Fracture risk as determined by prospective and retrospective study designs

Area of Science:

• Gerontology
• Osteoporosis Research
• Biomedical Engineering

Background:

• Low bone mass and bone mineral content (BMC) are established risk factors for low-trauma fractures in postmenopausal women.
• Epidemiological studies using BMC measurements show overlap between fracture and non-fracture groups, questioning their discriminatory power.

Purpose of the Study:

• To evaluate the discriminatory ability of bone mineral content (BMC) measurements in predicting low-trauma fractures.
• To contrast the concepts of risk assessment with the sensitivity and specificity of BMC measurements.

Main Methods:

• A nested case-control study was conducted within a 24-year cohort of women at risk for osteoporosis.
• Relative risk and odds ratios were calculated for BMC z-scores.
• Receiver operating characteristic (ROC) curve analysis was used to assess sensitivity and specificity.

Main Results:

• For each 1.0 decrement in BMC z-scores, the adjusted relative risk for fracture was 1.67 in the prospective design.
• The odds ratio from recent BMC z-score measurements was 1.87.
• ROC curve analysis revealed low sensitivity and low specificity of BMC z-scores in detecting existing fracture status.

Conclusions:

• While low BMC is a risk factor, BMC z-scores alone demonstrate limited ability to accurately discriminate between women who will and will not experience fractures.
• The sensitivity and specificity of BMC measurements are insufficient for reliably predicting fracture status in postmenopausal women.