Femoral components used in total hip arthroplasties (THA) are known to affect the loads placed in the proximal femur.1,32,33,42 The femoral load reduction associated with stress shielding after THA has resulted in bone loss that is apparent on radiographs.6,42 The measurable changes associated with stress shielded bone include cortical thinning,26,46 decreased bone mineral content (BMC; g),3,13,21,28,43 and/or decreased bone mineral density (BMD; g/cm2).3,22,26,35,39,44 The clinical consequences that result from stress shielding after primary THA are unclear. However, the quality and quantity of cortical bone have been shown to be important if revision should become necessary.30,42,47 Therefore, bone maintenance after THA may improve clinical outcomes.
Factors such as patient activity and demographics may influence stress shielding. Previously, greater bone loss attributed to stress shielding was associated with decreased walking ability.14 The presence of preoperative osteopenia as measured by BMD and BMC has been shown to correlate with greater bone loss because of stress shielding after THA.26,43 In addition to general bone loss, results of one small (four patients) study suggested changes in cortical bone shape after THA.46
We hypothesized that bone loss attributed to stress shielding in the implanted femurs would correlate with lower mechanical usage scores (MUS), advanced age at death, decreased weight, larger implant size, increased time in situ, and less bone stock in the nonimplanted femurs as measured by lower cross-sectional area, BMD, and BMC. We further postulated that the implanted femurs would be reduced in cortical bone cross-sectional areas and polar moments of inertia, increased in inner and/or outer cortical bone diameters, decreased in principal axes rigidity values, and increased in centroid and principal axes locations as compared with the nonimplanted femurs.
MATERIALS AND METHODS
To answer our hypotheses, human femurs were analyzed for bone loss by measuring differences in cortical bone cross-sectional area, BMD, and BMC between implanted and nonim-planted bilateral femurs. These bone loss measures were correlated to patient activity and demographics. To further understand bone adaptation attributed to stress shielding, cortical bone shape measurements were compared between the implanted and nonimplanted femurs.
We obtained 20 femurs from 10 postmortem donors who had primary THAs and consented to donation. Each pair of bilateral femurs contained one femoral component (Fig. 1). The study was performed with IRB approval.
All femoral components were of the second generation Zweymüller design: press-fit, grit-blasted, titanium alloy taper stem with rectangular cross section (Alloclassic®, Zimmer, Warsaw, IN). Implantation was performed by one surgeon (DKL) using the same surgical technique in which there was no reaming, but there was broaching and cementless fixation.24 All patients were advised to be full weightbearing on the first postoperative day. All implants appeared radiographically stable at death because no implant migration had occurred compared with the immediate postoperative radiograph, and a portion of bone was in contact with the implant in at least three of the four sections for all 10 femurs (Fig 1). The cross-sectional shape, BMC, and BMD were compared between the implanted and nonimplanted femurs. To determine the variables related to bone loss we correlated the percent difference in cortical bone area, BMC, and BMD between implanted and nonimplanted femurs to patient activity, patient age, weight, implant size, and time in situ (Table 1).11
To estimate patient activity levels, we developed a MUS from modified Harris hip scores (HHS)19 (Table 2). The HHS was recorded by the surgeon (DKL) before the study. The MUS was calculated by summing the scores of four components in the function category of the HHS (Table 3). The ideal measure of ambulatory activity would directly record the frequency of hip usage and the loads on the hip.2 The MUS estimated the patient's frequency of use and loads in the implanted hip at the last followup before death. For example, patients who can walk 2 to 3 blocks were thought to use their hips less frequently than patients who can walk more than 6 blocks. Patients with a moderate limp applied a smaller load to the implanted hip relative to their nonimplanted hip than patients who walked without a limp. Thus, patients walking shorter distances with a limp had lower MUS than patients walking longer distances with no limp.
The patient demographics tested for correlation to bone loss were patient age at death, weight, implant size, and time in situ (Table 1).11 Cortical bone area, BMC, and BMD of the donors' nonimplanted femurs also were tested for correlations to stress shielding in the implanted femurs as performed in prior studies.26,43 In these studies, BMC and BMD of the nonimplanted femurs were used to estimate preoperative osteopenia.26,43
Bone loss measurements (BMC and BMD) were measured for seven of the 10 femur pairs by using a dual energy xray absorptiometry (DEXA) scanner (Lunar Expert - XL, Lunar Corporation, Madison, WI).3 The additional three pairs of femurs used in the geometric analyses had been embedded and sectioned before the authors had access to a DEXA scanner. Therefore, BMC and BMD could not be performed on these three femur pairs. The percent difference in BMC and BMD of the seven paired femurs3 were not tested previously for correlations with patient activity and demographics.
We examined the cortical bone at 25% (trochanter region), 45%, 65%, and 85% (referred to as Levels 1-4, respectively) of implant length (Fig 1). By using the percent length of the femoral component, we normalized the analysis locations for all donors.23,26,40,46 We inquired about changes to the Alloclassic® femoral component's proportions with a change in size, but Zimmer, Inc, stated this information was considered confidential and proprietary. The implant size for this donor population is listed in Table 1.
To determine BMC and BMD, the femurs were DEXA scanned in an acrylic tank filled with a 15-cm radius of water to simulate soft tissue. The implanted and nonimplanted bilateral femurs were oriented at the identical anteversion angle. The femurs were oriented perpendicular to the DEXA beams to limit rotational effects.5 The DEXA machine scanned the aspects of the medial and lateral femur in 1-cm high sections for each level (Fig 1).
To determine the cross-sectional shape at the four levels (Fig 1), the femurs were dehydrated and embedded in polymethylmethacrylate (PMMA).12 After embedding, 3-to 5-mm sections were cut using a custom, cut-off saw (Rockazona, Peoria, AZ) at the previously defined levels (Fig 1).
The cross sections were oriented according to anatomic regions. High-resolution contact-section radiographs were taken (Torrex 120D; Scanray, Hawthorne, CA) with settings at 57 kV at 5 minutes and 23 seconds. The radiographs were scanned into a computer using true color, RGB, resolution 200 dpi settings (Powerlook 2100XL (UMAX), Dallas, TX). Cortical bone boundaries from the scanned images were defined automatically based on contiguous areas of similar bone density (Adobe Photoshop 5.5, San Jose, CA). In locations of dense cancellous bone, manual tracing was needed (Fig 2). All cortical bone outlines were performed by one investigator (TGR). This investigator was not blinded to which sections were implanted and nonim-planted.
We used the following measurements to quantify cross-sectional shape of cortical bone. These measurements have been defined.37 The cross-sectional area was defined as the quantified amount of cross-sectional cortical bone area (mm2). The centroid (x, y) was defined as the center of mass of the cross-section examined. The centroid quantified cortical bone distribution of the transverse cross section. Centroid (x) defined the cortical bone distribution in the medial and lateral quadrants (mm), and centroid (y) defined the cortical bone distribution in the anterior and posterior quadrants (mm). The cortical bone density was not included in this measurement. The principal axes (Imax, Imin) were defined as the second moment of area, also referred to as the area moment of inertia (I). They were used to evaluate beam rigidity under bending. The maximum and minimum second moments of area (Imax, Imin) indicated the relative magnitudes of greatest and least bending rigidity of a cross section, respectively. Cortical bone quantity and distribution contributed to the rigidity values (mm4). The polar moment of inertia (J) represented the ability of the cross sections to resist torsional moments. Cortical bone quantity and distribution contributed to J (mm4). The locations of principal axes (?1, ?2) were designated as the locations of Imax and Imin. This indicated the axes about which the cross section were the most (?1) and least (?2) rigid when bending. The principal axes (°) were always 90° apart and intersected at the cortical bone's centroid (Fig 2). The principal axes depended only on cortical bone distribution. The inner and outer cortical bone diameters quantified the differences in endosteal and periosteal cortical bone boundaries along the principal axes between the implanted and nonimplanted femurs (mm). Thus, cortical bone diameters quantified the cortical bone thickness along the axes of greatest (?1) and least (?2) bending rigidity.
Cortical bone measurements were performed in Scion Image Beta 4.02 (Scion Corporation, Frederick, MD). Code from the published SLICE program31 was converted into a modified macro to be used in Scion Image to calculate cortical bone area, centroid (x, y), maximum and minimum second moments of inertia (Imax, Imin), and location of principal axes (?1, ?2). Phi1 (?1) was the location of Imax (minor axis), and Phi2 (?2) was the location of Imin (major axis). The angle ? was measured counterclockwise from the mediolateral axis intersecting the centroid. Moments of inertia equations, the parallel-axis theorem, and the first derivative of transformation equations were used in the program to determine Phi1 (?1) and Phi2 (?2).36 The program was validated with standard shapes, such as rings and double triangles. The program showed less than 2% error in accuracy. The geometric analysis showed less than 5% error in precision after repeating the tracing and geometry program three times on a worst-case image. A worst-case image required manual tracing on at least 75% of the inner cortical bone boundary.
Once the program completed the calculations for principal axes and location of principal axes, the polar moment of inertia (J) was calculated by adding Imax and Imin (J = Imax + Imin).25,36 Cortical bone inner and outer diameters were measured along Imax and Imin of each cross section (Fig 2).
The measurements of cortical bone shape in the implanted and the contralateral nonimplanted femurs were compared using a paired t test at each level (Levels 1-4). The p values were adjusted for four multiple comparisons using the Tukey-Ciminera-Heyse procedure.38 The paired t test results of BMD and BMC have been previously reported.3 We determined all variables were normal using the Shapiro-Wilks test for normality (Intercooled Stata 8.0, College Station, TX,). Simple linear regression was used to correlate bone loss (cross-sectional area, BMD, BMC) with mechanical usage scores and patient demographics. All statistical tests were for a two-sided comparison, with significance set at p ≤ 0.05 (StatView 5.0, Cary, NC).
Lower MUS correlated (r2 = 0.88; p < 0.001) with greater bone loss as measured by cortical bone cross-sectional area (Fig 3). However, the MUS did not correlate with bone loss as measured by BMD and BMC. Thus, mineral levels in our patients were not as affected by mechanical usage as cortical bone cross-sectional area.
Lower patient weight correlated (r2 = 0.58; p = 0.045) with greater bone loss as measured by BMC. However, patient weight did not correlate with bone loss as measured by cortical bone cross-sectional area or BMD. Age, implant size, and time in situ did not correlate with bone loss as measured by cortical bone cross-sectional area, BMC, or BMD.
Lower cross-sectional area values of the nonimplanted femurs correlated (r2 = 0.45; p = 0.035) with greater bone loss in the implanted femurs as measured by the difference in cross-sectional area between the implanted and nonimplanted femurs (Fig 4). In other words, patients with thinner cortices in their nonimplanted femur experienced greater bone loss in their implanted femur. However, BMD and BMC values of the nonimplanted femurs did not correlate with bone loss in the implanted femurs as measured by the differences in BMD and BMC between the implanted and nonimplanted femurs (Fig 4).
The geometric analyses determined cortical bone cross-sectional areas were less (p < 0.05) in the implanted femurs in Levels 1 through 4. Polar moments of inertia (J) were less (p = 0.038) in the implanted femurs at Level 3 (Table 4).
The implanted femurs increased (p < 0.05) in inner and outer diameters. Levels 2 to 4 of the implanted femurs' inner diameters were greater along the Imax axis (p = 0.006, p = 0.036, respectively). Levels 2 and 3 were greater the along Imin axis (p < 0.001, p = 0.030, respectively). The outer diameters expanded (p = 0.018) along the Imax axis at Level 2 (Table 4).
The implanted femurs' principal axis, Imin, was reduced (p = 0.021) at Level 3. Because the implanted femurs at Levels 1, 2, and 4 were similar in Imax, Imin, or J, the rigidities of those levels were more consistently maintained than at Level 3. Maintaining cross-sectional rigidity in Levels 1, 2, and 4 with reductions in cross-sectional area would require changes in the cortical bone shape, such as periosteal expansion. Periosteal expansion (p = 0.018) occurred at Level 2 (Table 4).
The centroid (y) in the implanted femurs shifted (p = 0.031) toward the anterior quadrant at Level 1 compared with the nonimplanted femurs. The shift toward the anterior quadrant indicated the majority of the cortical thinning occurred in the posterior quadrant, causing the centroid to shift toward the anterior quadrant. Centroid (x) did not change in the implanted femurs; therefore a similar amount of endosteal resorption occurred in the medial and lateral quadrants. Like the centroid, the location of principal axes helped quantify changes in cortical bone distribution. The location of principal axes showed no differences between the implanted and nonimplanted femurs (Table 4). The changes seen between each implanted femur seemed random as to the amount of change and the direction of rotation.
There has been interest in characterizing bone adaptation after THA to gain better understanding of bone loss attributed to stress shielding and bone maintenance. We sought to characterize bone loss by investigating the influence of patient activity and demographics on reductions in cortical bone cross-sectional area, BMD, and BMC. Additionally, bone adaptation was investigated by measuring changes in cortical bone shape. To our knowledge, this study was unique in comparing three bone loss measures with the patients' MUS and selected demographics and in analyzing changes in bone shape to the extent performed in this study.
Our study has several limitations. First, we could not perform a power analysis because the minimum clinically relevant changes15 for cortical bone shape, BMD, and BMC are not yet known. Therefore, our nonsignificant results are inconclusive because a Type II error may have occurred.15 Second, the study may not reflect bone adaptation after THA in patients of other age groups because bone remodeling has been shown to be affected by age.8,18,20,34,45 Third, we assumed the MUS reflected hip usage. However, neither the MUS nor the previously used D'Aubigné and Postel's walking score9 (Table 5) has been validated for representing ambulatory activity. Even without validation, the MUS seemed to quantify ambulation activities in more detail than the D'Aubigné and Postel's walking score9 because it separately scores limp, support, distance walked, and the ability of the patient to climb stairs. Fourth, our patients' activity between their last MUS and death was unknown. Additional study is needed to learn how activity and bone remodeling affects bone loss attributable to stress shielding during the entire life of patients who have THAs.
Lower MUS had a high correlation with greater bone loss as measured by cortical bone cross-sectional area, supporting our first hypothesis. This supported a previous study, which found bone loss after THA was affected by stress shielding and patients applying smaller loads to the implanted limb relative to the nonimplanted limb.4 In addition, a lower walking ability from D'Aubigné and Postel's walking scores9 has been associated with stress shielding at 2 years postoperatively (Table 5).14 Another study using the Zweymüller implant found greater walking ability9 was associated with stress shielding at 10 years postoperatively.17 The discrepancy may be related to bone loss occurring from age or other morbidities in the 10-year study because the implanted femurs were compared over time instead of being compared with the contralateral nonimplanted femurs.8,18,20,34,45 The majority of evidence supports a relationship between decreased activity and greater bone loss. This relationship may be explained with Frost's mechanostat theory.16 Frost's theory stated a minimum stress or strain value must be maintained to prevent bone loss.16 One study suggests a minimum effective strain for bone maintenance because patients who have THAs retained more bone when the estimated stress in the cortical bone of the implanted femurs was above a threshold value.40 Our data supported Frost's concept of a minimum effective strain16 because greater use of the implanted hip seemed to sustain strains above the theoretical threshold value for bone maintenance (Fig 5).
The lower patient weight correlated with greater bone loss as measured from BMC, supporting our second hypothesis. The contribution of patient weight to bone loss remains unclear, as other studies showed fair to no correlations.26,43 It may be a combination of factors, such as patient weight and ambulatory activities, that play a role in keeping the strains on the implanted femur above Frost's minimum effective strain for bone maintenance (Fig 5).16
Lower cross-sectional area in the nonimplanted femurs moderately correlated with greater bone loss, supporting our third hypothesis. Previously, the bone quality of the nonimplanted femurs has been used to estimate preoperative osteopenia.26,43 Thus, thin femoral cortices before THA may increase the risk of greater bone loss after THA. BMD and BMC of the nonimplanted femurs did not correlate to bone loss, although such correlations have been reported.26,43 This inconsistency between our data and those of previous studies may be because of small sample size, differences in quality of fixation, differences in femoral components, or our patients' cortical bone area being more sensitive to stress shielding than the bone remodeling factors which influence BMD and BMC.
Cortical bone cross-sectional area and polar moment of inertia (J) were less in the implanted femurs, supporting the fourth hypothesis. The reduction in cross-sectional area in Levels 1 to 4 was consistent with previous studies using Zweymüller femoral components as bone loss generally decreased from the proximal to distal end of the cortical bone.17,48 Previous studies quantifying cortical bone loss around femoral components have shown a range of 2.8-45%.13,26,35,43,46 This was consistent with our findings, where bone loss ranged from 13-28% in Levels 1 to 4 (Fig 6). Results of a study analyzing cortical bone cross-sectional rigidity after THA also were consistent with our findings as J was reduced in the proximal portion of the implanted femurs compared with the nonimplanted femurs.46
The data supported the fifth hypothesis because the inner and outer diameters of the cortical bone increased. The inner diameter increases in Levels 2 to 4 were responsible for the cortical bone loss because the outer cortices expanded or remained unchanged. The larger increase in the cortical bone inner diameter compared with the outer diameter resulted in cortical thinning, which was consistent with results in a previous study.26 However, we analyzed the inner and outer cortical bone diameters separately and were able to verify that cortical bone loss occurred preferentially along the endosteal surface.
The rigidity value from the Imin principal axis was reduced in the implanted femurs, supporting the sixth hypothesis. A previous study showed mediolateral bending rigidity (Iy) typically was reduced in the implanted femur at the levels comparable to those in our study.46 However, only mechanical testing of femurs with the implant removed can determine actual changes in bending rigidity because other material properties can contribute to bone rigidity (eg, mineralization, porosity, and collagen fiber orientation).7,10,27,29,41 Therefore, the results based only on cortical bone shape showed implanted femurs became more flexible after THA.
The cortical bone distribution only changed at Level 1 because the centroid shifted toward the anterior quadrant, supporting the seventh hypothesis. Cortical thinning in the posterior quadrant may have caused this shift. A previous study showed the majority of cortical thinning occurred in the medial and posterior quadrants (above our Level 1 location), and in the medial and anterior quadrants (between our Level 1 and Level 2 locations).26 These differences could have been from different donor populations because the previous study combined donors with cemented (n = 13) and cementless (n = 11) femoral components and included nine different femoral components with varying cross-sectional shapes. We analyzed one type of cementless femoral component implanted by the same surgeon. In addition, variations in shape at Level 1 could be related to intrinsic differences between bilateral femurs.37
In summary, greater cortical thinning attributed to stress shielding strongly correlated to lower MUS. This indicated patients with greater ambulatory activity maintained stresses and strains above a threshold value16 to limit bone loss after THA. The patient weight and preoperative osteopenia had less influence on bone loss than ambulatory activity. The geometric measurements depending on cortical bone distribution did not change as consistently as those measuring cortical bone quantity. Cortical bone in the implanted femur was more likely to diminish in cross-sectional area uniformly than to consistently change in distribution. The reduction in cross-sectional area was minimized when patients who had THAs achieved higher activity levels.
We thank Derek Liau for assisting with the tissue processing, Bettina Willie for helpfulness in the laboratory, and Greg Stoddard for assisting with the statistical analysis.
1. Aamodt A, Lund-Larsen J, Eine J, Andersen E, Benum P, Husby OS. Changes in proximal femoral strain after insertion of uncemented standard and customised femoral stems: an experimental study in human femora. J Bone Joint Surg Br
2. Bergmann G, Deuretzbacher G, Heller M, Graichen F, Rohlmann A, Strauss J, Duda GN. Hip contact forces and gait patterns from routine activities. J Biomech
3. Bloebaum RD, Liau DW, Rosenbaum TG, Lester KD. Dual energy x-ray absorptiometry measurement and accuracy of bone mineral after unilateral total hip arthroplasty. J Arthroplasty
. 2006;21: 612-622.
4. Bryan JM, Sumner DR, Hurwitz DE, Tompkins GS, Andriacchi TP, Galante JO. Altered load history affects periprosthetic bone loss following cementless total hip arthroplasty. J Orthop Res
. 1996;14: 762-768.
5. Cohen B, Rushton N. Accuracy of DEXA measurement of bone mineral density after total hip arthroplasty. J Bone Joint Surg Br
6. Cristofolini L. A critical analysis of stress shielding evaluation of hip prostheses. Crit Rev Biomed Eng
7. Currey JD. The effect of porosity and mineral content on the Young's modulus of elasticity of compact bone. J Biomech
8. Currey JD, Brear K, Zioupos P. The effects of ageing and changes in mineral content in degrading the toughness of human femora. J Biomech
9. D'Aubigne RM, Postel M. Functional results of hip arthroplasty with acrylic prosthesis. J Bone Joint Surg Am
10. Dickenson RP, Hutton WC, Stott JRR. The mechanical properties of bone in osteoporosis. J Bone Joint Surg Br
11. Dorr LD, Faugere M-C, Mackel AM, Gruen TA, Bognar B, Malluche HH. Structural and cellular assessment of bone quality of proximal femur. Bone
12. Emmanual J, Hornbeck C, Bloebaum RD. A polymethyl methacrylate method for large specimens of mineralized bone with implants. Stain Technol
13. Engh CA, McGovern TF, Bobyn JD, Harris WH. A quantitative evaluation of periprosthetic bone-remodeling after cementless total hip arthroplasty. J Bone Joint Surg Am
14. Engh CA Jr, Young AM, Engh CA Sr, Hopper RH Jr. Clinical consequences of stress shielding after porous-coated total hip arthroplasty. Clin Orthop Relat Res
15. Freedman KB, Bernstein J. Sample size and statistical power in clinical orthopaedic research. J Bone Joint Surg Am
16. Frost HM. Bone “mass” and the “mechanostat”: a proposal. Anat Rec
17. Garcia-Cimbrelo E, Cruz-Pardos A, Madero R, Ortega-Andreu M. Total hip arthroplasty with use of the cementless Zweymuller Alloclassic system: a ten to thirteen-year follow-up study. J Bone Joint Surg Am
18. Grynpas MD, Holmyard D. Changes in quality of bone mineral on aging and in disease. Scanning Microsc
19. Harris WH. Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty: an end stage results using a new method of evaluation. J Bone Joint Surg Am
20. Kaptoge S, Dalzell N, Loveridge N, Beck TJ, Khaw KT, Reeve J. Effects of gender, anthropometric variables, and aging on the evolution of hip strength in men and women aged over 65. Bone
21. Kerner J, Huiskes R, van Lenthe GH, Weinans H, van Rietbergen B, Engh CA, Amis AA. Correlation between pre-operative periprosthetic bone density and post-operative bone loss in THA can be explained by strain-adaptive remodelling. J Biomech
22. Kilgus DJ, Shimaoka EE, Tipton JS, Eberle RW. Dual-energy x-ray absorptiometry measurement of bone mineral density around porous-coated cementless femoral implants. J Bone Joint Surg Br
23. Kuo TY, Skedros JG, Bloebaum RD. Comparison of human, primate, and canine femora: implications for biomaterials testing in total hip replacement. J Biomed Mater Res
24. Lester DK, Helm M. Mini-incision posterior approach for hip arthroplasty. Orthop Traumatol
25. Lieberman DE, Polk JD, Demes B. Predicting long bone loading from cross-sectional geometry. Am J Phys Anthropol
. 2004;123: 156-171.
26. Maloney WJ, Sychterz C, Bragdon C, McGovern T, Jasty M, Engh CA, Harris WH. The Otto Aufranc Award: Skeletal response to well fixed femoral components inserted with and without cement. Clin Orthop Relat Res
27. Martin RB, Ishida J. The relative effects of collagen fiber orientation, porosity, density, and mineralization on bone strength. J Biomech
28. McAuley JP, Sychterz CJ, Engh CA Sr. Influence of porous coating level on proximal femoral remodeling: a postmortem analysis. Clin Orthop Relat Res
29. McCalden RW, McGeough JA, Barker MB, Court-Brown CM. Age-related changes in the tensile properties of cortical bone: the relative importance of changes in porosity, mineralization, and microstructure. J Bone Joint Surg Am
30. Murphy SB, Rodriguez J. Revision total hip arthroplasty with proximal bone loss. J Arthroplasty
. 2004;19 (suppl 1):115-119.
31. Nagurka ML, Hayes WC. An interactive graphics package for calculating cross-sectional properties of complex shapes. J Biomech
32. National Institutes of Health. NIH Consensus Development Conference, September 12-14, 1994. Bethesda, MD: Nationals Institutes of Health; 1994;12:1-31.
33. Oh I, Harris WH. Proximal strain distribution in the loaded femur: an in vitro comparison of the distributions in the intact femur and after insertion of different hip-replacement femoral components. J Bone Joint Surg Am
34. Parfitt A. Age-related structural changes in trabecular and cortical bone: cellular mechanisms and biomechanical consequences. Calcif Tissue Int
. 1984;36 (suppl 1):S123-S128.
35. Pitto RP, Schramm M, Hohmann D, Schmidt R. Clinical outcome and quantitative evaluation of periprosthetic bone-remodeling of an uncemented femoral component with taper design. A prospective study. Chir Organi Mov
36. Pytel A, Kiusalaas J. Engineering Mechanics: Statics
. 2nd ed. Pacific Grove, CA:Brooks/Cole Publishing Company; 1999.
37. Rosenbaum TG, Hamblin T, Bloebaum RD. Determining the degree of cortical bone asymmetry in bilateral, nonpathological, human femur pairs. J Biomed Mater Res A
38. Sankoh AJ, Huque MF, Dubey SD. Some comments on frequently used multiple endpoint adjustment methods in clinical trials. Stat Med
39. Schmidt R, Muller L, Kress A, Hirschfelder H, Aplas A, Pitto RP. A computed tomography assessment of femoral and acetabular bone changes after total hip arthroplasty. Int Orthop
40. Silva MJ, Reed KL, Robertson DD, Bragdon C, Harris WH, Maloney WJ. Reduced bone stress as predicted by composite beam theory correlates with cortical bone loss following cemented total hip arthroplasty. J Orthop Res
41. Skedros JG, Mason MW, Nelson MC, Bloebaum RD. Evidence of structural and material adaptation to specific strain features in cortical bone. Anat Rec
42. Smolinski P, Rubash H. Bone remodeling around total hip implants. Crit Rev Biomed Eng
43. Sychterz CJ, Engh CA. The influence of clinical factors on periprosthetic bone remodeling. Clin Orthop Relat Res
44. Sychterz CJ, Topoleski LD, Sacco M, Engh CA Sr. Effect of femoral stiffness on bone remodeling after uncemented arthroplasty. Clin Orthop Relat Res
45. Thompson DD. Age changes in bone mineralization, cortical thickness, and Haversian canal area. Calcif Tissue Int
46. Torchia ME, Ruff CB. A quantitative assessment of cross-sectional cortical bone remodeling in the femoral diaphysis following hip arthroplasty in elderly females. J Orthop Res
47. Weeden SH, Paprosky WG. Minimal 11-year follow-up of extensively porous-coated stems in femoral revision total hip arthroplasty. J Arthroplasty
. 2002;17 (suppl 1):134-137.
48. Wick M, Lester DK. Radiological changes in second- and third-generation Zweymuller stems. J Bone Joint Surg Br