|Home | About | Journals | Submit | Contact Us | Français|
Objective: While there is evidence to suggest that the magnitude of the quadriceps (Q) angle changes with alterations in foot position, a detailed quantitative description of this relationship has not been reported. Our purpose was to determine the effect of varying foot placement on the magnitude of the Q angle.
Design and Setting: A mixed between-within, repeated-measures design was used to compare Q angles derived under static weight-bearing conditions with the feet positioned in self-selected versus standardized stance positions.
Subjects: Twenty healthy young-adult men and women with no history of acute injury to or chronic dysfunction of the lower limbs.
Measurements: We placed light-emitting diodes bilaterally on the left and right anterior superior iliac spines, the tibial tuberosities, and the midpoints of the patellae to bilaterally define the Q angles. An OPTOTRAK motion-measurement system was used to capture x,y coordinate data at a sampling rate of 60 Hz. These data were subsequently filtered and used to calculate the magnitude of the left and right Q angles.
Results: A repeated-measures analysis of variance revealed that when measured statically, Q angles differed significantly between stance positions (P < .001) and limbs (P < .05). Depending on the stance adopted, mean Q angles varied from 7.2° to 12.7° and 11.0° to 16.1° in the left and right lower limbs, respectively. Q-angle measurements taken in conjunction with the Romberg foot position most closely resembled those gathered with the feet in a self-selected stance (Pearson r = 0.86 to 0.92).
Conclusions: Q-angle magnitude varies with changes in foot position, increasing or decreasing as the foot rotates internally or externally, respectively. These data demonstrate the need for a standardized foot position for Q-angle measurements.
The quadriceps (Q) angle is a clinical measure of the alignment of the quadriceps femoris musculature relative to the underlying skeletal structures of the pelvis, femur, and tibia. It provides a reasonable estimate of the lateral force vector acting on the patella with quadriceps contraction1 and the tibial-tuberosity position relative to the midline of the trochlea.2 An excessive Q angle is considered indicative of extensor mechanism malalignment and has been associated with anterior knee pain,3–8 patellar subluxation or dislocation,9–12 and lower limb overuse injuries.13–16 It is often used as a requisite measure to identify candidates for surgery17–22 and as a means to assess surgical outcome.20,23–26 Most recently, an excessive Q angle has been implicated as a potential risk factor for noncontact anterior cruciate ligament injuries in female athletes.27–28
Given the purported significance of the Q angle, it is problematic that there is a lack of agreement within the literature as to what might be considered its “pathologic” limit.29–31 Some researchers17,21 regard Q angles in excess of 20° to be pathologic, while others have suggested that values as low as 10° to 14° are problematic.12,18,32 This lack of consensus may be due in part to the absence of a standardized measurement position.5,33 Methods often differ with respect to whether the subject is standing or supine and whether the feet are in a self-selected or controlled position.30 Empirical investigations31,33,34 clearly demonstrate that the magnitude of the Q angle increases slightly (0.2° to 1.3°) when an individual is standing rather than supine. Less understood is the effect of foot position on Q-angle magnitude, even though it is viewed as a factor that must be controlled during measurement.7,33,35
Olerud and Berg's36 photographic analysis of the variation in the standing Q angle with inward and outward foot rotation represents the only systematic investigation of its kind. However, 4 aspects of this study are problematic. First, measures were derived from one limb only. It is unclear which limb (ie, left or right) was measured. This is problematic given that recent authors37,38 have observed asymmetric Q angles in the right and left lower limbs. It is also not apparent whether the foot of the unmeasured limb was constrained in the same position as the measured limb or if it was simply not controlled. Second, the foot positions used were described only in terms of the rotational position of the long axis of the foot and without any indication of the distance between heel centers; therefore, they are irreproducible. Third, the 3 positions described (the long axis of the foot at 0°, 15° inward rotation, and 15° outward rotation) vary considerably from what has been described as an average preferred- or natural-stance position (7° external rotation).39,40 This leads us to question the generalizability of the results. Fourth, the results simply revealed that the Q-angle magnitude increased or decreased by approximately 5° with 15° of internal or external rotation of the foot, respectively. The actual data, in the form of the raw data set or descriptive statistical measures, were not reported.
Our primary purpose was to compare the Q-angle magnitude when measured with the feet positioned in self-selected versus standardized stances. The standardized foot positions included placing the medial borders of the feet together, as in the Romberg test of balance,40 and the average preferred-stance position reported by McIlroy and Maki.39 While the former position is easily replicated by subjects and may be viewed as reliable, the latter arguably represents an average preferred foot position for an adult population. Secondary purposes included comparing Q-angle measurements of the right and left lower limbs and assessing the degree of association between Q-angle measurements taken with the feet in self-selected versus constrained-stance positions.
All methods were approved before data collection by the University of Western Ontario's Review Board for Health Sciences Research Involving Human Subjects. Each subject provided informed written consent. Exclusion criteria included a history of lower-limb acute injury or chronic dysfunction. The sample of healthy active young adults included 6 men (height = 1.80 ± 0.09 m, mass = 80.6 ± 11.1 kg) and 14 women (height = 1.68 ± 0.06 m, mass = 62.4 ± 8.3 kg), ranging in age from 19 to 30 years (age = 22.1 ± 3.5 years).
We defined right and left Q angles by placing infrared light-emitting diodes (LEDs) bilaterally on the anterior superior iliac spines, midpoints of the patellae, and centers of the tibial tubercles while each subject stood with the feet in the Romberg position. The anatomical landmarks were located through palpation, visual estimation, and measurement by a single examiner. Three-dimensional coordinate data were collected at a sampling rate of 60 Hz using a single bank of OPTOTRAK motion-measurement sensors (Northern Digital Inc, Waterloo, ON, Canada). The system's calibration was verified before data collection using a rigid 3-dimensional (x,y,z) orthogonal jig. A mean accuracy of 0.5 mm was determined. This high degree of accuracy is consistent with that of other active optical-tracking systems.41 Participants stood on an elevated platform approximately 3 m from the position sensors, which afforded a viewing area of approximately 1.5 m2, with knees extended and the quadriceps muscle group relaxed, and their feet in each of the following stances: self-selected, Romberg (ie, medial borders of the feet touching), and average preferred stance (ie, 0.17 m between heel centers, with a 14° angle between the long axes of the feet). Using predefined lines and landmarks on the testing platform surface, we carefully placed the subject's feet in the average preferred-stance position. Five data-collection trials were completed for each foot position to control for the possibility of variation in the Q-angle measure due to body sway. The data were filtered at 10 Hz using a low-pass, fourth-order, recursive Butterworth filter. Using trigonometric algorithms for the filtered x,y coordinate data, we calculated the angular orientation of the quadriceps and patellar tendon rays and the magnitude of the resultant Q angle in the left and right lower limbs (Figure).
A preliminary investigation determined intratester reliability of the Q-angle measure. On 2 separate occasions separated by 1 week, LEDs were placed bilaterally by one investigator on the same 10 individuals while they stood with their feet aligned in the Romberg position. An intraclass correlation coefficient (2,1) procedure42 yielded an intratester reliability of 0.92, and the standard error of measurement was 1.4°. Differences in static Q angles by foot position (self-selected, Romberg, average preferred stance), limb (right, left), and trial were analyzed with a mixed between-within, repeated-measures 3-factor analysis of variance procedure using post hoc Scheffe F tests to distinguish the source, if any, of identified effects. By specifying the measurement trial as a within-subjects factor, data from all 5 trials per foot position were entered into the statistical analysis. We also calculated Pearson product moment correlation coefficients (r) to assess the degree of association between Q-angle measures derived in the 3 different positions.
Descriptive statistics for the Q-angle measurement are summarized in Table Table1.1. When measured statically, significant differences in Q angles between foot positions (F2,38 = 34.09, P < .001) were observed, with values different among all 3 stance conditions. Q angles in the right and left legs were greatest when measured in the Romberg position and least when the feet were placed in the average preferred-stance position. A significant difference in Q-angle magnitude between limbs (F1,19 = 4.14, P < .05) was also observed. Mean values derived from the right limb were 3.0° to 3.8° larger than those in the opposite limb, with the exact magnitude of the difference varying with the foot position adopted. We found no significant differences by measurement trial (F4,76 = 0.67, P < .62), for the interaction effects of foot position by limb (F2,38 = 0.68, P < .51), or for foot position by limb by trial (F8,152=1.46, P < .18). Correlational analysis revealed that Q-angle measurements taken in conjunction with the Romberg foot position most closely resembled those gathered under self-selected stance conditions (Table (Table22).
In their systematic investigation of changes in Q angle with changes in foot position, Olerud and Berg36 reported that the Q angle increased or decreased by 5° with 15° internal or external rotation of the foot, respectively. Others13,33,43 have studied the Q-angle magnitude while controlling foot position, yet they have done so with the aim of studying the relationship between knee conditions and lower limb structural variables only. Cowan et al,13 for example, in a study of overuse injury among 294 male infantry trainees, had participants stand with the heels spaced 7.5 cm apart and the medial borders of the feet 60° divergent. In this position, observed Q angles ranged from 0° to 26° and averaged 10° (SD = 5°). Reider et al,43 in contrast, observed a mean Q angle of 15.9° in healthy young controls who stood with the medial borders of the feet placed together, side by side, as is prescribed in the Romberg test of balance. In another study,33 Q angles averaged 11.1° (SD = 4.9°) and ranged from 1.0° to 25.0° for 60 men and women who stood with the long axes of their feet positioned perpendicular to the coronal plane but with their feet set an unknown distance apart. The lack of consistent subject positioning in these studies is striking. It is also interesting to note that none of these studies provided a rationale or justification for the foot positions used.
If a measurement such as the Q angle is to be a criterion in determining an individual's risk for injury or candidacy for surgery, then it must be accurate, valid, and reliable.44 We used x,y coordinate data captured with an OPTOTRAK active optical-tracking system to calculate the magnitude of the frontal-plane Q angle. This method is of value in that it is comparable with the photographic methods described in previous investigations13,36 yet methodologically preferable because it has been shown to effectively reduce measurement variability.41,45 Our methods of positioning the subject in an upright, weight-bearing posture with controlled-stance positions were also purposely chosen to enhance the ecologic validity and reliability of the resulting measurements, respectively. The average preferred- or natural-stance position was used because it may, as McIlroy and Maki39 suggested, meet the need to standardize while minimizing the extent of constraint on an individual's self-selected foot position. The decision to use the Romberg position was more arbitrary, yet it is easily replicated and has been used by others43 when measuring the Q angle.
Our static observations generally agreed with those of Olerud and Berg36: the magnitude of the angle increased or decreased as the foot rotated internally or externally, respectively. However, we observed somewhat smaller static Q-angle changes with alterations in foot position. These differences are most likely accounted for by our use of less extreme foot positions, leading to smaller amounts of internal or external rotation of the lower limb. The observed differences in Q-angle patterns between stances may be primarily attributed to the transverse-plane positioning (ie, internal or external rotation) of the femur and tibia imposed by foot position. The Romberg position, for example, requires greater lower-limb internal rotation, while the average preferred stance requires greater lower-limb external rotation. For most, the self-selected foot position was represented by a stance position that fell between these 2 extremes. The self-selected stance was clearly identified as the most comfortable experimental position for the completion of the task; participants frequently mentioned their discomfort with the average preferred-stance position and, to a much lesser extent, the Romberg stance.
Given the cost, time, and expertise required, it is unlikely that the motion-measurement device we used will be available in most clinical settings. Nonetheless, these findings have important clinical implications for practitioners and patients alike. First and foremost, the Q-angle magnitude changes with alterations in foot position. The practitioner must recognize the influence of foot positioning on the Q angle and, therefore, ensure that individuals are always similarly positioned when measures are gathered. The methods used should be accurately and completely described to increase the generalizability of studies reported within the literature. Second, practitioners should consider the ecologic validity of the measurement position used. We purposely chose a standing, weight-bearing position because individuals are more likely to experience patellar dysfunction when the knee joint is loaded. It is important to recognize, moreover, that while self-selected foot positions provide the greatest degree of ecologic validity, an individual's inability to replicate the position over time may limit the reliability of the resulting Q-angle measurement. Understanding the degree of association between Q angles measured under self-selected versus controlled foot positions may make the latter a meaningful alternative to the former. Our descriptive data clearly demonstrate that measurements taken with feet side by side yielded mean values that were larger than, but most closely resembled, those found under self-selected conditions. It is for this reason that we recommend the use of the Romberg position to standardize stance in future studies measuring the Q angle. Measures taken in the average preferred-stance position clearly underestimate those derived with feet in a self-selected stance. Correlational analyses, moreover, suggested that left Q-angle measures in the 2 calibrated stances yielded reasonable estimates (r2 = 85% to 87%) of the left Q angle in a self-selected stance position, while right Q-angle measures were somewhat less predictive (r2 = 67% to 74%). Why Q-angle measures gathered under controlled foot positions were less predictive in the right than the left lower limb is unknown. Similarly, why the mean Q-angle measures we report in this study differed significantly between the lower limbs is unclear. This is not the first investigation in which bilateral Q-angle asymmetry has been observed, although similar reports37,38 have only been published since 1997.
The Q-angle measurements reported in this study were derived from a healthy, young adult population. Whether similar results would be observed in those symptomatic for anterior knee pain or patellar subluxation or dislocation is unknown. Their generalizability, moreover, is limited to similar investigations in which the primary purpose is to understand how methodologic variation may affect measurement outcome. The inclusion of men and women in our sample of interest, while acceptable for this methodologic study, yielded mean data that are not appropriate for clinical interpretation or comparison. Our methods have been thoroughly reported so that future investigators can replicate our efforts. If the differences in reported Q angles over time within an individual or from study to study are to be interpreted as true differences or as a product of the measurement method used, a standardized method of measurement must be established and methods accurately reported. Only with such efforts can the enigmatic nature of the Q angle be better understood.