Air resistance modulates horizontal ground reaction forces during treadmill running

This study examined how air resistance alters ground reaction forces (GRFs) and stride characteristics during constant-speed treadmill running. Twenty-four trained male runners (sub-17 min 5 k equivalent or better) ran at two speeds (3.35 and 4.46 m·s^-1) under four airflow conditions: headwind (HW), tailwind (TW), no wind (NW), and no fan (NF). Aerodynamic drag was simulated using a custom fan-based system calibrated to match overground aerodynamic drag. In group-averaged data, HW versus NF increased propulsive impulse and reduced braking impulse at both speeds: 4.46 m·s^-1: Δpropulsion + 0.487 N·s (95 % CI 0.282-0.693); Δbraking + 0.817 N·s (0.638-0.997). 3.35 m·s^-1: Δpropulsion + 0.310 N·s (0.178-0.441); Δbraking + 0.583 N·s (0.452-0.713). TW and NW also showed smaller but consistent differences versus NF (Δpropulsion + 0.20-0.48 N·s; Δbraking + 0.39-0.69 N·s across speeds). Vertical GRFs and spatiotemporal parameters (contact time, step frequency, stride length) did not differ by airflow condition (all p > 0.245) but each scaled with speed (p < 0.001). Peak force effects by condition were limited. These experimental data support the hypothesis that trained runners overcome air resistance primarily by modulating anteroposterior (AP) GRFs, rather than altering vertical loading or spatiotemporal parameters; effects were more pronounced at the faster speed. Individual strategies varied, some runners emphasized greater propulsion, others reduced braking, yet the group-level pattern was consistent, increased propulsion and less braking under HW. Because belt speed was fixed, conclusions apply to scenarios in which pace is constrained. The observation that HW differs from NF emphasizes the relevance of simulating airflow in laboratory treadmill studies to better approximate outdoor running.