Accounting for Surgical Confounding Factors Affecting Dual-Energy X-ray Absorptiometry in a Large Clinical Trial

Joint replacements are among the most common orthopedic procedures performed in the U.S. and will continue to increase with the aging population. It is therefore necessary to account for these and other confounding factors, such as breast implants, when performing dual-energy X-ray absorptiometry (DXA) measurements. Whole-body DXA scans were performed in 771 participants (men ≥50 yr and women ≥55 yr) to assess bone mineral density (BMD) and body composition (fat and lean mass). In the DXA scan analyses of participants with internal metal, these affected regions of interest were replaced with measures from the unaffected, contralateral side, consistent with recommendations of the International Society for Clinical Densitometry. T-scores and Z-scores were recalculated using default sex and ethnicity-matched databases. We also explored effects of breast implants on bone density and body composition analyses. Approximately 13.1% of participants had internal metal artifacts at baseline. Replacing metal artifacts with the unaffected, contralateral side decreased the whole-body BMD by an average of 8.1% (SEM 0.84%; n = 67). In participants with unilateral hip (n = 17) and knee replacements (n = 20), BMD was decreased by an average of 14.1% (SEM 1.7%) and 11.2% (SEM 1.1%), respectively. Fat and lean mass were not significantly affected by metal artifacts, as differences between values with and without metal were within 1%. Two participants had bilateral breast implants, and in a separate trial, one participant had a unilateral breast implant. Bone mineral content (BMC) in the region with the breast implant was 5.8 times higher than the contralateral side, and whole-body BMC was increased by 4.7%. Metal artifacts and breast implants can confound DXA whole-body bone but not fat and lean results. It is therefore important in clinical studies to account for these factors to detect physiologically relevant differences in bone measures.