Skip to content
3 Pilates exercises to avoid for hip pain_Blog Banners_3

As a physiotherapist with a special interest in the hip, I see many patients with hip pain who have done, are currently participating in, or are considering joining up to a Pilates class. I find myself regularly providing similar advice, so today I am going to share my recommendations on 3 Pilates exercises to avoid for hip pain and why.

Pilates Method classes are one of the most popular choices for group exercise, particularly in western cultures, with millions of people worldwide practicing and teaching these techniques. Joseph Pilates developed this form of exercise in the earlier part of the 19th century, aiming to optimise both physical and mental health. In recent times, ‘Clinical Pilates’ has incorporated more biomechanical principles and adapted the method for musculoskeletal rehabilitation purposes. Exercises may be performed on the floor – matwork classes – or with various pieces of equipment or ‘apparatus.’ While there is much to be gained from this form of exercise, there are also many exercises that appear to be performed ‘just because’ … with little to no consideration of what is ‘normal’ or efficient muscle function, or how exercises might impact underlying joints or various painful pathologies with which a person might present.

No. 1 Pilates Exercise to Avoid for Hip Pain: The Clam or Clamshell Exercise

The clam or clamshell exercise is perhaps the most widely used exercise for ‘gluteus medius strengthening,’ not only within Pilates classes but prescribed by health and exercise professionals. This exercise is often a staple of Pilates mat classes, in which participants may perform high volume repetitions. For those blissfully unfamiliar with this exercise, it involves lying on the side with various degrees of lower limb flexion and then lifting the top hip into abduction and external rotation, with the heels remaining together.

Thanks to Lisa Jackson from Core Pilates, Brisbane, for modelling this exercise

For anyone who has attended my courses or lectures or read my previous blogs, you might already know of my mission to ‘Ban the Clam!’ There are a number of reasons for my lack of affection for the clam, generally fitting into 2 categories:

  1. adverse effects on hip pathologies and
  2. suboptimal effects on hip abductor muscle function.

The first of these I discussed in a previous blog – you can read about this here. That blog discussed the reasons why you should not prescribe clams for gluteal tendinopathy/Greater Trochanteric Pain Syndrome (GTPS) and why you should not prescribe clams for hip joint-related pain or after hip arthroscopy.  In today’s blog, I will focus on the potential effects on the hip abductor synergy.

‘The clam exercise appeared least favourable in terms of recruiting GMed (gluteus medius) muscle activity.’

A recent systematic review of the literature on gluteus medius and minimus recruitment in common therapeutic exercises stated that ‘the clam exercise appeared least favourable in terms of recruiting GMed (gluteus medius) muscle activity.’1 Available studies have shown only low recruitment of anterior and middle segments of gluteus medius, moderate activity in the posterior gluteus medius and low activation of both segments of gluteus minimus1. It was concluded that the clam exercise is unlikely to elicit sufficient activity for strengthening, particularly in the gluteus minimus and anterior and middle segments of gluteus medius. The relatively poor activation levels were suggested to be associated with the short lever arm these muscles have to work with in this exercise. Such a situation may in some circumstances lead to high excitation but poor force production, due to the inefficient lever arm. However, if there are synergists with a more advantageous lever arm, the work may be redistributed.

In the clam action, the gluteus maximus provides an external rotation torque, with the upper portion of the muscle also being a primary abductor. The upper gluteus maximus, exerting its effect via its insertion into the iliotibial band, has a relatively longer lever arm and is likely to be producing more force than the gluteus medius in this exercise. Selkowitz and colleagues found that, although Tensor Fascia Lata (TFL) activation levels are relatively low, upper gluteus maximus was almost twice as active as the gluteus medius in the clam exercise.2 I have heard many clinicians defend the clam by noting the strong contraction of the gluteus medius they feel on palpation of the mid-buttock region. It’s important to remember that the gluteus maximus overlies the gluteus medius and when the muscles are active in the clam action, it is not possible to palpate the underlying gluteus medius or judge its level of activation in this manner. What will be easily palpable, is the strong action of the upper gluteus maximus muscle, a superficial abductor exerting its effect via the ITB, just like TFL.

The other muscle that has a long lever arm for the clam action is the Sartorius muscle, whose action is hip flexion, abduction and external rotation – a perfect fit for the clam! I have noted that patients who have done high volumes of clams and open chain hip abduction exercises, will commonly present with disproportionate hypertrophy in the anterolateral corner of the hip. This appears to be a combination of Sartorius hypertrophy (likely associated with the clam) and TFL hypertrophy (likely associated with the open chain abduction – more on this in the next exercise).

Clams don't target the region of greatest atrophy

The stranglehold that the clam has on therapeutic exercise prescription for the hip appears to be linked with an obdurate clinical belief that atrophy or dysfunction of the posterior portion of gluteus medius is the primary issue associated with abductor dysfunction. With the scientific evidence mounting, it would appear that if we were to target any particular region of the abductors, the anterior aspects of gluteus minimus and medius might be due greater attention than the posterior aspects.

A recent case-control study demonstrated higher levels of fatty infiltration in the anterior segment of the gluteus minimus and the anterior and middle segments of the gluteus medius (but not the posterior segments) in those with hip OA, compared with control participants. 3 Fatty infiltration of the anterior gluteus minimus is common with ageing 3,4 but more common in those with hip OA3 and GTPS.4 The evidence would suggest that its time to re-examine our monogamous love affair with posterior gluteus medius and share the love a little more anteriorly. Again, I suggest that the clam should NOT be the go-to exercise for abductor training and hip pain. Time to move on!

No. 2 Pilates Exercise to Avoid for Hip Pain: Sidelying Hip Abduction

While the sidelying hip abduction exercise produces substantially more activity in the gluteus medius and minimus, TFL activation is also much higher2 and this exercise is so commonly performed poorly in group classes or as part of a home exercise program.

Thanks to Lisa Jackson from Core Pilates, Brisbane, for modelling this exercise

In addition, we need to remember that the available EMG evidence for therapeutic exercises for the hip abductor synergy has overwhelmingly been performed in young healthy adults without pain. As we just discussed, those with hip pain tend to present with fatty atrophy of the gluteals,3,4 while TFL is not affected in GTPS4 or hip OA5, even in advanced stages of disease.5 These means that the TFL will occupy a relatively larger proportion of the overall abductor volume, and reasonably then contribute to a larger degree to abductor force production.

If we hope to target the deeper abductors, exercise selection is key. My PhD work showed that young healthy males on bed rest experienced a rapid loss of gluteus minimus bulk, on average 23% loss in 2 weeks, while the TFL showed no significant loss in size even over 8 weeks of bed rest.6 Different members of the abductor synergy are very preferentially affected by weightbearing stimulus, even in healthy individuals. It is not surprising then that most studies report that weightbearing exercises are superior for stimulus of the trochanteric abductors (gluteus medius and minimus).1 This is likely to be even more important for those with hip pain who have atrophy in these deeper abductors.

If prescribing this exercise, pay attention to technique and relative muscle activation strategies, but getting the foot to the floor in a weightbearing exercise generally makes gluteal activation a whole lot easier!

In addition, if this exercise is performed with a return of the top ankle to touch the bottom ankle on each repetition, conditions that don't enjoy repetitive hip adduction are unlikely to respond well, e.g. gluteal tendinopathy/GTPS, ischiofemoral impingement and some joint conditions.

No. 3 Pilates Exercise to Avoid for Hip Pain: Legs in Straps Circumduction

The third exercise I try to steer Pilates instructors away from, is long lever circumduction with the feet in straps. This exercise is usually most provocative for people with proximal hamstring tendinopathy or those with hip joint related pain.

Thanks to Lisa Jackson from Core Pilates, Brisbane, for modelling this exercise

Legs in straps with proximal hamstring tendinopathy

If we consider that compressive loads are particularly provocative for those with insertional tendinopathy,7 the position that will create greatest compression for the proximal hamstring tendons is one that combines high ranges of hip flexion with knee extension (a position where the hamstrings are on stretch). In this position, the hamstring tendons wrap firmly around the ischium, compressing the deepest parts of the tendon against the bone. The compression will be even greater if the hamstring muscles are active. In this Pilates exercise, the tendon is under high compressive load both due to the position and the high level of hamstring muscle activation. Now add some ‘circles’ (circumduction) and this will serve to add friction to the mix, rubbing the compressed tendon medially and laterally across the ischium. My advice -  avoid this exercise like the plague (or Covid-19!) for anyone with proximal hamstring tendinopathy and consider whether there is something else you could substitute for those who might be in a higher risk category for developing this condition.

Learn more about proximal hamstring tendinopathy in my ebook, understanding tendinopathies course, or lateral hip & buttock pain course

Legs in straps with hip joint related pain or impaired stability mechanisms

Legs in straps, particularly long lever exercises (with knees straight) are also very challenging for those with hip joint related pain and/or those with reduced passive stability of the hip joint – acetabular dysplasia, labral injury or capsular laxity. Performing hip actions with straight knees strongly engages the biarticular thigh musculature, creating large translational and rotational forces across the hip joint. If there is inadequate control of these forces by local passive and dynamic mechanisms (deep uniarticular muscles), excessive shear forces and abnormal femoro-acetabular contact may occur. These forces are unfriendly to cartilage, the labrum and the capsule. The multiplanar circles require high levels of local control around the hip joint. In addition, for those with early bony impingement related to cam or pincer morphology (FAI) or OA related osteophytosis, circumduction actions may serve to rub the femoral head neck junction around the anterior and anterolateral acetabular rim. This is commonly provocative for those with hip joint related pain but may also be potentially injurious for anyone with either inadequate passive and/or dynamic stability mechanisms, or FAI morphology.

Learn more about hip joint related pain, dysplasia and FAI here

There are so many excellent Pilates exercises that provide much greater benefit with much lower risks. Anyone prescribing exercises for others has a duty-of-care to do no harm and to consider the relative risk-reward for the individual. Selecting exercises that optimise musculotendinous health and function while minimising adverse loads across the joint, the soft tissues and any pathologies, is a win-win for everyone!

Happy (thoughtful) programming!

Subscribe to our Newsletter

Sign up to our newsletter to receive updates on upcoming courses, news and special offers.


  1. Moore, D., Semciw, A. and Pizzari, T., 2020. A systematic review and meta-analysis of common therapeutic exercises that generate highest muscle activity in the gluteus medius and gluteus minimus segments. International Journal of Sports Physical Therapy, 15(6), pp.856-881.
  2. Selkowitz, D., Beneck, G. and Powers, C., 2013. Which Exercises Target the Gluteal Muscles While Minimizing Activation of the Tensor Fascia Lata? Electromyographic Assessment Using Fine-Wire Electrodes. Journal of Orthopaedic & Sports Physical Therapy, 43(2), pp.54-64.
  3. Kivle, K., Lindland, E., Mjaaland, K., Svenningsen, S. and Nordsletten, L., 2021. Gluteal atrophy and fatty infiltration in end-stage osteoarthritis of the hip. Bone & Joint Open, 2(1), pp.40-47.
  4. Cowan, R.M., Semciw, A.I., Pizzari, T., et al., 2019. Muscle Size and Quality of the Gluteal Muscles and Tensor Fasciae Latae in Women with Greater Trochanteric Pain Syndrome. Clinical Anatomy, 33(7), pp.1082-1090.
  5. Zacharias, A., Green, R., Semciw, A., English, D., Kapakoulakis, T. and Pizzari, T., 2018. Atrophy of hip abductor muscles is related to clinical severity in a hip osteoarthritis population. Clinical Anatomy, 31(4), pp.507-513.
  6. Grimaldi A, 2008. MRI Investigations of the Muscles involved in Lateral Stability of the Hip, PhD thesis, University of Queensland, Brisbane.
  7. Cook, J.L., Purdam, C., 2012. Is compressive load a factor in the development of tendinopathy? British Journal of Sports Medicine, 46(3), pp.163-168.