The Knee 21 (2014) 529–533

Contents lists available at ScienceDirect

The Knee

Clinical relevance

Gender analysis of the anterior femoral condyle geometry of the knee

Pingyue Li a,b, Tsung-Yuan Tsai a, Jing-Sheng Li a, Shaobai Wang a, Yu Zhang b, Young-Min Kwon a,Harry E. Rubash a, Guoan Li a,⁎a Bioengineering Laboratory, Department of Orthopaedic Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114, USAb Department of Orthopaedic Surgery, Guangzhou General Hospital of Guangzhou Military Command, Guangzhou, Guangdong 510010, China

⁎ Corresponding author at: Bioengineering LaboratorSurgery, Massachusetts General Hospital/Harvard MedicaBoston, MA 02114, USA. Tel.: +1 617 726 6472; fax: +1

E-mail address: gli1@partners.org (G. Li).

0968-0160/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.knee.2013.12.001

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 23 May 2013Received in revised form 12 November 2013Accepted 18 December 2013

Keywords:Anterior femoral condyleTKAGender-specific knee

Background: No study has used 3-D anatomic knee models to investigate the gender differences in anterior fem-oral condyles. Therefore, this study aims to determine the morphologic differences between genders in anteriorfemoral condyles of the knees using 3-D anatomic knee models.Methods:Ninety-six male and sixty-five female 3D anatomic kneemodels were used tomeasure lateral andme-dial anterior condyle heights, anterior trochlear groove heights, and anterior condyle width, whichwere normal-ized by the anterior–posterior and medial–lateral dimensions of the knee, respectively. The shape of anteriorcondyle groove was also analyzed.Results: The mean lateral anterior condyle height, medial anterior condyle height and anterior condyle width of

females were 6.6 ± 1.8 mm, 2.0 ± 2.3 mm, and 44.7 ± 4.2 mm, respectively. These data were significantlysmaller (p b 0.05) than those of males (7.7 ± 1.8 mm, 2.9 ± 2.0 mm and 50.0 ± 3.4 mm). However, after nor-malizing by the femur size, the aspect ratios had no gender differences. Both the ranges of lateral andmedial con-dyle of females were significantly smaller than those of males, and the geometry curve of anterior condyle wasdifferent between genders.Conclusion: Although the gender differences in anterior femoral condyle sizes no longer existed afternormalization with the femur size, the shape and the peak position of anterior condyle groove still havegender differences. The data may have important implications on the current debate of gender-specificTKAs.Clinical relevance: This study provides a better understanding of gender differences in anterior femoralcondyle geometry.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Anatomic variations between the knees of men and women havebeen reported in many studies [1–3]. Some studies reported that the fe-male knee has a less pronounced anterior femoral condyle height thanthe male knee [4,5], leading to the development of gender-specificknee implant designs that have a thinner and narrower anterior flangeto avoid the patellofemoral joint being overstuffed in female knees [6].

However, others arguedwhether there is a gender difference in ante-rior femoral condyles of the knee. Fehring et al. [7] analyzed magneticresonance images of the knee and found that there is no significant dif-ference between men and women when the anterior condyle heightswere normalized by the femoral medial–lateral dimensions. Brattstroem[8] performed a radiographic-anatomic study and found that womenhad lower lateral and medial anterior condyles than men, but that wasattributed to the smaller size of the women. A gender-specific anteriorflange design for femur component was therefore questioned in recent

y, Department of Orthopaedicl School, 55 Fruit St., GRJ-1215,617 724 4392.

ghts reserved.

literature [7,9,10]. Merchant et al. reviewed the evidence presentedto support that females have less prominence of the anterior medialand lateral condyles and concluded that these differences do notexist [9]. Greene et al. also concluded that there was little evidenceto support that gender-specific changes in TKA components lead tobetter outcomes and suggested further research to assess thegender-specific designs [10].

A literature review indicated that most recent studies used plane ra-diographic orMRI images to study the gender differences of anterior fem-oral condyle morphology [4,7,8]. No study has used 3-D anatomic kneemodels to investigate the gender differences in anterior femoral condyles.Therefore, the aim of this study is to determine the gender morpho-logic difference in anterior femoral condyles of the knees using 3-Danatomic knee models. The heights of the lateral anterior femoralcondyle, medial anterior femoral condyle, anterior trochlear groove,and the width of anterior condyle were measured in both male andfemale knee models. These parameters were then normalized bythe anterior–posterior and medial–lateral dimensions of the kneeto calculate the aspect ratios of the anterior condyle and trochleargroove heights as well as the anterior condyle width. We hypothe-size that there is a gender difference in anterior femoral condylemorphology between the male and female knees.

530 P. Li et al. / The Knee 21 (2014) 529–533

2. Materials and methods

2.1. Subjects

We analyzed 161 Caucasian knees (96 males and 65 females) thatwere collected from our previous studies with approval of institutionalreview board. The knees were examined with no anatomical abnor-mality. The males were 38.3 ± 10.2 years old and the females were35.5 ± 10.1 years old. The average height and weight were179.8 ± 6.3 cm and 80.9 ± 16.3 kg for males, and 166.3 ± 4.8 cmand 68.3 ± 9.5 kg for females, respectively.

2.2. Three-dimensional knee models

Among these knees, 34were scanned using a computer tomographic(CT) scanner (GEMedical System,Milwaukee,WI). The imageswere ac-quired along the axial direction between themid shafts of the femur andtibia and with a slice thickness of 0.625 mm and an in-plane resolutionof 0.3 mm × 0.3 mm. The rest 127 knees were MRI scanned using a3.0 Tesla magnet (Siemens, Erlangen, Germany) with a fat suppressed3D spoiled gradient-recalled sequence. Parallel sagittal plane imagesof 1 mm in thickness with no space between were obtained with afield of view of 180 mm × 180 mm and an in-plane resolution of512 × 512 pixels.

The CT images of the kneeswere segmented using a region-growingmethod, and the MRI images of the knees were segmented using amanual segmentation method [11] to construct 3D bony modelsof the knee. Each segmented bony contours were reviewed by an expe-rienced surgeon and corrected on slice-by-slice basis. A 3D surfacemodel of the distal femur was reconstructed using the contours of theknee. Although two imaging techniques were used to compare the geo-metric characteristics, a 3Ddeviation analysis betweenMRI and CT fem-oral models of one subject revealed that the average ± standarddeviation of the distance in between was −0.39 ± 0.64 mm for thefemur. A compatible accuracy of MRI models with respect to CTmodelswas also reported [12].

2.3. Measurement of anterior condyle geometry

To measure the anterior femoral condyle geometry and to simulateTKA anterior condyle cuts, a trans-epicondylar axis (TEA) was definedon each knee model [13]. An orthogonal coordinate system was then

LAC

Z

Y

7 mm

ACW X

A B

6.0°

Fig. 1. The femoral model and bony cuts to simulate a standard TKA procedure. A 2 mm cartila(LACH), medial anterior condyle height (MACH), anterior trochlear groove height (ATGH), antdial–lateral dimension (fML)were shown. (A) The anterior femoral condyle cut viewed in coronplane.

established at the lateral end point of the TEA on the surface model(Fig. 1). The X axis was defined as the TEA; the Y axis was defined as aline parallel with the femur mechanical axis which is 6° valgus withthe anatomical axis (the line along the femoral shaft) in the coronalplane [14]; and the Z axis was vertical to the X and Y axes, passingthrough the lateral end point of the TEA (Fig. 1). A customized scriptusingMATLAB (TheMathworks Inc., Natick, MA) was developed to cre-ate cutting planes through the distal femur [15]. The cutting planes ro-tated around the femoral TEA in 1° increments from −45° to 30°(Fig. 2A). On each cuttingplane, themiddle lowest pointwas used to de-fine the height of the anterior trochlear groove; the medial and lateralhighest points were selected to define the heights of themedial and lat-eral anterior condyles (Fig. 2B).

The position of each selected point could be determined using its co-ordinate values (x, y, z). The z-value of each point is vertical to the fem-oral shaft and defined as the height of the point: the anterior lateralcondyle (H1), anterior medial condyle (H2) and anterior trochleargroove height (H3) (Fig. 2B). An anterior condyle flush cut [16] whichwas along the anterior femur cortex surface, parallel to the femoralshaft and without anterior femoral notching was performed on eachfemur (Fig. 1A). The distance between the TEA and the anterior cuttingplane wasmeasured as H0 (Fig. 2A). Therefore, the lateral anterior con-dyle height (LACh) with respect to the anterior cortex was obtained asH1–H0, medial anterior condyle height (MACh) as H2–H0 and anteriortrochlear groove height (ATGh) as H3–H0 (Fig. 2B). A positive value in-dicates that the point is higher than the anterior femur cortex, and anegative value means that the point is lower than the anterior femurcortex.

2.4. Anterior and distal condyle resection

A distal femur cut of 7 mm (considering a 2 mm thickness of carti-lage [17]) above the lowest point of the medial femoral condyle wasmade on each femur (Fig. 1A). The femoral medial–lateral dimension(fML) was measured on the distal femoral cut surface in medial–lateraldirection that is parallel to the TEA in the transverse plane. The femoralanterior–posterior dimension (fAP) was defined as the distance be-tween the lowest lateral condyle with the lateral anterior cortex of thedistal femur (Fig. 1B). Thewidth of anterior condyle (ACW)was definedas the largest width of the resected anterior condyle which was parallelto the distal cut (Fig. 1A).

MACHH

ATGH

fAPfML X

TEA

ge thickness was considered in the resection. Definitions of lateral anterior condyle heighterior condyle width (ACW), femoral anterior–posterior dimension (fAP) and femoral me-al plane; and (B) the anterior femoral condyle cut anddistal femur cut viewed in transverse

Fig. 2. (A) The anterior femoral condyle was analyzed the cross sections through the transepicondylar line. (B) Definition of lateral andmedial condyle heights and trochlear groove depthon each cutting plane.

531P. Li et al. / The Knee 21 (2014) 529–533

2.5. Data analysis

In order to determine whether the femur size has an effect on thesizes of the anterior femoral condyles, the anterior condyle geometrywas normalized using the dimensions of the knee [7]. We calculatedthe aspect ratio (R) by dividing the anterior condylar heights using thefAP and noted as LACh-R, MACh-R, ATGh-R, respectively, and dividingthe anterior condylar width (ACW) using the fML and noted as ACW-R.

Using the LACh-R, MACh-R and ATGh-R values on cutting planesaround the TEA (Fig. 2A), we obtained LACh-R, MACh-R and ATGh-Rcurves which represent the geometric shapes of the anterior femoralcondyle (Fig. 3). We then determined the ranges of the anterior lateralandmedial condyles and the positions of the highest points of the later-al, medial condyles and anterior trochlear groove from these curves. Thegeometric curves of the lateral and medial condyles only contain thepoints with values of LACh-R or MACh-R N0, because we defined themedial or lateral condyle as the points that are higher than the anteriorfemur cortex (Fig. 1). Since most of the trochlear groove points werelower than anterior femur cortex, we defined the range of the anteriorgroove the same as the anterior lateral condyle (Fig. 3). We definedthe peak values of LACh, MACh and ATGh of each distal femur asLACH, MACH and ATGH.

2.6. Statistical analysis

An independent student-t test [7] was used to determine if themor-phological measurements of the anterior femoral condyle, including

-0.12

-0.08

-0.04

0

0.04

0.08

0.12

0.16

0.2

-45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30

Asp

ect

Rat

io (

1/fA

P)

Rotate Angle of Cut Plane(°)

MaleFemale

LACH-RMACH-RATGH-R

LACh-RMACh-R

ATGh-R Anterior cut line

Anterior

Posterior

Fig. 3. Lateral anterior condyle height (LACH),medial anterior condyle height (MACH) andanterior trochlear groove depth height (ATGH) on the distal femur of the Caucasianmalesand females. LACH, MACH and ATGH are the peak values of the LACh, MACh and ATGh. Rmeans normalization by the femoral anterior–posterior dimension (fAP).

MACH (MACH-R), LACH (LACH-R), ATGH (ATGH-R) and ACW (ACW-R),were statistically different between the genders. Based on the mean andstandard deviation of the knee width and depth of recruited subjects, apriori power analysis (α = 0.05) indicated that 28 subjects in total willhave more than 95% power to detect the differences in the knee. In thisstudy, 96 male and 65 female knee models were recruited for measure-ment of the anterior femoral condylar geometry. The sample size is suffi-cient to detect a difference in anatomy.

3. Results

3.1. Anterior condyle geometry curve

The fAP of the males is 59.2 ± 3.4 mm that was significantly larger than that of thefemales (54.6 ± 3.0 mm). The fML of the males was 75.1 ± 4.5 mm that was also signif-icantly larger than that of the females (65.7 ± 2.8 mm) (Table 1).

The range of the cutting angles through the anterior condyle curve revealed that ingeneral, there was a gender difference in the lateral and medial condyles (Table 1). Theranges of the cutting angles of the lateral condyle were 67.7° ± 4.3° and 64.1° ± 6.3°for the males and females, respectively (p b 0.05). The ranges of the medial condylewere 38.2° ± 14.9° and 32.3° ± 20.1°, respectively for males and females (p b 0.05).The ranges of the anterior trochlear groove were the same with the lateral condyle forboth males and females (Table 1). In general, the anterior groove curves were under theanterior femur cortex in all females and most males. The range where the ATGH valuesof males were positive was only from −15° to−4° (Fig. 3).

The position of the LACH (LACH-R) was −1.3° ± 4.1° and 0.38° ± 4.7° formales and females respectively and has no gender difference (p N 0.05). The posi-tions of the MACH (MACH-R) were 2.7° ± 5.8°,4.6° ± 6.6° and of ATGH (ATGH-R)were −9.3° ± 5.1°, −4.4° ± 5.1° for males and females, respectively (p b 0.05)(Table 1).

3.2. Anterior condyle parameters

The LACH, MACH and ACW values of females were 6.6 ± 1.8 mm, 2.0 ± 2.3 mm,44.7 ± 4.2 mm, respectively, and were significantly smaller (P b .05) than those ofmales (7.7 ± 1.8 mm, 2.9 ± 2.0 mm, 50.0 ± 3.4 mm). The ATGH dimension of thefemales (−0.2 ± 2.2 mm) has no significant difference with that of males(0.4 ± 1.7 mm) (Table 2).

3.3. Aspect ratio

After normalized by the femur size, the aspect ratios of the lateral andmedial condyleheights (LACH-R, MACH-R) of females (0.121 ± 0.035, 0.037 ± 0.043) have no signifi-cant difference with those of males (0.131 ± 0.033, 0.049 ± 0.034). The aspect ratios ofthe trochlear groove height (ATGH-R) were −0.002 ± 0.041 and 0.006 ± 0.029 for thefemales and males, respectively (p N 0.05). The aspect ratios of the condyle width,ACW-R, were 0.665 ± 0.033 and 0.673 ± 0.047 for the females and males, respectively(p N 0.05) (Table 2).

4. Discussion

This paper presented a comprehensive comparison of the 3D geo-metrical parameters of the anterior femoral condyles of normal maleand female knees. The data indicated that the female knees have

Table 1Average± standard deviation (minimum–maximum) of knee sizes, the range and peak of the cutting angles through the lateral anterior condyle heights (LACh-), medial anterior condyleheight (MACh-) and anterior trochlear groove height ratio (ATGh-R).

LACh-R curve MACh-R curve ATGh-R curve Femur parameter

Range LACH-R position Range MACH-R position Range ATGH-R position fAP (mm) fML (mm)

Male 67.7° ± 4.3° −1.3° ± 4.1° 38.2° ± 14.9° 2.7° ± 5.8° 67.7° ± 4.3° −9.3° ± 5.1 59.2 ± 3.4 75.1 ± 4.5(55–76°) (−9–11°) (0–76°) (−10–16°) (55–76°) (−24–2°) (51.8–65.6) (64.1–83.9)

Female 64.1° ± 6.3°a 0.38° ± 4.7° 32.3° ± 20.1°a 4.6° ± 6.6°a 64.1° ± 6.3°a 4.4° ± 5.1°a 54.6 ± 3.0a 65.7 ± 2.8a

(50–76°) (−13–11°) (0–73°) (−9–19°) (50–76°) (−15–5°) (49.1–60.6) (60.5–75.4)

a Statistic difference with male.

532 P. Li et al. / The Knee 21 (2014) 529–533

significantly smaller anterior femoral condyle heights andwidth (LACH,MACH and ACW) than the male knees, but, the normalized valuesshowed no significant gender differences. The trochlear groove heightwas found close to or slightly below the anterior femoral cortex andno significant gender difference was found in both ATGH and ATGH-Rvalues.

Several studies had reported the gender differences of the anteriorfemoral condyle using different methods (Table 3). Poilvache et al. [4]reported a significant gender difference in anterior condyle heights(13.74 mm laterally and 10.63 mm medially for males, and 12.26 mmlaterally and 8.96 mm medially for females). These values were largerthan ours because they used the lowest point of the anterior trochleargroove to measure the lateral and medial condyle heights and ourdata also did not include the cartilage thickness (~2.75 mm) [17]. Thedata of Zimmer gender knee solutions [6] used the anterior condyleheights of 10.9 mm laterally and 6.4 mm medially for males, and10.1 mm laterally and 5.1 mm medially for females. The trend ofthese data was similar to our data too (Table 3). These studies did notreport the aspect ratios of the anterior femoral condyle heights. Fehringet al. [7] analyzed transverse planeMR Images of the knee and reportedan average anterior condyle height of 7.3 mm laterally and 5.7 mmme-dially for males, and 6.8 mm laterally and 4.6 mmmedially for females.They indicated that after normalizing the data by medial–lateral (fML)size of the femur, both lateral and medial aspect ratios are not signifi-cantly different between the males and females, similar to our observa-tions. The different values of theirmeasurements and our resultsmay becaused by the measurement methods as they measured the anteriorcondyle heights using MR images in a transverse plane of the kneethat was 10 mm above the lowest trochlear groove point while ourmeasurement used 3D models and simulated TKA cuttings. Kwak [13]reported an average ACW of 51.2 mm and 44.3 mm for males and fe-males respectively of a Korean population,whichwere similar to our re-sults. However, no data has been reported on the ACW of Caucasianpopulation.

The data reported in literature as well as obtained from our studyshowed that the LACH and MACH are significantly lower in femalesthan in males [4,6–8]. However, the aspect ratios, LACH-R and MACH-R, are similar between genders, indicating similar shape in anterior fem-oral condyle heights between the males and females [7]. Therefore, tobetter fit the anterior femoral condyles of female patients, lower anteri-or femoral condyle heights of TKA componentsmay be necessary, as ad-vocated by the gender knee concepts [5,6]. However, these loweranterior femoral condyle heights for female knees could be achievedusing smaller sized components of contemporary TKAs if the

Table 2Morphological data of the anterior femoral condyle (mean ± SD).

LACH(mm) LACH-R MACH(mm) MACH-R

Male 7.7 ± 1.8 0.131 ± 0.033 2.9 ± 2.0 0.049 ± 0.034(Range) (2.9–12.1) (0.045–0.210) (−1.4–8.2) (−0.026–0.147Female 6.6 ± 1.8a 0.121 ± 0.035 2.0 ± 2.3a 0.037 ± 0.043(Range) (3.2–12.3) (0.063–0.239) (−2.5–7.9) (−0.043–0.152

a Means statistic difference with male.

components have the same aspect ratio in anterior femoral condylessince the males and females have similar aspect ratios of anterior femo-ral condyle heights.

Our data indicated that the females have significantly smaller rangeof lateral anterior and medial anterior condyle heights than males.Therefore, the females may have anterior flange overhang when usingcontemporary TKA components. Results also showed that the femaleshave shorter ACW than males. However, the ACW-R was similar be-tween themales and females, femalesmaynot need a narrower anteriorflange as proposed by the gender-knee design [6]. Smaller sized compo-nents may well fit the female knees in the anterior condyles if differentsize components have similar aspect ratios.

Our study revealed an interesting phenomenon in the anteriortrochlear groove heights. A higher groove can cause “overstuffing” ofthe knee anteriorly [12,18–20]. Most recent studies focused on the ori-entations of the trochlear groove [3,21]. Few data has been reportedon the trochlear groove. Our results showed that males and femaleshave similar ATGH and ATGH-R values. The cartilage thickness was as-sumed as 2 mm [17] during the bony resection. Although the baselinefor measurements of LACH, MACH and ATGH would be changed withdifferent cartilage thickness, it doesn't affect the curves and peak pointson the anterior condyle heights. The ATGH was almost on or slightlybelow the anterior femoral cortex. The ATGH of the females wasunder the anterior femur cortex in the whole range (Fig. 3), whichmeans that the patients may be overstuffed using contemporary TKAimplants since the anterior femoral condyles of the TKAs are all abovethe anterior femoral cortex of the knee. The ATGH positions of themales were significantly more proximal than the females (Table 1,Fig. 3). Therefore, as the results showed in this study, the shape of thetrochlear groovemay need to be designed gender-specific andmore an-atomic in TKA components if the anatomic trochlear groove is to bereproduced.

There are some limitations in this study. We analyzed the bony sur-face models of the knee, so we did not include the cartilage thickness inthe analysis. Both CT and MRI based models were analyzed since thesemodels were obtained from our previous investigations. Although twodifferent imaging techniques were used, the small differences resultedfrom these techniques have been confirmed [12]. Due to the concernof radiation of CT scan and the limitation of feasible region of MRI,only distal femur models were obtained. The distal femoral cut of7 mm above the lowest point of the medial femoral condyle wasmade along the estimatedmechanical axis whichwas 6° valguswith re-spect to the anatomical axis based on the clinical observations of previ-ous studies [22,23]. The distal femoral cut may be slightly changed if

ATGH(mm) ATGH-R ACW(mm) ACW-R

0.4 ± 1.7 0.006 ± 0.029 50.0 ± 3.4 0.665 ± 0.033) (−3.0–5.8) (−0.055–0.100) (42.1–59.9) (0.594–0.743)

−0.2 ± 2.2 −0.002 ± 0.041 44.7 ± 4.2a 0.673 ± 0.047) (−5.1–6.2) (−0.087–0.111) (35.5–51.1) (0.587–0.763)

Table 3Comparison of the morphological data with those reported in literature.

Methods Cartilage(Y/N)

LACH(mm) MACH(mm) Normalize(Y/N)

LACH sex difference (mm) MACH sex difference (mm) ACW(mm)

Poilvache PL [4] Radiologic N 13.74 ± 2.76 (M) 10.63 ± 2.48 (M) N 1.48 1.67 –

(American) 12.26 ± 2.26 (F) 8.96 ± 2.56 (F)Fehring TK [7] MRI slice Y 7.3 ± 3.0 (M) 5.7 ± 2.4 (M) Y 0.5 1.1 –

(American) 6.8 ± 3.1 (F) 4.6 ± 2.6(F)Brattstrom[8] Radiologic N – – N 1.5 1.1 –

– – –

ZIMMER [6] – – 10.9 (M) 6.4 (M) N 0.8 1.3 –

– 10.1 (F) 5.1 (F)Kwak DS [13] CT N – – N – – 51.2 ± 3.2(Korean) – – 44.3 ± 3.1Our study(Caucasian)

3-D model N 7.7 ± 1.8(M)6.6 ± 1.8(FM)

2.9 ± 2.0(M)2.0 ± 2.3(FM)

Y 1.1 0.9 50.0 ± 3.444.6 ± 4.2

M (male), FM (female), Y (YES), N (NO).

533P. Li et al. / The Knee 21 (2014) 529–533

there is an individual difference between the femoral mechanical axisand anatomical axis. However, this may only affect the measures ofthe fML and the corresponding ratio normalized using fML. The mea-sures based on anterior condyle cuts (i.e. fAP, LACH, MACH and ATGH)would not be changed. In future investigation, we will keep addingmore knee models into the analysis and to compare our data with con-temporary total knee replacement components. In addition, we are alsointerested in investigation of race effect on the anterior femoral condylegeometries.

5. Conclusion

Our study found that the gender differences in anterior femoral con-dyle sizes (including LACH, MACH, ATGH and ACW) disappeared whennormalized by the anterior–posterior and medial–lateral dimensions ofthe knee. However, the ATGH and MACH positions of the males weresignificantly more proximal than the females. Although the anteriorgroove height and aspect ratio have no significant gender differences,the shape and the peak positions of anterior condyle groove have gen-der differences. The data may have important implications on the cur-rent debating on gender-specific TKAs.

6. Conflict of interest

The authors of this manuscript have nothing to disclose that wouldbias our work.

References

[1] Chin PL, Tey TT, Ibrahim MY, Chia SL, Yeo SJ, Lo NN. Intraoperative morphometricstudy of gender differences in Asian femurs. J Arthroplast 2011;26:984–8.

[2] Clarke HD, Hentz JG. Restoration of femoral anatomy in TKA with unisex andgender-specific components. Clin Orthop 2008;466:2711–6.

[3] Yue B, Varadarajan KM, Ai S, Tang T, Rubash HE, Li G. Gender differences in the kneesof Chinese population. Knee Surg Sports Traumatol Arthrosc 2011;19:80–8.

[4] Poilvache PL, Insall JN, Scuderi GR, Font-Rodriguez DE. Rotational landmarks andsizing of the distal femur in total knee arthroplasty. Clin Orthop 1996:35–46.

[5] Conley S, Rosenberg A, Crowninshield R. The female knee: anatomic variations. J AmAcad Orthop Surg 2007;15(Suppl. 1):S31–6.

[6] Zimmer. Zimmer Gender Solutions Knee. http://genderkneecom/micro/ctl?op=global&action=1&id=9508&global=1&template=mn; Jan, 3, 2013.

[7] Fehring TK, Odum SM, Hughes J, Springer BD, Beaver JrWB. Differences between thesexes in the anatomy of the anterior condyle of the knee. J Bone Joint Surg Am2009;91:2335–41.

[8] Brattstroem H. Shape of the intercondylar groove normally and in recurrent disloca-tion of patella. A clinical and x-ray-anatomical investigation. Acta Orthop ScandSuppl 1964;68(Suppl. 68):1–148.

[9] Merchant AC, Arendt EA, Dye SF, Fredericson M, Grelsamer RP, Leadbetter WB, et al.The female knee: anatomic variations and the female-specific total knee design. ClinOrthop 2008;466:3059–65.

[10] Greene KA. Gender-specific design in total knee arthroplasty. J Arthroplast2007;22:27–31.

[11] Yue B, Varadarajan KM, Ai S, Tang T, Rubash HE, Li G. Differences of kneeanthropometry between Chinese and white men and women. J Arthroplast2011;26:124–30.

[12] Dejour D, Ntagiopoulos PG, Saffarini M. Evidence of trochlear dysplasia in femoralcomponent designs. Knee Surg Sports Traumatol Arthrosc 2012 Epub ahead ofprint, http://www.ncbi.nlm.nih.gov/pubmed/23229383.

[13] Kwak DS, Han S, Han CW, Han SH. Resected femoral anthropometry for design of thefemoral component of the total knee prosthesis in a Korean population. Anat CellBiol 2010;43:252–9.

[14] Cheng FB, Ji XF, Lai Y, Feng JC, ZhengWX, Sun YF, et al. Three dimensional morphom-etry of the knee to design the total knee arthroplasty for Chinese population. Knee2009;16:341–7.

[15] Varadarajan KM, Rubash HE, Li G. Are current total knee arthroplasty implants de-signed to restore normal trochlear groove anatomy? J Arthroplast 2011;26:274–81.

[16] Lonner JH, Jasko JG, Thomas BS. Anthropomorphic differences between the distalfemora of men and women. Clin Orthop 2008;466:2724–9.

[17] Connolly A, FitzPatrick D, Moulton J, Lee J, Lerner A. Tibiofemoral cartilage thicknessdistribution and its correlation with anthropometric variables. Proc Inst Mech Eng HJ Eng Med 2008;222:29–39.

[18] BongMR, Di Cesare PE. Stiffness after total knee arthroplasty. J Am Acad Orthop Surg2004;12:164–71.

[19] Scranton Jr PE. Management of knee pain and stiffness after total knee arthroplasty. JArthroplast 2001;16:428–35.

[20] Barink M, Van de Groes S, Verdonschot N, DeWaal Malefijt M. The difference in troch-lear orientation between the natural knee and current prosthetic knee designs; towardsa truly physiological prosthetic groove orientation. J Biomech 2006;39:1708–15.

[21] Iranpour F, Merican AM, Dandachli W, Amis AA, Cobb JP. The geometry of the troch-lear groove. Clin Orthop 2010;468:782–8.

[22] Brooks P. Seven cuts to the perfect total knee. Orthopedics 2009:32.[23] Kharwadkar N, Kent RE, Sharara KH, Naique S. 5 degrees to 6 degrees of distal fem-

oral cut for uncomplicated primary total knee arthroplasty: is it safe? Knee2006;13:57–60.