Series
Bariatric surgery 3 Bone loss after bariatric surgery: causes, consequences, and management Emily M Stein, Shonni J Silverberg
Bariatric surgery is an effective and increasingly common treatment for severe obesity and its many comorbidities. The side-effects of bariatric surgery can include detrimental effects on bone and mineral metabolism. Bone disease in patients who have had bariatric surgery is affected by preoperative abnormalities in bone and mineral metabolism related to severe obesity. Changes that arise after bariatric surgery are specific to procedure type: the most pronounced abnormalities in calciotropic hormones and bone loss are noted after procedures that result in the most malabsorption. The most consistent site for bone loss after all bariatric procedures is at the hip. There are limitations of dual-energy x-ray absorptiometry technology in this population, including artefact introduced by adipose tissue itself. Bone loss after bariatric surgery is probably multifactorial. Proposed mechanisms include skeletal unloading, abnormalities in calciotropic hormones, and changes in gut hormones. Few data for fracture risk in the bariatric population are available, and this is a crucial area for additional research. Treatment should be geared toward correction of nutritional deficiencies and study of bone mineral density in high-risk patients. We explore the skeletal response to bariatric surgery, potential mechanisms for changes, and strategies for management.
Introduction Bariatric surgery has become an increasingly common treatment for severe obesity1,2 because it results in substantial, sustained weight loss,3 reverses many complications of obesity,4,5 and decreases mortality.6,7 However, several detrimental effects are associated with these procedures such as deleterious effects on bone and mineral metabolism, including vitamin D deficiency, hyperparathyroidism, and bone loss.8–10 Bone loss after bariatric surgery is probably multifactorial. Proposed mechanisms include skeletal unloading, abnormalities in calciotropic hormones, and changes in gut hormones. Increased bone turnover might be associated with physiological adaptation of the skeleton to unloading, or with pathophysiological changes such as increased parathyroid hormone. Whether the skeletal changes that arise after bariatric surgery are pathological and are associated with skeletal fragility is unknown. In the third paper in this Series, we explore the skeletal response to bariatric surgery, potential mechanisms for these changes, and strategies for management. Because other recent reviews have specifically explored the association between gut hormones and adipokines and bone, we will not address that topic in detail. The rate of weight loss and resolution of comorbidities after bariatric surgery vary by procedure.11 The skeletal effects are also procedure specific. Surgery involves the reduction in stomach size (restriction), delayed mixing of food with bile salts and pancreatic juices (malabsorption), or a combination of both (figure 1). Laproscopic adjustable gastric banding (LAGB) is a purely restrictive procedure, in which a silicone band placed around the proximal stomach creates a pouch that holds only a small amount of food. In vertical sleeve gastrectomy (VSG), more than 80% of the stomach is transected. Nutrients rapidly pass www.thelancet.com/diabetes-endocrinology Vol 2 February 2014
Lancet Diabetes Endocrinol 2014; 2: 165–74 This is the third in a Series of four papers about bariatric surgery Columbia University College of Physicians and Surgeons, Columbia University, New York 10032, USA (E M Stein MD, Prof S J Silverberg MD) [email protected]
through the gastric conduit, resulting in altered gut hormones and metabolism.1 Roux-en-Y gastric bypass (RYGB) has both restrictive and malabsorptive features. Restriction arises through the creation of a small gastric pouch from the proximal stomach, which is then anastamosed to the proximal jejunum to form an alimentary tract. Food content mixes with bile and pancreatic secretions in the distal jejunum. In biliopancreatic diversion with duodenal switch, food bypasses most of the small intestine; a gastric sleeve is anastamosed to the distal ileum where food mixes with digestive enzymes.8 In RYGB and biliopancreatic diversion with duodenal switch, the intestinal surface area available for caloric absorption is reduced, leading to malabsorption of minerals and fat-soluble vitamins. Hormonal changes that arise as a result of VSG, RYGB, and biliopancreatic diversion with duodenal switch are important mediators of weight loss and can also affect bone loss.
Causes of abnormal bone metabolism in obese individuals Before discussing the effects of bariatric surgery on bone, it is important to review the skeletal consequences of obesity. Vitamin D deficiency is widespread in obese people,12,13 with potential causes including inadequate dietary intake of foods and supplements containing vitamin D despite high overall caloric intakes,14 insufficient sunlight exposure,15 and decreased bioavailability of vitamin D secondary to sequestration of the fat-soluble vitamin in excess adipose tissue.16 Individuals who are severely obese, African Americans, and those who have low amounts of sun exposure seem to be at greatest risk of vitamin D deficiency.12 Hyperparathyroidism is common, and although this disorder might be secondary to 165
Series
A
B
Laproscopic adjustable gastric banding
Verticle sleeve gastrectomy
Tube to carry fluid
Pylorus Gastric band Gastric sleeve
Limitations of bone studies in the bariatric population
Subcutaneous injection port
C
D
Roux-en-Y gastric bypass
Biliopancreatic diversion with duodenal switch
Alimentary limb Transverse mesocolon Biliopancreatic limb
Alimentary limb
Biliopancreatic limb Common channel Common channel
Figure 1: Common bariatric surgery procedures Reproduced from Jones and colleagues,11 by permission of Cine-Med.
vitamin D deficiency, an independent association between parathyroid hormone and obesity has been reported.17,18 Emerging data also support a direct relation between bone metabolism and fat mass and distribution.19,20 Findings from one study21 showed an inverse association between total fat mass and bone mineral density. Osteoblasts and adipocytes differentiate from the same mesenchymal precursor, and increased marrow fat is one proposed mechanism for low bone mineral density in obese individuals.22 Visceral fat in particular might have negative effects on bone formation, structure, and strength, whereas subcutaneous fat might be protective.23,24 In a recent study,17 obese women had higher levels of parathyroid hormone, bone-specific alkaline phosphatase, leptin, fibroblast growth factor-23 (FGF-23), and lower 1,25-dihydroxyvitamin D than controls. Leptin levels predicted both parathyroid hormone and FGF-23.17 The association between BMI and fracture risk is complex and differs across skeletal sites.25 Recent data 166
question the long-held notion that obesity protects against osteoporosis. Compared with bodyweight, bone microarchitecture and strength are lower in obese women than in controls.26 Although few data are available, obese patients might be at increased fracture risk, especially at peripheral sites,27–30 which might partly relate to an increased risk of falls.27 Increased intramuscular adipose tissue might result in impaired mobility and muscle strength.31 Additionally, these constraints on mobility in obese individuals might result in altered patterns of falling, with increased propensity toward backward and sideways falls. Men and women might be at risk for fractures at different anatomical sites.32
There are several limitations common to many studies of bone disease in individuals who have undergone bariatric surgery. Most studies are small, and have substantial dropout rates, especially the few that study patients beyond 1 year postoperatively. Substantial heterogeneity exists within studies relating to the age, sex, and race of patients, surgical techniques, sites assessed by dualenergy x-ray absorptiometry and associated dual-energy x-ray absorptiometry techniques. In most studies, supplementation with calcium and vitamin D is done as part of clinical care and is ancillary to the study protocol. For this reason, compliance is not assessed. In many studies, information about supplementation is not reported at all. A limitation of dual-energy x-ray absorptiometry technology is that obesity itself and changes in fat mass introduce artifacts that might compromise accuracy and precision.33–36 Measurements by quantitative CT (QCT)37 and high resolution peripheral QCT (HR-pQCT)38 might be less affected by changes in body fat. Although newer machines can accommodate patients up to about 200 kg or 450 pounds, the lower weight limit of 136 kg or 300 pounds of many dual-energy x-ray absorptiometry machines means that many studies were only able to obtain axial dual-energy x-ray absorptiometry measurements (of the spine and hip) in a subset of patients, making the results less powerful and less reliable.
Skeletal consequences of bariatric surgery LAGB After LAGB, patients typically lose between 20 and 30% of initial bodyweight and 41–54% of excess bodyweight.39,40 However, weight regain is common.41,42 Although vitamin D deficiency has been documented before LAGB, postoperatively, vitamin D remains stable or increases and parathyroid hormone remains stable.43,44 Increased bone resorption, measured by C-terminal telopeptide, is evident by 6 months, and persists for at least 24 months after surgery.43 Although it is beyond the scope of this Review to discuss in detail, circulating www.thelancet.com/diabetes-endocrinology Vol 2 February 2014
Series
oestrogen and leptin concentrations decrease, which are both related to decreased fat tissue.45 Bone density studies after LAGB are complicated by artefacts in spine dual-energy x-ray absorptiometry measurements that might be introduced if the band overlies the spine. Spuriously high follow-up values arise if affected vertebrae are not properly identified and excluded. Investigators have reported that lumbar spine bone mineral density does not change, or increases slightly.43,44,46,47 Estimates of hip bone loss have been variable, from no change at 1 year in a study of premenopausal women only, to 6% at 24 months.43,44 In vertical banded gastroplasty, another restrictive procedure, decreases in bone density at the femoral neck, but not at the lumbar spine have been reported.46,48
VSG VSG is becoming increasingly common; the degree of weight loss with this procedure falls between that of the purely restrictive and malabsorptive procedures (1–2 year weight loss: 20–30%; excess BMI loss: 45–80%).39,49–53 Data for bone metabolism after VSG are scarce. In one study,49 in which 95% of patients had vitamin D deficiency and 43% had raised parathyroid hormone at baseline, 25-hydroxyvitamin D increased and parathyroid hormone decreased postoperatively, although no information is available regarding supplement use or adherence. A prospective study51 with higher baseline concentrations found 25-hydroxy vitamin D increased but parathyroid hormone did not change after VSG. Data for skeletal outcomes are scarce. Increased bone turnover markers have been reported at 1 year.51 Bone loss has been reported as early as 6 months after VSG (total hip 5·2% and femoral neck 7·0%, with a small decrease at the lumbar spine).54 In a small prospective study, eight women who underwent VSG had substantial bone loss at the spine (4·6%) and hip (total hip 8·3% and femoral neck 7·1%) at 1 year.51 By contrast, in a retrospective study49 in which most patients were vitamin D-deficient preoperatively, lumbar spine bone mineral density increased over 2 years postoperatively. The contradictory findings in the latter study might be explained by its retrospective study design, selection bias in the patients referred for dual-energy X-ray absorptiometry, and treatment of vitamin D deficiency. Larger, prospective studies are needed to elucidate the changes in bone density that happen after VSG.
RYGB Most available data for changes in bone after bariatric surgery come from studies of RYGB. After RYGB, patients typically lose 35% of initial weight, or 62–75% of excess bodyweight,8–10,50 which is a substantially greater loss than with LAGB.55,56 Calcium malabsorption has been documented as early as 3 months after RYGB,9,10 with reduced true fractional absorption of calcium57 and secondary hyperparathyroidism.8,9,58,59 Findings from early www.thelancet.com/diabetes-endocrinology Vol 2 February 2014
studies showed high rates of vitamin D deficiency after RYGB.60,61 Although aggressive vitamin D repletion has led to less postoperative vitamin D deficiency, improvements have not been commensurate with supplementation, suggesting substantial impairment in vitamin D absorption after RYGB. We found stable 25-hydroxyvitamin D concentrations despite more than 200% increases in vitamin D intake (mean of 658 IU per day at baseline, 1698 IU per day at 12 months),9 and no increase in 25-hydroxyvitamin D concentrations on doses of roughly 5000 IU daily.8 Further reductions in 25-hydroxyvitamin D and increases in parathyroid hormone have been reported at 3 years compared with at 1 year after RYGB,62 but no information on compliance with supplements was reported. Bone turnover increases as early as 3 months after RYGB.9,10 Resorption markers increase steadily by up to 200% in the first 12–18 months.8,63,64 Increases in bone formation markers are less exuberant,8,9 and have not been uniformly reported.10 No long-term longitudinal data are available that elucidate the duration of increased bone turnover. In one cross-sectional study, osteocalcin and bone-specific alkaline phosphatase were still raised compared with obese controls 10 years after RYGB.65 Although findings from cross-sectional comparisons did not show consistent differences in bone mineral density in patients who had had RYGB compared with non-surgical controls,65,66 findings of prospective studies have shown clear decreases in bone mineral density in the first year after RYGB. Decreases in hip bone mineral density range between 8% and 11%.8–10,20,62,67–69 Few investigators have assessed changes beyond 1 year; in one study, hip bone mineral density decreased by 3% between years 1 and 3.62 Changes at the spine are more variable; some studies report small decreases,10,27,63,67–69 whereas others have not.8,9,65,66 One study62 reported a further small decrease in spine bone mineral density between 1 and 3 years. Some66 but not all8 researchers have reported that menopausal status affects change in bone mineral density after RYGB, with postmenopausal women losing the greatest amounts of bone. Bone mineral density at the radius has only been measured in a few longitudinal studies, and most have found no change,8–10,37 although one study did report a decrease.69 Other studies were restricted by inclusion of patients who underwent other malabsorptive procedures.70,71 A recent study in which patients were studied after RYGB with dual-energy x-ray absorptiometry and central QCT reported decreases at the lumbar spine by both imaging modalities. However, decreased bone density at the hip as measured by dual-energy x-ray absorptiometry was not recorded with QCT, raising the possibility that artifactual changes might substantially affect hip measurements.63 In other work, when apparent bone mineral density was calculated to compensate for potential artifact related to changes in bone size, bone mineral apparent density did not decrease at the spine.9 167
Series
Additional work is needed to clarify what the most accurate method is in this population. Few studies have compared bone loss after RYGB with other procedures, and those that have are limited by small sample sizes. Typically, patients who have had RYGB lose more weight and more bone at the hip and femoral neck at 1 year than do those who undergo LAGB8,47 or vertical banded gastroplasty.72 The one study73 that found similar bone loss at the spine and hip after VSG and RYGB also had similar excess weight loss and calciotropic hormone levels, eliminating the differences in those factors that might be expected to affect bone density.
Biliopancreatic diversion with duodenal switch Biliopancreatic diversion with duodenal switch is typically reserved for extremely obese patients (BMI >50 kg/m²); mean excess weight loss is 70–80%.74–76 After this procedure, more than 50% of patients have vitamin D deficiency and estimates of secondary hyperparathyroidism range from 60 to 100% even with aggressive supplementation.75,77–79 Markers of bone turnover increase substantially.74,80 Findings of a randomised trial81 showed greater postoperative 25-hydroxyvitamin D deficiency in biliopancreatic diversion with duodenal switch versus RYGB, and although parathyroid hormone concentrations did not differ, supplement use was far greater in patients with biliopancreatic diversion with duodenal switch. Decreased bone formation and increased resorption as measured by histomorphometry was reported after an early form of biliopancreatic diversion with duodenal switch.82 Most patients (30 of 41, 73%) had defective mineralisation. A small decrease in lumbar spine bone mineral density and stable hip density was seen 4 and 10 years after biliopancreatic diversion with duodenal switch,80 but results were very heterogeneous, with bone mineral density increasing and decreasing in similar numbers of patients. In another study,74 in which investigators measured only lumbar spine bone mineral density, significant decreases were recorded at 1 year, unattenuated by high-dose calcium supplementation.
Changes in bone quality and microarchitecture after bariatric surgery In addition to bone density and bone turnover, microarchitecture is an important factor contributing to bone strength and fracture risk.83 Assessment of changes in skeletal microarchitecture can better elucidate the nature of bone loss after bariatric surgery, but data are limited. Bone biopsy samples at baseline and 4 years postoperatively in patients who underwent biliopancreatic diversion with duodenal switch showed decreased cortical thickness and increased trabecular bone volume.80 Mineralisation decreased over the 4 year follow-up, with an increase in osteoid volume and bone formation rate. These data raise the possibility that a decrease in mineralisation could explain some of the 168
postoperative reduction in bone mineral density. Although this mechanism might play a part in individuals who have had biliopancreatic diversion with duodenal switch, defective mineralisation is unlikely to have a big role in the decreased bone mineral density recorded in recent studies of other procedures (LAGB, VSG, or RYGB), given that in response to robust supplementation, most patients had 25-hydroxyvitamin D concentrations close to 50 nmol/L. In our recent prospective study8 of 22 women who underwent RYGB, VSG, and LAGB (14 patients had RYGB, two had VSG, and two had LAGB), with HR-pQCT, we recorded a decrease of cortical area, density, thickness, and total density at the tibia 1 year after surgery. Decreases in cortical bone were predicted by the increase in parathyroid hormone. The subgroup of patients who underwent RYGB lost markedly more weight, had more cortical bone loss than did those with LAGB or VSG and had decreases in cortical load share estimated by finite element analysis.84 HR-pQCT might also be affected by changes in subcutaneous fat.
Mechanisms of bone loss after bariatric surgery Unloading Mechanical loading of bone is an important determinant of bone size, mass, and biomechanical properties. Changes in loading can induce compensatory increases in localised bone remodelling,85 probably mediated through osteocytes and the sclerostin pathway. Bone loss has been reported in the setting of skeletal unloading in other populations, including patients with spinal cord injury,86 and those restricted to bed rest.87 Hip bone loss has been documented in individuals who lose even small amounts of weight from caloric restriction.88–92 The hip typically carries a load roughly two to three times bodyweight;93 therefore, drastic weight loss after bariatric surgery might especially affect this site. A strong association between extent of weight loss after bariatric surgery and amount of bone loss has been documented by most,9,10,43,48,62,67 but not all, studies.63 We found that after RYGB, declines at the total hip (r=0·65, p=0·02) and femoral neck (r=0·90, p
Bariatric surgery 3 Bone loss after bariatric surgery: causes, consequences, and management Emily M Stein, Shonni J Silverberg
Bariatric surgery is an effective and increasingly common treatment for severe obesity and its many comorbidities. The side-effects of bariatric surgery can include detrimental effects on bone and mineral metabolism. Bone disease in patients who have had bariatric surgery is affected by preoperative abnormalities in bone and mineral metabolism related to severe obesity. Changes that arise after bariatric surgery are specific to procedure type: the most pronounced abnormalities in calciotropic hormones and bone loss are noted after procedures that result in the most malabsorption. The most consistent site for bone loss after all bariatric procedures is at the hip. There are limitations of dual-energy x-ray absorptiometry technology in this population, including artefact introduced by adipose tissue itself. Bone loss after bariatric surgery is probably multifactorial. Proposed mechanisms include skeletal unloading, abnormalities in calciotropic hormones, and changes in gut hormones. Few data for fracture risk in the bariatric population are available, and this is a crucial area for additional research. Treatment should be geared toward correction of nutritional deficiencies and study of bone mineral density in high-risk patients. We explore the skeletal response to bariatric surgery, potential mechanisms for changes, and strategies for management.
Introduction Bariatric surgery has become an increasingly common treatment for severe obesity1,2 because it results in substantial, sustained weight loss,3 reverses many complications of obesity,4,5 and decreases mortality.6,7 However, several detrimental effects are associated with these procedures such as deleterious effects on bone and mineral metabolism, including vitamin D deficiency, hyperparathyroidism, and bone loss.8–10 Bone loss after bariatric surgery is probably multifactorial. Proposed mechanisms include skeletal unloading, abnormalities in calciotropic hormones, and changes in gut hormones. Increased bone turnover might be associated with physiological adaptation of the skeleton to unloading, or with pathophysiological changes such as increased parathyroid hormone. Whether the skeletal changes that arise after bariatric surgery are pathological and are associated with skeletal fragility is unknown. In the third paper in this Series, we explore the skeletal response to bariatric surgery, potential mechanisms for these changes, and strategies for management. Because other recent reviews have specifically explored the association between gut hormones and adipokines and bone, we will not address that topic in detail. The rate of weight loss and resolution of comorbidities after bariatric surgery vary by procedure.11 The skeletal effects are also procedure specific. Surgery involves the reduction in stomach size (restriction), delayed mixing of food with bile salts and pancreatic juices (malabsorption), or a combination of both (figure 1). Laproscopic adjustable gastric banding (LAGB) is a purely restrictive procedure, in which a silicone band placed around the proximal stomach creates a pouch that holds only a small amount of food. In vertical sleeve gastrectomy (VSG), more than 80% of the stomach is transected. Nutrients rapidly pass www.thelancet.com/diabetes-endocrinology Vol 2 February 2014
Lancet Diabetes Endocrinol 2014; 2: 165–74 This is the third in a Series of four papers about bariatric surgery Columbia University College of Physicians and Surgeons, Columbia University, New York 10032, USA (E M Stein MD, Prof S J Silverberg MD) [email protected]
through the gastric conduit, resulting in altered gut hormones and metabolism.1 Roux-en-Y gastric bypass (RYGB) has both restrictive and malabsorptive features. Restriction arises through the creation of a small gastric pouch from the proximal stomach, which is then anastamosed to the proximal jejunum to form an alimentary tract. Food content mixes with bile and pancreatic secretions in the distal jejunum. In biliopancreatic diversion with duodenal switch, food bypasses most of the small intestine; a gastric sleeve is anastamosed to the distal ileum where food mixes with digestive enzymes.8 In RYGB and biliopancreatic diversion with duodenal switch, the intestinal surface area available for caloric absorption is reduced, leading to malabsorption of minerals and fat-soluble vitamins. Hormonal changes that arise as a result of VSG, RYGB, and biliopancreatic diversion with duodenal switch are important mediators of weight loss and can also affect bone loss.
Causes of abnormal bone metabolism in obese individuals Before discussing the effects of bariatric surgery on bone, it is important to review the skeletal consequences of obesity. Vitamin D deficiency is widespread in obese people,12,13 with potential causes including inadequate dietary intake of foods and supplements containing vitamin D despite high overall caloric intakes,14 insufficient sunlight exposure,15 and decreased bioavailability of vitamin D secondary to sequestration of the fat-soluble vitamin in excess adipose tissue.16 Individuals who are severely obese, African Americans, and those who have low amounts of sun exposure seem to be at greatest risk of vitamin D deficiency.12 Hyperparathyroidism is common, and although this disorder might be secondary to 165
Series
A
B
Laproscopic adjustable gastric banding
Verticle sleeve gastrectomy
Tube to carry fluid
Pylorus Gastric band Gastric sleeve
Limitations of bone studies in the bariatric population
Subcutaneous injection port
C
D
Roux-en-Y gastric bypass
Biliopancreatic diversion with duodenal switch
Alimentary limb Transverse mesocolon Biliopancreatic limb
Alimentary limb
Biliopancreatic limb Common channel Common channel
Figure 1: Common bariatric surgery procedures Reproduced from Jones and colleagues,11 by permission of Cine-Med.
vitamin D deficiency, an independent association between parathyroid hormone and obesity has been reported.17,18 Emerging data also support a direct relation between bone metabolism and fat mass and distribution.19,20 Findings from one study21 showed an inverse association between total fat mass and bone mineral density. Osteoblasts and adipocytes differentiate from the same mesenchymal precursor, and increased marrow fat is one proposed mechanism for low bone mineral density in obese individuals.22 Visceral fat in particular might have negative effects on bone formation, structure, and strength, whereas subcutaneous fat might be protective.23,24 In a recent study,17 obese women had higher levels of parathyroid hormone, bone-specific alkaline phosphatase, leptin, fibroblast growth factor-23 (FGF-23), and lower 1,25-dihydroxyvitamin D than controls. Leptin levels predicted both parathyroid hormone and FGF-23.17 The association between BMI and fracture risk is complex and differs across skeletal sites.25 Recent data 166
question the long-held notion that obesity protects against osteoporosis. Compared with bodyweight, bone microarchitecture and strength are lower in obese women than in controls.26 Although few data are available, obese patients might be at increased fracture risk, especially at peripheral sites,27–30 which might partly relate to an increased risk of falls.27 Increased intramuscular adipose tissue might result in impaired mobility and muscle strength.31 Additionally, these constraints on mobility in obese individuals might result in altered patterns of falling, with increased propensity toward backward and sideways falls. Men and women might be at risk for fractures at different anatomical sites.32
There are several limitations common to many studies of bone disease in individuals who have undergone bariatric surgery. Most studies are small, and have substantial dropout rates, especially the few that study patients beyond 1 year postoperatively. Substantial heterogeneity exists within studies relating to the age, sex, and race of patients, surgical techniques, sites assessed by dualenergy x-ray absorptiometry and associated dual-energy x-ray absorptiometry techniques. In most studies, supplementation with calcium and vitamin D is done as part of clinical care and is ancillary to the study protocol. For this reason, compliance is not assessed. In many studies, information about supplementation is not reported at all. A limitation of dual-energy x-ray absorptiometry technology is that obesity itself and changes in fat mass introduce artifacts that might compromise accuracy and precision.33–36 Measurements by quantitative CT (QCT)37 and high resolution peripheral QCT (HR-pQCT)38 might be less affected by changes in body fat. Although newer machines can accommodate patients up to about 200 kg or 450 pounds, the lower weight limit of 136 kg or 300 pounds of many dual-energy x-ray absorptiometry machines means that many studies were only able to obtain axial dual-energy x-ray absorptiometry measurements (of the spine and hip) in a subset of patients, making the results less powerful and less reliable.
Skeletal consequences of bariatric surgery LAGB After LAGB, patients typically lose between 20 and 30% of initial bodyweight and 41–54% of excess bodyweight.39,40 However, weight regain is common.41,42 Although vitamin D deficiency has been documented before LAGB, postoperatively, vitamin D remains stable or increases and parathyroid hormone remains stable.43,44 Increased bone resorption, measured by C-terminal telopeptide, is evident by 6 months, and persists for at least 24 months after surgery.43 Although it is beyond the scope of this Review to discuss in detail, circulating www.thelancet.com/diabetes-endocrinology Vol 2 February 2014
Series
oestrogen and leptin concentrations decrease, which are both related to decreased fat tissue.45 Bone density studies after LAGB are complicated by artefacts in spine dual-energy x-ray absorptiometry measurements that might be introduced if the band overlies the spine. Spuriously high follow-up values arise if affected vertebrae are not properly identified and excluded. Investigators have reported that lumbar spine bone mineral density does not change, or increases slightly.43,44,46,47 Estimates of hip bone loss have been variable, from no change at 1 year in a study of premenopausal women only, to 6% at 24 months.43,44 In vertical banded gastroplasty, another restrictive procedure, decreases in bone density at the femoral neck, but not at the lumbar spine have been reported.46,48
VSG VSG is becoming increasingly common; the degree of weight loss with this procedure falls between that of the purely restrictive and malabsorptive procedures (1–2 year weight loss: 20–30%; excess BMI loss: 45–80%).39,49–53 Data for bone metabolism after VSG are scarce. In one study,49 in which 95% of patients had vitamin D deficiency and 43% had raised parathyroid hormone at baseline, 25-hydroxyvitamin D increased and parathyroid hormone decreased postoperatively, although no information is available regarding supplement use or adherence. A prospective study51 with higher baseline concentrations found 25-hydroxy vitamin D increased but parathyroid hormone did not change after VSG. Data for skeletal outcomes are scarce. Increased bone turnover markers have been reported at 1 year.51 Bone loss has been reported as early as 6 months after VSG (total hip 5·2% and femoral neck 7·0%, with a small decrease at the lumbar spine).54 In a small prospective study, eight women who underwent VSG had substantial bone loss at the spine (4·6%) and hip (total hip 8·3% and femoral neck 7·1%) at 1 year.51 By contrast, in a retrospective study49 in which most patients were vitamin D-deficient preoperatively, lumbar spine bone mineral density increased over 2 years postoperatively. The contradictory findings in the latter study might be explained by its retrospective study design, selection bias in the patients referred for dual-energy X-ray absorptiometry, and treatment of vitamin D deficiency. Larger, prospective studies are needed to elucidate the changes in bone density that happen after VSG.
RYGB Most available data for changes in bone after bariatric surgery come from studies of RYGB. After RYGB, patients typically lose 35% of initial weight, or 62–75% of excess bodyweight,8–10,50 which is a substantially greater loss than with LAGB.55,56 Calcium malabsorption has been documented as early as 3 months after RYGB,9,10 with reduced true fractional absorption of calcium57 and secondary hyperparathyroidism.8,9,58,59 Findings from early www.thelancet.com/diabetes-endocrinology Vol 2 February 2014
studies showed high rates of vitamin D deficiency after RYGB.60,61 Although aggressive vitamin D repletion has led to less postoperative vitamin D deficiency, improvements have not been commensurate with supplementation, suggesting substantial impairment in vitamin D absorption after RYGB. We found stable 25-hydroxyvitamin D concentrations despite more than 200% increases in vitamin D intake (mean of 658 IU per day at baseline, 1698 IU per day at 12 months),9 and no increase in 25-hydroxyvitamin D concentrations on doses of roughly 5000 IU daily.8 Further reductions in 25-hydroxyvitamin D and increases in parathyroid hormone have been reported at 3 years compared with at 1 year after RYGB,62 but no information on compliance with supplements was reported. Bone turnover increases as early as 3 months after RYGB.9,10 Resorption markers increase steadily by up to 200% in the first 12–18 months.8,63,64 Increases in bone formation markers are less exuberant,8,9 and have not been uniformly reported.10 No long-term longitudinal data are available that elucidate the duration of increased bone turnover. In one cross-sectional study, osteocalcin and bone-specific alkaline phosphatase were still raised compared with obese controls 10 years after RYGB.65 Although findings from cross-sectional comparisons did not show consistent differences in bone mineral density in patients who had had RYGB compared with non-surgical controls,65,66 findings of prospective studies have shown clear decreases in bone mineral density in the first year after RYGB. Decreases in hip bone mineral density range between 8% and 11%.8–10,20,62,67–69 Few investigators have assessed changes beyond 1 year; in one study, hip bone mineral density decreased by 3% between years 1 and 3.62 Changes at the spine are more variable; some studies report small decreases,10,27,63,67–69 whereas others have not.8,9,65,66 One study62 reported a further small decrease in spine bone mineral density between 1 and 3 years. Some66 but not all8 researchers have reported that menopausal status affects change in bone mineral density after RYGB, with postmenopausal women losing the greatest amounts of bone. Bone mineral density at the radius has only been measured in a few longitudinal studies, and most have found no change,8–10,37 although one study did report a decrease.69 Other studies were restricted by inclusion of patients who underwent other malabsorptive procedures.70,71 A recent study in which patients were studied after RYGB with dual-energy x-ray absorptiometry and central QCT reported decreases at the lumbar spine by both imaging modalities. However, decreased bone density at the hip as measured by dual-energy x-ray absorptiometry was not recorded with QCT, raising the possibility that artifactual changes might substantially affect hip measurements.63 In other work, when apparent bone mineral density was calculated to compensate for potential artifact related to changes in bone size, bone mineral apparent density did not decrease at the spine.9 167
Series
Additional work is needed to clarify what the most accurate method is in this population. Few studies have compared bone loss after RYGB with other procedures, and those that have are limited by small sample sizes. Typically, patients who have had RYGB lose more weight and more bone at the hip and femoral neck at 1 year than do those who undergo LAGB8,47 or vertical banded gastroplasty.72 The one study73 that found similar bone loss at the spine and hip after VSG and RYGB also had similar excess weight loss and calciotropic hormone levels, eliminating the differences in those factors that might be expected to affect bone density.
Biliopancreatic diversion with duodenal switch Biliopancreatic diversion with duodenal switch is typically reserved for extremely obese patients (BMI >50 kg/m²); mean excess weight loss is 70–80%.74–76 After this procedure, more than 50% of patients have vitamin D deficiency and estimates of secondary hyperparathyroidism range from 60 to 100% even with aggressive supplementation.75,77–79 Markers of bone turnover increase substantially.74,80 Findings of a randomised trial81 showed greater postoperative 25-hydroxyvitamin D deficiency in biliopancreatic diversion with duodenal switch versus RYGB, and although parathyroid hormone concentrations did not differ, supplement use was far greater in patients with biliopancreatic diversion with duodenal switch. Decreased bone formation and increased resorption as measured by histomorphometry was reported after an early form of biliopancreatic diversion with duodenal switch.82 Most patients (30 of 41, 73%) had defective mineralisation. A small decrease in lumbar spine bone mineral density and stable hip density was seen 4 and 10 years after biliopancreatic diversion with duodenal switch,80 but results were very heterogeneous, with bone mineral density increasing and decreasing in similar numbers of patients. In another study,74 in which investigators measured only lumbar spine bone mineral density, significant decreases were recorded at 1 year, unattenuated by high-dose calcium supplementation.
Changes in bone quality and microarchitecture after bariatric surgery In addition to bone density and bone turnover, microarchitecture is an important factor contributing to bone strength and fracture risk.83 Assessment of changes in skeletal microarchitecture can better elucidate the nature of bone loss after bariatric surgery, but data are limited. Bone biopsy samples at baseline and 4 years postoperatively in patients who underwent biliopancreatic diversion with duodenal switch showed decreased cortical thickness and increased trabecular bone volume.80 Mineralisation decreased over the 4 year follow-up, with an increase in osteoid volume and bone formation rate. These data raise the possibility that a decrease in mineralisation could explain some of the 168
postoperative reduction in bone mineral density. Although this mechanism might play a part in individuals who have had biliopancreatic diversion with duodenal switch, defective mineralisation is unlikely to have a big role in the decreased bone mineral density recorded in recent studies of other procedures (LAGB, VSG, or RYGB), given that in response to robust supplementation, most patients had 25-hydroxyvitamin D concentrations close to 50 nmol/L. In our recent prospective study8 of 22 women who underwent RYGB, VSG, and LAGB (14 patients had RYGB, two had VSG, and two had LAGB), with HR-pQCT, we recorded a decrease of cortical area, density, thickness, and total density at the tibia 1 year after surgery. Decreases in cortical bone were predicted by the increase in parathyroid hormone. The subgroup of patients who underwent RYGB lost markedly more weight, had more cortical bone loss than did those with LAGB or VSG and had decreases in cortical load share estimated by finite element analysis.84 HR-pQCT might also be affected by changes in subcutaneous fat.
Mechanisms of bone loss after bariatric surgery Unloading Mechanical loading of bone is an important determinant of bone size, mass, and biomechanical properties. Changes in loading can induce compensatory increases in localised bone remodelling,85 probably mediated through osteocytes and the sclerostin pathway. Bone loss has been reported in the setting of skeletal unloading in other populations, including patients with spinal cord injury,86 and those restricted to bed rest.87 Hip bone loss has been documented in individuals who lose even small amounts of weight from caloric restriction.88–92 The hip typically carries a load roughly two to three times bodyweight;93 therefore, drastic weight loss after bariatric surgery might especially affect this site. A strong association between extent of weight loss after bariatric surgery and amount of bone loss has been documented by most,9,10,43,48,62,67 but not all, studies.63 We found that after RYGB, declines at the total hip (r=0·65, p=0·02) and femoral neck (r=0·90, p
