Hip Anatomy and Squat Form

Hip Anatomy and Squat Foot Placement

by Dr. Brookbush DPT, PT, MS, CPT, HMS, IMT

This article and systematic research review is an excerpt from the course:

Lower Body Exercise Progressions

Squat Depth Recommendations: Based on All Available Research

Introduction

Modifying foot position or stance width during a barbell squat to accommodate "unique hip anatomy" is a questionable approach for several reasons. For example, the majority of individuals have similar proportional anatomy. Further, it is a logical leap to imply changes in hip joint structure should alter foot position, bypassing the knee and ankle joint completely. Additionally, why not first address other potential contributing factors, such as mobility limitations (e.g., restricted knee flexion, hip rotation, or ankle dorsiflexion), altered muscle activity (e.g., gluteus medius or tibialis anterior activation), or weaknesses in the primary muscles involved (e.g., gluteus maximus, quadriceps, or soleus)? Note that this reasoning applies to a wide range of lower-extremity strength exercises, including variations of the back squat (e.g., high bar and low bar), leg press, hack squat, Smith machine squat, as well as lunges, split squats, and step-ups.

Brookbush Institute's Position Statement: Research demonstrates that the number of individuals exhibiting significant hip structure differences is relatively small (less than 10%). Further, for the few individuals who exhibit significantly different hip morphology, there is no evidence to suggest that changing foot position during a squat will increase performance or decrease the risk of injury. Research also suggests that changes in mobility and muscle activity are correlated with an inability to maintain optimal alignment or perform a full range of motion during lower extremity exercises with toes forward and feet hip-to-shoulder-width apart. Discomfort when performing lower extremity exercises should not be considered evidence of a structural issue and instead should be considered an opportunity to address mobility or muscle activity issues with manual therapy and/or corrective exercise techniques, with the potential to significantly improve performance.

Summary of the Issues with Correlating Hip Anatomy and Foot Position

Normally Distributed: Research suggests that variations in hip anatomy are normally distributed and could be plotted on a "bell curve.” That is, the gross majority of individuals exhibit bone shape, angles, and alignment that are within a relatively small range of variation.
Little to No Correlation: There may be no correlation between hip morphology and foot placement. In fact, it is likely that other structural angles compensate for normal variations in hip morphology during development.
A Logical Error: Excessive hip retroversion and excessive hip anteversion cannot both be addressed by the same recommendation (feet wider and/or feet turned out). Further, if the standard deviation in hip anteversion is about 10°, how does anyone justify 20 - 50° of feet turn out during squats?
A Functional Anatomy Error: Feet turn out is not hip external rotation; it is tibia (knee) external rotation.
Correlated with Pain, Dysfunction, and Injury: Knee pain is correlated with less hip and knee internal rotation ROM, more knee varus, and more tibia and foot external rotation during a variety of activities. This implies that the recommendation to turn the feet out and/or "drive" the knees into varus (knees bow out) may contribute to an increased risk of pain and dysfunction.
A Better Solution: Before accepting small imperfections in movement as evidence of permanent structural differences, it is recommended that these issues (e.g., feet turn out, knees bow in, etc.) are treated as soft-tissue issues and targeted with an "integrated manual therapy," "corrective exercise," or "movement prep" routine.

Corrective Exercise Routine or Movement Prep Warm-up

Overhead Squat Assessment : Feet Turn Out

Release:
Mobilize
- Ankle Mobilization
Lengthen
- Hip Flexor Stretch
- Static Calve Stretch
Activate
- Quick Glute Activation Circuit (1-2 sets, 12-20 reps, 4:2:maxV tempo)
- Tibialis Anterior Activation (1-2 sets, 12-20 reps, 4:2:maxV tempo)
Core Integration
- Gali-peds Quadruped Progression (1-2 sets, 12-20 reps, 4:2:maxV tempo)
Subsystem Integration
- Single-leg Touchdowns with Posterior Pull (1-2 sets, 12-20 reps, 4:2:maxV tempo)

Summary of Research Review: Anteversion and Retroversion Angle of the Hip Joint

Evidence-Based Summary Statement: The average adult version angle is 7° - 19°, with a standard deviation (SD) ± 9.66°, a range of -4° - 36°, and is normally distributed. Note that individual research studies generally report similar SD and ranges, which likely implies there are some differences in measurement methodology from study to study. This may have resulted in a relatively wide range of averages. Toogood et al. (2009) demonstrated that as the neck version angle or inclination angle increased, translation in the same direction at the head-neck junction tended to decrease, demonstrating some compensatory adaptation. Further, some, but not all, dysfunctions may be correlated with a higher prevalence of extreme version angles. Last, research has demonstrated that knee pain is correlated with significantly less hip and knee internal rotation and more knee varus, tibia external rotation, and foot external rotation during a variety of activities. This implies that the recommendation to turn the feet out and/or "drive" the knees into varus (knees bow out) may contribute to an increased risk of pain and dysfunction.

Research Summary:

Cadaver Studies: Cadaver femurs exhibit an average version angle between 7° - 19° with a range of -4° - 36°. Some of the variations can be correlated with age, gender, and/or ethnicity. Further, the femoral head is slightly anterior, inferior, and abducted on the femoral neck, and the femoral head-neck junctions may have more concavity at the back and bottom than at the front and top. Note that Toogood et al. (2009) demonstrated that as the neck version angle or inclination angle increased, translation in the same direction (anterior for the neck version angle, superior for the angle of inclination) at the head-neck junction tended to decrease, demonstrating some compensatory adaptation.
Healthy Hips in Living Subjects: Again, these studies demonstrate an average anteversion angle for an adult of between 18.2° and 18.64° with a range of 10°–33°, and adolescents have an average anteversion angle of 4.5° - 4.8° with an SD of ±10.0°.
Pathological Hips in Living Subjects: These studies demonstrate an average anteversion angle of 8°-9° (range -5° - 31°, SD ± 9.66°). Further, some dysfunctions may not be correlated with a higher prevalence of extreme version angles, for example, patients receiving hip resurfacing and patients who have fractured their femur. However, we do see higher rates of extreme version (14%) for individuals receiving a hip replacement for osteoarthritis, and a higher prevalence of retroversion is correlated with developmental dysplasia and Legg-Calvé-Perthes disease.
Recommendations Correlated with Dysfunction: Knee pain is correlated with significantly less hip and knee internal rotation and more knee varus, tibia external rotation, and foot external rotation during a variety of activities. This implies that the recommendation to turn the feet out and/or "drive" the knees into varus (knees bow out) may contribute to an increased risk of pain and dysfunction.

Caption: Range of version angle of femur (not distribution).

Your hip anatomy is probably NOT unique:

This image above, which is commonly used by "gurus" to support a hyperbolic argument that we are all unique, is an example of the "range" of the femoral version angle, not the distribution. The range represents the extremes of a variable. However, the range may have very little to do with the distribution. The distribution is the number of individuals who exhibit any portion of a given range.

Citing extreme examples as if they are common and/or representative of the average (mean) is a logical error. It is the equivalent of selecting a 3'6" individual with achondroplasia and a 7'1" basketball player and presenting them as common variations in the heights of men. Although these heights exist, we cannot say they are representative of the average or of the majority of individuals. A little digging and you will soon find that nearly 70% of men are between 5'7" and 6'1", 96% are between 5'4" and 6'4", and the heights in our first example likely represent less than 1% of 1%.

Height, like many measures of human anatomy, can be plotted on a "bell curve" (a.k.a. normal distribution). Hip morphology (a.k.a. angles or version) appears to be another measure that can be plotted on a bell curve. Although extremes exist, we find that the gross majority of individuals fall within a relatively narrow range. Based on the available research, the average adult version angle is 7° - 19°, with a standard deviation (SD) ± 9.66° and a range of -4° to 36°, and is normally distributed. Note that individual research studies generally report similar SD and ranges, which likely implies there are some differences in measurement methodology from study to study that resulted in the relatively wide range of averages.

Caption: Bell curve distribution of femoral version angle from Pierrepont (2020).

This bar graph from a study by Pierrepont et al. (9) illustrates an increase in the prevalence of extreme anteversion among patients requiring total hip replacement (this would be a reasonable expectation). Note that even with this increase, the values for anteversion are still normally distributed.

There may be no correlation between hip morphology and foot placement.

There is no research correlating hip anatomy and ideal foot placement during a squat. To be more specific, there is no research to imply that the few individuals who exhibit relatively large deviations from average hip anatomy would benefit from altered foot placement. One issue that significantly decreases the relationship between foot placement and hip position is the number of structural angles between the hip and foot. Further, there is evidence of changes in anatomy to compensate for altered hip version angle. Toogood et al. (2009) demonstrated that as the neck version angle or angle of inclination increased, translation in the same direction (anterior for the neck version angle, superior for the angle of inclination) at the head-neck junction tended to decrease, demonstrating compensatory adaptation. Additional angles that may contribute to maintaining ideal alignment are included below. Note that a 5° increase in structural anteversion could be compensated for with a 1-2° change in the transverse plane component of the following structural angles (which would be within normal variation):

Femoral head translation (as mentioned above by Toogood et al. (2009).
Acetabular version angle
Tibial torsion
The angle of the axis of the ankle
- For more on these structural angles, you may want to check out "Hip (Pelvifemoral) Joint "

A practical example of how foot placement and hip morphology are not correlated can be observed by comparing an individual's squat foot placement and lunge foot placement. If there was a correlation between hip anatomy and foot placement, and a person exhibited a significant difference in hip anatomy, then the same stance would need to be adopted during squats and lunges (both exercises require similar hip ROM). However, most individuals can perform a full ROM lunge with feet, knees and hips in alignment in the sagittal plane, even when their squat form is wide with feet turned out. Note that lunges are generally easier to perform with neutral alignment because the support from the rear foot allows for an individual to move their center of mass posteriorly, which reduces the need for optimal dorsiflexion.

Caption: An illustration of the range of variation in pelvic structure. This is an example of the extremes, not normal distribution.

Again, this image of two pelves is representative of "range," not "normal distribution." This is cherry-picking the extremes, not accurately depicting normal variation (within 1 standard deviation of the mean).

If everyone is so unique, than why does everyone get the same recommendation?

If the argument supporting altered foot position during a squat is, "Everybody is unique, so ideal foot placement does not exist," then why does everybody get the same recommendation? Why is the recommendation always to turn the feet out and/or take a wider stance? It is unlikely that excessive hip anteversion and excessive hip retroversion could both be addressed by the same recommendation. Would it not be logical to assume that one of these structural differences (e.g., retroversion) would result in "turning the feet in" and/or standing with the "feet closer together," resulting in greater comfort during squatting comfortably?

Further, how does any professional justify cuing 20 - 50° of feet turn out when variations in hip version angle are unlikely to exceed 10°? If someone actually had 45° of additional anteversion, they would likely exhibit significant disability, and surgical intervention would likely be recommended. The difference between 10° and 45° is significant. For a visual example, when the minute hand on a clock has traveled from the top position (12:00) to the 2-minute hash, it has traveled 12°. When it has traveled from the top position (12:00) to the 8-minute hash, it has traveled 48°. That is a difference of 4 times.

A Functional Anatomy Error

The following error makes some otherwise smart professionals look pretty dumb. Feet turn out is NOT hip external rotation; it is tibia (knee) external rotation. The assertion that turning the feet out will alter hip position is actually a functional anatomy error. Feet turn out, when the knees are flexed, is achieved by external rotation of the tibia and not the external rotation of the hips. Hip external rotation at the bottom of a squat would NOT cause your feet to turn out. If your hips externally rotate a small amount at the bottom of a squat, your femurs rotate outward, forcing your knees to bow out, resulting in a "functional knee varus (knees bow out) ." Significant external rotation of the hips would result in your ankles crossing and you sitting in the yoga position that is known as the "lotus position." If the hips internally rotate a small amount during a squat, your femurs turn inward, resulting in a "functional knee valgus (knees bow in) ." And, a significant amount of internal rotation would cause your ankles to splay outward, resulting in a sitting position commonly observed in children, referred to as "W sitting." In summary, "feet turn out" during squats is the result of the rotation of the tibia at the knee, not the rotation of the hips. For a visual demonstration of hip internal and external rotation during hip flexion (as exhibited at the bottom of the squat) consider this image of seated hip internal rotation and hip external rotation goniometry. In the author's humble opinion, this one simple fact should have stopped this myth from ever developing.

Caption: Seated hip external rotation and hip internal rotation.

Caption: Seated hip internal and external rotation

Research Review of Version Angle of the Hip

Evidence-Based Summary Statement: The average adult version angle is 7° - 19°, with a standard deviation (SD) ± 9.66°, a range of -4° - 36°, and is normally distributed. Note that individual research studies generally report similar SD and ranges, which likely implies there are some differences in measurement methodology from study to study. This may have resulted in a relatively wide range of averages. Toogood et al. (3) demonstrated that as the neck version angle or inclination angle increased, translation in the same direction at the head-neck junction tended to decrease, demonstrating some compensatory adaptation. Further, some, but not all, dysfunctions may be correlated with a higher prevalence of extreme version angles. Last, research has demonstrated that knee pain is correlated with significantly less hip and knee internal rotation and more knee varus, tibia external rotation, and foot external rotation during a variety of activities. This implies that the recommendation to turn the feet out and/or "drive" the knees into varus (knees bow out) may contribute to an increased risk of pain and dysfunction.

Research Summary:

Cadaver Studies: Cadaver femurs exhibit an average version angle between 7° - 19° with a range of -4° - 36°. Some of the variations can be correlated with age, gender, and/or ethnicity. Further, the femoral head is slightly anterior, inferior, and abducted on the femoral neck, and the femoral head-neck junctions may have more concavity at the back and bottom than at the front and top. Note that Togood et al. (2009) demonstrated that as the neck version angle or inclination angle increased, translation in the same direction (anterior for the neck version angle, superior for the angle of inclination) at the head-neck junction tended to decrease, demonstrating some compensatory adaptation.
Healthy Hips in Living Subjects: Again, these studies demonstrate an average anteversion angle for an adult of between 18.2° and 18.64° with a range of 10°–33°, and adolescents have an average anteversion angle of 4.5° - 4.8° with an SD of ±10.0°.
Pathological Hips in Living Subjects: These studies again demonstrate an average anteversion angle of 8°-9° (range -5° - 31°, SD ± 9.66°). Further, some dysfunctions may not be correlated with a higher prevalence of extreme version angles, for example, patients receiving hip resurfacing and patients who have fractured their femur. However, we do see higher rates of extreme version (14%) for individuals receiving a hip replacement and retroversion for osteoarthritis, developmental dysplasia, and Legg-Calvé-Perthes disease.
Recommendations Correlated with Dysfunction: Knee pain is correlated with significantly less hip and knee internal rotation and more knee varus, tibia external rotation, and foot external rotation during a variety of activities. This implies that the recommendation to turn the feet out and/or "drive" the knees into varus (knees bow out) may contribute to an increased risk of pain and dysfunction.

Research Review

Cadaver Studies

Several studies measured version angles on cadaver femurs. Hoagland et al. measured and compared the femurs of Caucasian and Chinese cadavers. The study states that the axis between the neck and the shaft averages 135°, and the femoral neck is anteverted from the transcondylar plane an average of 8° for adults. However, the findings demonstrate that Caucasian cadaver femurs exhibited an average anteversion angle of 7.0° (range: -2° - 35°) for males and 10.0° (range: -2° - 25°) for females. Hong Kong Chinese cadavers exhibited an average of 14.0° (range: -4° - 36°) for males and 16.0° (range: 7° - 28°) for females (1). Eckhoff et al. investigated variations in femoral anteversion in an adult African skeletal population. This study demonstrated greater average anteversion (19°) and a significant right‐left variation (21° vs. 17°) when compared to previously reported Caucasian and Asian populations (average anteversion 9°) (2). Toogood et al. compared 200 male and female adult skeletons from cadavers older than 18 years with no history of underdeveloped femora at the time of death or anatomical abnormality on gross visual inspection (i.e., those affected by diseases, such as arthritis, osteonecrosis, or another deformity). The samples were diversified based on race, age, and gender. Specimens were digitally photographed in standardized anterior-posterior and lateral positions, and measurements were obtained using ImageJ software. Outcome measures included the anterior, posterior, superior, and inferior shifts of the femoral head on the axis of the femoral neck. The findings demonstrated that the average anterior/posterior shift ratio for the entire sample was higher than 1 (mean, 1.14), indicating that, on average, the femoral heads of the population were shifted forward. Additionally, the average superior/inferior shift ratio for the sample was less than 1 (mean, 0.90), suggesting that, on average, the femoral heads of the population were shifted downward. Measures of sphericity of the femoral head and concavity of the acetabulum were measured. The average alpha and gamma angles (averages: 45.61° and 53.46°, respectively) were higher than the average beta and delta angles (averages: 41.85° and 42.95°, respectively), which suggests that, on average, the femoral head-neck junctions had more concavity at the back and bottom than at the front and top. Further, the physeal angle was also measured, which could be defined as the amount of rotation of the femoral head on the femoral neck. The average ante/retro physeal angle for the sample was less than 90° (mean, 74.33°), indicating that, on average, the femoral heads of the population were anteverted. Similarly, the average abduction/adduction physeal angle was less than 90° (mean, 81.83°), indicating that, on average, the femoral heads of the population were abducted. Based on this data, males and females differed in translation, rotation, and concavity. Males had more inferior offset in translation and more abduction and anteversion in rotation. Males and those older than 50 also had less concavity of the anterior head-neck junction. No significant differences based on gender or age were found for neck-shaft relationship measurements. Additionally, it was not that, "...as either neck version or angle of inclination increased, translation in the same direction (anterior for neck version, superior for angle of inclination) at the head-neck junction tended to decrease" (3). These studies demonstrate that cadaver femurs exhibit an average version angle between 7° - 19° with a range of -4° - 36°. Some of the variations can be correlated with age, gender, and/or ethnicity. Further, the femoral head is slightly anterior, inferior, and abducted on the femoral neck, and the femoral head-neck junctions may have more concavity at the back and bottom than at the front and top. Note that Togood et al. (2009) demonstrated that as the neck version angle or inclination angle increased, translation in the same direction (anterior for the neck version angle, superior for the angle of inclination) at the head-neck junction tended to decrease, demonstrating some compensatory adaptation.

Healthy Hips in Living Subjects

Additional studies compare version angles of healthy individuals. Philippon et al. collected data on 61 asymptomatic youth ice hockey players (age: 10-18 years) and 27 youth skiers (controls) (age: 10-18 years) who underwent a clinical hip examination, including hip angle measured by magnetic resonance imaging (MRI). The findings demonstrated that the mean femoral anteversion of hockey players was 4.8° (± 10.0°), and skiers 4.5° (±8.8) (4). Saikia et al. compared 104 healthy males and females (age: 20-70 years) with normal hip joints and no history of dysplasia, skeletal immaturity, or surgery. 184 normal hips were evaluated with a physical exam, plain X-ray, and CT scan, with measured parameters including the center edge angle (CE) angle of Wiberg, the acetabular angle of Sharp, the neck shaft angle, the acetabular version, the femoral neck anteversion, the acetabular depth, and the joint space width. The findings demonstrated that the average parameters observed were as follows: acetabular angle 39.2°, center edge angle 32.7°, neck shaft angle 139.5°, acetabular depth 2.5 cm, joint space width 4.5 mm, acetabular version 18.2°, and femoral neck anteversion 20.4° (5). Sengodan et al. collected data on 200 South Indian individuals (400 hips) (age: 20-70 years) with normal hip joints who received abdominal (and hip) computed tomography (CT) scans between 2012 and 2014 for other reasons. The average value of the neck-shaft angle (NSA) was 135.4°. The average value for men was 136.7 (range: 128°–147°), and the average value for women was 134.18° (range: 122°–145°). The average value for the right side was 134.60° (range: 122° to 147°), and the average value or the left side 136.26° (range: 124°–147°). The average value of the acetabular angle of sharp (AA) is 35.5°. The average value for men was 35.33°, the average value for women was 35.73°, the average value for right sides was 35°, and the average value for left sides was 36.07° (the range for all was the same 24°–42°). The average value of hip anteversion was 18.64°. Further, the average value among males was 17.84° (range: 10°–33°), for females was 19.45° (range: 11°–33°), for right hips was 18.05° (range: 11°–33°) and for left hips was 19.25° (10°–29°) (6). Again, these studies demonstrate an average anteversion angle for an adult of between 18.2° and 18.64° with a range of 10°–33°, and adolescents have an average anteversion angle of 4.5° - 4.8° with an SD of ±10.0°.

Pathological Hips in Living Subjects

Several studies have also compared the anteversion angle (and various other angles) of the femurs of individuals with diagnosed dysfunction. Atkinson et al. compared 100 males and females (age: 52-54 years) undergoing pelvic CT scanning as part of a routine workup for hip resurfacing arthroplasty surgery for hip osteoarthritis. All participants underwent radiographic measurements for the following parameters: femoral neck angle, femoral torsion, femoral offset, acetabular version, and acetabular inclination. The findings demonstrated that gender differences in femoral torsion/anteversion, femoral neck angle, and acetabular inclination were statistically insignificant. Males had a mean femoral torsion/anteversion of 8° (range -5° - 26°), a mean femoral neck angle of 129° (range 119°-138°), and a mean acetabular inclination of 55° (range 40°-86°). Females had a mean femoral torsion/anteversion of 9° (range -2° - 31°), a mean femoral neck angle of 128° (range 121°-138°), and a mean acetabular inclination of 57° (range 44°-80°). Females had a significantly greater acetabular version of 23° (range 10°-53°) compared with 18° in males (range 7°-46° (p = 0.02), and males had a significantly greater femoral offset of 55 mm (range 42 to 68 mm) compared with 48 mm (range 37 to 57 mm) in females (p = 0.00) (7). Koerner et al. compared 417 consecutive patients with femur fractures treated with an intramedullary nail and who had no history of bilateral injuries, previous injuries, or previous deformities. All participants were scanned to obtain axial images through the proximal and distal femurs of the injured and uninjured femurs, and the images of the uninjured femurs were included in this study. The findings demonstrated that the average version angle of the uninjured femur for all the patients was 8.84° ± 9.66° of anteversion. There was no significant difference in the mean version between males and females. Differences in ethnicity did not reach statistical significance either, with 21.4% of white males exhibiting retroversion of the uninjured femur, compared with 15.1% of African Americans and 7.1% of Hispanics. Additionally, differences in the prevalence of > 10° retroversion did not reach statistically significant differences, being exhibited by 5.7% of African American males and 5.9% of females, 0% of white males and females, and 1.8% of Hispanic males and 0% of females (8). Pierrepont et al. compared the native femoral anteversion angle of 1215 patients prior to hip replacement surgery. As implied by the study by Ezoe et al. (2006), it may be reasonable to conclude that this population would have a higher proportion of individuals who fall outside of the normal range (beyond 2 standard deviations). Native femoral neck anteversion was measured from 3-dimensional computed tomography (CT) reconstructions in subjects. The findings demonstrated that the median femoral anteversion across all 1215 subjects was 14.4° (−27.1 – 54.5°, IQR 7.4 – 20.9°) (Figure 2). The median anteversion in males was 12.7° (range: −27.1 – 45.5°, Interquartile range (IQR): 6.0 – 19.1°), compared to female anteversion of 16.0° (range: −14.0° - 54.5°; IQR 9.7° - 22.4°). Gender differences were statistically significant, p < 0.0001.14% of patients had extreme anteversion (<0° or >30°) (9). Ezoe et al. retrospectively examined anteroposterior radiographs of the pelvis of 250 patients (342 hips), including 56 patients (112 hips) with normal findings, 66 patients (70 hips) with osteoarthritis, 64 patients (74 hips) with developmental dysplasia, 30 patients (36 hips) with osteonecrosis of the femoral head, and 34 patients (50 hips) with Legg-Calvé-Perthes disease. The findings demonstrated that the prevalence of acetabular retroversion was 6% (7 of 112 hips) in the normal group, 20% (14 of 70 hips) in the osteoarthritis group, 18% (thirteen of seventy-four hips) in the developmental dysplasia group, 6% (2 of 36 hips) in the group with osteonecrosis of the femoral head, and 42% (21 of 50 hips) in the group with Legg-Calvé-Perthes disease (10). These studies again demonstrate an average anteversion angle of 8°-9° (range -5° - 31°, SD ± 9.66°). Further, some dysfunctions may not be correlated with a higher prevalence of extreme version angles, for example, patients receiving hip resurfacing and patients who have fractured their femur. However, we do see higher rates of extreme version (14%) for individuals receiving a hip replacement and retroversion for osteoarthritis, developmental dysplasia, and Legg-Calvé-Perthes disease.

Caption: Bell curve distribution of hip anteversion and retroversion from Pierrepont et al. (2009)

Caption: Bell curve distribution of findings from Pierrepont et al. (9)

Recommendations Correlated with Dysfunction

Studies have demonstrated that both feet turn out and knee varus (knees bow out) are correlated with knee pain. Willson et al. compared 20 active females (age:18-35 years) with patellofemoral pain syndrome (PFPS) and no history of meniscus pathology, ligamentous pathology, or traumatic injury during the previous 6 months. PFPS was confirmed with at least a 3/10 pain in the patellar tendon, knee joint, or peripatellar area (and not solely at the iliotibial band) while squatting, sitting, running, or jumping, or provoked by compression of the patella at 15° knee flexion, palpation of the posterior surface of the patella, or a score less than 85/100 points on the Anterior Knee Pain Scale, Participants were assigned to a control group (normal) or a PFPS group for 1 session. The exercise protocol included running for 23 m at 3.7 m/s for 5 sets, single-leg squats deeper than 60° of knee flexion for 1 set, 5 reps, with only bodyweight, at a moderate 15 reps/min tempo, and single-leg jumps as high as possible for 1 set, 5 reps/set, with only bodyweight. The findings demonstrated that the PFPS group exhibited significantly less hip internal rotation and knee internal rotation, greater knee external rotation, foot external rotation, and tibial external rotation across all activities, and less hip internal rotation excursion during the single-leg squats (measured using motion capture software). No group exhibited significant differences in vertical ground reaction forces, knee extension moments, knee flexion angle at peak knee external moment, knee transverse plane rotation, knee transverse plane excursion, hip adduction, hip adduction excursion, and hip internal rotation (measured using motion capture software) (11). Winslow et al. compared 24 university-aged female ballet dancers with patellofemoral pain syndrome in front of or under the knee cap satisfying 3 of the following 5 criteria: associated with kneeling, associated with squatting, during stair climbing, associated with sensations of cracking/grinding in the knee or associated with incidents of knee joint locking or “catching,” who danced a minimum of 4 hours/week. Participants were assigned to a control group (no pain) or a PFPS group. All participants’ legs were examined for iliotibial band (ITB) tightness with the Ober Test, followed by 10 dancers randomly selected to measure foot turn out in first position (dancers "first position"), knee flexion during demi plie in first position, and tibial external rotation in first position and during demi plie. The findings demonstrated that the PFPS group exhibited a statistically significant correlation between a positive Ober's test and knee pain, including ITB tightness in 11/14 knees with PFPS compared to 9/34 knees in the pain-free control group. Additionally, the z-score was significant for foot turnout and demi-plié. Dancers with a positive Ober's test had a mean tibial external rotation of 69.1° (54.0°) during demi-plié, compared to 62.3° (23.8°) for those with a negative Ober's test (12). Lo et al. compared 82 males and females diagnosed with symptomatic knee osteoarthritis (OA) of at least a grade 2 in the same knee for at least 1 month and no history of using an assistive device for ambulation. Participants were recruited from an RCT of vitamin D treatment for osteoarthritis. They were further classified by the presence of varus during gait, including the first appearance of varus or abrupt increase of existing varus of the weight-bearing limb during ambulation with a return of the limb to a less varus alignment during the swing phase of gait based on the first and last 5-6 strides walking away from and towards the camera. The findings demonstrated that participants with OA exhibited significantly larger varus thrust on radiographic assessment and increased pain (measured with the WOMAC) (13). In summary, knee pain is correlated with significantly less hip and knee internal rotation and more knee varus, tibia external rotation, and foot external rotation during a variety of activities. This implies that the recommendation to turn the feet out and/or "drive" the knees into varus (knees bow out) may contribute to an increased risk of pain and dysfunction.

A Better Recommendation:

Before accepting small imperfections in movement as evidence of permanent structural differences, it is recommended that these issues (e.g., feet turn out, knees bow in, etc.) are treated as soft-tissue issues and targeted with an "integrated manual therapy," "corrective exercise," or "movement prep" routine. (Note that "movement prep" is a term often used to imply a circuit of corrective exercises used as a warm-up.). There are many potential benefits of corrective exercise and manual therapy, and they are very low risk. The worst outcome from using these interventions is likely that nothing happens, at which time foot placement could be altered. Although a full review of manual therapy and corrective exercise is beyond the scope of this course, the following categories of courses are recommended, and we have included a corrective exercise routine to address feet turnout below: