Abstract
Fifty-nine revisions that were done for aseptic acetabular loosening after 6,128 total hip arthroplasties for degenerative arthritis or traumatic arthritis were studied. These revisions were in forty-four (approximately 1 per cent) of 4,576 hips that had a twenty-two-millimeter femoral-head component, in two of 520 that had a twenty-eight-millimeter femoral-head component, and in thirteen (approximately 2.5 per cent) of 487 that had a thirty-two-millimeter femoral-head component. Therefore, the thirty-two-millimeter femoral component was associated with the highest rate of acetabular revision (p less than 0.001). The dimensions of the acetabular wall were thinner in the hips that had the thirty-two-millimeter component than in those that had the twenty-two-millimeter component (p less than 0.05). Multivariate analysis demonstrated a significantly increased risk of acetabular loosening in men and in patients who were less than sixty years old.