Purpose: Purpose one was to characterize the consistency of intra-limb hip/knee (H/K) coordination according to a measure of average coefficient of correspondence (ACC) across strides. Purpose two was to investigate H/K ACC validity and ability to discriminate pre-/post-treatment change in stroke survivors.

Methods: Five healthy controls and 32 chronic (>12 mos) stroke survivors were enrolled, and H/K ACC was calculated for both groups. Comparison between controls and stroke was made using the Mann-Whitney Test. Convergent validity of H/K ACC was tested using the Pearson Correlation model with gait speed and the 6 min Walk Test (6MWT). Stroke survivors were randomized to either: (1) gait training with functional neuromuscular stimulation (FNS) using intramuscular (IM) electrodes or (2) gait training without FNS. Both groups had treatment 1.5 h/day, 5 days/week, for 12 weeks, including .5 h coordination exercise, .5 h body weight supported treadmill training (BWSTT), and .5 h over ground gait training. The FNS-IM group used FNS-IM for all treatment components; the No-FNS group did not. Pre-/post-treatment comparisons were made using ANOVA.

Results: H/K ACC detected a significant difference between controls versus stroke involved limb (p=.0001) and controls versus stroke uninvolved limb (p=.042). The H/K ACC measure was well-correlated with gait speed (r=.70) and 6MWT (r=.69). H/K ACC showed a significant treatment response to FNS-IM (p=.003), but not No-FNS (p=.747).

Conclusions: H/K ACC sensitively discriminated between controls versus stroke involved or uninvolved limbs. H/K ACC was valid, with significant correlations with both walking speed and 6MWT. FNS-IM produced a significant gain in H/K ACC, and No-FNS did not.