What determines the metabolic cost of human running across a wide range of velocities?

The 'cost of generating force' hypothesis proposes that the metabolic rate during running is determined by the rate of muscle force development (1/t(c), where t(c)=contact time) and the volume of active leg muscle. A previous study assumed a constant recruited muscle volume and reported that the rate of force development alone explained similar to 70% of the increase in metabolic rate for human runners across a moderate velocity range (2-4 m s(-1)). We hypothesized that over a wider range of velocities, the effective mechanical advantage (EMA) of the lower limb joints would overall decrease, necessitating a greater volume of active muscle recruitment. Ten high-caliber male human runners ran on a force-measuring treadmill at 8, 10, 12, 14, 16 and 18 km h(-1) while we analyzed their expired air to determine metabolic rates. We measured ground reaction forces and joint kinematics to calculate contact time and estimate active muscle volume. From 8 to 18 km h(-1), metabolic rate increased 131% from 9.28 to 21.44 W kg(-1). t(c) decreased from 0.280 s to 0.190 s, and thus the rate of force development (1/t(c)) increased by 48%. Ankle EMA decreased by 19.7 +/- 11%, knee EMA increased by 11.1 +/- 26.9% and hip EMA decreased by 60.8 +/- 11.8%. Estimated active muscle volume per leg increased 52.8% from 1663 +/- 152 cm(3) to 2550 +/- 169 cm(3). Overall, 98% of the increase in metabolic rate across the velocity range was explained by just two factors: the rate of generating force and the volume of active leg muscle.