Maïmoun Laurent, Alonso Sandrine, Mahadea Krishna Kunal, Dubois Julien, Paunet Tom, Kucharczak Florentin, Nande Lisa Maïmoun, Boudousq Vincent, Mura Thibault, Mariano-Goulart Denis
Service de Médecine Nucléaire, Hôpital Lapeyronie, CHU Montpellier, France; Physiologie et Médecine Expérimentale du Cœur et des Muscles (PhyMedEx), INSERM, CNRS, Université de Montpellier (UM), France.
Service BESPIM (Biostatistique, Epidémiologie clinique, Santé Publique et Information Médicale), CHU Nîmes, Nîmes, France.
J Clin Densitom. 2025 Jan-Mar;28(1):101553. doi: 10.1016/j.jocd.2024.101553. Epub 2024 Dec 13.
The aim of this study was to investigate the correlations between areal bone mineral density (aBMD) and body composition measured by two dual-energy X-ray absorptiometers (DXA), the DMS Stratos® (STR) and the Hologic Horizon A® (HRZ), and then generate cross-calibration equations between the two scanners.
Repeat scans were obtained from 251 adults (85 % female), 36 ± 14 years old with mean body mass index (BMI) of 28.7 ± 11.1 kg/m2, using HRZ (fan-beam technology) and STR (pencil-beam technology). aBMD was measured at whole body [WB], femoral neck [FN], total hip [TH], lumbar spine [LS] and radius, while fat mass [FM] and lean tissue mass [LTM] were determined at whole body and at android and gynoid subregions.
Compared to HRZ, STR underestimated both aBMD at WB and radius and LTM at WB and android and gynoid regions. Conversely, STR overestimated aBMD at FN, TH, LS and FM at WB android and gynoid regions. Except for WB bone mineral content (r = 0.87) and WB aBMD (r = 0.84), there were strong correlations of aBMD and body composition between the two DXAs (r > 0.91; p < 0.0001). Several of the parameters that required the determination of specific cross-calibration equations because of the significant bias between the two DXAs were found to be influenced by BMI.
Although the data from the STR and the HRZ were highly correlated for aBMD and body composition parameters, a systematic measurement bias between two DXAs was observed. The development of cross-calibration equations fully corrected these differences and they may thus be useful for multicenter studies when scans are performed with STR and HRZ.