The hip is a ball-and-socket joint that serves to unite the upper and lower halves of the body and provide segmental range of motion and force transmission for ambulation1. The femoral neck is a bony process that attaches the head of the femur to its shaft. The angle of the femoral neck relative to its shaft in the transverse plane (i.e., the degree of femoral torsion) is known as the angle of femoral version and is the result of torquing within the bone as it develops. At birth, the angle of femoral version is approximately 30° and later decreases to 15° once a person reaches adulthood2. However, the amount of femoral version exhibited among adults does vary and appears to be influenced by mechanical loading2. In such cases, the angle of femoral version is not consistently 15° throughout the adult population but instead can range from -23° (retroversion) to 63° (anteversion)3.

The greater trochanter, a bony landmark situated opposite the femoral head, is a thickening of the lateral femur that is used as a proxy to assess the angle of femoral version due to its inherent palpability. A well-known application of this anatomical landmark to evaluate femoral version is the Craig's test, or trochanteric prominence angle test4,5,6. The Craig's test is administered with the patient lying prone. The examiner flexes the patient's knee to 90°, then palpates the greater trochanter by manually rotating the patient's hip medially and/or laterally until the most prominent position of this bony landmark is felt6 (Figure 1). The angle of femoral version is then measured with a goniometer as the angle between the long axis of the lower leg in this position and the vertical axis (Figure 2). It is thought that when the Craig's test is administered properly, the patient's examining hip will be placed in a manner (medial or lateral) and degree of rotation that precisely corresponds with their degree of femoral version. For example, an individual with a femoral version angle of 15° would have had the most lateral prominence of their greater trochanter felt by the examiner when their hip was internally rotated 15°.

In this manner of administration, the Craig's test is frequently used to identify and quantify femoral anteversion, defined as excessive femoral version such that the femoral head torques anteriorly relative to the shaft >30° in the transverse plane, and femoral retroversion, defined as insufficient femoral version where the femoral head torques ≤ 8° posteriorly in the transverse plane4,6 (Figure 3). Femoral anteversion and retroversion are of clinical relevance because both conditions are linked to musculoskeletal disorders along the kinematic chain4,6,7. Notably, femoral anteversion is linked to anterior acetabular labral tears4,8 and in-toeing gait2,7. Femoral retroversion is a main cause of femoroacetabular impingement, a potentially painful precursor syndrome to osteoarthritis9.

Although the Craig's test is desirable due to its cost-effectiveness and non-invasive administration, its validity has been called into question. In a study comparing magnetic resonance imaging (MRI) calculations of femoral anteversion and the Craig's test, Souza et al.6 found that, while inter- and intra-tester reliability of the Craig's test were high, there was only moderate agreement between MRI measures of anteversion and measures produced by the Craig's test. As such, the wide confidence interval of this moderate relationship gave the researchers pause in recommending the Craig's test for widespread clinical use. In a later study, Uota et al.10 used computed tomography (CT) to examine the validity of the Craig's test. They found that measures of femoral anteversion obtained using CT, the gold standard in the field, failed to explain over 75% of the variance of the measures obtained from the Craig's test. While the Craig's test did produce somewhat true values of femoral anteversion, it could not be considered valid, nor did it appear to exclusively measure femoral version despite its ubiquitous usage for that method10. This study was the first of its kind to assess validity of the Craig's test in healthy adults, highlighting the need for further investigation.

Our study aim was to indirectly assess the validity of the Craig's test using dual-energy X-ray absorptiometry (DXA) in healthy adults. DXA primarily measures bone mineral density (BMD). From measurements obtained by DXA, clinicians can diagnose osteoporosis, a progressive disease leading to the breakdown of bone, and assess a patient's disease progression once they have already been diagnosed with osteoporosis11. As a diagnostic tool, DXA is considered to be accurate and is therefore regarded as the gold standard for the identification and management of osteoporosis and its precursor, osteopenia12. In order to obtain accurate BMD readings of the femoral neck, the patient lies supine with their hips positioned in about 15° of internal rotation, matching the average of 15° of femoral anteversion in the adult population. With the femoral neck set parallel to the examination table in this manner, it is believed that the X-ray beams on the DXA measuring arm hit the femoral neck at a perpendicular angle to produce an accurate measurement of BMD13,14. However, while this conventional hip placement method is effective for those people who have the expected amount of femoral anteversion in the adult population, it does not account for people with more anteversion than normal or with femoral retroversion. In these cases, the X-ray beams do not hit the femoral neck in a perpendicular manner and may produce a falsely elevated measurement of BMD14.

Our aim was to determine if femoral neck BMD values were lower in participants whose hips were placed in a customized position, using femoral version values obtained from the Craig's test, compared to the conventional hip placement in 15° of internal rotation. This is based on the presumption that BMD measurements obtained via DXA are lowest when the femoral neck is scanned in a position where it is precisely parallel to the scanning table13,14.