The Role of the Pelvis in Hamstring Injuries & Posterior Thigh Pain

Diane Lee & LJ LeeThe following article about the pelvis’ role in hamstring injuries & posterior thigh pain was written by Diane Lee,BSR and LJ Lee, BScPT, PhD (candidate). Diane and LJ are internationally recognized physiotherapists, authors, and educators. The article is reproduced with their permission. Please see Diane’s site and LJ’s site for more excellent articles.

Hamstring injuries are common in many sports, especially those involving running and rapid acceleration and deceleration such as track, soccer / football, cricket, rugby and tennis. In the Australian Football League (AFL) hamstring injuries are the most common injury, resulting in the most games missed, with an average of 6 hamstring injuries per club per season (Orchard & Seward 2002) and are responsible for 16% of missed playing time (Seward et al 1993). The musculotendinous junction of biceps femoris is cited as being most commonly involved (53%; Woods et al 2006, & 74%; Orchard & Seward 2002). It is thought that the injury most likely occurs during terminal swing phase, just before foot strike, during sprinting (Heiderscheit et al 2005). Re-injury is common; 30% will recur on return to sport (Verral et al 2001) and 12% of these will recur in the first week back to sport. According to Verral et al (2001) the risk factors for sustaining a hamstring strain include:

  1. A past history of posterior thigh injury (4.9 times increased risk)
  2. A past history of knee or groin injur
  3. A higher age (> 23 years)

Warren et al (2008) identified several clinical predictors of return to competition following hamstring strain; if it takes longer than 24 hours to be able to walk without pain and if there is a history of previous injury, it will take longer than three weeks to return to play.

Differentiating a true hamstring strain from posterior thigh pain is often reported to be a clinical challenge (Brukner & Kahn 2007). Referred posterior thigh pain is common with or without a previous hamstring strain and Verral et al (2001) notes that a history of injury to the low back does not increase the risk for a hamstring injury. A low back injury does, however, correlate with increased risk of referred pain to the posterior thigh. They also note that 10% of all posterior thigh “injuries” show no structural changes on magnetic resonance imaging

Why do we believe it is critical to consider the role of the pelvis in all hamstring injuries and presentations of posterior thigh pain?

The pelvis is a key area for load transfer between the lower extremity and the spine and a stable “platform” is essential for optimal function of the lower extremity (Lee 2004, Lee & Lee 2008a). Non-optimal strategies for load transfer and movement can result in poor control of the joints of the pelvis and consequently an “unstable” or poorly controlled platform. This can result in apparent weakness and loss of power in one or both lower extremities during multiple tasks (Lee & Lee 2008a). For the athlete involved in sports requiring running, this is particularly evident during hip extension task

What should the pelvic girdle do during single or double leg loading tasks?

Pelvis in Hamstring Injuries single leg stand - Figure 1Fig. 1 One leg standing, intrapelvic motion control. The innominate is palpated with one hand while the sacrum is palpated at either the spinous process of S2 or the ipsilateral ILA. The two bones should move as a unit as the pelvic girdle shifts laterally over the weight bearing lower extremity. Watch and feel for early, middle or late anterior rotation of the innominate relative to the ipsilateral sacrum on the weight bearing side as the patient transfers their weight (this is non-optimal). Remember that the amplitude of motion for the SIJ is very small in weight bearing (4 – 6°) and even when the SIJ unlocks completely during this task the movement is very subtle, yet palpable.

Rotation of the innominate relative to the sacrum occurs during tasks that load the pelvis asymmetrically. Using reflective surface markers on 15 bony landmarks of the femur, innominate and sacrum and a motion analysis imaging system (six-camera expert vision motion analysis hi res 5.0 system), Hungerford et al (2004) investigated the osteokinematic motion of the innominate relative to the sacrum during single leg standing and contralateral hip flexion to 90° in both non-painful and pelvic girdle pain populations. They found that when a healthy subject stood on one leg and flexed the contralateral hip (Fig. 1), the supporting innominate (weight bearing side) either posteriorly rotated or did not move relative to the ipsilateral sacrum. It is thought that in upright standing the sacrum is nutated relative to the innominates (Sturesson et al 2000) and thus if the innominate on the weight bearing side did not move or posteriorly rotated, the sacrum either remained relatively nutated or had relatively increased nutation

The SIJ is thus close-packed in preparation for load transfer. The non-weight bearing innominate (side of hip flexion) also posteriorly rotated relative to the ipsilateral sacrum during this motion. NB, a difference between the weight bearing and non-weight bearing sides was that the bones (innominate and sacrum) moved towards each other on the weight bearing side (increased compression) but moved away from each other on the non-weight bearing side (decompression). This highlights that although the right and left SIJs were similarly positioned (i.e. nutated), there were different forces acting across the joints to provide optimum mechanics suited to the different demands on each leg in this task.

Fig. 2 Mean onset timing (measured via surface EMG) of various muscles of the supporting leg during one leg standing. When the subjects with pelvic girdle pain were compared to healthy controls, differences were noted in the onset of certain muscle activity during this task. In the pelvic girdle pain group there were delays in the time of onset of IO / TrA, MF and GMax and early activation of the biceps femoris. Note that the medial hamstring muscles (semintendinosus and semimembranosus) were not measured in this study and that the inclusion criteria meant that a specific subgroup of subjects with pelvic girdle pain was studied (Hungerford et al 2003). Other patterns of muscle activity (onset timing) likely exist in other patient subgroups.
Mean onset timing single leg stand - Figure 2

Anterior rotation and / or decompression of the weightbearing innominate relative to the sacrum is non-optimal if it occurs during this task since this is the loose-packed or unlocked position for the SIJ and therefore a less robust position for the transference of loads. Anterior rotation of the weight bearing innominate occurred in the subjects with unilateral pelvic girdle pain (Hungerford et al 2004). In this same series of studies surface EMG was used to measure the activation of internal / oblique, lumbar multifidus, biceps femoris, adductor longus, gluteus maximus, medius and tensor fascia lata in relationship to initiation of motion of weight shift (Fig. 2) (Hungerford et al 2003)

In the non-painful population an increase in activity of the low horizontal fibres of IO / TrA as well as the lumbar multifidus occurred before the onset of motion (as defined by the transfer of weight on a force plate). In addition gluteus maximus onset was noted to occur before biceps femoris, although this occurred after the onset of motion. A different pattern of muscle onset timing was noted in the group with pelvic girdle pain. In this group there was a delay in the activation of the low horizontal fibres of IO / TrA and lumbar multifidus (activity occurred after the onset of motion) and a reversal in the timing of activation between biceps femoris and gluteus maximus (BF activated before GMax). We believe that this altered neural patterning could predispose the biceps femoris muscle to injury in conditions of either repetitive use or sudden loading

What does the research tell us about the best way to manage hamstring injuries?

Mason et al (2008) conducted a systematic review of all RCTs in the Cochrane database up to February 2006 that investigated the effect of one rehabilitation strategy in isolation or in combination with another compared to another strategy or control performed on individuals with a hamstring injury. The objective was to evaluate the effectiveness of all rehabilitation protocols to restore the return to full strength, range of motion and function in those individuals presenting with all forms of hamstring injury regardless of site, severity, onset or level of chronicity. Effectiveness was determined by the time it took to return to play. Only three trials met the selection criteria; no trial considered intervention versus rest. Essentially, their findings were that there was no evidence to support anything other than perhaps increasing the frequency of stretching has an impact on the rate of return to function. Interestingly, one of the three RCTs (Sherry & Best 2004) found a significant difference in the re-injury rates at both two weeks and 1 year post return to play when programmes which included progressive agility, trunk stabilisation and icing were compared to traditional stretching, strengthening and icing. Since these outcomes were not being measured as part of this systematic review it was not considered a significant finding.

What do the experts in the field suggest?

Fig. 3 From Brukner and Khan (2007), key differential findings of a hamstring tear vs hamstring pain.
Hamstring Tear & Hamstring Pain - Figure 3

In evidence-based practice it is important to consider not only the best available scientific evidence but also the opinions of clinical experts in the related field. Brukner and Khan (2007) suggest that posterior thigh pain that is not the result of hamstring strain will require skilful clinical reasoning to determine the cause. Figure 3 tabulates their opinion on how to differentiate a torn hamstring from hamstring pain.

While there are some notable differences, they suggest that both the torn hamstring and the painful hamstring (referred pain) may have abnormal lumbosacral and SIJ signs and thus cannot be differentiated by these findings alone. Unfortunately, they do not elaborate on what these signs could be and from clinical experience we feel that it is possible to differentiate the two entities – true hamstring muscle injury vs. posterior thigh pain – with specific objective tests.

The system-based classification approach to differential diagnosis and management of hamstring injury and posterior thigh pain

Using a system-based classification approach (Lee & Lee 2007, 2008a,b) (Fig. 4) we believe that it is possible to determine when the pelvis is a key contributor to both single or recurrent hamstring injury and / or posterior thigh pain. A major and common complaint of those with either recurrent hamstring strain or posterior thigh pain is a loss of power during hip extension tasks. This is often reported during the subjective examination and becomes one of the meaningful tasks examined in the system-based classification approach.

Fig.4 This graphic is a useful tool for organising the key information gained through the examination process and for reflection and interpretive reasoning of the findings.
Lee & Lee Clinical Puzzle - Figure 4

The “clinical puzzle” is a graphic that conceptualises the system-based classification approach for disability and pain. It is used clinically as a tool for clinical reasoning and decision-making. The outer circle represents the meaningful tasks that the patient is having difficulty performing and is where the relevant strategies they use for function and performance are recorded. These tasks are determined from listening to the patient’s story. Impairment in any piece(s) or loss of congruence and synergy between the pieces of the puzzle (the “systems”) within the outer circle can “drive” non-optimal strategies for function and performance. The centre piece of the puzzle represents several systems that relate to the person and the sensorial (sensations, perceptions), cognitive (beliefs, attitudes, motivations) and emotional (fears, anger, anxiety) dimensions of their current experience. It also includes systemic systems (such as endocrine balance, immune function) and genetic factors. It is the place where primary symptoms, goals and barriers to recovery are noted. The four other pieces of the puzzle represent the various systems in which impairments are assessed and noted during the clinical examination. During this process, the therapist also considers and reflects upon the relationship of these impairments to the person in the middle of the puzzle, e.g. the meaning these impairments may or may not have and the impact these impairments may have on the non-optimal strategies for function and performance during meaningful tasks. All clinical puzzles are unique since no two individuals have the same life experiences

Meaningful task analysis of hip extension – prone knee bend and hip extension

Fig. 5 Prone knee bend. In this figure the therapist is palpating the right SIJ and notes whether the innominate anteriorly rotates as the patient bends the ipsilateral knee. Optimally, the pelvic girdle should remain in a stable, self-locked position (innominate relatively posteriorly rotated compared to the sacrum) and should not unlock during this task
Prone knee bend - Figure 5

These two tasks were originally developed to assess the role of the pelvis in patients with recurrent hamstring injuries and runners with hamstring or ischial symptoms (possible referred pain) during the mid-stance to toe-off phases of gait. Some of the athletes did not have pain but simply reported decreased power on push off. These tests help to determine when the pelvis needs to be addressed in order for full return to function and decreased risk of recurrence.

From the prone position ask the patient to bend one knee to 90° flexion (a prone knee bend) and then to do the same with the other leg. Note any reproduction of pain and if pain is present in the posterior thigh, note whether it is medial or lateral, and / or mid-belly. Ask the patient to then think about the effort required to initiate the movement of the leg off the table as they repeat the movements and to report if it feels more difficult to bend / lift one leg than the other. Note which leg is heavier and repeat the prone knee bend with this leg while you palpate and note:

  • Any intrapelvic torsion
  • Any segmental hinging into extension or flexion in the lumbar spine
  • Any loss of control of the femoral head in the acetabulum
  • Any loss of control (anterior rotation, also termed ‘unlocking’) of the ipsilateral or contralateral SIJ (Fig. 5)

Fig. 6 Prone knee bend. Poor control of the joints of the pelvic girdle is often responsible for apparent hamstring weakness. In this illustration the therapist is comparing the subjective “strength” response of the hamstrings when the pelvis is passively compressed posteriorly (force closure is increased). The location of compression can be varied to note the best response. A subjective change in effort is used to guide and interpret the test
Prone knee bend - Figure 6

Optimally, the pelvis should remain neutral, the lumbar spine should not hinge into extension at any segment, the femoral head should remain centered in the acetabulum and the innominate should not anteriorly rotate relative to the sacrum at either SIJ. The following modifications are then applied if failed load transfer is noted in any of the areas above. Note if there is a change in effort to perform the task while:

  1. the sacrum is nutated passively (focused to the SIJ where unlocking was noted)
  2. the pelvic girdle is aligned & compressed
    1. bilaterally across the anterior aspec
    2. bilaterally across the posterior aspect (Fig. 6)
    3. obliquely from left anterior to right posterior
    4. obliquely from right anterior to left posterior
  3. the femoral head position is corrected and manually controlled

Fig. 7 Resisted hamstrings test. In this illustration the therapist is comparing the ‘strength’ response of the hamstrings with the sacrum passively nutated at the right SIJ (thereby augmenting force closure). If increased strength and / or decreased pain is noted when the sacrum is passively nutated it is likely that pelvic girdle control is a factor in the original weakness noted, i.e. when the test is performed without nutation of the sacrum
Resisted hamstrings test - Figure 7

The test and its modifications can be performed at a higher load by using isometric resistance (manual muscle test of the hamstrings) (Fig. 7)

Note any tendency for the tibia to rotate (an effort to or indication of bias between the medial or lateral hamstrings), any reproduction of pain as well as any loss of control in the lumbar spine, pelvis or hip.

This resisted test is useful for those patients who cannot feel a difference in effort between the right and left sides in the prone knee bend alone.

Subsequently, ask the patient to extend the hip while keeping the knee extended and note:

  • Any intrapelvic torsion
  • Any segmental hinging into extension in the lumbar spine
  • Any loss of control of the femoral head in the acetabulum
  • Any loss of control of the ipsilateral or contralateral SIJ
  • The timing of loss of control of either the lumbar spine, SIJ, hip

Optimally, the pelvis should remain neutral, the lumbar spine should not hinge into extension at any segment, the femoral head should remain centered in the acetabulum and the innominate should not anteriorly rotate relative to the sacrum at the contralateral SIJ. Ask the patient to quantify the effort required to lift the extended leg from the prone position (0 = no problem, 5 = unable) and then to note the difference in effort when:

  1. The sacrum is nutated passively
  2. The pelvic girdle is compressed
    1. bilaterally across the anterior aspect
    2. bilaterally across the posterior aspect
    3. obliquely from left anterior to right posterior
    4. obliquely from right anterior to left posterior
  3. The femoral head position is manually corrected and controlled

A higher load can be added by providing isometric resistance to hip extension. Note the strength on the right and left sides and use the weaker side to continue the test. Repeat the test while passively nutating the sacrum and note any change in hip extension strength.

When an optimal strategy for these tasks are used the patient will not notice any difference in effort between the right and left legs to perform the prone knee bend / hip extension tests (+/- resistance) and no differences will be noted with any manual modifications (nutation of the sacrum, pelvic compressions, femoral head centering).

However, if areas of failed load transfer were noted initially (SIJ, lumbar spine or hip), then one or more of the manual modifications may make the effort experienced on testing decrease. If this occurs then a non-optimal strategy for stabilisation of the lumbar spine, pelvic girdle and / or hip can be considered a key contributor to the loss of hip extension power noted both subjectively by the athlete and objectively by the examiner. Further tests are required to determine why the strategy is non-optimal (specific system impairment); a full examination of the lumbopelvichip region is now indicated.

Based on this clinical test, Takasaki et al (2008) designed a study to determine if different amounts of compression across the pelvis affected timing of activity of the erector spinae, gluteus maximus and semitendinosus muscles during a prone hip extension task in asymptomatic subjects. Timing differences between the onset of gluteus maximus compared to semitendinosus occurred in the different pelvic compression conditions which provides some support to the idea that changing the forces across the pelvis can impact neural patterning.

It is important to note that there are likely to be many different reasons for a positive response to a certain pattern of compression across the pelvic girdle in these tests and although the different underlying mechanisms are yet to be studied, further clinical tests (examination of the systems) combined with a critical clinical reasoning process can provide solid hypotheses to direct treatment.

How do these tests aid in differential diagnosis?

The patient with a true structural deficit in the hamstring muscle will demonstrate minimal or no changes in perceived effort, pain or ability / muscle strength output when the prone knee bend / hip extension tests (+/- resistance) are performed with and without the modifications to augment control and stability of the lumbar spine, pelvis and hip joints. In certain patients poor lumbopelvic-hip control may have been a predisposing factor to the hamstring injury. It is our experience that even these patients do not demonstrate marked differences in effort or strength output when the affected hamstring is tested with and without lumbopelvic support if the main impairment is a structural deficit in the hamstring. A small improvement may be felt but the painful lesion will still test weak and painful, especially in an acute or sub-acute injury. If poor lumbopelvic control is present as the tissue deficit heals these tests will then reveal that it is time to include proximal control and treatment, i.e., the test modifications will make a change to the effort and / or strength output on the tests. It is our opinion that these findings indicate that the patient will not fully recover from the hamstring injury without treating impairments in the proximal systems.

Patients with referred posterior thigh pain will have significant and marked changes in effort, strength and pain responses to the modifications to the prone knee bend / hip extension tests (+/- resistance) applied to the lumbopelvic-hip region. Note that if any of the proximal compressions or joint control modifications make the effort or strength output worse, this is also supportive of a lumbopelvic “driver” and indicates that the system is under too much compression. Manual tests to relieve compression on the relevant components of the lumbopelvic complex (Lee 2004, Lee & Lee 2004) will then improve the responses to the tests. In either case, the significant change in symptoms, perceived effort and strength output with alterations to the lumbopelvic-hip complex indicate that the “hamstring pain” has significant proximal drivers and is most likely not a true structural deficit of the hamstring muscle.


Loss of motion control of the lumbopelvic-hip complex is a common finding in patients / athletes presenting with single or recurrent hamstring injuries and / or posterior thigh pain. A functional “platform” is essential for optimal performance of not only the lower extremity but the body in general. Therefore, whenever impairments that can be shown to relate to non-optimal strategies during meaningful tasks are found, whether painful or not, these impairments must be addressed. The system-based classification approach is an integrated, evidence-based model that considers both pain and disability, i.e., it relates impairments found in systems, pain and the impact of these impairments on the synergistic function required for optimal strategies for function and performance and ultimately, for health. It is a classification that applies to the whole person, rather than to a specific type of pain presentation. Thus, it can be used across pain and disease populations and is not only used for patients with lumbopelvic pain. This approach is evidence-based (Sacket et al 2000) in that it is a patient centered model which considers the best available research evidence in combination with clinical expertise the definition of which is having the ability “to do the right thing at the right time” and requires the ability to think critically and be critical about your thinking (reflection and metacognition) (Jones & Rivett 2004)

The prone hip extension task analysis with its multiple modifications / variations is an example of how clinical reasoning is used to determine when / if the pelvis is a contributor to the patient’s experience of either single or recurrent hamstring strains and / or posterior thigh pain. Exactly why the pelvis is failing to transfer load would require further analysis of the “pieces of the puzzle” – the systems – i.e. the physiological, genetic and psychological makeup of the person in the middle of the puzzle, articular system, myofascial system, neural system and visceral system – which is beyond the scope of this article.

The system-based classification approach forms the foundation for courses taught by Diane Lee and Linda-Joy (LJ) Lee. We hope this article stimulates your interest in this integrated, evidence-based approach and we look forward to meeting you at one of our courses.


Brukner P, Khan K 2007 Clinical sports medicine, 3 rd edn.McGraw-Hill, Australia

Heiderscheit BC, Hoerth DM, Chumanov ES, Swanson SC, Thelen BJ, Thelen DG 2005 Identifying the time of occurrence of a hamstring strain injury during treadmill running: a case study. Clin Biomech (Bristol, Avon). 20(10):1072-8. also E-pub Aug 31 2005

Hungerford B, Gilleard W, Hodges P 2003 Evidence of altered lumbopelvic muscle recruitment in the presence of sacroiliac joint pain. Spine 28(14):1593

Hungerford B, Gilleard W, Lee D 2004 Alteration of pelvic bone motion determined in subjects with posterior pelvic pain using skin markers. Clinical Biomechanics (19):456

Jones MA, Rivett D 2004 Introduction to clinical reasoning. Ch 1 In: Clinical reasoning for manual therapists. Eds Jones M A, Rivett D A, Elsevier, Edinburgh

Jull GA, Bogduk N, Marsland A 1988 The accuracy of manual diagnosis

Lee DG 2004 The Pelvic Girdle, 3rd edn. Elsevier, Edinburgh

Lee DG, Lee L J 2004 An integrated approach to the assessment and treatment of the lumbopelvic-hip region. Available online at

Lee DG, Lee LJ 2007 Bridging the Gap: the role of the pelvic floor in musculoskeletal and urogynecological function. Proceedings of the World Physical Therapy Conference, Vancouver, Canada

Lee DG, Lee LJ 2008(a) Integrated, multimodal approach to the treatment of pelvic girdle pain and dysfunction. Ch 14 In: Magee ed. Volume III Treatment of Pathology and Injuries WB Saunders p473-484

Lee LJ, Lee DG 2008(b) Integrated, multimodal approach to the thoracic spine and ribs. Ch 8 In: Magee ed. Volume III Treatment of Pathology and Injuries WB Saunders p306-337

Mason DL, Dickens VA, Vail A 2008 Rehabilitation for hamstring injuries (review) The Cochrane Collaboration Library, Issue 4

Orchard J, Seward H 2002 Epidemiology of injuries in the Australia football league, seasons 1997-2000. Br J. Sports Med. 36(1):39-44

Sackett DL, Straus S, Richardson W S, Rosenberg, Haynes RB 2000 Evidence-based medicine. How to practice & teach EBM. Elsevier Science, New York

Seward H, Orchard J, Hazard H, Collinson D 1993 Football injuries in Australia at the elite level. Med J Aust Sep 6; 159(5):298-301

Sherry MA, Best TM 2004 A comparison of 2 rehabilitation programs in the treatment of acute hamstring strains. J Orthop Sports Phys 34(3):116-25

Sturesson B, Uden A, Vleeming A 2000 A radiosteriometric analysis of movements of the sacroiliac joints during the standing hip flexion test. Spine 25(3):364

Takasaki H, Iizawa T, Hall T, Nakamura T, Kaneko S 2008 The influence of increasing sacroiliac joint force closure on the hip and lumbar spine extensor muscle firing pattern. Man Ther Dec 30 Epub ahead of print

Verrall GM, Slovotinek JP, Barnes PG, Fon GT, Spriggins AJ 2001 Clinical risk factors for hamstring muscle strain injury: a prospective study with correlation of injury by magnetic resonance imaging. Br J Sports Med 35:435-439

Warren P, Gabbe BJ, Schneider-Kolsky, Bennell KL 2008 Clinical predictors of time to return to competition and of recurrence following hamstring strain in elite Australian footballers. Br J Sports Med doi:10.1136/bjsm.2008.048181 published online

Woods C, Hawkins RD, Hulse M, Thomas A, Hodson A 2004 The football association medical research programme: an audit of injuries in professional football-analysis of hamstring injuries. Br J Sports Med 38:36-41

This article was originally published in In Touch, Summer 2009, No. 127

You may also be interested in:

The Pelvic Girdle: An integration of clinical expertise and research

The Pelvic Girdle continues to provide the busy clinician with the latest evidence and clinical tools/knowledge to immediately impact and enhance daily practice for the management of lumbopelvic-hip pain and disability. This fourth edition has changed fundamentally in presentation and content to provide the clinician with the evidence and clinical tools for effective practice.

The Thorax: An Integrated Approach

This clinical textbook will update the reader on the relevant anatomy, known biomechanics, clinical assessment, musculoskeletal conditions and treatment of the thorax and how these relate to the function of the whole body. An integrated biopsychosocial model (the Integrated Systems Model – ISM) will be highlighted in this text and used as a foundation for clinicians to organize their knowledge from multiple sources. 

Leave a Reply

Scroll to Top