Obesity Surgery

, Volume 18, Issue 9, pp 1134–1143

The Bone-Adipose Axis in Obesity and Weight Loss


  • J. Gómez-Ambrosi
    • Metabolic Research Laboratory, Clínica Universitaria de NavarraUniversity of Navarra
    • CIBER de Fisiopatología de la Obesidad y Nutrición (CIBEROBN)Instituto de Salud Carlos III
  • A. Rodríguez
    • Metabolic Research Laboratory, Clínica Universitaria de NavarraUniversity of Navarra
    • CIBER de Fisiopatología de la Obesidad y Nutrición (CIBEROBN)Instituto de Salud Carlos III
  • V. Catalán
    • Metabolic Research Laboratory, Clínica Universitaria de NavarraUniversity of Navarra
    • CIBER de Fisiopatología de la Obesidad y Nutrición (CIBEROBN)Instituto de Salud Carlos III
    • Metabolic Research Laboratory, Clínica Universitaria de NavarraUniversity of Navarra
    • CIBER de Fisiopatología de la Obesidad y Nutrición (CIBEROBN)Instituto de Salud Carlos III
    • Department of Endocrinology, Clínica Universitaria de NavarraUniversity of Navarra

DOI: 10.1007/s11695-008-9548-1

Cite this article as:
Gómez-Ambrosi, J., Rodríguez, A., Catalán, V. et al. OBES SURG (2008) 18: 1134. doi:10.1007/s11695-008-9548-1


Body fat and lean mass are correlated with bone mineral density, with obesity apparently exerting protection against osteoporosis. The pathophysiological relevance of adipose tissue in bone integrity resides in the participation of adipokines in bone remodeling through effects on deposition and resorption. On the other hand, the skeleton has recently emerged as an endocrine organ with effects on body weight control and glucose homeostasis through the actions of bone-derived factors such as osteocalcin and osteopontin. The cross-talk between adipose tissue and the skeleton constitutes a homeostatic feedback system with adipokines and molecules secreted by osteoblasts and osteoclasts representing the links of an active bone–adipose axis. Given the impact of bariatric surgery on absorption and the adipokine secretory pattern, to focus on the changes taking place following surgical-induced weight loss on this dynamic system merits detailed consideration.


Bone massObesityAdipose tissueAdipokinesOsteokines


The skeleton is submitted to complex structural renewal processes throughout life to guarantee its correct physiology. The balance between bone deposition and resorption is determinant for an adequate development and maintenance of bone size, shape, and integrity [1]. While osteoclasts resorb preexisting mineralized bone, osteoblasts construct an extracellular matrix that will be subsequently mineralized. Imbalance between bone deposition and resorption leads to pathologic conditions of either high (osteopetrosis) or low bone mass (osteopenia and osteoporosis) [1]. Osteoporosis, the most common bone remodeling disease, is characterized by enhanced bone fragility as a consequence of a reduction in both bone quantity and quality [2].

Obesity is defined medically as a state of increased adipose tissue of sufficient magnitude to produce adverse health consequences [3]. Abundant epidemiological evidence has established a link between body weight and composition, with bone mass regulation in humans [2, 4]. Both body fat mass and fat free mass are directly correlated with bone mineral density (BMD). Overweight and obesity have been proposed to exert a protective role in the development of osteoporosis [5]. On the contrary, low body mass index (BMI) confers a risk for low bone quality, osteoporosis, and all fractures that is largely independent of age and sex [6]. The protective effect of obesity on bone integrity has been explained through several mechanisms. The increase in BMD associated to obesity may be firstly attributed to the mechanical load. However, the protective effects of weight have also been observed in non-weight-bearing bones suggesting the implication of other factors [7]. Another potential explanation is the increased secretion of bone-protecting hormones from the pancreatic β-cell, such as insulin and amylin [2, 4]. In addition, adipokines secreted from adipose tissue seem also to participate in bone mass regulation as detailed below. Finally, adipose tissue is the major place of conversion of androgens to estrogens in both elderly men and women [8]. Therefore, the bone-preserving action of estrogens may underlie the increased effect of body weight in women and why obese women do not lose bone as quickly as their lean counterparts after menopause [9]. Although the protective effect of overweight and obesity seems clear, there is still controversy as to which compartment, fat mass or lean mass, is more important in determining BMD [2, 5]. It is likely that both fat mass and lean mass contribute to this relation [4].

Cross-talk Between Fat Mass and Bone

Until recently, the adipocyte had been only considered a passive cell for the storage of excess energy as triglycerides. It has, however, been clearly established that adipocytes secrete a wide variety of biologically active molecules, thus representing an extremely active endocrine organ [3, 10, 11]. Adipose tissue releases a wide variety of proteins, called adipokines, including leptin, adiponectin, tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), visfatin, and resistin, among others, which are known to be involved in the complex regulation of bone physiology [4, 1214]. On the other hand, the skeleton has recently emerged as an endocrine organ with potential effects on body weight, energy expenditure and glucose homeostasis [15]. Figure 1 represents the putative “endocrine” interplay between adipose tissue and bone with adipokines and molecules secreted by osteoblasts and osteoclasts (which potentially could have endocrine actions and we term “osteokines”) being the links of a bone–adipose axis.
Fig. 1

Schematic diagram showing the “endocrine” interplay between adipose tissue and bone with adipokines and osteokines as the active players

Adipokines Influencing Bone Physiology

The pathophysiological relevance of adipokines in bone physiology resides in the participation of these molecules beyond body weight balance in bone homeostasis through effects on bone deposition and resorption (Table 1). In this sense, adipokines have been shown to be implicated either directly or indirectly in the regulation of bone remodeling [4, 1214].
Table 1

Main adipokines with relevance in bone metabolism


Effect on bone physiology



Restores skeletal growth in animals under caloric restriction. Conflicting results indicate that leptin may induce either bone formation or resorption which could depend on a bimodal threshold response or in the different central vs peripheral effects of leptin.

[17, 21, 26]


Negative effect on bone formation by an indirect induction of osteoclast formation via RANKL stimulation and inhibition of osteoprotegerin production in osteoblasts. However, a positive effect through the suppression of osteoclasts and the stimulation of osteoblasts has also been described. In humans, most published studies support its role as a negative regulator of bone mass in both men and women, possibly by the promotion of excessive bone resorption.

[13, 14, 32, 3437, 39, 40]


Stimulates osteoclastogenesis and the proliferation of osteoblasts. Significant negative determinant of lumbar spine BMD in humans. Other studies find no effect.

[12, 14, 43]


Stimulates glucose uptake, proliferation and production of type I collagen by human osteoblasts closely resembling the effect of insulin in these cells. No apparent relation to BMD in humans.

[14, 45]


May be considered as a bone-resorptive factor. Associated either with lower or unchanged bone mass.



Stimulator of bone resorption in vitro and in vivo. Evidence of a direct stimulatory effect on osteoclastogenesis.


IL-6 interleukin 6; TNF-α tumor necrosis factor-α.


Leptin is a hormone mainly produced by adipocytes in proportion to fat size stores [10, 11]. It was originally thought to be only involved in food intake and body weight regulation acting at its hypothalamic receptors. The full-length leptin receptor, OB–Rb, activates the JAK/STAT signal transduction pathway [16]. Leptin receptors are expressed in almost all tissues stressing its high pleiotropism and involvement in energy expenditure, reproduction, angiogenesis, immunity, wound healing, and cardiovascular function [10, 16]. Plasma leptin concentrations are increased in obese patients being strongly correlated with the percentage of body fat [10, 11].

Several reports have shown that leptin directly stimulates bone growth in vitro and increases bone density in leptin-deficient animals [1719], which would be in accordance with its mitogenic effect [16]. However, central leptin administration to mice causes bone loss via the hypothalamus [20, 21]. This effect seems to take place through the regulation of bone formation and resorption by the β2-adrenergic receptor [22], the neuropeptide cocaine and amphetamine-regulated transcript (CART) [23] and the clock genes, which determine circadian rhythms [24]. These findings could depend on a bimodal threshold response, on the different central vs peripheral effects of leptin or on the skeletal region [25, 26]. Expression of leptin by osteoblasts together with the fact that leptin receptors are present in bone marrow stem cells, osteoblasts and osteoclasts provide evidence for a potential direct effect of leptin on these cells [19, 27]. Leptin treatment has shown to inhibit osteoclastogenesis through an increase in osteoprotegerin and a decrease in receptor activator of nuclear factor-κB ligand (RANKL) and to induce osteoblasts differentiation [27].

Epidemiological studies have reported controversial results on the relation between leptin and bone density in humans [26]. Plasma leptin concentrations have been shown to be strongly related to BMD in men and women [28, 29]. However, when these analyses are adjusted for fat mass leptin is either positively [30] or negatively [29] associated with BMD. Thus, future interventional studies are needed to fully elucidate the role of leptin in bone metabolism in humans.


Adiponectin, which is highly expressed in adipose tissue, is downregulated in obesity and exerts a wide variety of physiological effects including a direct insulin sensitizing activity, a protective role in the development of cardiovascular disease, and a regulatory action on food intake and energy expenditure [31].

Adiponectin and its receptors are expressed in osteoblasts and osteoclasts indicating the possibility of adiponectin acting on bone not only as a hormone, but also in an autocrine/paracrine fashion [32, 33]. Circulating adiponectin has a negative effect on bone formation by an indirect induction of osteoclast formation stimulating RANKL and inhibiting osteoprotegerin production in osteoblasts [34]. However, mice lacking adiponectin exhibit a normal bone phenotype. This finding has been explained by a compensatory positive autocrine effect [33]. However, other authors have described a positive effect on bone formation through the suppression of osteoclasts [35, 36] and the stimulation of osteoblasts [32, 37]. In humans, most of the published studies indicate adiponectin as a negative regulator of bone mass in women and men [13, 14, 3840]. This negative effect may be mediated by the promotion of bone resorption [14]. However, other studies have shown a positive effect of adiponectin on BMD [41] or no effect [12].


Resistin is an adipokine which was initially proposed as a link between increased adiposity and T2DM. Some studies indicate that T2DM patients have increased circulating concentrations of resistin, although these results are not univocal [42]. Increased resistin concentrations have been described in patients with severe inflammatory disease. However, the precise physiological role of resistin in the pathogenesis and perpetuation of inflammation and cardiovascular disease remains unclear [42].

Resistin is expressed in osteoblasts and osteoclasts [43]. In vitro murine studies have demonstrated that recombinant resistin exerts a stimulatory effect on osteoclastogenesis, at the same time as increasing the proliferation of osteoblasts. This seems to indicate that resistin may play a role in bone remodeling [43]. In humans, circulating resistin is a significant negative determinant of lumbar spine BMD in middle-aged men [12], while another study also including men failed to replicate this finding [14]. Further research is needed to better understand resistin’s true participation in bone homeostasis.


Visfatin, previously identified as colony-enhancing factor of pre-B cells, is highly secreted by visceral fat of both mice and humans with its expression level in serum increasing during the development of obesity [44]. It has been reported that visfatin seems to have insulin-like activity and to bind to the insulin receptor, thereby lowering blood glucose concentrations. However, the role of visfatin in obesity and glucose homeostasis in humans remains to be fully elucidated [44].

Data regarding the effect of visfatin on bone metabolism are scarce. Visfatin stimulates glucose uptake, proliferation and the production of type I collagen by human osteoblasts closely resembling the effect of insulin in these cells [45]. However, circulating concentrations of visfatin are apparently unrelated to BMI in men [14].


Interleukin-6 is an inflammatory mediator with effects on a variety of tissues, including regulation of glucose and lipid metabolism [10]. It has been reported that adipose tissue produces IL-6 with circulating concentrations being proportional to the fat mass and the degree of insulin resistance [46]. IL-6-deficient mice develop mature-onset obesity, with the obese phenotype being only partly reversed by IL-6 replacement [47]. Interestingly, acute IL-6 administration produces an increase in insulin-stimulated glucose uptake in humans in vivo and induces fatty acid oxidation in vitro [46].

IL-6 has been considered as an osteoresorptive factor [48, 49]. In this sense, it has been suggested that high IL-6 concentrations may induce lower bone mass [50] and that serum IL-6 is a predictor of postmenopausal bone loss [51]. However, other authors have reported that IL-6 concentrations are unchanged in osteoporotic women [52]. Analysis of the role of IL-6 in bone homeostasis has to take into account the functionality of its soluble receptors as well as the interaction with other related factors sharing the same signaling pathways such as leptin [49].


TNF-α is a cytokine implicated in the metabolic disturbances of chronic inflammation with its biological actions including induction of insulin resistance, anorexia, and weight loss. It is overexpressed in adipose tissue, which is both a source of and a target for TNF-α [10, 46].

TNF-α has been shown to stimulate bone resorption in vitro and in vivo being also implicated in diseases with increased osteoclastic bone resorption [53, 54]. In addition, direct evidence of a stimulatory effect by TNF-α on osteoclastogenesis has been found in both human [55] and mouse [56] bone marrow cells. The increased circulating concentrations of TNF-α found in obesity [57] are difficult to reconcile with the protective effect of obesity on BMD. Further studies analyzing the precise role of TNF-α in bone physiology are warranted.

Osteokines Exerting an Impact on Adipose Biology

The skeleton has emerged as a previously ignored player in energy balance control. The regulation of bone remodeling by leptin leads to hypothesize that bone exerts a role in the feedback control of energy homeostasis [15]. In this sense, it has been shown that mice lacking the protein tyrosine phosphatase OST-PTP are hypoglycemic and are protected against obesity. In contrast, mice lacking the osteoblast-secreted molecule osteocalcin (OCN) exhibited an increased adiposity and insulin resistance [15, 58]. These findings suggest that bone-derived factors (osteokines) exert an endocrine regulation on glucose homeostasis and body weight (Table 2).
Table 2

Main osteokines with relevance in adipose tissue metabolism


Effect on adipose tissue physiology


Osteopontin (OPN)

Increased adipose tissue expression in human and murine obesity. Plays a role in the development of obesity-associated low-grade chronic inflammation and insulin resistance.


Osteoprotegerin (OPG)/RANKL

Expressed in rat adipose tissue from different locations and increasingly expressed during adipocyte differentiation. Potential protective role against inflammation and angiogenesis in adipose tissue.


Osteocalcin (OCN)

Regulates body adiposity. Improves glucose tolerance increasing insulin expression in β-cells and adiponectin in adipose tissue.


Osteonectin (SPARC)

Serum levels and expression in adipose tissue is increased in obesity. Expression correlates with adipocyte hyperplasia suggesting that it may play a role in adipocyte growth and differentiation.

[77, 79, 80]

RANKL receptor activator of nuclear factor-κB ligand; SPARC secreted protein, acidic, and rich in cysteine.


Osteopontin (OPN), also known as early T lymphocyte activation (Eta-1), secreted phosphoprotein-1 (SPP1), and bone sialoprotein-1, is a phosphoprotein expressed by a wide variety of cell types, such as osteoclasts, macrophages, hepatocytes and vascular smooth muscle cells, among others [59]. OPN has important roles in bone turnover serving as attachment for osteoclasts activating the resorption cascade [59, 60]. In addition to bone remodeling, OPN is also involved in several pathophysiological processes including immunity, inflammation, neoplastic transformation, progression of metastases, promotion of cell survival, wound healing, and cardiovascular function. OPN has recently been shown to be also produced by adipocytes and to be involved in obesity and T2DM [11, 6163]. Moreover, OPN has been suggested to play a pivotal role linking obesity to insulin resistance development by promoting inflammation and the accumulation of macrophages in adipose tissue [62].


OCN is a noncollagenous protein marker of osteoblastic activity thought to play a role in mineralization and calcium homeostasis [64]. OCN is secreted mainly by osteoblasts and is reduced in starvation, malnutrition, and anorexia nervosa. OCN has been traditionally considered as a biological marker of bone formation [64].

As mentioned above, mice lacking OCN exhibited an increased body fat as well as insulin resistance. OCN is able to improve glucose tolerance in vivo through the stimulation of the expression of insulin and β-cell proliferation and the induction of the expression of adiponectin and genes involved in energy expenditure in adipocytes [15, 65].


Osteoprotegerin (OPG) is a secreted glycoprotein member of the TNF-α receptor superfamily [66]. OPG together with RANKL and its receptor RANK have been involved in the control of bone resorption [67]. Exogenous administration of OPG to normal mice increases BMD and bone volume leading to profound osteopetrosis accompanied by a decrease in osteoclast differentiation. OPG inhibits osteoclasts differentiation and osteoclastic bone resorption interfering with RANKL binding to RANK [67]. OPG has been used in clinical trials to decrease bone resorption in postmenopausal osteoporosis. The role of OPG as a regulator of bone metabolism is underscored by the observation of osteoporosis in OPG-deficient mice [68].

Serum OPG concentrations are significantly lower in obese patients than in normal-weight controls [69]. Furthermore, OPG has been related to insulin resistance and inflammation in obese patients [70]. In addition, serum OPG is associated with the presence and severity of coronary artery disease [71], T2DM [72], and atherosclerosis [73]. Interestingly, OPG/RANKL are expressed in rat adipose tissue from different locations, and its expression increases during adipocyte differentiation [74]. Moreover, it has been demonstrated that OPG is detected also in human adipose tissue, and that its levels are increased in obesity [75]. It is suggested that OPG may have a protective role against inflammation and angiogenesis in adipose tissue [74, 75]. However, the role of OPG in adipose biology remains to be fully clarified.


Osteonectin, also known as secreted protein, acidic, and rich in cysteine (SPARC) is an extracellular matrix glycoprotein secreted by osteoblasts during bone formation [76]. In vitro studies indicate that osteonectin binds collagen and hydroxyapatite and can regulate cell proliferation and cell–matrix interactions, particularly during matrix remodeling, to regulate cell interaction with the extracellular milieu during development and in response to injury [76]. It is abundant in bone and is expressed also by other cell types in areas of active remodeling outside the skeleton [77]. Osteonectin-deficient mice exhibit decreased bone formation and osteopenia [78].

Serum levels and expression in adipose tissue of osteonectin is upregulated in obesity [79]. Furthermore, osteonectin-null mice exhibit increased adiposity without significant differences in overall body weight [77]. Finally, osteonectin expression correlates with adipocyte hyperplasia suggesting that it may play a role in adipose tissue physiology regarding adipocyte growth and differentiation [80].

Effect of Body Weight-Regulating Drugs on Bone

Weight loss affects bone mass depending on the way weight reduction is achieved. Caloric restriction-induced weight loss seems to be a risk factor for rapid bone loss [81]. On the contrary, physical activity-induced weight loss preserves BMD. However, because the amount of exercise needed to successfully decrease weight is high, the preferred clinical recommendation for weight reduction is a combination of diet and exercise [81].

Several antiobesity drugs are currently licensed for long-term use [82]. Orlistat, a gastrointestinal lipase inhibitor which reduces dietary fat absorption, induces small but significant weight loss of around 3 kg on average; adverse gastrointestinal effects are common [82]. Short-term orlistat treatment does not affect mineral balance and bone turnover in obese men [83]. In the long term, orlistat induces a small increase in bone resorption, possibly due to malabsorption of vitamin D and/or calcium [84]. Given the reduction in the absorption of fat-soluble vitamins, concomitant administration of vitamin and mineral supplements is recommended. However, no changes in bone mass or density are seen 1 year after orlistat treatment apart from those explained by the weight loss itself [84].

Sibutramine, a serotonin-reuptake inhibitor, results in mean weight losses of 4–5 kg but is associated with increases in blood pressure and pulse rate [82]. There is no scientific evidence of an effect of sibutramine on bone homeostasis. However, the use of other selective serotonin-reuptake inhibitors has been associated with low BMD [85] and with an increased risk of fracture [86].

A third antiobesity drug, rimonabant is approved in Europe but has not been approved in the USA [82]. Rimonabant is a selective CB1 endocannabinoid receptor antagonist with appetite-inhibiting effects, but also with peripheral effects through its ability to directly target adipocytes, the gastrointestinal tract, and skeletal muscle [87]. The CB1 cannabinoid receptor has been involved in the regulation of bone remodeling [88, 89]. Thus, interventional studies are needed to address the potential effect of rimonabant on bone homeostasis.

The insulin-sensitizing thiazolidinediones (TZDs), which are selective ligands of the peroxisome proliferator-activated receptor γ are drugs frequently used for the treatment of insulin resistance in patients with T2DM [90]. Its use is generally accompanied by weight gain and an increase in the subcutaneous adipose-tissue mass [90]. Paradoxically, it has been shown that TZDs contribute to bone loss and can lead to excess incidence of fractures in T2DM [91, 92] in part by inhibiting osteoblastic differentiation [93] and promoting osteoclast differentiation and bone resorption [94].

Finally, there is no evidence of a significant effect on adipose tissue physiology and body weight regulation of bone mass-regulating drugs [95].

Effect of Bariatric Surgery on Bone

Metabolic bone disease is a well-recognized complication in patients who had undergone bariatric surgery [96, 97]. Although bone disease is more likely to take place after operations with a malabsorptive component, alterations in mineral metabolism may also be observed in patients undergoing restrictive procedures.

Vertical banded gastroplasty (VBG) and adjustable gastric banding (AGB) are the most frequent restrictive operations among bariatric surgery procedures [98, 99]. A significant reduction in BMD in the femoral neck and trochanter 1 year after VBG has been described [100]. This decrease was apparently not accompanied by changes in parathyroid hormone (PTH), with some authors even evidencing a significant decrease with a concomitant increase in vitamin D [101]. A lack of effect of AGB on BMD or bone mineral content is consistently found in the literature [102, 103]. However, a decrease in femoral and trochanter BMD with no significant changes in total BMD has been reported in one study 24 months after AGB [104]. This effect was suggested to be linked to a decrease in insulin-like growth factor-binding protein 3 [104].

Bone loss after restrictive–malabsorptive Roux-en-Y gastric bypass (RYGBP) is frequently reported [96, 105]. The main proposed mechanisms to explain this problem in mineral metabolism pertain to alterations related to secondary hyperparathyroidism, often caused by a deficiency of dietary calcium or vitamin D, and decreased intestinal absorptive surface area following bariatric procedures [106110]. Although in several studies, an increase in PTH concentrations is the mechanism suggested to explain the bone loss observed after RYGPB [105, 107, 111], this increase is not univocally observed [96, 112]. In order to avoid calcium and vitamin D deficiencies, the intake of supplements containing those nutrients are routinely recommended, especially after malabsorptive procedures [98, 110]. However, calcium and vitamin D deficiencies after RYGBP are rarely noted [96, 97, 112], and preoperative values should be taken into account [113]. Finally, there is no evidence of an increased incidence of fractures among patients who have undergone RYGBP in spite of a potentially altered calcium homeostasis following the operation. Similar side effects have been proposed to take place after biliopancreatic diversion with the possibility of developing alterations being more likely due to the more pronounced malabsorptive component of this technique [114116].

In summary, bone loss after bariatric surgery is frequently observed. However, whether this phenomenon is attributable to the surgical intervention and its gastrointestinal modifications together with the associated hormonal and nutrient alterations or merely to the weight loss itself remains to be better clarified [100].

Concluding Remarks

The adipocyte, considered for a long time a mere energy storage organ, has now been shown to play an active role in a wide variety of homeostatic processes including bone physiology through the effects of a plethora of biologically active adipokines. On the other hand, the skeleton has emerged as an interesting key regulator of energy balance. Recent data seem to indicate that several osteokines may exert an endocrine regulation on body weight and glucose homeostasis. The reviewed evidence suggests that adipose tissue and bone regulate each other in a complex feedback loop. The full understanding of the whole array of molecules involved in this regulatory system and deciphering the signaling pathways implicated will certainly continue to provide a fertile area of research.


The authors gratefully acknowledge the funding of their experimental work by grants from the Spanish Instituto de Salud Carlos III (FIS PI030381, FIS PI061458, and FIS PI06/90288) from the Ministerio de Sanidad y Consumo, as well as by grants 20/2005 and 3/2006 from the Department of Health of the Gobierno de Navarra, Spain and from the PIUNA Foundation. CIBER de Fisiopatología de la Obesidad y Nutrición (CIBEROBN) is an initiative of the Instituto de Salud Carlos III, Spain.

Copyright information

© Springer Science + Business Media, LLC 2008