Table 1.

Relationship between lower-limb joint and muscle-level neuromechanics and whole-body metabolic rate

Gait Cycle Phase	1.25 m s− 1	1.50 m s− 1	1.75 m s− 1
Δ Exoskeleton Positive Work Rate (J kg − 1 s− 1)	p = 0.3887	p = 0.1490	p = 0.0703
Δ Biological Positive Work Rate (J kg − 1 s− 1)	p = 0.0099, R2 = 0.47	p = 0.0730	p = 0.1107
Δ Biological Moment Rate (N kg − 1 s− 1)	p = 0.6574	p = 0.0502	p = 0.1273
Δ Soleus + Tibialis Anterior Activation Rate (unitless s− 1)	p = 0.0009, R2 = 0.56	p = 0.0497, R2 = 0.63	p = 0.0002, R2 = 0.69

The relationships between changes in ankle joint and muscle-level neuromechanical rate metrics (x) and the change in net metabolic rate (W kg^− 1) (y) during walking with exoskeletons over a range of walking speeds and stiffness values, k_exo. For each metric, Δ represents the difference from the k_exo = 0 condition (i.e., the effect of increasing exoskeleton stiffness) with positive values indicating an increase. We used a within-participant linear regression analysis to determine whether relationships were significant (*p < 0.05), and we report the line of best fit to the data and the R² value of the fit. Data showing actual lines of best fit can be found in Fig. 6. Note that work rate (J s^− 1) and power (W) are synonymous