Rates, Risks, and Time to Fracture in Patients Undergoing Laparoscopic Vertical Sleeve Gastrectomy versus Roux-en-Y Gastric Bypass

Supplemental Digital Content is available in the text.

Abstract

Objective:

To assess the rates, risks, and time to fracture in patients undergoing laparoscopic vertical sleeve gastrectomy (VSG) versus those undergoing Roux-en-Y gastric bypass (RYGB).

Summary Background Data:

Metabolic and bariatric surgery has been implicated in significant bone loss and may increase fracture risk. Preoperative patient characteristics that might impact fracture risk and the time to fractures have not been established. Furthermore, the patient characteristics that might impact fracture risk and the time to fractures by surgical approach are unknown.

Methods:

This population-based retrospective cohort analysis used Humana claims data from January 1, 2007 to March 31, 2017, and included 4073 patients undergoing laparoscopic RYGB and VSG as a first surgical intervention for weight loss. The primary outcomes were the incidence of fractures (Humeral, Radial or Ulnar, Pelvic, Hip, and Vertebral) within 48 months after laparoscopic VSG versus RYGB and days to these fractures.

Results:

An analysis of total fractures (odds ratio [OR] 0.53; 95% confidence interval [CI], 0.38–0.73), vertebral fractures (OR 0.61; 95% CI, 0.38–0.99), hip fractures (OR 0.36; 95% CI, 0.15–0.84), and humeral fractures (OR 0.44; 95% CI, 0.22–0.90) demonstrated a reduction in fracture risk in patients undergoing VSG versus RYGB. Furthermore, postmenopausal status was independently associated with increased odds of total fractures and hip fractures (OR 2.18; 95% CI, 1.06–4.50; OR 5.83; 95% CI, 1.16–29.27; respectively). Likewise, osteoporosis at the time of surgery was associated with increased odds of total fractures (OR 1.61; 95% CI, 1.09–2.37), vertebral fractures (OR 2.01; 95% CI, 1.19–3.39), and hip fractures (OR 2.38; 95% CI, 1.19–4.77). Except for a significantly decreased odds of vertebral fractures in osteoporotic patients undergoing VSG versus RYGB (OR 0.41; 95% CI, 0.18–0.95), osteoporotic or postmenopausal status at the time of surgery was not found to increase odds of fracture depending on surgical intervention. However, time to fracture (total) and for all site-specific fractures, except for pelvic fractures, was significantly reduced in postmenopausal women undergoing RYGB versus VSG. Time to fracture (total) and for all site-specific fractures except pelvic and radial or ulnar fractures was significantly reduced in osteoporotic patients undergoing RYGB versus VSG.

Conclusions and Relevance:

Though bariatric surgery is associated with several health-related benefits, increased fracture risk is an important factor to discuss with patients undergoing bariatric surgery. Bariatric surgery strategy, RYGB versus VSG, carries a differential risk of fracture, with RYGB carrying a higher risk of fracture and decreased time to fracture. Furthermore, patients who are postmenopausal or osteoporotic at the time of surgery carry an increased risk of total fractures, independent of bariatric surgery strategy. Being mindful of patient-specific fracture risk after bariatric surgery may help anticipate, identify, and prevent fractures.

INTRODUCTION

As obesity rates have risen in the United States and around most of the developed world, bariatric surgery has become an increasingly common intervention. Bariatric surgery has demonstrated lasting substantial weight loss and amelioration of and protection from obesity-related diseases.^1–9 However, bariatric surgery has recently been associated with loss of bone mineral density (BMD) and increased rates of bone fracture.^10–12 Bredella et al found that lumbar spine, total hip, and femoral neck BMD declined significantly in patients who underwent Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy.¹² Rousseau et al demonstrated that the risk of fracture in patients who underwent bariatric surgery was 1.44 times the risk of nonobese patients (relative risk 1.44; 95% confidence interval [CI], 1.29–1.59).¹⁰ In a propensity-matched cohort study, Lu et al found that bariatric surgery patients had a 1.21-fold increase in fracture risk compared with matched obese controls (adjusted hazard ratio 1.21; 95% CI, 1.01–1.44).¹¹ These results are supported by several other studies.^13–23 The importance of bariatric surgery’s impact on bone fracture risk is highlighted by research showing that bone fractures negatively impact patients’ health-related quality of life and increase healthcare utilization.^24–26

To date, there is a paucity of information available quantifying the risk of site-specific fracture based upon type of bariatric surgery or patient characteristics. Similarly, little is known about whether the surgical approach or patient characteristics impact the time to fracture. Only one study has evaluated the risk fracture in patients undergoing the two most common surgical approaches, laparoscopic RYGB and vertical sleeve gastrectomy (VSG).¹⁵ However, the results of that study may have been confounded by the inclusion of traumatic fractures (including face/skull, ribs, and sternum). Moreover, site-specific fracture risk was not assessed. Studies of this nature aid in risk-benefit conversations with patients and could provide a basis for critical decision making when approaching bariatric surgery.

To this end, here we compare site-specific fracture risk in patients undergoing laparoscopic RYGB versus those undergoing VSG, quantify the potential contribution of patient characteristics to this risk, and explore how time to fracture is impacted by surgical approach and a patient’s characteristics.

METHODS

Data Source

This study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines. Longitudinal Analytic Files containing 100% of all inpatient, outpatient, drug, and laboratory claims of the 22 million patients insured by Humana, an American health insurance company, between January 1, 2007 and March 31, 2017 were utilized in this cohort study. The study was approved by the Rush University Medical Center Institutional Review Board with a waiver of patient informed consent, as the nature of this analysis posed minimal risk to participating individuals, and the data were presented in aggregate to minimize any risk of loss of confidentiality of medical data.

Study Cohort

Adult patients who underwent a primary RYGB or VSG procedure were identified using International Classification of Diseases, Ninth Revision (ICD-9) and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) diagnostic and procedure codes, and Current Procedural Terminology codes. In this study that investigated the potential impact of osteoporosis (among other patient comorbidities) on bone fracture risk, patients of age 30–74 were included in the study cohort, considering peak bone mass is achieved at approximately 30 years of age and the diagnosis of osteoporosis is made by comparing patient bone density to healthy 30-year-old individuals of the same sex.^27–29 Exclusion criteria were as follows: patients without an active insurance record for at least 48 months following their index procedure, those with missing demographic information, those who had a prior bariatric procedure, and those who underwent surgery for indications other than weight loss. The patient cohort selection procedure is summarized in Figure 1 and further detailed in Supplementary Table 2, http://links.lww.com/AOSO/A75. Supplementary Table 3, http://links.lww.com/AOSO/A75 details the diagnostic and procedural codes used to define patient comorbidities at the time of surgery (i.e., hypertension, history of tobacco use, nonalcoholic fatty liver disease [NAFLD], hyperlipidemia, type 2 diabetes [T2D], osteoporosis, osteoarthritis, postmenopausal status, and obstructive sleep apnea [OSA]). The procedures and definitions used for cohort selection and comorbidity identification have been previously utilized in similar investigations.^30,31

Figure 1.

Patient selection procedure.

Outcome Definition

The primary outcomes of this study were the incidence of fractures within 48 months after laparoscopic bariatric surgery and time to fracture (days). Although incidence of fractures may be indicative of the magnitude of loss of BMD, time to fracture may be indicative of the rate of loss of BMD. Types of fractures most attributable to osteoporosis/loss of BMD were identified using ICD-9 and ICD-10 codes, these included: Humeral, Radial and Ulnar, Pelvic, Hip, and Vertebral fractures (Supplemental Table 4, http://links.lww.com/AOSO/A75).^32,33 This strategy utilizing claims-based algorithms to identify fractures has a high positive predictive value for these kinds of fractures and has been previously employed in this manner to study fracture risk.^13,31,34,35 Secondary outcomes included the assessment and evaluation of site-specific fracture risk, time (in days) to these fractures, and patient risk factors to fracture type.

Statistical Analysis

χ² tests were used to compare categorical variables including age ranges, sex, race, and comorbidity status (hypertension, history of tobacco use, NAFLD, hyperlipidemia, T2D, osteoporosis, osteoarthritis, postmenopausal status, and OSA), and Student’s T-tests were used for quantitative variables. Odds ratios were calculated to compare fracture events based upon surgical procedure: RYGB versus VSG. Logistic regression models were developed to assess the risk factors independently associated with each fracture type (humeral, radial and ulnar, pelvic, hip, vertebral, and total fractures), and included age ranges, race, gender, surgical procedure, and comorbidity status. Interquartile ranges were calculated and plotted to demonstrate the distribution of time to fracture of each fracture type and assessed against surgical procedure and risk factors as identified by our logistic regression model. Data were analyzed using R statistical software version 3.6 (The R Foundation).

RESULTS

Descriptive Characteristics

A total of 14,965 patients were identified as having undergone laparoscopic RYGB or VSG. Following application of the exclusion criteria, 982 (24.1%) and 3091 (75.9%) patients were identified as having undergone laparoscopic VSG and RYGB, respectively (Fig. 1). The demographic distribution and comorbidity status at the time of surgery are summarized in Table 1.

TABLE 1.

Descriptive Characteristics of Patients Undergoing Laparoscopic Roux-en-Y Gastric Bypass or Vertical Sleeve Gastrectomy

Parameters	Total (N = 4073)	Roux-en-Y Gastric Bypass (N = 3091)	Vertical Sleeve Gastrectomy (N = 982)	P value
Age				<0.01*
30–39, n (%)	306 (7.5)	243 (7.9)	63 (6.4)
40–49, n (%)	778 (19.1)	594 (19.2)	184 (18.7)
50–59, n (%)	1402 (34.4)	1085 (35.1)	317 (32.3)
60–69, n (%)	1370 (33.6)	1025 (33.3)	342 (34.8)
70–74, n (%)	217 (5.3)	141 (4.6)	76 (7.7)
Sex				0.31
Male, n (%)	1128 (27.7)	869 (28.1)	259 (26.4)
Female, n (%)	2945 (72.3)	2222 (71.9)	723 (73.6)
Race				0.13
White, n (%)	3067 (75.3)	2337 (75.6)	730 (74.3)
Black, n (%)	544 (13.4)	394 (12.7)	150 (15.3)
Hispanic, n (%)	64 (1.6)	52 (1.7)	12 (1.2)
Other, n (%)	24 (0.6)	24 (0.8)	—
Unknown, n (%)	374 (9.2)	284 (9.2)	90 (9.2)
Comorbidities
Hypertension, n (%)	3627 (89.0)	2753 (89.1)	874 (89.0)	>0.99
History of tobacco use, n (%)	471 (11.6)	338 (10.9)	133 (13.5)	0.03*
NAFLD, n (%)	43 (1.1)	30 (1.0)	13 (1.3)	0.44
Hyperlipidemia, n (%)	3195 (78.4)	2417 (78.2)	778 (79.2)	0.52
Osteoporosis, n (%)	294 (7.2)	221 (7.1)	73 (7.4)	0.82
Osteoarthritis, n (%)	2067 (50.7)	1543 (49.9)	524 (53.4)	0.065
Postmenopausal, n (%)	2157 (53.0)	1609 (52.1)	548 (55.8)	0.04*
OSA, n (%)	2225 (54.6)	1664 (53.8)	561 (57.1)	0.077
DM (II), n (%)	2622 (64.4)	2017 (65.3)	605 (61.6)	0.04*
Fractures
Humerus, n (%)	72 (1.8)	63 (2.0)	9 (0.9)	0.03*
Radius or Ulnar, n (%)	46 (1.1)	37 (1.2)	9 (0.9)	0.58
Pelvic, n (%)	18 (0.4)	15 (0.5)	3 (0.3)	0.64
Hip, n (%)	58 (1.4)	52 (1.7)	6 (0.6)	0.02*
Vertebral, n (%)	112 (3.0)	102 (3.3)	20 (2.0)	0.055
Total fractures, n (%)	316 (7.8)	269 (8.7)	47 (4.8)	<0.001*
Total patients with fractures, n (%)	273 (6.7)	230 (7.4)	43 (4.4)	<0.001*

*Significant values (P < 0.05).

NAFLD, nonalcoholic fatty liver disease; OSA, obstructive sleep apnea.

The majority of patients were female (72.3%) and did not differ between procedure type (VSG 73.6% vs RYGB 71.9%, P = 0.31). Patients undergoing VSG tended to be older at the time of their surgery compared to RYGB patients (P ≤ 0.001). There was no difference in the distribution of race between procedure type (P = 0.13).

Rates of T2D (61.6% vs 65.3%, P = 0.04), postmenopausal status (55.8% vs 52.1%, P = 0.04), and history of tobacco use (13.5% vs 10.9%, P = 0.03) were significantly different between patients undergoing VSG versus those who underwent RYGB, respectively. No significant differences between rates of hypertension (89% vs 89.1%, P = 1), NAFLD (1.3% vs 1%, P = 0.44), hyperlipidemia (79.2% vs 78.2%, P = 0.52), osteoporosis (7.4% vs 7.1%, P = 0.82), osteoarthritis (53.4% vs 49.9%, P = 0.065), or OSA (57.1% vs 53.8%, P = 0.077) were observed (Table 1).

Rates and Risk of Fracture

The incidence of total fractures in patients who underwent RYGB was 18.6 per 1000 patient-years, whereas the incidence in patients who underwent VSG was 10.9 per 1000 patient-years (Table 1). The odds of total fractures within 48-months after bariatric surgery were lower in patients undergoing VSG compared to RYGB (OR 0.53; 95% CI, 0.38–0.73). Likewise, odds of humeral (OR 0.44; 95% CI, 0.22–0.90), hip (OR 0.36; 95% CI, 0.15–0.84), and vertebral fractures (OR 0.61; 95% CI, 0.38–0.99) were significantly lower in patients undergoing VSG as compared to RYGB (Supplemental Table 5, http://links.lww.com/AOSO/A75).

In multivariable analysis, patients who had VSG had reduced odds of total fractures (OR 0.56; 95% CI, 0.40–0.79), vertebral fractures (OR 0.59; 95% CI, 0.36–0.96), hip fractures (OR 0.35; 95% CI, 0.15–0.82), and humeral fractures (OR 0.44; 95% CI, 0.22–0.89) compared with patients who had RYGB. No significant difference was observed between the two surgical procedures in the odds of pelvic or radial and ulnar fractures. These analyses also demonstrated an increased odds of total fractures (OR 2.18; 95% CI, 1.06–4.50) and hip fractures in postmenopausal women (OR 5.83, 95% CI 1.16 to 29.27); an increased odds of total fractures (OR 1.61; 95% CI, 1.09–2.37), vertebral fractures (OR 2.01; 95% CI, 1.19–3.39), and hip fractures (OR 2.38; 95% CI, 1.19–4.77) in patients who had a reported diagnosis of osteoporosis at the time of surgery; and an increased odds of vertebral fractures (OR 2.09; 95% CI, 1.05–3.80) in patients who had a reported diagnosis of hyperlipidemia at the time of surgery. Patients whose race was black were found to have lower odds of total fractures compared to patients whose race was white (OR 0.42; 95% CI, 0.25–0.69). All other relationships were not found to be significant (Fig. 2, Supplemental Table 1 http://links.lww.com/AOSO/A75).

Figure 2.

Logistic regression of factors’ contributing to fracture within 48 months after either vertical sleeve gastrectomy vs Roux-en-Y gastric bypass (Reference values).

Time to Fracture

Days to total fractures (559.15 ± 489.06 vs 788.68 ± 361.45 days, P ≤ 0.001), hip fractures (585.56 ± 481.06 vs 64.61 ± 245.94 days, P = 0.045), humeral fractures (554.6 ± 497.70 vs 941.17 ± 350.77 days, P ≤ 0.001), and vertebral fractures (515.56 ± 486.59 vs 807.75 ± 375.86 days, P ≤ 0.001) were found to be significantly higher in patients undergoing VSG as compared to RYGB, respectively. No significant difference was observed in days to pelvic or radial and ulnar fractures when comparing procedure type (Fig. 3).

Figure 3.

Distribution of time to fracture within 48 months after either vertical sleeve gastrectomy versus Roux-en-Y gastric bypass *Significant values (P < 0.05).

Fracture Risk and Timing in Postmenopausal and Osteoporotic Patients

Multivariable analysis of postmenopausal patients undergoing VSG as compared to those undergoing RYGB demonstrated no independent increased odds of site-specific fracture or total fractures associated with bariatric surgery type (Fig. 4A). Time to fracture (days) was significantly increased in postmenopausal patients who had VSG compared to RYGB, for total fractures (VSG: 793.96 ± 359.25 vs RYGB: 580.34 ± 481.34, P ≤ 0.001), and in all site-specific fractures except pelvic fractures (Fig. 4B).

Figure 4.

(A) Multivariable interaction and (B) Distribution of time to fracture within 48 months after either vertical sleeve gastrectomy versus Roux-en-Y gastric bypass in postmenopausal females and patients who had a diagnosis of osteoporosis at the time of surgery. *Significant values (P < 0.05).

Patients who were osteoporotic at the time of RYGB were found to have higher odds of vertebral fractures compared to osteoporotic patients undergoing VSG (OR 0.41; 95% CI, 0.18–0.95). Otherwise, osteoporotic patients undergoing RYGB were not found to have higher odds of site-specific fracture or total fractures based upon surgery type (Fig. 4A). However, time to fracture (days) was significantly increased in osteoporotic patients who had VSG compared to RYGB, for total fractures (VSG: 705.03 ± 379.66 vs RYGB: 521.00 ± 470.23, P = 0.002), and in all site-specific fractures except for pelvic and radial or ulnar fractures (Fig. 4B).

DISCUSSION

The effect of bariatric surgery on bone health is complex. Calcium and vitamin D deficiencies can cause secondary hyperparathyroidism and lead to excessive bone resorption.¹⁹ Bariatric procedures cause several hormonal changes that may impact bone health. Increased levels of adiponectin after bariatric surgery have been correlated with lower BMD.^36,37 Peptide YY inhibits the activity of osteoblasts and postprandial levels have been found to increase after bariatric surgery.^38,39 Other postbariatric surgery hormonal changes that may impact bone health include decreased postprandial leptin, increased postprandial glucagon-like peptides 1 and 2, decreased ghrelin, decreased fasting insulin, and differential changes in gonadal steroids.^{19,36,38,40–47} In addition, mechanical unloading due to weight loss contributes to bone loss.¹⁹

In a 2018 meta-analysis, Zhang et al demonstrated an increased risk of total fractures in bariatric surgery patients versus nonsurgical controls (relative risk 1.29; 95% CI, 1.18–1.42).²³ In the present study of 4073 privately insured patients, fracture risk was compared between the two most common bariatric procedures: laparoscopic RYGB and VSG. Patients who underwent RYGB had a higher incidence and odds of humeral, hip, vertebral, and total fractures. Time to fracture was also significantly reduced in patients undergoing RYGB as compared to those who underwent VSG. These findings are consistent with previous studies that have shown an increase in bone turnover markers and a decrease in areal BMD in patients undergoing malabsorptive procedures such as RYGB versus restrictive procedures such as VSG.^12,18,48 Indeed, the malabsorptive nature of RYGB may explain the increased risk of fracture and decreased time to fracture found in our study.

The current study provides highly valuable information to the clinical practice of metabolic and bariatric surgery by specifying laparoscopic RYGB and VSG-specific fracture outcomes and fracture timing. To the knowledge of the authors, only one study has been published that evaluated the comparative risk of any fracture in patients undergoing RYGB or VSG between 1985 and 2015 at a single center by Fashandi et al.¹⁵ In this study of 3339 patients, an all-fracture risk was assessed, which included upper and lower extremities, hip, facial/skull, pelvis, spinal, rib, or sternal fractures, finding that patients undergoing RYGB had a higher odds of fracture compared to those undergoing VSG (OR 2.18; 95% CI, 1.05–4.53). However, this study had several limitations. First, given the inclusion of fractures of the face/skull, ribs, spine, and sternum, the outcome of that study may have been confounded by its inclusion of these traumatic fractures. Second, the associated patient-follow-up rates in the Fashandi et al study were limited to those patients who would have presented back to a single center for fracture care. Third, VSG is a relatively newer procedure and little data have historically been available to characterize its effects on bone health.¹³ This was the case in the Fashandi et al study, where the proportion of VSG cases was low (6.8%). Our study is likely more reflective of surgical care today (24.1% of cases were VSG). As laparoscopic VSG has become increasingly common, eclipsing RYGB in popularity, the skeletal effects associated with VSG are emerging.^{2,10,19,20,49}

Previous studies have examined the potential contributions of patient demographics (age, sex, and race), comorbidity status, and insurance status to the risk of the development of fractures after bariatric surgery.^{11,13,15,16,50} A recent study of almost 30,000 Medicare beneficiaries demonstrated no differential risk between older and younger adults undergoing RYGB, but described a 73% increased risk of nonvertebral fracture after RYGB, which was preserved across age groups, proportionally.¹³ Likewise, in a recent study by Yu et al assessing fracture risk in 23,263 patients undergoing RYGB or AGB, age was not an independent risk factor for fracture.⁵⁰ Similarly, the present study did not find age to be independently associated with increased odds of fracture. Demographics, insurance status, and comorbidity status were evaluated by Fashandi et al as potential risk factors for fracture after bariatric surgery. Specifically, their analysis demonstrated a potential decreased independent risk associated with white race (OR 0.68; 95% CI, 0.46–0.996) and private insurance status (OR 0.69; 95% CI, 0.29–0.91), and potential increased risk of fracture with tobacco use (OR 3.12; 95% CI, 1.63–5.97).¹⁵ The present study demonstrated an increased odds of fracture in patients with osteoporosis or who were postmenopausal, regardless of bariatric surgery type.

Postmenopausal status and osteoporosis, in the context of medical weight loss and bariatric surgery, have been thought to precipitate increased fracture risk secondary to resulting changes in hormones (estrogens and androgens), elements of malabsorption of micronutrients (vitamin D, protein, and calcium), and the further reduction of loading stress placed upon bones.^{16,19,21,51,52} In the present study, osteoporotic status was independently associated with hip, vertebral, and total fractures, and postmenopausal status was independently associated with hip fractures and total fractures. Osteoporotic patients undergoing RYGB were significantly more likely to experience vertebral fractures compared to those undergoing VSG. Otherwise, patients undergoing RYGB who had a pre-existing diagnosis of osteoporosis or who were postmenopausal at the time of surgery were not significantly more likely to experience fractures compared to patients undergoing VSG. However, postmenopausal patients undergoing RYGB experienced hip, humeral, radial or ulnar, vertebral, and total fractures earlier compared to those undergoing VSG. Similarly, osteoporotic patients experienced hip, humeral, vertebral, and total fractures earlier compared to those undergoing VSG. The increased risk of fracture seen in osteoporotic and postmenopausal patients regardless of the type of surgery may be explained by these patients’ decreased bone reserve at baseline, while decreased time to fracture in RYGB versus VSG may be explained by the malabsorptive nature of RYGB. Although our study did not find an increased risk of fracture in osteoporotic and postmenopausal patients undergoing RYGB versus VSG, further investigation may be required to rule out a potential association. Taken as a whole, these findings suggest a potential need for thoughtful procedural selection and tailored postoperative management for patients with osteoporosis or who are postmenopausal.

Current recommendations regarding the evaluation and monitoring of skeletal parameters in patients undergoing bariatric surgery vary widely (Table 2).^53–58 The majority of guidelines recommend baseline measurement of calcium, 25 (OH) vitamin D, and parathyroid hormone, with follow-up measurements at various intervals. Guidelines from the American Association for Clinical Endocrinologists, The Obesity Society, American Society for Metabolic & Bariatric Surgery, Obesity Medicine Association, and American Society of Anesthesiologists (AACE/TOS/ASMBS/OMA/ASA) recommend monitoring 24-hour calcium urine excretion after biliopancreatic diversion (BPD) or to evaluate for bone loss after any bariatric procedure, and the ASMBS recommends consideration of 24-hour calcium urine excretion preoperatively or to evaluate for bone loss in all bariatric procedures.^53,58 Regarding monitoring bone turnover markers, the Interdisciplinary European (IDE) guidelines recommend monitoring bone alkaline phosphatase after BPD, and the ASMBS recommends considering perimenopausal and postmenopausal women and patients at risk of osteoporosis for monitoring C-terminal telopeptide of type 1 collagen (CTX).^56,58 The AACE/TOS/ASMBS/OMA/ASA guidelines recommend follow-up BMD measurement for all bariatric procedures at 2 years, and the ASMBS guidelines are in agreement.^53,58 Only the IDE guidelines recommend baseline BMD measurement.⁵⁶ With few exceptions detailed above, current guidelines do not tailor recommendations for screening and monitoring bone health to specific patient risk factors.

TABLE 2.

Guidelines for Serum Calcium, 25 (OH) Vitamin D, PTH, and BMD Screening and Monitoring Postbariatric Surgery

CPG	Calcium	25 OH Vitamin D	PTH	BMD
AACE/TOS/ASMBS/OMA/ASA (2019)53	Baseline: All	Baseline: All	Baseline: NA	Baseline: Same as the NOF indications
	Follow-up: All at each visit*	Follow-up: RYGB and BPD/DS at 1, 3, and 6–12 months thereafter	Follow-up: RYGB and BPD/DS at 1, 3, and 6–12 months thereafter	Follow-up: All at 2 years*
BOMS (2014, 2020)54,55	Baseline: All	Baseline: All	Baseline: All	Baseline BPD/DS: Consider in specific patient groups
	Follow-up: All at 3, 6, 12 months then annually	Follow-up: All at 3, 6, 12 months then annually	Follow-up: All if not checked before surgery	Follow-up: NA
IDE (2014)56	Baseline: NA†	Baseline: NA	Baseline: NA	Baseline: All
	Follow-up RYGB: annually	Follow-up RYGB: annually	Follow-up RYGB: annually	Follow-up: NA‡
	BPD: at 1, 4, and 12 months then annually	BPD: at 1, 4, and 12 months then annually	BPD: at 1, 4, and 12 months then annually
OBN (2016)57	Baseline: All	Baseline: All	Baseline: All	NA
	Follow-up: All at 6, 12 months then annually up to 5 years	Follow-up: All at 3, 6, 12 months then annually up to 5 years	Follow-up: All at 6, 12 months then annually up to 5 years
ASMBS (2020)58	Baseline: All	Baseline: All	Baseline: All	Baseline: Same as the NOF indications
	Follow-up: All at least annually§	Follow-up: All every 3–6 months in the first year and then annually	Follow-up: All at least annually	Follow-up: RYGB, BPD-DS at 2 years§,‖,¶

*The AACE/TOS/ASBMS/OMA/ASA guidelines recommend 24-hour calcium urine excretion after BPD or to evaluate for bone loss.

†The IDE guidelines recommend routine preoperative assessment, but do not specify the specific tests.

‡The IDE guidelines include monitoring bone alkaline phosphatase at 1, 4, and 12 months post-BPD, and annually thereafter.

§The ASBMS guidelines recommend consideration of 24-hour urinary calcium in relationship to dietary intake preoperatively, and postoperatively to monitor bone loss.

‖The ASBMS guidelines state agreement with bone density monitoring recommendations set forth by the AACE/TOS/ASBMS/OMA/ASA Clinical Practice Guidelines.

¶The ASBMS guidelines include recommendations to monitor for bone loss with C-terminal telopeptide of type 1 collagen (CTX) in perimenopausal and postmenopausal women and patients at risk for osteoporosis.

AACE/TOS/ASMBS/OMA/ASA, American Association of Clinical Endocrinologists, The Obesity Society, American Society For Metabolic & Bariatric Surgery, Obesity Medicine Association, and American Society of Anesthesiologists; AGB, Adjustable Gastric Band; ASBMS, American Society for Metabolic and Bariatric Surgery; BMD, bone mineral density; BOMSS, British Obesity and Metabolic Surgery Society; BPD/DS, biliopancreatic diversion/duodenal switch; CPG, Clinical Practice Guidelines; ES, Endocrine Society; IDE, Interdisciplinary European; NOF, National Osteoporosis Foundation; OBN, Ontario Bariatric Network; PTH, parathyroid hormone; RYGB, Roux-en-Y gastric bypass; SG, sleeve gastrectomy.

Similarly, recommendations for postsurgical supplementation are variable and not tailored to specific patient risk factors (Table 3).^53,55,56,58 Calcium and vitamin D supplementation postbariatric surgery has been shown to decrease bone turnover markers, and vitamin D supplementation has been shown to decrease bone loss.^59–61 Recommendations for calcium intake range from 1200 mg/d from dietary sources to 2400 mg/d calcium citrate supplementation. Where specified, guidelines agree on vitamin D supplementation with target 25(OH)D > 30 ng/ml, recommending a minimum of 2000 to 3000 IU/d to reach goal.^53,55,58 The AACE/TOS/ASMBS/OMA/ASA guidelines recommend that antiresorptive agents (bisphosphonates or denosumab) only be considered in osteoporotic patients and after vitamin D and calcium repletion; other guidelines do not mention antiresorptive agents. The findings presented in this study indicate that osteoporotic and postmenopausal patients may benefit from bone health screening, monitoring, and supplementation regimens that more specifically address their increased bone fracture risk.

TABLE 3.

Guidelines for Calcium and Vitamin D Supplementation Postbariatric Surgery

CPG	Calcium	Vitamin D
AACE/TOS/ASMBS/OMA/ASA (2019)53	LAGB, SG, RYGB: 1200–1500 mg/d (citrate preferred)	All: Vitamin D3, at least 3000 IU/d, titrate to 25(OH)D > 30 ng/ml
	BPD/DS: 1800–2400 mg/d (citrate preferred)
BOMSS (2020)55	AGB, SG, RYGB: 1200–1500 mg calcium/d from food and supplements.	All: Adjust vitamin D3 supplementation to maintain 25(OH)D > 75 nmol/L. “Maintenance levels 2000–4000 IU/d may be required after SG and RYGB and higher after malabsorptive procedures such as BPD/DS”
	BDP/DS: 1800–2400 mg/d
IDE (2014)56	BPD: Calcium 2000 mg/d (citrate preferred)	AGB, RYGB: “vitamin and micronutrient supplements (oral) should routinely be prescribed to compensate for their possible reduced intake and absorption
		BPD: “Lifelong daily vitamin and micronutrient supplementation (vitamins should be administered in water-soluble form): Vitamins A, D, E, and K”
ASMBS (2020)58	LAGB, SG, RYGB: 1200–1500 mg/d (citrate preferred)	All: Vitamin D3, at least 3000 IU/d, titrate to 25(OH)D > 30 ng/ml
	BPD/DS: 1800–2400 mg/d (citrate preferred)

Sentences in quotations are taken verbatim from original documents.

AACE/TOS/ASMBS/OMA/ASA, American Association of Clinical Endocrinologists, The Obesity Society, American Society For Metabolic & Bariatric Surgery, Obesity Medicine Association, and American Society of Anesthesiologists; AGB, adjustable gastric band; ASBMS, American Society for Metabolic and Bariatric Surgery; BOMSS, British Obesity and Metabolic Surgery Society; BPD/DS, biliopancreatic diversion/duodenal switch; CPG, Clinical Practice Guidelines; ES, Endocrine Society; IDE, Interdisciplinary European; LAGB, laparoscopic adjustable gastric band; NOF, National Osteoporosis Foundation; OBN, Ontario Bariatric Network; RYGB, Roux-en-Y gastric bypass; SG, sleeve gastrectomy.

The current study provides valuable information that should be included in shared decision-making when choosing between bariatric surgical options. These points include: (1) RYGB is associated with a greater risk of postoperative humeral, hip, vertebral, and total fractures compared to VSG; (2) osteoporotic patients are at increased risk of postoperative hip, vertebral, and total fractures regardless of the type of surgery; and (3) postmenopausal patients are at increased risk for postoperative hip and total fractures regardless of the type of surgery.

Limitations

Although the retrospective nature of the study is a limitation, as retrospective studies may not allow for the full control of selection bias in the use of different surgical approaches for different patients, the large sample size functions as a notable strength of the study. The analysis was restricted to comparing RYGB to VSG, and comparisons between bariatric surgical patients and nonsurgical controls were not possible. Administrative data allow access to more medical visits nationwide and longitudinal tracking of patients through distinct identifiers based on a standardized coding system. Important limitations in the use of this data must be considered. Primary administrative data do not provide enough specific details on the severity of disease states (e.g., degree of osteoporosis or months of amenorrhea when accessing postmenopausal status), patient-reported outcome scores, or allow for standardization of treatment protocols or surgeon technique and experience, which may mask certain confounding factors. In addition, administrative data limit the assessment of baseline weight, BMI, amount of weight lost as a result of bariatric surgery, or specific bone density measurements and so we are unable to directly assess any potential associations between these variables and fracture risk. Furthermore, patients undergoing VSG and RYGB on average lose 18% and 25% of their body weight at 5 years after surgery, but do so at different weight loss velocities, and so the complex relationship between absolute weight loss, weight loss velocity, and body mass composition cannot be rigorously explored using this kind of data.^7,9,62–64

Conclusions

Although bariatric surgery is associated with several health-related benefits, increased fracture risk is an import factor to discuss with prospective bariatric patients, especially those who are postmenopausal or osteoporotic at the time of surgery. Patients who undergo RYGB for weight loss should be managed with a specific focus on their bone health, as the risk of fracture seems to be higher compared to VSG. This finding expands on previous studies that demonstrate increased bone turnover markers in patients undergoing RYGB versus VSG; however, the exact mechanism underlying this difference remains unclear and further investigation is warranted. Furthermore, additional investigation of site-specific fractures and associated risk factors, as well as patient surveillance paradigms are required to further guide recommendations for supporting optimal bone health after metabolic and bariatric surgery.

Supplementary Material

as9-2-e099-s001.pdf ^{(662.2KB, pdf)}