Assessing lower limb position sense in stroke using the Gradient Discrimination Test (GradDT™) and Step-height Discrimination Test (StepDT™): a reliability and validity study

Purpose: To evaluate the psychometric properties of two novel tests of lower limb position sense. Methods: Our newly developed tests assess discrimination thresholds of under-foot slope and step height perception using a two alternative forced choice approach. Stroke participants (n=32) and age matched controls (n=32) were tested. Inter- and intra-rater reliability and agreement, sensitivity and specificity, discriminant and convergent validity were evaluated. Results: Intra-rater reliability for both variants of the gradient discrimination test was excellent: intraclass correlation coefficients (ICC) =0.91 and 0.89. The step height discrimination test had excellent intra-rater reliability and agreement: ICC=0.95. Inter-rater reliability was also excellent in both tests (ICC= 0.85-0.93). Discriminant validity was demonstrated with significant differences in test performance between stroke and control participants (p<0.001). Our novel tests did not significantly correlate with the proprioceptive component of the Erasmus modified Nottingham Sensory Assessment. Receiver Operating Characteristic curve analysis indicated both novel tests to have greater sensitivity and specificity than the proprioceptive component of the Erasmus modified Nottingham Sensory Assessment in predicting the presence of self-reported sensory impairments. Functional Reach Test, 10 metre walk test, Centre of Pressure measurement and reported falls showed significant and moderate to strong correlations with novel test performance (r=0.40-0.60); the Erasmus modified Nottingham Sensory Assessment did not. Conclusion(s): Our novel, functionally oriented tests of lower limb position sense are reliable, valid and feasible for use in an ambulatory chronic stroke and elderly population.