Functional pelvic tilt frequently differs from the anterior pelvic plane in non-arthritic hip disorders

Abstract

Aims

Hip preservation surgeons are increasingly using commercially available 3D motion analysis software to investigate areas of impingement and quantify femoral head coverage. Variations in functional pelvic tilt will affect the position of the acetabular rim and projected femoral head coverage, but currently the majority of available software standardizes sagittal rotation to the anterior pelvic plane (APP). The study hypothesis was that the APP does not correlate well with patient-specific pelvic position.

Methods

In total, 60 patients were selected from an institutional database: 20 with acetabular retroversion (AR), 20 with developmental dysplasia of the hip (DDH), and 20 with cam-type femoroacetabular impingement (FAI). Multiplanar CT reformats were created and the sagittal rotation was aligned to the APP. The sagittal pelvic orientation was then corrected until the anteroposterior (AP) projection mirrored that of their preoperative supine and standing plain radiographs. The change in sagittal pelvic tilt angle required was measured.

Results

The mean deviation from the APP in the AR group was 11.55° (SD 4.96°) for supine radiographs and 3.28° (SD 8.66°) for standing radiographs. The mean deviation from the APP in the DDH group was 12.2° (SD 4.26°) for supine radiographs and 6.96° (SD 3.43°) for standing radiographs. The main deviation from the APP in the FAI group was 8.63° (SD 5.21°) for supine radiographs and -1.28° (SD 7.31°) for standing.

Conclusion

There is a wide variation in patients’ functional pelvic positioning in both supine and standing radiographs, in all different subgroups, which rarely correlates with the APP. Commercial 3D motion analysis may therefore give misleading results for both the extent and location of hip impingement as well as femoral head coverage, which may affect surgical decision-making. Consideration should be given to incorporating this into the software algorithms.

Cite this article: Bone Jt Open 2025;6(6):651–657.

Introduction

Non-arthritic hip pain is a significant cause of morbidity in younger patients, and may be caused by a wide variety of pathology leading to either situations of instability, such as in acetabular dysplasia and high femoral torsion, or femoroacetabular impingement (FAI) from cam or pincer lesions as well as acetabular retroversion (AR) and femoral retrotorsion.^1-3 Operative management options for these conditions can include periacetabular osteotomy (PAO), proximal femoral osteotomy (PFO), hip arthroscopy, or combinations thereof. Decision-making around choice of surgical procedure is based around clinical presentation, examination findings, and interpretation of imaging to evaluate femoral head coverage, acetabular shape and orientation, and areas of bony impingement.

The effect of sagittal pelvic tilt on the radiological appearances of developmental dysplasia of the hip (DDH), FAI, and AR have previously been studied. The change in pelvic tilt between supine and standing positions has been shown to be between 6° and 8° in DDH patients,^4,5 and between 5° and 8° in AR patients with the pelvis tilting posteriorly when moving from supine to standing.⁶ The absolute values of pelvic tilt in supine patients have also previously been reported for acetabular retroversion (1° to 4°),^6,7 osteoarthritis patients before (4.4°) and after (6.3°) total hip arthroplasty,⁴ FAI (8°),⁸ and DDH (9°).⁹ The extent of the variation in pelvic tilt between patients within a group, however, has not previously been assessed and reported in detail.

To aid in the interpretation of imaging, as well as surgical planning, hip preservation surgeons are increasingly using commercially available 3D motion analysis software to investigate areas of impingement and to quantify femoral head coverage and acetabular version.^10,11 The sagittal rotation of the pelvis is an important factor here as variations in pelvic tilt will affect the position of the acetabular walls, the projected femoral head coverage, and the potential for modelled bony impingement through a simulated range of motion. Currently, however, the majority of available software standardizes sagittal pelvic rotation to the anterior pelvic plane (APP). This may not reliably represent the functional position of an individual patient,¹² and therefore may subsequently lead to erroneous software analysis of acetabular position and coverage, as well as predicted impingement areas through a simulated range of motion. The extent of changes to pelvic tilt required to align imaging to the APP in various clinical scenarios, along with how this could affect assumptions made by motion analysis software, has not been previously studied.

The hypothesis of this study was that the APP would not correlate well with patient-specific sagittal pelvic position in patients presenting with non-arthritic hip disorders.

Methods

This was a single-centre, retrospective radiological study on consecutive patients who had undergone hip preservation surgery for DDH or FAI. Following institutional review board approval, 60 patients with adequate available imaging were selected from the institutional database: 20 with AR, 20 with DDH, and 20 with cam-type FAI. Equal numbers of male and female patients were included for each diagnosis. The mean age was 26 years for AR patients (SD 3.30), 26 years for DDH patients (SD 3.03), and 27 years for FAI patients (SD 3.36). Patients were excluded if they did not undergo surgery for their diagnosis, or if they had inadequate imaging. Their preoperative anteroposterior (AP) standing, supine pelvic radiographs, and pelvic CT imaging were analyzed from the perspective of their functional sagittal pelvic positioning as follows.

All AP pelvic radiographs were taken following our hospital protocol: the beam was directed perpendicular to the table for supine radiographs, perpendicular to the standing patient for standing radiographs, and towards a point midway between the pubic symphysis and the line connecting the anterior superior iliac spines, with a focus distance of 115 cm from the film and the lower limbs internally rotated 15°. Radiographs were considered adequate for coronal plane rotation if the coccyx was in line with the pubic symphysis, and the iliac wings, obturator foramina, and radiological teardrops appeared symmetrical.

CT scans were performed on a Canon Aquilion PrimeSP CT scanner (Canon Medical Systems, Japan) using a pelvic protocol with 1 mm slices. Multiplanar CT reformatted projections were created and corrected for axial and coronal rotation before the sagittal rotation was aligned initially to the APP. The APP is defined as a vertical line connecting the anterior superior iliac spines and the pubic tubercle.¹³ The sagittal pelvic orientation was then corrected until the AP projection mirrored that of the supine and standing plain radiographs. This was achieved using the pubic symphysis to sacroiliac (PS-SI) index, the ratio of the length of a line drawn from the superior border of the centre of the pubic symphysis to its intersection with the sacroiliac line (a line drawn between the inferior aspect of the sacroiliac joints), and the length of the sacroiliac line¹⁴ and the obturator foramen ratio¹⁵ (Figure 1). Figure 2 demonstrates the change in sagittal rotation in an indicative case. The change in sagittal pelvic tilt angle required to change the CT image from APP to the supine and standing positions could then be measured (Figure 3).

Fig. 1.

Supine and standing radiographs and equivalent CT projections of a 25-year-old female, with sagittal pelvic projection matched by the pubic symphysis to sacroiliac (PS-SI) index and obturator foramen ratios.

Fig. 2.

CT-generated images of a 25-year-old female aligned in the sagittal plane to standing and supine radiological projections after correction. The angle depicts a vertical line and the anterior pelvic plane.

Fig. 3.

CT-generated images of a 25-year-old female aligned in the sagittal plane to the anterior pelvic plane (APP), standing, and supine positions.

A subset of measurements were performed by two separate observers (MRJJ, JH) at two separate timepoints at least 14 days apart to evaluate for inter and intraobserver reliability. Inter-rater reliability as measured by Cronbach α was 0.919.

Statistical analysis

Results were reported as means with standard errors (SEM). Normal distribution was assessed using D’Agostino & Pearson test. Each set of results was compared to the APP using an independent-samples t-test. Variance was assessed using an ordinary, one-way analysis of variance (ANOVA) test. Analysis was performed using Prism-GraphPad (GraphPad, USA). A p-value < 0.05 was considered significant.

Results

The measurements of pelvic tilt from the APP to supine and standing positions were all normally distributed (Table I). There was a significant difference between the mean pelvic tilt and APP in AR patients, both supine and standing, in DDH patients both supine and standing, and in FAI patients in the supine position. This did not occur in FAI patients in the standing position who had pelvic tilt either side of the APP (Table II).

Table I.

Normal distribution analyses using D'Agostino & Pearson test.

Group	APP to standing	APP to supine
AR	K2: 0.295, p = 0.863	K2: 0.692, p = 0.708
DDH	K2: 1.089, p = 0.580	K2: 0.415, p = 0.813
FAI	K2: 2.425, p = 0.297	K2: 0.546, p = 0.761

APP, anterior pelvic plane; AR, acetabular retroversion; DDH, developmental dysplasia of the hip; FAI, femoroacetabular impingement.

Table II.

Comparison of pelvic tilt to anterior pelvic plane (APP) using independent-samples t-test.

Group	APP to standing	APP to supine
AR	t = 4.656, df = 19, p < 0.001	t = 13.86, df = 19, p < 0.001
DDH	t = 9.078, df = 19, p < 0.001	t = 17.06, df = 19, p < 0.001
FAI	t = 1.315, df = 19, p = 0.204	t = 11.39, df = 19, p < 0.001

AR, acetabular retroversion; DDH, developmental dysplasia of the hip; df, degrees of freedom; FAI, femoroacetabular impingement.

The mean deviation from the APP in the AR group was 11.55° (SD 4.96°) for supine radiographs and 3.28° (SD 8.66°) for standing radiographs. The mean deviation from the APP for standing radiographs in the DDH group was 6.96° (SD 3.43°) and -1.28° (SD 7.31°) in the FAI group (Figure 4). Ordinary, one-way ANOVA analysis demonstrated that there was significant variance between the diagnoses from APP to standing (F = 23.5, p < 0.001). The mean deviation from the APP for supine radiographs in the DDH group was 12.2° (SD 4.26°) and 8.63° (SD 5.21°) for the FAI group (Table III, Figure 5). Ordinary, one-way ANOVA analysis demonstrated that there was significant variance between the diagnoses from APP to supine (F = 5.99, p = 0.004).

Fig. 4.

Mean deviation from anterior pelvic plane (APP) to standing. AR, acetabular retroversion; DDH, developmental dysplasia of the hip; FAI, femoroacetabular impingement.

Table III.

Mean measurements of pelvic tilt.

Group	APP to standing	APP to supine
AR	3.28° (SD 8.66°)	11.55° (SD 4.96°)
DDH	6.96° (SD 3.43°)	12.2° (SD 4.26°)
FAI	-1.34° (SD 5.79°)	8.63° (SD 5.21°)

APP, anterior pelvic plane; AR, acetabular retroversion; DDH, developmental dysplasia of the hip; FAI, femoroacetabular impingement.

Fig. 5.

Mean deviation from anterior pelvic plane to supine. AR, acetabular retroversion; DDH, developmental dysplasia of the hip; FAI, femoroacetabular impingement.

Discussion

This is the first study to demonstrate the wide variance of functional sagittal pelvic tilt angles for DDH, FAI, and AR from the APP, and the subsequent implications for the clinical application of commercial 3D motional analysis software. There is significant interest in the arthroplasty world in how pelvic tilt, resulting both from functional positioning and spinopelvic motion variance, affects the positioning of the acetabular component and thus the potential outcome of the surgery.^16-19 Patients undergoing total hip arthroplasty (THA) have also been shown to have a wide range of functional pelvic tilt (−19.0° to 17.9°) which can affect the functional positioning of acetabular components following THA.²⁰ This study has demonstrated significant variability in functional pelvic positioning in both supine and standing radiographs in patients with non-arthritic hip disease. These values are not consistent or predictable and importantly do not correlate with the APP. They also vary significantly depending on the diagnosis. These findings suggest that commercial 3D motion analysis software in its current iteration may give misleading results in many cases for both the extent and location of hip impingement, as well as femoral head coverage and acetabular version. This is important as both the accuracy and the interpretation of the information they provide is designed to aid both surgical decision-making and intraoperative correction parameters. This may be mechanistic, such as deciding whether the main issue is one of impingement or instability or more technical, as this information may be used for planning of the positioning of the acetabular fragment in a PAO or the extent of bony debridement during hip arthroscopy, either at the acetabular rim or from the femoral head-neck junction. This latter aspect is often included in the software package output in addition to the overall analysis.

In non-arthritic hip disorders, converting all patients’ sagittal pelvic tilt to the APP prior to analysis risks misinterpretation of their functional anatomy. This affects not only the absolute measurements by which we define morphology, but also the potential consequences. The relationship between pelvic tilt and measured acetabular version has previously been described, with one degree in change in pelvic tilt causing a one-degree change in acetabular version in patients with dysplasia, AR, and a normal control group.^12,21-23 Uemura et al²⁴ demonstrated when tilting the pelvis from posterior to anterior, every one-degree change in tilt led to a 0.5% increase in anterior coverage and a 0.3% decrease in posterior coverage. Ross et al²¹ used 3D models from CT scans to demonstrate how dynamic changes in pelvic tilt influence the functional position of the acetabulum, as well as the range of motion to impingement during various movements.

The pelvic tilt measurements in this study are similar to those reported in previous studies, with the exception of the supine pelvic tilt measurements in AR, which has a similar change in pelvic tilt measurements from supine to standing.^4-9 This observation confirms the high variation in pelvic tilt that we observed in this study, and that the APP does not correlate well with the functional patient position, nor is the variation predictable.

Kitamura et al⁵ showed that individual and postural variations in the physiological pelvic tilt have previously been demonstrated to affect joint contact pressure in the hip. These two studies evaluated joint contact area, contact pressure, and equivalent stress of the acetabular cartilage at three pelvic tilt positions – when compared to a control group, it was found that sagittal pelvic tilt altered the loading environment and joint stress distribution of the hip joint, potentially impacting the degeneration process in dysplastic hips. The effects on contact pressure and equivalent stress were more significantly affected in dysplastic as opposed to normal hips, which further reinforces the importance of fragment positioning during periacetabular osteotomy (PAO) correction and thus how we utilize our preoperative information to inform our surgical process.

The findings in this study of the large variation in physiological pelvic tilt between patients, independent of their underlying pathology, therefore suggests that these patients would have a wide variation in functional acetabular version and femoral head coverage. This observation is independent of other anatomical variations which further affect version and coverage independently of the effect of pelvic tilt.¹²

This study has limitations. It is a retrospective study, which means the availability and standard of some data is limited, but prospective collection of data would have increased the length of time it took to complete the study considerably, without providing any benefit with regard to validity, reliability, or reproducibility of the findings. The number of patients we have included remains a small sample. This is partly because it was important to have sufficiently high-quality imaging for analysis. Equally, regular collection of both supine and standing AP pelvic radiographs has been a relatively recent addition to the non-arthritic hip pain assessment in our unit.⁶ This, combined with the overall small number of AR patients (a consequence of the relative infrequency with which patients present with this diagnosis) is why the numbers per group were chosen. Equal numbers of patients in each group were prioritized over larger numbers of one group. FAI can be caused by abnormalities on both the acetabular and femoral side of the hip joint, and this study only included patients with cam-type FAI or AR as distinct groups. Pincer-type FAI which was not caused by AR, and patients with a mixed picture of both cam type and pincer-type FAI, were therefore missed from the study. We did not include additional components to both instability and impingement factors, such as femoral torsion, although these would have additional influences on the outcome of any 3D analysis. Overall, however, even with the small numbers, we have still demonstrated a significant variation throughout the cohort, which is clinically relevant in the context of modern computational analysis.

This study’s hypothesis, that the APP does not correlate well with patient-specific pelvic position, has been supported by the results. The wide variation in patients’ functional pelvic positioning in both supine and standing radiographs may cause commercial 3D motion analysis software to give misleading results for both the extent and location of hip impingement as well as femoral head coverage and acetabular version. This may affect surgical decision-making, both in terms of causation for situations of both instability and impingement as well as surgical planning for PAO fragment position or arthroscopic bone resection. This could lead to under- or over-correction of deformities during PAO or hip arthroscopy. In turn, this may cause suboptimal patient outcomes with ongoing pain and ultimately failure of the hip joint requiring arthroplasty. The dynamic nature of the causative pathology suggests that future iterations of analysis software should take into account the individual patient’s functional pelvic tilt and correct to a functional pelvic position, preferably more than one, rather than the APP. This will allow a better assessment of acetabular version, femoral head coverage, and impingement zones, increasing the reliability of the findings provided to the surgeon.

Take home message

- There is a wide variation in patients’ functional pelvic positioning in both supine and standing radiographs which rarely correlates with the anterior pelvic plane.

- Commercial 3D motion analysis may therefore give misleading results for both the extent and location of hip impingement, as well as femoral head coverage, which may affect surgical decision-making.

Data Availability

The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request.