Fay B. Horak PhD, PT
Invited Oral Presenation
In a Theme: Peripheral and Central Somatosensory Control of Posture
Antonio Nardonie: Effects of Peripheral and Spinal Somatosensory Loss Richard Poppele: Global Somatosensory Encoding in the Spinocerebellar Track Baastian Bloem: Somatosensory Control of Posture in Parkinsonian Patients
Contact for Correspondence:
Fay B Horak Ph.D., PT
Senior Scientist and Professor
Neurological Sciences Institute of Oregon Health Sciences University
505 NW 185th Ave.
Beaverton, OR 97006
Despite the importance of somatosensory information for postural control, most patients with somatosensory loss in the feet and legs can stand and walk independently, although their fall rate is 23 times higher than in healthy control subjects. Somatosensory loss in the legs results in absent or delayed postural responses, reduced scaling of postural response magnitude for increasing perturbation magnitude, and increased sway in stance (Horak and Macpherson, 1995). Recent studies in our laboratory support the hypothesis that postural compensation for somatosensory loss can involve sensory substitution, predictive mechanisms, and increased sensitivity to remaining sensory information. For example, an unusual patient with total body loss of large fiber sensory afferents demonstrated that auditory cues indicating perturbation onset can trigger functional postural responses when the direction of perturbation is predictable (Horak, et al, 1996). We also show that patients with partial loss of somatosensory information from the feet from diabetic peripheral neuropathy can substitute light touch from a fingertip to reduce sway and improve scaling of postural response magnitude (Dickstein, et al, in press).
Auditory substitution in a subject with total body somatosensory loss
The deafferented subject was a 47-year-old female (GL) who suddenly lost somatosensory function from a viral sensory polyneuropathy 15 years prior to testing. She had loss of all tendon reflexes, light and crude touch, vibration, kinesthesia, and position sense. Bilateral sural nerve biopsies revealed near total loss of fibers >6.5µm. Muscle strength, motor nerve conduction and needle electromyographic activity was normal. Postural muscle responses were measured to forward and backward surface translations (6 cm ramp at 35 cm/s) and to toes-up and toes-down rotations (5 deg at 20 deg/s) in a block of 5 trials with eyes open or eyes closed. To eliminate the possibility of using auditory cues to trigger a postural response, a block of trials was tested with hearing masked via earphone playing waterfall music.
Light fingertip touch substitution for somatosensory loss due to diabetic neuropathy
This study determined the extent to which subjects with somatosensory loss in the feet can substitute light touch of a fingertip on a stationary support to improve postural stability. Eight subjects with chronic diabetic peripheral neuropathy and 8 age-matched healthy control subjects (mean age 58) participated in these studies. Electrodiagnostic testing showed no plantar nerve responses and sural nerve responses in 6 out of the 8 diabetic subjects. Vibration sense, joint position sense and cutaneous sensation were significantly diminished at the feet and ankles in all diabetic subjects. Motor nerve conduction, joint range of motion, leg muscle strength and vestibulospinal reflexes in the horizontal plane were normal in all subjects. Trials with and without light touch (<100 grams), eyes open and eyes closed, were compared in 3 tasks: quiet stance on a firm surface stance on a compliant foam surface (7.5 cm Temper foam), and stance on a slowly rotating surface (± 4 deg at .05 Hz). Anterior-posterior (AP) and mediolateral (ML) center of pressure (CoP) mean sway area and mean angular trunk velocity was compared between groups and between touch, surface and visual conditions (MANOVA). Another study examined the scaling of postural response magnitude (initial CoP rate of change) with 5 backward platform translation velocities (5 -55 cm/s) with no fingertip touch, light touch (<100 grams) and heavy touch (as much as desired).
The subject with complete, total body loss of somatosensory function shows absent postural responses to surface perturbations, even with eyes open and adequate vestibular function Figure 1 A). Auditory cues associated with onset of platform movement could apparently be used to trigger these responses because postural responses were absent whenever hearing was masked. Availability of visual information from the ambient environment was not sufficient to trigger postural responses when hearing was masked. This subject with total somatosensory loss was able to trigger late, but functional, postural responses when hearing was available. Although postural responses were delayed between 20 and 500 ms and quite variable in latency from trial to trial, postural muscle activity occurred in the appropriate muscles for each direction of surface perturbation. For example, gastrocnemius was the first muscle active in response to a backward surface translation whereas tibialis was the first muscle activated in response to a toes-up rotation. Prediction of platform perturbation direction was necessary for direction-specific postural responses since auditory cues indicating platform onset were non-specific for direction. Although remarkably rapid and functionally preventing falls, these direction-specific postural responses were likely voluntarily-triggered because they required intention by the subject as well as auditory cues and predictability of perturbation characteristics.
Light fingertip touch substitution
The ML CoP sway and trunk velocity was larger in subjects with somatosensory loss than in control subjects, especially when standing on the foam surface. Fingertip somatosensory input through light touch significantly attenuated both AP and ML CoP and trunk velocity in both healthy subjects and subjects with diabetic neuropathy (p<0.0001). Since the subjects with neuropathy had more sway with eyes closed on foam without touch than the control subjects but similar sway as controls with light fingertip touch, this ability to use fingertip somatosensory information to control postural sway appears to be more sensitive in the subjects with neuropathy. Fingertip somatosensory input through light touch attenuated trunk velocity as much as heavy touch although the decrease in CoP sway was less effective with light than heavy touch. Figure 2 shows that light touch is even more effective in reducing postural sway than vision. Fingertip touch was also helpful in increasing the scaling of postural response magnitude for increasing velocities of surface translations, although more than 100 grams of touch was required.
Studies have shown that patients with somatosensory loss rely on a variety of compensatory mechanisms to maintain balance including sensory substitution, increase sensory sensitivity, and predictive mechanisms. The delays and depression of postural responses in patients with partial somatosensory loss and the absent postural responses in a patient with total body somatosensory loss, despite the presence of visual and vestibular information, suggests that somatosensory information from the lower extremities is critical for control of posture in stance. Nevertheless, many patients with severe somatosensory loss can stand and walk independently because of their ability to use compensatory mechanisms. One natural, compensatory mechanism for loss of somatosensory information is to substitute alternative, intact sensory information. The subject with total body somatosensory loss was able to trigger a functional, although somewhat delayed, postural response based on auditory cues regarding onset of surface perturbations when the direction was predictable. Subjects who have remaining somatosensory information in the upper extremities naturally touch walls and other stationary objects to assist balance control. We have shown that light touch incapable of providing mechanical support can significantly reduce body sway and that patients with chronic sensory loss in the feet show more benefit from this light touch than healthy subjects. An increase in sensitivity of remaining senses for postural control is consistent with the increased vestibulospinal sensitivity using galvanic stimulation in patients with somatosensory loss (Horak and Hlavacka, in press). Examining postural control in patients with loss of sensory information reveals not only the normal role of the missing sense but also the mechanisms used by the nervous system to compensate for the loss.
Supported by NIH grants AG 06457 and DC 01849.
Figure 1. Average of 5 EMG responses to backward surface translations in a healthy adult (A), lack of response in a subject with total body somatosensory loss with hearing masked (B) and functional response in the same subject with somatosensory loss with hearing of surface perturbation available (C).
Figure 2. Effect of light touch on CoP during very slow, sinusoidal platform rotations over 4 minutes was more effective than vision in a subject with somatosensory loss due to peripheral neuropathy. Anterior/posterior center of pressure traces from single trials without touch or vision, with vision but no touch and with fingertip touch (< 100 grams) but not vision are shown.
Dickstein R, Shupert, C L and Horak FB. Finger touch improves postural stability in patients with peripheral neuropathy. Gait and Posture, in press.
Horak FB and Hlavacka F. Somatosensory loss increases vestibulospinal sensitivity. Exp Brain Res in press.
Horak FB, Lamarrre Y, Macpherson JM, Shupert C, Henry SM, Macpherson. Postural control associated with total body somatosensory loss. Soc Neurosci Abstracts 1996; 22:1632.
Horak FB and Macpherson JM. Postural orientation and equilibrium. In: Shepard J. and Rowell L. (eds.), Handbook of Physiology: Section 12, Exercise: Regulation and Integration of Multiple Systems. New York: Oxford University Press 1996: 255-292.