Abstract
Background:
The effect of hip arthroscopy on the natural progression of osteoarthritis is still unclear.
Purpose:
To clarify the relationship between unilateral hip arthroscopy, the size of the cam lesion, and its excision on the progression of hip osteoarthritis.
Study Design:
Cohort study; Level of evidence, 3.
Methods:
A total of 105 patients who underwent unilateral hip arthroscopy due to femoroacetabular impingement and had a minimum of 2 years of follow-up were included. Joint space width (JSW) measurements were made at the medial (MJW), lateral (LJW), and central (CJW) aspects of the sourcil line on anteroposterior supine pelvis radiographs. The alpha angle (α) was measured using preoperative and early postoperative Dunn 45° radiographs.
Results:
LJW decreased and was more obvious in men than in women in the operated hips. All LJWs, MJWs, and CJWs decreased on the contralateral nonoperated hips (P < .05). The operated sides had higher MJW compared with the contralateral nonoperated sides (3.58 ± 0.75 mm vs 3.34 ± 0.6 mm; P = .02) and a lower decrease of medial joint space compared with contralateral sides (Δ1, −0.05 mm; Δ3, −0.3 mm; P = .00). At the final follow-up, central joint space was greater in the operated hips compared with the contralateral hips in patients with higher preoperative alpha angle (α0) (≥0.1 mm vs <0.1 mm; preoperative α, 81 vs 79; P = .03).
Conclusion:
Patients treated with unilateral hip arthroscopy demonstrated a decrease in the LJW, while nonoperated sides had a decrease in all 3 widths, leading to a greater joint space in the operated hip, especially at the MJW. Furthermore, patients having more severe cam deformity preoperatively had better preservation of CJW.
More than 20 years ago, Ganz et al 7 proposed femoroacetabular impingement (FAI) as a leading cause of hip osteoarthritis and the silent killer of the hip, secondary to chondrolabral tissue injury via acetabulum-based pincer or femoral-based cam or mixed pathologies. Hip arthroscopy has become more popular in recent decades, as it has been proven to be an effective procedure to preserve the joint and alleviate symptoms. The main components of the procedure are labral repair and cam lesion excision.12,23,29
However, the effect of hip arthroscopy on the natural progression of osteoarthritis is still unclear. Ramkumar et al 29 reviewed unilateral hip arthroscopy patients and reported advancement of Tönnis grade in 28% of the operated hips compared with 48% in the nonoperated contralateral hips. Similar to this study, Husen et al 12 reported a 26.5% rate of progression of osteoarthritis in the operatively treated cohort compared with a 35.2% rate in the control nonoperatively treated cohort. Both of these studies used the Tönnis grading system to assess osteoarthritis, which has been reported to be subjective with questionable interobserver and intraobserver reliability.4,33,37
Another controversial issue is whether the correction of the cam lesion prevents the development of hip osteoarthritis.6,9,16 A number of studies have reported no benefit to the progression of hip osteoarthritis with the correction of the alpha angle or cam deformity.1,6,9 These conflicting findings may be partly explained by limitations of categorical grading systems. Classifications such as the Tönnis grade may lack sensitivity for detecting early degenerative changes, highlighting the need for quantitative radiographic measures. 12 In this regard, joint space width (JSW) has been shown to be a sensitive and prognostically relevant parameter for assessing osteoarthritis progression. Reduced JSW predicts conversion to total hip arthroplasty after hip arthroscopy, and side-to-side JSW differences between the operative and contralateral asymptomatic hip further refine this risk assessment.20,26 However, several previous studies assessed JSW only on preoperative radiographs, without evaluating longitudinal changes over time. 20 In addition, preoperative CJW measurements have demonstrated prognostic value for postoperative outcomes, particularly in patients aged ≥35 years. 8 Therefore, assessing JSW changes both longitudinally and relative to the contralateral hip may provide a more robust framework for evaluating osteoarthritis progression and the potential disease-modifying effect of cam lesion excision.
Accordingly, in the present study, the contralateral asymptomatic hip was used as an internal control to minimize patient-related confounding factors and allow a within-patient assessment of JSW changes after unilateral hip arthroscopy.
The purpose of the present study was to clarify the effect of cam lesion size and its excision on the progression of hip osteoarthritis after unilateral hip arthroscopy, using quantitative changes in JSW as a radiographic marker of disease progression.
We hypothesized that the decrease in JSW at 3 points of the sourcil would be smaller in the operated hip compared with the contralateral nonoperated hip, reflecting a relative slowing of osteoarthritis progression, and that this effect would be dependent on both preoperative and postoperative alpha angles, with a more pronounced effect in larger cam lesions.
Methods
A total of 105 patients who underwent unilateral hip arthroscopy due to femoroacetabular impingement and who had a minimum of 2 years of follow-up were included in this retrospective study.
Ethics approval was obtained from the local ethics board. The research was conducted in accordance with the Declaration of Helsinki. Consent to participate was obtained from all participants. FAI was diagnosed based on clinical symptoms and radiographic findings. Alpha angles >55° using the 45° Dunn view were considered cam deformity,21,27 and LCEA (lateral center-edge angle) >40° was considered pincer deformity.17,41
Surgery was performed on patients with FAI who had persistent hip pain refractory to nonoperative treatment for ≥3 months. Patients who had bilateral symptoms, avascular necrosis, advanced-level hip osteoarthritis (Tönnis grade >1), 36 any previous ipsilateral hip surgery, revision hip arthroscopy, incomplete radiographs, hip dysplasia (LCEA <250°) 8 or could not be reached (n = 13) were excluded from this study.
Data on continuous and categorical demographic and clinical variables, including age, sex, body mass index (BMI), and duration of symptoms before surgery, were collected.
Anteroposterior (AP) supine pelvic radiographs and 45° Dunn radiographs were obtained and evaluated in all patients. LCEA was measured using the method described by Wiberg 41 on the AP pelvic radiograph. The alpha angle (α) was measured using 45° Dunn radiographs.21,27
Joint space measurements were performed on AP supine pelvic radiographs. The medial and lateral measurements were performed with regard to the most medial and lateral aspects of the sourcil (medial joint width [MJW]; lateral joint width [LJW]). For the central joint space measurement, a perpendicular line was drawn from the centers of the femoral heads, and the measurement was made orthogonal to this horizontal line (central joint width [CJW]) 30 (Figure 1).
Joint width measurements on an anteroposterior supine radiograph: (A) lateral joint width; (B) central joint width; (C) medial joint width.
Four different calculations were made using LJW, CJW, and MJW.
Δ1: Joint width at the final follow-up minus the preoperative joint width of the operated hips.
Δ2: Joint width difference at the final follow-up between operated and nonoperated hips.
Δ3: Joint width difference between the first and last radiograph of the nonoperated hips.
Δ4: The absolute difference between Δ1 and Δ3.
Patient-reported outcomes, which included the Hip Outcome Score Activities of Daily Living Scale (HOS ADL), 22 modified Harris Hip Score (mHHS), 35 and visual analog scale for pain, were collected via direct contact with the patients.
Reported Patient Acceptable Symptom State (PASS) values ≤5 years after hip arthroscopy for FAI were accepted as 83.6 for mHHS and 86 for HOS ADL.25,38 The minimal clinically important difference was not used because patients had been assessed at different time intervals.
Radiographic assessments were performed by 2 orthopaedic consultants. One month after the initial radiographic assessments, the measurements were repeated to assess intrarater reliability. Intraclass correlation coefficient (ICC) values were calculated for interrater and intrarater reliability. ICC values >0.80 indicated excellent reliability, 0.61 to 0.80 substantial reliability, 0.41 to 0.60 moderate reliability, 0.21 to 0.40 fair reliability, and ≤0.20 poor reliability. 19
Surgical Technique
The patient was placed supine on a hip arthroscopy–specific traction table to obtain appropriate hip distraction. A horizontal interportal capsulotomy was used to improve the visualization and access to the central compartment. A 4.5-mm arthroscopic bur was used to perform acetabuloplasty. Degenerative labral tears or those with multiple cleavage planes were considered irreparable, and unstable flaps were selectively debrided. Tears that involved the base of the labrum with chondrolabral disruption were repaired using 1 to 3 suture anchors. Traction was then released, the peripheral compartment was entered, and decompression of the cam deformity was performed and confirmed by intraoperative fluoroscopy and arthroscopic dynamic examination. The capsule was routinely left open at the end of the procedure.
Rehabilitation
All patients were instructed to use crutches to limit weightbearing for 2 weeks. Daily passive range of motion exercises were begun on the first postoperative day. At 3 weeks, active range of motion and full weightbearing were commenced. After 6 weeks, strengthening exercises and light treadmill walking were begun. Daily oral anti-inflammatory medication was prescribed for the first 4 weeks.
Statistical Analysis
Paired t test was used to assess differences between preoperative and postoperative data. Independent sample t test was used to make comparisons between operated and nonoperated contralateral sides or between changes in joint space (Δ) and frequency of achieving PASS values for functional scores or categorical variables. The Pearson correlation analysis was then used to analyze the correlation between measured variables and Δ values.
Results
Of the 105 patients, 63 were male, and 42 were female. They had a mean age of 35 ± 11 years and a mean follow-up time of 5 years. Differences in the JSWs (Δ1, Δ2, Δ4) were not correlated with age, follow-up time, symptom duration, BMI, status of active smoking, type of labral treatment (repair or debridement), LCEA, or preoperative and postoperative α angles (Table 1).
Demographics, Patient-Reported Outcomes and Radiological Measurements of the Patients (N = 105) a
Data are presented as mean ± SD or n. HOS ADL, Hip Outcome Score Activities of Daily Living; LCEA, lateral center-edge angle; mHHS, modified Harris Hip Score; pVAS, visual analog scale for pain.
Patients who reached the PASS score for mHHS had a smaller decrease only in their CJW that approached statistical significance (PASS [+]: Δ1: 0.04 mm vs PASS(–): Δ1: –0.2 mm; P = .06).
Men had a greater decrease in LJW in their operated hips compared with women (Δ1, –0.5; Δ1, –0.1; P = .01).
The operated hips only had a statistically significant decrease in LJW (Δ1 LJW: –0.4± 0.7; P = .00), while the nonoperated contralateral hips had a decrease in LJW, CJW, and MJW (Δ3 LJW: –0.3 ± 0.4, P = .00; Δ3 CJW: –0.1 ± 0.3, P = .00; Δ3 MJW: –0.3 ± 0.4, P = .00) (Table 2).
Joint Width Measurements a
CJW, central joint width; LJW, lateral joint width; MJW, medial joint width.
The operated sides had higher MJW compared with the contralateral nonoperated sides (3.58 ± 0.75 mm vs 3.34 ± 0.6 mm; P = .02) and a smaller decrease in the medial joint space (Δ1, −0.05 mm; Δ3, −0.3 mm; P = .00) at the final follow-up.
The difference in the magnitude of the decrease of the CJWs between operated and nonoperated hips approached statistical significance (CJW: Δ1, –0.001 mm; Δ3, –0.1 mm; P = .05).
When Δ values were grouped into 2 according to their median values and further analyzed for possible α0 differences, only CJW was found to be greater in the operated hips compared with the contralateral sides in patients with higher preoperative α0 (Δ2: ≥0.1 mm vs <0.1 mm; preoperative α0 = 81 vs 79; P = .03). There was no difference for postoperative α0 (460 vs 470) or the change of α0 (preoperative–postoperative) (350 vs 320) (P > .05).
All variables had excellent ICCs (>0.8) except preoperative α0 (intrarater, 0.74; interrater, 0.61), postoperative α0 (interrater, 0.74), postoperative LJW (interrater, 0.75), and MJW (interrater, 0.72) having substantial reliabilities.
Discussion
The main findings of the present study were that in patients treated with unilateral hip arthroscopy, only lateral joint space was decreased, which was more apparent in men than in women compared with contralateral nonoperated hips, while in the nonoperated hips, all 3 joint widths demonstrated a decrease (lateral, central, and medial). Medial joint space measurements were higher, and the decrease in this space was lower in the operated hips. Central joint space was greater in patients with a more severe cam deformity (higher preoperative α0) at the time of the final follow-up.
There are controversial reports about the effect of hip arthroscopy on the progression of hip osteoarthritis. However, evidence does exist showing the preventive effect of hip arthroscopy.12,18,29,31 These controversial reports may be due to the usage of Tönnis grading of hip osteoarthritis in most studies,12,29,31 which was reported to have only fair reliability.11,28,40
Husen et al 12 compared 132 operatively treated hips with nonoperatively treated hips (n = 982 hips) and found that the rate of osteoarthritis progression was lower in the operative group (27% vs 35%) while conversion to hip arthroplasty was similar between groups, which were associated with the presence of cam deformity and male sex, similar to the present study. This differs from the present study in a number of ways. Husen et al included a different subset of patients as a control group rather than the contralateral hip. They also included borderline dysplastic patients and graded osteoarthritis with the Tönnis grading system.
Another study of 100 patients treated with unilateral hip arthroscopy, similar to the present study, using the contralateral hip as the internal control group, reported progression of Tönnis grade in 48% of the nonoperated hips, while there was progression in only 28% of the operated hips. 29 At follow-up, Tönnis grades were equal in 70% of the hips. The operative hips had a better grade 25% of the time, and the nonoperative hips had a better grade 5% of the time. They also used Tönnis grading and included borderline dysplastic patients in their study.
The only study in the literature measuring JSW in unilateral hip arthroscopy patients (mean age, 42; follow-up, 2-155 months) and comparing them with contralateral nonoperated hips reported a decrease in central, medial, and lateral joint width, whereas there was no change reported in the nonoperated side. 18 This was explained by using the analogy of opening Pandora's box. The hypothesis was that exposing subchondral bone and producing cartilage and bone detritus at the impingement zone during femoral osteoplasty elevates the metabolic state and stimulates an inflammatory cascade.10,18 Similar to this, a study by Schmaranzer et al 31 reported a paradoxical delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) index decrease 1 year after joint-preserving hip surgery compared with a smaller dGEMRIC index decrease in symptomatic patients who did not undergo surgical treatment. In contrast to these studies, the present study found a lesser degree of joint space narrowing, which could be a result of longer-term follow-up (mean, 5 years) and/or younger age population (mean age, 35 vs 42 years).
Another controversial issue is the effect of preoperative and postoperative alpha angles on the changes in JSW. As reported, the preoperative alpha angle is associated with the severity of acetabular chondral damage 34 and a greater postoperative alpha angle is associated with a higher frequency of conversion to total hip arthroplasty without reporting Tönnis grades. 39 Another similar study reported higher functional scores and higher total hip arthroplasty–free survivorship rates in patients with cam deformity treated arthroscopically for labral or chondral pathology with femoral osteoplasty compared with patients without femoral osteoplasty. Both groups had similar preoperative Tönnis grades. However, alpha angles and changes in the Tönnis grades during the follow-up were not mentioned. 24 In the present study, only a lesser degree of CJW decrease and/or preservation of joint width approached statistical significance at PASS (+) hips for mHHS (P = .06).
CJW was greater at the final follow-up compared with contralateral hip joint space in patients who had higher preoperative alpha angles (more severe cam deformity). There are 3 possible explanations for this. The first is that cam excision could be more effective in patients with a more severe FAI, leading to better preservation of joint space. Second, although patients had higher preoperative alpha angles, the change in the α0 (preoperative–postoperative) tended to be higher, and the postoperative α0 tended to be lower, without statistical significance (P > .05), which was enough to make a difference and lead to greater JSW. Last, impingement could be symmetric, and patients could have the same lesion on the contralateral hip. In other words, patients operated for more severe cam lesions could possibly have the same severe lesion on the opposite side, leading to more joint space narrowing in the contralateral nonoperated hip. 15
It has been reported that the progression of osteoarthritis after hip arthroscopy is influenced by various patient- and joint-related factors beyond cam morphology. Previous studies have demonstrated that cam-type morphology, advanced age, high BMI, reduced JSW, and more advanced Tönnis grade are important predictors of postoperative osteoarthritis development or conversion to total hip arthroplasty.5,14 In this context, the present study used an intrapatient comparative design with the contralateral hip serving as an internal control, providing a framework to reduce the influence of patient-related and joint-related factors. Nevertheless, further studies are required to more clearly elucidate the relative contributions of these factors to osteoarthritis progression after hip arthroscopy.
Limitations
This study has some limitations. First, it is a single-surgeon (O.H.), small-sized case series without a control group and with the use of retrospective data. Second, the severity of chondrolabral lesions was not graded. Instead, indirect signs of cartilage damage, such as JSW and its change, preoperative alpha angle and the type of labral treatment (debridement or repair), were reported.2,13,30,32,34 Third, JSW measurements were made on supine pelvic radiographs instead of weightbearing radiographs. As reported before, measurements did not differ between the 2, 3 and instead of analyzing the effect of raw JSW, we assessed the changes in the preoperative and postoperative JSW, and JSW differences between operated and nonoperated hips. Last, although the contralateral sides did not receive surgery during the follow-up, we did not score those hips or measure other radiographic parameters such as alpha angle; thus, we cannot be certain that the risk of further generation was the same in both hips.
Conclusion
Patients treated with unilateral hip arthroscopy demonstrated a decrease in the lateral joint width, while nonoperated sides had a decrease in all 3 widths, leading to a greater joint space in the operated hip, especially at the MJW. Furthermore, patients having more severe cam deformity preoperatively had better preservation of CJW.
Footnotes
Final revision submitted January 13, 2026; accepted February 23, 2026.
The authors declared that they have no conflicts of interest in the authorship and publication of this contribution. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethics approval was obtained from the Dokuz Eylul University ethics board (file No. 9145-GOA; decision No. 2024/29-23).
