This article is more than 5 years old. Information may no longer be current.
I am going to keep blogging until we get it right
Click Here to Manage Email Alerts
Every morning my inbox has a number of links to the latest relevant literature — a service I enjoy and usually find helpful. It doesn’t take much time to find and store links that might be of use now or sometime soon. But two articles this week made me see red.
The first reported on bone density in postpartum women within a few months of delivery. The study found that 17% of the women had the dreaded osteopenia and 0.3% had osteoporosis.
These terms were, and to my knowledge still are, intended to be applied to postmenopausal women and possibly to women who have a clinical situation (disease or therapy) known to be associated with bone loss.
At any age, bone density is normally distributed around a mean value. Ninety-five percent of the population will lie between +2.0 and –2.0 standard deviations (ie, Z-score in standard statistical parlance, T-score in the bone field) from the mean value. By statistical definition, 16% will be between –1.0 and –2.0, so the finding of 17% having “osteopenia” is simply misstating that they were normal and not entitled to any disease label. Two and a half percent of the population will have a value below –2.0 with less than 1% below –2.5. In other words, that 0.3% labeled as having “osteoporosis” were more than likely normal, healthy, recently postpartum women.
I know I have ranted before about misuse or misinterpretation of DXA data in healthy premenopausal women, but to repeat myself: DXA is not a meaningful tool in healthy premenopausal women. More specifically, the method of reporting as T- or Z-scores that may include a negative value 50% of the time does not indicate either bone loss or bone disease. One simply cannot make a dynamic statement on a single time point measurement.
DXA issues are not just related to clinical care but are sometimes fodder for strange clinical research. Even stranger — much of this gets published, which actually says more about the reviewers and editors than it does about the investigators.
The second article looked at a possible association between microscopic hematuria and bone density. I could not really work out why the question was even asked, but that is not my issue here. They found a statistically significant association between the red blood cell count in the urine and bone density. For all analyses, the RBC count accounted for less than 1% of the variance in bone density. Significant to be sure (P<0.01 or better), but biologically relevant? No way!!