Evaluating Two Measures of Postural Stability in Response to Perturbations in People with Diabetic Neuropathy

Diabetic neuropathy is a common complication of diabetes. It is characterized by a marked decrease in proprioception, particularly in the lower body. This reduced proprioception leads to an increase in falls risk in this population. The purposes of this study were to compare two measures of postural stability, sway excursion and virtual time-to-contact, with disease severity in persons with diabetic neuropathy in response to anteroposterior and mediolateral support-surface perturbations and to identify the relationship between virtual time-to-contact and disease severity in persons with diabetic neuropathy in response to oblique support-surface perturbations. We hypothesized that virtual time-to-contact would provide a more sensitive and robust measure of postural stability for people with diabetic neuropathy in response to anteroposterior and mediolateral perturbations. We also hypothesized that as disease severity increased, postural stability would decrease in response to oblique perturbations. We expected a direct relationship between disease severity and sway excursion and an inverse relationship between disease severity and virtual time-to-contact.  Postural kinematics and force plate data were collected for ten adults with diabetes and a range of neuropathy from none to moderate-severe. Postural kinematics were collected using an eight-camera Qualisys motion capture system. Perturbations were controlled by and force plate data was collected using a NeuroCom Research Module. Each participant was perturbed in eight directions, at two speeds per direction (10 cm/sec and 20 cm/sec), and for two trials for each condition for a total of thirty two perturbations. Nine statistically significant correlations were found between disease severity and virtual time-to-contact, while one statistically significant correlation was found between disease severity and sway excursion. Interestingly, the nine correlations between disease severity and virtual time-to-contact were all positive correlations.  We had support for our first hypothesis in that virtual time-to-contact had more correlations with disease severity than sway excursion across anteroposterior and mediolateral perturbations at both perturbation speeds. Our second hypothesis was not supported, in that virtual time-to-contact increased in response to oblique perturbations as disease severity increased. On average, virtual time-to-contact can explain 58% of the variation in disease severity. Further research is needed into why virtual time-to-contact was unexpectedly directly correlated with disease severity of diabetic neuropathy.  M.S