Bone marrow composition, diabetes, and fracture risk: More bad news for saturated fat

In this month’s issue, Patsch and colleagues (1) offer ample food for thought for anyone interested inmarrow adipocytes, the enigmatic endosteal fat cells whose developmental origins, function, and impact on bone mass are currently subjects of active investigation. Marrow fat is often associated with low bonemass, as in anorexia nervosa, aging, and osteoporosis, but is also present in normal bone in complex age‐, sex‐, and site‐ specific patterns. Although marrow adipocytes were historically assumed to be neutral, quiescent space fillers, recent studies have demonstrated that marrow adipose tissue has a role in systemic energy metabolism and that bone marrow fat mass and bone mass are inversely correlated even within normal subjects. Marrow fat is now understood as a metabolically active depot that can be mobilized in starvation and that exerts potential beneficial as well as deleterious effects on skeletal and overall metabolism. However, many questions remain about exactly howmarrow fat might increase fracture risk, and how this risk varies with age, sex, and comorbidities. Patsch and colleagues tackle two key outstanding questions about marrow fat: whether the composition of marrow fat, as opposed to the quantity, is related to fracture risk; and whether there is an interaction betweenmarrow fat, diabetes, and skeletal fragility. To address these questions, the authors measured vertebral bone marrow adiposity via magnetic resonance imaging (MRI), areal bone mineral density (aBMD) by dual‐ energy X‐ray absorptiometry (DXA), and volumetric BMD (vBMD) by quantitative computed tomography (QCT) in the lumbar spine (L1–L3) in postmenopausal women with and without prior fracture history and with or without type 2 diabetes. Marrow fat quantification included both the total lipid and the proportions of saturated, unsaturated, and residual lipid, differentiated by the presence or absence of double bonds. Overall, the results corroborate patterns seen in previous studies: aBMD at the hip and spine was highest in women with type 2 diabetes but no fracture, and lowest in nondiabetic womenwith prior fracture, whereas vBMD of the spinewas lower in women with fracture versus women without fracture, both within controls and within diabetics. Interestingly, there were no differences in total marrow fat content among groups, regardless of diabetes or fracture status, in contrast to at least one recent study that found higher vertebral marrow fat in subjects with prevalent vertebral fractures. However, both diabetes and fracture history were correlated with differences in marrow fat composition. As shown in Table 2 in Patsch and colleagues, the degree of unsaturation was lower in individuals with both diabetes and fractures (DMFx) compared to controls without fracture (Co) and diabetics without fracture (DM), and the degree of saturation was higher in DM and DMFx versus Co. In other words, people with diabetes hadmore saturatedmarrow fat, and people with diabetes and history of fracture had less unsaturated marrow fat, compared to controls without fracture. However, while type 2 diabetes and fracture were each related to bone marrow adiposity, these correlations were independent of one another. After adjustment for age, ethnicity, and aBMD, therewas no interaction between type 2 diabetes, fracture, and total marrow fat, nor any of the subsets of marrow fat—saturated, unsaturated, or residual. Patsch and colleagues’ interpretation was that diabetes and fracture have additive, negative effects on bone marrow composition. Given the cross‐sectional study design, further studies will be needed to validate the direction of causality. These findings challenge our current understanding of the relationship between marrow fat and bone in several key ways, and generate a host of intriguing questions for future work. First and foremost is the provocative conclusion that it is the composition, rather than the amount, of marrow fat that is correlated with fracture history and diabetes. This finding corroborates previous studies that found lower unsaturation in osteoporosis and in diabetes, both of which involve elevated fracture risk, but contrasts with studies that have shown an inverse relationship between the quantity of marrow fat and BMD or bone mass. Additionally, this finding raises the issue of causality—does marrow fat become less unsaturated before a fracture, or afterward? Does altered metabolism in diabetes change marrow lipid composition and thereby increase