Clinical application of dual photon absorptiometry (DPA) at the lumbar spine (LS) in the diagnosis of osteoporosis

This study evaluates the effectiveness of DPA to separate patients with osteoporosis (greater than 2 spinal fractures, normal Ca, P, absence of drugs, and metabolic bone disease) from a normal population. Performance criteria for the instrument have been described previously. Data was obtained from a prospective study of 105 normal women, 75 patients with osteoporosis and a retrospective study of 300 patients with osteoporosis seen in 1982/83. The results were as follows: (1) Area density (gm/cm/sup 2/) was found superior to mass (gm) due to the occasional problem to clearly identify the boundaries of L1-4. (2) Separation of the two populations was best when L1-L4,L2-L4, L3 alone or 10 paths over the mid lumbar area were used. One pass was not acceptable. (3) Compression fractures (CF) in the LS showed an increase in area density initially but area density may be undistinguishable from intact vertebrae later. To correct for this loss of bone area a factor predicting the area of lumbar vertebrae and based on patients actual height and weight was introduced and tested. (4) In the retrospective study a negative correlation was found between number of thoracic spine CF and bone mineral values in the LS. (5) A fracture threshold value of BM defined as the level below which 95% of all patients with CF were found was determined to be 0.98 g/cm2. Sixty-five percent of patients with two or more spinal CF could be separated from the normal population (outside 2SD). By using a correction factor for height loss this could be further increased to about 70%. CF in the LS may falsely elevate bone mineral values