- Email: [email protected]
Vestibulospinal adaptation to microgravity
Vestibulospinal adaptation to microgravity WILLIAM H. PALOSKI, PhD, Houston, Texas
Human balance control is known to be transiently disrupted after spaceflight; however, the mechanisms responsible for postflight postural ataxia are still under investigation. In this report, we propose a conceptual model of vestibulospinal adaptation based on theoretical adaptive control concepts and supported by the results from a comprehensive study of balance control recovery after spaceflight. The conceptual model predicts that immediately after spaceflight the balance control system of a returning astronaut does not expect to receive gravity-induced afferent inputs and that descending vestibulospinal control of balance is disrupted until the central nervous system is able to cope with the newly available vestibular otolith information. Predictions of the model are tested using data from a study of the neurosensory control of balance in astronauts immediately after landing. In that study, the mechanisms of sensorimotor balance control were assessed under normal, reduced, and/or altered (sway-referenced) visual and somatosensory input conditions. We conclude that the adaptive control model accurately describes the neurobehavioral responses to spaceflight and that similar models of altered sensory, motor, or environmental constraints are needed clinically to predict responses that patients with sensorimotor pathologies may have to various visual-vestibular or changing stimulus environments. (Otolaryngol Head Neck Surg 1998;118:S39-S44.)
Vestibulospinal or, perhaps more accurately, sensorimotor balance control systems automatically regulate biomechanical stability of the body while performing normal lifetime activities. Early in life, the central nervous system (CNS) learns to avoid falls and fall-related injuries by maintaining the body center of mass over the base of support during quiet activities, in anticipation of balance disturbances created by planned voluntary movements and in response to those caused by unexpected external perturbations. Later, the CNS learns to adapt these systems to altered environmental constraints such as the reduced support surface friction encountered while walking across a frozen pond, the altered base of support dimensions and visual flow fields encountered while riding a bicycle, and, for some, the altered relations among visual, vestibular,
From Life Sciences Research Laboratories, NASA/Johnson Space Center. Supported by the NASA Extended Duration Orbiter Medical Project (EDOMP DSO 605), the NASA Microgravity Vestibular Investigations Experiment, and the NASA Shuttle-Mir Science Program. Presented at “Vestibular Dysfunction: Lessons and Legacies from Space,” jointly sponsored by the American Neurotology Society, American Academy of Otolaryngology–Head and Neck Surgery, and NASA, September 28, 1996, Washington, D.C. Reprint requests: William H. Paloski, PhD, Mail Code SD3, NASA/Johnson Space Center, Houston, TX 77058. Copyright © 1998 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/98/$5.00 + 0 23/0/85850
and somatosensory inputs associated with traveling on maritime vehicles, piloting high-performance aircraft, scuba diving, or participating in orbital spaceflight missions. For others, the CNS must learn to adapt to (compensate for) traumatic, pathological, or therapeutic loss of sensory information from the vestibular, visual, or somatosensory systems. Balance control is a fundamental task performed by the CNS on a continuous, normally subconscious basis. Virtually every voluntary body movement we make, from swinging a golf club to taking a deep breath, is accompanied by automatic sets of anticipatory and compensatory muscle activations designed to maintain stable body posture while performing the focal task. Patients with degraded balance control function have significantly reduced abilities to perform normal activities of daily living. Therefore, assessment of balance control function is an integral part of most clinical neurologic examinations. Since at least 1846, when Romberg1 showed that static posture testing could identify patients with proprioceptive loss, one of the most widely used and well-studied balance control tasks has been the maintenance of stable upright stance. The Romberg test continues to be used clinically for standard neurologic examinations, whereas more sophisticated dynamic posturography tests have been developed for detailed evaluation of balance disorders.2 Using static and dynamic posturography techniques, scientists have now unraveled many of the underlying principles of neurosensory control of S39
S40 PALOSKI
Fig. 1. Sensorimotor control of balance. The central nervous system (CNS) estimates the current state of the body’s orientation from sensory feedback provided primarily by the visual, vestibular, and proprioceptive systems. Motor output commands are generated and adjusted by the CNS to minimize differences between desired and estimated body orientations. Desired orientation commands presumably originate in higher-level brain centers.
Otolaryngology– Head and Neck Surgery March 1998
human subjects during the adaptation process is studies of the recovery of balance control in astronauts after spaceflight.3-9 In this report, we propose a conceptual model of vestibulospinal adaptation based on the results from a comprehensive study of balance control recovery after spaceflight. Adaptive control models that accurately describe neurobehavioral responses to altered sensory, motor or environmental constraints are needed to aid our understanding of the responses to be expected when astronauts, high-performance aircraft pilots, and underwater explorers venture into and return from their unusual working environments. Such models are also needed clinically to predict responses that patients with sensorimotor pathologies may have to various environments, to guide physical therapists in designing rehabilitation programs to aid CNS adaptation (compensation) in their patients, and to aid engineers in designing appropriate adaptive controllers for functional electrical stimulation of paraplegic patients. CONCEPTUAL MODEL OF CNS ADAPTATION
Fig. 2. Adaptive control scheme for maintaining balance. Feedback control of upright postural stability depends on context-dependent internal models of the sensory, motor, and environmental characteristics. A copy of the multidimensional efferent motor command signal triggers a stored memory (internal model) that is used to predict the multidimensional, spatiotemporal sensory afferent information expected to be received (reafference) in response to the commanded motor activity. Comparisons between the expected and actual reafference form afference error signals that are used to modify the motor command generation and/or inform higherlevel centers of significant or sustained incongruences between expected and actual reafference.
upright stance. However, the mechanisms by which the vestibulospinal system adapts to environmental challenges are not well understood. Most of what is currently known about the mechanisms underlying adaptive changes in balance control has been learned from studies of patients with irreversible loss of sensory or motor function. Unfortunately, these observations have generally been made after compensation was already complete, so details of the adaptation process were lost. The primary source of systematic data detailing the responses of the balance control system in normal
To control balance during upright stance, the CNS uses sensory information obtained from the visual, vestibular (both canal and otolith), and somatosensory receptors (Fig. 1). From this information, the CNS estimates the true biomechanical state of the body (i.e., the physical relations among arms, legs, trunk, head, and space), compares this state to the currently desired body state (which presumably originates in higher-level centers), and then selects and commands the most appropriate motor control synergies and/or strategies10,11 to return the body to the desired orientation. Adaptive changes in vestibulospinal control of posture can be explained with the use of a conceptual model (Fig. 2) that is based on the efference copy hypothesis proposed by von Holst and Mittlestaedt,12 as later modified by Hein and Held.13 The latter authors proposed that the CNS stores a copy of each efferent motor command generated to effect a voluntary body movement. The efference copy triggers a memorized internal model of the body mechanics and sense organ transfer functions, which leads to a prediction of the sensory feedback (reafference) expected from the commanded voluntary movement. Any errors (incongruencies) between the received afferent input and the expected reafference are thought to trigger conscious awareness of the incongruency. Although such errors are usually caused by unexpected external perturbations, they can also be caused by changes in sensory, motor, or environmental characteristics that were not predicted by the internal model. We propose that adap-
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 2
tive responses are triggered by persistent afference errors, driven by higher level perception centers, and realized through trial and error by altering the internal models to provide new predictions of reafferent signals that are congruent with the actual reafference associated with voluntary movements. A generalized scheme describing CNS adaptation to microgravity has evolved from the concepts of reafference (Fig. 3). Upon insertion into orbit spaceflight, an astronaut experiences sudden loss of the gravitational stimuli transduced by the vestibular otolith organs (and other gravireceptors) and thought to be used by the CNS as the primary vertical spatial reference for controlling upright stance. During orbital spaceflight, absence of this continuous, earth-vertical spatial reference causes incongruence between the expected and actual sensory afference resulting from voluntary body movements. This incongruence creates a persistent afference error that raises conscious awareness, may lead to space motion sickness,14 and, most importantly, triggers the central adaptive processes that create new internal models, resulting in reinterpretation15 or neglect16 of gravity-induced sensory input signals. Over time, new internal models develop, compensating for the lost spatial reference. The end result of this adaptive response is that the CNS no longer seeks gravitational sensory inputs for use in estimating spatial orientation. This ameliorates space motion sickness and optimizes central neural control of coordinated body movements in the absence of gravity. Unfortunately for the space traveler, the in-flight adaptation also appears to significantly impact motor coordination and control of body movements immediately after return to earth.17 Our conceptual adaptation model can be used to explain how the reinterpretation of gravity-induced sensory inputs during spaceflight would disrupt balance control after flight. On orbit, the sensory input signals resulting from a voluntary body movement would initially be incongruent with those expected to be received, based on the terrestrial internal models. Because the incongruence persists, adaptive changes would be triggered by higher-level centers, leading, through trial and error, to development of new efference copy models optimized for the microgravity environment. These microgravity internal models would not expect sustained changes in vestibular otolith inputs to result from head tilts, as would the terrestrial internal models. Immediately after spaceflight, when the terrestrial gravity-induced sensory inputs return, the CNS would continue to use the microgravity efference copy models developed during flight, and the unexpected afferent signals caused by gravitational stimulation
PALOSKI S41
Fig. 3. Schematic representation of central nervous system adaptation to microgravity.
would disrupt sensorimotor control. These unexpected afferent signals would also create persistent afference errors that would trigger readaptive modification, leading eventually to a reemergence of the original, terrestrial efference copy models. Although in-flight adaptive changes in the CNS processing of sensory information may optimize central neural control of coordinated body movements in microgravity, they may also maladapt the astronaut for control of coordinated body movements immediately after return to earth.17 EXPERIMENTAL SUPPORT AND DISCUSSION
The conceptual model was developed on the basis of responses observed in more than 40 astronauts who participated in balance control experiments after spaceflight.8,9,18-20 The experimental stimuli were administered with a modified computerized dynamic posturography system, commercially available for clinical assessment of balance disorders (Equitest, NeuroCom International, Clackamas, Ore.). In these experiments, the subject stood erect on the system forceplate and was directed to “maintain stable upright posture.” The subject was then presented with multiple sequential trials of four types of postural perturbations created by suddenly translating or rotating the support surface (forceplate) under computer control. The perturbation sequence began with three trials of backward translation and con-
Otolaryngology– Head and Neck Surgery March 1998
S42 PALOSKI
tinued with five trials of toes-up rotation, three trials of forward translation, and five trials of toes-down rotation. After the sudden perturbation tests, the subject was presented with three randomized trials of each of a set of six sensory organization tests. The six sensory organization test conditions were (1) eyes open, (2) eyes closed, (3) sway-referenced vision, (4) sway-referenced support surface, (5) sway-referenced support surface with eyes closed, and (6) sway-referenced support surface with sway-referenced vision. Tests 1 and 2 were the standard Romberg tests. Tests 4 through 6 examined how well the balance control system resolved sensory information conflicts created by sway-referencing the visual and/or proprioceptive reference surfaces. (Sway-referencing was accomplished by servo-controlling the pitch orientations of the forceplate and/or visual surround to follow the subject’s center of mass sway.) In tests 5 and 6, the only accurate orientation information was provided by the vestibular system. Landing Day Balance Control Deficits
The conceptual model predicts that immediately after spaceflight the balance control system of a returning astronaut would use microgravity internal models (developed during spaceflight) and therefore would not expect the gravity-induced afferent signals that reappear upon landing. These unexpected afferent signals would create afference errors that would confuse the balance control system and disrupt postural stability, especially under conditions 5 and 6, when the vestibular system is the only accurate source of orientation reference information. In our experiments,8,9,18 balance control decrements were observed within 5 hours after flight in every astronaut studied. Under the standard Romberg conditions (tests 1 and 2), the sway amplitudes increased after flight by less than 0.5 degrees. Under the simple sensory conflict conditions (tests 3 and 4), the sway amplitudes increased after flight by 0.50 to 1.0 degrees. However, under the complex sensory conflict conditions (tests 5 and 6), the sway increased after flight by 2.0 to 3.0 degrees. Whereas the sway was increased on all sensory organization tests after flight, the increased sway was only stabilitythreatening under the postflight test conditions requiring accurate CNS interpretation of vestibular inputs (tests 5 and 6). These results clearly demonstrated that astronauts were unable to fully use vestibular sensory information in balance control immediately after landing. Indeed, the most severely affected returning crew members had performance patterns identical to those previously reported in vestibular-deficient patients exposed to this test battery.21
Recovery of Preflight Stability Levels
The conceptual model also predicts that the persistent afference errors created by the unexpected gravityinduced afferent signals that appear upon landing would trigger a readaptive response that would lead, eventually, to a new set of internal models structured to minimize afference errors, presumably by accounting accurately for the new gravitational inputs. The outward manifestation of this adaptive process would be a gradual improvement in balance control that would continue until the astronaut’s performance on the experimental test battery returned to preflight levels. Our experimental results were consistent with these model predictions because we found that all crew members gradually returned to preflight stability levels (or above) over the first 4 to 8 days after return from space.8 The recovery time course followed a double exponential path: a very rapid stability improvement over the first 8 to 10 hours after flight, followed by a more gradual return to preflight stability levels over the next 4 to 8 days. Furthermore, because the performance recovery was associated primarily with improvements in conditions 5 and 6, the results also suggest that postflight postural ataxia is mediated primarily by alterations in the vestibular (presumably otolithic) feedback loop. Detailed sensory analyses showed that the ankle proprioceptive feedback loop was also altered in some subjects, perhaps as a result of decreased postural loading or new movement strategies developed during spaceflight, and that most subjects had increased reliance on the visual feedback loop during the recovery process, perhaps in partial compensation for the degraded performance of the other two feedback systems during spaceflight.20 Rookies Versus Veterans
A further prediction of the conceptual model is that shifting from one set of internal models to another (e.g., from microgravity internal models to terrestrial internal models) becomes easier each time the CNS makes the transition between two different sets of environmental stimuli (e.g., from orbital spaceflight to earth). If true, then balance control deficits after spaceflight would become less severe as the number of previous spaceflights increased. Comparisons between the rookie and veteran astronaut groups in our experiments demonstrated significant differences between subjects having previous spaceflight experience and those having none.18 Preflight performances were statistically indistinguishable between these groups on every sensory organization test. Similarly, postflight performances on tests 1, 2, 3, and 4 were not different between rookies
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 2
and veterans. On the postflight conditions in which vestibular inputs provided the only reliable spatial orientation reference cues (tests 5 and 6), however, rookies exhibited significantly higher sway than veterans after flight. In a separate analysis of overall postural stability (based on the Equitest composite equilibrium score), we found that first-time fliers had the lowest scores on landing day, that second-time fliers had significantly higher scores than the first-time fliers, and that third-time fliers had significantly higher scores than second-time fliers. There were no differences between the groups before flight. Head-trunk Coordination Strategies
A final prediction of the conceptual model is that sustained afference errors cause neurobehavioral changes aimed at identifying the source of the error and appropriately adjusting the internal models to minimize it. If true, then new biomechanical strategies would be expected to emerge during the adaptation period after a change in sensory, motor, or environmental characteristics. One possible manifestation of this prediction in our astronaut subjects is that postflight postural biomechanics would be affected by adopted head-trunk coordination strategies aimed at minimizing confusing vestibular inputs. Data from our sudden perturbation translation trials were analyzed to assess such changes.19 We found that immediately after flight, the sway responses to the translational perturbations were exaggerated relative to before flight in all subjects. The center of pressure and hip sway trajectories were generally more labile, or underdamped, after flight than before flight, and the learning associated with successive sequential perturbations disappeared in some subjects on landing day. Before flight, specific head-trunk coordination strategies were not systematically observed. After flight, most subjects attempted to minimize head movements with respect to trunk movements during their active response to the support surface translation; however, some subjects attempted to minimize head movements with respect to space. The postflight strategy of maintaining the orientation of the head fixed with respect to the torso may have been adopted to minimize conflicts between neck proprioceptive and vestibular otolith sensory inputs. Alternatively, it may have been adopted to simplify the segmental motor control task by reducing the number of independent body segments to be controlled. A drawback of this strategy is that it would have also complicated the sensory task required to separate gravity from linear head accelerations. The postflight strategy of fixing the head in space may have been adopted to simpli-
PALOSKI S43
fy the sensory task of separating gravity from linear accelerations by maintaining constant the gravitational component of otolith stimulation. A drawback of this strategy is that it would also have increased the complexity of the segmental motor control task. In either case, these findings support the notion that balance control after spaceflight is driven by a top down program; however, in contrast to Berthoz and Pozzo,22 the top down program may have been organized around minimizing confounding vestibular inputs rather than stabilizing vision. Thus as predicted by the conceptual model, it appears that gravity-induced sensory inputs are confusing to the posture control system immediately after spaceflight and that either to minimize the confusion caused by these inputs or to test the source of the afference error, many of the returning astronauts adopted new biomechanical movement strategies designed to aid the adaptation process. CONCLUSIONS
We conclude that sensory feedback and accurate internal models are critical to vestibulospinal adaptation. Our conceptual model of vestibulospinal adaptation, based on the pioneering theoretical work of Hein and Held13 and supported by our results from a comprehensive study of balance control recovery after spaceflight, appears to accurately describe the adaptive neurobehavioral responses to the altered environmental constraints caused by orbital spaceflight. After spaceflight, the sensorimotor control of balance is initially disrupted because the new, microgravity internal models developed during spaceflight do not respond appropriately to the reemergent sensory inputs associated with gravitational stimulation. Although the conceptual model presented here may be a good starting point for more robust models in the future, it needs to be further validated before it can be used clinically to predict responses that patients with sensorimotor pathologies may have to various environments. Nevertheless, our postflight observations on readaptation of central posture control systems to the terrestrial gravitational environment provide us with a better understanding of the basic physiology of otolith function, sensorimotor organization, and central adaptation. They may also be relevant to clinical medicine insofar as they provide us with a better understanding of patient responses to acute vestibular loss, because vestibular loss patients are likely to undergo similar adaptive changes during their compensation processes. Improved understanding of the mechanisms and time courses of these adaptive processes is as crucial to designing effective rehabilitative physical therapy
S44 PALOSKI
regimes for patients as it is to defining effective countermeasures for returning astronauts. The author wishes to thank Drs. F. O. Black and M. F. Reschke for their help in formulating the conceptual model and interpreting the experimental results, E. Allen for providing invaluable assistance in preparing the manuscript, and all of the participating crew members for volunteering their time and efforts. REFERENCES 1. Romberg MH. Lehrbuch der nervenkrankheiten des menschen. Berlin: A. Dunker, 1846. 2. Nashner LM, Black FO, Wall III C. Adaptation to altered support and visual conditions during stance: patients with vestibular deficits. J Neurosci 1982;2:536-44. 3. Homick JL, Reschke MF. Postural equilibrium following exposure to weightless space flight. Acta Otolaryngol 1977;83:455-64. 4. Kozlovskaya IB, Kreidich YV, Oganov VS, et al. Pathophysiology of motor functions in prolonged manned space flights. Acta Astronautica 1981;8:1059-72. 5. Anderson DJ, Reschke MF, Homick JL, et al. Dynamic posture analysis of Spacelab-1 crew members. Exp Brain Res 1986;64: 380-91. 6. Clément G, Gurfinkel VS, Lestienne F, et al. Adaptation of posture control to weightlessness. Exp Brain Res 1984;57:61-72. 7. Kenyon RV, Young LR. MIT/Canadian vestibular experiments on Spacelab-1 mission, V: postural responses following exposure to weightlessness. Exp Brain Res 1986;64:335-46. 8. Paloski WH, Reschke MF, Black FO, et al. Recovery of postural equilibrium control following space flight. In: Cohen B, Tomko DL, Guedry F, editors. Sensing and controlling motion: vestibular and sensorimotor function. New York: Ann N Y Acad Sci 1992;682:747-54. 9. Paloski WH, Bloomberg JJ, Reschke MF, et al. Spaceflight-induced changes in posture and locomotion. J Biomech 1994;27:812.
Otolaryngology– Head and Neck Surgery March 1998
10. Nashner LM, Berthoz A. Visual contribution to rapid motor responses during postural control. Brain Res 1978;150:403-7. 11. Massion J. Postural changes accompanying voluntary movements: normal and pathological aspects. Hum Neurobiol 1984;2: 261-7. 12. von Holst E, Mittlestaedt H. Das Reafferenzprinzip. Naturwissenschaften 1950;37:464-76. 13. Hein A, Held R. Dissociation of the visual placing response into elicited and guided components. Science 1967;158:390-2. 14. Reason JT, Brand JJ. Motion Sickness. London: Academic Press, 1975. 15. Parker DE, Reschke MF, Arrott AP, et al. Otolith tilt-translation reinterpretation following prolonged weightlessness: implications for preflight training. Aviat Space Environ Med 1985;56: 601-6. 16. Young LR. Adaptation to modified otolith input. In: Berthoz A, Melvill Jones G, editors. Adaptive mechanisms in gaze control. Amsterdam: Elsevier, 1985. 17. Reschke MF, Bloomberg JJ, Harm DL, et al. In: Nicogossian AE, Huntoon CL, Pool SL, editors. Neurophysiological aspects: Sensory and sensory-motor function. Space physiology and medicine. 3rd ed. Philadelphia: Lea & Febiger, 1994:261-85. 18. Reschke MF, Harm DL, Bloomberg JJ, et al. Neurosensory and sensory-motor function. In: Genin AM, Hunton CL, editors. Space biology and medicine. Vol. 3: Humans in spaceflight. Book 1: Effects of microgravity. Washington, D.C.: AIAA, 1997. 19. Paloski WH. Control of posture and movement. In: Winters J, Crago P, editors. Neuro-musculo-skeletal interaction and organizational principles. Heidelberg: Springer-Verlag, 1998. 20. Black FO, Paloski WH, Doxey-Gasway DD, et al. Vestibular plasticity following orbital space flight: recovery from postflight postural instability. Acta Otolaryngol (Stockh) 1995;520:450-4. 21. Black FO, Shupert CL, Horak FB, et al. Abnormal postural control associated with peripheral vestibular disorders. Prog Brain Res 1988;76:263-75. 22. Berthoz A, Pozzo T. Intermittent head stabilization during postural and locomotory tasks in humans. In: Amblard B, Berthoz A, Clarac F, editors. Posture and gait: development, adaptation and modulation. Amsterdam: Elsevier, 1988:189-98.
Human balance control is known to be transiently disrupted after spaceflight; however, the mechanisms responsible for postflight postural ataxia are still under investigation. In this report, we propose a conceptual model of vestibulospinal adaptation based on theoretical adaptive control concepts and supported by the results from a comprehensive study of balance control recovery after spaceflight. The conceptual model predicts that immediately after spaceflight the balance control system of a returning astronaut does not expect to receive gravity-induced afferent inputs and that descending vestibulospinal control of balance is disrupted until the central nervous system is able to cope with the newly available vestibular otolith information. Predictions of the model are tested using data from a study of the neurosensory control of balance in astronauts immediately after landing. In that study, the mechanisms of sensorimotor balance control were assessed under normal, reduced, and/or altered (sway-referenced) visual and somatosensory input conditions. We conclude that the adaptive control model accurately describes the neurobehavioral responses to spaceflight and that similar models of altered sensory, motor, or environmental constraints are needed clinically to predict responses that patients with sensorimotor pathologies may have to various visual-vestibular or changing stimulus environments. (Otolaryngol Head Neck Surg 1998;118:S39-S44.)
Vestibulospinal or, perhaps more accurately, sensorimotor balance control systems automatically regulate biomechanical stability of the body while performing normal lifetime activities. Early in life, the central nervous system (CNS) learns to avoid falls and fall-related injuries by maintaining the body center of mass over the base of support during quiet activities, in anticipation of balance disturbances created by planned voluntary movements and in response to those caused by unexpected external perturbations. Later, the CNS learns to adapt these systems to altered environmental constraints such as the reduced support surface friction encountered while walking across a frozen pond, the altered base of support dimensions and visual flow fields encountered while riding a bicycle, and, for some, the altered relations among visual, vestibular,
From Life Sciences Research Laboratories, NASA/Johnson Space Center. Supported by the NASA Extended Duration Orbiter Medical Project (EDOMP DSO 605), the NASA Microgravity Vestibular Investigations Experiment, and the NASA Shuttle-Mir Science Program. Presented at “Vestibular Dysfunction: Lessons and Legacies from Space,” jointly sponsored by the American Neurotology Society, American Academy of Otolaryngology–Head and Neck Surgery, and NASA, September 28, 1996, Washington, D.C. Reprint requests: William H. Paloski, PhD, Mail Code SD3, NASA/Johnson Space Center, Houston, TX 77058. Copyright © 1998 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/98/$5.00 + 0 23/0/85850
and somatosensory inputs associated with traveling on maritime vehicles, piloting high-performance aircraft, scuba diving, or participating in orbital spaceflight missions. For others, the CNS must learn to adapt to (compensate for) traumatic, pathological, or therapeutic loss of sensory information from the vestibular, visual, or somatosensory systems. Balance control is a fundamental task performed by the CNS on a continuous, normally subconscious basis. Virtually every voluntary body movement we make, from swinging a golf club to taking a deep breath, is accompanied by automatic sets of anticipatory and compensatory muscle activations designed to maintain stable body posture while performing the focal task. Patients with degraded balance control function have significantly reduced abilities to perform normal activities of daily living. Therefore, assessment of balance control function is an integral part of most clinical neurologic examinations. Since at least 1846, when Romberg1 showed that static posture testing could identify patients with proprioceptive loss, one of the most widely used and well-studied balance control tasks has been the maintenance of stable upright stance. The Romberg test continues to be used clinically for standard neurologic examinations, whereas more sophisticated dynamic posturography tests have been developed for detailed evaluation of balance disorders.2 Using static and dynamic posturography techniques, scientists have now unraveled many of the underlying principles of neurosensory control of S39
S40 PALOSKI
Fig. 1. Sensorimotor control of balance. The central nervous system (CNS) estimates the current state of the body’s orientation from sensory feedback provided primarily by the visual, vestibular, and proprioceptive systems. Motor output commands are generated and adjusted by the CNS to minimize differences between desired and estimated body orientations. Desired orientation commands presumably originate in higher-level brain centers.
Otolaryngology– Head and Neck Surgery March 1998
human subjects during the adaptation process is studies of the recovery of balance control in astronauts after spaceflight.3-9 In this report, we propose a conceptual model of vestibulospinal adaptation based on the results from a comprehensive study of balance control recovery after spaceflight. Adaptive control models that accurately describe neurobehavioral responses to altered sensory, motor or environmental constraints are needed to aid our understanding of the responses to be expected when astronauts, high-performance aircraft pilots, and underwater explorers venture into and return from their unusual working environments. Such models are also needed clinically to predict responses that patients with sensorimotor pathologies may have to various environments, to guide physical therapists in designing rehabilitation programs to aid CNS adaptation (compensation) in their patients, and to aid engineers in designing appropriate adaptive controllers for functional electrical stimulation of paraplegic patients. CONCEPTUAL MODEL OF CNS ADAPTATION
Fig. 2. Adaptive control scheme for maintaining balance. Feedback control of upright postural stability depends on context-dependent internal models of the sensory, motor, and environmental characteristics. A copy of the multidimensional efferent motor command signal triggers a stored memory (internal model) that is used to predict the multidimensional, spatiotemporal sensory afferent information expected to be received (reafference) in response to the commanded motor activity. Comparisons between the expected and actual reafference form afference error signals that are used to modify the motor command generation and/or inform higherlevel centers of significant or sustained incongruences between expected and actual reafference.
upright stance. However, the mechanisms by which the vestibulospinal system adapts to environmental challenges are not well understood. Most of what is currently known about the mechanisms underlying adaptive changes in balance control has been learned from studies of patients with irreversible loss of sensory or motor function. Unfortunately, these observations have generally been made after compensation was already complete, so details of the adaptation process were lost. The primary source of systematic data detailing the responses of the balance control system in normal
To control balance during upright stance, the CNS uses sensory information obtained from the visual, vestibular (both canal and otolith), and somatosensory receptors (Fig. 1). From this information, the CNS estimates the true biomechanical state of the body (i.e., the physical relations among arms, legs, trunk, head, and space), compares this state to the currently desired body state (which presumably originates in higher-level centers), and then selects and commands the most appropriate motor control synergies and/or strategies10,11 to return the body to the desired orientation. Adaptive changes in vestibulospinal control of posture can be explained with the use of a conceptual model (Fig. 2) that is based on the efference copy hypothesis proposed by von Holst and Mittlestaedt,12 as later modified by Hein and Held.13 The latter authors proposed that the CNS stores a copy of each efferent motor command generated to effect a voluntary body movement. The efference copy triggers a memorized internal model of the body mechanics and sense organ transfer functions, which leads to a prediction of the sensory feedback (reafference) expected from the commanded voluntary movement. Any errors (incongruencies) between the received afferent input and the expected reafference are thought to trigger conscious awareness of the incongruency. Although such errors are usually caused by unexpected external perturbations, they can also be caused by changes in sensory, motor, or environmental characteristics that were not predicted by the internal model. We propose that adap-
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 2
tive responses are triggered by persistent afference errors, driven by higher level perception centers, and realized through trial and error by altering the internal models to provide new predictions of reafferent signals that are congruent with the actual reafference associated with voluntary movements. A generalized scheme describing CNS adaptation to microgravity has evolved from the concepts of reafference (Fig. 3). Upon insertion into orbit spaceflight, an astronaut experiences sudden loss of the gravitational stimuli transduced by the vestibular otolith organs (and other gravireceptors) and thought to be used by the CNS as the primary vertical spatial reference for controlling upright stance. During orbital spaceflight, absence of this continuous, earth-vertical spatial reference causes incongruence between the expected and actual sensory afference resulting from voluntary body movements. This incongruence creates a persistent afference error that raises conscious awareness, may lead to space motion sickness,14 and, most importantly, triggers the central adaptive processes that create new internal models, resulting in reinterpretation15 or neglect16 of gravity-induced sensory input signals. Over time, new internal models develop, compensating for the lost spatial reference. The end result of this adaptive response is that the CNS no longer seeks gravitational sensory inputs for use in estimating spatial orientation. This ameliorates space motion sickness and optimizes central neural control of coordinated body movements in the absence of gravity. Unfortunately for the space traveler, the in-flight adaptation also appears to significantly impact motor coordination and control of body movements immediately after return to earth.17 Our conceptual adaptation model can be used to explain how the reinterpretation of gravity-induced sensory inputs during spaceflight would disrupt balance control after flight. On orbit, the sensory input signals resulting from a voluntary body movement would initially be incongruent with those expected to be received, based on the terrestrial internal models. Because the incongruence persists, adaptive changes would be triggered by higher-level centers, leading, through trial and error, to development of new efference copy models optimized for the microgravity environment. These microgravity internal models would not expect sustained changes in vestibular otolith inputs to result from head tilts, as would the terrestrial internal models. Immediately after spaceflight, when the terrestrial gravity-induced sensory inputs return, the CNS would continue to use the microgravity efference copy models developed during flight, and the unexpected afferent signals caused by gravitational stimulation
PALOSKI S41
Fig. 3. Schematic representation of central nervous system adaptation to microgravity.
would disrupt sensorimotor control. These unexpected afferent signals would also create persistent afference errors that would trigger readaptive modification, leading eventually to a reemergence of the original, terrestrial efference copy models. Although in-flight adaptive changes in the CNS processing of sensory information may optimize central neural control of coordinated body movements in microgravity, they may also maladapt the astronaut for control of coordinated body movements immediately after return to earth.17 EXPERIMENTAL SUPPORT AND DISCUSSION
The conceptual model was developed on the basis of responses observed in more than 40 astronauts who participated in balance control experiments after spaceflight.8,9,18-20 The experimental stimuli were administered with a modified computerized dynamic posturography system, commercially available for clinical assessment of balance disorders (Equitest, NeuroCom International, Clackamas, Ore.). In these experiments, the subject stood erect on the system forceplate and was directed to “maintain stable upright posture.” The subject was then presented with multiple sequential trials of four types of postural perturbations created by suddenly translating or rotating the support surface (forceplate) under computer control. The perturbation sequence began with three trials of backward translation and con-
Otolaryngology– Head and Neck Surgery March 1998
S42 PALOSKI
tinued with five trials of toes-up rotation, three trials of forward translation, and five trials of toes-down rotation. After the sudden perturbation tests, the subject was presented with three randomized trials of each of a set of six sensory organization tests. The six sensory organization test conditions were (1) eyes open, (2) eyes closed, (3) sway-referenced vision, (4) sway-referenced support surface, (5) sway-referenced support surface with eyes closed, and (6) sway-referenced support surface with sway-referenced vision. Tests 1 and 2 were the standard Romberg tests. Tests 4 through 6 examined how well the balance control system resolved sensory information conflicts created by sway-referencing the visual and/or proprioceptive reference surfaces. (Sway-referencing was accomplished by servo-controlling the pitch orientations of the forceplate and/or visual surround to follow the subject’s center of mass sway.) In tests 5 and 6, the only accurate orientation information was provided by the vestibular system. Landing Day Balance Control Deficits
The conceptual model predicts that immediately after spaceflight the balance control system of a returning astronaut would use microgravity internal models (developed during spaceflight) and therefore would not expect the gravity-induced afferent signals that reappear upon landing. These unexpected afferent signals would create afference errors that would confuse the balance control system and disrupt postural stability, especially under conditions 5 and 6, when the vestibular system is the only accurate source of orientation reference information. In our experiments,8,9,18 balance control decrements were observed within 5 hours after flight in every astronaut studied. Under the standard Romberg conditions (tests 1 and 2), the sway amplitudes increased after flight by less than 0.5 degrees. Under the simple sensory conflict conditions (tests 3 and 4), the sway amplitudes increased after flight by 0.50 to 1.0 degrees. However, under the complex sensory conflict conditions (tests 5 and 6), the sway increased after flight by 2.0 to 3.0 degrees. Whereas the sway was increased on all sensory organization tests after flight, the increased sway was only stabilitythreatening under the postflight test conditions requiring accurate CNS interpretation of vestibular inputs (tests 5 and 6). These results clearly demonstrated that astronauts were unable to fully use vestibular sensory information in balance control immediately after landing. Indeed, the most severely affected returning crew members had performance patterns identical to those previously reported in vestibular-deficient patients exposed to this test battery.21
Recovery of Preflight Stability Levels
The conceptual model also predicts that the persistent afference errors created by the unexpected gravityinduced afferent signals that appear upon landing would trigger a readaptive response that would lead, eventually, to a new set of internal models structured to minimize afference errors, presumably by accounting accurately for the new gravitational inputs. The outward manifestation of this adaptive process would be a gradual improvement in balance control that would continue until the astronaut’s performance on the experimental test battery returned to preflight levels. Our experimental results were consistent with these model predictions because we found that all crew members gradually returned to preflight stability levels (or above) over the first 4 to 8 days after return from space.8 The recovery time course followed a double exponential path: a very rapid stability improvement over the first 8 to 10 hours after flight, followed by a more gradual return to preflight stability levels over the next 4 to 8 days. Furthermore, because the performance recovery was associated primarily with improvements in conditions 5 and 6, the results also suggest that postflight postural ataxia is mediated primarily by alterations in the vestibular (presumably otolithic) feedback loop. Detailed sensory analyses showed that the ankle proprioceptive feedback loop was also altered in some subjects, perhaps as a result of decreased postural loading or new movement strategies developed during spaceflight, and that most subjects had increased reliance on the visual feedback loop during the recovery process, perhaps in partial compensation for the degraded performance of the other two feedback systems during spaceflight.20 Rookies Versus Veterans
A further prediction of the conceptual model is that shifting from one set of internal models to another (e.g., from microgravity internal models to terrestrial internal models) becomes easier each time the CNS makes the transition between two different sets of environmental stimuli (e.g., from orbital spaceflight to earth). If true, then balance control deficits after spaceflight would become less severe as the number of previous spaceflights increased. Comparisons between the rookie and veteran astronaut groups in our experiments demonstrated significant differences between subjects having previous spaceflight experience and those having none.18 Preflight performances were statistically indistinguishable between these groups on every sensory organization test. Similarly, postflight performances on tests 1, 2, 3, and 4 were not different between rookies
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 2
and veterans. On the postflight conditions in which vestibular inputs provided the only reliable spatial orientation reference cues (tests 5 and 6), however, rookies exhibited significantly higher sway than veterans after flight. In a separate analysis of overall postural stability (based on the Equitest composite equilibrium score), we found that first-time fliers had the lowest scores on landing day, that second-time fliers had significantly higher scores than the first-time fliers, and that third-time fliers had significantly higher scores than second-time fliers. There were no differences between the groups before flight. Head-trunk Coordination Strategies
A final prediction of the conceptual model is that sustained afference errors cause neurobehavioral changes aimed at identifying the source of the error and appropriately adjusting the internal models to minimize it. If true, then new biomechanical strategies would be expected to emerge during the adaptation period after a change in sensory, motor, or environmental characteristics. One possible manifestation of this prediction in our astronaut subjects is that postflight postural biomechanics would be affected by adopted head-trunk coordination strategies aimed at minimizing confusing vestibular inputs. Data from our sudden perturbation translation trials were analyzed to assess such changes.19 We found that immediately after flight, the sway responses to the translational perturbations were exaggerated relative to before flight in all subjects. The center of pressure and hip sway trajectories were generally more labile, or underdamped, after flight than before flight, and the learning associated with successive sequential perturbations disappeared in some subjects on landing day. Before flight, specific head-trunk coordination strategies were not systematically observed. After flight, most subjects attempted to minimize head movements with respect to trunk movements during their active response to the support surface translation; however, some subjects attempted to minimize head movements with respect to space. The postflight strategy of maintaining the orientation of the head fixed with respect to the torso may have been adopted to minimize conflicts between neck proprioceptive and vestibular otolith sensory inputs. Alternatively, it may have been adopted to simplify the segmental motor control task by reducing the number of independent body segments to be controlled. A drawback of this strategy is that it would have also complicated the sensory task required to separate gravity from linear head accelerations. The postflight strategy of fixing the head in space may have been adopted to simpli-
PALOSKI S43
fy the sensory task of separating gravity from linear accelerations by maintaining constant the gravitational component of otolith stimulation. A drawback of this strategy is that it would also have increased the complexity of the segmental motor control task. In either case, these findings support the notion that balance control after spaceflight is driven by a top down program; however, in contrast to Berthoz and Pozzo,22 the top down program may have been organized around minimizing confounding vestibular inputs rather than stabilizing vision. Thus as predicted by the conceptual model, it appears that gravity-induced sensory inputs are confusing to the posture control system immediately after spaceflight and that either to minimize the confusion caused by these inputs or to test the source of the afference error, many of the returning astronauts adopted new biomechanical movement strategies designed to aid the adaptation process. CONCLUSIONS
We conclude that sensory feedback and accurate internal models are critical to vestibulospinal adaptation. Our conceptual model of vestibulospinal adaptation, based on the pioneering theoretical work of Hein and Held13 and supported by our results from a comprehensive study of balance control recovery after spaceflight, appears to accurately describe the adaptive neurobehavioral responses to the altered environmental constraints caused by orbital spaceflight. After spaceflight, the sensorimotor control of balance is initially disrupted because the new, microgravity internal models developed during spaceflight do not respond appropriately to the reemergent sensory inputs associated with gravitational stimulation. Although the conceptual model presented here may be a good starting point for more robust models in the future, it needs to be further validated before it can be used clinically to predict responses that patients with sensorimotor pathologies may have to various environments. Nevertheless, our postflight observations on readaptation of central posture control systems to the terrestrial gravitational environment provide us with a better understanding of the basic physiology of otolith function, sensorimotor organization, and central adaptation. They may also be relevant to clinical medicine insofar as they provide us with a better understanding of patient responses to acute vestibular loss, because vestibular loss patients are likely to undergo similar adaptive changes during their compensation processes. Improved understanding of the mechanisms and time courses of these adaptive processes is as crucial to designing effective rehabilitative physical therapy
S44 PALOSKI
regimes for patients as it is to defining effective countermeasures for returning astronauts. The author wishes to thank Drs. F. O. Black and M. F. Reschke for their help in formulating the conceptual model and interpreting the experimental results, E. Allen for providing invaluable assistance in preparing the manuscript, and all of the participating crew members for volunteering their time and efforts. REFERENCES 1. Romberg MH. Lehrbuch der nervenkrankheiten des menschen. Berlin: A. Dunker, 1846. 2. Nashner LM, Black FO, Wall III C. Adaptation to altered support and visual conditions during stance: patients with vestibular deficits. J Neurosci 1982;2:536-44. 3. Homick JL, Reschke MF. Postural equilibrium following exposure to weightless space flight. Acta Otolaryngol 1977;83:455-64. 4. Kozlovskaya IB, Kreidich YV, Oganov VS, et al. Pathophysiology of motor functions in prolonged manned space flights. Acta Astronautica 1981;8:1059-72. 5. Anderson DJ, Reschke MF, Homick JL, et al. Dynamic posture analysis of Spacelab-1 crew members. Exp Brain Res 1986;64: 380-91. 6. Clément G, Gurfinkel VS, Lestienne F, et al. Adaptation of posture control to weightlessness. Exp Brain Res 1984;57:61-72. 7. Kenyon RV, Young LR. MIT/Canadian vestibular experiments on Spacelab-1 mission, V: postural responses following exposure to weightlessness. Exp Brain Res 1986;64:335-46. 8. Paloski WH, Reschke MF, Black FO, et al. Recovery of postural equilibrium control following space flight. In: Cohen B, Tomko DL, Guedry F, editors. Sensing and controlling motion: vestibular and sensorimotor function. New York: Ann N Y Acad Sci 1992;682:747-54. 9. Paloski WH, Bloomberg JJ, Reschke MF, et al. Spaceflight-induced changes in posture and locomotion. J Biomech 1994;27:812.
Otolaryngology– Head and Neck Surgery March 1998
10. Nashner LM, Berthoz A. Visual contribution to rapid motor responses during postural control. Brain Res 1978;150:403-7. 11. Massion J. Postural changes accompanying voluntary movements: normal and pathological aspects. Hum Neurobiol 1984;2: 261-7. 12. von Holst E, Mittlestaedt H. Das Reafferenzprinzip. Naturwissenschaften 1950;37:464-76. 13. Hein A, Held R. Dissociation of the visual placing response into elicited and guided components. Science 1967;158:390-2. 14. Reason JT, Brand JJ. Motion Sickness. London: Academic Press, 1975. 15. Parker DE, Reschke MF, Arrott AP, et al. Otolith tilt-translation reinterpretation following prolonged weightlessness: implications for preflight training. Aviat Space Environ Med 1985;56: 601-6. 16. Young LR. Adaptation to modified otolith input. In: Berthoz A, Melvill Jones G, editors. Adaptive mechanisms in gaze control. Amsterdam: Elsevier, 1985. 17. Reschke MF, Bloomberg JJ, Harm DL, et al. In: Nicogossian AE, Huntoon CL, Pool SL, editors. Neurophysiological aspects: Sensory and sensory-motor function. Space physiology and medicine. 3rd ed. Philadelphia: Lea & Febiger, 1994:261-85. 18. Reschke MF, Harm DL, Bloomberg JJ, et al. Neurosensory and sensory-motor function. In: Genin AM, Hunton CL, editors. Space biology and medicine. Vol. 3: Humans in spaceflight. Book 1: Effects of microgravity. Washington, D.C.: AIAA, 1997. 19. Paloski WH. Control of posture and movement. In: Winters J, Crago P, editors. Neuro-musculo-skeletal interaction and organizational principles. Heidelberg: Springer-Verlag, 1998. 20. Black FO, Paloski WH, Doxey-Gasway DD, et al. Vestibular plasticity following orbital space flight: recovery from postflight postural instability. Acta Otolaryngol (Stockh) 1995;520:450-4. 21. Black FO, Shupert CL, Horak FB, et al. Abnormal postural control associated with peripheral vestibular disorders. Prog Brain Res 1988;76:263-75. 22. Berthoz A, Pozzo T. Intermittent head stabilization during postural and locomotory tasks in humans. In: Amblard B, Berthoz A, Clarac F, editors. Posture and gait: development, adaptation and modulation. Amsterdam: Elsevier, 1988:189-98.