Hypothesis / aims of study
Despite similar external loads during running and no differences in pelvic floor muscle function (1,2), previous research showed that incontinent runners (IR) had lower ankle and greater hip energy absorption at a single time point (30 minutes) during standardized high-speed running, suggesting increased mechanical demands on pelvic structures (2). This study aims to investigate how ankle and hip energy absorption changes across multiple time points during self-selected speed running, and whether these differ between continent runners (CR) and IR. We hypothesize that IR will show impaired load management over time.
Study design, materials and methods
This cross-sectional study included women aged 18-45 with more than six months of running experience, no injuries or recent surgery in the lower limbs or pelvis, no pregnancy in the past year, and no prior physiotherapy treatment for urinary incontinence. IR had incontinence during running, whereas CR had no leakage on the International Consultation on Incontinence Questionnaire-Short Form. Participants ran for 25 minutes on an instrumented treadmill (Bertec, OH, USA) at a self-selected speed. Three-dimensional kinetics were recorded using a motion analysis system (Oqus 7, Qualisys, Sweden) at 5, 10, 20, and 25 minutes. Joint energy absorption was defined as the time integral of negative joint power during stance, calculated as net joint moment (normalized to body mass) × angular velocity at the ankle and hip. We assessed between-group differences in self-selected speed with an independent t-test and analyzed time and group effects on energy absorption using a two-way repeated-measures ANOVA (α = 0.05), with Bonferroni correction for post hoc comparisons.
Results
Thirty women were included (15 per group). Groups did not differ in self-selected speed (IR: 2.28m/s; CR:2.23m/s; p=0.687, d=0.14). The ANOVA revealed a significant main time effect only for ankle energy absorption (F=3.547, p=0.018, η²p=0.11), whereas no group effect was found (F=2.116, p=0.157, η²p=0.07). Post hoc analyses supported these results, revealing that the temporal effect at the ankle was driven by a difference between 10 and 25 min (p=0.014). At the hip, a group effect was observed (F=5.825, p=0.023, η²p=0.17), with greater energy absorption in IR than CR, indicating that the groups are different, regardless of time (F=0.514, p=0.674, η²p=0.02) (Figure 1). No significant time × group interactions were found (ankle: F=2.136, p=0.102, η²p=0.07; hip: F=1.470, p=0.229, η²p=0.05).
Interpretation of results
Despite no differences in self-selected running speed, IR showed greater hip energy absorption than CR across the entire run. Therefore, the load distribution observed in IR, with greater energy reaching the hip, may increase local mechanical demands on the pelvic floor (2). Although no statistically significant group differences were observed at the ankle, IR presented lower ankle absorption than CR at all time points, with moderate effect sizes. This pattern suggests a possible shift of load toward proximal structures (3), however, given the absence of statistical significance, these findings should be interpreted with caution. Thus, future studies should investigate other biomechanical factors that can be related to the increased demand observed in the pelvic area (3). Sustained over repetitive loading, this increased demand may exceed the capacity of the pelvic floor to respond adequately, potentially contributing to incontinence symptoms (2).