Reflex Interactions during Whole Head-and-Body Tilts are Modified by Age in Humans

The effects of vibrotactile biofeedback training on trunk sway in Parkinson's disease patients

BACKGROUND

Postural instability in Parkinson's disease (PD) can lead to falls, injuries and reduced quality of life. We investigated whether balance in PD can improve by offering patients feedback about their own trunk sway as a supplement to natural sensory inputs. Specifically, we investigated the effect of artificial vibrotactile biofeedback on trunk sway in PD.

METHODS

Twenty PD patients were assigned to a control group (n = 10) or biofeedback group (n = 10). First, all patients performed two sets of six gait tasks and six stance tasks (pre-training assessment). Subsequently, all subjects trained six selected tasks five times (balance training). During this training, the feedback group received vibrotactile feedback of trunk sway, via vibrations delivered at the head. After training, both groups repeated all twelve tasks (post-training assessment). During all tasks, trunk pitch and roll movements were measured with angular velocity sensors attached to the lower trunk. Outcomes included sway angle and sway angular velocity in the roll and pitch plane, and task duration.

RESULTS

Overall, patients in the feedback group had a significantly greater reduction in roll (P = 0.005) and pitch (P < 0.001) sway angular velocity. Moreover, roll sway angle increased more in controls after training, suggesting better training effects in the feedback group (P < 0.001).

CONCLUSIONS

One session of balance training in PD using a biofeedback system showed beneficial effects on trunk stability. Additional research should examine if these effects increase further after more intensive training, how long these persist after training has stopped, and if the observed effects carry over to non-trained tasks.

Improving impaired balance function: real-time versus carry-over effects of prosthetic feedback

This study investigated whether training with realtime prosthetic biofeedback (BF) of trunk sway induces a carry-over improvement in balance control once BF is removed. 12 healthy older adults and 7 uncompensated unilateral vestibular loss patients were tested. All participants performed a battery of 14 balance and gait tasks (pre-test) upon their initial lab visit during which trunk angular sway was measured at L1-3. They then received balance BF training on a subset of 7 tasks, three times per week, for two consecutive weeks. BF was provided using a multi-modal biofeedback system with graded vibrotactile, auditory, and visual cues in relation to subject-specific angular displacement thresholds. Performance on the battery of the 14 balance and gait tasks (without BF) was re-assessed immediately after the 2 week training period, as well as 1 week later to examine BF carry-over effects. Significant reductions in trunk angular displacement were observed with the real-time BF, compared to the pre-test trials. The effects of BF persisted when BF was removed immediately after the final training session. BF carry-over effects were less evident at one week post-training. This evidence supports the potential short-term effects of BF training in a limited number of tasks after the BF is removed in healthy elderly subjects and those with vestibular loss. However, the prospect for longer term (>1 week) effects of prosthetic training on balance control remains currently unknown.

Trunk sway reductions in young and older adults using multi-modal biofeedback

This study investigated whether real-time biofeedback of angular trunk displacement could alter balance performance among healthy older and young adults. Healthy community-dwelling older adults (n=32) and healthy young adults (n=32) were included in the randomized control trial study. The intervention group received combined vibrotactile, auditory and visual biofeedback of angular trunk displacement in real-time during training on a battery of static and dynamic balance tasks and during the subsequent post-training balance re-assessment. The control group received balance training and were re-assessed in the absence of real-time biofeedback of their trunk displacement. The 90% range of angular trunk displacement was calculated for each balance task pre- and post-training. Significant age-related differences were observed independent of the intervention. Biofeedback intervention significantly changed the angular displacement of the trunk for both young and older participants on a number of balance tasks compared to control treatment (40-60% reduction in angular displacement). In some cases, biofeedback influenced balance in older adults, but not younger adults.

Responses to whole head-and-body tilts with and without passive ankle dorsiflexion in the absence of visual feedback

OBJECTIVES

We have investigated lower limb responses in seven blindfolded healthy subjects to well controlled tilts in the standing position. Our aims were (1) to determine the effect of head acceleration magnitude on responses evoked by whole head-and-body tilts, and (2) to establish whether tilt-evoked responses are modifiable by passive ankle dorsiflexion. Whole head-and-body tilts evoked responses in the biceps femoris, medial gastrocnemius and tibialis anterior muscles.

METHODS

Seven young healthy subjects stood on a spring-activated tilting apparatus and underwent sudden whole head-and-body tilts of about 15 degrees from the vertical position, with or without passive ankle dorsiflexion. Head acceleration was recorded with a linear accelerometer and ankle angular displacement with a potentiometer. Surface EMG signals were recorded in the right biceps femoris, medial gastrocnemius and tibialis anterior muscles.

RESULTS

As the peak of head acceleration was increased from 0.5 g to 1.8 g, the frequency of occurrence of tilt-evoked responses increased from 7% to 60% of trials in the biceps femoris muscle during whole head-and-body tilts. In general, the more proximal muscle (biceps femoris) was activated before the more distal muscle (medial gastrocnemius) during whole head-and-body tilts, while the opposite pattern was found during tilt with dorsiflexion.

CONCLUSIONS

Our results indicate that the occurrence of tilt-evoked responses increases with an increase in the amplitude of tilting acceleration. This suggests that tilt-evoked responses are dependent, at least in part, on vestibular stimulation. In addition, the spatio-temporal pattern of biceps femoris and medial gastrocnemius muscle activation was opposite during whole head-and-body tilts and tilts with dorsiflexion. This finding suggests that foot/ankle somatosensory inputs can modify tilt-evoked responses.