissue (Tesch and Berg, 1990; Convertino, 1991; Dudley et al., 1991; Berg and Tesch, 1998) and from the limited MRI data, we concluded that that the muscle volume increased more for the thighs (9.2 ± 1.3%) than for the calves (4.2 ± 1.0%). Although the heel raise exercise was the exercise with the highest loads, the two other exercises targeted the thighs. It is thought that during a squat and deadlift, both the quadriceps muscles as well as the hamstrings muscles are activated (Escamilla et al., 1998). While the knee extends, the quadriceps muscles shorten and the hamstring muscles lengthen, but while the hip extends, the quadriceps muscles lengthen and the hamstrings shorten. The result is a simultaneous concentric and eccentric contraction at the opposite ends of each muscle (Baechle, 1994; Escamilla, 2000). Furthermore, electromyography (EMG), used to measure of the amount of muscle activity during a squat exercise in previous studies (Escamilla, 2000; Escamilla et al., 2001), showed that the squat elicited significant quadriceps EMG activity. However, using the Smith Machine eliminates the need to balance the weight, which will likely decrease the involvement of many synergistic muscles that are likely to be used during exercises that do not make use of the Smith Machine, like the deadlift in our study.

Observations of whole body composition changes where undetected, possibly because the magnitude of the changes was not great enough to induce a training effect on a whole body scale.

Changes in Bone

Because bone and connective tissue are thought to respond to mechanical forces that threaten the supporting structures of the contracting musculature, the principle of progressive overload; (progressively placing greater-than-normal demands of the exercising musculature), would apply to increase bone mass as well as improve muscle strength (Beachle and Earle, 1994; Turner, 2000). The cross-sectional area of a muscle is thought to reflect the total amount of force that the muscle is capable of exerting on the bones to which it is attached (Beachle and Earl, 1994). Since the forces of larger and better-trained muscles are greater, Wolff’s law suggests that related regions of connective tissue and bone should increase their mass and strength to provide a sufficient support structure for the hypertrophied muscles. This is most plausibly caused by a decrease in bone resorption, or an increase in bone formation, or a combination of both.

The observation that regional BMD increased in L4 suggests that bone turnover is increased coupled with an increase in bone formation. Indeed several studies have shown that serum concentrations of osteocalcin, a specific bone formation marker that indicates osteoblast activity, are elevated as a consequence of strength training (Fujimura et al., 1997; Menkes et al., 1993; Fiore et al., 1991).

Based on a previous study (McCarthy et al., 1994), in which the investigators did observe increases in spine bone mineral density (unpublished) after 16 weeks of training, it was expected to see some regional changes during this study as well. However, it has been suggested that bone takes considerably more time to remodel than muscle (Dalsky et al., 1988). After the 16 week training period, the only significant change was found in bone mineral density for the lumbar spine, lumbar vertebrae 4, to be more precise. The fact that there was not an increase in BMC for lumbar 4, or a change in bone area for L4, makes it likely that the variation among subjects in change in bone mineral content was larger than the variation among subjects in change in density. Although, not significant, at least 5 out of the 7 subjects for the free weight group were showing increases in both BMC and BMD in L1- L4, whereas the effects for the other groups were more diverse.

Although not statistically analyzed, we observed that the subjects with lower initial bone mineral density prior to the start of the training benefited the most from resistive exercise training. The individual increases were higher than the group means. Controversially, the effects of the training regime on subjects that started with a high (higher than average) initial high bone mineral density are a bit more diverse. We observed both increases as well as maintenances in bone mineral density. Some decreases were observed, but these were within the precision error of the DEXA machine (approximately 1%). The increases are particularly marked in the spine. The subjects with moderate to high initial bone mineral densities that did show a substantial increase were the subjects with the highest training loads. Although statistically not different, we observed individual regional lumbar spine bone mineral density changes up to approximately 9%. For the hip no such observations were made, which makes it likely that the spine responds faster to loading than the hip. On the other hand it is also possible that the exercises loaded the spine more than the hip.

The fact that the FW group increased in Lumbar bone mineral density in L4 may be due to the forces, experienced by the skeleton when lifting free weights. In lifting free weights, there are two main peak forces. The fist peak force occurs at the bottom of the range of motion to overcome the inertia. There is, however, a second peak force, which is basically an impact force. As one reaches terminal knee and hip extension, the body slows down, but the weight continues to be accelerated upward. The bar displaces slightly from the shoulders at the top of the movement and then falls back on the shoulders. This momentary displacement creates a high impact force distributed throughout the entire spine and hips and could be beneficiary for developing new bone tissue.

The finding of no apparent change in total spine BMD despite increased regional BMD also supports the hypothesis of a localized rather than a systemic effect of strength training (Menkes et al., 1993). On the other hand, it is also likely that the magnitude of the increase in total spine was too small to be detected. The percent increase in regional BMD represents such a small percentage of total BMD that it is perhaps not surprising that changes of this magnitude in total spine BMD could not be detected.

Most studies that report changes in bone mineral density are long-term studies, with the majority of the studies comprising at least one year (Tsuzuku et al., 1998; Chilibeck et al., 1995). We think that our study duration was too short to induce significant changes in bony tissue.

IRED3 Results

Changes in Strength

After 16 weeks of strength training, the iRED3 group had increases in strength for all the exercises, as evidenced by the increased loads. Moreover, we detected no differences in final 1RM values between the free weight group and the iRED3 group after 16 weeks of training. The time course of strength increase was, however, different when compared to the free weight group. The iRED3 group did not increase in strength in any of the exercises during the first 8 weeks of the training study, whereas the free weight group showed the biggest increase during the first 8 weeks. Several observations could explain why the iRED3 group showed a delay in training response.

First, the iRED3 group experienced lower training intensities at the beginning of the study. The training intensity for the iRED3 subjects were, like the free weight group, based on their 1RM values. However, for the FW group, the 1RM value was directly applicable for determining training intensity since intensity (load) and the number or repetitions are inversely related (Heyward, 1998). However, since the iRED has different loading characteristics, we assigned the training loads for the iRED as peak loads for the first three weeks. After the first three weeks, the peak load was adjusted accordingly so that the subjects were performing 6-8, 8-10 and 10-12 repetitions on the heavy, medium and light intensity days, respectively. However, the adjustments up to repetition maximum for each intensity day, took approximately 2-3weeks, which means that the protocols were not exactly the same during the first 5-6 weeks at the beginning of the study. During the remainder of the study, the training protocols were identical in that all the 3-set subjects were exercising to the extent where the last set could be completed with the appropriate number of repetitions for each training day.

Second, The iRED3 subjects experienced a reduced training stimulus with the iRED when compared to free weight training. Based on the concept of elastomer training, loading characteristics (total work and eccentric component) were less for the iRED3 subjects. Total work was also reduced for the iRED groups due to the limited ROM available for training.

A third factor that may have contributed to a lower training status could have resulted from hardware failures during the first phase of the training study. At times, we had to reschedule subjects for a new session, thereby missing the training stimulus for that day. On other occasions (approximately 5% of the total number of training sessions), the iRED3 subjects trained with CRES. On average the hardware failures occurred between session 18 and 22 and between session 31 and 33, approximately. CRES training provides lower loads during the lower part of the range of motion, contributing to yet another possible lowered training progression.

During the course of the study, the first phase subjects experienced hardware failures. The three subjects in the fist phase group trained, on average 5 times with the CRES system, whereas the phase two subjects never trained with CRES. By looking at the individual data, we found that training intensity for the deadlift was compromised for the first phase group. Training intensity for the second phase group increased 68.4%, whereas the increase in training intensity for the first phase subjects was only 34.3%. However, the impact on the 1RM was only about 10 percent (18.8% vs. 29.9% for the fist phase and second phase, respectively). A greater difference was observed in the 1RM for the squat exercise, where the first phase subjects increased a mere 8.3%. The second phase group increased approximately 21.4%. This data suggests that hardware failure and subsequent training with CRES reduced the training effect.

A fourth factor might be due to a testing artifact. Increases in strength are likely to be dependent on, and specific to, the mode of training, known as the concept of specificity (Fleck and Kraemer, 1997). Sale and colleagues (1992) noted that a (isometric) strength task that was different from the training task, but involved the same muscle groups did not lead to an increase in strength, even when muscles were hypertrophied. We cannot rule out the possibility that the iRED3 subjects did not show the same strength increase based on the fact that training and testing were performed on two different devices.

The FW group was highly familiarized with the equipment and lifting high loads during the four pre-training 1RM sessions, as well as during the training sessions, before the mid- and post-training 1RM sessions occurred. The iRED subjects, on the other hand, had likely developed a totally new neuromuscular pattern of lifting for the same exercises on the iRED. The different loading characteristics (muscle effort was trained at different knee and hip joint angles), specific for the iRED, when compared to lifting free weights, likely prevented a similar increase in free weight 1RM, as observed with the FW group. The iRED3 group did not train the specific neuromuscular pathways necessary for maximal free weight lifting.

For the heel raise, the basic movement during iRED3 training and free weight 1RM testing seems to be highly similar, where the basic movement only occurs at the ankle joint. However, there is a difference between the placing of the load, when comparing the heel raise performed on the iRED compared to the one with free weights. With the iRED, the squat harness evenly distributes the heavy load across the shoulders of the subject, without high pressure points. When lifting heavy weights on the Smith Machine, however, the total load is placed on the upper parts of the Trapezius muscle. We cannot exclude the possibility that there is a psychological effect for the iRED subjects, who are apparently not accustomed to this localized loading. There is a possibility that the subjects did not push themselves as hard as did the free weight subjects who were accustomed to these higher pressures.

To test whether an artifact could have prevented an increase in free weight 1RM value for the iRED subjects, we attempted to perform 1RM tests on the iRED as well. The device proved to be unsuitable for this purpose. The major factor was the limit of approximately 136kg of resistance that could be generated by the iRED. The capacities of the devices were maximized by a majority of the iRED subjects.

During the training sessions, if necessary, we augmented the iRED with the TVIS (treadmill vibration isolation system) bungee cords. For the heel raise exercise, the iRED and the TVIS bungees had similar loading characteristics, since the range of motion was only around 12cm. We could lower the resistance of the iRED and augment it with a bungee to get to a desired peak force. However, for the deadlift and squat, the addition of the bungees proved of little use in the lowest part of the range of motion, since the bungees were not being stretched until half way through the ascending phase. At this point, the biomechanical advantage was increased so that some of the subjects were able to stand erect with several bungees attached. The problem arose when subjects were not able to stand up with the bungees. We had to lower the resistance on the iRED, which makes lifting less difficult in the lower part of the range of motion. A comparison between the different subjects became impossible since the loading characteristics were now far from equal. After the mid-training 1RM session, we abandoned the attempt.

Based on the fact that the training intensity was recorded for the iRED in peak load and the intensity for the FW group was recorded as the total amount of lifted, a direct comparison in changes in the intensities is impossible. However, the iRED subjects had to also learn to lift on the iRED and they had to control the weight to a greater extent during the squat and heel raise exercise compared to the Smith Machine-based exercises. This makes it likely to assume that the absolute amount of weight could not be increased to the same extent as for the free weight group throughout the first half of the study, as evidenced by the training data that showed no increases in training intensity for the squat and heel raise exercises.

The deadlift exercise on the iRED is the exercise that is most comparable to free weight lifting, since they both allow movements in x-, y- and z-axis. The deadlift on the iRED was actually less difficult than the deadlift for free weights. The way that the iRED cables were positioned at the bottom of the iRED canisters created a dynamic pivot to which the subjects could "lean". However, the loading characteristics were still very different.

Apart from this, the limited range of motion that was available for the whole body exercises on iRED was especially noticeable for the deadlift. The FW group trained with a greater range of motion (55.8 ± 0.9) than the iRED3 (50.4 ± 1.0) and iRED6 (51.0 ± 0.9) group. If necessary, French clips were added in line with the iRED cables in order to prevent the cable to be extended beyond the maximal deflection of 56cm. For the taller subjects, this meant that the length of the cable stroke did not allow these subjects to reach parallel at the bottom of the ROM, because the cable stoppers touched the iRED canisters. As a consequence, the range of motion was less for the iRED groups. Training with a reduced range of motion likely affected the results of the maximal strength test, where a full range of motion, especially at the lowest part of the range of motion, is needed to successfully complete the exercise.

Changes in Muscle

Lean tissue (muscle) and total mass increased in the legs, which indicates that the iRED 3-set training program was capable of eliciting hypertrophied muscles at the end of the study.

The DEXA scans revealed no significant differences between the iRED3 and the FW group. Although not statistically tested, the MRI analysis revealed smaller increases in thigh muscle volume (4.2 ± 1.0% vs. 9.2 ± 1.3%), but the calf muscles hypertrophied more (6.1 ± 0.7% vs. 4.7 ± 1.0%). The reason why we think that iRED training affects the calf muscles more than free weight training is based on the assumption that the total center of gravity is shifted forward while performing the exercises on the iRED. A study by Wretenberg et al. (1996), postulated that shifting the center of gravity forward during a squat, likely increases the need for compensatory (eccentric) ankle plantar flexion. If this holds true, than performing the exercises with the squat harness would lead to increased activity in the Gastrochnemius and Soleus muscles.

Changes in Bone

There were no observations that support our hypothesis that the IRED was capable of producing increases in bone mass. There were no differences in total or regional bone mass or density between the iRED3 and the free weight group at the end of the training study. We assume that the training duration was too short to detect differences in bone mass or density between the two groups. Typically, increases in bone mineral densities (or mass) are found in longitudinal, or long-duration ( ³ 1 year), or cross-sectional studies (Tsuzuku et al., 1998; Chilibeck et al., 1995).

Based on the concepts of elastomer training, we speculate that marked differences will be observed between the two groups when training duration is extended for longer durations. Based on unpublished force plate data, we assume that the stresses applied by the active muscles on the bones are likely to be higher with Free weights than with iRED. Torque is created around the hip joints by the muscles pulling the femur around the joint to extent the hips. Based on the fact that peak forces are higher in the lower part of the range of motion, the torque would be much higher around the hip with the Smith machine. This is extremely important since the femur is the greatest load bearing bone in the body. With the iRED the peak forces occurred at the end of the range of motion, i.e. while almost standing up, the muscles have a biomechanical advantage. The high muscular pull needed to lift a free weight during a squat and deadlift is not needed for the iRED lifting.

The fact that the iRED subjects trained with the squat harness eliminates the impact force observed while training with the Smith Machine bar, and could have important consequences for increases in spinal bone density after longer durations of training.

iRED6 Results

Changes in Strength

We hypothesized that doubling the training volume for the iRED6 group, while keeping the training intensities equal to the other groups, would provide more similar results when compared to the FW group and an enhanced training response when compared to the iRED3 group.

Apparent from the 1RM data is that the iRED6 subjects did not show any increases in 1RM strength for the squat exercise, however, increases (in percent) in heel raise and deadlift 1RM strength were almost identical to the FW group.

The major differences between the iRED3 and the iRED6 were that the percent increase for the iRED6 in heel raise 1RM was larger than observed for the iRED3 group (23.0 ± 6.6% vs. 16.9 ± 2.6%). Similar, the increase in deadlift 1RM was larger for the iRED6 group when compared to the iRED3 group (23.0 ± 7.3% vs. 25.1 ± 5.0%). Like the FW group, the iRED6 group showed an increase in deadlift 1RM after 8 weeks of training. This data suggest that in fact a doubling of training volume may have a small beneficiary effect on strength development for both the heel raise and the deadlift exercise. However, no increases in squat strength were observed for the iRED6 group.

There is however a discrepancy between the changes in 1RM strength and changes in training intensity for the iRED6 group. The iRED6 group showed increases in squat and deadlift training intensity, but not in heel raise intensity, when comparing the first and last representative time point. The deadlift also showed an increase during the first half of the study, but there were no differences in these exercises between the two iRED groups. Since intensity and number of repetitions are inversely related, it is likely that the iRED6 group did not increase as much in training intensity, as did the iRED3 group. The number of sets likely prevented significant increases in training intensity during the heel raise exercise. Why this did not occur with the squat and deadlift exercise may be related to the force characteristics of the iRED. Because of the small range of motion during the heel raise, the force was greater and more constant throughout the deflection than with the other exercises. For the other exercises it was easier for the subjects to increase the peak load, while still successfully lifting the load. Another factor that may be related to this is the total amount of muscle mass used in the exercise and the onset of muscle fatigue. The protocol was such that the last set had to be with the same weight as the starting weight. While 3 sets of heel raises might be enough to be able to increase the load, 6 sets of the same exercise could well prevent an increase from happening. In this sense, the training of the iRED6 group would more specific for increasing muscular endurance rather training for increases in strength.

However, we cannot rule out the possibility that the injuries, allegedly caused by overtraining, could have also had a negative impact on the training progressions made by the iRED6 group.

Changes in Muscle

Muscle mass increased for the iRED6 group as well but the increases were not superior when compared to the iRED3 group (4.3 ± 0.6% vs. 3.1 ± 05%) or the FW group, which increased 5.4 ± 1.2%. Muscle mass had also significantly increased at the end of the training period. Based on the MRI data a similar observation was made: thigh muscle volume increases more than that of the iRED3 group (6.7 ± 2.3% vs. 4.7 ± 1.0%), but is still less than the increase observed in the FW group (9.2 ± 1.3%). There are, however, no differences in calf muscle volumes between the iRED groups.

Combining the MRI, the DEXA, and the training data, we suggest a causal relationship between these factors. The iRED6 subjects did not increase their training intensities, whereas the other groups did, even though 1RM values did increase. Apparent from the data is that the extra three training sets did not contribute to extra gains in muscle volume, or strength. A study by Hass et al., 2000 came to the same results, increasing training volume over a 13 wk period did not lead to significantly greater improvements in fitness (muscular strength) for adult recreational weight lifters.

Changes in Bone

The iRED6 group did not support our hypothesis in relationship with increases in bone mineral density or bone mineral content in any of the examined areas. Doubling the training volume had no beneficiary effect on any of the sites. This observation is confirmed in the literature. If the magnitude of the load, the rate of force application, or both are sufficient there is typically not a need to perform more than a total of 30 to 35 repetitions (3-4 sets), as a greater volume of loading is not likely to improve any additional stimulus for bone growth (Beachle and Earle, 1996). When looking at the individual data it becomes clear that a subject with very low bone mineral density will benefit the most from the iRED training.

Limitations of the Study

Subjects vs. Astronauts

By limiting the subject pool to untrained male subjects we hoped to maximize the training effect and to reduce the variability in the training response. However, there may be a difference between training adaptation between individuals with a low initial physical or muscular fitness and individuals with high muscular fitness levels (Häkkinen, 1985). Astronauts are typically assigned a 12-month training program, including high intensity resistive exercise before a long duration space flight. Furthermore their goal is to maintain musculoskeletal function during space flight rather than to increase muscular strength as in the present study.

Increased Risk of Injury

While doing the exercises on the iRED, the cables coming out of the iRED canisters pull the subjects towards the center of the base plate, creating a dynamic "pivot" point on the base plate. The squat harness was engineered to distribute exercising load evenly across the shoulders. Specifically, the iRED harness is designed from modified American football shoulder pads, with metal braces across the top to prevent the pads from compressing under high loads on the subjects chest. Off each corner of the brace, a nylon cord is attached at anterior and posterior points and drops down along the mid-axillary line of the subject. At the distal end of the cord, a pivot pulley attaches the cord to a clip witch clips into the iRED cords. Due to this configuration the line of force application is different than the line of force application when lifting free weights. Most noticeably, the pull was shifted anterior, i.e. towards the chest. The subjects forced themselves into increased trunk flexion, in order to do a successful lift. However, by doing so, this likely created more force over the knee joint than over the hip joint. Unfortunately, the knee is more prone to injury than the hip because of the knee’s location between two long levers (the femur and the tibia). Of the various components of the knee, the patella and surrounding tissue are most susceptible to the kinds of forces encountered in resistance training. The patella’s main function is to hold the quadriceps tendon away from the knee axis, thereby increasing the moment arm of the quadriceps group and its mechanical advantage. Because of the squat harness, we suspect that patellar tendon and the anterior cruciate ligament (ACL) encountered higher forces during lifting on iRED than lifting free weights. This manifested itself by patellar tendonitis in some of our iRED subjects, which was characterized mostly by tenderness. Of the two iRED groups, the iRED6 subjects suffered most from tendonitis and we believe this to be a direct cause of overtraining as a result of the increased training volume encountered by this group. As a consequence, two iRED6 subjects were unable to complete the post-training 1RM test and were subsequently excluded from the analysis.

Overtraining might have a negative impact on training compliance of the iRED6 group. On average, this group had the least amount of training sessions per week and the least amount of total training sessions at the end of the study. This comprised training compliance likely had a negative impact on strength development and we assume that this might be a reason why we didn’t observe the increases in strength, muscle hypertrophy and bone mass in this group as we hypothesized.

This finding has a direct impact on the way astronauts onboard the ISS should use the iRED during squats and deadlifts. As much weight as possible should be placed around the hips. Not only will this prevent knee injuries, but his also puts strain on the Trochanter Major of the hip, which is prone to decreases in bone mineral density during space flight (Oganov, 1997). A new squat harness is currently being tested that shifts the load more posterior. The data implies that 6 training sets are likely to induce signs of overtraining. This might be even more so during space flight. Astronauts aboard ISS work out six times a week, instead of the three times that the subjects trained during our study.

Another risk of injury could manifest itself in the lack of force feedback on ISS. No exact calibration procedures can be carried out while on board of the station. Calibration becomes critical when shifts in force production occur as a consequence of wear to the polymer spokes as the number of repetitions increase. A potential misbalance between the iRED canisters could occur and this might have a negative effect.

Shift in force production also occurs with increasing temperatures. Since there is no effective heat transportation in the node where the iRED is stored, the heat accumulation might have negative effects on the loading characteristics as well. Lack of convection in microgravity may result in greater heat accumulation inside the iRED canisters

Resistance Training in Micro-Gravity

Conclusions from our data reveal that training with the iRED will lead to increases in muscular strength and lean muscle tissue. However, since ambulatory ground-based studies suffer from incomplete validation and discrepancies with space flight, it is not completely possible to extrapolate the data gathered in our ground-based study to a microgravity environment. This leads to the question whether a protocol that has been proven to increase strength and muscle mass on Earth, will also prove efficient in a microgravity. One of our primary hypotheses was that resistive overload training protocols in space would provide similar increases in muscle volume and strength and bone mineral density when compared to traditional weightlifting programs in terrestrial conditions. This, however, remains to be proven.

During our ambulatory ground-based study the subjects were always influenced by the constant gravitational pull on the body. Antigravity muscles worked against this force during their daily lives as well as during the training sessions. While weightlifting, skeletal muscles do not solely work to overcome the gravitational pull on the weights that are being lifted; the subject body is part of the lifted weight as well. Gravity, acting on the body is, however, absent in microgravity. The gravitational pull is reduced to one-millionth of that experienced on Earth, with the resistance now being entirely dependent of the device used for resistance exercise. With the squat and deadlift exercise the strongest pull on the bones (femur) is likely to be where the biomechanical disadvantage is the highest, i.e. in the parallel position. As observed in unpublished data from our lab, this holds true for the Free weights, where the (ground) peak forces are indeed observed in the bottom of the range of motion. Peak force with iRED is dependent on acceleration as well as deflection. However, the biomechanical disadvantage is likely to be the highest in the parallel position as well, simply because the momentum arm is maximal in this position.

In microgravity, there is likely to be virtually no load in the lower part of the range of motion provided by the iRED and no gravitational pull is being exerted on the astronaut. The lack of gravity would result in little torque about the hip at the bottom of the movement; subsequently dramatically lessen the muscular pull on the bones needed to stand up from the parallel position.

The limitation of 136kg for the iRED had already an impact during our study for the heel raise exercise. For a majority of the iRED subjects we had to augment the resistance with bungees. Each bungee has a limited capacity of 48kg at a full 94cm extension. However, the range of motion is only around 12cm with the heel raise, leaving only a fraction of the total capacity that can be used for resistance training. Elimination of the terrestrial gravitational background strain would call for even more augmentation of bungees. With a maximum of six bungees some astronauts might max out these capacities.

Eccentric Contractions

Eccentric muscle actions provide force development during muscle stretching, which is natural occurrence in terrestrial conditions. If total force development is critical to preserve muscle size and strength, eccentric resistance provides greater force development than concentric actions (Fleck and Kraemer, 1997). With each repetition, a dynamic approach results in concentric and eccentric muscle actions, respectively (Berg and Tesch, 1998; Fleck and Kraemer, 1997). Exercises that use both eccentric and concentric (and isometric) muscle actions are thought to be superior in inducing training effects than exercises that target concentric muscle actions only (Dudley et al., 1991).

In exercise programs developed for space flight, where a gravitational "back ground strain" is absent, the capability of providing eccentric muscle actions (during the descending phase) is solely dependent on the design and implementation of the unique equipment that can generate eccentric resistance without the help of a gravitational pull. Elastic cords and spring-based equipment can provide resistance during both concentric and eccentric muscle actions, however, the eccentric load profile is very different than the gravitational pull acting on individuals in terrestrial conditions (Berg and Tesch, 1998) as the elasticity "cushions " the eccentric force when compared to FW lifting. Based on unpublished data from our lab it becomes obvious that the eccentric to concentric ratios were lower (adjusted for body mass) than those observed during the DCRE. It is to be expected that in micro-G, these forces will reduce to dramatically low values.

This has a serious impact on the expected adaptation of training on the iRED on board the ISS. The fact that iRED was capable of increasing muscular strength and mass in our study does not provide an answer to whether the iRED is also capable of producing increases muscle strength and mass during space flight.

Limitations of the Protocol

Another limitation of the study was believed to be the training protocol itself. Although increases in muscle strength were observed after 16 weeks of training the majority of the increase occurred during the first half. This was believed to be the result of an early training adaptation. After the first half, most of the neuromuscular adaptation had occurred. Although the amount or repetitions and the load that can be lifted are inversely related, our subjects trained to the same level of muscular fatigue every training day. This leads us to believe that the training paradigm actually did not consist of a high-, medium- and low-intensity days, but that all days were high-intensity days, only organized differently.

This was most visible in the FW group after the mid-1RM values were established. Although, the subjects were able to perform the exercises with the same range of motion, the number of training repetitions decreased indicating that the training was too intense. The reason why this didn’t happen for the iRED groups is that their mid-1RM did not change significantly. In accordance with the protocol, the training intensities were not increased. However, as iRED subjects became stronger, training intensity was increased, leading to a significant increase in training intensity at the end of 16 weeks for the iRED3 group. No changes were detected in the Heel Raise strength for the iRED6 group, which is likely affected by the small subject number (n=5). The small increase in 1RM values during the second half of the study was thought to be mediated by increases in muscle hypertrophy.

One way to study the effects of training with the iRED in microgravity is during bed rest. Plans a currently being developed to conduct this kind of research. During this study the longitudinal gravitational "background strain" will be minimal, which will give us the opportunity to test the effects of training with the iRED when resistance becomes solely dependent on the hardware.

Limitations of Available Data

In order to compare our current data to prescribe training regimes to be used during space flight, we have to keep in mind that the reliability of evaluating the effectiveness of in-flight exercises so far has, for several reasons, is difficult to demonstrate from a scientific point of view.

A number of uncontrolled variables limit the conclusions that can be drawn from the in-flight data. Adequate scientific research requires a well-documented, reproducible, controlled experiment setting and availability of sufficient subjects and longitudinal research to detect any problems that might occur only after a distinct period of time (Convertino, 1996). So far, this kind of research has been extremely difficult to perform during actual space flight (Bishop, 1993).

The first factor that comes into play is the small number of subjects (the astronauts) and the pre-training training status of the astronauts before launch (Edgerton et al., 1995; Dambrink et al., 1989; Greenleaf et al., 1989). The fact that astronauts change from a state of high activity to one described as sedentary during most of the flight time will result in deconditioning, in addition to any specific effects of microgravity. Each crewmember carries out an individualized and complex series of flight assignments (including exercise) requiring variable amounts of neuromuscular activity (Edgerton et al., 1995). This lack of standardization, coupled with the limited data shows us that the responses of the crewmembers are highly individualized. It is therefore extremely difficult, if possible at all, to draw conclusions from previous space flights that exercise, per se, was and will be an effective countermeasure for the functional and anatomical changes of the musculoskeletal system associated with prolonged weightlessness.

One factor, which may have played a critical role in preventing maintenances in both bone and muscle mass during space flight is diet. Nutritional factors haven’t yet been established for space flight. Adequate diets could be provided to maintain body mass and minimize mineral excretion. However, it is known that dietary intake of the astronauts decrease, resulting in a severe negative energy balance (Stein et al., 1999a, 1999b) as well as negative calcium balance (Smith et al., 1998).

Typically, astronauts will perform aerobic exercises (30 min, 6 days/week) in combination with lower and upper body resistive exercises (1 ½ hour, 6 days/week). It is unknown at this point how the combination of iRED resistance training, the aerobic training protocols and dietary intake will affect the musculoskeletal responses.

To date, the US Space Program recommends, but does not require specific exercise-training regimes for astronauts to use on the ground to attain and/or maintain a high level of physical fitness. NASA has never formally identified job-related personal fitness targets, determined the level of crew performance capability or established minimal standards for muscular strength and bone mineral density. Pre-flight recommendations for exercise may actually turn out to be counterproductive for space flight, when dietary intake is insufficient. The establishment and maintenance of an elite physical fitness status may increase the metabolic requirements of an individual and make that individual more susceptible to changes that occur under conditions of reduced biomechanical loading if these requirements are not met.

Future Developments

Based on the data provided in this report, the iRED is capable of inducing strength gains and muscular hypertrophy. Increasing the training volume by increasing the number of sets will not bring about extra gains in training effects, but does increase the risk of overtraining and injuries and it is therefore not recommend for space flight. We think that the duration of the study was too short to elicit consistent changes in bone mineral content and density.

Further research is necessary to test the effects of factors that were not controlled in this study, i.e., diet, daily life-style (including sports). Optimally this study would be repeated with a larger subject population for a longer period. The study would also mimic the training protocols performed on ISS. Astronauts will train both upper and lower extremities, six days a week. Biomechanical analyses, combined with bone modeling studies would provide detailed information about the exercises and their impact on the musculoskeletal system.

Since it is highly unlikely that there is an all encompassing type of exercise capable of counteracting microgravity-induced deficiencies in the cardiovascular, vestibular, neuromotor and musculoskeletal systems studies should be undertaken to take advantage of the technologies currently being developed to produce forms of artificial gravity via the use of exercise-cycling (DiPramprero, 1994 and DiPramprero and Antonutto, 1997) and other devices that have the potential to transiently generate g-forces known to positively impact on these systems, while simultaneously providing an exercise stimulus mimicking that currently seen with most conventional devices.

For a Resistive Exercise Device this implies that the current design should be "up-graded" towards, or replaced by, a device that will meet the requirements originally set by the Human Research Facility. A next-generation Resistance Exercise device should have a higher maximal capacity, should allow a more functional ROM, should provide more force during the lower parts of the range of motion and should allow higher eccentric forces during the descending phases of the exercises.

In order to assure astronauts safety, the line of force should be directed more posterior during the squat exercise by for example using a Smith Machine-like configuration, or by improving the design of the currently used squat harness.

Furthermore, an instrumented RED with both manual and software control (including uplink and downlink) capabilities should be developed to monitor angles, velocities and forces/torques as a function of time during a specific exercise, to safely and adequately adjust resistive parameters, if needed.


Go to the INDEX page