Distribution of Z-scores in a University cohort with an emphasis on "high" bone mineral density.

High bone mineral density (BMD) is currently not defined by the International Society for Clinical Densitometry with a specific Z-score cutoff; however, it has been suggested that a Z-score greater than or equal to 2.5 is not normal. Institutional Review Board approval was obtained. We evaluated a University dual-energy X-ray absorptiometry database over the previous 24 mo to define Z-score distributions. A Z-score greater than or equal to 2.5 was selected as the outcome event of interest in a logistic regression for adjusted odds ratio. The covariates were height; weight; body mass index (BMI); gender; menopausal status; use of female hormones; presence of insufficiency fractures after 50 yr of age; previous fractures; previous surgeries (back surgeries, vertebroplasty, or kyphoplasty); transplant history; presence of long-term chronic conditions (asthma, lupus, rheumatoid arthritis, or cystic fibrosis); eating disorder; current use of glucocorticoids; smoking status; and current and past use of osteoporosis pharmacological therapies. The study included a total of 8216 patients; 7212 (87.8%) were females, and 1044 (12.2%) were males. In the total population, 13.6% had a Z-score greater than or equal to 2.5 at the lumbar spine, femoral neck, or total hip. Only 0.2% of the males and 0.8% of the females had a Z-score greater than or equal to 2.5 at all 3 sites. The 97.5th percentiles for Z-scores in our population for men and women, respectively, were 3.4 and 3.9 at the lumbar spine, 1.5 and 2.1 at the femoral neck, and 1.6 and 2.2 at the total hip. The 99th percentile for Z-scores for men and women, respectively, were 4.9 and 4.7 at the lumbar spine, 2.4 and 2.7 at the femoral neck, and 2.2 and 2.7 at the total hip. At the lumbar spine, female gender and weight were found to be risk factors for a high Z-score (≥ 2.5). The use of glucocorticoids, bone-active medications, BMI, and smoking were significantly less likely to predict a lumbar spine Z-score greater than or equal to 2.5. A high total-hip Z-score is predicted by increasing weight, whereas those patients using bone-active medications were less likely to have high BMD at the total hip. At the femoral neck, there were no significant risk factors related to a Z-score greater than or equal to 2.5; those taking bone-active medications were significantly less likely to have a high Z-score. These data suggest that a high Z-score is common at 1 or more sites. Further research about the criteria for the diagnosis of high BMD is warranted.