Dual X-ray absorptiometry (DXA) of the lumbar spine in a clinical paediatric setting: does the method of size-adjustment matter?

Abstract

Dual X-ray absorptiometry (DXA) is increasingly used in a clinical setting to evaluate bone mass in children. Areal Bone Mineral Density (aBMD) measurements are known to be influenced by body size, but there is no consensus on the optimal way to deal with this for individual patients. To compare parameters of bone mass with varying degrees of size correction and to determine the effect on the categorisation of patients as normal or abnormal. Healthy children (n = 78) and 4 groups of patients (n = 194) underwent DXA scans of the lumbar spine (L2-4, GE Lunar Prodigy). Five measures of bone mass were derived, all adjusted for age and sex: aBMD, BMAD (BMC/BA (1.5)), BMCh (BMC/height3), BMCa (BMC adjusted for BA), BMCt (BMC adjusted for BA and height). SD scores were calculated for each parameter for patients using data from healthy controls. Compared to healthy children, all patient groups had significantly reduced BMD SD scores (P < 0.001). Mean BMAD, BMCa and BMCt SD scores were significantly lower in only 2/4 patient groups, whilst BMCh SD scores were low only in one group. BMCt showed no advantage over BMCa. The proportion of patients with SD scores <-2 was 27% for aBMD but between 10-13% for BMAD, BMCh and BMCa. All size-corrected parameters of bone mass performed similarly and classified significantly fewer patients as abnormal than did aBMD. The use of one of these parameters should reduce the number of patients diagnosed inappropriately with 'low bone mass'. However, without validation against an outcome measure or 'gold standard' of bone density or structure, it is not possible to determine which parameter is most correct.