Altered load history affects periprosthetic bone loss following cementless total hip arthroplasty

Dual energy x-ray absorptiometry was used to measure periprosthetic, distal femoral, and proximal tibial bone mass in the affected and contralateral limbs of eight patients 10 years after unilateral total hip arthroplasty with a cementless, porous-coated titanium alloy femoral stem. Gait analyses to assess the presence of asymmetries in loading of the lower extremities were also performed 10 years postoperatively. The patients had excellent clinical results and no other significant lower extremity pathology. On the basis of comparison of the affected and unaffected proximal femora, bone loss adjacent to the proximal medial aspect of the femoral stem was determined to be 34% (p < 0.001). However, the patients also had 16% less bone in the ipsilateral proximal tibia (p = 0.003) and 15% less bone in the ipsilateral femur 3 cm distal to the prosthesis (p = 0.007) compared with the contralateral limb. When normalized to the asymmetry in tibial bone mineral content, the estimated proximal medial periprosthetic bone loss was still statistically significant, but the magnitude was reduced from 34 to 17% (p = 0.009). The gait analyses indicated that several measures that influence the loads at the hip and knee joints were reduced in the involved limb compared with the contralateral limb. Furthermore, the bilateral difference in the vertical component of the external force acting on the proximal tibia was correlated with the bilateral difference in tibial bone mineral content (r = 0.80, p = 0.02). These data suggest that two mechanical factors, the local stress-shielding effect of the prosthesis and the global effect of decreased loading of the limb, can both make significant contributions to periprosthetic bone loss. It is apparent that the magnitude of the periprosthetic bone loss related to stress-shielding has been overestimated by as much as 50% in retrospective studies.