Long-term variability of markers of bone turnover in postmenopausal women and implications for their clinical use: The OFELY study

Bone marker variability has raised concern for its use in individual patients. Serum osteocalcin (formation) and CTX (resorption) were measured every year for 4 years in 268 postmenopausal women. Seventy percent to 80% of women classified as having high bone turnover at baseline were similarly classified by the same methods 4 years later. Introduction: High bone marker levels are a risk factor for osteoporosis in postmenopausal women, but variability of measurements has raised doubts about their clinical use in an individual patient. Methods: We studied 268 untreated postmenopausal women (50-81 years of age) belonging to a population-based prospective cohort. We collected fasting morning blood samples every year for 4 years to measure serum intact osteocalcin (OC) and serum C-terminal cross-linked telopeptide of type I collagen (CTX) as bone formation and resorption markers, respectively. Results: Serum OC and CTX remained stable during follow-up (+1.2%/year, p = 0.003 and -0.13%/year, p = 0.70 for OC and S-CTX, respectively). At baseline, women were classified as having low (tertile 1), intermediate (tertile 2), or high (tertile 3) bone turnover. Agreement of classification between baseline and 4-year measurements was moderate (kappa [95% CI]: 0.51 [0.43-0.59] and 0.52 [0.44-0.60] for OC and S-CTX, respectively). Less than 10% of women in tertile 1 or 3 of either marker at baseline were found in the opposite tertile 4 years later. When the two markers were combined, only 2% of women at high turnover at baseline - defined as OC and/or S-CTX in tertile 3 - were classified at low turnover 4 years later. Among women classified at high bone turnover at baseline (tertile 3), 70 - 80% were also found at high turnover 4 years later. Among women in tertile 2, only 51% and 43% for OC and CTX, respectively, remained in the same tertile at the second measurement. Conclusions: Serum levels of bone formation and resorption markers are stable over 4 years in postmenopausal women, on average. The majority of women classified as having high bone turnover were similarly classified by the same methods 4 years later. However, 20-30% of these women at risk for fracture would be incorrectly classified, suggesting that further investigation would be required to reduce the number of patients who would be treated unnecessarily if the decision was made on bone marker measurement. For women with intermediate levels, classification may be improved by a second measurement or by combining two markers.