Written by manual therapists Thomas T. Ødegaard and Roar Syltebø

The German surgeon Friedrich Trendelenburg originated Trendelenburg’s test. In 1895 he became chief surgeon at the University of Leipzig, and it was around this time that he published his hip test. Some of the literature also refers to the Trendelenburg gait, which forms part of Trendelenburg’s full test battery. Trendelenburg’s test is an active functional test intended to assess the stability of the pelvis and hip joint (Trendelenburg, 1998).

Indication

Trendelenburg’s test is indicated in all examinations of the back, pelvis and hip. It is very quick and easy to perform and should be a natural part of the initial active functional examination.

Procedure

The literature describes two variants of the test: single-leg stance without pelvic elevation, and a standardised variant in which the patient actively elevates the pelvis on the side of the swing leg. Both variants are performed on both sides, and gait is assessed as well, since abnormal pelvic and hip function often produces a limp (Trendelenburg, 1998).

Single-leg stance without pelvic elevation

The patient stands with their back to the examiner and is asked to stand on one leg. The examiner assesses whether the patient can keep the pelvis level in the horizontal plane when the other leg is lifted (Trendelenburg, 1998). Grimaldi and colleagues had the patient stand side-on to a wall with the affected leg furthest from the wall and one finger against the wall for balance. The patient lifted the opposite foot with the knee flexed to 90 degrees and the hip in neutral, and held the position for up to 30 seconds (Grimaldi et al., 2017). During the test the examiner asks whether the patient feels pain, and the patient indicates where the pain is located (Grimaldi et al., 2017; Lequesne et al., 2008). Because the test is judged by pain in this variant, light support for balance does not affect the result.

Single-leg stance with active pelvic elevation

In 1985 Hardcastle and Nade described an alternative method. The examiner stands behind the patient and observes the line between the iliac crests relative to the floor. The patient lifts the foot on the side not being tested, with the hip between neutral and 30 degrees of flexion and the knee flexed enough for the foot to clear the floor. According to the authors, the knee is flexed to prevent rectus femoris from contributing. The patient then elevates the pelvis on the swing-leg side as high as possible and holds the position for 30 seconds. With 90 degrees of hip flexion of the swing leg, healthy subjects elevated the pelvis less, because the iliac crest then came closer to the rib cage (Hardcastle & Nade, 1985).

For balance, the patient may use a stick in the hand on the stance-leg side, or the examiner may hold the arm on the stance-leg side or support both shoulders. The patient must not be given support on the swing-leg side. If the patient leans the trunk too far over the stance leg, the examiner corrects this with light pressure on the shoulders so that the vertebra prominens (C7) lies approximately over the vertical line through the stance leg. The examiner records the time before the pelvis begins to drop, if it does (Hardcastle & Nade, 1985).

Which variant is recommended?

In our clinical experience, the variant without pelvic elevation, in which reproduction of the patient’s familiar pain constitutes a positive test, is the easiest to perform and interpret. Pain within 30 seconds also has documented diagnostic value for gluteal tendinopathy, and the location of the pain helps distinguish lateral hip pain from intra-articular pathology (Grimaldi et al., 2017; Fearon et al., 2013).

In our experience it is more difficult to judge when the pelvis begins to drop than to record the patient’s report of pain. Pelvic drop is also seen in healthy individuals and in a range of hip disorders, and it has not shown diagnostic value for greater trochanteric pain syndrome (Hardcastle & Nade, 1985; Kinsella et al., 2023). Body Examination therefore recommends the variant without pelvic elevation. The variant with active pelvic elevation can be used when a more specific assessment of abductor function is wanted.

Interpretation

In single-leg stance without pelvic elevation, the test is positive if the patient’s familiar pain is reproduced within 30 seconds (Grimaldi et al., 2017; Lequesne et al., 2008). The examiner also records whether the pelvis drops on the swing-leg side. In Trendelenburg’s original description, such a pelvic drop was a positive test, and the same is seen in the patient’s gait (Trendelenburg, 1998).

In the variant with active pelvic elevation, the test is negative (normal) if the patient can elevate the pelvis on the swing-leg side as high as abduction of the stance hip allows, and hold the position for 30 seconds with the vertebra prominens (C7) over the hip and foot. The test is positive if the patient cannot do this. This also applies when the pelvis is elevated above the horizontal but not maximally, and when the pelvis is elevated but begins to drop before 30 seconds have passed. A test that becomes positive only after a while is called a delayed positive Trendelenburg’s test, and the time before the pelvis drops provides a measure that can be compared over time (Hardcastle & Nade, 1985).

Pain and the position of the pelvis provide different information and should be assessed separately. The position of the pelvis says something about abductor function but discriminates poorly between different hip disorders. The location of the pain is therefore central to the diagnosis. Single-leg stance places both tensile and compressive load on the gluteal tendons at the greater trochanter, and lateral hip pain over the greater trochanter during the test points to gluteal tendinopathy (Grimaldi et al., 2017). Groin pain, by contrast, is the typical symptom of intra-articular pathology such as a labral tear (Burnett et al., 2006). Lateral hip pain on clinical examination is also among the findings that have best distinguished greater trochanteric pain syndrome from hip osteoarthritis (Fearon et al., 2013). Whatever the findings, Trendelenburg’s test should be followed by a thorough examination, including the passive tests of the hip joint.

The hip joint is a ball-and-socket joint with a large range of motion but, owing to its anatomical structure, also good stability. The hip joint is central to normal gait, and various painful conditions and diagnoses may affect gait. Standing stably with a level pelvis on a single weight-bearing leg requires active, well-functioning gluteal muscles, particularly gluteus medius and minimus.

Friedrich Trendelenburg became a leading surgeon in Germany after spending his first years in the army. He is described as skilled and very practically minded, with extensive experience. Trendelenburg had worked with children and young people and followed them over time. The basis for the test was that he had observed an abnormal gait pattern in young patients with dislocated hips. A dislocated hip changed the anatomy, with increased lumbar lordosis, a markedly tilted pelvis and weak muscles, and thus reduced pelvic stability. Trendelenburg was convinced that an intact gluteus medius was essential for good hip and pelvic function. He developed and described the test on the basis of long clinical experience, but no measurements of the test’s properties were made in the original article (Trendelenburg, 1998).

According to Solberg and Kirkesola, an altered movement pattern or pelvic drop may be due to impaired coordination or recruitment of muscle force because of pain, dysfunction or altered neuromuscular responses. If the patient cannot keep the pelvis level, this may be due to weakness of the abductor muscles of the stance leg (Solberg & Kirkesola, 2007).

Hardcastle and Nade examined 50 healthy subjects and 103 patients with neurological or mechanical disorders of the hip or spine. Healthy subjects who were simply asked to stand on one leg responded in three different ways: the pelvis was elevated, stayed level or dropped on the swing-leg side. Electromyography showed that little abductor activity is needed to keep balance with the pelvis lowered, and that abductor activity increases when the pelvis is elevated with the trunk over the stance hip. The authors therefore considered that the patient must actively elevate the pelvis if the test is to assess abductor function, and that merely observing the position of the pelvis gives little information about it. In patients with neurological disorders, a normal test required full muscle power (MRC grade 5), and all those with grade 4 or lower had a positive test, either immediately or within 25 seconds. A delayed positive test was also seen in hip subluxation and avascular necrosis. In hip osteoarthritis the response varied between patients but was consistent in the individual patient (Hardcastle & Nade, 1985).

The authors also described several sources of error. False-positive findings may be caused by pain, poor balance and the patient not understanding or not cooperating. Nerve root irritation could produce false-positive findings, whereas back pain in itself did not. Two patients with marked scoliosis had a false-positive test because the lower costal margin abutted the iliac crest when they tried to elevate the pelvis, even though abductor strength was normal. Kyphosis did not affect the test (Hardcastle & Nade, 1985).

False-negative findings occur when the patient elevates the pelvis using muscles other than the hip abductors. With the hand supported on the swing-leg side, patients could elevate the pelvis using latissimus dorsi and possibly psoas major on the same side, together with quadratus lumborum on the stance-leg side. Psoas and rectus femoris can also contribute, which is why the knee of the swing leg should be flexed. In addition, the patient may shift the trunk far out over the stance hip, so that the demand on the abductors becomes small. Such compensations were particularly evident in patients with neurological disorders, and electromyography then showed minimal abductor activity. Patients with a painful hip also shifted their centre of gravity over the hip to unload the abductors (Hardcastle & Nade, 1985).

The authors emphasise that a positive test may be misleading in the presence of back or leg pain, deformity, or when the patient does not understand the instructions or does not cooperate, whereas a negative test is significant because it indicates normal hip mechanics. The test cannot be assessed reliably in children under four years of age (Hardcastle & Nade, 1985).

The relationship between the test and abductor strength is nevertheless uncertain. In 2020 McCarney and colleagues examined 36 healthy young adults using 3D motion analysis and isometric and isokinetic dynamometry. They found no significant relationship between abductor strength and frontal-plane pelvic motion during the test, and agreement between a clinician’s visual assessment and the 3D measurements was poor (kappa 0.22–0.25). The study included only healthy individuals, and the results cannot simply be generalised to patients with hip disorders (McCarney et al., 2020). In patients with low back pain, however, visual assessment of the test has shown good test-retest reliability (kappa above 0.75) (Roussel et al., 2007).

There is now a fair amount of research on clinical tests of the hip joint, but its quality is generally low. The studies are often small, frequently compare patients with healthy controls and use uncertain reference standards. This also applies to Trendelenburg’s test, even though it has been used throughout the world for well over 100 years. The most relevant studies of the test’s diagnostic value in different hip disorders are presented below.

In their 2012 book, Cook and Hegedus describe the test as single-leg stance without pelvic elevation, in which the examiner assesses whether the pelvis drops on the swing-leg side. The finding is confirmed during gait, and the test is positive in the presence of asymmetric pelvic drop. They have given this variant an overall utility score of 2, meaning that the evidence moderately supports use of the test (on a scale from 1, strong support, to 3, minimal or no support) (Cook & Hegedus, 2012). The score is based on the studies by Bird, Keeney and Burnett, discussed below, but only Bird and colleagues investigated the test’s diagnostic value. As the studies below show, newer research, particularly the 2023 meta-analysis, calls into question whether pelvic drop on its own has moderate diagnostic value in lateral hip pain (Kinsella et al., 2023). When the same position is instead interpreted by pain within 30 seconds, the test has shown good diagnostic value for gluteal tendinopathy (Grimaldi et al., 2017; Lequesne et al., 2008).

Greater trochanteric pain syndrome

Greater trochanteric pain syndrome (GTPS) is a clinical umbrella term for pain over the lateral aspect of the greater trochanter, and was previously often called trochanteric bursitis. Tendinopathy of the gluteus medius and minimus tendons is now regarded as the main local cause of this pain, and bursitis is often seen together with the tendinopathy. Greater trochanteric pain syndrome and gluteal tendinopathy are therefore largely two sides of the same coin: the former describes the clinical picture, the latter the underlying tissue change. The terms are often used interchangeably in the literature (Grimaldi & Fearon, 2015).

In 2001 Bird and colleagues published a study of clinical testing and MRI in patients with greater trochanteric pain syndrome. 24 patients were examined according to a standard protocol with three clinical tests in addition to MRI. 12 of the patients were retested after two months to determine intra-examiner reliability. All patients were women, with a mean age of 58 years and a mean symptom duration of 12 months (Bird et al., 2001).

MRI showed that 46% of the patients had a gluteus medius tear and 68% gluteus medius tendinopathy. Two patients had trochanteric bursitis, and one had avascular necrosis of the femoral head. The MRI findings were compared with three clinical tests, of which Trendelenburg’s test performed best. For gluteus medius pathology (tear or tendinopathy) the test had a sensitivity of 72.7% and a specificity of 76.9%. Trendelenburg’s test was also the most reliable of the tests, with an intra-examiner kappa of 0.68, corresponding to substantial agreement (Bird et al., 2001).

The study is interesting. The examiners were blinded to other examinations, and Trendelenburg’s test was performed with both single-leg stance and gait. The study nevertheless has clear limitations. There was no control group, and all participants were middle-aged women with the same diagnosis. The radiologists examined only the symptomatic hips. We therefore know little about whether the results can be generalised to other populations. The authors conclude that Trendelenburg’s test is a good clinical test for detecting gluteus medius pathology in this patient group, with acceptable intra-examiner reliability (Bird et al., 2001).

In 2008 Woodley and colleagues examined 40 patients with unilateral lateral hip pain using clinical examination and MRI of both hips. Trendelenburg’s test was performed according to Hardcastle and Nade’s standardised method, with 30 seconds of single-leg stance. None of the clinical tests could reliably detect gluteal tendon pathology. Some tests had likelihood ratios above 2 or below 0.5, but the confidence intervals were very wide, and the authors conclude that no firm conclusions can be drawn about the tests’ diagnostic value (Woodley et al., 2008).

In 2008 Lequesne and colleagues examined patients with refractory greater trochanteric pain syndrome (at least four months) without radiographic osteoarthritis, and a control group without hip pain. MRI showed tendinopathy of the gluteus medius and/or minimus tendons, and/or bursitis adjacent to the tendons, in all patients. Standing on one leg for 30 seconds was interpreted as positive if it produced pain corresponding to the patient’s familiar pain, and had a sensitivity of 100% and a specificity of 97.3%. Here the test was interpreted by pain, not by the position of the pelvis. Specificity was calculated against people without hip pain rather than against patients with other hip disorders, and is therefore probably too high for clinical practice (Lequesne et al., 2008).

In 2017 Grimaldi and colleagues investigated clinical tests in 65 patients with lateral hip pain, using MRI-confirmed gluteal tendinopathy as the reference standard. The patient stood on the affected leg for up to 30 seconds, and the test was positive if the familiar lateral hip pain was reproduced within this time. The test had a specificity of 100% and a positive likelihood ratio of about 12, and shifted an assumed pre-test probability of 50% to a post-test probability of 98%. Absence of tenderness on palpation over the greater trochanter made the diagnosis unlikely. The sample was small, and the authors emphasise that the estimates are uncertain (Grimaldi et al., 2017).

Fearon and colleagues are among the few who have compared patients with lateral hip pain with patients with an intra-articular condition. In 2013 they examined 41 patients with greater trochanteric pain syndrome, 20 with hip osteoarthritis and 23 healthy controls, matched for age and sex. The examination comprised a battery of clinical tests and single-leg stance, and participants indicated where the reproduced pain was located. Lateral hip pain on the FABER test distinguished greater trochanteric pain syndrome from the other groups with a sensitivity of 0.81 and a specificity of 0.82, whereas non-specific hip pain gave a specificity of only 0.28. Of the items in the Harris Hip Score, only maximum walking distance and the ability to put on shoes and socks distinguished greater trochanteric pain syndrome from osteoarthritis. The authors conclude that patients with lateral hip pain who have no difficulty putting on shoes and socks, who are tender on palpation of the greater trochanter and who have lateral hip pain on the FABER test probably have greater trochanteric pain syndrome (Fearon et al., 2013).

A 2023 systematic review with meta-analysis found that Trendelenburg’s test, assessed by the position of the pelvis, did not significantly shift the probability of GTPS in either a positive or a negative direction. The authors recommend palpation of the greater trochanter together with resisted hip abduction as a starting point, and a painful 30-second single-leg stance to confirm the diagnosis (Kinsella et al., 2023).

Labral tears

In 2004 Keeney and colleagues compared MR arthrography with arthroscopy in the evaluation of intra-articular hip pathology. 100 patients (102 hips) with clinically diagnosed labral pathology were included. The mean age was 37.6 years, and 70% were women. The clinical diagnosis was based on anterior groin pain, positive labral provocation tests, exclusion of extra-articular pathology and minimal degenerative changes on plain radiographs. The study was not designed to evaluate Trendelenburg’s test, but it shows which clinical findings are typical in this patient group. MR arthrography detected a labral tear in 76% of the hips. 40 hips, or 39.6% of the patients, had a positive Trendelenburg’s test, and 38% of the patients had a limp. The article does not describe how Trendelenburg’s test was performed (Keeney et al., 2004).

In 2006 Burnett and colleagues published a study of the clinical presentation of patients with arthroscopically confirmed hip labral tears. 66 patients were included, 71% women, with a mean age of 38 years. The typical clinical picture was moderate to severe groin pain. The study was not designed to evaluate Trendelenburg’s test, but 38% of the patients had a positive test and 39% had a limp over a short distance. The study has many weaknesses and is classified as level IV evidence. There is no description of how Trendelenburg’s test was performed (Burnett et al., 2006). The results suggest that Trendelenburg’s test is of limited use in this young patient group. That around 40% of patients with intra-articular pathology had a positive test in both studies also shows that pelvic drop is not specific to lateral hip pain.

Body Examination suggests the following use and interpretation of Trendelenburg’s test

• Trendelenburg’s test is very easy to perform and should be part of the routine active functional examination in all patients with complaints of the hip, back and lower limbs. Both standing function and gait should be assessed.
• Body Examination recommends the variant without pelvic elevation: the patient stands on one leg for 30 seconds, and the test is positive if the familiar pain is reproduced. In our clinical experience this variant is the easiest to perform and interpret.
• The position should be held for 30 seconds, since both pain and pelvic drop often appear only after a while.
• If a more specific assessment of abductor function is wanted, the variant with active pelvic elevation can be used. The patient must then keep the trunk centred over the stance hip and must not lean on support on the swing-leg side, otherwise compensations may mask weak abductor function.
• Pain and the position of the pelvis should be assessed separately, and the patient should always indicate where the pain is located. The location of the pain is more important for the diagnosis than the position of the pelvis.
• Lateral hip pain over the greater trochanter within 30 seconds strongly suggests gluteal tendinopathy, whereas absence of pain does not rule out the diagnosis. Groin pain points to an intra-articular cause.
• Pelvic drop is seen both in lateral hip pain and in intra-articular pathology, and does not distinguish between them.
• Whatever the findings, the test should be followed by a thorough examination, including the graded routine of passive tests of the hip joint.
• Trendelenburg’s test is a useful but very general active functional test. Research on the diagnostic value of pelvic drop is sparse and inconsistent, and the test should not be used on its own to make a specific diagnosis. When greater trochanteric pain syndrome is suspected, it should be combined with palpation of the greater trochanter, resisted hip abduction and the FABER test with assessment of where the pain is located.

Literature search, updating and translation were carried out with the assistance of AI (Claude, Anthropic). The professional content has been reviewed and quality-assured by the editor-in-chief, Roar Syltebø.

References

• Bird PA, Oakley SP, Shnier R, Kirkham BW. Prospective evaluation of magnetic resonance imaging and physical examination findings in patients with greater trochanteric pain syndrome. Arthritis Rheum. 2001 Sep;44(9):2138-45.
• Burnett RS, Della Rocca GJ, Prather H, Curry M, Maloney WJ, Clohisy JC. Clinical presentation of patients with tears of the acetabular labrum. J Bone Joint Surg Am. 2006 Jul;88(7):1448-57.
• Cook CE, Hegedus EJ (2012): Orthopedic Physical Examination Tests: An Evidence-Based Approach (2nd Edition). New Jersey: Prentice Hall.
• Fearon AM, Scarvell JM, Neeman T, Cook JL, Cormick W, Smith PN. Greater trochanteric pain syndrome: defining the clinical syndrome. Br J Sports Med. 2013 Jul;47(10):649-53.
• Grimaldi A, Fearon A. Gluteal tendinopathy: integrating pathomechanics and clinical features in its management. J Orthop Sports Phys Ther. 2015 Nov;45(11):910-22.
• Grimaldi A, Mellor R, Nicolson P, Hodges P, Bennell K, Vicenzino B. Utility of clinical tests to diagnose MRI-confirmed gluteal tendinopathy in patients presenting with lateral hip pain. Br J Sports Med. 2017 Mar;51(6):519-24.
• Hardcastle P, Nade S. The significance of the Trendelenburg test. J Bone Joint Surg Br. 1985 Nov;67(5):741-6.
• Keeney JA, Peelle MW, Jackson J, Rubin D, Maloney WJ, Clohisy JC. Magnetic resonance arthrography versus arthroscopy in the evaluation of articular hip pathology. Clin Orthop Relat Res. 2004 Dec;(429):163-9.
• Kinsella R, Semciw A, Hawke L, Stoney J, Choong PFM, Dowsey MM. Diagnostic accuracy of clinical tests for assessing greater trochanteric pain syndrome: a systematic review with meta-analysis. J Orthop Sports Phys Ther. 2023. doi:10.2519/jospt.2023.11890.
• Lequesne M, Mathieu P, Vuillemin-Bodaghi V, Bard H, Djian P. Gluteal tendinopathy in refractory greater trochanter pain syndrome: diagnostic value of two clinical tests. Arthritis Rheum. 2008 Feb 15;59(2):241-6.
• McCarney L, Andrews A, Henry P, Fazalbhoy A, Selva Raj I, Lythgo N, Kendall JC. Determining Trendelenburg test validity and reliability using 3-dimensional motion analysis and muscle dynamometry. Chiropr Man Therap. 2020 Oct 19;28(1):53.
• Roussel NA, Nijs J, Truijen S, Smeuninx L, Stassijns G. Low back pain: clinimetric properties of the Trendelenburg test, active straight leg raise test, and breathing pattern during active straight leg raising. J Manipulative Physiol Ther. 2007 May;30(4):270-8.
• Solberg AS, Kirkesola G (2007): Klinisk undersøkelse av ryggen. 2nd ed. Kristiansand: Høyskoleforlaget AS.
• Trendelenburg F. Trendelenburg’s test: 1895. Clin Orthop Relat Res. 1998 Oct;(355):3-7.
• Woodley SJ, Nicholson HD, Livingstone V, Doyle TC, Meikle GR, Macintosh JE, Mercer SR. Lateral hip pain: findings from magnetic resonance imaging and clinical examination. J Orthop Sports Phys Ther. 2008 Jun;38(6):313-28.