AN INVESTIGATION OF SPINAL BONE-MINERAL DENSITY MEASURED LATERALLY - A NORMAL RANGE FOR UK WOMEN

A UK normal range for 250 volunteers was established for bone mineral density (BMD) of the lumbar spine (L2-L4) measured laterally in the de cubitus position. Two software defined regions of interest (''Body'' a nd ''Mid'') within the vertebral body were analysed throughout. As exp ected, a negative correlation of BMD with age was found for Body (r = -0.55, p < 0.001) and Mid (r = -0.56, p < 0.001). The age related bone loss from young to old (20-80 years) was 40% in L3 (Body) and 45% in L3 (Mid). In 22% of the cases only L3 could be measured owing to the i nfluence of rib over L2 and interference of the iliac crest over L4. A ge related normal ranges (+/-2 standard deviations) for the three lumb ar vertebra L2, L3, L4 for young normals (age 19-39 years) were found to be 0.54 to 1.02, 0.49 to 1.05 and 0.5 to 1.14 g cm(-2) respectively for the Body region and 0.49 to 0.97, 0.45 to 1.01 and 0.45 to 1.13 g cm(-2) respectively for the Mid region. These ranges can now be used as reference values for patients with suspected osteoporosis and possi bly for future fracture prediction. The in vivo precision in 19 volunt eers was found to be 4.2% and 5.6% on Body and Mid respectively. The s hort term (less than 1 week) in vitro precision was 3.1% and 2.7% resp ectively. From these data it appears that there is a greater measured age related drop in BMD in the vertebral body (measured laterally) tha n in the entire vertebra (measured anterior-posteriorly) indicating th at the lateral measurement may prove to be more sensitive in predictin g fracture. The precision of these results indicates that lateral meas urements of the spine are not yet useful for monitoring individuals ov er short term periods and are less useful for studying the effects of drug treatment than the more traditional anterior-posterior measuremen t of the spine and femur.