The strains produced in the proximal femur by noncemented and cemented PCA femoral implants have been compared to each other and to the strains in the same intact femur. The effect on the strain pattern of a hip-abductor muscle simulator was also tested. Nine embalmed femora were tested; two were used for development of the protocol, one was covered with a reflective photoelastic coating, and six were instrumented with eight strain gauge rosettes on each femur. For a given body weight on the photoelastically coated intact bone, the abductor-simulator increased the mean shear strains on both the medial (132%) and lateral (153%) aspects, with standard deviations of 13% and 20%, respectively; however, no strain-concentrated areas were observed. With an abductor simulator on the strain-gauged femurs, calcar shear strains were significantly reduced (P less than .01), from those on the intact bone, by both noncemented and cemented implants. These reductions were 74.2% and 91.8%, respectively, with no significant difference between the two techniques. Reductions in shear strain were noted at midstem on the medial side for the noncemented (20.5%) and cemented (35.9%) implants; however, only the cemented implants produced significantly less (P less than .05) strain than the intact bone at this location. At midstem on the medial aspect of the femur, there was a significant difference (P less than .05) between the data for the cemented and noncemented implants. Analysis of variance identified no other regions of significant change.
