Humans trade-off energetic cost with fatigue avoidance while walking

Metabolic cost minimization is widely regarded as the principal optimality criterion that governs walking. Minimizing muscle activation has, nevertheless, outperformed energy optimization in simulating human gait and predicting certain gait behaviors. The highly coupled nature of metabolic and muscle activation costs makes it difficult to empirically discern the interrelationship between these objectives. We implemented a unique experimental design that pits metabolic cost against muscle activation costs estimated from electromyography of seven lower limb muscles. Healthy adults (N=10) selected between walking on a treadmill incline versus walking in a crouched posture (that disproportionately affected activation cost), forcing a choice between minimizing metabolic cost or activation cost. When experiencing these Competing-Cost-Pairs, participants systematically protected their activation cost at the expense of high metabolic power (19% penalty, p<0.05). This held true when activation cost was expressed as the sum of the muscle activations squared (66% saving, p<0.05) and as the maximal activation across muscles (44% saving, p<0.05), both of which penalize overburdening any individual muscle and thus indicate fatigue avoidance. Activation cost, expressed as the sum of muscle volume-normalized activation, more closely models energy use and was also protected by the participants' decision (23% saving, p<0.05) demonstrating that activation was, at best, an inaccurate proxy signal for metabolic energy. Energy minimization was only observed when there was no adverse effect on muscle activation. By decoupling metabolic and activation costs, we provide the first empirical evidence of humans embracing non-energetic optimality in favor of a clearly defined alternate neuromuscular objective. ### Competing Interest Statement The authors have declared no competing interest.