BACKGROUNDIn children with unexplained fractures who are below the age of two years, it may be difficult to distinguish those with low bone mineral density (BMD) due to conditions such as osteogenesis imperfecta (OI) from those who have been abused. Currently, no imaging modality can readily or reliably assess BMD or evaluate bone strength in this age group.AIMTo investigate whether bone health index (BHI) and bone health index standard deviation scores (SDS) are sufficiently sensitive to distinguish between children under two years old with and without OI.METHODSIn this retrospective pilot study, we measured BHI and BHI SDS from 122 radiographs (33 OI, 89 suspected abuse) using BoneXpert software. Standard statistical methods (t-test, Pearson's correlation) were applied in addition to clinical diagnostics, sensitivity, specificity, and receiver operating characteristic (ROC) curves. An arbitrary level of p < 0.05 was assumed.RESULTSBHI was significantly greater in the group without OI compared to the group with OI, 3.75 and 3.41, respectively (p = 0.003). The percentage of children in the OI/non-OI groups with BHI ≤ 2.49, 2.5-2.99, 3-3.49, and ≥4 was 0 %/0 %, 27 %/7 %, 58 %/28 %, 18 %/29 %, and 12 %/36 %, respectively. While BHI SDS was significantly greater in the group without OI compared to the group with OI, -0.039 and -0.451, respectively (p = 0.01), BHI SDS was within the normal range (±2) for both groups.CONCLUSIONAlthough BHI SDS is lower in OI children, it remained within the normal range. Infants without OI had better volumetric bone mineral density, associated with stronger bones. This suggests BHI might be used to differentiate between young children with low BMD and those with healthy bones. Clinicians may find the cut-points established in this study useful for assessing the sensitivity and specificity of BHI in detecting OI and identifying individuals without OI. Further research is needed to assess BHI's clinical utility in this age group.