To analyze pelvic asymmetry between the hemiplegic and non-hemiplegic sides in stroke patients and to analyze the progressive changes in pelvic asymmetry during continuous walking.

A total of 25 stroke patients admitted to the Neurology Department of Beijing Bo'ai Hospital, China Rehabilitation Research Center, from October 2020 to September 2022 were enrolled. All participants underwent three-dimensional gait analysis, during a task involving the maximum continuous walking distance. The entire gait cycle was divided into three phases: initial, middle, and terminal. The first three gait cycles were excluded from analysis. The pelvic angle (angle 1, A₁) in the initial phase was calcuated as the mean of the 4th to 6th gait cycles, while the terminal phase angle (A₃) represented the mean of the last three gait cycles before stopping. The middle phase angle (A₂) was defined as the mean of three consecutive pelvic angles between A₁ and A₃. Based on walking distance, patients were divided into an indoor walking group (50-200 m) and an outdoor walking group (>200 m). Differences in pelvic joint angles (X-axis: anterior/posterior tilt; Y-axis: up/down obliquity; Z-axis: internal/external rotation) between the hemiplegic and non-hemiplegic sides were analyzed. The progressive trends in pelvic joint angles during continuous walking were assessed, and differences in these trends between patients with varying walking abilities were examined. Inter-group comparisons were conducted using t-test or double-associated sample rank-sum test. The repeated measures ANOVA of variance was used to analyze the changing trends of the same index of the pelvic joint at different time points.

Significant differences were observed between the hemiplegic and non-hemiplegic sides in the following parameters: maximum pelvis anterior tilt angles [PTA_max;（31.19±0.67）° vs （31.55±0.70）°， t=2.614, P=0.011], maximum up obliquity angles [POA_max; 4.78°（3.33°，6.74°） vs 4.54°（1.77°，5.93°），Z=2.115, P=0.034], the range of motion (ROM) of pelvic tilt (PTA_ROM1) in the initial phase [（4.27±0.37）° vs （4.91±0.43）°，t=2.829, P=0.009], the ROM change of pelvic tilt between terminal and initial phases [PTA_ROM3-1; （0.50±0.15）° vs （-0.53±0.31）°，t=3.263, P=0.003]. During continuous walking, the hemiplegic side showed statistically significant progressive increases in both PTA_max (F=3.295, P=0.046) and PTA_ROM (F=4.374, P=0.018). Subgroup analysis revealed statistically significant progressive increases in PTA_max specifically in the indoor walking group (F=5.830, P=0.009), while significant progressive increases in PTA_ROM were observed in the outdoor walking group (F=3.500, P=0.046). The hemiplegic side demonstrated statistically significant progressive reductions in both maximum pelvic internal rotation (PRA_max; F=3.306, P=0.046) and the ROM of pelvic rotation (PRA_ROM; F=4.896, P=0.012), with the non-hemiplegic side also showing significant progressive decrease in PRA_ROM (F=4.454, P=0.017). Further analysis identified that the significant progressive reduction in hemiplegic-side PRA_max was particularly evident in the indoor walking group (F=3.484, P=0.048).

Stroke patients demonstrate asymmetric alterations in pelvic tilt and obliquity between the hemiplegic and non-hemiplegic sides during the entire gait cycle. During continuous walking, progressive changes occur in anterior pelvic tilt, range of motion of pelvic tilt, and internal rotation on the hemiplegic side, without corresponding symmetrical changes on the non-hemiplegic side. Patients with poorer walking ability demonstrate more significant changes in pelvic asymmetry during continuous walking.