Bone marrow adipose tissue and bone metabolism: Bad to the bone?

In this thesis we explored the association between bone marrow adipose tissue (BMAT) and bone metabolism, to ultimately determine if BMAT might be a potential new imaging biomarker or potential new treatment target for osteoporosis. As patients with osteoporosis have low bone mineral density (BMD) combined with high BMAT, and bone marrow adipocytes (BMAds) and osteoblasts share a common progenitor, it is hypothesized that preferential differentiation towards adipocytes causes increased bone marrow adiposity and decreased bone formation. A second hypothesis is that BMAds have paracrine effects on bone metabolism. BMAT is a dynamic tissue. Neonates have little BMAT and during our life BMAT increases from the extremities in a centripetal way. In chapter 2 we show the specific pattern of BMAT distribution within the spine, pelvis, femur and tibia in a group of healthy subjects. BMAT increased from cranial to caudal within the spine, from proximal to distal within femora and showed a small, but consistent decrease from proximal to distal within tibiae. Furthermore, we show that the age-related increase in BMAT is gender and location dependent. As BMAT fatty acid unsaturation is associated with fractures, we also quantified BMAT fatty acid unsaturation. The association between BMAT and BMAT fatty acid unsaturation was opposed, depending on the location. Within the spine, BMAT and unsaturation were negatively correlated, while in the femora and tibiae BMAT and unsaturation were positively correlated. The gender specific differences in BMAT patterns observed in chapter 2 could possibly be mediated by sex steroids as estrogen treatment decreases BMAT and increases BMD. In chapter 3 we found no effect of raloxifene (a selective estrogen receptor modulator, registered for treatment of osteoporosis) on BMAT, adipocyte size or number, quantified in bone biopsies of postmenopausal women with osteoporosis. We found that BMAT volume and bone marrow adipocyte size both were negatively associated with osteoclast number, suggesting an association between BMAT and bone resorption. Furthermore, we found that women with osteoporosis and vertebral fractures had higher BMAT compared to women with osteoporosis without vertebral fractures, while there was no difference in bone volume between these groups, which could possibly indicate that BMAT is associated with fracture risk independent of bone volume. The negative association between BMAT and bone resorption could potentially be mediated by receptor activator of nuclear factor κ-B ligand (RANKL). Bone resorption is regulated by RANKL, which is expressed by osteocytes, osteoblasts and bone marrow precursor cells. In chapter 4 we showed that mature bone marrow adipocytes also express RANKL in a mouse model of postmenopausal osteoporosis (ovariectomy). Furthermore, we showed that estrogen deficiency caused by ovariectomy, not only increased BMAT, but also increased the percentage of RANKL positive BMAds. In chapter 5, we sought to inhibit adipogenesis by administration of a PPARγ antagonist, in the same mouse model of postmenopausal osteoporosis, to determine if we could prevent bone loss caused by estrogen deficiency. However, in these animals, administration of the PPARγ antagonist had no effect on BMAT, bone turnover, bone volume nor on bone strength. Postmenopausal estrogen deficiency is associated with increased visceral adipose tissue (VAT), and increased VAT is associated with increased BMAT and decreased BMD, suggesting an association between body composition and BMAT. Furthermore, patients with anorexia nervosa have high BMAT and low BMD. In chapter 6 we explore the effect of weight loss by gastric bypass surgery on BMAT and BMD in morbidly obese postmenopausal women. We found that both BMAT and BMD decreases after weight loss due to gastric bypass surgery, suggesting that BMAT does not contribute to bone loss after gastric bypass in postmenopausal women.