to the editor: The article, “Radiologic Bone Assessment in the Evaluation of Osteoporosis,”¹ was an excellent review of the subject. However, I question the logic used to formulate the authors' algorithm in (Figure 4) for assessing bone mineral density in women.

Discordance in bone mass measurements refers to the significant differences that can be found in bone density at various skeletal sites. Discordance is more prevalent in early post-menopausal women and becomes less of a problem as women age. Several studies² in women 50 to 59 years of age have shown that 8 to 12 percent are normal or osteopenic at one site, but osteoporotic at another site. This is not surprising considering that the composition and metabolism of bone varies from site to site. Early postmenopausal bone loss, mainly caused by estrogen deficiency, will affect the cancellous skeleton first and would be expected to be manifested initially at a central site (lumbar spine).³ In persons who are more than 65 years of age, bone density varies less between different skeletal sites, and peripheral measurements, such as calcaneal ultrasonography, will accurately reflect the condition of the skeletal areas of most concern (the hips and spine).

In my opinion, the author has gotten it backwards. Calcaneal ultrasonography is more helpful, not less helpful, in patients over 65 years of age. Conversely, I would be hesitant to reassure a woman who is early post-menopausal that her risk of spinal fracture is low based on a normal calcaneal ultrasonography, knowing that the chance this result is a false-negative is approximately 10 percent. Because dual-energy x-ray absorptiometry (DXA) scanning of central sites is readily available in my community at a reasonable cost, I do not order peripheral bone mass determinations in women under 65 years of age. In rural communities, this technology may be less readily available and peripheral bone mass measurements may need to be used to assess early postmenopausal women. A positive (osteoporotic) peripheral measurement would be highly predictive of low bone mass at the spine or hip. However, a negative (normal) calcaneal ultrasonography would not rule out the possibility of significant osteoporosis at a central site.

In reply: We appreciate the opinion of Dr. Wulfers. Many areas in the management and screening of osteoporosis remain uncertain. The vast majority of patients who will require therapy for osteoporosis are over 65 years of age. Only DXA allows for serial measurements to assess the efficacy of therapy. For patients under 65 years of age who do not have any risk factors for osteoporosis, we believe that calcaneal ultrasound is adequate for screening. While discordancy exists between central and peripheral bone sites, the majority of clinically relevant discordancy requiring therapy will be in patients over 65 years of age and in those under 65 years of age who have risk factors. We do not disagree with Dr. Wulfers about the limitations of a modality such as calcaneal ultrasound. However, to advocate calcaneal ultrasound in those with the greatest risk of developing fractures but consider it inappropriate for the lowest risk group, is a position we would disagree with.

Finally, we prefer DXA to calcaneal ultrasound because of its ability to assess central bone mass and to conduct serial measurements. However, the cost for DXA is higher, and no study that we are aware of has shown that DXA scanning is more predictive of fracture risk than calcaneal ultrasound.