Common biochemical markers of bone turnover predict future bone loss: a 5-year follow-up study

Bone mineral density (BMD) is used to follow gain or loss of bone mass but cannot detect changes within a short period of time. Biochemical markers of bone turnover may be of value for prediction of individual bone loss.

We studied the relation between common inexpensive markers of bone turnover (serum alkaline phosphatase (ALP), osteocalcin (OC), urinary hydroxyproline (OHPr), and calcium (Ca)), BMD, age, and menopause in a combined cross-sectional and longitudinal design comprising 429 pre- and postmenopausal randomly selected women aged 21-79 years (mean 50 years). A follow-up was initiated after 5 years (including 192 of these women), which focused on changes in bone mass and the ability of these four common markers of bone turnover (sampled at baseline) to predict future bone loss.

A marked increase was observed for all markers at the beginning of menopause. During the postmenopausal period ALP and Ca decreased to near premenopausal levels, while OC and OHPr remained high even 15 years after menopause. We also found inverse correlations at baseline between the bone markers and BMD, independent of the selected marker or skeletal site, r=-0.14 to -0.46, P<0.05. The correlations between ALP, OC, OHPr, and subsequent bone loss over 5 years, was significant for arm, r=-0.23 to -0.36, P<0.01. Baseline levels of all bone markers correlated significantly at group level with the 5-year follow-up of BMD for all sites. The ability of markers to predict individual bone loss was estimated by a multivariate regression model, which included baseline BMD, age, and body mass index as independent variables. ROC analysis showed a validity of approximately 76% for the forearm model, but was lower for the hip (55%) and lumbar spine (65%).

These data show that the common inexpensive biochemical markers of bone turnover ALP, OC, OHPr, and Ca were related to the current bone mass and, moreover, provides information about future bone loss at the individual level. Future investigations should include an evaluation of the clinical relevance of markers of bone turnover in relation to fracture risk.