Anatomical variations in Asian knees: one size does not fit all

Asian knees differ from Caucasian knees in many ways. For example, size, shape, thickness, angle and growth rates. How do anatomical differences influence a surgeon's approach to TKR in these populations?

TKR requires precision. Accurate bone cutting, soft tissue balance and adequate resected surface coverage can all influence TKR success. This is why a number of practitioners [1] have started to question the use of prosthetics designed for Caucasian knees with Asian patients. Ethnicity, it seems, creates very different types of knees that may require equally variable treatment.

Part 1 in this series of articles on TKR in the Asian knee looked at cultural influences upon usage as well as patient expectations. Here we will examine the inherent anatomical differences of Asian TKR patients.


Different from the start

Knee alignment and growth rates can vary widely across regions. For example, the majority varus (inward angulation) TFA (*) observed [2] in Chinese children as young as 2 years old differs markedly from the reportedly majority valgus (outward) alignment of Indian, Korean, Nigerian, and white children of the same age [3].

At the same time, Turkish children have even more pronounced valgus TFA, so that up to 11 degrees physiological valgus is considered within the normal range for Turkish children between 3 and 17 years of age [4].

Also, the mean IMD of Chinese children has been shown to decrease after 3 years of age, returning to zero at roughly 8 years of age. White children, meanwhile, were found to be maximally bow-legged at 6 months, returning on average to neutral by 18 months. Bowlegs are considered unusual amongst white children after 2 years of age. Whereas in Korea a similar knee angle development process was found to occur at a different pace, so that the varus alignment of Korean children neutralized at 1.5 years of age and valgus alignment peaked at 4 years of age [5].

In Nigeria, the distribution of knee angles becomes bi-modal after 6 months of age, with half varus and half valgus, returning to majority valgus after two years of age [6]. Nigerian children became maximally and uniformally knock-kneed between 3 and 3.5 years of age [6], while Indian children reached maximum valgus between 5 and 6 years of age [3].

Body weight does not appear to significantly influence these variable growth rates [4], so it is likely that they reflect the ethnical and racial differences between regions [1]. 



The fact that knees are so innately variable has important implications for TKR procedures. Most of the currently available knee prosthetics are generally designed for the Caucasian knee. In studies being undertaken to assess the impact of this historic, one size fits all approach it appears that the success of TKR in Asia has been compromised by what is, effectively, a component mismatch. Japanese patients, for example, have significantly less postoperative ROM than white patients [7]. At least 4.1% have required revision within 7 years, whilst only 2.6% of American patients required similar work within 9 years [7].

Even the smallest Caucasian specific prosthetics might be too large for the Asian knee and size is not the only factor that needs to be taken in to account.

In overview, differences between Caucasian and Asian knees include (1) the asymmetrical tibial anatomical axis of Chinese, Korean, and Iranian knees (requiring a change to the tibial entry point), (2) the variable rate of change of the femoral aspect ratio and tibial aspect ratios of Chinese, Japanese, Korean, and Indian patients, (3) significant differences in angle parameters including the distal femoral condylar angle and the posterior femoral condylar angle, and (4) structurally distinct patella.


1) Tibial placement

There is strong evidence to suggest that the center of the tibial plateau should not be used as a landmark of the tibial component [1: 59]. Studies show that Chinese [8], Korean [9] and Iranian [1] patients tend to have a more asymmetrical tibial plateau, so that the central point (Cp) runs medial to the central shaft line (Cs). This requires that the tibial entry point be adjusted, because otherwise the tibial component is likely to be inserted in varus which can cause the tibial medial cortex to fracture.

In other words, tibial base-plates designed for more symmetrical, Caucasian knees may not always be suitable.


2) Aspect ratios

Femoral aspect ratio

Also, the risk of component oversizing is increased if related aspect ratios are not taken in to consideration. The smaller distal femoral diameter of Chinese [10], Japanese [11], Korean [12] and Indian [13] populations compared with their Caucasian counterparts also creates a significant disparity amongst related parameters. The femoral aspect ratio, for example, increases in the context of these smaller dimensions. In general this ratio will be higher for smaller knees and lower for larger knees [1].

Tibial aspect ratio

Similarly, the tibial aspect ratio (tML/tAP %) is negatively correlated with tAP, which means that the smaller the tAP, the larger the aspect.

In a Chinese study [10] of relationships between the tibial mediolateral (tML), the femoral mediolateral (fML) dimension, the tibial mediolateral (tML) and the femoral anteroposterior (fAP) measurements it was found that the fML and fAP were, likewise, strongly correlated with the tML. As the tML increased, the fML and fAP also increased [1].

In Korea, where tibial dimensions appear to correlate to height and smaller AP dimensions again reveal higher aspect ratios, every increase of the AP dimension of the proximal tibia thus means that the anticipated tibial component becomes less oval [9, 12].


3) The angles

The rotational configuration of the Asian distal femur differs from its Caucasian counterpart. The inferolateral angle between the knee joint surface and the mechanical axis of the tibia of both Chinese women and men is more oblique than that of Caucasians, creating a rectangular flexion gap.

These differences extend to the tibial varus angle, the posterior tibial slope [10, 14] and the mechanical versus anatomical axes of the Whiteside-epicondylar angle [15] in the Chinese population.

Similarly in Japan, where 78% of patients have exhibited a varus alignment, torsion angles are generally lower than Caucasian subjects and ACL laxity is generally higher [11]. 

Failing to take these differences in to account when preparing and cutting the distal femur and proximal tibia can mean that “soft tissue tension, ligament balancing and the ROM of the joint may be disturbed” [1: 58]


The whiteside-epicondylar (a) and posterior condylar (b) angles.


4) Patellar dimensions

Patellar dimensions change significantly between ethnic groups. Variables such as thickness, height/width ratio, and relative position of the median ridge are important considerations in the selection of prosthetic patellar components [16].

A thin patella can soften the impact of patellofemoral contact, but is also more susceptible to stress fracture and anteroposterior instability. A thick patella, meanwhile, can increase effective quadriceps moment arm at low flexion angles, but can also decrease motion. As a result it is generally believed to be preferable if a resurfaced patella maintains its original thickness. Thus, in the case of thin patella (associated with many Asian knees [17]) specifically designed patellar prosthesis with less thickness is recommended [17, 18].

When selecting the optimal patella size it is also important to consider the position of the median ridge. A well placed median ridge can help to support restoration of normal movement after TKA. Again, it is generally considered more helpful to replace the median ridge in its original position. In order to restore the median ridge of Asian patients – whose patella can be thinner and smaller than Caucasians [17] - orthopedic surgeons may need to select smaller patellar components that also reduce the patellofemoral articulation contact area.

Yin and Yang: gender considerations

Gender differences further complicate regional variations. In India, for example, it has been observed that the mean TFA of boys and girls varies significantly between 3 and 6 years of age when girls become noticeably more valgus, returning to similarity after 6 years [3]. In Turkey meanwhile, girls experience maximum valgus at 6, whereas boys reach their peak valgus a year later.

Naturally, female and male adult bodies retain distinct characteristics. Studies show ongoing, ethnic specific variations between the knees of Thai [19], Korean [12], Japanese [20], and Chinese [21] men and women.


Yin and Yang symbol.


In China, despite their more varus knee alignment [15, 22], the distal femur of Chinese women are also distinctly narrower than white women (so that their knees are both smaller and narrower), whilst Chinese men tend to have a wider proximal tibia than their white counterparts [21].

The size and shape of Chinese female knees differ significantly from white female knees [1]. The fML dimension of Chinese women is generally smaller than that of white women. Similarly, the fML and fAP ratios (**) of the two differ so that those of Chinese women are again much smaller. Despite the petite Chinese tibia, however, there is no significant difference in the tibial aspect ratio of the two. Nor are there any marked differences in their media/lateral posterior tibial slope in either sex [21].

Yet, differences between the tibial aspect ratio of Chinese men and white men are pronounced. The inferolateral angle in both the right and left extremity in Chinese men is significantly larger than that of white men [23]. At the same time their fML and fAP (femoral aspect ratio) dimensions are significantly smaller. It is suspected that, in general, a smaller tAP (tibial) dimension accounts for a larger tibial aspect ratio [1].

Other examples of gender and ethnicity variation include the greater FTA of most Japanese subjects compared to their Australian counterparts [20]. The difference is more marked in Japanese men, however, than Japanese women. Yet, Japanese women enjoy greater femoral antetorsion than men of either ethnicity. Also, despite no significant ethnic differences in tibiofibular torsion, Japanese women experience a noticeable decrease in tibiofibular torsion as they age, whilst Japanese men do not.

Gender also creates similarity as much as difference. It appears that Caucasian women, for example, can achieve knee flexion greater than 150 degrees, on a par with both Japanese women and men. Caucasian men on the other hand trail behind them all. The medial condyles of Caucasian men reveal roll above 120 degrees of flexion. Similarly, in deep flexion they have the least amount of external rotation. [18].


Thai Chi class.


Design recommendations

Judging by the normalized ratios and non-linear shape analysis presented in multiple studies [14, 21, 24] of the Asian knee, researchers have concluded that the sorts of differences canvassed in this article are generally “independent of any scale factor” [1: 35]. Yet still today many Asian patients have prosthetic implants that do not account for these adjustments [2, 10]. Either they are too big, or they employ relatively constant aspect ratios [1] so that tML is either undersized with the smaller tAP, or overhang with the larger tAP – whereas the tibial aspect ratio of a prosthesis designed especially for the Korean population, for example, would decrease with a corresponding increase in the anteroposterior dimension [12].

Often too, the tibial stem is medially offset to suit the anteromedial tibial shaft of Western populations, whereas for TKA in Chinese populations, for example, it is preferable to employ a variable and long-stemmed offset stem that can accommodate a variety of asymmetrical alignments [25].

Just as smaller angles of cutting blocks and shorter intramedullary rods entering the femur can optimize a distal femoral cut depending on the size and shape of the patient [26], so asymmetrical femoral components might also prevent soft tissue irritation [10].


As evidence mounts of the need for more prostheses designed especially for the Asian knee, researchers are calling [1: 43] for customizable prosthetics that approach the tibia and femur as a whole (particularly given the strong correlations between tML and fML) and employ individual tML and fAP as their design criteria. Likewise, it is considered “imperative” [1: 58] by many that patellar prosthesis be specifically designed to accommodate the thinner, Asian patella.

(*) TFA is the angle defined by the mechanical axis of the femur intersecting the
mechanical axis of the tibia 1. Hosseinzadeh, H.R.S., et al., Special Considerations in Asian Knee Arthroplasty. 2013, Chapter.: 36

(**) As opposed to their fAP measurements which appear relatively similar.

Contributing experts

This series of articles was created with the support of the following specialists (in alphabetical order):

Myung Chul Lee MD, PhD

Seoul National University Hospital
Seoul, South Korea

This issue was created by Word+Vision Media Productions, Switzerland


Additional Resources

Additional AO resources

Access videos, tools, and other assets to learn more about this topic.


  1. Hosseinzadeh HRS, Tarabichi S, Shahi AS, et al (2013) Special Considerations in Asian Knee Arthroplasty. Chapter. Intech Creative Commons Licence.
  2. Cheng N, Shi Q (2010) [Rehabilitation exercises after single total knee replacement: a report of 38 cases]. Zhongguo gu shang= China journal of orthopaedics and traumatology. 23(3): p. 220–221.
  3. Saini UC, Bali K, Sheth B, et al (2010) Normal development of the knee angle in healthy Indian children: a clinical study of 215 children. Journal of children's orthopaedics. 4(6): p. 579–586.
  4. Arazi, M, Ögün TC, Memik R (2001) Normal development of the tibiofemoral angle in children: a clinical study of 590 normal subjects from 3 to 17 years of age. Journal of Pediatric Orthopaedics. 21(2): p. 264–267.
  5. Yoo JH, Choi IH, Cho TJ, et al (2008) Development of tibiofemoral angle in Korean children. J Korean Med Sci. 23(4): p. 714–717.
  6. Oginni LM, Badru OS, Sharp CA, et al (2004) Knee angles and rickets in Nigerian children. J Pediatr Orthop. 24(4): p. 403–407.
  7. Iorio R, Kobayashi S, Healy WL, et al (2007) Primary posterior cruciate-retaining total knee arthroplasty: a comparison of American and Japanese cohorts. J Surg Orthop Adv. 16(4): p. 164–170.
  8. Chiu KY, Zhang SD, Zhang GH (2000) Posterior slope of tibial plateau in Chinese. J Arthroplasty. 15(2): p. 224–227.
  9. Yoo JH, Kang YG, Chang CB et al (2008) The relationship of the medially-offset stem of the tibial component to the medial tibial cortex in total knee replacements in Korean patients. J Bone Joint Surg Br. 90(1): p. 31–36.
  10. Cheng FB, Ji XF, Lai Y (2009) Three dimensional morphometry of the knee to design the total knee arthroplasty for Chinese population. Knee. 16(5): p. 341–347.
  11. Hovinga KR, Lerner AL (2009) Anatomic variations between Japanese and Caucasian populations in the healthy young adult knee joint. J Orthop Res. 27(9): p. 1191–1196.
  12. Kwak DS, Surendran S, Pengatteeri YH, et al (2007) Morphometry of the proximal tibia to design the tibial component of total knee arthroplasty for the Korean population. Knee 14(4): p. 295–300.
  13. Vaidya SV, Ranawat CS, Aroojis A, et al (2000) Anthropometric measurements to design total knee prostheses for the Indian population. J Arthroplasty. 15(1): p. 79–85.
  14. Mahfouz M, Abdel Fatah EE, Bowers LS, et al (2012) Three-dimensional morphology of the knee reveals ethnic differences. Clin Orthop Relat Res. 470(1): p. 172–185.
  15. Yip DK, Zhu YH, Chiu KY et al (2004) Distal rotational alignment of the Chinese femur and its relevance in total knee arthroplasty. J Arthroplasty. 19(5): p. 613–619.
  16. Baldwin JL, House CK (2005) Anatomic dimensions of the patella measured during total knee arthroplasty. J Arthroplasty. 20(2): p. 250–257.
  17. Kim TK, Chung BJ, Kang YG, et al (2009) Clinical implications of anthropometric patellar dimensions for TKA in Asians. Clin Orthop Relat Res. 467(4): p. 1007–1014.
  18. Hitt K, Shurman JR 2nd, Greene K, et al (2003) Anthropometric measurements of the human knee: correlation to the sizing of current knee arthroplasty systems. J Bone Joint Surg Am. 85(suppl 4): p. 115–122.
  19. Chaichankul C, Tanavalee A, Itiravivong P (2011) Anthropometric measurements of knee joints in Thai population: correlation to the sizing of current knee prostheses. Knee. 18(1): p. 5–10.
  20. Tamari K, Tinley P, Briffa K et al (2006) Ethnic-, gender-, and age-related differences in femorotibial angle, femoral antetorsion, and tibiofibular torsion: cross-sectional study among healthy Japanese and Australian Caucasians. Clin Anat. 19(1): p. 59–67.
  21. Yue B, Varadarajan KM, Ai S, et al (2011) Gender differences in the knees of Chinese population. Knee Surg Sports Traumatol Arthrosc. Jan;19(1):80–8.
  22. Tang W, Zhu Y, Chiu K (2000) Axial Alignment of the Lower Extremity in Chinese Adults*. J Bone Joint Surg Am. Nov;82-A(11):1603–8.
  23. Moreland JR, Bassett L, Hanker G (1987) Radiographic analysis of the axial alignment of the lower extremity. J Bone Joint Surg Am. Jun;69(5):745–9.
  24. Nelson AE, Braga L, Braga-Baiak A, et al (2009) Static knee alignment measurements among Caucasians and African Americans: The Johnston County Osteoarthritis Project. J Rheumatol. Sep;36(9):1987–90.
  25. Abraham R1, Malkani AL, Lewis J, et al (2007) An anatomical study of tibial metaphyseal/diaphyseal mismatch during revision total knee arthroplasty. The Journal of arthroplasty. J Arthroplasty. Feb;22(2):241–4.
  26. Fang Y, Fu X, Chi L, Wang G, et al (2006) [The biomechanical study of rotating-arm self-locking intramedullary nails in comminuted femoral shaft fractures]. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. Oct;23(5):1041–4
Cookies help us improve your website experience.
By using our website, you agree to our use of cookies.