Functional principal component analysis of H-reflex recruitment curves.

The primary purpose of this study was to use functional principal component analysis (FPCA) to analyze Hoffman-reflex (H-reflex) recruitment curves. Smoothed and interpolated recruitment curves from 38 participants were used for analysis. Standard methods were used to calculate three discrete variables (i.e., H(max)/M(max) ratio, H(th), H(slp)). FPCA was then used to extract principal component functions (PCFs) from the processed recruitment curves. PCF scores were calculated to determine how much each PCF contributed to an individuals' recruitment curve. The analysis extracted three PCFs, and three sets of PCF scores. Correlation analyses and systematic variation in the PCF scores indicated that the scores for the first PCF were primarily correlated to H-reflex threshold (H(th)) and that the scores for the second and third PCFs were correlated to H-reflex magnitude (H(max)/M(max) ratio) and slope (H(slp)), respectively. In addition, results from the FPCA indicated that the first PCF explained 56.0% of the variance between all H-reflex recruitment curves, whereas the second and third PCFs explained 24.1% and 13.0%, respectively. The high correlations indicate FPCA-derived PCFs capture similar physiological information as the standard discrete variables and suggest that application of FPCA to H-reflex recruitment curves could be used in future studies to complement traditional analyses that investigate excitability of the motoneuron pool.