Bone Marrow Adipose Tissue Composition Following High-Caloric Feeding and Fasting

Abstract

Background:

Bone marrow adipose tissue (BMAT) plays a role in systemic energy metabolism and responds to nutritional changes. Chronic starvation as well as visceral adiposity are associated with BMAT accumulation. Two types of BMAT have been described which differ in anatomic location (proximal – regulated -rBMAT vs distal-constitutive-cBMAT) and composition (higher unsaturated lipids of cBMAT compared to rBMAT).

Objective:

To determine the response of BMAT composition to short-term high-caloric feeding and fasting. We hypothesized that high-feeding and caloric restriction would be associated with differences in BMAT composition according to the skeletal site.

Materials and Methods:

We examined 23 healthy subjects (13m, 10 f, mean age 33±7yrs, BMI 26±1.5kg/m²) who were admitted for a 10-day high-caloric stay (caloric intake with goal to achieve 7% weight gain) followed by discharge home for 13–18 days to resume normal diet (stabilization period), followed by a 10-day fasting stay (no caloric intake). Subjects underwent single voxel proton MR spectroscopy (1H-MRS) at 3T of the lumbar spine (L4) (rBMAT), the femoral diaphysis and distal tibial metaphysis (cBMAT) to determine BMAT composition (unsaturation index, UI and saturation index, SI). Within group comparisons were performed by the Wilcoxon signed rank test.

Results:

After the high-calorie visit, SI of L4 increased compared to baseline (0.62±0.27 to 0.70±0.28, p=0.02), and there was a trend of an increase in femoral SI and UI (p ≥0.07), while there was no significant change in tibial BMAT (p≥0.13). During the stabilization period, SI of L4 decreased (0.70±0.28 to 0.57±0.21, p<0.0001) and SI of the femoral diaphysis decreased (5.37±2.27 to 5.09±2.43, p=0.03), while there was no significant change in UI or tibial BMAT (p ≥0.14). During the fasting period, SI of L4 increased (0.57±0.21 to 0.63±0.30, p=0.03), while there was no change in UI (p=0.7). SI and UI of femoral diaphysis decreased (5.09±2.43 to 4.68±2.15, p=0.03, and 0.62±0.42 to 0.47±0.37, p=0.02, respectively) and UI of the tibial metaphysis decreased (1.48±0.49 to 1.24±0.57, p =0.04).

Conclusion:

1H-MRS is able to quantify BMAT composition during short-term nutritional challenges, showing a significant increase in SI of rBMAT during high caloric feeding and a differential response to fasting with an increase in SI of rBMAT and a decrease in SI and UI of femoral cBMAT and decrease in UI of tibial cBMAT.

Introduction

Recent studies have recognized the role of bone marrow adipose tissue (BMAT) as a novel biomarker of skeletal integrity and a regulator of systemic energy metabolism [1–4]. While bone mineral density and bone microarchitecture have traditionally been thought of as the main methods of estimating bone strength, recent studies have highlighted the role of the bone marrow microenvironment in skeletal integrity. In this context BMAT has been identified as a diagnostic marker and a potential target for the treatment of bone loss [5–9]. Bone marrow is one of the largest organs in the human body and contains mesenchymal stem cells which give rise to multiple cell types, including adipocytes, hematopoietic cells, osteoblasts, osteoclasts, and others [10] and studies have suggested that BMAT functions as an energy source for hematopoiesis and maintenance of bone homeostasis [11]. Bone marrow adipocytes have distinct lipid compositions, secretory and metabolic properties, depending on their location within the bone marrow [12].

Studies in animal models and humans have described two distinct types of BMAT which differ in their anatomic location and composition [12]. Regulated BMAT (rBMAT) is found in proximal locations, such as the axial skeleton. This fat depot is dynamic and responds to nutritional challenges and contains more saturated lipids. Constitutive BMAT (cBMAT) is found in distal anatomic locations, such as the appendicular skeleton, and is more inert and less responsive to nutritional challenges. cBMAT contains more unsaturated lipids [12].

Advances in MR imaging allow for the non-invasive longitudinal assessment of BMAT quantity and composition [13, 14]. Studies using proton MR spectroscopy (1H-MRS) have identified composition of BMAT as a novel imaging biomarker for fractures in postmenopausal women [15] and for metabolic risk in patients with type 2 diabetes [4].

BMAT is responsive to nutritional challenges, but its response differs compared to white adipose tissue (WAT). In contrast to WAT, BMAT accumulates in states of chronic caloric restriction, such as anorexia nervosa [1] and decreases with weight recovery [16]. In adolescents undergoing bariatric surgery, we have found differential response of BMAT. One year after sleeve gastrectomy, BMAT of the axial skeleton (lumbar spine, rBMAT) increased, while BMAT of the appendicular skeleton (femur and tibia, cBMAT) decreased [17]. However, chronic over-feeding, resulting in visceral obesity, is also associated with accumulation of BMAT [3]. We have recently demonstrated increased BMAT of L4 following both short-term high-calorie feeding and fasting and a reduction in BMAT of the lumbar spine and the femoral diaphysis during regular diet [18], but there are no data on BMAT composition following short-term controlled nutritional challenges.

The purpose of our study was to determine the response of BMAT composition to short-term high-caloric feeding and fasting. We hypothesized that high-caloric feeding and fasting would be associated with differences in BMAT composition according to the skeletal site.

Materials and Methods

This study was Institutional Review Board approved and Health Insurance Portability and Accountability Act compliant. Written informed consent was obtained from all subjects.

Subjects and Study Design

Subjects were recruited from the community through advertisements. Inclusion criteria were ages from 18 to 45 years and eumenorrhea in women. Exclusion criteria were a history of an eating disorder, diabetes mellitus, or other chronic illnesses, use of medication that could influence bone metabolism, estrogen use within the last three months, and contraindications to MRI such as the presence of a pacemaker or metallic implant. The study subjects and the protocol have been described previously [18]. In brief, subjects were admitted to the clinical translational research center for a 10-day high-caloric stay (day 1 to day 10) with the goal to achieve 7% weight gain. The macronutrient content of the diet consisted of 45–55% carbohydrates, 30–40% fat, and <25% protein. After the high-caloric stay, subjects were discharged home for 13–18 days to resume normal diet (stabilization period) (day 10–25), followed by a 10-day in-patient fasting stay (no caloric intake, water ad libitum) (day 25 – 35). Subjects received a multivitamin, cholecalciferol, and 20 meq of potassium chloride per day. Clinical characteristics and total BMAT of the lumbar spine and femur have been reported previously [18], however, no data on BMAT composition or tibial BMAT have been reported.

Proton MR Spectroscopy

Subjects underwent proton MR spectroscopy (1H-MRS) of the 4th lumbar vertebral body, the femoral mid-diaphysis, and the distal tibial metaphysis. All studies were performed on a 3.0 Tesla MR imaging system (Siemens Trio; Siemens Medical Systems, Erlangen, Germany) after an overnight fast. Single-voxel 1H-MRS data were acquired by using a point-resolved spatially localized spectroscopy pulse sequence with and without water suppression (TR/TE 3000/30, eight acquisitions, 1024 data points. For each voxel placement, automated optimization of gradient shimming was performed [2].

Fitting of all 1H-MRS data was performed using LCModel (version 6.3–0K, Stephen Provencher, Oakville, Canada). Metabolite quantification was performed using eddy current correction and water scaling. A customized fitting algorithm for bone marrow analysis provided estimates for all lipid signals combined (0.9, 1.3, 1.6, 2.3, 5.2 and 5.3 ppm). Olefinic protons at 5.2 and 5.3 ppm were used as an estimate of fatty acid unsaturated bonds; methylene protons at 1.3 ppm were used as an estimate of fatty acid saturated bonds. Total BMAT content was determined from the unsuppressed spectra by combining all lipid peaks (0.9, 1.3, 1.6, 2.0, 5.2 and 5.3 ppm). Saturation index (SI) was determined by obtaining a ratio between the methylene protons at 1.3 ppm and total lipid content Unsaturated lipid estimates were obtained from water suppressed spectra. The unsaturation index (UI) was determined by obtaining a ratio between the olefinic resonance at 5.2 and 5.3 ppm and total lipid content as previously described [15] (Figure 1).

Figure 1.

Lipid composition of bone marrow adipose tissue the lumbar spine at L4, the femur, and distal tibia. 1H-MR spectroscopy obtained at 3.0 T using water suppression demonstrates combined olefinic protons at 5.2 and 5.3 ppm, an estimate of unsaturated lipids (UL), and methylene protons at 1.3 ppm, an estimate of saturated lipids (SL).

Statistical Analysis

Statistical analyses were performed using JMP Statistical Discovery Software (Version 12, SAS Institute, Carey, NC). Changes within groups during high caloric feeding, stabilization, and fasting, were assessed using the Wilcoxon signed rank test. P < 0.05 was used to denote significance and p<0.1 was used to denote a trend. Data are presented as mean ± SD.

Results

Clinical Characteristics

Our study group comprised 23 healthy subjects (13 males, 10 females, mean age 33±7 years, range 22–44 years) who were of normal weight or overweight (mean BMI 26±1.5 kg/m², range 23.3–27.9 kg/m²). None of the subjects had a history of an eating disorder or metabolic disease, such as diabetes mellitus.

As previously reported, during the high caloric period, weight increased by 6.3±1.7% (p<0.0001). During the stabilization period weight decreased by −2.7±1.9% (p<0.0001), and during the fasting period weight decreased by −8.8±1.2% (p<0.0001) (Table 1).

Table 1:

Weight and bone marrow adipose tissue composition during the high-calorie visit, the stabilization period, and the fasting visit.

	High-Caloric Visit			Stabilization Period				Fasting Visit
Variable	Day 1	Day 10	p-value	Day 10	Day 25	p-value	p-value*	Day 25	Day 35	p-value
Weight (kg)	74.7±8.5	79.3±8.4	<0.0001	79.3±8.4	77.3±8.7	<0.0001	<0.0001	77.3±8.7	70.4±8.1	<0.0001
L4 Saturation Index	0.62±0.27	0.70±0.28	0.02	0.70±0.28	0.57±0.21	<0.0001	0.01	0.57±0.21	0.63±0.30	0.03
L4 Unsaturation Index	0.055±0.039	0.046±0.061	0.16	0.046±0.061	0.032±0.017	0.45	0.12	0.032±0.017	0.035±0.026	0.7
Femur Saturation Index	5.31±2.34	5.37±2.27	0.09	5.37±2.27	5.09±2.43	0.03	0.09	5.09±2.43	4.68±2.15	0.03
Femur Unsaturation Index	0.63±0.44	0.77±0.58	0.07	0.77±0.58	0.62±0.42	0.14	0.92	0.62±0.42	0.47±0.37	0.02
Tibia Saturation Index	9.30±1.84	9.41±2.07	0.47	9.41±2.07	9.11±1.34	0.60	0.99	9.11±1.34	8.49±2.22	0.28
Tibia Unsaturation Index	1.46±0.58	1.61±0.72	0.13	1.61±0.72	1.48±0.49	0.36	0.95	1.48±0.49	1.24±0.57	0.04

^*p-value compared to baseline

Saturation index was calculated as the ratio between saturated lipids and total lipids and unsaturation index was calculated as the ratio between unsaturated lipids and total lipids.

Bone Marrow Adipose Tissue

During the high caloric period, there was a significant increase in the saturation index of the axial skeleton (L4 - rBMAT) (p=0.03), while there was no significant change in the unsaturation index (p=0.16). In the appendicular skeleton (femur - cBMAT) there was a trend of an increase in the saturation index (p=0.09) and the unsaturation index (p=0.07). In the distal tibia (cBMAT), there was no significant change in total BMAT (p=0.41) or BMAT composition (p≥0.13).

During the stabilization period there was a significant reduction in the saturation index of L4 (p<0.0001), while there was no significant change in the unsaturation index (p=0.45). In the femur, saturation index decreased (p=0.03), while there was no significant change in the unsaturation index (p=0.14). There was no change in the total amount of BMAT (p=0.53) or composition of BMAT (p≥0.36) of the distal tibia. Of note, the saturation index of L4 decreased to lower values compared to baseline (p=0.01) and there was a trend of a lower reduction of the femoral saturation index compared to baseline (p=0.09).

During the fasting period there were opposite changes of the axial and peripheral skeleton. Saturation index of L4 increased (p=0.03), while there was no change in unsaturation index (p=0.7). However, in the femur saturation index and unsaturation index decreased (p ≤ 0.03). In the distal tibia, there was a significant decrease in the unsaturation index (p=0.04), while there was no significant change in total BMAT (p=0.32) or saturation index (p=0.28) (Table 1 and Figure 2).

Figure 2.

Bone marrow adipose tissue (BMAT) composition at multiple skeletal sites during the high caloric visit, the stabilization period, and the fasting visit. BMAT index was calculated as the ratio between saturated lipids or unsaturated lipids and total lipids.

SI: saturation index

UI: unsaturation index

*p-value compared to baseline

Saturation index was calculated as the ratio between saturated lipids and total lipids and unsaturation index was calculated as the ratio between unsaturated lipids and total lipids.

Discussion

Our study demonstrates that BMAT is a dynamic organ that changes its composition following short-term nutritional challenges, and these changes differ by skeletal site. We showed more pronounced changes in the proximal skeleton, at the 4^th lumbar vertebra, a regulated fat depot that has been shown to be more dynamic and responsive to systemic challenges, compared to the appendicular skeleton (lower extremity), which contains more constitutive BMAT which is more inert, and increases the more distal the anatomic site [12]. During the 10-day high caloric feeding visit we observed a significant increase in the saturation index of BMAT in the lumbar spine and a trend of an increase in the saturation and unsaturated index of the more proximal appendicular skeleton (mid femur), while there was no change in the distal skeleton (distal tibia), consistent with the previously reported responsiveness to nutritional changes of rBMAT and cBMAT. Of note, saturation index of the lumbar spine and femur dropped to below baseline values during the stabilization period. During the fasting visit we observed opposite changes in the axial and appendicular skeleton. In the lumbar spine, there was an increase in saturation index, while in the femur saturation index decreased. In the femur and tibia, there was also a significant decrease in unsaturation index. These results might shed further light on the potential role of BMAT in the regulation of energy metabolism.

While BMAT has long been hypothesized to be an inert filler of the bone marrow space in response to ageing or decreased demand for erythrocyte production, recent studies have revealed its important role in skeletal integrity and in the regulation of energy metabolism [11]. Studies have suggested that BMAT functions as an energy source for hematopoiesis and maintenance of bone homeostasis [11]. The quantity and composition of BMAT have been shown to be an important indicator of bone health, with high BMAT and low unsaturation index being a risk factor for fractures [15, 19]. BMAT also functions as an endocrine organ, secreting adiponectin and may play a role in the development of metabolic disease [4, 20].

Recent studies have shown that BMAT is a heterogeneous fat depot and has a different composition dependent on the anatomic site: proximal rBMAT is composed of single adipocytes interspersed with active hematopoiesis, while distal cBMAT contains larger adipocytes, has low hematopoiesis, develops earlier and remains preserved upon systemic challenges. These fat deports also show a different composition, with cBMAT showing a higher level of unsaturated lipids, increasing the more distal the anatomic site [12].

BMAT plays an important role in energy regulation and responds to nutritional challenges [1, 2, 17, 21]. For instance, the accrual of BMAT in response to caloric restriction is a unique behavior of this fat depot [3,4]. Of note, the response of BMAT is often reciprocal to WAT. In fact, chronic starvation, such as anorexia nervosa, or rapid weight loss following bariatric surgery is associated with an increase of BMAT [1, 17, 22], while weight recovery in patients with anorexia nervosa is associated with a reduction of BMAT [16]. Attane et al. have recently identified specific markers of bone marrow adipocyte lipid metabolism that might explain why BMAT behaves like a preserved lipid source during caloric restriction [23]. However, chronic over feeding, resulting in visceral adiposity, is also associated with high BMAT [2, 3].

We have recently shown that total BMAT increases following short-term high-caloric feeding and after fasting [18] but there are no data on BMAT composition. A study in adolescents undergoing sleeve gastrectomy has identified differential response of proximal rBMAT (lumbar spine) compared to distal cBMAT (femur and tibia). While the saturation and unsaturation index of the lumbar spine (rBMAT) increased 12 months after sleeve gastrectomy, the saturation and unsaturation index of the femur and tibia (cBMAT) decreased [17]. These results are similar to our results during the fasting visit which showed differential response of the axial and appendicular skeleton, with an increase in saturation index of the lumbar spine and a reduction of saturation and unsaturation index in the femur and reduction in unsaturation index in the tibia. Interestingly, the opposing condition to fasting, high-caloric intake, was also associated with an increase in saturation index of the lumbar spine and femur (trend), while there was no change in BMAT of the more inert distal tibia. Murine studies have shown increased BMAT following a high-fat diet [24]. We have previously shown that women with obesity and type 2 diabetes have a lower unsaturation index compared to non-diabetic controls [4], and that BMAT correlates positively with HbA1c [4, 20], highlighting the role of BMAT as an endocrine organ. Of note, during the stabilization period there was a significant decrease in saturated lipids of L4 and a trend of a decrease of saturated lipids of the femur below the bassline levels. This underscores the dynamic nature of BMAT and may reflect redistribution following episodes of acute nutritional challenges.

BMAT has been recently identified as a regulator of lipid synthesis, storage and release and insulin signaling. In fact, lipids in the bone marrow microenvironment may affect the accumulation of BMAT at different skeletal sites [11]. Our observed differential changes in BMAT composition following the different nutritional challenges, also further supports the notion that BMAT is a heterogenous fat depot with a unique response to specific nutritional challenges depending on the anatomic location. In addition, BMAT has been shown to play a role in the pathogenesis of skeletal fragility in obesity and anorexia nervosa. Our study suggests that BMAT functions beyond a regulator of bone homeostasis to include whole-body energy homeostasis, highlighting its role at the interphase of bone and fat metabolism. Given the known BMAT accumulation in metabolic diseases and conditions associated with bone loss, future therapeutic options for metabolic bone diseases could target BMAT content and composition.

Limitations of our study include the relatively small study population. However, with 23 subject we were able to detect significant changes in BMAT composition. In addition, the fasting visit followed the high calorie visit (after the stabilization period), per study design, and we do not know whether the observed changes would have been different if we had switched the nutritional interventions. Strengths of our study were the controlled in-patient caloric intervention and the use of advanced MR techniques to assess the composition of BMAT with quantification of saturated and unsaturated lipids at different skeletal sites.

In conclusion, 1H-MRS is able to quantify BMAT composition during short-term nutritional challenges, showing a significant increase in saturation index of rBMAT during high caloric feeding and a differential response to fasting with an increase in saturation index of rBMAT and a decrease in saturation and unsaturation index of femoral cBMAT and decrease in unsaturation index of tibial cBMAT. These results further support the notion that BMAT is a heterogenous fat depot with a unique response to specific nutritional challenges depending on the anatomic location.

Highlights.

• MR spectroscopy can quantify BMAT composition after short term nutritional changes

• High caloric feeding increases saturated lipids in the spine and femur

• Fasting increases saturated lipids in the spine

• Fasting decreases saturated and unsaturated lipids in the femur

• Fasting decreases unsaturated lipids in the tibia