Implications in the use of T-scores for the diagnosis of osteoporosis in men.

Abstract

Osteoporosis is recognized as a disorder of both men and women. However, the World Health Organization's (WHO) definition of osteoporosis (a bone mineral density [BMD] T-score of -2.5 or less) was formulated for use with postmenopausal women only. In the absence of a BMD-based definition for male osteoporosis, the WHO definition is often applied to men as well. Several important questions exist when considering the use of T-scores in men. First, is the WHO definition appropriate for men? What is the impact of using a -2.5 criteria, in terms of the number of men that would be identified as osteoporotic? When calculating T-scores in men, should male or female young normal values be used? Can the same T-score criteria be used for all skeletal sites and technologies? To address these questions, osteoporosis prevalence estimates for men aged 50 yr and over were generated using WHO methods and manufacturer normative data from dual-energy X-ray absorptiometry (DXA), quantitative computed tomography (QCT), and ultrasound. Estimates were determined for several skeletal sites and technologies using both male and female young normal values. Prevalence estimates were compared to published fracture risk estimates. Mean T-scores declined with age at all measurement sites. Discrepancies were found between the different skeletal sites and techniques, similar to the previously reported differences in women. A -2.5 criterion (based on young normal males or females) appeared to underestimate the prevalence of osteoporosis, except for QCT, which seemed to overestimate risk. Depending on the technique used, 0 to 12.5 million US men 50 yr of age and older would be classified as osteoporotic using the WHO definition. T-Scores based on male norms were less discordant across skeletal sites than female-based T-scores. Male-based T-scores between -1.8 and -2.3 using DXA and ultrasound and -3.1 for QCT provided osteoporosis prevalence estimates that approximated the likelihood of common fractures in men 50 and over. We conclude that the use of single T-score-based criterion for the diagnosis of osteoporosis in men has many potential difficulties. BMD measurement techniques provide discrepant estimates of prevalence and may underestimate the size of the male population at risk for fracture. Based on available normative data, a -2.5 criterion underestimates osteoporosis prevalence in men, whether based on male or female norms. Prospective studies are needed to further refinement to the BMD definition of osteoporosis in men.