ECTS2013 Poster Presentations Osteoporosis: pathophysiology and epidemiology (49 abstracts)
1University of California, San Francisco, California, USA; 2Icelandic Heart Association, Kopavogur, Iceland; 3National Institute on Aging, Bethesda, Maryland, USA; 4Maine Medical Center Research Institute, Scarborough, Maine, USA; 5University of Iceland, Reykjavik, Iceland; 6Landspitali University Hospital, Reykjavid, Iceland.
Previous studies found a negative correlation between vertebral bone marrow fat (MF) and bone density (BMD). Proposed mechanisms for this include i) a shift in stem cell lineage allocation from osteoblasts towards adipocytes, and ii) an increase in osteoclast-promoting cytokines with greater MF. However, little is known about the relationship between MF and bone markers. To assess these relationships in older adults, we used data from the AGES-Reykjavik cohort. MF was measured in 301 participants with magnetic resonance spectroscopy (MRS; 1.5 Tesla) at L1L4 and expressed as ratio of fat to water plus fat (%). After excluding subjects with diabetes (n=17), inadequate serum (n=2), or bone-active medication use (n=44), analyses included 111 men and 127 women (mean age=79 years, mean BMI=27.6 kg/m2). Hip and spine scans were obtained using quantitative computed tomography (QCT). Blood was drawn fasting. Serum CTX, PINP, and sclerostin were batch assayed and compared to MF and BMD using Spearman rank correlations. There was a trend towards a negative correlation between MF and CTX in men (r=−0.18, P=0.063), but none of the correlations between MF and PINP or CTX were statistically significant. MF was correlated with sclerostin in men only (Table). Sclerostin correlated negatively with bone turnover markers and positively with BMD. The results for MF and sclerostin suggest a gender-dependent relationship between bone formation and marrow fat.
Vertebral MF | PINP | CTX | Trabecular spine vBMD | Trabecular hip vBMD | Cortical hip vBMD | |
Men (n=111) | 0.27* | −0.24* | −0.28* | 0.27* | 0.42* | 0.36* |
Women (n=127) | −0.06 | −0.29* | −0.32* | 0.54* | 0.27* | 0.41* |
Participants with diabetes excluded. Spearman rank correlations. *P<0.01. |