Abstract
BACKGROUND: The aim of this study was to compare quantitative ultrasound (QUS) and dual energy X-ray absorptiometry (DXA) for determining bone mineral density (BMD) among children living with HIV (CLWH) who were switching to second-line antiretroviral therapy (ART). METHODS: We conducted a cross-sectional study among CLWH as a sub-study of the CHAPAS-4 trial. Total body less head (TBLH) BMD and lumbar spine (LS) BMD were determined by DXA while the sound of speed (SOS), broad band ultrasound attenuation (BUA) and bone quality index (BQI) were determined by QUS. We evaluated the correlation between the DXA and QUS measurements using spearman correlation coefficient. RESULTS: A total of 167 children were enrolled; the median age was 9.4 (interquartile range = 6.0-12.0) years. Eighty-five (50.9%) were male. The median weight- for- age Z- score (IQR) was - 1.29(-2.16, -0.49) and height for age Z-score was - 1.03(-1.56, 0.01). SOS was weakly correlated with TBLH BMD R = 0.35, P < 0.001), lumbar spine bone apparent density (LSBMAD) (R = 0.19, P = 0.01) and LS bone mineral content (BMC) (R = 0.31, P < 0.001). BUA was moderately correlated with TBLH BMD (R = 0.50, P < 0.001), BMC (R = 0.47, P = 0.001), and LS BMC (R = 0.43, P < 0.001) but weakly correlated with LSBMAD (R = 0.28, P < 0.001). BQI was moderately correlated with TBLH BMC (R = 0.46, P < 0.001), TBLH BMD (R = 0.46, P < 0.001) and LSBMC (R = 0.41, P < 0.001). There was weak correlation between LSBMAD and BUA (R = 0.28, P < 0.001) BQI (R = 0.29, P < 0.001). QUS Z-score was weakly correlated with TBLH BMD and (R = 0.30, P < 0.01) and no correlation with LSBMAD (R = 0.07, P = 0.35). CONCLUSION: In CLWH, there was moderate correlation observed for TBLH measurements when comparing DXA to QUS, and weak correlation was found between LSBMAD and QUS measurements. There was a moderate correlation detected between the LSBMC and QUS.QUS may not be an appropriate substitute for DXA scan.