Clinical Biomechanics 14 ( 1999) 7- 13

Bilateral Q angle asymmetry and anterior knee pain syndrome

Lori A. Livingston,* James L. Mandigo
Department ofKkesio/ogy and Physical Education, Wilfid Laker University. 75 Univerxity Avenue West, Waterloo, ON, Canada N2L 3C5 Received 1 December 1997; accepted 1 May 1998

Abstract
Objective. To determine whether quadriceps (Q) angles were bilaterally symmetric in individuals asymptomatic vs symptomatic for anterior knee pain. Design. Cross-sectional study. Background. Previous attempts to link excessive Q angles to the occurrence of knee pain have yielded equivocal results. Deriving unilateral .rather than bilateral measures of the Q angle and thereafter analysing data using traditional between-group analysis-of-variance structural models may, however, play a role in obscuring the true nature of the Q angle-knee pain relationship. Methods. Left and right Q angles were goniometrically measured in 75 subjects (37 males, 38 females) while they adopted a static, standing position with quadriceps relaxed. The majority (n = 50) were asymptomatic, while the remainder were unilaterally (n = 11) or bilaterally (n = 14) symptomatic for anterior knee pain. A questionnaire was used to determine the extent and magnitude of pain experienced in each of the symptomatic subjects. Results. Significant right vs left lower limb differences in Q angles were observed by group (p < 0.001) and group by gender (p<O.O5). Mean values, however, did not always reflect the true variation of data within the sample. Forty-seven percent of the subjects studied demonstrated a minimum 4 bilateral Q angle difference, while in 13 of 75 subjects, this difference ranged from 8 to 12. Only a weak yet significant relationship between right and left Q angles (r = 0.53, p <O.OOl) was noted. While there were no correlations between Q angle measures and the magnitude of discomfort experienced in unilateral knee pain sufferers, these relationships were weak yet significant in bilateral knee pain sufferers. Conclusion. Q angles are not bilaterally symmetric, with the magnitude and direction of the observed asymmetry varying according to whether an individual is asymptomatic, unilaterally symptomatic, or bilaterally symptomatic for anterior knee pain.

Relevance Future investigations of the Q angle must ensure that measures are derived bilaterally and analysed appropriately. Data from unilateral vs bilateral symptomatic subjects should be evaluated separately, and the shortcomings of standard ANOVA structural models must be recognized. 0 1999 Elsevier Science Ltd. All rights reserved.
Keywords: Q angle; Bilateral asymmetry; Anterior knee pain

1. Introduction A quadriceps (or Q) angle in excess of 15-20 continues to be regarded [l-8] as an anatomic risk factor in the etiology of overuse injuries of the knee, despite the contention that it is a less reliable clinical measure than wa,s previously believed [9-121. The inability of researchers to reach a consensus on what might be considered a normal angle [13], or to consistently link excessive Q angles with the occurrence of knee pathologies [14-191, has raised doubt about its
Corresponding author. E-mail: Ilivings@machl.wlu.ca.

diagnostic value. The lack of a standardized measurement protocol is at least partially to blame since the size of the angle can differ dramatically when measured using differing procedures [12,20]. Indeed, variations in subject posture [21], the contractile status of the quadriceps muscle group [21,22], knee position [15,23], and foot position [23,24] during measurement are known to alter the magnitude of the angle. Direct comparisons between studies using differing measurement procedures are therefore not feasible [14] as it is impossible to discern whether observed changes in the Q angle are attributable to independent variable manipulation or the methodology that was utilized.

0268-0033/99/$ -- see front matter 0 1999 Elsevier Science Ltd. All rights reserved PII: SO268-00~~3(98)oOO4S-X

L. A. Livingston, J. L. MandigolClinical

Biomechanics 14 (1999) 7-13

Less well understood is the extent to which the common practice of deriving data from one lower limb only [1,19-21,23,25-281, or the reporting of a single, ill-defined value [13,14,29-321, has confounded our understanding of the Q angle. The measuring and reporting of a lone value is problematic for it implies that Q angles are bilaterally symmetric which is an assumption that has not been rigorously tested. The lack of attention to bilateral Q angle measurement is surprising given that bilateral asymmetry in structural measures [33-351 has been implicated in the etiology of knee pathologies. Moreover, the limited generalizability of studies which have concluded that Q angles are bilaterally symmetric is easily recognizable. Hvid and Andersen [3], for example, reported no significant differences in median Q angles observed in the right vs left legs of 29 patients with patellofemoral complaints. Messier and co-authors [36] similarly observed no significant bilateral imbalances in Q angles in a group of pain free controls (n = 20) or in runners with patellofemoral pain (n = 16). The findings of these studies are questionable, however, because of the small samples that were studied and the failure to distinguish data derived from unilaterally vs bilaterally symptomatic subjects. Others have measured and reported bilateral Q angle data using larger samples, but they have done so for descriptive purposes only [l&31,37] or for the purposes of determining the relationship between lower limb structural measures and injury [16,35] without determining the significance of observed differences between limbs. Hahn and Foldspangs [22] recent investigation is the first to provide reliable evidence to suggest that the common assumption of bilateral Q angle symmetry may be erroneous. In a study of 339 active athletes aged 14-24 yr, right Q angles were on average 3 greater in magnitude than, and significantly different @r <O.OOl) from, those in the left lower limb. A gender difference (p <O.OOl) was also noted, with female subjects tending to display larger mean Q angles than their male counterparts. The strength of this study lies in the large sample that was studied. In the absence of similar investigations against which to compare their findings, however, the authors could only speculate that the observed asymmetry was the product of a real phenomenon such as increased muscle tonus in the dominant limb, measurement error, or both. Further investigations are warranted to either validate or invalidate these findings and, thereafter, to determine why bilateral Q angle asymmetry exists. It would be of interest, moreover, to determine if the magnitude and direction of the asymmetry is similar or different in those asymptomatic vs symptomatic for selected knee pathologies. An understanding of the relationship between left and right Q angles may yield new insights into the etiology of such conditions.

The purpose of this investigation, therefore, was to compare measured values of left and right Q angles to examine whether bilateral similarities or differences existed between a group of asymptomatic controls and those symptomatic for anterior knee pain. The null hypotheses were that there would be no significant between-group (asymptomatic vs symptomatic, or, male vs female) or within-subject (right vs left limb) differences, or interaction effects.
2. Methods

Seventy-five young adult males (n = 37) and females (n = 38) ranging in age from 15 to 48 yr volunteered to participate in this study. Participants were recruited from the general university population and all methods were approved, prior to beginning data collection, by the Wilfrid Laurier University Research Ethics Committee. Upon entering the lab, each subject completed a general information and an anterior knee pain screening questionnaire. The former was used to gather information on the physical activity patterns and lower limb injury history of each subject. Individuals with a history of traumatic lower limb injury, surgery, or impairment were excluded from participation in the study. The knee pain questionnaire, with the use of visual analog scales, elicited information about the magnitude and extent of the anterior knee pain experienced, if any, during static (e.g. prolonged standing) and dynamic (e.g. walking, jogging, stairclimbing) weightbearing activities. With this information, the subjects were then assigned to one of three groups: asymptomatic (n = 50) unilateral knee pain sufferers (n = 11) or bilateral knee pain sufferers (n = 14). Table 1 shows average characteristics of the individuals tested. Q angle measurements were taken bilaterally with the subject in a standing position, with the knees extended and the quadriceps muscle group relaxed. A transparent, flexible plastic full circle goniometer (Sammons Inc, Burr Ridge, IL, USA), with two 0.18 m arms and 1 increments, was used to measure the Q angle with the anterior superior iliac spine (ASIS), midpoint of the patella, and midpoint of the tibia1 tubercle as landmarks (Fig. 1). This method has previously been described [13,14,38]. All goniometric measurements for a given subject were taken by the same investigator. At the time of measurement, the investigator was not knowledgeable as to whether the subject was asymptomatic or symptomatic for anterior knee pain. Intertester reliability and the measurement error of the gonionmeter were established in a preliminary study. An r = 0.67 (ICC(2,l)) [39] indicates that reliability was good [40], while measurement error (i.e. 1.5) was limited.

L. A. Livingston, J. L. MandigolClinical Biomechanics 14 (1999) 7-13

Table 1 Mean and SD. of the age, mass and height of the sample populations Asymptomatic Male n 26 21.6 (7.5) 80.2 (10.5) 1.77 (0.05) control Female 24 24.2 (6.4) 66.1 (12.5) 1.66 (0.06) Unilateral Male knee pain Female Bilateral knee pain Male Female

5
28.3 (5.3) 82.6 (6.7) 1.81 (0.10)

6
26.4 (7.0) 65.6 (7.2) 1.66 (0.05) -.

2i.4 (I 1.0)
92.3 (14.5) 1.84 (0.07)

8
25.0 (7.1) 62.4 (10.3) 1.69 (0.09)

Age W
Mass (kg) Height (m)

Differences in mean Q angle by group (asymptomatic vs unilateral symptomatic vs bilateral symptomatic), limb (right vs left), and gender (male vs female) were analysed using a mixed between-within threefactor analysis of variance (ANOVA) procedure. Differences in right vs left Q angles for each subject were also calculated (AQ = Right Q Angle -Left Q Angle) so that the magnitude and direction of withinsubject bilateral asymmetry could be evaluated. Correlations [i.e. Pearson product-moment (r), Spearman rank (rJ, point&serial (rpb)] were used to determine the degree of association between the Q angle variables and other dependent (i.e. magnitude of pain

experienced in symptomatic knees) and independent (i.e. group, gender) measures.

3. Results

ASIS

For the entire sample, there was only a weak relationship between right and left Q angle magnitudes (r = 0.53, p < 0.001). The lack of correlation is evident in Fig. 2, which displays the raw data. The same correlation was weak, yet significant for the asymptomatic group (r = 0.46, p <O.OOl) and insignificant for the unilateral knee pain sufferers. In contrast, for the bilateral knee pain group, there was a strong linear association (r = 0.89, p < 0.001) between right and left Q angle magnitudes. Descriptive statistics for the Q angle are presented in Table 2. While the 1.2 difference in mean values for the left lower limb between the asymptomatic and symptomatic groups may be explained by measurement error alone, those derived from the right limb showed more variation with between-group differences
30

25 -

L---.-l
Asymptomatic Unilateral Bilateral A
l

AKPS w

AKPS

0,

0,
5

2I
10

d I
15

I
20

I
25

I
30

Right

Q Angle 1)

Fig. 1. Q angle formed by the intersection of a line, from the anterior superior iliac spine (ASIS) to the midpoint of the patella, with a line from the midpoint of the patella to the midpoint of the tibia1 tubercle.

Fig. 2. A scatterplot of right Q angle versus left Q angle values for each subject. Linear correlations for the entire sample (r = 0.53, p < 0.001) and asymptomatic controls (r = 0.46, p <O.OOl) were significant but weak, while that for the unilateral knee pain sufferers (r = 0.54, p > 0.05) was insignificant. In contrast, for those in the bilateral knee pain group, there was a strong linear relationship (Y = 0.89, p < 0.001) between right and left Q angle magnitude.

10

L. A. Livingston, J. L. MandigolClinical

Biomechanics 14 (1999) 7-13

Table 2 Descriptive statistics for right and left Q angles and AQ* Asymptomatic control Male Female 24 10.5 4.2 0.9 21.0 4.0 12.2 5.2 1.1 20.0 2.0 -1.7 4.8 1.0 10.0 -9.0 Unilateral knee pain Male 5 10.9 6.7 3.0 19.0 0.0 14.4 1.3 3.3 22.0 6.0 -3.5 5.1 2.3 3.0 - 11.0 Female 6 14.4 5.4 2.2 19.0 5.0 10.8 6.1 2.5 17.0 3.0 3.6 4.9 2.0 13.0 0.0 Bilateral knee pain Male 6 15.6 3.4 1.4 22.0 13.0 10.3 1.2 0.5 12.0 9.0 5.3 3.5 1.4 12.0 3.0 Female

n
Right Q angle () Mean S.D.t S.E.M.$ Maximum Minimum Left Q angle () Mean SD. S.E.M. Maximum Minimum

26 9.5 4.6 0.9 23.0 1.0 10.4 5.7 1.1 23.0 1.0 -0.9 5.6 1.1 9.0 - 10.0

8
15.7 9.4 3.3 30.0 3.0 12.2 7.1 2.5 23.0 2.5 3.5 3.5 1.2 1.0 -2.0

AQ ()

Mean SD. S.E.M. Maximum Minimum

*AQ = right Q angle-left Q angle tS.D. = standard deviation $S.E.M. = standard error of mean

reaching a maximum of 5.7. In addition, mean right lower limb Q angles in females were greater than those observed in males in all three of the groups tested. However, significant main effect differences by group [F(2,69) = 2.16, p ~0.121 or gender [F(1,69) = 0.29, p ~0.591 were not observed. When the within-subject main effect of limb was factored into the analysis, significant group by limb [F(2,69) = 7.27, p <O.OOl] and group by limb by gender [F(2,69) = 3.22, p <O.OS] interaction effects were noted. Differences in Q angle magnitude between the right and left lower limb (AQ) for each subject (Fig. 3), and average AQ values for each group (Table 2) were calculated. When averaged across all subjects (n = 75) the AQ value was small at -0.02 (SD. 5.4). Mean AQ values for the asymptomatic males and females were negative and slightly larger in magnitude than those for the overall sample. These negative AQ values would be expected given that 72% of the asymptomatic subjects had left Q angles that were larger in magnitude than those in their right lower limb. In addition, 36 of the 50 demonstrated a minimum 4 bilateral Q angle absolute difference, while in 10 of the 50 this ranged from 8 to 10.3. For those with bilateral knee pain, the AQ results were quite different with positive mean values of 5.3 and 3.5 for males and females, respectively. Eighty-six percent of the subjects studied in this group had larger right as compared to left Q angles, while 8 of 14 (57%) had absolute bilateral Q angle differences

of 4 or more. No trends in AQ values for the unilateral group were apparent. When grouped together, 72% of the symptomatic subjects had greater right than left Q angles. AQ and group membership were significantly, yet weakly correlated (rs = -0.34, p ~0.002) while there was no significant association between AQ and gender (rpi,= -0.04). Interestingly, absolute AQ values for left hand dominant subjects (n = 3) did not exceed 4. Dependent Q angle measures and data from the anterior knee pain screening questionnaire were used to generate the Pearson product-moment correlation matrix presented in Table 3. The magnitude of the Q angle in symptomatic knees, AQ, and the absolute value of AQ were positively and significantly correlated with pain, although the relationships were weak at best and varied between the two symptomatic groups. Q angle measures were linearly associated with the extent of pain experienced by those in the bilateral knee pain group, but not for those in the unilateral knee pain group.
4. Discussion

The present findings suggest that the common assumption of bilateral Q angle symmetry is erroneous despite the observation that, for the entire sample, the mean right Q angle (M 11.5, S.D. 5.6) did not differ

L. A. Livingston, J. L. MandigoiClinical

Biomechanics 14 (1999) 7-13

11

significantly from the mean left Q angle (M 11.5, S.D. 5.5). The apparent agreement in these averaged values was not reflected in either the correlation coefficient (Y= 0.53, p < 0.001) or in the individual AQ values calculated for each subject. Almost half (47%) of the

Asymptomatic

h=50)

Symptomatic h=251

Left

Hand Dominant

Fig. 3. AQ values calculated and plotted by group. Raw data is individually displayed for (A) asymptomatic and (B) symptomatic subjects. lnverted arrows indicate those subjects who were left hand dominant (n = 3).

Table 3 Pearson product-moment correlation coefficients (r) between pain experienced and Q angle variables in symptomatic knees Magnitude of knee pain experienced Q angle in AQ affected limb IAQI

All symptomatic knees (n = 39) 0.52** Unilaterally symptomatic knees (n = 1 I) 0.42 Bilaterally symptomatic knees (n = 28) 0.52*
lp<O.Ol; **p~:0.001

0.35 0.40* -0.08 0.14 o.s1* 0.51*

individuals studied demonstrated a minimum 4 difference between right and left Q angles. In 13 of 75 subjects, these differences ranged from 8 to 12. These results provide a rather salient example of how the act of summarizing data can lead to a compromise between simplicity and accuracy [41], and leads us to question whether meaningful inferences about the Q angle can be derived from aggregate scores alone [42]. It would be erroneous in this instance, moreover, to draw inferences based on results calculated for the overall sample, since such a practice clearly disregards the heterogeneity of the groups and individuals studied. The true variation of the data was more readily apparent when bilateral minimum, maximum, and mean values, calculated on the basis of gender and group membership, were compared. While Q angles in excessof 15 in men and 20 in women are thought to predispose individuals to knee pathologies [l-8], these arbitrary limits were exceeded in 11 of 100 asymptomatic knees. In 13 of the 39 symptomatic knees studied, Q angles would be classified as excessive. However, in 12 of the 39 symptomatic knees, Q angles were less than lo. With the exception of a 30 Q angle observed in the right knee of one bilaterally symptomatic female, maximum values for right and left Q angles did not appear to differ between the groups. Greater variation was apparent in the minimum Q angles observed for each group, with values for the symptomatic groups being generally greater than those seen in asymptomatic controls. Differences in calculated group mean values may, therefore, be the result of larger minimum as opposed to larger maximum values. In asymptomatic controls, left Q angles were on average l-2 greater than those observed in the right limb for males and females, respectively. Bilateral knee pain sufferers, in contrast, had right Q angles that were on average 3-5 greater than those in the left limb. In unilateral knee pain sufferers, the absolute difference in mean right and left Q angles was approximately 3.5 in both males and females. Direct comparisons between the present findings and those of others [18,22,35] regarding the magnitude of observed bilateral Q angle imbalances are not possible because of differences in the measurement protocols utilized. These investigations do provide evidence, however, that the assumption of bilateral Q angle symmetry is erroneous and that, on average, females tend to have larger Q angles than their male counterparts. The calculation and illustration of AQ values for each subject provided a method by which to view bilateral Q angle imbalances, while overcoming the inadequacies of aggregate scores frequently used to analyse group design data [42]. The data in Fig. 3 clearly illustrate why the magnitude and direction of

12

L. A. Livingston, J. L. MandigoiClinical

Biomechanics 14 (1999) 7-13

mean right vs left Q angles differed between the asymptomatic and bilaterally symptomatic, with strong trends apparent for both groups. The two groups are clearly differentiated by the direction of their Q angle imbalances. The random scattering of AQ values for the unilateral group, however, provided no evidence of a meaningful bilateral Q angle relationship. It might be argued that the limited number of individuals studied in this latter group may have prevented a trend from emerging. An alternate explanation may be, however, that unilateral knee pain is as some have suggested [30,43] best described as a product of non-anatomic (e.g. exercise intensity, exercise frequency) factors. While the aforementioned observations yield new information on bilateral Q angle imbalances, the question of why such imbalances exist remains. Measurement error may account for some of the observed variability in right vs left Q angles [22]. There is also some suggestion within the literature that increased muscle tonus [20,22,44] in the dominant limb may lead to a reduction in the magnitude of the Q angle, in comparison to that found in the opposite non-dominant limb. Another possible explanation is that bilateral asymmetry in lower limb bone dimensions contributes to the observed differences. While human right upper limb bones tend to be more robust than left upper limb bones, a reverse pattern is seen in lower limb skeletal structures. Left lower limb bones in the human skeleton, and in particular the femur, have been shown to be on average slightly longer and heavier than right lower limb bones with females showing more asymmetry than males [45,46]. The left lower limb is thought to be used to counteract the often dominant actions of the right upper limb in what is known as the crossed symmetry pattern [47]. Limb dominance leading to changes in the muscular and skeletal components of the lower limb may, therefore, contribute to bilateral differences in Q angle magnitude. In future studies of the Q angle, investigators would be well advised to determine the upper and lower limb dominance patterns of their sample, as well as to gather functional quadriceps group strength measurements in an attempt to examine these hypothetical limb dominance-Q angle relationships. It might be of interest, moreover, to determine if homolateral or contralateral patterns of upper limblower limb dominance are associated with Q angle magnitudes, AQ values, or resulting knee pain symptoms. The results of this study suggest that the notion of an excessive Q angle-knee pain relationship is questionable at best. To view Q angles from this perspective is particularly problematic, moreover, for such an approach blatantly disregards answering the question of why those with so-called normal Q angles experience knee pain. An in vitro study [17] of patello-

femoral contact pressures completed well over a decade ago suggested that both increased and decreased Q angles must be considered as potential etiological factors in conditions leading to anterior knee pain. An extensive review of literature, however, has revealed no in vivo investigation whereby the moderate Q angle-knee pain relationship has been explained on the basis of the data collected, with investigators preferring instead to fall back on the notion that knee pain has a multi-faceted etiology and thereafter, that the observed pain must be the result of other precipitating factors [14]. The AQ value provides a new approach to studying the Q angle, whether excessive or moderate in nature, and its role as a predisposing factor for knee extensor mechanism dysfunction and pain. While the results of our correlation analysis suggest that it is a no better predictor of the magnitude of knee pain that an individual is likely to experience than is the Q angle itself, its role in predicting predisposition to unilaterally versus bilaterally symptomatic conditions is worthy of further study.

5. Conclusions The results of this study suggest that the assumption of bilateral Q angle asymmetry is erroneous and that it is not just the magnitude of the Q angle, but rather the magnitude and direction of bilateral imbalances in the measure that may predispose an individual to anterior knee pain. In future investigations of the Q angle, bilateral measurement is warranted as is the differential analysis of data collected from unilateral and bilateral anterior knee pain sufferers. Investigators must also recognize that reliance upon standard ANOVA structural models for the purposes of analysing Q angle data may be inappropriate.

Acknowledgement The authors gratefully acknowledge that financial support for this research was provided by a Short-Term Research Grant from Wilfrid Laurier University.

References
[l] Aglietti P, Insall JN, Cerulli G. Patellar pain and incongruence. LMeasurements of incongruence. Clin Orthop 1983;176:217-24. [2] Brody DM. Running injuries. Clin Symposia 1980;32(4):2-36. [3] Hvid I, Andersen LI. The quadriceps angle and its relation to femoral torsion. Acta Orthop &and 1982;53:577-9. [4] Hvid I, Andersen LI, Schmidt H. Chondromalacia patellae:the relation to abnormal patellofemoral joint mechanics. Acta Orthop Stand 1981;52:661-6.

L. A. Livingston, J. L. MandigolClinical

Biomechanics 14 (1999) 7-13

13

J. Chondromalacia patellae:patellar malalignment syndrome. 0rtho;p Clin North Am. 1979;10:117-27. [6] James SL. Chondromalacia of the patella in the adolescent. 1n:Kennedy JC, editor. The injured adolescent knee. Baltimore, MD:Williams and Wilkins, 1979:205-51. [7] Goldberg B. Patellofemoral malalignment. Pediatr Ann
[5] Insall 1997;26:32-5.

[8] Papagelopoulos PJ, Sim FH. Patellofemoral pain syndrome: diagnosis and management. Orthopedics 1997;20:148-57. [9] Schulthies SS, Francis RS, Fisher AG, Van de Graaff KM. Does the Q angle reflect the force on the patella in the frontal plane? Phys Ther 1995;75:24-30. [lo] Strother RT, Samoil D. Patellofemoral syndrome:therapeutic regimen based on biomechanics. Can Fam Physician 1989;35:1649-54. [ll] Tiberio D. The effect of subtalar joint pronation on patellofemoral mechanics. JOSPT 1987;9:160-5. [12] Woodall W, Welsh J. A biomechanical basis for rehabilitation programs involving the patellofemoral joint. JOSPT 1990;11:535-42. [13] Horton MG, Hall TL. Quadriceps femoris muscle angle:normal values and relationships with gender and selected skeletal measures. Phys Ther 1989;69:897-901. [14] Caylor D, Fites R, Worrell TW. The relationship between quadriceps angle and anterior knee pain syndrome. JOSPT 1993;17:11-6. [15] Dzioba RB. Diagnostic arthroscopy and longitudinal open lateral release:a four year follow-up study to determine predictors of surgical outcome. Am J Sports Med 1990;18:343-8. [16] Fairbank JCT, Pynsent PB, van Poortvliet JA, Phillips H. Mechanical factors in the incidence of knee pain in adolescents and young adults. J Bone Joint Surg Br 1984;66:685-93. [17] Huberti HH, Hayes WC. Patellofemoral contact pressures:the influence of Q-angle and tendofemoral contact. J Bone Joint Surg Am 1984;66:715-24. [18] McConnell J. The management of chondromalacia patellae:a long term solution. Austral J Physiother 1986;32:215-23. [19] Messier SP, Pittala KA. Etiologic factors associated with selected running injuries. Med Sci Sports Exert 1988;20:501-5. [20] Guerra JP, Arnold MJ, Gajdosik RL. Q angle:effects of isometric quadriceps contraction and body position. JOSPT 1994;19:200-4. [21] Woodland LH, Francis RS. Parameters and comparisons of the quadriceps angle of college-aged men and women in the supine and standing positions. Am J Sports Med 1992;20:208-11. [22] Hahn T, Foldspang A. The Q angle and sport. Stand J Med Sci Sports 1997;7:43-8. [23] Ando T, Hirose H, Inoue M, Shino K, Doi T. A new method using computed tomographic scan to measure the rectus femoris-patellar tendon Q-angle comparison with conventional method. Clin Orthop 1993;289:213-9. [24] Olerud C, Berg P. The variation of the Q angle with different positions of the foot. Clin Orthop 1984;191:162-5. [25] Brown DE, Alexander AH, Lichtman DM. The Elmslie-Trillat procedure:evaluation in patellar dislocation and subluxation. Am J Sports Med 1984;12:104-9. [26] Doucette SA, Goble EM. The effect of exercise on patellar tracking in lateral patellar compression syndrome. Am J Sports Med 1992;20:434-40.

[27] Kernozek TW, Greer NL. Quadriceps angle and rearfoot motion:relationships in walking. Arch Phys Med Rehabil 1993;74:407-10. [28] Reider B, Marshall JL, Warren RF. Clinical characteristics of patellar disorders in young athletes. Am J Sports Med 1981;9:270-4. [29] Boucher JP, King MA, Lefebvre R, Pepin A. Quadriceps femoris muscle activity in patellofemoral pain syndrome. Am J Sports Med 1992;20:527-32. [30] Cowan DN, Jones BH, Frykman PN. et al. Lower limb morphology and risk of overuse injury among male infantry trainees. Med Sci Sports Exert 1996;28:945-52. [31] Hsu RWW, Himeno S, Coventry MB, Chao EYS. Normal axial alignment of the lower extremity and load-bearing distribution at the knee. Clin Orthop 1990;255:215-27. [32] Messier SF,Davies AB, Moore DT, Davis SE, Pack RJ, Kazmar SC. Severe obesity:effects on foot mechanics during walking. Foot Ankle 1994;15:29-34. [33] Bloedel PK, Hauger B. The effects of limb length discrepancy on subtalar joint kinematics during running. JOSPT 1995;22:60-4. [34] Kujala UM, Friberg 0, Aalto T, Kvist M, Osterman K. Lower limb asymmetry and patellofemoral joint incongruence in the etiology of knee exertion injuries in athletes. Int J Sports Med 1987;8:214-20. [35] Shambaugh JP, Klein A, Herbert JH. Structural measures as predictors of injury in basketball players. Med Sci Sports Exert 1991;23:522-7. [36] Messier SP, Davis SE, Curl WW, Lowery RB, Pack RJ. Etiologic factors associated with patellofemoral pain in runners. Med Sci Sports Exert 1991;23:1008-15. [37] Skalley TC, Terry GC, Teitge RA. The quantitative measurement of normal passive medial and lateral patellar motion limits. Am J Sports Med 1993;21:728-32. [38] Roy S, Irvin R. Sports medicine:prevention, evaluation, management, and rehabilitation. Englewood Cliffs, NJ:Prentice-Hall Inc., 1983:317. [39] Shrout PE, Fleiss JL. Intraclass correlations:uses in assessing rater reliability. Psych Bull 1979;86:420-8. [40] Fleiss JL. Reliability of measurement. New York, NY:Wiley and Sons, 1986:1-32. [41] Triola M. Elementary statistics. Don Mills, Ont:AddisonWesley, 1992:37. [42] Bouffard M. The perils of averaging data in adapted physical activity research. Adapted Phys Activity Q 1993;10:371-91. [43] Beck JL, Wildermuth BP. The female athletes knee. Clinics Sports Med 1985;4:345-66. [44] Biedert RM, Gruhl C. Axial computed tomography of the patellofemoral joint with and without quadriceps contraction. Arch Orthop Trauma Surg 1997;116:77-82. [45] Ruff CB. Biomechanical analyses of archaeological human skeletal samples. 1n:Saunders SR, Katzenberg MA, editors. Skeletal biology of past peoples:research methods. New York, NY:Wiley-Liss, Inc, 1992:37-58. [46] Ruff CB, Jones HH. Bilateral asymmetry in cortical bone of the humerus and tibia - sex and age factors. Human Biol 1981;53:69-86. [47] Schaeffer AA. Spiral movement in man. J Morph Physiol 1928;45:293-398.