Exercise will almost certainly play an integral part in minimizing the bone mineral loss and muscular atrophy that occur during spaceflight. It has been hypothesized that an effective exercise regimen can be developed to elicit loads on the lower extremities and require muscle actions which resemble those encountered on Earth (Cavanagh, 1986;​ Convertino and Sandler, 1995).

The Penn State Zero-Gravity Simulator (PSZS Davis et al. 1996) is a device which suspends subjects horizontally from multiple latex cords, with each cord negating the weight of a limb segment. A treadmill mounted on the wall under the PSZS enables subjects to run in simulated OG. Subjects wear a harness to which a number of springs, which provide a gravity replacement load, are connected. The opposite end of each spring is connected to the side of the treadmill. During exercise, astronauts currently wear a similar harness in which the spring tethering load pulls at both the waist and shoulders (Greenisen and Edgerton, 1994).

Ground reaction forces, muscular activations, and joint angles of the left leg during overground, treadmill, and fully-loaded zero-gravity simulated locomotion were assessed in order to gain insight into the effectiveness of the exercise regimen used by NASA to prevent the muscular atrophy and bone demineralization which occur in weightlessness.

There were three hypotheses to this research. It was hypothesized that there will be no differences in peak ground reaction forces and peak loading rates between overground gait and gait in the full bodyweight loaded conditions in the ZLS. A second hypothesis was that that there will be no differences in hip, knee, and ankle joint positions between walking or running overground, on a standard treadmill, and in full bodyweight loaded conditions in the ZLS. The third hypothesis was that the muscular activations, as a percentage of maximal voluntary contraction, will be similar between walking or running overground, on a standard treadmill, and in full body-weight loaded conditions in the ZLS.

Methods

Sixteen individuals (age 22.9±6.9 yrs, height 178.1±6.68 cm, and mass 72.8±5.8 kg) were studied at two speeds (1.35 m/s walking and 2.68 m/s running). Data were collected during overground locomotion, standard treadmill locomotion, ZLS "shoulder springs only" locomotion, and ZLS “waist and shoulder springs" locomotion. Ground reaction forces were assessed using a force plate mounted within the ZLS treadmill belt and another mounted in the laboratory floor. Angles of each subject's left ankle, knee, and hip were measured with electrogoniometers. Electromyographic data were collected of each subject's left tibialis anterior, gastrocnemius, rectus femoris, vastus lateralis, biceps femoris, and gluteus maximus muscles. Spring tensions in the Subject Load Devices (SLDS) were measured in the ZLS conditions using load cells mounted at the spring attachment sites on the treadmill.

Results and Discussion

The motion of the ankles and hips of subjects in the ZLS were representative of those observed in 1G locomotion. However, the knee was significantly more flexed in the ZLS conditions than in 1G. Maximum knee extension in stance was -0.84+1.85° in overground walking (negative value indicates knee extension), -2.44+1.31° in treadmill walking, 3.23±1.32° in "shoulder springs only" walking, 1.76±1.14° in "waist and shoulder spring" walking, 6.95±1.75° in overground running, 11.44+0.98° in "shoulder springs only" running, and 12.01±1.00° in "waist and shoulder springs" running. This greater degree of knee flexion was an attempt either to reduce the discomfort felt by the subjects at the shoulders and hips due to the tethering load, or to provide the subjects with a longer flight time, which was shorter in the ZLS.

The large magnitude of the fluctuation of subject load was the most notable finding in the study of the spring tension. The fluctuations were as follows:​ 13.52+1.70%BW in "shoulder springs only" walking, 21.97±1.49%BW in "waist and shoulder springs" walking, 17.89±1.25%BW in "shoulder springs only" running, and 36.83±1.30%BW in "waist and shoulder springs" running. The differences between the "shoulder springs only" condition and the "waist and shoulder springs" condition can be attributed to the fact that twice as many springs were used in the latter condition. The fluctuations during walking were predictable based upon normal 1G walking center of mass (COM) oscillations (Cavagna et al., 1963). However, the fluctuations during running were less than predicted based upon 1G COM oscillations measured by Morgan et al. (1990). The COM oscillations during running were less in the ZLS than in the study by Morgan et al. (1990) because the subject load, or the "apparent gravity" was greater than 1G during the flight phase of the ZLS trials because the springs were more stretched during flight that during standing. In the ZLS trials, the peak to peak oscillation of the COM was only 6.4cm, not the 10 cm measured in 1G by Morgan et al. Because of the decreased flight phase duration and the fact that subjects ran with a greater degree of knee flexion during stance in the ZLS than in 1G, the average subject load was 96.37±1.59%BW in "shoulder springs only" running and 88.95±1.66%BW in "waist and shoulder springs" running.

The maximum active ground reaction force peaks were significantly larger in the overground conditions than in their ZLS counterparts. The active force peak values were as follows:​ 124.37±7.10%BW in overground walking, 87.84+5.23%BW in "shoulder springs only" walking, 81.14±4.51%BW in "waist and shoulder springs" walking, 240.61±7.04%BW in overground running, 180.04±3.77%BW in "shoulder springs only" running, and 159.75±3.97%BW in "waist and shoulder springs” running. Two distinct mechanisms may be responsible for these results:​ the increased amount of knee flexion during stance in the ZLS and the fluctuation of subject load due to the normal mechanics of running. McMahon et al. (1980) noted lower maximum active force peaks when the subjects in their study were told to "groucho run", or run with an increased amount of knee flexion. Also, when the ground reaction forces of this study are normalized to subject load instead of body weight, the forces in all conditions appear remarkably similar.

The loading rate was significantly different between conditions (p<0.05), with overground running resulting in a lower loading rate, at 40.60 ± 2.85% BW/sec, than the ZLS conditions, at 51.95±1.53 for the "shoulder springs only" condition and 51.81 ± 1.57 for "waist and shoulder springs" condition.

The magnitude of the passive force peak was not significantly different between conditions, but the passive force peak occurred earlier in the running cycle in the ZLS conditions than in the overground condition (15.01±1.11% of cycle overground, 10.64±0.59% of cycle "shoulder springs only", and 10.07±0.60% of cycle "waist and shoulder springs"). The flight phase impulse (flight time* gravity or the subject load) for the overground data was an average of 93.2% of the flight impulse obtained from the "shoulder springs only" condition and only 83.2% of the flight impulse from the "waist and shoulder springs" condition. The impulse-momentum relationship states that the product of force and time equals product of mass and change in velocity. Therefore, if the impulse was greater in one condition for a subject with a given mass, the impact velocity would be proportionately greater.

The tibialis anterior, rectus femoris, and vastus lateralis were significantly more activated, in terms of their activation integrals, in the ZLS than in the 1G conditions. The gastrocnemius, biceps femoris, and gluteus maximus produced the same levels of activation during all four conditions.

Conclusion

The influence of the gravity replacement load on the perceived comfort, knee kinematics, and ground reaction force variables was the most important finding of this study. This influence was present in several different aspects:​ the mechanical aspects of load fluctuation and the psychological aspects of discomfort or heightened consciousness of the load. The load fluctuation had a dramatic effect on the ground reaction forces. The effects of the load fluctuation were stated above, but the desirability of this large fluctuation will remain to be determined until after it is known whether large forces or large loading rates are more important in the maintenance of bone density.

The discomfort of the subject load affected subjects such that they bent their knees to reduce this load. In order to replicate the knee kinematics of overground locomotion, particular attention should be paid to harness design. Maximization of harness comfort is of utmost importance if subjects are to run for extended bouts of exercise over a period of several months.

While the ground reaction force results were shown to be relatively ambiguous in terms of maintenance of bone mass, the electromyographic data were much more encouraging, especially in light of the fact that the health of the muscles also will influence the density of the bone (Currey, 1984). The activation integrals of the tibialis anterior, rectus femoris, and vastus lateralis were greater in the zero-gravity simulated conditions, while there were no significant differences between any of the conditions with regard to the activations of the gastrocnemius, biceps femoris, and gluteus maximus. This was surprising considering that while the subject in the ZLS was under a full body weight gravitational load, the actual limb segments were unweighted. Subjects thus did not have to resist gravity in each of the body segments as they raised their legs off of the surface to run;​ rather, it was the stance leg had to resist the pull of the springs. Nevertheless, a resulting increase in eccentric contractions was not noted in the ZLS conditions. It therefore must be concluded that, within the limitations of extrapolating to force from EMG patterns, fully-loaded tethered treadmill locomotion for sufficient duration could be a satisfactory countermeasure to both bone loss and muscular atrophy during long duration space flight. Issues of the duration of exercise for effective countermeasures remain to be answered and are outside the scope of this thesis.