Reliability of overground running measures from 2d video analyses in a field environment

Abstract

Two-dimensional running analyses are common in research and practice, and have been shown to be reliable when conducted on a treadmill. However, running is typically performed outdoors. Our aim was to determine the intra-and inter-rater reliability of two-dimensional analyses of overground running in an outdoor environment. Two raters independently evaluated 155 high-speed videos (240 Hz) of overground running from recreationally competitive runners on two occasions, seven days apart (test-retest study design). The reliability of foot-strike pattern (rear-foot, mid-foot, and fore-foot), foot-strike angle (◦ ), and running speed (m/s) was assessed using weighted kappa (κ), percentage agreement, intraclass correlation coefficient (ICC), typical error (TE), and coefficient of variation (CV) statistics. Foot-strike pattern (agreement = 99.4%, κ = 0.96) and running speed (ICC = 0.98, TE = 0.09 m/s, CV = 2.1%) demonstrated excellent relative and absolute reliability. Foot-strike angle exhibited high relative reliability (ICC = 0.88), but suboptimal absolute reliability (TE = 2.5◦, CV = 17.6%). Two-dimensional analyses of overground running outdoors were reliable for quantifying foot-strike pattern, foot-strike angle, and running speed, although foot-strike angle errors of 2.5◦ were typical. Foot-strike angle changes of less than 2.5◦ should be interpreted with caution in clinical settings, as they might simply reflect measurement errors.