The Gluteus Medius

Chris Mallac looks at the gluteus medius – and the most common dysfunctions and how they’re best rehabilitated…

Center fielder Kevin Kiermaier runs home, 2018  Credit: Kim Klement-USA TODAY Sports


The Gluteus Medius (GMed) is a muscle that has received a significant amount of interest amongst the correctional exercise fraternity and physical therapy world. It is a muscle that is often implicated in playing a crucial role in stabilising the pelvis during stance phase of gait and controlling the sagittal, frontal and coronal planes of movement of the lower limb during stance phase. Dysfunction in the GMed has often been associated with a range of musculoskeletal pain syndromes including knee, back and hip problems.

Table 1
Absorption movementControlling muscles
Pelvic lateral tilt (Trendelenburg Sign)Hip abductors (TFL, GMed, Glute Min, Glute Max – superior fibres)
Anterior pelvic tiltPelvic posterior tilters such as gluteals and hamstrings.
Hip joint flexion, adduction, internal rotationControlled by gluteus medius and other hip joint external rotators such as the gemellus muscles, quadrutus femoris, obturator muscles and piriformis.
Knee joint flexionQuadriceps
Ankle dorsiflexionSoleus
Midfoot pronationTibialis Posterior, FHL, FDL

Relevant anatomy and biomechanics

During single-limb weight bearing movements such as stance phase of walking/running, lunging, landing from a jump amongst others, the natural tendency for the lower limb joints is to absorb the impact of gravity on the body. The force of gravity will generate joint moments into certain directions and muscles are required to work to counteract these forces (usually these muscles work isometrically and/or eccentrically). Table 1 describes the ‘absorbing’ joint movments and which muscles control these joint moments.

The GMed is a proximal hip muscle that is purported as one of the muscles that controls proximal pelvic/hip joint motion that in turn controls lower limb kinetics around the knee and ankle. Most anatomy textbooks will describe the GMed as attaching to the iliac crest and inserting onto the greater trochanter. Its function is often described as being a hip abductor, hip external rotator and stabilising the pelvis on the femur during stance phase of gait. However, as Gottschalk et al (1989) suggest, its most significant role may in fact be to compress the femoral head into the acetabulum during the stance phase of gait1. The muscle is divided into three equal components: anterior, middle and posterior.

The fibres of the posterior portion run almost parallel with the neck of the femur, while the middle and anterior parts run vertically from the iliac crest to the anterosuperior aspect of the greater trochanter. Each of the three parts of GMed has its own nerve supply running from the superior gluteal nerve, suggesting that the muscle actions of the three heads are independent of each other.

Gottschalk et al (1989)2 also conducted EMG studies and they found that the GMed is not all that active in isolated abduction of the hip. This finding may well surprise most readers, as it is contrary to what has been taught for years in anatomy and biomechanics lectures and textbooks. They observed that the tensor fascia lata (TFL) is significantly more active in isolated hip abduction. They went on to suggest that the three heads of the GMed have a phasic muscle action during stance phase of gait. The posterior directed fibres are more active at heel strike, and then the muscle becomes progressively recruited from posterior to anterior as movement occurs from early stance to late stance of gait. In other words, the front portion of the muscle (which is anatomically similar to the TFL) is most active at full stance and single-leg support phase, while the rear fibres fire strongly at initial heel strike.

Gottschalk et al suggested that the main role of the GMed is to compress the head of the femur into the acetabulum (hip socket) during locomotion and to assist in stabilising the pelvis on the femur in single leg stance. They then put forward the notion that each of the three distinct heads of the muscle performs a unique role in locomotion:

The posterior fibres contract at early stance phase to lock the ball into the socket. This idea is supported by the observation that the posterior fibres have an almost parallel fibre alignment with the neck of the femur. The posterior fibres therefore essentially perform a stabilising or compressing function for the hip joint.

The middle/anterior fibres, which run in a vertical direction, help to initiate hip abduction, which is then completed by the TFL. These fibres work synergistically with TFL in stabilising the pelvis on the femur, to prevent the other side dropping (or Trendelenburg). The researchers point out that the TFL has the more important role in stabilising the pelvis on the supporting hip; the GMed simply assists this action, analogous to how the supraspinatus in the shoulder assists the more powerful deltoid in shoulder abduction. The anterior fibres allow the femur to internally rotate in relation to the hip joint at mid-to-end stance phase. This is essential for pelvic rotation, so that the opposite side leg can swing forward during gait. The anterior fibres perform this role with TFL. So Gottschalk et al postulated that the primary functions of the GMed are:

  1. To stabilise the femur on the ilium (pelvic stability)
  2. To act as hip rotators
  3. To approximate the head of the femur into the acetabulum, in effect creating a very tight and stable hip joint during gait.

Other researchers have found on cadaver studies that GMed has a large physiological cross-sectional area and short fibre lengths and therefore is able to generate large forces over a narrow range of lengths3(16). They postulate the GMed really only works in neutral hip/pelvic postures as it would when stabilising the pelvis and hip during single leg stance. Exercises that force the GMed into lengthened or overly shortened positions may in fact not target the GMed but other hip abductors/external rotators. GMed has the largest CSA of the hip abductors so it should be the more dominant of the hip abductors. It can produce a lot of force for its size as it has short fibres which are packed tightly together. But is does not produce large forces over a wide range of lengths. It is designed to work isometrically and stabilise the hip on the femur and vice versa.

Table 2: Rehabilitaion exercises for the Gluteus Medius
Distefano et al (2009)(7)Bolga and Uhl (2005)(4)Ayotte et al (2007)(2)Boren et al (2018)(5)
Sideways hip abduction 81%Pelvic drop 57%Wall squats 52%Side plank abduction with
dominant leg on bottom 103%
Single limb squat 64%WB with flexion left hip abduction 46%Front step ups 44%Side plank abduction with
dominant leg on top 89%
Lateral band walk 61%WB left hip abduction 42%Lateral step-ups 38%Single limb squat 82%
Single-limb deadlift 58%NWB side lie hip abduction 42%Retro step-ups 37%Clamshell (hip clam) 77%
Sideways hop 57%NWB standing hip abduction 33%Mini squat 36%Front plank with hip extension 75%

Injuries to the Gluteus Medius

Dysfunction in the GMed has often been associated with a range of musculoskeletal pain syndromes. It is believed that these injuries are a result of the inability of the GMed to control movement and alignment at the pelvis, femur and tibia. These injuries include but are not limited to:

  1. Patellofemoral pain syndromes4
  2. Lumbar spine problems5
  3. ITB friction syndromes6
  4. Hip joint pathology7.

It has been assumed for some time that hip internal rotation is an unwanted pathomechanic of the hip joint as hip joint rotation will allow the femur to migrate inwards and create valgus collapse at the knee. The assumption has then been further extrapolated to suggest that this unwanted hip internal rotation is a consequence of weak GMed and other hip joint external rotators. However, the work by Ward et al (2011)8 suggest that in fact the GMed seems t o work better physiologically if the hip is placed in some internal rotation. Therefore is this hip internal rotation a compensation to allow the GMed to be recruited better in the presence of weakness of other hip muscles such as the gluteus maximus and other deep hip rotators?

Direct injuries to the GMed such as trigger points, strain injuries, tendon tears and associated trochanteric bursitis have also been attributed to having weak GMeds91011.

Rehabilitation Exercises for the Gluteus Medius

A wide range of studies have examined the function of the GMed whilst performing a variety of lower limb exercises12 13 14 15 1617. These studies base their conclusions on relative electromyographic (EMG) data during certain exercises. The top five exercises in each study are presented in Table 2 and readers can be directed to the reference l i s t for clarification on the exact exercises used in the respective studies. The % displayed alongside the exercise is the % of Maximal Voluntary Contraction (MVC) of the GMed. It needs to be pointed out that differences in activity between authors may be different due to a number of factors such as:

  1. EMG placement
  2. Electrical interference from other muscles
  3. Differences in the exact mechanics of the chosen exercises.

Also the EMG data does not necessarily tell us if the GMed is creating the action or simply stabilising the hip joint and pelvis whilst other muscles are working, similar to how the rotator cuff muscles work during active shoulder abduction and flexion movements.

In a more recent study, researchers looked at the relative contribution between the GMedand the TFL and identified these five exercises that best used GMed with minimal TFL18:

  1. Clam with theraband
  2. Sidestep with theraband
  3. Unilateral bridge
  4. Quadruped hip extension, knee extending
  5. Quadruped hip extension, knee flexing.

With so many variations in the possible beneficial exercises that may be used to strengthen the GMed, often the choice of exercise used by the therapist may simply be a ‘horses for courses’ approach. If the individual feels pain on weight bearing movements then non-weight bearing variations may be used. Often the client simply may not ‘feel’ the movement required therefore other exercises need to be chosen. Often it may simply be the therapist’s personal choice as to what they perceive to be the most effective GMed exercise. Furthermore, it could be argued that what a client feels in and around their postero-lateral hip may be GMed and/or other hip abductors such as gluteus minimus and/or other deep hip rotators such as piriformis, the obturator group, quadrutus femoris and gemellus muscles. Studies are needed using both surface EMG and fine wire EMG on deep muscles to truly elucidate the contribution of the complex interactions between these muscles.

The three exercises shown below are variations on other exercises listed in Table 2. The reason they have been included is to satisfy the work by Gottschalk et al19 that shows that GMed works in varying ways during hip flexion to extension as demonstrated in the gait cycle, and also the work by Ward et al20 that suggests the muscle works through very neutral hip/pelvic positions and works essentially isometrically or through very short ranges of movement. Furthermore, the three exercises attempt to either directly weight bear through the hip joint or simulate weight bearing through the hip joint, making them more functional in terms of activation in weight bearing positions.

1. Standing short range hip abduction.

This exercise works both the stance limb (isometric) and non-stance limb (concentrically).

a. Stand with a band around the foot and the hand on the same side supported by a broomstick for balance.

b. Gently move the banded leg into abduction/external rotation/extension.

c. The stance limb is in slight hip flexion and remains in this position.

d. Perform 8-10 repetitions of slow hip abduction/external rotation/extension.

e. This will be felt in both the stance side GMed (in slight hip flexion) as well as the abducting side GMed (into slight hip extension).

Standing short range hip abduction (start)

Standing short range hip abduction (start)

Standing short range hip abduction (finish)

Standing short range hip abduction (finish)

2. Kneeling clam.

This is one variation on the ever popular clam exercise that has been shown in numerous studies to activate the GMed muscle. Again this is performed in weight bearing as the limb accepts axial load via kneeling.

a. Kneel on a bench with a band wrapped around the knees. Feet are kept together.

b. Hold onto a broomstick for balance.

c. Gently move the knees apart whilst maintaining the foot contact. This moves the hip into slight abduction/external rotation.

d. Perform sets of 10-15 repetitions and ensure the movement is kept small (2-3 inches only).

Kneeling clam – start. Notice feet together and knees together

Kneeling clam – start. Notice feet together and knees together

Kneeling clam – finish. Notice feet together and knees apart

Kneeling clam – finish. Notice feet together and knees apart

3. Modified clam.

This is another variation on the clam exercise that resembles the traditional clam except with some variations. The first important difference is that the heels push into a wall or box to simulate weight bearing through the limb. Next the exercise is performed as an isometric hold and not an active abduction/adduction movement. Finally the exercise is performed in two positions: 1. Slight hip flexion and 2. Slight hip extension. A light weight is placed on the knee to act as external resistance. The aim is to hold the limb static for a prescribed period of time.

Modified clam – hip in slight flexion

Modified clam – hip in slight flexion

Modified clam – hip in slight extension. Note the contralateral hip is flexed to assist the pelvis to maintain a neutral tilt. If the knees are kept together, this encourages unwanted anterior pelvic tilt

Modified clam – hip in slight extension. Note the contralateral hip is flexed to assist the pelvis to maintain a neutral tilt. If the knees are kept together, this encourages unwanted anterior pelvic tilt


The GMed is a muscle that has received a lot of research and attention in the last few decades. Research into the GMed using EMG, biomechanical modelling and cadaveric studies have concluded that it is an important muscle that needs to be strengthened to assist in pelvic control, hip joint stability and lower limb kinetic control. This article presents the relevant and up-to-date anatomy and biomechanical knowledge on the muscle, the existing research on activation in exercise and also the author’s suggestions on some new exercises that can work the GMed muscle.

  1. Journal of Anatomy. 1989. 166: 179-189.
  2. Journal of Anatomy. 1989. 166: 179-189.
  3. J Orthop Sports Phys Ther. 2009. 39:12-19.
  4. Arch Phys Med Rehabil; 1998. 79:412-417.
  5. J Orthop Sports Phys Ther; 2010. 40:52-58.
  6. Man Ther. 2009;14:611-617.
  7. Journal of Orthopaedic and Sports Physical Therapy. 2011. 40(2); 95-102.
  8. The Iowa Orthopedic journal. 2003. 23; pp57-
  9. AJR. 173(4); 1123-1126.
  10. Eur Radiol. 2006. 17(7); pp 1172-83.
  11. Arch Phys Med Rehabil; 1999. 80:842-850.
  12. J Orthop Sports Phys Ther; 2007. 37: 48-55.
  13. J Orthop Sports Phyl Ther; 2005. 35: 488-494.
  14. International Journal of Sports Physical Therapy. 2011. 6(3). 206-223.
  15. J Sport Rehabil; 2009. 18:91-103.
  16. J Orthop Sports Phys Ther; 2009. 39: 532-540.
  17. Journal of Orthopaedic and Sports Physical
    Therapy. 2013. 43(2); 54-65.
  18. Journal of Anatomy. 1989. 166: 179-189.
  19. Man Ther. 2009;14:611-617.
Share this
Follow us