We read with interest the report on the effect of chemotherapy on bone mineral density (BMD) in children with acute lymphoblastic leukemia (ALL) using quantitative computed tomography (QCT) by Kaushik, et al.(1). Children with ALL are known to have lower BMD and a higher risk of fractures. Canadian STeroid-associated Osteoporosis in the Pediatric Population (STOPP) Research Program documented a 16% prevalence of vertebral fractures and every 1 SD reduction in lumbosacral BMD Z-score increased the odds for fracture by 80%(2). Thus, their results showing low BMD in 81% Indian children on treatment are interesting. We, however, would like to highlight our concerns regarding presentation and interpretation of data. Reference data are not sufficient for the clinical use of QCT for fracture prediction or diagnosis of low bone mass(3).

Authors have reported T-scores in their results and used the same to compare BMD and define osteoporosis. It would be incorrect to use cut-offs of T-scores (WHO criteria) used for postmenopausal women to diagnose osteoporosis in children. International Society for Clinical Densitometry (ISCD) guidelines(3) state that T-score has no relevance in a growing child who has not yet attained peak bone mass and as such should not be used in children. BMD values and z-scores, which are provided in the study, are rightly informative. T-score cannot be used to compare BMD either in between individuals or even in same individuals over time. BMD can be compared at diagnosis and follow-up of a child, which they have rightly provided. The appropriate method for comparing BMD would have been to measure change in BMD (Ä BMD) in each individual over 6 months; if greater than the least significant change (LSC), compare with zero and test for significance. In the same context, their statement "...81% had decrease in BMD and remaining had increase….", has no validity if they do not state that it was more than the LSC. The authors also mention that BMD increased in girls, but they have not mentioned ages of these girls. Was it that they entered puberty and had more increase in BMD surpassing the decrease resulting from disease and therapy?

We totally agree with authors that these children with ALL had decrease in BMD due to various factors described and understand the significance of the same in this group of children, but feel that the work could have been presented in a better way.