The Relationship Between Stroke-Related Cognitive Impairment and Robotic Proprioceptive Assessment

The interplay between proprioceptive and cognitive impairment following stroke is not well understood, along with the implications it may have on the design of robotic proprioception assessments. To examine the relationship between proprioceptive and cognitive impairments, we used a robotic device to measure ankle proprioception in 37 chronic stroke patients. Participants also completed the Montreal Cognitive Assessment (MoCA) and NIH stroke scale (NIHSS). We found that overall MoCA score was correlated with both active ($\mathrm{R}_{\mathrm{s}}=-0.32 \mathrm{p}=0.034$) and passive proprioception ability ($\mathrm{R}_{\mathrm{s}}=-0.52, \mathrm{p}=0.01$). When we examined the MoCA subscores, we found that the Visuospatial/Executive subscore was moderately correlated with measures of both active ($R_{s}=-0.57$, $\mathrm{p}=0$) and passive ($\mathrm{R}_{\mathrm{s}}=-0.42, \mathrm{p}=0.01$) ankle proprioception ability. When we removed 5 individuals who were identified as having unilateral neglect using the NIHSS, the correlations remained significant. The Attention subtest was also weakly correlated with active proprioception ability ($\mathrm{R}_{\mathrm{s}}=-0.36, \mathrm{p}=0.03$). Thus, proprioception ability is interrelated with specific cognitive impairments after stroke besides neglect. This suggests that cognitive tests should be conducted alongside robotic proprioceptive tests after stroke, to identify where cognitive factors may be contributing to measured proprioceptive error.