William D. Leslie, MD, MSc; Sumit R. Majumdar, MD, MPH; Suzanne N. Morin, MD, MSc; and Lisa M. Lix, PhD

Ann Intern Med. Published online 19 July 2016 doi:10.7326/M15-2937

Background
Whether change in bone mineral density (BMD) is an accurate indicator of antifracture effect in clinical practice is unknown.

Objective
To evaluate repeated BMD testing as an indicator of treatment-related fracture risk reduction.

Design
Registry-based cohort study.

Setting
Manitoba, Canada.

Patients
6629 women aged 40 years or older initiating osteoporosis treatment with 2 consecutive dual-energy x-ray absorptiometry scans (mean interval, 4.5 years).

Measurements
Change in BMD between the first and second dual-energy x-ray absorptiometry scans categorized as stable, detectable decrease, or detectable increase. Incident fractures were ascertained from health services data.

Results
During a mean of 9.2 years, 910 (13.7%) women developed incident fractures, including 198 with hip fractures. After adjustment for baseline fracture probability, women with a detectable decrease in total hip BMD compared with stable BMD had an absolute increase of 2.9% (95% CI, 1.5% to 4.4%) and 5.5% (CI, 2.8% to 8.1%) in the 5- and 10-year cumulative incidence of any fracture, respectively. In contrast, risk for any fracture in women with a detectable increase in total hip BMD was 1.3% (CI, 0.4% to 2.2%) and 2.6% (CI, 0.7% to 4.5%) lower after 5 and 10 years, respectively. Consistent results were seen for change in femoral neck and lumbar spine BMD and across a range of subgroup analyses.

Limitation
Lack of standardization in the BMD testing interval.

Conclusion
Treatment-related increases in total hip BMD are associated with reduced fracture risk compared with stable BMD, whereas decreases in BMD are associated with greater risk for fractures. Monitoring BMD in clinical practice may help to identify women with a suboptimal response to osteoporosis treatment.