Speed and cadence adaptations during overground sloped running under real-world conditions

Background: Biomechanical adaptations to sloped running have been widely studied in laboratory settings, but these are limited by artificial conditions and constrained speeds. Advances in wearable technology now allow for analysis of running biomechanics in real-world environments. Aims: This study examined the relationship between surface gradient, running speed, and cadence in recreational runners using field-based data. Methods: Data were extracted from the We-TRAC database, comprising GPS-enabled Garmin watch records. Runs were included if they spanned at least 5 km, featured elevation changes over 100 m, and averaged speeds above 1.2 m/s. Each run was segmented into 100 m intervals and categorized by slope: uphill (+ 3 to 15%), level (− 2 to + 2%), and downhill (− 3 to − 15%). Results: A total of 148 participants (3001 runs) were included. Uphill segments showed significantly reduced cadence and speed compared to level segments (p < 0.001, Cohen’s d = 0.30–1.13). Downhill segments were associated with significantly higher speed (p = 0.013, Cohen’s d = 0.213) but no change in cadence (p = 0.694). Within individual runners, increases in slope were associated with decreases in both cadence and speed during uphill running, though this pattern was less consistent during downhill running. Conclusion: These findings underscore slope-dependent adaptations in real-world running and highlight the utility of wearable data in capturing ecological biomechanics. Recreational runners naturally adjust their cadence and speed according to gradient, suggesting that training programs and wearable feedback systems should account for slope to better monitor performance and reduce injury risk.