Effects of crutch-assisted walking on pain, gait, and hip function in early and middle-stage osteonecrosis: a randomized trial

Abstract

Background

Early-to middle-stage osteonecrosis of the femoral head (ONFH) is commonly associated with pain, gait impairment, and functional limitation. Although protective weight-bearing is frequently recommended, high-quality evidence supporting standardized crutch use as a non-surgical intervention remains limited.

Objective

To evaluate the effects of daily crutch-assisted walking on pain, hip function, gait performance, health-related quality of life, and bone metabolic markers in patients with early-to middle-stage ONFH.

Methods

In this single-center randomized controlled trial, 56 patients (96 hips) with ARCO stage I–IIIA ONFH were randomly assigned to a crutch-assisted walking group (3 months of daily use) or a control group without walking aids. Both groups received standardized pharmacological treatment. Outcomes were assessed at baseline, 45 days, and 90 days. The primary outcome was pain intensity measured by the visual analog scale (VAS). Secondary outcomes included the Harris Hip Score (HHS), hip range of motion, spatiotemporal gait parameters, SF-36 scores, and serum biomarkers related to inflammation and bone metabolism. Linear mixed-effects models were applied to account for repeated measurements and bilateral hip clustering.

Results

At 90 days, the crutch-assisted walking group showed lower VAS scores compared with controls (adjusted mean difference −3.1, 95% CI −3.8 to −2.5; P < 0.001). Improvements were also observed in HHS, selected gait parameters, and several exploratory serum biomarkers, including osteocalcin, total procollagen type I amino-terminal propeptide, and interleukin-6. No significant between-group difference was observed in NSAID consumption. No serious adverse events were reported.

Conclusion

Standardised daily crutch-assisted walking may provide short-term symptomatic and functional benefits in patients with early-to middle-stage ONFH. However, given the modest sample size, relatively short follow-up duration, and exploratory nature of several secondary outcomes, these findings should be interpreted cautiously. Further studies are required to determine whether these short-term improvements translate into long-term structural or clinical benefits.

Trial registration

This study was registered at the Chinese Clinical Trial Registry on January 8, 2025, with registration number ChiCTR2500095502.

The translational potential of this article

The findings of this randomized controlled trial provide translational evidence supporting standardized crutch-assisted walking as a feasible and low-cost conservative intervention for patients with early-to middle-stage osteonecrosis of the femoral head. By demonstrating improvements in pain, gait performance, and hip function without activity restriction, this study informs everyday clinical decision-making and supports the integration of biomechanical unloading strategies into routine non-surgical management.

Graphical abstract

1. Introduction

Osteonecrosis of the femoral head (ONFH) is a progressive condition that commonly affects adults in early to middle age and is frequently associated with pain, impaired gait, and functional limitation [[1], [2], [3]]. In the early and intermediate stages of the disease, preservation of hip function and delay of functional deterioration are major clinical priorities, particularly for patients who are not immediate candidates for surgical intervention [4,5].

Mechanical loading plays a critical role in the pathophysiology of ONFH [6,7]. Reduced trabecular integrity and altered stress distribution within the necrotic femoral head increase susceptibility to microstructural strain during routine weight-bearing activities, contributing to pain and functional decline [8,9]. Accordingly, protective weight-bearing strategies are often recommended in clinical practice to reduce mechanical stress on the affected hip.

Mechanical unloading is commonly recommended for early ONFH, but the optimal degree, duration, and clinical value of protected weight-bearing remain uncertain [10]. Existing studies have reported heterogeneous findings, partly because conservative interventions are often combined with pharmacologic therapy, activity restriction, or postoperative rehabilitation protocols, making the specific contribution of crutch-assisted walking difficult to isolate [11,12]. In addition, many previous studies have focused on structural or surgical outcomes rather than on the integrated assessment of pain, gait performance, and functional recovery in patients managed conservatively [13,14].

Therefore, this randomized controlled trial was designed to evaluate a standardised crutch-assisted walking strategy in preoperative patients with early-to middle-stage ONFH. Compared with prior studies, the present trial sought to isolate a practical conservative unloading intervention and to assess its short-term effects on pain, hip function, gait performance, health-related quality of life, and exploratory serum biomarkers. By integrating these clinical and functional domains within a randomized design, this study aimed to provide more clinically relevant evidence regarding the role of biomechanical unloading in non-surgical ONFH management.

2. Methods

2.1. Study design and participants

This study was designed as a single-center, parallel-group randomized controlled trial with a 3-month follow-up period. Participants were consecutively recruited from the orthopedic outpatient clinic.

The study protocol was approved by the Medical Ethics Committee of the General Hospital of the People’s Liberation Army of China (approval number: 2024KY0147-KS001; approval date: January 6, 2025). The trial was registered with the Chinese Clinical Trial Registry on January 8, 2025 (ChiCTR2500095502). The first participant was randomized on January 10, 2025, and the last participant was randomized on May 6, 2025. Follow-up was completed on August 30, 2025. The study was conducted in accordance with the Declaration of Helsinki, and all participants provided written informed consent prior to participation.

Patients were eligible if they were 20–65 years of age, diagnosed with osteonecrosis of the femoral head (ONFH) based on radiography, computed tomography, and magnetic resonance imaging, and classified as ARCO stage I–IIIA [15]. For patients with bilateral ONFH, the inter-hip difference in ARCO stage was required to be ≤ 1 stage, and the difference in baseline pain intensity (VAS) ≤3 points, to ensure comparable disease severity within individuals. Additional inclusion criteria were a baseline VAS score between 1 and 7 and the ability to participate in follow-up assessments. Exclusion criteria included use of crutches within the previous 3 months, receipt of lower-limb physical therapy within 6 months, prior hip surgery, major medical comorbidity judged by the treating clinicians to preclude safe participation or follow-up, or psychiatric conditions judged to impair informed consent, adherence, or reliable outcome reporting.

2.2. Randomization and blinding

Eligible participants were randomly assigned in a 1:1 ratio to either the crutch-assisted walking group or the control group using a computer-generated randomization sequence prepared by an independent statistician. Allocation concealment was ensured using sequentially numbered, opaque, sealed envelopes. Because of the nature of the intervention, participants and treating therapists could not be blinded to group allocation. However, outcome assessors responsible for clinical evaluation, gait assessment, and data collection were not involved in treatment delivery and were kept unaware of group assignment throughout follow-up. Data analysts were also blinded to treatment allocation during the primary statistical analyses by using masked group coding until the main analyses had been completed.

2.3. Intervention

Participants in the intervention group received standardized instruction from a licensed physical therapist on proper crutch-assisted walking. Training included crutch height adjustment, arm positioning, and safe ambulation techniques to ensure effective biomechanical unloading of the affected hip during daily walking [16,17]. Participants in the intervention group were instructed to use bilateral axillary crutches during routine ambulation for 3 months, regardless of whether the disease was unilateral or bilateral, in order to standardise the unloading strategy across the intervention group.

Participants in the control group were instructed to maintain their usual daily activities without the use of any walking aids. Both groups received the same standardised background pharmacological treatment throughout the study period according to institutional clinical practice guidelines. The medication regimen included oral alendronate sodium (70 mg once weekly), calcium carbonate D3 tablets (600 mg calcium plus 400 IU vitamin D daily), and celecoxib capsules (200 mg once daily as needed for pain control). No group-specific pharmacological modifications were introduced during follow-up unless clinically required.

2.4. Adherence monitoring

Adherence to crutch use was monitored using daily self-reported logs documenting duration and context of use. Weekly telephone follow-up was conducted by the physical therapist to verify adherence and address practical difficulties related to crutch use. No additional therapeutic counselling or rehabilitation intervention was provided during these calls. Adequate adherence was defined as crutch use for at least 4 h per day on ≥80% of study days during the 90-day intervention period.

2.5. Outcome measures

Assessments were conducted at baseline, 45 days, and 90 days by trained assessors who were blinded to group allocation.

2.6. Primary outcome

The primary outcome was pain intensity, measured using the visual analog scale (VAS), ranging from 0 (no pain) to 10 (worst imaginable pain) [18].

2.7. Secondary outcomes

Secondary outcomes included hip function assessed by the Harris Hip Score (HHS) [19], hip range of motion measured with a standard goniometer [20], and health-related quality of life evaluated using the Short Form-36 (SF-36) questionnaire [21]. Gait performance was assessed using a validated wearable gait analysis system, which provided spatiotemporal parameters including gait speed, cadence, stride length, turning angle, and single-leg support stability [22,23]. Gait assessments were performed without crutches under assessor supervision in order to evaluate underlying walking function rather than device-assisted performance.

Serum biomarkers were assessed as exploratory outcomes. Blood samples were collected in the morning after an overnight fast at baseline, 45 days, and 90 days. Inflammatory status was evaluated by serum interleukin-6 (IL-6) [24,25]. Bone metabolism markers included osteocalcin, total procollagen type I amino-terminal propeptide (tP1NP), β-C-terminal telopeptide of type I collagen (β-CTX), intact parathyroid hormone, and 25-hydroxyvitamin D. Biomarker measurements were performed using standardized laboratory assays.

2.8. Sample size calculation

Sample size estimation was based on the primary outcome (VAS). Assuming a minimal clinically important difference of 2 points and a standard deviation of 2 points, with a two-sided significance level of 0.05 and 80% power, the required sample size was calculated. To account for repeated measurements and bilateral hip clustering, an intraclass correlation coefficient was incorporated. Allowing for potential attrition, a total of 96 hips from 56 patients were included in the final analysis. Secondary outcomes were considered supportive, and biomarker analyses were prespecified as exploratory; therefore, the trial was not specifically powered for all secondary and exploratory endpoints.

2.9. Statistical analysis

Randomization was performed at the patient level rather than the hip level. Patients with bilateral ONFH contributed both affected hips to hip-level analyses where applicable. Outcomes were analysed according to their measurement level. Patient-level outcomes, including SF-36 scores, serum biomarkers, NSAID use, and adverse events, were analysed using the patient as the unit of analysis and were not duplicated across hips in participants with bilateral disease. Hip-level outcomes, including VAS pain scores, HHS scores, hip range-of-motion measures, gait parameters, and other hip-specific clinical outcomes, were analysed at the hip level. For bilateral cases, both eligible hips were included in hip-level analyses, and within-patient correlation was accounted for using patient-specific random effects in linear mixed-effects models.

No formal imputation procedures were applied for missing data. Primary longitudinal analyses were conducted using linear mixed-effects models, which included all available observations under the missing-at-random (MAR) assumption. Accordingly, the analyses did not represent a strict intention-to-treat framework.

Descriptive statistics are presented as mean (standard deviation) or frequency (percentage), as appropriate. Between-group comparisons at baseline were performed using independent-sample t tests or Mann–Whitney U tests for continuous variables and chi-square or Fisher's exact tests for categorical variables, as appropriate. Longitudinal outcomes were analysed using linear mixed-effects models including fixed effects for group, time, and group-by-time interaction, with baseline values entered as covariates where appropriate. Adjusted mean differences at follow-up were derived from model-estimated marginal means. Multiple comparisons for prespecified secondary outcomes were adjusted using the Holm–Bonferroni method. Secondary and exploratory outcomes were interpreted cautiously and considered supportive or hypothesis-generating rather than definitive confirmatory evidence. Sensitivity analyses using complete-case and per-protocol datasets were additionally performed for the primary outcome to assess the robustness of the findings. All statistical tests were two-sided, and a P value < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS (version 26.0).

3. Results

Participant enrollment, randomization, and follow-up are summarized in Fig. 1. Randomization was conducted at the patient level.

Fig. 1.

Participant flow through the study. A total of 88 patients were assessed for eligibility, of whom 56 patients (96 hips) were randomized to the control group or the crutch-assisted walking group. Outcome assessments, including pain, hip function, gait performance, range of motion, quality-of-life measures, and serum biomarkers, were conducted at baseline (T0), 45 days (T45), and 90 days (T90). Losses to follow-up and withdrawals at each time point are shown. Most incomplete follow-up assessments were related to logistical factors, including work-related scheduling conflicts and participants residing outside the local study area, rather than clinical deterioration. The numbers of patients and hips included in the analyses at each follow-up time point are indicated.

3.1. Baseline characteristics

Baseline demographic and clinical characteristics of the study participants are summarized in Table 1. A total of 56 patients (96 hips) were included, with 28 patients assigned to each group. No statistically significant differences were observed between the control and intervention groups with respect to age, sex distribution, body mass index, or disease duration.

Table 1.

Baseline demographic characteristics of patients.

Variable	Control (n = 28)	Intervention (n = 28)	P value
Age (years)	36.3 (9.7)	40.4 (11.2)	0.142
Male/Female	20/8	23/5	0.401
BMI (kg/m2)	24.3 (3.4)	24.9 (3.8)	0.525
Disease duration (month)	9.6 (13.4)	11.5 (16.1)	0.668
Hip involved			0.400
Unilateral	6 (21.4%)	10 (35.7%)
Bilateral	22 (78.6%)	18 (64.3%)
Total number	50	46
Etiology			0.440
Idiopathic	12 (42.9%)	14 (50.0%)
Alcohol	8 (28.6%)	8 (28.6%)
Steroid	8 (28.6%)	6 (21.4%)
ARCO Stage			0.299
Stage I	3 hips (6.0%)	9 hips (19.6%)
Stage Ⅱ	14 hips (28.0%)	13 hips (28.3%)
Stage ⅢA	33 hips (66.0%)	24 hips (52.2%)
JIC type			0.668
A	2 hips (4.0%)	3 hips (6.5%)
B	5 hips (10.0%)	4 hips (8.7%)
C1	15 hips (30.0%)	16 hips (34.8%)
C2	28 hips (56.0%)	23 hips (50.0%)
25(OH)D (ng/mL)	20.3 (4.5)	19.8 (4.2)	0.681
iPTH (pg/mL)	63.5 (12.3)	65.2 (11.8)	0.552

Abbreviations: BMI = body mass index; ARCO = Association Research Circulation Osseous; JIC = Japanese Investigation Committee classification.

a Baseline characteristics are reported at the patient level unless otherwise specified; ARCO stage and JIC classification are reported at the hip level.

b Values are presented as mean (SD) or number (%). Between-group comparisons were performed using independent t-tests, Mann–Whitney U tests, χ² tests, or Fisher's exact tests, as appropriate.

The distribution of unilateral and bilateral hip involvement was comparable between groups, resulting in 50 hips in the control group and 46 hips in the intervention group. Etiological factors, including idiopathic, alcohol-related, and steroid-related osteonecrosis, were similarly distributed.

At the hip level, no significant between-group differences were observed in ARCO stage or Japanese Investigation Committee (JIC) classification. Baseline serum levels of 25-hydroxyvitamin D and intact parathyroid hormone were also comparable between groups.

Overall, baseline characteristics were well balanced between the two groups, indicating adequate comparability prior to intervention.

3.2. Primary outcome

Pain intensity, assessed using the visual analog scale (VAS), was comparable between the two groups at baseline (Table 2). Over the 3-month follow-up period, pain progressively worsened in the control group, whereas patients in the crutch-assisted walking group experienced a marked reduction in pain. At 45 days, the intervention group demonstrated significantly lower VAS scores compared with controls (adjusted mean difference −2.4, 95% CI −2.9 to −2.0; P < 0.001). This between-group difference further increased at 90 days (adjusted mean difference −3.1, 95% CI −3.8 to −2.5; P < 0.001), exceeding the minimal clinically important difference for pain.

Table 2.

Primary and key secondary outcomes.

Outcome	Time point	Control (Mean ± SD)	Intervention (Mean ± SD)	Treatment Effect (95% CI)	P value
VAS	Baseline	3.7 ± 1.1	3.3 ± 1.4	−0.4 (−0.9 to 0.1)	0.218
	45 days	4.7 ± 1.1	2.3 ± 1.1	−2.4 (−2.9 to −2.0)	<0.001∗
	90 days	5.7 ± 1.6	2.6 ± 1.6	−3.1 (−3.8 to −2.5)	<0.001∗
HHS	Baseline	77.4 ± 15.2	79.3 ± 11.6	1.8 (−3.6 to 7.4)	0.505
	45 days	68.1 ± 13.1	73.8 ± 10.1	5.7 (1.0 to 10.8)	0.027∗
	90 days	62.5 ± 13.2	72.7 ± 10.0	10.2 (5.0 to 15.5)	0.003∗
Gait speed (m/s)	Baseline	1.1 ± 0.2	1.0 ± 0.3	0.1 (−0.1 to 0.2)	0.258
	45 days	0.9 ± 0.3	1.1 ± 0.3	0.2 (0.1 to 0.3)	0.013∗
	90 days	0.8 ± 0.2	1.0 ± 0.3	0.2 (0.1 to 0.3)	<0.001∗
Cadence (steps/min)	Baseline	107.7 ± 11.9	103.2 ± 11.2	−4.4 (−9.1 to 0.3)	0.065
	45 days	98.0 ± 11.8	98.4 ± 13.6	0.4 (−5.7 to 5.2)	0.905
	90 days	83.0 ± 17.9	96.2 ± 20.1	13.3 (8.6 to 17.9)	0.004∗
Turning angle (°)	Baseline	13.1 ± 9.9	14.2 ± 11.2	1.1 (−3.1 to 5.4)	0.614
	45 days	14.4 ± 10.1	9.1 ± 8.9	−5.3 (−9.4 to −1.2)	0.012∗
	90 days	14.3 ± 10.9	9.0 ± 7.1	−5.3 (−9.3 to −1.3)	0.008∗
Single-leg support stability	Baseline	0.8 ± 0.3	0.9 ± 0.3	0.1 (−0.1 to 0.2)	0.132
	45 days	0.7 ± 0.3	0.8 ± 0.3	0.1 (−0.1 to 0.2)	0.216
	90 days	0.6 ± 0.3	0.9 ± 0.4	0.3 (0.2 to 0.4)	<0.001∗

∗P < 0.05 indicates statistical significance after correction.

Abbreviations: VAS = visual analog scale; HHS = Harris hip score.

Additional secondary and exploratory outcomes are presented in the Supplementary Material.

^a Values are presented as mean ± standard deviation.

^b Treatment effects were estimated using linear mixed-effects models adjusting for baseline values and bilateral-hip clustering.

Sensitivity analyses for the primary outcome were consistent with the primary mixed-effects model. In the complete-case analysis, the group-by-time interaction for VAS remained statistically significant (F = 10.776, P < 0.001). Similarly, in the per-protocol analysis, the group-by-time interaction also remained significant (F = 12.908, P < 0.001). In both analyses, the estimated pain trajectories showed the same pattern as in the primary analysis, with worsening pain over time in the control group and lower follow-up VAS scores in the intervention group (Supplementary Table S5).

3.3. Key secondary outcomes

Hip function, measured by the Harris Hip Score (HHS), did not differ significantly between groups at baseline. At 45 days, patients in the crutch-assisted walking group showed significantly higher HHS scores than controls (adjusted mean difference 5.7, 95% CI 1.0 to 10.8; P = 0.027). This improvement was more pronounced at 90 days, with a between-group difference of 10.2 points (95% CI 5.0 to 15.5; P = 0.003), indicating clinically meaningful functional benefit. Temporal trends in VAS and HHS are illustrated in Fig. 2.

Fig. 2.

Mean VAS (A) and HHS (B) scores over time in the control and intervention groups.

Gait performance outcomes demonstrated consistent improvements associated with crutch-assisted walking. Gait speed was similar between groups at baseline but was significantly higher in the intervention group at both 45 days (adjusted mean difference 0.2 m/s, 95% CI 0.1 to 0.3; P = 0.013) and 90 days (adjusted mean difference 0.2 m/s, 95% CI 0.1 to 0.3; P < 0.001). In addition, turning angle during walking was significantly reduced in the intervention group at 45 days and 90 days, reflecting improved directional control. Single-leg support stability also improved over time in the crutch-assisted walking group and was significantly higher than in controls at 90 days (adjusted mean difference 0.3, 95% CI 0.2 to 0.4; P < 0.001).

All key secondary outcomes remained statistically significant after Holm–Bonferroni correction for multiple comparisons. Treatment effects were estimated using linear mixed-effects models adjusting for baseline values and bilateral-hip clustering. These secondary findings should be interpreted as supportive rather than definitive, given that the trial was powered for the primary outcome only (Table 2).

3.4. Hip range of motion

Changes in hip range of motion are summarized in Supplementary Table S1. Baseline values for hip flexion, extension, adduction, and abduction did not differ significantly between the two groups.

At 45 days, patients in the crutch-assisted walking group demonstrated significantly greater hip flexion and abduction compared with the control group. No significant between-group differences were observed for hip extension or adduction at this time point.

At 90 days, significantly greater hip flexion and abduction were maintained in the intervention group. In addition, hip adduction was significantly higher in the intervention group at 90 days. No significant between-group differences were observed for hip extension at any time point.

Overall, improvements in hip range of motion associated with crutch-assisted walking were primarily observed in flexion and abduction, whereas changes in extension were not significant.

3.5. Gait outcomes

Gait performance outcomes are summarized in Supplementary Table S2. No significant between-group differences were observed in gait parameters at baseline.

At 45 days, patients in the crutch-assisted walking group demonstrated significantly greater gait speed and a significantly smaller turning angle compared with the control group. No significant between-group differences were observed in stride length, cadence, or single-leg support stability at this time point.

At 90 days, gait speed remained significantly higher in the intervention group. In addition, cadence and single-leg support stability were significantly greater in the crutch-assisted walking group, while turning angle remained significantly reduced. No significant between-group difference was observed for stride length at any time point.

Overall, crutch-assisted walking was associated with improvements in gait speed, directional control, and single-leg support stability, whereas stride length did not differ significantly between groups.

3.6. Exploratory biomarker outcomes

Changes in serum biomarkers were evaluated as exploratory outcomes and are summarized in Supplementary Table S3. Baseline levels of inflammatory and bone metabolism markers did not differ significantly between the two groups.

At 45 days, no statistically significant between-group differences were observed for any of the assessed biomarkers. Serum interleukin-6 (IL-6) levels showed a trend toward reduction in the crutch-assisted walking group compared with controls; however, this difference did not reach statistical significance.

At 90 days, patients in the crutch-assisted walking group demonstrated significantly lower IL-6 levels than those in the control group. In addition, markers associated with bone formation, including osteocalcin and total procollagen type I amino-terminal propeptide (tP1NP), were significantly higher in the intervention group at 90 days. In contrast, no significant between-group differences were observed for β-C-terminal telopeptide of type I collagen (β-CTX) at any time point.These biomarker findings were exploratory and were not intended to infer structural repair.

Use of nonsteroidal anti-inflammatory drugs over the 90-day period did not differ significantly between groups.

3.7. Secondary health-related quality of life outcomes (SF-36)

Health-related quality of life was assessed using the SF-36 questionnaire as a secondary outcome (Supplementary Table S4). Baseline SF-36 domain scores were comparable between the two groups.

At 45 days, patients in the crutch-assisted walking group demonstrated significantly higher scores in the bodily pain, general health, and social functioning domains compared with the control group. Improvements in the other SF-36 domains were not statistically significant.

At 90 days, the intervention group continued to show significantly higher bodily pain and general health scores. In addition, social functioning scores were significantly higher in the crutch-assisted walking group at this time point. No significant between-group differences were observed in the remaining SF-36 domains.

3.8. Intervention compliance and safety

Intervention compliance and safety outcomes are summarized in Table 3. Among patients assigned to the crutch-assisted walking group, 24 of 28 participants (85.7%) achieved the predefined compliance criteria. The mean daily duration of crutch use among compliant participants was 312.0 ± 78.0 min.

Table 3.

Intervention compliance and crutch-related adverse events.

Index/Event	Intervention (n = 28)	Control (n = 28)	Notes (Definition/Treatment/Outcome)
Crutch use compliance
Compliance achieveda	24 (85.7%)	−	Compliance defined as crutch use ≥4 h/day on ≥80% of intervention days (≥72 of 90 days)
Non-compliance	4 (14.3%)	−	Reasons: axillary discomfort (n = 2), upper limb muscle soreness (n = 1), minor fall (n = 1)
Mean daily crutch use (min/d, ±SD)	312.0 ± 78.0	−	Calculated among compliant participants only (n = 24)
Adverse Events (AEs)
Axillary skin irritation	2 (7.1%)	0 (0.0%)	Mild; resolved within 1 week after crutch height adjustment
Upper limb muscle soreness	1 (3.6%)	0 (0.0%)	Mild; resolved within 3 days after stretching guidance
Minor fall	1 (3.6%)	0 (0.0%)	No fracture or joint injury; no recurrence after safety instruction
Total AEs (CTCAE Grade 1)	4 (14.3%)	0 (0.0%)	All events were self-limited and did not require discontinuation

^b Compliance was defined as crutch use for ≥4 h per day on ≥80% of intervention days.

^c Adverse events were descriptively summarized and graded according to the Common Terminology Criteria for Adverse Events (CTCAE).

^aValues are presented as number (%) unless otherwise specified.

Four participants (14.3%) in the intervention group did not meet the compliance criteria. Reasons for non-compliance included axillary discomfort (n = 2), upper limb muscle soreness (n = 1), and a minor fall (n = 1).

No serious adverse events occurred in either group during follow-up. The mild adverse events summarized in Table 3 were crutch-related events observed in the intervention group and included axillary skin irritation (n = 2), upper-limb muscle soreness (n = 1), and one minor fall without fracture or joint injury. All of these events were self-limited and resolved with conservative management. No participant required discontinuation of the intervention because of an adverse event.

4. Discussion

This randomized controlled trial suggests that standardised daily crutch-assisted walking was associated with reduced pain and improved hip function and gait performance over 90 days in patients with early-to middle-stage osteonecrosis of the femoral head. Compared with usual care alone, patients who used crutches for 3 months showed clinically meaningful reductions in pain intensity, accompanied by improvements in spatiotemporal gait parameters and functional hip scores. These findings suggest that crutch-assisted walking may function as a practical biomechanical unloading strategy that could help support functional mobility without substantial restriction of daily activity. However, given the modest sample size, relatively short follow-up duration, and exploratory nature of several secondary outcomes, these findings should be interpreted cautiously.

The observed improvements in pain and gait performance may plausibly be related to biomechanical unloading of the affected hip during daily ambulation [26,27]. In early-to middle-stage ONFH, necrotic bone exhibits reduced trabecular strength and altered stress distribution, making routine weight-bearing activities prone to inducing microstructural strain and pain [28,29]. Crutch-assisted walking redistributes axial load away from the affected femoral head and enhances external stability during stance and turning, which may reduce peak mechanical stress on vulnerable regions [8,30,31]. This unloading effect likely attenuates nociceptive input during gait, enabling patients to walk with greater confidence and efficiency rather than restricting movement.

Importantly, the concurrent improvements in gait speed, turning angle, and single-leg support stability suggest that crutch use does not merely reduce pain through activity avoidance, but instead facilitates safer and more symmetrical locomotion [32,33]. This distinction is clinically relevant, as excessive activity restriction may exacerbate muscle deconditioning and functional decline [34]. By contrast, crutch-assisted walking appears to provide a protective biomechanical environment that supports continued mobility, potentially interrupting the cycle of pain, altered gait, and functional deterioration commonly observed in ONFH [35,36].

In addition to functional improvements, changes in selected serum biomarkers were observed following crutch-assisted walking. Specifically, increases in osteocalcin and total procollagen type I amino-terminal propeptide, along with a reduction in interleukin-6 levels, were detected in the intervention group at 90 days. However, these findings should be interpreted cautiously. Circulating biomarkers of bone turnover and inflammation may reflect short-term systemic responses associated with changes in pain, mobility, inflammatory status, or load redistribution, but they do not constitute direct evidence of local structural repair within the femoral head [37].

It is possible that crutch-assisted walking altered the balance between pain, ambulation, inflammation, and systemic bone turnover during follow-up. Reduced pain and improved gait performance may have enabled patients to maintain safer levels of physical activity, which could in turn influence systemic biomarker profiles [38,39]. However, the relationship between mechanical unloading and bone metabolism in ONFH is biologically complex, and the present study was not designed to establish a specific mechanistic pathway [40,41]. The absence of a significant between-group difference in β-CTX does not suggest increased bone resorption in the intervention group, although this finding should also be interpreted cautiously [42]. Overall, the observed changes in P1NP, osteocalcin, and IL-6 should be regarded as exploratory and hypothesis-generating rather than as evidence that crutch-assisted walking directly promotes bone formation or modifies disease progression. This interpretation is further limited by the absence of imaging-based structural endpoints.

From a translational perspective, the present findings should be interpreted in the context of other non-surgical treatment strategies for ONFH, including extracorporeal shock wave therapy, bisphosphonate-based treatment, and broader multimodal conservative management [[43], [44], [45]]. Compared with these approaches, standardised crutch-assisted walking may represent a simpler biomechanical unloading strategy that is low-cost, practical, and readily implementable in routine care. By reducing mechanical load while preserving daily mobility, crutch use may help alleviate pain and maintain functional gait without imposing excessive activity restriction [8]. This approach may be particularly relevant for patients who are not immediate candidates for surgery or who prefer conservative management options.

Several limitations should be acknowledged. First, this was a single-center study with a relatively short follow-up period, and imaging-based outcomes, such as femoral head collapse or lesion progression, were not assessed. Second, bone metabolic markers were evaluated as exploratory outcomes and therefore cannot be interpreted as evidence of structural repair. Despite adjustment for multiple comparisons in prespecified secondary outcomes, the possibility of type I error for exploratory analyses cannot be excluded. Third, the study was powered for the primary pain outcome rather than for the full set of secondary and exploratory endpoints.

An additional limitation relates to discrepancies between the original trial registration entry and the final implemented study protocol. The trial registration was completed during the early planning phase of the study, and several protocol elements were subsequently refined prior to completion of participant recruitment, including aspects of the study population, intervention implementation, outcome assessment framework, and analytical approach. However, these modifications were not fully updated in the public trial registration record, which reduced protocol transparency and may increase the risk of perceived reporting bias. Although the primary study objective and overall intervention strategy remained unchanged throughout the study period, this inconsistency should be considered when interpreting the findings.

Weekly telephone contact was used to monitor adherence in the intervention group, whereas the control group did not receive an equivalent intensity of follow-up contact. This imbalance may have introduced attention-related or expectancy effects, particularly for patient-reported outcomes such as VAS and the subjective components of the HHS. In addition, adherence to crutch-assisted walking was primarily monitored through self-reported daily logs and scheduled telephone follow-up rather than objective monitoring methods. Future studies should incorporate objective monitoring approaches to improve adherence verification and reduce potential reporting bias.

Most missing follow-up data were related to incomplete gait assessments due to logistical factors, including participants residing in other cities and work-related scheduling conflicts, rather than clinical deterioration or worsening pain symptoms. Nevertheless, because missingness cannot be assumed to be completely random, the possibility of attrition-related bias cannot be entirely excluded.

Future studies should include multicentre designs, longer follow-up, and imaging-based structural endpoints to determine whether the short-term symptomatic benefits observed here translate into long-term preservation of the femoral head. Additional research is also needed to evaluate the role of crutch-assisted walking in broader patient populations, including older adults, patients with obesity, and those receiving combined conservative therapies, and to clarify its place among other non-surgical treatments for ONFH.

5. Conclusion

Standardised daily crutch-assisted walking may provide short-term symptomatic and functional benefits in patients with early-to middle-stage osteonecrosis of the femoral head. As a simple and low-cost biomechanical unloading strategy, crutch-assisted walking may help reduce mechanical loading while preserving daily mobility. However, given the modest sample size, relatively short follow-up duration, and exploratory nature of several outcomes, these findings should be interpreted cautiously. Further studies with longer follow-up, larger sample sizes, and imaging-based structural outcomes are required to determine whether these preliminary functional improvements translate into long-term clinical or structural benefits.

Trial registration

This study was registered at the Chinese Clinical Trial Registry on January 8, 2025, with registration number ChiCTR2500095502.

Author contributions

Junming Zhang and Yanbin Wu contributed equally to this work. Junming Zhang, Yanbin Wu, and Yazhou Li were involved in study conception and design. Tingjie Zhao, Jiazhou Wu, and Zexian Liu contributed to patient recruitment and data collection. Tao Qian, Biao Ma, Yun Bai, and Jialiang You performed clinical assessments and follow-up evaluations. Endong Luo, Xiaohan Sun, and Aiyuan Wang conducted data analysis and interpretation. Dawei Zhang and Jiang Peng provided methodological guidance and critical revisions of the manuscript. Junming Zhang drafted the manuscript. All authors reviewed and approved the final manuscript.

Ethics approval and consent to participate

This study was approved by the Medical Ethics Committee of the General Hospital of the People's Liberation Army of China.e (Approval No: 2024KY0147-KS001,approval date: 6 January 2025). The experiment adheres to the principles set forth in the Declaration of Helsinki, and all participants were formally informed of the purpose and content of the research and signed an informed consent form.

This study was registered at the Chinese Clinical Trial Registry on January 8, 2025, with registration number ChiCTR2500095502.

Data and materials availability

The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.

Declaration of artificial intelligence (AI) and AI-assisted technologies in the writing process

We confirm that AI tools were used appropriately during manuscript preparation to assist with grammar correction and language polishing. These tools helped improve the text's clarity, accuracy, and fluency. The authors have nevertheless carefully reviewed and verified all AI-generated modifications to ensure scientific accuracy and consistency with the intended meaning.

Funding

This work was funded by National Key Research and Development Program (2024YFA1108600).

additional Information

ARCO

Association Research Circulation Osseous

CI

Confidence Interval

CTCAE

Common Terminology Criteria for Adverse Events

HHS

Harris Hip Score

IL-6

Interleukin-6

JIC

Japanese Investigation Committee

NSAID

Nonsteroidal Anti-inflammatory Drug

ONFH

Osteonecrosis of the Femoral Head

ROM

Range of Motion

SF-36

36-Item Short Form Health Survey

tP1NP

Total Procollagen Type I Amino-terminal Propeptide

VAS

Visual Analog Scale

β-CTX

β-Isomerized C-terminal Telopeptide of Type I Collagen

Appendix A. Supplementary data

The following are the Supplementary data to this article.