Fig. 6. Results of multiple linear regression of muscle power output (W kg^–1) and the two predictor variables log₁₀ %MHC155 and PC1 TnT (see also Table 1), showing the relationship between actual muscle power output and that predicted by the two parameter model (adjusted r²=0.979, P=0.001, N=6) for healthy individuals (A). The same model explains only 38.3% (adjusted r²=0.295, P=0.034, N=17) of the variation in muscle power output when data for infected individuals (filled circles) are included (B).