Author(s): LUCIANO, F., MESQUITA, R.M.2, CATAVITELLO, G.2,3, DEWOLF, A.H.2, NATALUCCI, V.1, PAVEI, G.1, WILLEMS, P.A.2, MINETTI, A.E.1, Institution: UNIVERSITY OF MILAN, Country: ITALY, Abstract-ID: 1093

INTRODUCTION:
The metabolic cost (C) of walking and running is reported to increase with headwind and decrease with tailwind (1). In analogy with gradient locomotion, air drag may increase C by impacting the mechanical work done on the body center of mass and the proportion between its positive and negative fractions (2-4), although limited evidence exists on the underlying mechanisms. Elucidating the relationship between wind speed and C would shed further light on the energy demands of overground locomotion, where people move through air, and the generalizability of treadmill studies, where relative wind speed is nil. This study aimed to assess how air drag affects the metabolic and mechanical demands of walking and running.
METHODS:
After sample size estimation, eight male endurance athletes (age: 32±6 y, mass: 63.2±6.6 kg; height: 1.77±0.05 m, PB 10000 m: 31:20±01:12 min:s) were recruited. Participants walked at 1.5 m/s and ran at 4.0 m/s on an instrumented treadmill in a wind tunnel. Wind speeds (v) ranged from −12.5 to 12.5 m/s, where negative and positive signs indicate wind from the back (tailwind) or the front (headwind) of participants, respectively. A portable metabograph measured steady-state gas exchanges, and an eight-camera optoelectronic system recorded the position of reflective markers on the main body segments. This allowed calculating C (J/(kg*m)), drag force (Fd, N/kg), internal kinetic mechanical work (Wintk), positive and negative external mechanical work (Wext+ and Wext−). Mixed-effects models regressed such variables over v and v^2: those with the lowest Akaike Information Criterion were reported with their fixed effects and t-values.
RESULTS:
Headwind increased C for walking (C=2.72−0.05*v+0.01*v^2; t_v=−1.5; t_v^2=3.1) and running (C=4.28−0.08*v+0.01*v^2; t_v=−3.3; t_v^2=6.4), while tailwind decreased it (C=2.72+0.12*v; t_v=14.7 and C=4.28+0.15v; t_v=19.3, respectively). Similarly, Wext+ increased with headwind and decreased with tailwind; the opposite was observed for Wext−, whereas Wintk was negligibly affected by wind. Across the whole range of wind speeds (−12.5 to 12.5 m/s), variations in C followed linearly those in Fd (C=2.52+1.9*Fd; t_Fd=17.2 for walking, and C=4.0+2.5*Fd; t_Fd=20.1 for running).
CONCLUSION:
Our study confirms that the C of walking and running increases with headwind and decreases with tailwind; variations in C have similar magnitude in walking and running, and parallel those in Fd. The relations between C, Wext+ and Wext− with a head- and tailwind align with those observed in uphill and downhill locomotion, respectively (2). As for this case, variations in C may be determined by the partitioning between positive and negative work, together with their different efficiencies.

REFERENCES:
(1) Davies, J Appl Physiol, 1980
(2) Minetti et al., J Physiol, 1993
(3) Mesquita et al., Eur J Appl Physiol, 2020
(4) Dewolf et al., Eur J Appl Physiol, 2020