Postural control characteristics as reflected by center of pressure measurements in anterior cruciate ligament reconstruction patients at varying post-operative stages: a cross-sectional comparative study

Abstract

Objective

To compare center of pressure (COP) characteristics and postural control changes in anterior cruciate ligament reconstruction (ACLR) patients across distinct post-operative phases and to establish quantitative evidence for post-operative functional assessment.

Methods

This cross-sectional controlled study enrolled participants prospectively. One hundred ACLR patients (n = 100) were stratified into five post-operative cohorts: Group A (1.5–3 months, n = 20), Group B (3–6 months, n = 20), Group C (6–9 months, n = 20), Group D (9–12 months, n = 20), and Group E (≥ 12 months; mean ± SD: 16.9 ± 3.3 months, n = 20). A control group of 20 healthy individuals without history of lower limb injury (Group F) was included. All ACLR patients underwent standardized rehabilitation protocols according to institutional guidelines. The Zebris plantar pressure analysis system was used to measure center of pressure trajectory parameters at self-selected comfortable walking speed, including gait line length, single-limb support line length, anteroposterior displacement, and mediolateral displacement. Limb symmetry indices were calculated by comparing the operative and non-operative limbs, and postural control differences across post-operative stages were determined through between-group comparisons.

Results

Gait line length and single-limb support line length were significantly shorter on the operative limb compared with the non-operative limb (P < 0.05), whereas no significant bilateral differences were observed in the control group (P >  0.05). Between-group comparisons revealed significant differences in center of pressure parameters across the post-operative stages (P < 0.05). Gait line length in Groups A through E was reduced by 30.0%, 15.3%, 11.1%, 6.8%, and 2.0%, respectively, relative to the control group. Single-limb support line length demonstrated corresponding reductions of 83.4%, 54.0%, 34.8%, 20.6%, and 16.1%, respectively, from Groups A through E. Anteroposterior displacement in Groups B and C was significantly greater than in the control group, whereas mediolateral displacement in Groups A, B, and C was significantly greater than in the control group (P < 0.01).

Conclusion

Patients in the post-operative ACLR cohorts demonstrated abnormal center of pressure parameters and significant functional asymmetry between limbs. Center of pressure parameters across all post-operative timepoint groups were lower than in the healthy control group, indicating that center of pressure-based postural control assessment provides objective quantitative metrics for post-operative functional evaluation.

Introduction

Anterior cruciate ligament reconstruction (ACLR) is the standard surgical treatment for complete anterior cruciate ligament rupture, effectively restoring the anatomical structure and mechanical stability of the knee joint [1–5]. However, postural control dysfunction persists post-operatively, reflecting the complexity of the physiological recovery process [6–9]. The anterior cruciate ligament contains numerous mechanoreceptors and serves as a critical source of knee joint proprioceptive feedback [10, 11]. Ligamentous injury results in loss of mechanoreceptors, disruption of proprioceptive feedback, and impaired neuromuscular control [12, 13]. Moreover, multiple factors—including slow graft reinnervation, adaptive changes in periarticular soft tissues, and compensatory reorganization of the central nervous system—perpetuate the recovery process [14–22]. Although standardized rehabilitation protocols incorporating strength training, proprioceptive training, and functional training are implemented, a considerable proportion of patients continue to demonstrate persistent postural control dysfunction following rehabilitation completion, which compromises athletic performance and increases re-injury risk [23–27].

Center of pressure (COP) trajectory analysis represents an objective methodology for quantifying postural control during gait [28]. COP represents the projection of the body’s center of mass onto the supporting surface; dynamic variations in COP trajectory reflect the neuromuscular system’s capacity for balance regulation. Compared with strength testing and subjective rating scales, COP parameters—including gait line length, single-limb support line length, anteroposterior displacement, and mediolateral displacement—provide objective quantitative characterization of postural control subtleties [29–31]. Extant literature predominantly examines COP parameters at single post-operative timepoints, with limited systematic comparison across different post-operative stages [32–34]. Elucidating COP parameter characteristics across different post-operative periods is essential for identifying critical time windows in functional recovery, optimizing staged rehabilitation protocols, and establishing evidence-based return-to-sport criteria [8, 34].

Therefore, this study employed a cross-sectional multi-timepoint controlled design utilizing the Zebris plantar pressure analysis system to measure COP parameters in ACLR patients across distinct post-operative phases. Systematic comparisons between each post-operative stage and the healthy control group were conducted to provide objective evidence for post-operative functional assessment, optimization of rehabilitation strategies, and establishment of evidence-based return-to-sport criteria.

Methods

Study design

This study employed a cross-sectional controlled design in accordance with the STROBE reporting guidelines. The research protocol received approval from the Medical Ethics Committee of Binzhou Medical University Hospital (Approval No.: KYLL-239) and conforms to the ethical principles outlined in the Declaration of Helsinki. All participants provided written informed consent after receiving complete information regarding study objectives and procedures. Data collection was performed in the Biomechanics Laboratory, Department of Rehabilitation Medicine, Binzhou Medical University Hospital, between August 2024 and August 2025.

Sample size calculation

Sample size was estimated based on a meta-analysis by Yu et al. [35], which reported medium effect sizes (SMD = 0.47) for center of pressure parameters in ACLR populations. Using G*Power software (version 3.1.9.7) with a medium effect size (Cohen’s f = 0.35), α = 0.05, and 90% power for six-group comparison, the minimum sample size was 108 participants. Accounting for 10% dropout, the final sample size was 120 participants (20 per group).

Participants

Patients who underwent anterior cruciate ligament reconstruction (ACLR) were recruited from the outpatient orthopedic clinic, Department of Rehabilitation Medicine, and inpatient wards of Binzhou Medical University Hospital using consecutive sampling. Initial screening was performed through electronic medical record review, with attending physicians assessing post-operative timepoint and functional status, followed by detailed evaluation of inclusion and exclusion criteria by the research team. All patients underwent arthroscopic single-bundle reconstruction using autogenous hamstring tendon (semitendinosus and gracilis tendons) graft and received standardized post-operative rehabilitation at our institution. The rehabilitation program comprised range of motion exercises, progressive resistance strengthening, balance and proprioceptive training, core stability exercises, and functional training, with individualized progression based on functional recovery. To ensure homogeneity in post-operative management, patients who did not receive rehabilitation at our institution were excluded. A total of 100 patients were stratified into five post-operative cohorts according to time since surgery: Group A (1.5–3 months, n = 20), Group B (3–6 months, n = 20), Group C (6–9 months, n = 20), Group D (9–12 months, n = 20), and Group E (≥ 12 months; mean ± SD: 16.9 ± 3.3 months, n = 20). These time intervals were designed to capture the continuum of functional recovery following ACLR, informed by previous studies demonstrating variations in neuromuscular and proprioceptive function across different post-operative stages [36, 37]. Group allocation was determined solely by time elapsed since surgery at enrollment. A healthy control group (Group F, n = 20) was recruited through community announcements and the hospital health examination center, matched for age, sex, and BMI with the patient cohorts.

• Inclusion criteria:

  (1) age between 18 and 45 years; (2) unilateral complete anterior cruciate ligament rupture confirmed by magnetic resonance imaging and arthroscopic assessment; (3) arthroscopic single-bundle reconstruction using autogenous hamstring tendon graft; (4) standardized post-operative rehabilitation received at Binzhou Medical University Hospital; (5) capable of independent ambulation for ≥ 10 min without obvious gait deviation; (6) absence of marked knee joint swelling; (7) Visual Analog Scale pain score ≤ 3; (8) normal cognitive function with ability to comprehend and complete study instructions.

• Exclusion criteria:

  (1) bilateral anterior cruciate ligament injury or history of contralateral knee pathology; (2) concomitant procedures including grade III–IV cartilage damage treatment, meniscectomy > 1/3 of meniscus, or meniscus repair; (3) prior lower extremity injury or previous knee surgery; (4) post-operative complications such as infection, joint stiffness, or deep vein thrombosis; (5) neurological disorders affecting postural control or vestibular dysfunction; (6) body mass index > 30 kg/m²; (7) spinal, hip, or ankle joint pathology; (8) vigorous exercise within 48 h prior to testing or concurrent use of medications affecting postural control.

Data collection

The Zebris Medical FDM-T plantar pressure analysis system (Zebris Medical GmbH, Germany) was used for assessment. The system features a high-precision pressure-sensitive platform (dimensions: 212.2 × 60.5 × 2.5 cm) containing 11,360 capacitive pressure sensors (density: 1.4 sensors/cm²), with a sampling frequency of 100 Hz, pressure measurement range of 1–120 N/cm², and measurement accuracy of ± 5%. Accompanying WinFDM-T software (version 2.5.2) calculates center of pressure trajectory coordinates in real-time based on plantar pressure distribution data.

Testing was performed in a quiet, well-lit biomechanics laboratory with a 6-meter straight walking corridor. The pressure platform was positioned at the corridor center with 2-meter acceleration and deceleration zones at each end to ensure participants achieved stable gait within the data acquisition area. The system underwent standardized calibration prior to all testing. Participants wore light athletic clothing, walked barefoot, rested in a seated position for 10 min, and then performed 5 min of adaptive walking on the pressure platform to familiarize themselves with the testing environment. During formal testing, participants walked continuously back and forth along the corridor at their self-selected comfortable walking speed for 60 s. To control for speed effects, optical timing systems measured the self-selected walking speed during the adaptation phase, and formal testing required maintenance of this speed within ± 5% variation. Gait quality and speed were monitored in real-time during testing; if gait deviation or speed variance exceeding 10% occurred, the test was terminated and repeated to ensure acquisition of at least 20 complete, stable gait cycles. All testing was performed by the same trained rehabilitation medicine specialist to ensure standardized testing procedures (Fig. 1).

Fig. 1.

Data collection setup and procedure

The system software employed pressure threshold methodology (> 5 N/cm² defined as heel contact; ≤ 5 N/cm² defined as toe-off) to automatically identify gait events and delineate gait cycles and support phases. The system automatically separated bilateral center of pressure trajectories based on the lateral coordinate and extracted center of pressure parameters during the support phase for the operative and non-operative limbs separately. For each participant, the first and last 2 gait cycles of each recording were excluded, and at least 15 consecutive stable gait cycles from the middle walking phase were selected for analysis. The software automatically removed abnormal gait cycles based on the following criteria: gait cycle time deviation > 3 standard deviations from the mean, or support phase duration < 0.3 s or > 1.2 s. Mean values for each parameter across all valid gait cycles were calculated and used as representative measurements.

Outcome measures

Four center of pressure trajectory parameters were selected as primary indices of postural control function. These parameters have been widely applied in post-operative ACLR functional assessment and characterize postural control across multiple dimensions during ambulation. The system software automatically identified the support phase for each gait cycle and calculated corresponding parameters; the final measurement value was derived from the mean of at least 15 valid gait cycles.

1. Length of gait line: The cumulative displacement distance of the center of pressure from heel contact to toe-off during the support phase, reflecting anteroposterior postural control trajectory (Fig. 2) [29, 38].

2. Single limb support line: The cumulative trajectory length of the center of pressure during the single-limb support phase, reflecting postural control capacity during unilateral weight-bearing (Fig. 2) [29, 39].

3. Anterior/posterior position: The maximum excursion amplitude of the center of pressure in the sagittal plane during the support phase, reflecting the dynamic range of anteroposterior postural control (Fig. 2) [29, 40].

4. Lateral symmetry (mm): The maximum excursion amplitude of the center of pressure in the coronal plane during the support phase, reflecting the dynamic range of mediolateral postural control (Fig. 2) [29, 41].

Fig. 2.

Dynamic balance parameters measured by COP trajectory analysis. Note: Schematic representation of a Length of gait line, b Single limb support line, c Anterior/posterior position, and d Lateral symmetry

Statistical analysis

Data analysis was performed using SPSS software (version 27.0; IBM Corp., Armonk, NY, USA), with continuous variables presented as mean ± standard deviation (SD). The normality of data distribution was assessed using the Shapiro-Wilk test. For normally distributed data, paired t-tests were used to compare operative and non-operative limbs, and one-way analysis of variance (ANOVA) was performed for between-group comparisons; when variances were unequal, Welch correction was applied, and the least significant difference (LSD) method was used for post-hoc pairwise comparisons. For non-normally distributed data, Wilcoxon signed-rank tests were used to compare operative and non-operative limbs, and Kruskal-Wallis H tests were performed for between-group comparisons, with post-hoc pairwise comparisons conducted using Bonferroni-corrected Mann-Whitney U tests. All statistical tests were two-tailed with a significance level of α = 0.05, and statistical significance was defined as P < 0.05.

Results

Baseline characteristics

Age, height, weight, sex, affected side distribution, and Visual Analog Scale (VAS) pain scores were comparable across all groups (P > 0.05). Walking speed differed significantly among the six groups (F = 109.51, P < 0.01). Group A demonstrated the lowest walking speed (0.99 ± 0.41 m/s), which was significantly lower than Groups C, D, E, and F. Group B (1.99 ± 0.18 m/s) was significantly lower than Groups D, E, and F (P < 0.05). Group C (2.43 ± 0.11 m/s) was significantly lower than Groups E and F (P < 0.05). No significant differences were observed among Groups D, E, and F (Table 1).

Table 1.

Comparison of baseline characteristics among groups

Variables	Group A	Group B	Group C	Group D	Group E	Group F	F/χ2	P
Age (years)	32.80 ± 8.043	33.10 ± 6.63	30.25 ± 7.49	30.55 ± 8.18	31.65 ± 8.28	33.75 ± 7.59	0.68	0.64
Height (cm)	168.60 ± 9.83	167.25 ± 9.25	171.80 ± 9.91	173.05 ± 9.59	171.80 ± 9.01	171.55 ± 9.05	1.12	0.35
Weight (kg)	69.20 ± 14.59	68.05 ± 15.68	71.70 ± 13.88	67.70 ± 14.67	69.45 ± 14.10	70.40 ± 15.79	0.20	0.96
Walking speed (m/s)	0.99 ± 0.41cdef	1.99 ± 0.18def	2.43 ± 0.11ef	2.98 ± 0.21	3.58 ± 0.24	3.55 ± 0.55	109.51	< 0.01
VAS	1.85 ± 0.67	1.65 ± 0.59	1.45 ± 0.51	1.35 ± 0.49	1.20 ± 0.52	–	8.46	0.08
Sex (M/F)	12/8	13/7	9/11	13/7	7/13	10/10	5.89	0.32
Affected side (L/R)	8/12	13/7	9/11	6/14	8/12	–	5.44	0.25

p < 0.05; c indicates statistically significant difference compared with Group C, p < 0.05; d indicates statistically significant difference compared with Group D, p < 0.05; e indicates statistically significant difference compared with Group E, p < 0.05; f indicates statistically significant difference compared with Group F, p < 0.05

Gait line length

Gait line length was significantly shorter on the operative limb compared with the non-operative limb across all patient groups (P < 0.05). In Group A, operative limb measurements were 155.66 ± 34.40 mm versus non-operative limb 181.27 ± 23.14 mm (P < 0.01; d = − 0.87). Group B demonstrated operative limb values of 188.32 ± 24.98 mm versus non-operative limb 210.58 ± 15.39 mm (P < 0.01; d = − 1.07). Group C demonstrated operative limb values of 197.60 ± 23.50 mm versus non-operative limb 209.53 ± 20.78 mm (P = 0.03; d = − 0.54). Group D demonstrated operative limb values of 207.27 ± 21.22 mm versus non-operative limb 218.16 ± 15.61 mm (P < 0.01; d = − 0.58). Group E demonstrated operative limb values of 217.82 ± 17.73 mm versus non-operative limb 224.56 ± 18.21 mm (P < 0.01; d = − 0.38). Group F demonstrated no significant bilateral differences (222.29 ± 15.87 mm vs. 222.51 ± 15.94 mm; P = 0.89; d = 0.02) (Table 2; Fig. 3).

Table 2.

Comparison of gait line length and single-limb support line length between operative and non-operative limbs in ACLR patients across post-operative stages and healthy controls

Parameter	Group	Affected side/Left side		Unaffected side/Right side	t/z	P	Cohen’s d
Length of gait line (mm)	Group A	155.66 ± 34.40		181.27 ± 23.14	− 2.84	< 0.01	− 0.87
	Group B	188.32 ± 24.98		210.58 ± 15.39	− 4.63	< 0.01	− 1.07
	Group C	197.60 ± 23.50		209.53 ± 20.78	− 2.33	0.03	− 0.54
	Group D	207.27 ± 21.22		218.16 ± 15.61	− 3.64	< 0.01	− 0.58
	Group E	217.82 ± 17.73		224.56 ± 18.21	− 2.99	< 0.01	− 0.38
	Group F	222.29 ± 15.87		222.51 ± 15.94	− 0.14	0.89	0.02
Single limb support line (mm)	Group A	19.05 ± 14.31		38.06 ± 19.29	− 3.49	< 0.01	− 1.12
	Group B	52.67 ± 17.18		83.00 ± 14.04	− 7.39	< 0.01	− 1.93
	Group C	74.62 ± 20.87		87.38 ± 22.26	− 2.71	0.01	− 0.59
	Group D	90.87 ± 15.40		103.49 ± 12.56	− 3.35	< 0.01	− 0.90
	Group E	96.01 ± 22.96		103.94 ± 15.20	− 2.25	0.04	− 0.41
	Group F	114.42 ± 11.13		114.60 ± 11.11	− 0.13	0.90	− 0.02

Fig. 3.

Intragroup comparison of length of gait line between affected and unaffected sides across different rehabilitation stages post-ACLR. Note: * indicates statistically significant difference, p < 0.05; ** indicates statistically significant difference, p < 0.01

Between-group comparisons of operative limb measurements revealed significant differences (F = 50.81, P < 0.01; η² = 0.48). Compared with Group F, Groups A, B, and C demonstrated significantly shorter operative limb measurements (P < 0.05), whereas Groups D and E showed no significant differences (P > 0.05). (Fig. 4) Regarding non-operative limb measurements, significant between-group differences were observed (F = 43.08, P < 0.01; η² = 0.36). Group A’s non-operative limb measurements were significantly shorter than all other groups (P < 0.05), whereas Groups B, C, D, and E demonstrated no significant differences from Group F (P > 0.05) (Table 3; Fig. 5).

Fig. 4.

Intragroup comparison of single limb support line between affected and unaffected sides across different rehabilitation stages post-ACLR. Note: * indicates statistically significant difference, p < 0.05; ** indicates statistically significant difference, p < 0.01

Table 3.

Between-group comparisons of center of pressure parameters in ACLR patients across post-operative stages and healthy controls

Parameter	Group A	Group B	Group C	Group D	Group E	Group F	F/H	P	η2
Length of gait line(affected side)	155.66 ± 34.40cdef	188.32 ± 24.98ef	197.60 ± 23.50f	207.27 ± 21.22	217.82 ± 17.73	222.29 ± 15.87	50.81	< 0.01	0.48
Length of gait line (unaffected side)	181.27 ± 23.14bcdef	210.58 ± 15.39	209.53 ± 20.78	218.16 ± 15.61	224.56 ± 18.21	221.90 ± 23.75	43.08	< 0.01	0.36
Single limb support line (affected side)	19.05 ± 14.31cdef	52.67 ± 17.18def	74.62 ± 20.87f	90.87 ± 15.40	96.01 ± 22.96	114.42 ± 11.13	87.75	< 0.01	0.77
Single limb support line (unaffected side)	38.06 ± 19.29cdef	83.00 ± 14.04def	87.38 ± 22.26f	103.49 ± 12.56	103.94 ± 15.20	114.60 ± 11.11	75.96	< 0.01	0.71
Anterior/posterior position	8.99 ± 7.09	14.26 ± 6.81f	15.83 ± 10.63f	8.98 ± 3.64	8.65 ± 4.78	7.41 ± 4.22	20.28	< 0.01	0.19
Lateral symmetry	25.94 ± 18.13ef	21.41 ± 13.26ef	14.20 ± 12.00f	11.27 ± 7.67	7.37 ± 6.15	3.72 ± 2.62	35.79	< 0.01	0.33

a indicates statistically significant difference compared with Group A, p< 0.05; b indicates statistically significant difference compared with Group B, p< 0.05; c indicates statistically significant difference compared with Group C, p< 0.05; d indicates statistically significant difference compared with Group D, p< 0.05; e indicates statistically significant difference compared with Group E, p< 0.05; f indicates statistically significant difference compared with Group F, p< 0.05

Fig. 5.

Intergroup comparison of length of gait line across different rehabilitation stages post-ACLR. Note: b indicates statistically significant difference compared with Group B, p< 0.05; c indicates statistically significant difference compared with Group C, p< 0.05; d indicates statistically significant difference compared with Group D, p< 0.05; e indicates statistically significant difference compared with Group E, p< 0.05; f indicates statistically significant difference compared with Group F, p< 0.05

Single support line length

Single-limb support line length was significantly shorter on the operative limb compared with the non-operative limb across all patient groups (P < 0.05). In Group A, operative limb measurements were 19.05 ± 14.31 mm versus non-operative limb 38.06 ± 19.29 mm (P < 0.01; d = − 1.12). Group B demonstrated operative limb values of 52.67 ± 17.18 mm versus non-operative limb 83.00 ± 14.04 mm (P < 0.01; d = − 1.93). Group C demonstrated operative limb values of 74.62 ± 20.87 mm versus non-operative limb 87.38 ± 22.26 mm (P = 0.01; d = − 0.59). Group D demonstrated operative limb values of 90.87 ± 15.40 mm versus non-operative limb 103.49 ± 12.56 mm (P < 0.01; d = − 0.90). Group E demonstrated operative limb values of 96.01 ± 22.96 mm versus non-operative limb 103.94 ± 15.20 mm (P = 0.04; d = − 0.41). Group F demonstrated no significant bilateral differences (114.42 ± 11.13 mm vs. 114.60 ± 11.11 mm; P = 0.90; d = − 0.02) (Table 2; Fig. 4).

Between-group comparisons of operative limb measurements revealed significant differences (F = 87.75, P < 0.01; η² = 0.77). Compared with Group F, Groups A, B, and C demonstrated significantly shorter operative limb measurements (P < 0.05), whereas Groups D and E showed no significant differences (P > 0.05). Regarding non-operative limb measurements, significant between-group differences were observed (F = 75.96, P < 0.01; η² = 0.71). Groups A, B, and C demonstrated significantly shorter non-operative limb measurements than Group F (P < 0.05), whereas Groups D and E showed no significant differences from Group F (P > 0.05) (Table 3; Fig. 6).

Fig. 6.

Intergroup comparison of single limb support line across different rehabilitation stages post-ACLR. Note: b indicates statistically significant difference compared with Group B, p< 0.05; c indicates statistically significant difference compared with Group C, p< 0.05; d indicates statistically significant difference compared with Group D, p< 0.05; e indicates statistically significant difference compared with Group E, p< 0.05; f indicates statistically significant difference compared with Group F, p< 0.05

Anteroposterior displacement

Between-group comparisons revealed significant differences in anteroposterior displacement (F = 20.28, P < 0.01; η² = 0.19). Compared with the control group (7.41 ± 4.22 mm), Groups B (14.26 ± 6.81 mm) and C (15.83 ± 10.63 mm) demonstrated significantly greater anteroposterior displacement (P < 0.05), whereas Groups A (8.99 ± 7.09 mm), D (8.98 ± 3.64 mm), and E (8.65 ± 4.78 mm) showed no significant differences from the control group (P > 0.05) (Table 3; Fig. 7).

Fig. 7.

Intergroup comparison of anterior/posterior position and lateral symmetry across different rehabilitation stages post-ACLR. Note: e indicates statistically significant difference compared with Group E, p< 0.05; f indicates statistically significant difference compared with Group F, p< 0.05

Mediolateral displacement

Between-group comparisons revealed significant differences in mediolateral displacement (F = 35.79, P < 0.01; η² = 0.33). Compared with the control group (3.72 ± 2.62 mm), Groups A (25.94 ± 18.13 mm), B (21.41 ± 13.26 mm), and C (14.20 ± 12.00 mm) demonstrated significantly greater mediolateral displacement (P < 0.05), whereas Groups D (11.27 ± 7.67 mm) and E (7.37 ± 6.15 mm) showed no significant differences from the control group (P > 0.05) (Table 3; Fig. 7).

Discussion

This study demonstrated that ACLR patients exhibited substantially diminished postural control on the operative limb compared with the non-operative limb during gait, whereas healthy controls demonstrated no significant bilateral differences, confirming operative-limb-specific functional deficits in the post-operative period. Both gait line length and single-limb support line length on the operative limb were shorter than on the non-operative limb, with asymmetry persisting beyond 12 months post-operatively, reflecting functional asymmetry as a persistent, long-term clinical concern. This finding may be attributed to multiple factors. First, the anterior cruciate ligament contains numerous mechanoreceptors, and their loss disrupts knee joint proprioceptive feedback, thereby impairing postural reflex regulation at both the spinal and cortical levels [37, 42–44]. Second, ACLR patients demonstrate abnormal neuromuscular activation patterns that are closely associated with postural control asymmetry [45–47]. Furthermore, selective quadriceps atrophy in the post-operative period weakens the extensor chain function of the operative limb, whereas alterations in joint capsule and ligamentous structures further compromise passive joint stability [48, 49]. The integrated changes in peripheral sensory input, neuromuscular function, and joint structure collectively result in diminished postural control capacity during operative-limb weight-bearing. Notably, some patients demonstrated abnormalities on the non-operative limb as well, suggesting that adaptive changes may occur as the non-operative limb compensates for operative-limb dysfunction, or that interactive effects between bilateral limb neural control mechanisms exist [49–51] .

Gait line length and single-limb support line length reflect postural control characteristics during ambulation. Across different post-operative timepoint groups, operative-limb values for both parameters were lower than non-operative limb and control group measurements, with the 1.5–3 months post-operative cohort demonstrating lower values than the ≥ 9 months post-operative cohort, and single-limb support line length demonstrating greater between-group variability than gait line length. From a neuromuscular control perspective, parameter reduction may be correlated with quadriceps insufficiency [52]. The quadriceps muscle is critical for knee joint stability during the single-limb support phase; surface electromyography (EMG) studies demonstrate delayed quadriceps activation timing and reduced activation amplitude in ACLR patients, abnormalities that persist at 12 months post-operatively [36]. Patients may reduce single-limb support time duration and limit center of mass displacement magnitude to decrease demands on neuromuscular strength and motor control precision [53]. Gait line length reduction reflects a conservative postural control strategy to maintain overall gait stability. Previous longitudinal research has demonstrated that gait biomechanical alterations following ACLR do not resolve in a simple linear pattern. Erhart-Hledik et al. tracked knee joint mechanics from 2 to 8 years post-ACLR and found that kinematic and kinetic parameters exhibited dynamic temporal changes, with gait asymmetries at 2 years post-operatively significantly associated with deteriorating patient-reported outcomes at 8-year follow-up [54]. Our findings demonstrate asymmetry present in all post-operative cohorts with varying magnitudes, potentially attributable to differences in sample characteristics, assessment methodologies, and group stratification approaches. Although walking speed demonstrated between-group differences, operative-non-operative asymmetry was significant across all cohorts, and the asymmetry magnitude pattern did not completely correspond to speed differences, suggesting that asymmetry likely reflects inherent postural control characteristics independent of gait speed.

The changing patterns of anteroposterior and mediolateral displacements revealed differential characteristics of directional postural control. Anteroposterior displacement increased in the 3–9 months post-operative cohort and approximated control group levels in the 1.5–3 months and ≥ 12 months post-operative cohorts. Mediolateral displacement increased in the 1.5–6 months post-operative cohort and approached normal control levels by 9 months post-operatively. The differentiated changes in anteroposterior versus mediolateral displacement reflect distinct recovery trajectories for sagittal and coronal plane postural control. Previous research indicates that patients at 1–6 months and beyond 6 months post-operatively both demonstrate abnormal joint position sense, with 30° joint position more severely impaired than 60° position, reflecting persistent proprioceptive function deficits [37]. When proprioceptive function remains incomplete, the central nervous system may compensate for insufficient sensory information by enlarging the oscillation range in the sagittal or coronal planes; this strategy is primarily dependent on enhanced ankle joint control and visual-vestibular system compensation [8, 34, 55]. Research has demonstrated differing running characteristics between soccer athletes less than 9 months and at least 9 months post-operatively, with the former demonstrating greater vertical force asymmetry and the latter showing reduced asymmetry, consistent with the trend of mediolateral displacement reduction observed in the present study as post-operative time increased, suggesting progressive improvement in coronal plane lateral stability [56]. However, the increase in anteroposterior displacement during the 3–9 months post-operative period indicates that sagittal plane postural control strategy adjustments follow a distinct temporal trajectory. The increase in mediolateral displacement in the early post-operative period carries clinical significance, potentially indicating insufficient lateral stability control and an elevated re-injury risk. Previous research found enlarged mediolateral center of pressure range during single-leg squat tasks at 12 months post-operatively, with lateral stability deficits associated with reduced functional performance [57]. Integrated with the present findings, increased mediolateral displacement may serve as a potential indicator for assessing re-injury risk.

Comprehensive analysis of center of pressure parameters reveals that different post-operative timepoint cohorts demonstrate differentiated center of pressure characteristics, reflecting complex neuromuscular system changes. The 1.5–6 months post-operative cohort is characterized primarily by compensatory adjustments, manifested as increased mediolateral displacement and limited operative-limb support function, with patients predominantly maintaining balance through enhanced visual and vestibular system mechanisms. The 3–9 months post-operative cohort demonstrates specific anteroposterior displacement increase, potentially reflecting sagittal plane postural control adjustments related to quadriceps neuromuscular activation patterns and motor cortex plasticity changes. The ≥ 9 months post-operative cohort demonstrates mediolateral displacement approaching normal control levels, suggesting progressive recovery of coronal plane lateral stability. However, operative-non-operative asymmetry persists across all post-operative timepoint cohorts, with gait line length and single-limb support line length consistently demonstrating shorter operative-limb measurements. The persistent functional asymmetry on the operative limb may be attributed to incomplete graft reinnervation [53, 58]. Previous research indicates that slow and incomplete graft reinnervation may result in long-term proprioceptive function limitations. Even at extended post-operative timepoints, proprioceptive parameters remain incomplete, suggesting that compensatory mechanisms cannot fully restore normal neuromuscular control patterns, with the operative limb tending to adopt conservative postural control strategies [8, 11, 53, 59, 60].

In conclusion, this study systematically evaluated center of pressure parameters during gait in ACLR patients across different post-operative timepoints, revealing persistent operative-non-operative asymmetry and differentiated changing patterns of anteroposterior and mediolateral displacements, reflecting complex changes in neuromuscular and proprioceptive function. Center of pressure trajectory analysis, as an objective quantitative assessment method, can reflect functional recovery status at different post-operative stages and provide evidence-based guidance for individualized rehabilitation planning and return-to-sport decision-making. Clinical rehabilitation professionals should employ targeted training strategies based on post-operative timepoint and functional characteristics; return-to-sport decisions should be based on multidimensional functional assessment rather than time-based criteria alone, to facilitate the transition toward individualized, evidence-based rehabilitation approaches.

Limitations

This study has several limitations that warrant discussion. First, the cross-sectional design reflects only between-group differences and cannot track individual functional evolution trajectories or establish causal relationships, thereby restricting dynamic understanding of the functional recovery process. Second, measurement methodological limitations exist; test-retest reliability and intra-rater reliability were not evaluated, and although walking speed demonstrated significant differences among post-operative timepoint groups, statistical control (i.e., analysis of covariance) was not applied, potentially affecting the accuracy of between-group comparisons. Third, the assessment task was overly simplistic, involving only gait without incorporation of more challenging functional tasks such as single-leg stance or jump-landing; additionally, kinematic parameters including knee joint angle and angular velocity were not measured, limiting in-depth analysis of postural control abnormality mechanisms. Fourth, the sample was recruited from a single medical center with a single surgical technique (single-bundle autogenous hamstring tendon reconstruction), potentially limiting the generalizability of findings to other populations and surgical approaches. Fifth, the representativeness of the non-operative limb as an internal control reference is questionable; both previous research and the present findings demonstrate that bilateral adaptive changes may occur post-operatively, with the non-operative limb potentially not representing true healthy status. Despite these limitations, this study employed a standardized research design and objective measurement methodology to provide valuable quantitative evidence for postural control assessment in ACLR patients across different post-operative stages.

Conclusion

This study demonstrated that ACLR patients exhibit marked operative-non-operative functional asymmetry during gait. Operative-limb gait line length and single-limb support line length were shorter than the non-operative limb, whereas anteroposterior and mediolateral displacements demonstrated abnormal fluctuations at different post-operative timepoints. All center of pressure parameters failed to recover to healthy control levels, reflecting long-term postural control alterations in the post-operative period. Center of pressure trajectory analysis can objectively reflect functional characteristics at different post-operative stages, providing quantitative evidence-based reference for clinical assessment and individualized rehabilitation decision-making.

Author contributions

Chao Liu, and BaoHua Liu were responsible for the experimental design. The experiments were conducted by Chao Liu, Fuyi Liu and Zihao Sun,. Chao Liu handled data processing and manuscript writing. All authors contributed to the manuscript’s draft and approved the final manuscript.

Funding

Not applicable.

Data availability

Any data will be available at the corresponding author upon request.

Declarations

Ethics approval and consent to participate

The studies involving human participants were reviewed and approved by the Ethics Committee of the Affiliated Hospital of Binzhou Medical University. The patients/participants provided their written informed consent to participate in this study.

Consent for publication

Participants give their consent for publication of their image if require.

Competing interests

The authors declare no competing interests.