Obesity is a disease (1) and a global health challenge, now included among the global noncommunicable disease targets identified by the World Health Organization (2). Bariatric surgery constitutes a remarkable tool against obesity and its comorbidities (3). The current epidemic proportions reached by obesity make it a cornerstone treatment to contrast obesity if resistant to standard weight-loss approaches, especially when significant weight loss results are mandatory to control obesity-related impaired health conditions. A number of recent clinical practice guidelines exist on bariatric surgery in adults with obesity (4–6) and the evolution of surgical trends in the past 10 years shows similarities and disparities in the number and types of surgical and endoluminal interventions (7). As for trends of standard bariatric surgery, sleeve gastrectomy (SG) shows a continuous upward trend, while Roux-en-Y gastric bypass (RYGB) and laparoscopic adjustable gastric band (AGB) have trended downward (5). In parallel, an increasing number of gastroenterologists are performing bariatric endoscopic procedures that include placement of intragastric balloons, plications and suturing of the stomach, and insertion of a duodenal-jejunal bypass liner, among other emerging procedures (8).

To ensure long-term postoperative success, patients must be prepared to adopt comprehensive lifestyle changes, yet a number of endogenous and exogenous factors are known to influence bariatric surgery outcomes, as summarized in Table 1 (5, 9–25). It is known that postsurgical weight loss is associated with extended health benefits in terms of arterial hypertension, diabetes mellitus, cardiopulmonary problems, dyslipidemia, susceptibility to neoplasms, osteoarticular disabilities, gastroesophageal reflux disease, psychosocial wellbeing, quality of life, as well as decreased odds of cardiovascular fatalities, strokes and all-cause-mortality (26–29). Nevertheless, there is growing interest on the potential impact of bariatric surgery on bone health, as changes have accumulated to suggest the postsurgical development of accelerated bone loss and skeletal fragility (30). After gastric bypass, an increase in bone resorption markers occurs as early as 10 days postoperatively (31), then marker levels peak by 6 to 12 months (31, 32) and remain elevated thereafter (33). Biochemical markers of bone formation increase but to a lesser extent, suggesting a potential “uncoupling” of resorption from formation (34–37). Several post-bariatric surgery mechanisms have been claimed to occur and act globally on bone health, including mechanical unloading due to bodyweight loss, adipocyte-derived and gastro-enteric hormone changes, and malabsorptive factors (30, 34, 37, 38). Concerns exist on the key role of bariatric surgery on micronutrient intake, diminished calcium and vitamin D absorption leading to secondary hyperparathyroidism, and restricted energy delivery (30, 37). Secondary hyperparathyroidism, which often occurs before bariatric surgery due to the highly prevalent vitamin D deficiency in obesity, progressively increases its prevalence following bariatric surgery, going from 21% at baseline to 35.4% at 1 year and 63.3% at 5 years after surgery (39). The mechanism resides in calcium malabsorption which, in association with vitamin D deficiency, promotes secondary hyperparathyroidism leading to bone resorption. There are even claims that, in the very long term, post-bariatric surgery hypocalcemia may overstimulate parathyroid gland and lead to the anectodical development of a parathyroid adenoma (40). Further, intestinal adaptation mechanisms, local effect of the Wnt/β-catenin signaling pathway, changes in carrier proteins, and the effect of adipokines are other calcium-related mechanisms potentially involved in bone loss after bariatric surgery (40, 41).

While the skeletal response to the effects of bariatric surgery has been collectively confirmed in systematic reviews and meta-analyses, discrepancies still exist in terms of affected bone regions and timing of dynamic bone changes (30, 42, 43). As such, dual-energy X-ray absorptiometry (DXA) studies have shown significant BMD loss at the hip with less consistent results for the lumbar spine (44, 45). Despite weight stabilization and maintenance of metabolic parameters, bone loss and deterioration in bone strength continued years following bariatric surgery, supporting the hypothesis that fracture risk does not appear to be dependent on weight loss, and rather it increases with observation time (46). Moreover, an increased risk of fall injury has been reported after bariatric surgery, which contributes to increase the fracture risk in patients undergoing bariatric surgery procedures (47, 48). Supporting DXA studies, analyses by high-resolution peripheral quantitative computed tomography (HR-pQCT) also showed lower cortical density and thickness, higher cortical porosity, and lower trabecular density and number for up to 5 years after gastric bypass (49). Given these circumstances, it is conceivable that bone health should be accounted for before deciding on the type of bariatric surgery. Little is known about the role of lifestyle intervention, rehabilitation and pharmacological treatments to prevent or treat post-bariatric surgery bone loss and fracture outcome.

The aim of the present review is to update the current data regarding the mechanisms and determinants of bone damage after bariatric surgery and the strategies of prevention and treatment.

The crosstalk between adipose tissue and bone is regulated by a large number of interacting factors ( Figure 1 ). A common stromal cell origin of osteoblasts and adipocytes has been proposed as a possible link between adipose tissue and bone (50). Despite being acknowledged as beneficial to bone mineral density (BMD) due to the mechanical loading effect of weight excess (51), obesity is consistently emerging as a potential detrimental factor for bone health, particularly appendicular bones. Studies highlighted the impact of fat overload on bone strength (52–54) and cortical rearrangement through insulin resistance (55). Osteoporosis is associated with obesity in one out of three women (56), and nearly one out of four postmenopausal women with fractures can present with obesity, particularly in the case of ankle and femur fractures (57). Further, femoral neck BMD is reduced and risk of non-vertebral fragility fractures increased in obese postmenopausal women compared to lean counterparts (58). Collectively, potential mechanisms claimed to explain the interaction between obesity, menopause and bone metabolism (59) include: 1) visceral fat accumulation (60) and sarcopenia (61) compromise the mechanical loading effect (62); 2) lower vitamin D levels and secondary hyperparathyroidism can favor osteoporosis (63); 3) obesity affects endocrine somatotroph, adrenal and thyroid signals active on the bone (64–66); 4) increased risk of type 2 diabetes mellitus (T2DM) can impair femoral neck strength (67, 68).

It is noteworthy that several molecules can serve as signaling triggers between adipose and bone tissue. These include several products and determinants, i.e. adipocytokines released by the adipose tissue (AT) and its macrophage-rich stromal fraction, osteoblast- and osteoclast-derived proteins, as well as several vitamins (69–75). Leptin and other adipokines secreted by the adipose tissue can modulate bone cells through major inhibition of bone remodeling, whereas molecules activating the peroxisome proliferator-activated receptor-γ can drive mesenchymal stem cell differentiation from osteoblastic towards adipocyte lineage (76). Obesity-associated leptin resistance has been linked to decreased bone mass, as seen in case of hypoleptinemia due to extreme leanness (77). Leptin actions involve inhibition of osteoblastic bone formation through its binding to a specific receptor located in the hypothalamus (69, 70). A study conducted by Ducy et al. showed that leptin receptor expression is associated with noradrenalin release and activation of β2 adrenergic receptor in osteoblasts, thus reducing their activity (70). In mice, leptin also inhibits endocrine function of osteoblasts by sympathetic enhancement of the Esp gene expression, thereby decreasing osteocalcin bioactivity and leading to hyperinsulinemia and glucose intolerance (78).

Importantly, insulin partakes in the feedback loop between pancreas and osteoblasts (79), enhances bone remodeling and promotes the decarboxylation of osteocalcin. Osteoblast-derived osteocalcin circulates both as carboxylated (cOC) and undercarboxylated (ucOC) isoforms (69). ucOC possesses extra-skeletal effects, as it stimulates insulin expression and secretion, β-cells proliferation and adiponectin expression in adipocytes, thus resulting in improved glucose tolerance (69, 80). In 1998, Rosato et al. (81) reported that osteocalcin levels were lower in patients with T2DM than healthy subjects, while others observed lower serum osteocalcin levels in patients with poor metabolic control as compared to those with adequate metabolic control and to healthy subjects (82). In rodents and humans, ucOC has also been found to reduce fat mass accumulation through an increase in adiponectin production (83–88). In osteocalcin-deficient homozygous mice (Ocn -/-), Lee et al. (71) documented higher blood glucose and lower insulin levels than in wild-type or heterozygous mice. More, Ocn -/- mice showed an increase in fat mass, adipocyte number and serum triglyceride levels than wild-type mates. In children and adolescents with obesity, serum osteocalcin is reported to be inversely associated with markers of metabolic health and meta-inflammation, i.e. HOMA-IR, HbA1c, triglycerides, C-reactive protein, and fibrinogen, as well as with measures of adiposity, i.e. body mass index (BMI), body fat and waist circumference (87).

Recent studies emphasized a potential role for the osteocyte-secreted product of the SOST gene, sclerostin, in the relationship linking adiposity, T2DM and bone (89, 90). Biologically, sclerostin antagonizes the Wnt/β catenin signaling pathway, which regulates positively osteoprogenitor cell and osteoblast activity (91–93), and plays a role in adipocyte differentiation and metabolic homeostasis (94). Sclerostin predicts bone loss in relation to age, gender and menopause (95), prolonged immobilization (96), and postmenopausal hip fracture risk (97). Even if no difference appears to exist between obese and controls (98), serum sclerostin was found to be negatively associated with insulin sensitivity in obese but not lean subjects, suggesting a potential role for the Wnt/β-catenin pathway in regulating insulin sensitivity in obesity (99). Moreover, sclerostin is increased in states of unloading and possibly mediates changes in bone metabolism associated with weight loss and exercise (100). A negative association relates sclerostin to skeletal muscle mass after adjusting for multiple confounders (101). Finally, in pre- and postmenopausal women with obesity sclerostin positively predicted lumbar spine BMD. Although this relationship seems to conflict with the intrinsic osteopenic effects of sclerostin, this finding suggests a potential role of this hormone in the protective effects elicited by obesity on lumbar spine at menopause (102). Further studies are required to clarify this issue.

Another relevant link between bone, adipose tissue and glucose metabolism is vitamin D. It has been demonstrated that 25-hydroxyvitamin D (25(OH)D) concentrations are positively associated with adiponectin (103) and negatively associated with indices of insulin resistance (103, 104), BMI, and leptin (105). Debatedly, the adipose tissue (AT) is capable of storing vitamin D and, in case of AT excess, it is deemed as responsible of leading to a reduction in its circulating levels (106, 107). Moreover, there is suggestion that vitamin D deficiency promotes greater adiposity by elevating PTH release, which has been shown to increase intracellular calcium accumulation in adipocytes, thereby enhancing lipogenesis (108). After cholecalciferol administration, a change in multimeric adiponectin is also seen (109). Adiponectin is regulated by osteocalcin and has insulin-sensitizing effects (110), with a well-known negative correlation with parameters of the metabolic syndrome (111–114).

In summary, the crosstalk between bone metabolism and adipose tissue involves multiple factors, which could exert different regulatory mechanisms that affect the skeletal health. However, these mechanisms still need to be clarified.

Bariatric surgery is characterized by rapid and dramatic changes in body composition and nutritional factors that are paralleled by changes in bone turnover markers (37, 115). Bone turnover can be inspected and monitored through circulating markers that include serum C-terminal cross-linked telopeptide of type I collagen (CTX), procollagen type I N-terminal propeptide (PINP), bone-specific alkaline phosphatase (BALP), and osteocalcin.

These factors dramatically increase from the first 3-10 days, peak after 6-24 months, and remain elevated until 7 years following bariatric surgery (116–118). Both SG and RYGB promote increases in bone turnover markers, with the latter eliciting the strongest effects and leading to an increase in CTX by 50%-300% (37, 49, 119). Comparative analyses between RYBP and SG showed a significantly higher increase in CTX, P1NP, TRAcP5b with the former, and a greater increase in total OC and uOC with the latter, suggesting a predominating bone resorption over bone formation markers during RYGB (36). A randomized triple-blind trial showed an approximately 100% higher increase in P1NP and CTX-1 levels after RYGB than SG at 1-year post surgery (45). Likewise, studies after biliopancreatic diversion (BPD) showed that CTX increased significantly at 3 days (+ 66%), 3 months (+ 219%), and 12 months (+ 295%), while OC decreased at 3 days (- 19%) then increased at 3 months (+ 69%) and 12 months (+ 164%), suggesting an earlier and greater increase in bone resorption over bone formation markers with BPD (120). Although both CTX and P1NP start to decline after 12-24 months since surgery, they do not tend to return to presurgical levels (116, 121). Oppositely, BALP is a less varying bone turnover marker (10-25%) which predominantly changes during the first year after surgery and more after RYGB than SG (49, 118, 119). Figure 2 summarizes visually the changes occurring over time in bone turnover markers after bariatric surgery. Interestingly, the magnitude of variation in bone turnover markers is reportedly similar between diabetic and non-diabetic cohorts undergoing bariatric surgery, yet patients T2DM patients carry per se a higher risk of osteoporosis (122, 123). Because alterations in CTX and P1NP levels have also been reported in adolescents after RYGB (124) and SG (125), uncertainties remain on the final effect of bariatric surgery on bone mass peak and subsequent adult risk of osteoporosis.

A negative impact of bariatric surgery on bone mineral density (BMD) has been described since the 1980s. Hey et al. described the case of a 38-year old woman who underwent a jejunoileal bypass for severe obesity. Despite vitamin D supplementation, five years after the surgical procedure, the patient develop a fragility fracture of the distal forearm, which healed only after intestinal reanastomosis. The authors hypothesized that the surgery-related malabsorption could lead to alterations in bone metabolism with an increased fracture risk (126). Several studies and meta-analyses on the relationship between bariatric surgery and bone loss have clearly demonstrated that significant declines in BMD occur already within the first year after bariatric surgery (127). Following any type of bariatric surgery, the BMD at the femoral neck has been found significantly lower as compared to controls, with a mean difference (MD) of -0.05 g/cm2 (95% CI -0.07 to -0.02) (43). Oppositely, no difference in BMD was found at column between the two groups.

Growing evidence suggested that the negative effect of bariatric surgery on BMD is strictly dependent on the type of the surgical procedure. A systematic review and meta-analysis by Rodriguez-Carmona et al. demonstrated a significant decrease of -0.03 g/cm2 (95% CI -0.06 to 0.00) in total BMD in patients undergoing mixed restrictive-malabsorptive surgical procedures, but not in those undergoing restrictive surgery (128). In particular, BMD was reduced by -0.07 g/cm2 (95% CI -0.11 to -0.03) at the lumbar spine, -0.12 g/cm2 (95% CI -0.15 to -0.10) at the hip and -0.03 g/cm2 (95% CI -0.04 to -0.02) at the forearm.

BMD changes following bariatric procedures differ depending on skeletal sites and time passed since the operation. At the hip, bone loss following RYGB reaches -3 to -5% after 6 months (129–132) and -6 to -11% after 9-12 months (34, 35, 46, 133–135). Similar results were reported for bone loss at the femoral neck, with a decrease in BMD of -1 to -5% after 6 months and -2 to -9% after 9-12 months, mostly in patients undergoing malabsorptive procedures (34, 35, 38, 49, 130–138). A decline in BMD at the hip and femoral neck has also been observed after restrictive procedures, equivalent to -2 to -8% after 6-24 months (130, 131). At the lumbar spine, some authors did not observe significant variations in BMD at 6-12 month after RYGB (34, 129, 135), whereas others showed a significant reduction in BMD equivalent to -2 to -6% at 6 months and -3 to -7% at 9-12 months (33, 133, 134, 138). Nogues et al. observed a mild difference in BMD loss at the lumbar spine between SG and RYGB (-4.6% vs -6.3%, respectively) (139), while others reported a significant reduction of BMD at this level only after restrictive procedures (130, 140). Further, Maghrabi and co-workers conducted a randomized controlled trial on patients with T2DM to evaluate BMD after 2 years since RYGB and SG, as compared to intensive medical treatment (141). At the hip, BMD loss was significantly higher in patients undergoing SG (-9.2%) and RYGB (-9.5%) than intensive medical therapy group (-0.3%), whereas at lumbar spine a significant decrease in BMD was only observed in the SG group (-2.3%), without changes in RYGB (0.4%) and intensive medical therapy groups (0.8%). A subsequent randomized controlled trial in patients with obesity and T2DM demonstrated that subjects undergoing RYGB had a higher decrease in BMD at the femoral neck (mean between-group difference -2.8%, 95% CI -4.7 to -0.8), total hip (mean between-group difference -3.0%, 95% CI -5.0 to -0.9) and lumbar spine (mean between-group difference -4.2%, 95% CI -6.4 to -2.1) than patients undergoing SG (45).

Overall, the decrease in BMD mainly occurs in the first years after bariatric surgery during the period of rapid weight loss, but it continues even after reaching a stable weight, suggesting that the impact of bariatric surgery on BMD is not completely explained by weight loss (142). Losses appear to be heavier after malabsorptive surgery. Several mechanisms have been hypothesized to explain the negative effects of bariatric surgery on bone metabolism, including mechanical unloading, malabsorption of macro and micronutrients, alterations in gut-derived hormones and microbiota, and changes in body composition (30, 37), which have been extensively described in the following paragraphs.

As seen in many observational studies and clinical trials investigating bone loss associated with bariatric surgery, attention is primarily focused on areal BMD (aBMD) evaluated by using dual-energy X-rays absorptiometry (DXA). However, in the setting of a marked weight loss, DXA results could be influenced by changes in the composition of the bone surrounding tissues (37, 143). In addition, DXA is unable to assess bone microarchitecture, as it cannot discriminate trabecular from cortical bone compartments (37). With the aim of overcoming such limitations, more recent studies have focused on HR-pQCT to evaluate post-bariatric volumetric BMD (vBMD) and bone microarchitecture (38, 46, 119, 135, 144–146). A pioneer study by Stein and co-workers (38) assessed HR-pQCT at the distal radius and tibia in 22 women who underwent RYGB (n=14) and restrictive procedures (n=8). Compared to baseline, after 12 months from surgery trabecular parameters remained stable in both sites, while cortical bone deterioration was observed as exemplified by reductions at the tibia in cortical density (-1.7%), cortical thickness (-2.1%) and cortical area (-2.7%). These changes in cortical compartment were noticed to be more pronounced after RYGB and were independently predicted by the increase in PTH levels, thus suggesting a preferential endocortical bone resorption (38).

In a prospective cohort study on 30 obese adults and 20 non-surgical controls, Yu et al. evaluated the rate of bone loss and microarchitectural alterations occurring in 24 months after RYGB (144). Their results showed that total vBMD progressively decreased after surgery both at the radius and tibia, with a 9% decrease of bone strength as compared to the control group. At the radius, the decrease in bone strength was associated with a greater reduction in trabecular vBMD together with an increase in trabecular heterogeneity, while similar reductions in cortical and trabecular vBMD were seen at the tibia. In addition, the authors observed that the impairment in bone microarchitecture, density and strength observed after the first 12 months was maintained or even worsened after 24 months following surgery, despite a bodyweight plateau reached after 6 months. This temporal connection highlights the potential complex origin of bone loss associated with bariatric surgery.

Subsequent studies (46, 119, 135, 145), underscored an increased porosity in the bone cortical compartment associated, in some instances, with a significant decrease in cortical and trabecular vBMD both at the radius and tibia (46, 49, 119, 135, 145) ( Figure 3 ). The effects of bariatric surgery on bone microarchitecture and vBMD are early and occur within the first months, with a progressive deterioration over the following years (46, 119, 135, 145). It has been also hypothesized that the changes in the cortical compartment of weight bearing and non-weight bearing sites could be modulated by estrogen concentrations in the bone microenvironment (46, 135).

A recent comparative study on vBMD and bone microarchitecture 10 years after RYGB and adjustable gastric banding (AGB) in comparison to age, sex and BMI-matched non-surgical controls (146) documented that total vBMD at the tibia and radius were 17% and 19% lower than in controls, respectively. Alterations were prominent in trabecular microarchitecture and consisted of lower trabecular number and thinner trabeculae. Moreover, in RYGB group, trabecular bone was less axially aligned at both radius and tibia with a decrease in plate bone volume fraction and density as compared to matched controls. No significant differences were found in terms of bone morphology and microarchitecture between the AGB group and controls (146).

In summary, all the evidences confirmed that the decrease in vBMD after bariatric surgery is strongly related to a significant deterioration in bone microarchitecture and strength, thus predisposing to greater bone fragility.

Despite the large body of evidence suggesting that obesity is associated with an increase in BMD, both relating to higher estrogens levels in the adipose tissue and due to the mechanical effect of weight increment (115), this increase in BMD does not reflect a functional improvement of bone microarchitecture, and several studies have demonstrated a higher risk of fracture in obese patients (147). This “obesity paradox” (148) has been attributed to mechanisms involving an increase in bone fragility caused by adiposity and higher risk of falls (115, 149–153) ( Figure 4 ).

Opposed to what would be expected, weight loss is not associated with a decrease in bone fractures risk. In fact, some authors have demonstrated that weight loss, both unintentional and intentional, leads to a decrease in BMD at the proximal humerus and the hip (154–156), with a consequent increase of fracture risk in these sites (157–159). Ensrud and co-workers demonstrated that women achieving intentional and non-intentional weight loss had 1.8 times the risk of subsequent hip fracture (95% CI 1.43–2.24) than those with stable or increasing weight (154). Moreover, a decline of 35% in hip BMD for every 5 kg lost has been observed. The consequences of bariatric surgery are strictly connected with the effect of weight loss, and many studies suggested its potential negative effects on bone metabolism (160). A summary of the results of observational studies and interventional trials is reported in Table 2 (48, 141, 161–169).

One of the first studies designed to evaluate the increased fracture risk in patients undergoing bariatric surgery was published in 2014 by Nakamura and co-workers (162). The authors observed that the relative risk (RR) for any fracture was increased by 2.3-fold as compared to non-surgical controls and that the standardized incidence ratios (SIRs) for a first fracture in osteoporotic sites, including the spine, hip, wrist, or humerus, was nearly doubled (SIR, 1.9; 95% CI, 1.1-2.9). Many subsequent retrospective and interventional studies confirmed the strong association between bariatric surgery and bone fractures risk, also in long-term follow-up (47, 48, 164–166, 168, 170, 171). A recent meta-analysis of 10 observational studies was conducted to compare the fracture risk between 116,205 subjects who underwent bariatric surgery and 134,637 non-surgical patients (121). The results showed that the risk of any fractures was significantly increase by 20% in the group of bariatric surgery than the control counterpart. Despite the low rating on the risk of bias assessment scales, the analysis on three interventional trials shows a trend toward an increase in the fracture risk in patients who underwent bariatric surgery (RR 1.16, CI 95% 1.00-1.33).

Alternatively, Lalmohamed at al. failed to observe a significant increase in the fracture risk in patients undergoing bariatric surgery as compared to control group (adjusted relative risk 0.89, 95% CI 0.60-1.33). However, the authors found a trend towards an increased fracture risk after 3 to 5 years following surgery and in patients with a greater weight loss (161). A lack of association between bariatric surgery and an increased fracture risk was also observed in three subsequent retrospective cohort studies (163, 169, 172). The results of a randomized controlled trial aiming at investigating the 2-year outcomes of bariatric surgery vs intensive medical therapy, have reported a lower total and hip BMD in bariatric surgery but the number of bone fractures did not differ between groups (141). It is important to note that these studies predominantly included patients who underwent restrictive surgery with few cases of malabsorptive procedures (161, 163). Some suggested that the degree of damage to bone microarchitecture varies according to the type of surgery (48, 162). Paccou and co-workers, in a population-based cohort study including 81,948 patients (40,992 in the bariatric surgery group, and 40,992 matched controls), observed a 70% increased risk of fragility fractures only for RYGB within 10-year after surgery (170), while no association between the risk of fragility fractures and SG, AGB and vertical banded gastroplasty (VBG) was seen. An average 1.4-fold higher fracture risk was also observed in a Bayesian network metanalysis, with differences emerging across the various surgical procedures (173), as subjects receiving mixed restrictive/malabsorptive procedures tended to suffer from an increased risk of fracture as compared with those undergoing restrictive procedures (RR 1.54, 95% CI 0.96-2.46). A meta-analysis by Chaves et al. confirmed that malabsorptive procedures elicited a high fracture risk as compared to controls (RR 1.53, CI 95% 1.13-2.07), and RYGB group had a higher risk as compared to SG group (RR 1.77, CI 95% 1.48-2.12) (118). It has been suggested that the higher fracture risk related to malabsorptive or combined procedures could be attributable to neurohormonal changes and malabsorption (37).

A number of mechanisms have been hypothetically linked to postsurgical bone loss, which may involve nutrient absorption deficits, mechanical unloading, alterations in bone marrow adipose tissue (BMAT), as well as changes in adipokines and gut-derived hormones ( Table 3 ) (30, 92, 174–195).

In 1985, Parfitt et al. described bone histomorphometric changes in patients undergoing intestinal bypass and collected evidence from literature of 2.5-25% rate of osteomalacia after shunt operations, commenting that osteomalacia after intestinal bypass surgery has similar clinical, biochemical and histologic features as in other causes of net intestinal malabsorption (196). Bariatric surgery impairs the ability of the digestive tract to secrete hydrochloric acid required for digestion and absorption of nutrients which are required for bone formation and healthy bone remodeling, such as trace elements, essential minerals, water-soluble and fat-soluble vitamins. Many of these disorders are inadequately replaced after surgery, especially if extended gastric bypass surgery, duodenal switch or biliopancreatic diversion are involved (197). Among these, calcium is absorbed passively as well as actively in the small intestine, and a study using a dual stable calcium isotope method (198) demonstrated a significant reduction in calcium absorption following RYGB despite adequate vitamin D and calcium intake. With concern to vitamin D, even solely restrictive procedures like SG can lead to postoperative vitamin D deficiency in as much as 39% of patients despite daily multivitamin supplementation (199), while malabsorptive surgeries pose a higher risk for nutrient deficiencies (200). In a retrospective study, a 73% incidence of vitamin D deficiency was observed following BPD (201).

Weight loss following bariatric surgery instigates a condition of mechanical unloading, that causes a net loss of bone mass by reducing osteoblast function and bone formation. The mechanism likely involves the aforementioned sclerostin and its negative regulation of Wnt/β-catenin signaling pathway, which is important for osteoblast differentiation and function (202). Maintenance of bone strength and density is indeed dependent on adequate muscle mass and function, which reflects intake of high-quality protein (203). Bariatric patients ingest less than 60–120 g of protein recommended daily (197, 204). The combination of postsurgical malnutrition, negative skeletal muscle protein balance, and rapid weight loss initiates a net loss of fat-free mass especially during the first postoperative year, which can last up to 36 months or even up to 9 years post-surgery, and is correlated with loss of handgrip strength (205). Post-operative decrease in fat-free mass and fat mass does not differ between RYGBP and LSG (206). While the biomechanical interaction between nutrition, muscle and bone after bariatric surgery is of key relevance, also important is the biochemical communication existing between muscle and bone which involves secreted factors that act bidirectionally with autocrine/paracrine effects locally as well as through the endocrine system (207). Together, these mechanisms and cross-talks potentially represent the theoretical basis to explain the bone-directed benefit of protein supplementation and mechanical loading through exercise after bariatric surgery. In addition to the previous, a role for bone marrow adipose tissue (BMAT) has also suggested in this scenario. Identified over 100 years ago, BMAT represents up to 70% of bone marrow volume in humans and is a metabolically active, insulin-sensitive and molecularly distinct fat depot that may play a role in whole body energy metabolism (208). Several factors are known to impact bone marrow fat such as obesity, T2DM, estrogen deficiency, caloric restriction, aging, chronic kidney disease, radiotherapy and glucocorticoids. In obese individuals, BMAT fraction is negatively associated with aBMD and the change described for BMAT fraction following RYGB is positively associated with changes in BMI and total body fat (209). RYGB decreases both BMAT and vBMD, yet their changes are unrelated (210).

Finally, a role for adipokines and gut hormones through their central and peripheral receptors has been described (211–216), with evidence indeed coming mainly from animal and in vitro studies ( Figure 5 ).

In humans, bariatric surgery modulates a change in gut hormones, the magnitude of which is dependent of the surgical technique and has a predictive value on postsurgical weight loss and metabolic improvements by mechanisms other than restriction and malabsorption (24). In rodents, GLP-1 administration elicited an anabolic effect on bone, while GLP-2 was associated with reduction in bone resorption markers (217, 218). In vitro, it was further observed that ghrelin administration suppressed osteoclastogenesis and stimulated proliferation and differentiation of osteoblasts (219), while a study in bariatric surgery patients reported an association between ghrelin reduction and BMD loss after RYGB and SG (220). Also, PYY has been positively associated with bone resorption (221), whereas a positive correlation linking PYY to CTX and P1NP levels was observed in patients following RYGB and AGB. In a 12-month study on patients subjected to RYGB, SG or greater curvature plication, postsurgical bone mineral content at the lumbar spine was inversely correlated both with fasting ghrelin and GLP-1, while BMD was positively correlated with post-surgical fasting glucagon and insulin at the femoral neck and inversely with GLP-1 at the lumbar spine.

As previously mentioned, bariatric surgery induces a decline in muscle mass, which is responsible for the 10-28% of total body weight loss (237, 238). A prospective cohort study on 184 patients who underwent SG showed that the prevalence of sarcopenia increases from 8% up to 32% within one year after surgical procedure (239). Muscle waist and sarcopenia were found to be independently associated with several adverse outcomes, including functional decline, a higher rate of falls as well as a higher risk of hospitalization and mortality (OR 3.596, 95% CI 2.96-4.37) (240). Therefore, during follow-up of patients after bariatric surgery, the assessment of BMD, muscle mass and strength, physical activity and fitness should be considered (241). With the aim at improving these parameters, individual interventions of physical activity and rehabilitation have been proven to be effective in preserving muscle mass and endurance capacity, in reducing the risk of bone fractures and in improving quality of life (242–245).

In this context, several studies have demonstrated that physical activity during weight loss is able to prevent the reduction of BMD (100, 246, 247). In 2016, Muschitz et al. conducted an interventional study in 220 patients after RYGB and SG procedures with the aim of assessing the differences in serum markers of bone turnover and BMD between an intervention group (supplementation of vitamin D, calcium, protein, and physical exercise) and a non-intervention group after 2 years from surgical procedure (248). The physical activity intervention consisted of an aerobic and strength exercise program including Nordic walking for 45 minutes for at least 3 times a week and strength training for 30 minutes for at least 2 times a week, for two years. The results showed that the supplementation combined with physical exercise exert a positive effect on long-term outcome in bone protection after bariatric surgery, by modulating serum levels of sclerostin, CTX, DKK-1 and PTH, and counteracting the loss of the spine, hip and total body BMD. One year later, Campanha-Versiani and co-workers evaluated the role of physical exercise after RYGB in influencing BMD and bone turnover markers in a group of patients undergoing a regular and supervised exercise program compared to a control group (249). In this study, physical exercise combined weight-bearing and aerobic exercises two times a week for 36 weeks. One year following RYGB, the intervention group showed a lower decrease in total BMD and at the lumbar spine and hip than the control group, without significant differences in terms of serum concentrations of bone remodeling markers.

Recently, Diniz-Sousa et al. conducted a systematic review and meta-analysis to extensively evaluate the effect of physical exercise and training on BMD at clinically relevant skeletal sites during the first year after bariatric surgery (250). Meta-analysis showed that the decrease in BMD after bariatric surgery can be attenuated from 0.7 to 3.7 percentage points with an exercise training intervention. In particular, exercise training induced a positive effect on BMD at femoral neck [standardized mean difference (SMD)=0.63 (95% CI 0.19-1.06)], total hip [SMD=0.37 (95% CI 0.02-0.71)], lumbar spine [SMD=0.41 (95% CI 0.19-0.62)], and 1/3 radius [SMD=0.58 (95% CI 0.19, 0.97)] as compared to standard medical interventions.

Overall, the evidences suggest that physical exercise after bariatric surgery is fundamental, as it prevents bone loss and muscle depletion during the drastic weight reduction period (142, 205, 251). Exercise programs that include high-impact loading, resistance and strength training as well as aerobic exercises seem to be effective in counteracting the negative effects of bariatric surgery on BMD and bone microarchitecture (142, 252, 253).

Being bariatric surgery patients exposed to multiple systemic risks and particularly for cardiovascular diseases and musculoskeletal impairment, a tailored rehabilitation program through a multidisciplinary approach is key to optimize post-bariatric surgery management (254). The multidisciplinary approach integrates different clinical specialties including endocrinology, clinical nutrition, psychiatry, rehabilitation medicine, as well as health professionals such as nursing, physiotherapy and occupational therapy (255). A post-bariatric surgery rehabilitation project should comprise a number of goals: 1) preventing surgical-related complications, 2) enhancing physical function through adapted physical activity, 3) addressing bariatric-related disabilities as well as socio-environmental and psychological barriers, 4) promoting education on nutritional management, and 5) providing primary or secondary prevention for cardiovascular diseases (254). The rehabilitation program should thus combine musculoskeletal reconditioning, functional mobility, balance training, muscle strengthening, aerobic exercises or physical endurance, activity of daily living (ADL) training, nutritional and psychological support, weight management, and monitoring of clinical aspects (254).

Bone changes after bariatric surgery may have effects that go well beyond the acute phase of weight loss. Bariatric surgery, especially malabsorptive procedures, increases bone turnover, decreases bone mass, and enhances the fracture risk. These consequences advocate the need to adequately study, monitor and support skeletal health and micronutrient supply in bariatric surgery patients to avoid short- and long-term damage in bone density. There is collective evidence that bone density estimation by DXA can be improved by HR-pQCT to better classify patients at risk of osteoporosis. To compensate for nutritional and mechanical deficits after surgery, replenishment with calcium citrate and high-dose vitamin D plus scheduled exercise programs are mandatory. Noticeably, absorption problems and potential ulceration of anastomosis should be considered before prescribing oral bisphosphonates. In patients who fail to achieve BMD improvements post-surgically, intravenous bisphosphonates and/or denosumab should be considered, with calcium and vitamin D being critical to avoid hypocalcemia. Yet, RCTs are needed to determine whether anti-osteoporotic therapy is effective and safe for preventing high-turnover bone loss and treating osteoporosis in this population. Finally, it is of utmost importance to consider bone health before deciding the type of bariatric surgery a patient with obesity should be subjected to.

Conceptualization and methodology, CM, MC, and PM. Original draft preparation, CM, MC, AF, TD, BC, DS, and FP. Review and editing, CM, GA, and PM. Supervision, GA, AN, and PM. All authors contributed to the article and approved the submitted version.

Research founded by the Italian Ministry of Health (18C101_2011).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.