Gluteal tendon pathology, hip dysplasia, and femoroacetabular impingement

Preliminary descriptive report

Author

Lu Mao

Published

September 14, 2026

Study overview

The dataset contains 1,678 surgical hip records representing 1,674 distinct hips from 1,377 patients, with surgeries from October 2017 through April 2026. Both hips are represented for 297 patients.

Age refers to the earliest available imaging of the corresponding hip, ranges from 11 to 64 years, and may differ between hips within the same patient. All descriptive summaries are based on surgical hip records; patients with multiple records contribute more than once.

Statistical analysis

Categorical variables are summarized as counts and percentages, and continuous variables as medians and interquartile ranges. Tendon pathology was analyzed as an ordered outcome (none, tendinosis, partial tear, complete tear) and as any pathology versus none. Unadjusted associations were evaluated using patient-clustered score tests, with trend tests for imaging-defined dysplasia and overall tests for other findings. Comparisons were also stratified by age and sex. Opposite-hip pathology was summarized by index-hip findings, with confidence intervals obtained by resampling patients. All tests were two-sided; exploratory P values were not adjusted for multiple comparisons.

Descriptive findings

Gluteal tendon pathology was recorded in 244 (14.5%) of the surgical hip records: 190 had tendinosis, 47 had a partial tear, and 7 had a complete tear.

Cohort characteristics

Table 1 summarizes demographics and imaging findings by tendon status. Partial and complete tears are combined in the tear group.

Table 1. Cohort characteristics by tendon status
Overall (N = 1,678) None (N = 1,434) Tendinosis (N = 190) Tear (N = 54)
Age at earliest hip imaging, years
Median (Q1–Q3) 30.0 (20.0–39.0) 28.5 (20.0–38.0) 39.0 (29.0–46.0) 41.5 (35.0–48.0)
Sex
F 1,050 (62.6%) 864 (60.3%) 144 (75.8%) 42 (77.8%)
M 628 (37.4%) 570 (39.7%) 46 (24.2%) 12 (22.2%)
Operated side
Left 778 (46.4%) 672 (46.9%) 86 (45.3%) 20 (37.0%)
Right 900 (53.6%) 762 (53.1%) 104 (54.7%) 34 (63.0%)
Procedure
Arthroscopy 1,511 (90.0%) 1,288 (89.8%) 173 (91.1%) 50 (92.6%)
PAO + Arthroscopy 166 (9.9%) 146 (10.2%) 16 (8.4%) 4 (7.4%)
Open Procedure or Other 1 (0.1%) 0 (0.0%) 1 (0.5%) 0 (0.0%)
Hip dysplasia on imaging
Normal 1,297 (77.3%) 1,104 (77.0%) 150 (78.9%) 43 (79.6%)
Borderline 246 (14.7%) 211 (14.7%) 27 (14.2%) 8 (14.8%)
Dysplastic 135 (8.0%) 119 (8.3%) 13 (6.8%) 3 (5.6%)
FAI on imaging
Normal 360 (21.5%) 301 (21.0%) 47 (24.7%) 12 (22.2%)
Cam 1,271 (75.7%) 1,093 (76.2%) 140 (73.7%) 38 (70.4%)
Pincer 12 (0.7%) 10 (0.7%) 1 (0.5%) 1 (1.9%)
Mixed 35 (2.1%) 30 (2.1%) 2 (1.1%) 3 (5.6%)
Labral tear
Yes 1,620 (96.5%) 1,379 (96.2%) 188 (98.9%) 53 (98.1%)
Acetabular chondral defect
Yes 318 (19.0%) 252 (17.6%) 44 (23.2%) 22 (40.7%)
Femoral chondral defect
Yes 169 (10.1%) 136 (9.5%) 22 (11.6%) 11 (20.4%)
Trochanteric bursitis
Yes 138 (8.2%) 92 (6.4%) 31 (16.3%) 15 (27.8%)
Missing, n 1 1 0 0
Values are n (%) unless otherwise indicated. Percentages use nonmissing observations within each column. F = female; M = male; PAO = periacetabular osteotomy; FAI = femoroacetabular impingement. Tear includes partial and complete tears.

Tendon findings by imaging diagnosis

Table 2 shows the frequency of tendon pathology within each imaging-based dysplasia and FAI category. These are unadjusted comparisons.

Table 2. Tendon pathology by imaging diagnosis
Records, n None Tendinosis Partial tear Complete tear
Hip dysplasia on imaging — Trend: ordinal P = 0.428; any pathology P = 0.432
Normal 1,297 1,104 (85.1%) 150 (11.6%) 38 (2.9%) 5 (0.4%)
Borderline 246 211 (85.8%) 27 (11.0%) 6 (2.4%) 2 (0.8%)
Dysplastic 135 119 (88.1%) 13 (9.6%) 3 (2.2%) 0 (0.0%)
FAI on imaging — Overall: ordinal P = 0.763; any pathology P = 0.759
Normal 360 301 (83.6%) 47 (13.1%) 12 (3.3%) 0 (0.0%)
Cam 1,271 1,093 (86.0%) 140 (11.0%) 31 (2.4%) 7 (0.6%)
Pincer 12 10 (83.3%) 1 (8.3%) 1 (8.3%) 0 (0.0%)
Mixed 35 30 (85.7%) 2 (5.7%) 3 (8.6%) 0 (0.0%)
Values are n (row %). Ordinal tests use the ordering: none, tendinosis, partial tear, complete tear. Binary tests compare any pathology with none. Both assess an ordered trend for dysplasia and an overall association for FAI, using asymptotic score tests that account for repeated records within patients. Tests are unadjusted for age or sex; P values are exploratory.
Figure 1: Tendon-pathology composition by imaging diagnosis. Labels above bars indicate the number of surgical hip records.
  • Tendon-pathology severity and prevalence were similar across imaging-defined dysplasia and FAI categories, with no statistically significant associations.

Tendon findings by postoperative diagnosis

  • Table 3 summarizes tendon-pathology severity by postoperative diagnosis.
  • “Not documented” means the condition was not listed in the recorded postoperative diagnosis; it does not necessarily indicate normal anatomy.
Table 3. Tendon pathology by postoperative diagnosis
Records, n None Tendinosis Partial tear Complete tear
Postoperative hip dysplasia — Overall: ordinal P = 0.189; any pathology P = 0.205
Not documented 1,511 1,286 (85.1%) 174 (11.5%) 45 (3.0%) 6 (0.4%)
Documented 167 148 (88.6%) 16 (9.6%) 2 (1.2%) 1 (0.6%)
Postoperative FAI — Overall: ordinal P = 0.394; any pathology P = 0.356
Not documented 31 29 (93.5%) 1 (3.2%) 1 (3.2%) 0 (0.0%)
Cam 838 712 (85.0%) 94 (11.2%) 28 (3.3%) 4 (0.5%)
Pincer 48 42 (87.5%) 5 (10.4%) 1 (2.1%) 0 (0.0%)
Mixed 761 651 (85.5%) 90 (11.8%) 17 (2.2%) 3 (0.4%)
Values are n (row %). Ordinal tests use the ordering: none, tendinosis, partial tear, complete tear. Binary tests compare any pathology with none. P values are from overall asymptotic score tests accounting for repeated records within patients. Tests are unadjusted for age or sex; P values are exploratory. 'Not documented' does not necessarily indicate normal anatomy.
Figure 2: Tendon-pathology composition by postoperative diagnosis. Labels above bars indicate the number of surgical hip records.
  • These comparisons describe unadjusted associations within the surgical cohort.
  • Small diagnostic groups limit the precision of comparisons.

Tendon findings by age and sex

  • We repeated the comparisons separately by sex and by age group (<20, 20–39, and ≥40 years).
  • Bars show tendon-pathology composition; labels indicate record counts. P values assess associations within each stratum and are exploratory.

By sex

Figure 3: Tendon-pathology composition by imaging diagnosis and sex.
Figure 4: Tendon-pathology composition by postoperative diagnosis and sex.

By age

Figure 5: Tendon-pathology composition by imaging diagnosis and age at earliest hip imaging.
Figure 6: Tendon-pathology composition by postoperative diagnosis and age at earliest hip imaging.
  • Tests account for repeated records within patients. Sex-stratified comparisons are not adjusted for age, and age-stratified comparisons are not adjusted for sex.
  • Small groups and sparse events limit interpretation of stratum-specific P values. A category with no records is labeled n = 0.

Findings in the opposite hip

  • Among patients with both hips represented, we examined tendon pathology in the opposite hip according to dysplasia and tendon tears in the index hip.
  • Each hip served as the index hip in turn. We retained the earliest surgical record per hip and accounted for paired observations within patients.
  • Dysplasia denotes the imaging category “Dysplastic”; borderline hips are included among hips without definite dysplasia.
Tendon pathology in the opposite hip
Overall association: P = 0.152
Index-hip findings Index hips, n Patients, n Opposite-hip pathology, n (%) 95% CI
Neither 538 285 74 (13.8%) 10.5–17.3%
Dysplasia only 45 35 4 (8.9%) 2.0–18.8%
Tear only 11 9 6 (54.5%) 11.1–87.5%
Dysplasia and tear 0 0 Not estimable
Tear includes partial and complete tears. 'Neither' means neither definite dysplasia nor a tear; tendinosis may still be present in the index hip. Patients may contribute to more than one row. Confidence intervals use 2,000 patient-level bootstrap samples. The overall P value accounts for patient clustering and is exploratory. Intervals are not reported when bootstrap estimation is degenerate or too few resamples contain the group.
  • Opposite-hip tendon pathology was more frequent when the index hip had a tear (54.5% versus 13.8% with neither dysplasia nor a tear), but this estimate was based on few patients and was imprecise.
  • No index hips had both dysplasia and a tear, so their combined association with opposite-hip pathology could not be assessed

Chondral defects and tendon pathology

  • We compared tendon findings between hips with and without acetabular or femoral chondral defects.
  • Comparisons are unadjusted and account for repeated records within patients.
Tendon pathology by chondral-defect status
Records, n None Tendinosis Partial tear Complete tear
Acetabular chondral defect — Ordinal P = 0.002; any pathology P = 0.002
Absent 1,360 1,182 (86.9%) 146 (10.7%) 27 (2.0%) 5 (0.4%)
Present 318 252 (79.2%) 44 (13.8%) 20 (6.3%) 2 (0.6%)
Femoral chondral defect — Ordinal P = 0.070; any pathology P = 0.083
Absent 1,509 1,298 (86.0%) 168 (11.1%) 38 (2.5%) 5 (0.3%)
Present 169 136 (80.5%) 22 (13.0%) 9 (5.3%) 2 (1.2%)
Values are n (row %). Ordinal tests use the ordering: none, tendinosis, partial tear, complete tear. Binary tests compare any pathology with none. P values account for repeated records within patients but are unadjusted for age, sex, or multiple comparisons. Records missing the relevant finding are excluded.
Figure 7: Tendon-pathology composition by acetabular and femoral chondral-defect status. Labels indicate record counts.
  • Acetabular chondral defects were associated with more frequent and more severe tendon pathology (ordinal and binary P = 0.002).

Trochanteric bursitis and tendon pathology

We compared tendon findings between hips with and without trochanteric bursitis, accounting for repeated records within patients.

Tendon pathology by trochanteric-bursitis status
Ordinal P <0.001; any pathology P <0.001
Bursitis Records, n None Tendinosis Partial tear Complete tear
Absent 1,539 1,341 (87.1%) 159 (10.3%) 35 (2.3%) 4 (0.3%)
Present 138 92 (66.7%) 31 (22.5%) 12 (8.7%) 3 (2.2%)
Values are n (row %). Ordinal tests use the ordering: none, tendinosis, partial tear, complete tear. Binary tests compare any pathology with none. P values account for repeated records within patients but are unadjusted for age, sex, or multiple comparisons. Records missing bursitis status are excluded.
Figure 8: Tendon-pathology composition by trochanteric-bursitis status. Labels indicate record counts.
  • Trochanteric bursitis was associated with more frequent and more severe tendon pathology (ordinal and binary P < 0.001). Any pathology occurred in 33.3% of hips with bursitis versus 12.9% without bursitis.