The Association of Bone-related Biomarkers With Incident Hip Fracture: A Nested Case-control Study

Current osteoporosis risk stratification relies on clinical factors and bone mineral density alone.

To determine if osteocalcin, c-terminal cross-linking telopeptide of type 1 collagen, sclerostin, and bicarbonate ("bone-related biomarkers") are associated with future fracture risk or improve risk stratification.

Nested, matched case-control.

Longitudinal cohorts of health care workers.

Individuals with and without hip fracture in the Nurses' Health Study I and the Health Professionals Follow-up Study.

Hip fracture.

Among 642 women in Nurses' Health Study I (mean age, 70.3 years; 29% with osteoporosis), we found no consistent associations between bone-related biomarkers and incident hip fracture, and addition of biomarkers to clinical models predicting incident hip fracture did not improve model fit. Among 586 men in Health Professionals Follow-up Study (mean age, 63.8 years; <1% with osteoporosis), higher levels of osteocalcin (odds ratio, 0.37 [95% CI, 0.13-1.04] for quintile 5 vs quintile 1; P for trend = .02) and sclerostin (odds ratio, 0.22 [95% CI, 0.09-0.54] for quintile 5 vs quintile 1; P for trend < .001) were associated with lower risk of hip fracture; however, addition of sclerostin to clinical models predicting incident hip fracture provided limited additional predictive value.

Osteocalcin, c-terminal cross-linking telopeptide of type 1 collagen, sclerostin, and bicarbonate were not associated with incident hip fracture among older, predominantly White women. Osteocalcin and sclerostin were associated with hip fracture among men but did not meaningfully improve the predictive accuracy of models based on clinical risk factors alone.