Total-body versus knee-specific dual-energy X-ray absorptiometry scans for assessing bone mineral density surrounding the knee joint in collegiate athletes with and without an anterior cruciate ligament reconstruction

Abstract

Following anterior cruciate ligament reconstruction (ACLR), athletes demonstrate reduced bone mineral density (BMD) surrounding the knee that persists for years. These changes have been identified using total-body and site-specific dual-energy X-ray absorptiometry (DXA) scans. It is unknown whether these two scan types provide comparable measures of BMD. This study assessed the agreement between total-body and knee-specific DXA scans for measuring BMD of the distal femur and proximal tibia in Division I collegiate athletes with and without an ACLR. We further assessed the influence of surgical hardware on BMD with and without the use of the manual artifact correction tool in athletes post-ACLR. Collegiate athletes (healthy: 46; post-ACLR: 20) were included. Regions of interest (ROI) were placed at the 5 % and 15 % lengths of the distal femur and the proximal tibia to assess BMD of each limb on total-body and knee-specific scans. BMD values for each limb and between-limb asymmetry were compared between scan types at each ROI using Bland-Altman analyses. Mean differences are reported as total-body minus knee-specific values. Total-body scans provided significantly lower BMD than knee-specific scans for all ROIs (-0.102 to -0.038 g/cm², all p-values < 0.01) except for the tibia 5 % (0.047 g/cm², p < 0.01) in healthy athletes. No differences in between-limb asymmetry were detected between scan types in healthy athletes (-0.78 % to 0.56 %; all p-values > 0.11). Athletes post-ACLR had more pronounced differences in BMD between scan types (-0.245 to -0.064 g/cm², all p-values < 0.02) and wider limits of agreement when surgical hardware was present, but this effect was mitigated when the manual artifact correction tool was used to exclude hardware (-0.122 to -0.008 g/cm², p-values from <0.01 to 0.57). Despite differences in absolute BMD, this study supports the use of total-body scans for assessing between-limb asymmetry in knee-specific regions in healthy athletes and in athletes post-ACLR when excluding hardware.