pelvic obliquity

Rumba turning involves continuous whole-body reorientation, but discrete kinematic summaries do not localise where consistent directional deviations occur across the turn. This exploratory study examined rotational-progress-dependent patterns of pelvic obliquity, thoracic frontal-plane angle, and right support-foot progression angle during the initial 360° Rumba spiral turn in highly trained adolescent female Latin dancers. Fifteen dancers aged 14–16 years completed five valid trials each (75 trials). Three-dimensional kinematics were recorded using a 10-camera Vicon system at 100 Hz. For each trial, accumulated pelvic rotation was used to define rotational progress, and the three outcome waveforms were resampled to 101 points across 0%–100% of the turn. The five trials from each participant were averaged point by point, yielding 15 participant-level representative waveforms per outcome. Two-tailed one-sample statistical parametric mapping (SPM{t}) tests identified intervals in which group mean waveforms differed from their corresponding Plug-in Gait model-defined 0° references, using a Bonferroni-adjusted primary α = 0.0167 across the three outcomes. Pelvic obliquity showed a positive deviation at 11.3%–47.2% and a negative deviation at 56.5%–97.3% of rotational progress. Thoracic frontal-plane angle showed a negative deviation at 13.2%–41.9% and positive deviations at 57.6%–83.0% and 85.7%–92.0%. Right support-foot progression angle showed negative, positive, and negative deviations at 0.0%–4.8%, 10.4%–49.1%, and 56.2%–100.0%, respectively. Sensitivity analyses without the additional across-outcome Bonferroni adjustment preserved the principal directional patterns, with only modest changes in cluster extent and continuity. These findings identify where, across rotational progress, the measured segmental orientations consistently departed from model-defined references during a standardised Rumba spiral turn, without implying normative targets of optimal dance technique.