Search tips
Search criteria 


Logo of corrspringer.comThis journalToc AlertsSubmit OnlineOpen Choice
Clin Orthop Relat Res. 2009 April; 467(4): 1007–1014.
Published online 2008 October 15. doi:  10.1007/s11999-008-0557-0
PMCID: PMC2650045

Clinical Implications of Anthropometric Patellar Dimensions for TKA in Asians

Tae Kyun Kim, MD, PhD,corresponding author1,2 Byung June Chung, MD,1 Yeon Gwi Kang, BS,1 Chong Bum Chang, MD, PhD,1,2 and Sang Cheol Seong, MD, PhD2


Anthropometric patellar dimensions can influence implant design and surgical techniques in patellar resurfacing for TKA. We measured anthropometric patellar dimensions in 752 osteoarthritic knees (713 in females and 39 in males) treated with TKA in 466 Korean patients and compared them with reported dimensions for Western patients. We investigated the effects of postoperative overall thickness deviations, residual bony thickness after bone resection, and postoperative deviations of component center positions from median ridge positions versus clinical and radiographic outcomes evaluated 1 year after surgery. Korean patients undergoing TKA had thinner and smaller patellae than Western patients. We found no associations between preoperative to postoperative overall thickness differences and clinical and radiographic outcomes and no differences between knees with a residual bony thickness 12 mm or greater and knees with a residual thickness less than 12 mm, with the exception of WOMAC pain scores. We found no associations between postoperative deviations of component center position and clinical or radiographic outcomes. Our findings indicate bone resection for patellar resurfacing can be flexible without jeopardizing clinical outcome.

Level of Evidence: Level IV, prognostic study. See the Guidelines for Authors for a complete description of levels of evidence.


Patella-related problems are among the issues of concern after TKA [8, 13, 17, 18]. To prevent complications related to the patella, the use of appropriately designed prostheses and proper surgical techniques are mandatory. Information regarding anthropometric patellar dimensions can play an important role during the design of patellar prostheses and the development of surgical techniques [1, 7]. Thickness, height/width ratio, and relative position of the median ridge all have implications relating to the selection of patellar components, patellofemoral contact stress, and patellar tracking in the trochlear groove [8, 13, 16, 18].

Patellar thickness is a challenging consideration during patellar resurfacing for TKA. A thin patella can reduce patellofemoral contact force but also poses the potential risk of stress fracture and anteroposterior instability [8, 16]. Increasing patellar thickness might be expected to increase effective quadriceps moment arm at low flexion angles of the knee but potentially reduces range of motion and predisposes to patellar subluxation [9, 12]. It commonly is assumed that it is desirable for a resurfaced patella to be equal to its original thickness [1, 6, 8, 13, 16], and a bony patellar thickness of at least 15 mm should be maintained [16]. However, according to our experience with Korean patients, it is not uncommon to find, intraoperatively, the patella is too thin to simultaneously satisfy these criteria. Three surgical options should be considered with a thin patella: (1) leave the patella unresurfaced; (2) restore the original patellar thickness by removing bone to account for the thickness of the prosthesis, while accepting thin residual bone; and (3) leave residual bone of adequate thickness, while accepting an increase in overall patellar thickness. Despite frequently encountered challenging clinical scenarios, there is a paucity of literature regarding the effects of a thin residual patella or an increased overall patellar thickness on the clinical outcomes of patellar resurfacing in TKA [1, 6, 8, 10, 16].

Patellar height/width ratios and the position of the median patellar ridge have clinical implications for TKA with patellar resurfacing. Because the mediolateral width of a patella typically is larger than its height [1, 7], a dome-shaped patella component chosen by the height does not cover the entire resected surface of the patella. It has been recommended the patella be placed at the medial margin of a resected patella to help patellofemoral tracking by decreasing the Q angle [14]. In addition, the position of the median ridge is another factor to consider when selecting the size of a patellar component. The median ridge can act as a fulcrum for patellofemoral tracking and thus can influence restoration of normal kinematics after TKA with patellar resurfacing. If an orthopaedic surgeon attempts to restore the original position of the median ridge, he or she may have to select a smaller component, which reduces patellofemoral articulation contact area. To the best of our knowledge, no previous study has elaborated on the effect of the relative position of the median patellar ridge in the whole mediolateral width on clinical outcomes for TKA with patellar resurfacing [1].

Given this lack of information, we attempted to determine whether the patellar anthropometric features of Korean and Western patients differ. In particular, we sought to determine whether residual patellar thickness influences the clinical results of TKA with patellar resurfacing and whether the position of the median ridge should be considered when selecting a component size. We hypothesized that (1) anthropometric features of the patellae of Korean and Western patients differ, (2) thicknesses of residual patellae influence clinical results, and (3) considering ridge position when selecting component size aids optimal patellofemoral tracking.

Materials and Methods

Seven hundred fifty-two consecutive osteoarthritic knees (713 from females and 39 from males; 466 patients) that underwent TKA between October 1, 2003, and May 31, 2006, with intraoperatively measured patellar anthropometric dimensions, were considered for inclusion in this study. Sixty-eight knees were excluded because the patellae were too distorted to determine the normal anatomy. All 466 patients were Korean and their mean age was 68.6 years (standard deviation [SD] = 6.4 years). The mean height, weight, and body mass index were 151.2 cm (SD = 5.5 cm), 60.8 kg (SD = 9.1 kg), and 26.5 kg/m2 (SD = 3.4 kg/m2) for women and 163.6 cm (SD = 6.7 cm), 65 kg (SD = 7.9 kg), and 24.2 kg/m2 (SD = 2.3 kg/m2) for men. This study was approved by the institutional review board of our hospital, and informed consent for the use of medical information was obtained from all patients.

All surgeries were performed by one surgeon (TKK). Five hundred seventy-two TKAs were performed as bilateral procedures in 286 patients and the other 180 were performed as unilateral procedures. All cases were approached via medial parapatellar arthrotomy, and implant fixation was performed using cement. Three hundred eighty-seven knees were implanted with a fixed-bearing system (Genesis II®; Smith and Nephew, Inc, Memphis, TN) and 365 with a mobile-bearing system (e.motion®; B. Braun-Aesculap, Tuttlingen, Germany). Both systems had dome-shaped patellar components. The Genesis II® system had the same component thickness (9 mm) for all component sizes, whereas the e.motion® system had varying thicknesses (7–12 mm). Patellae were resurfaced in all cases and bone resection was performed using a free-hand technique, during which the saw blade was started from the medial side of the patella stabilized with two towel clips by an assistant. Usually the amount of bone resection was equal to the thickness of the patellar component to be placed (9 mm for Genesis II®; 7–12 mm for e.motion®), with the aim of restoring original patellar thicknesses. When a patella was too thin to restore the original thickness without leaving a residual bone thinner than 10 mm, we reduced the amount of bone resection to allow a residual bone thickness of at least 10 mm. A selected component was placed superomedially when it was smaller than the prepared bone surface.

Intraoperative measurements of patellar dimensions were performed by the operating surgeon (TKK) using a special caliper for neurosurgery (AA845R; B. Braun-Aesculap). To accurately define the intraarticular portion of the patella and to facilitate measurement, the synovial tissue around the patella was sharply incised and peripheral osteophytes were removed with a rongeur. Care was taken to avoid encroaching on the native structure of the patella. Measured dimensions before bone resection included thickness at the median ridge, the mediolateral width of whole patella, and the width of the medial facet (Fig. 1). After bone resection, we measured the bony thickness, mediolateral width, and superoinferior height of residual bone. We measured the overall thicknesses of resurfaced patellae after implantation (Fig. 2).

Fig. 1
The schematic drawings show anthropometric dimensions before bone resection, including medial facet width (MFW), whole patella width (WW), and thickness at the median ridge (RT).
Fig. 2
The schematic drawings show anthropometric dimensions after bone resection, including the widths (W) and heights (H) of the prepared surface, residual bone thicknesses (BT), and overall thicknesses (OT) after implantation.

All clinical information was collected prospectively using predesigned datasheets and maintained in our database by an independent investigator (YGK). Clinical information included demographic data, preoperative clinical statuses, and postoperative outcomes evaluated at 6 and 12 months postoperatively and annually thereafter. Preoperative clinical statuses and postoperative outcomes were evaluated using degree of maximum flexion, the patellofemoral scoring system [5], and WOMAC scales [2]. Degrees of maximum flexion were measured to the nearest 5° using a goniometer with patients in the supine position.

Radiographic outcomes were evaluated by the patellar tilting angle and patellar subluxation index measured using the method described by Bindelglass et al. [3]. All radiographs were viewed by one individual (YGK) who was blinded to the surgical procedures for patellar resurfacing. In the measurements, the patellar tilting angle was defined as the angle between the tangent line of patellar resection surface and the line connecting the highest points of the medial and lateral trochlear ridges of a femoral component (Fig. 3). A positive number was given to the values suggesting lateral tilting and a negative number was given to the values for medial tilting. The presence of lateral tilting was identified when the value was greater than 5° and the presence of medial tilting was noted when the value was smaller than −5°. The patellar subluxation ratio was defined as the percent of the distance between a patellar component center and the trochlear groove center in the total width of the patellar component (Fig. 3). The direction was noted by giving a positive value to the measurement indicating lateral subluxation and a negative value to medial subluxation. The presence of medial or lateral subluxation was identified when patellar subluxation index was greater than 10%. The measurements were made using axial radiographs taken in 45° knee flexion using a leg-holding device (the Merchant view) [11]. All radiographic images were acquired digitally through a picture archiving and communication system (PACS), and assessments were performed subsequently using PACS software (IMPAX; Agfa, Antwerp, Belgium).

Fig. 3A B
Axial radiographs show measurement of (A) patellar tilting angle and (B) patellar subluxation index. (A) The Merchant view of a left knee shows lateral tilting of 2°. (B) The Merchant view of a right knee shows lateral subluxation of 6%.

All statistical analyses were conducted using SPSS software (Version 15.0; SPSS Inc, Chicago, IL), and p values less than 0.05 were considered significant. To determine whether Korean and Western patients differed in terms of anthropometric dimensions, we compared our results with published values [1, 4, 7].

To investigate the effect of residual bony thickness, we compared clinical and radiographic outcomes of 561 knees (with available clinical and radiographic outcomes at 1-year followups) with residual bony thickness less than 12 mm or 12 mm or greater. The associations between preoperative and postoperative patellar thickness differences and clinical and radiographic outcome variables were investigated using the correlation analyses and comparative analyses between the subgroups divided by the amount of preoperative and postoperative patellar thickness deviations. For the correlation analyses, strength of associations was represented by correlation coefficients. For the comparative analyses, knees were classified into three groups: (1) knees with deviations 1 mm or less; (2) knees with a positive deviation of 2 mm or greater; and (3) knees with a negative deviation of 2 mm or greater.

Correlation and comparative analyses between subgroups categorized based on the presence of positional change of the patellar center were done to determine whether the central ridge position should be considered when selecting patellar component size. The presence of position change of the patellar center was determined by subtracting the proportion of patellar component radius in the mediolateral width of a prepared patella from the proportion of medial facet width in the entire mediolateral width of the original patella (Fig. 4).

Fig. 4
The schematic drawings show how deviations between the positions of median ridges and component centers were assessed. Position of the median ridges is represented by the proportion of the medial facet width (MFW) in the whole width of the patella (WW) ...

A priori power analysis, using chi square test or Student’s t test for the subgroup analyses and correlation analyses for the correlation, indicated the sample size of 561 knees with the 1-year clinical and radiographic information would provide greater than 90% power (α = 0.05) to detect a difference of 5% in the clinical and radiographic outcomes and a correlation greater than 0.3.


Comparisons of Western and Korean anthropometric dimensions in patients undergoing TKA showed Koreans have thinner and smaller patellae than Westerners. In our study, mean central ridge thickness was 21.2 mm in women and 23.1 mm in men, whereas corresponding reported mean thicknesses in Western patients are 21.8 to 22.5 mm and 23.9 to 26.1 mm (Table 1). In Koreans, mean postcut heights and widths were 33.1 mm and 41.0 mm in women and 36.2 mm and 45.6 mm in men, whereas corresponding values in Western patients are 35.0 mm and 42.7 mm in women and 39.4 mm and 49.5 mm in men [1].

Table 1
Comparisons of the anthropometric dimensions of the patellae of Korean and Western patients (based on published values)

Correlation analyses revealed no associations between preoperative and postoperative patellar thickness differences and clinical or radiographic results. The mean postoperative overall thickness deviation was a 0.1-mm decrease (range, 3-mm decrease to 3-mm increase). Correlation coefficients were smaller than 0.1 and p values were greater than 0.1 for all clinical and radiographic outcome variables. Subgroup analyses of knees with deviations of ± 1 mm or less (658 knees [87.5%]), knees with a deviation of +2 mm or greater (37 knees [4.9%]), and knees with a deviation of −2 mm or greater (57 knees [7.6%]) found no differences with respect to any clinical or radiographic parameter (Table 2). In addition, subgroup analyses on knees with a residual bony thickness less than 12 mm and with a residual bony thickness of 12 mm or greater showed no differences for any clinical and radiographic parameters, except WOMAC pain score (Table 3); knees with a residual bony thickness less than 12 mm had a poorer mean pain score (3.5 versus 2.4, p = 0.029). Mean residual bony thickness after bone resection and overall thickness after patellar resurfacing were 12.5 mm (range, 10–15 mm) and 21.1 mm (range, 17–25 mm) in women and 13.5 mm (range, 12–16 mm) and 22.7 mm (range, 20–26 mm) in men, respectively. Only one case of periprosthetic patellar fracture was recorded, and this occurred in a knee with a thickness of 12 mm or more. The patient, a 56-year-old woman, sustained direct injury to the patella after slipping 13 months after surgery. A radiographic evaluation at the time showed an undisplaced patellar fracture without component loosening. The original patellar thickness, residual bony thickness after bone resection, and overall thickness after implantation were 23 mm, 14 mm, and 23 mm, respectively. She recovered from this injury after 6 weeks immobilization wearing a brace.

Table 2
Comparisons of the clinical and radiographic parameters of patients with different postoperative thickness deviations*
Table 3
Comparison of the clinical and radiographic parameters of patients*

Correlation analyses and subgroup analyses indicated considerations of the original position of the central ridge are not necessary for patellar component selection based on patellar height when the component is positioned medially on a prepared surface. Correlation analyses revealed preoperative to postoperative differences between the center of a positioned component and the original position of the central ridge are not associated with clinical or radiographic outcome. Moreover, subgroup analyses on knees with a medialized or unchanged component center and knees with a lateralized component center revealed no differences (Table 4).

Table 4
Comparison of clinical and radiographic results of knees*


Patella-related morbidities are among issues of concern after TKA, and detailed anthropometric information can be helpful for development of proper surgical techniques and prosthesis designs. In this study, we aimed to document anthropometric features of the Korean adult patella receiving TKA for advanced osteoarthritis. Furthermore, we attempted to elucidate the clinical implications of postoperative patella thickness deviations, residual bony thickness, and deviations between preoperative positions of median ridges and postoperative positions of component centers. All patients were ethnic Koreans and all had an underlying diagnosis of osteoarthritis, which probably minimized the effects of many confounding factors during evaluations of clinical and radiographic outcomes.

Several limitations should be noted when interpreting our findings. First, the patients were typically elderly women (women composed 94.8% of the study population and had a mean age of 68.6 years) and, thus, physical activity levels would be expected to be low. Second, we found no associations between preoperative and postoperative overall thickness deviations, residual bony thicknesses, or alterations in the central positions of components and clinical or radiographic outcomes. Although we used an extensive set of clinical and radiographic parameters, which we deemed appropriate for evaluating TKA with patellar resurfacing, these measures may not have been adequate to detect subtle differences. Third, the clinical data and radiographs investigated were obtained 1 year after surgery, and although clinical status after TKA typically plateaus 1 year postoperatively, we cannot exclude the possibility that somewhat different results may have been found had followups been longer. Fourth, we used a caliper to measure the anthropometric dimensions of patellae, and although caliper measurements are commonly used in anthropometric studies, a question was raised regarding accuracy of smooth caliper measurements [15]. The authors of that study suggested a spiked caliper should be used to reduce measurement errors caused by intervening soft tissues [15]. Finally, the knees we evaluated had advanced osteoarthritis warranting TKA. Even though we excluded severely deformed and diseased knees that might have biased measurement findings, it is reasonable to assume that all patellae were diseased to some extent. Thus, anthropometric measures, including those of original patellae thicknesses and median ridge positions, may differ from those of patellae without osteoarthritis. However, we believe this study provides valuable information required for TKA where this pathologic anthropometric information is required.

Information of anthropometric patellar dimensions can be helpful for proper development of surgical techniques for patellar resurfacing and patellar prostheses that better suit the patient population. However, current TKA systems were largely developed based on anatomic data obtained from Western subjects and, thus, might not be optimized for Asian patients. Moreover, despite the need for accurate anthropometric information for Asian patients, no previous study has reported detailed anthropometric data of the patella for Asian patients. The primary goal of our study was to compare anthropometric dimensions of our Korean patients with data from Western patients. Our review of the English literature revealed several studies on this topic for Western patients. Nevertheless, it was difficult to identify comparable studies because the majority of studies involved small case numbers and provided no ethnic information on the studied patient population and little information on anthropometric dimensions [1, 4, 7]. Given these limitations of previous studies of Western patients, our study reveals Korean patellae are thinner and smaller than those of Western patients (Table 1). However, despite differences in thicknesses, widths, and heights, width/height ratios and ridge positions were similar in Koreans and Westerners.

Some studies have advocated reestablishment of original thickness and adequate residual bony thickness as key surgical guidelines [4, 6, 8, 10, 16]. However, in Korean patients, these two guidelines often cannot be satisfied. In such cases, the surgeon is forced to choose between restoring the original thickness while accepting a low residual bony thickness and retaining sufficient residual bony thickness while accepting an increased overall thickness. Our study indicates either of these two options can be chosen without causing adverse clinical or radiographic effects, providing postoperative overall thickness remains within 3 mm of the original thickness and residual bony thickness is between 10 mm and 15 mm. Our findings concur with those of several clinical studies [6, 10, 17]. A study comparing the incidence of lateral retinacular release in TKA found knees with a residual bony thickness less than 15 mm had a lower incidence of lateral retinacular release than knees with a residual thickness of 15 mm or greater [6]. Moreover, a study of 112 Chinese patients showed no differences in the maximum flexion and function scores (as described by the American Knee Society) between the knees with a residual bony thickness greater than 12 mm and 12 mm or less [10]. Another study of 1146 TKAs in which postoperative overall thickness deviations were ± 6 mm found no association between thickness change and several clinical parameters (patellar fracture, component loosening, the need for lateral retinacular release, and maximum flexion) [17]. The above findings and ours indicate no adverse clinical findings are likely to occur if overall and residual bony thicknesses are maintained in a reasonable range, despite the possibility that postoperative patella thickness may affect patellofemoral contact pressures in an experimental setting [8, 13, 16].

Because the median ridge of the patella plays an important role in patellofemoral kinematics, preoperative to postoperative deviations between the positions of the median ridge and the center of the patellar component may have clinical implications. The size of a patellar component typically is determined by the height of the prepared patellar surface, and in many cases, a component of suitable height can be chosen to maintain the position of the component center at the original position of the median ridge. However, our findings in this respect suggest alterations of ridge position do not affect clinical and radiographic results, providing the selected component is placed as medially as possible. Our findings lead us to propose the size of a patellar component can be determined based on width-height information, regardless of the original position of the median ridge.

Results of our study showed Korean patients scheduled for TKA for advanced osteoarthritis typically had thinner and smaller patellae than Western patients. Our findings indicate bone resections for patellar resurfacing can be flexible without jeopardizing clinical outcomes, providing postoperative overall thickness remains within 3 mm of the original thickness and residual bony thickness is between 10 mm and 15 mm. Based on our finding of no association between postoperative deviations of component centers and clinical and radiographic outcomes, we propose dome-shaped patellar component size can be determined by width-height information of the prepared surface, regardless of the original position of the median ridge.


We thank Dr. Sung Ju Kim (Department of Statistics, Korea University, Seoul, Korea) for help with statistical analyses.


One or more of the authors (TKK, CBC) have received funding from the clinical research fund (B-0708/048-103) of Seoul National University Bundang Hospital.

Each author certifies that his or her institution has approved the human protocol for this investigation, that all investigations were conducted in conformity with ethical principles of research, and that informed consent was obtained.


1. Baldwin JL, House CK. Anatomic dimensions of the patella measured during total knee arthroplasty. J Arthroplasty. 2005;20:250–257. [PubMed]
2. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol. 1988;15:1833–1840. [PubMed]
3. Bindelglass DF, Cohen JL, Dorr LD. Patellar tilt and subluxation in total knee arthroplasty: relationship to pain, fixation, and design. Clin Orthop Relat Res. 1993;286:103–109. [PubMed]
4. Chmell MJ, McManus J, Scott RD. Thickness of the patella in men and women with osteoarthritis. Knee. 1995;2:239–241.
5. Feller JA, Bartlett RJ, Lang DM. Patellar resurfacing versus retention in total knee arthroplasty. J Bone Joint Surg Br. 1996;78:226–228. [PubMed]
6. Greenfield MA, Insall JN, Case GC, Kelly MA. Instrumentation of the patellar osteotomy in total knee arthroplasty: the relationship of patellar thickness and lateral retinacular release. Am J Knee Surg. 1996;9:129–131; discussion, 131–132. [PubMed]
7. Hitt K, Shurman JR 2nd, Greene K, McCarthy J, Moskal J, Hoeman T, Mont MA. Anthropometric measurements of the human knee: correlation to the sizing of current knee arthroplasty systems. J Bone Joint Surg Am. 2003;85(suppl 4):115–122. [PubMed]
8. Hsu HC, Luo ZP, Rand JA, An KN. Influence of patellar thickness on patellar tracking and patellofemoral contact characteristics after total knee arthroplasty. J Arthroplasty. 1996;11:69–80. [PubMed]
9. Huberti HH, Hayes WC. Patellofemoral contact pressures: the influence of q-angle and tendofemoral contact. J Bone Joint Surg Am. 1984;66:715–724. [PubMed]
10. Koh JS, Yeo SJ, Lee BP, Lo NN, Seow KH, Tan SK. Influence of patellar thickness on results of total knee arthroplasty: does a residual bony patellar thickness of ≤12 mm lead to poorer clinical outcome and increased complication rates? J Arthroplasty. 2002;17:56–61. [PubMed]
11. Merchant AC, Mercer RL, Jacobsen RH, Cool CR. Roentgenographic analysis of patellofemoral congruence. J Bone Joint Surg Am. 1974;56:1391–1396. [PubMed]
12. Merkow RL, Soudry M, Insall JN. Patellar dislocation following total knee replacement. J Bone Joint Surg Am. 1985;67:1321–1327. [PubMed]
13. Oishi CS, Kaufman KR, Irby SE, Colwell CW Jr. Effects of patellar thickness on compression and shear forces in total knee arthroplasty. Clin Orthop Relat Res. 1996;331:283–290. [PubMed]
14. Police P, Lotke PA, Lonner JH. Principles of instrumentation and component alignment. In: Callaghan JJ, Rosenberg AG, Rubash HE, Simonian PT, Wickiewicz TL, eds. The Adult Knee. Philadelphia, PA: Lippincott Williams & Wilkins; 2003:1085–1093.
15. Rauh MA, Bayers-Thering M, Buyea CM, Phillips M, Krackow KA. Reliability and validity of a new caliper for measuring patellar thickness. J Arthroplasty. 2002;17:105–107. [PubMed]
16. Reuben JD, McDonald CL, Woodard PL, Hennington LJ. Effect of patella thickness on patella strain following total knee arthroplasty. J Arthroplasty. 1991;6:251–258. [PubMed]
17. Ritter MA, Pierce MJ, Zhou H, Meding JB, Faris PM, Keating EM. Patellar complications (total knee arthroplasty): effect of lateral release and thickness. Clin Orthop Relat Res. 1999;367:149–157. [PubMed]
18. Star MJ, Kaufman KR, Irby SE, Colwell CW Jr. The effects of patellar thickness on patellofemoral forces after resurfacing. Clin Orthop Relat Res. 1996;322:279–284. [PubMed]

Articles from Clinical Orthopaedics and Related Research are provided here courtesy of The Association of Bone and Joint Surgeons