Risedronate use may blunt appendicular lean mass loss secondary to sleeve gastrectomy: Results from a pilot randomized controlled trial

Abstract

Background:

Despite robust weight loss and cardiometabolic benefit, lean mass loss following sleeve gastrectomy (SG) confers health risk. Bisphosphonates are a potential therapeutic agent for lean mass maintenance. Thus, our objective was to explore the effect of six months of risedronate (vs placebo) on change in dual energy x-ray absorptiometry (DXA) and computed tomography (CT) derived lean mass metrics in the year following SG.

Methods:

24 SG patients were randomized to six months of 150 mg oral risedronate or placebo capsules (NCT03411902). Body composition was assessed at baseline and six months with optional 12-month follow-up using whole-body DXA and CT at the lumbar spine and mid-thigh. Group treatment effects and 95% CIs were generated from a mixed model using contrast statements at six and 12 months, adjusted for baseline values.

Results:

Of 24 participants enrolled [55.7±6.7 years (mean±SD), 79% Caucasian, 83% women, body mass index (BMI) 44.7±6.3kg/m²], 21 returned for six-month testing, and 14 returned for 12-month testing. Six-month weight loss was −16.3 kg (−20.0, −12.5) and −20.9 kg (−23.7, −18.1) in the risedronate and placebo groups, respectively (p=.057). Primary analysis at six-months revealed a non-significant sparing of appendicular lean mass in the risedronate group compared to placebo [−1.2 kg (−2.3, −0.1) vs −2.1 kg (−3.0, −1.2)]; p=.20. By 12-months, the risedronate group displayed no change in appendicular lean mass from baseline [−0.5 kg (−1.5, 0.6)]; however, the placebo group experienced significantly augmented loss [−2.9 kg (−3.6, −2.1)].

Conclusion:

Pilot data indicate risedronate treatment may mitigate appendicular lean mass loss following SG. Further study is warranted.

Introduction:

Due to increasing prevalence of severe obesity in the U.S., there has been dramatic growth in metabolic surgery.¹ Sleeve gastrectomy (SG) is the most popular surgical option for the treatment of obesity and related comorbidities, with a 450% increase in SG surgeries performed since 2011.¹ Although SG provides effective, long-term weight loss and cardiometabolic benefit, it also confers physiologic risk, including loss of bone and lean mass.^{2, 3}

Losing some lean body mass is an anticipated consequence following SG surgery due to the magnitude and speed of weight loss^{4, 5}; however, excessive lean mass loss can diminish functional capacity, slow resting metabolism, and hinder the ability to perform activities of daily living.^{3, 6} These adverse consequences are in direct opposition to the goals of bariatric surgery, which include weight loss to resolve comorbidities and, ultimately, improve quality of life.⁷ Thus, identification of therapeutic interventions that minimize lean mass loss following SG is important and timely, with the potential to improve long-term surgical outcomes.^{8, 9}

Bisphosphonates, which are traditionally used to treat low bone mass,¹⁰ have emerged as a potential therapeutic for maintaining lean mass. Data from several pre-clinical studies suggest that bisphosphonates can maintain lean mass in murine models of cancer cachexia, aging, immobility, and muscular dystrophy.^11–14 Likewise, observational data in humans show that among osteoporotic women taking weekly alendronate for one year, appendicular lean mass is increased 2.5-fold as compared to non-treated women.¹⁵ A variety of mechanisms have been proposed to explain the muscle preservation experienced with bisphosphonate use. Proposed mechanisms include downregulation of SIRT-3 expression, reduction of Smad2/3 protein levels, and anti-inflammatory effects of bisphosphonates themselves.^{11, 13, 16} However, bisphosphonate use for the mitigation of SG-associated lean mass loss has not yet been examined, presenting a gap. Although little has been done to investigate the role of bisphosphonates for lean mass preservation in humans, the findings from prior observational studies offer a starting point for trial work to begin to fill in gaps.

The primary purpose of this randomized controlled trial (RCT) was to explore the effect of six months of bisphosphonate use (versus placebo) on change in dual-energy x-ray absorptiometry (DXA) and computed tomography (CT) acquired muscle indices in the six months following SG. DXA measures included total lean mass, total fat mass, and appendicular lean mass, and CT measures included muscle size and muscle attenuation at the lumbar spine and mid-thigh. In secondary analyses, we examined extended effects of six months of bisphosphonate treatment on the same battery of bioimaging-acquired body composition parameters at 12-months post-SG. We hypothesized that participants in the risedronate group will better maintain lean mass during rapid weight loss, while those in the control group will lose lean mass during rapid weight loss, and that these findings will be maintained throughout the 12-month follow-up period.

Methods:

Population and Study Intervention

The Weight Loss with Risedronate for Bone Health (WE RISE; NCT03411902) pilot RCT design has been previously described.¹⁷ Briefly, middle-aged and older adults (≥40 years) undergoing SG, weighing <205 kg, without history of bone disorders including osteoporosis, met inclusion criteria for this study. Following informed consent and physician clearance, enrolled participants were randomized, via computer generated block randomization (stratified by sex) to 150 mg oral risedronate or placebo capsules. The risedronate and placebo capsules were identical (150mg in size) and dispensed by the study institution’s research pharmacy. Capsules were administered 3–7 days pre-operatively, with the first dose taken before surgery and monthly thereafter for six total doses. Bone measures and body composition were assessed at baseline and six months with optional 12-month follow-up, using DXA and CT scans. All eligible and interested participants read and signed an IRB-approved (protocol #48310) informed consent document prior to enrollment.

DXA Acquisition

All DXA-acquired body composition metrics (including total lean mass and total fat mass) were collected at baseline (N=24) and six months (N=21) with optional 12-month follow-up (N=14), and analyzed using a single scanner (iDXA; GE Medical Systems, Madison, WI, USA) in accordance with the International Society for Clinical Densitometry.¹⁸ Coefficients of variation (CV) from repeated measurements are 1.38% for lumbar spine areal bone mineral density (aBMD), 1.21% for total hip aBMD, and 1.82% for femoral neck aBMD. If the participant exceeded the field of view, a full-view scan of the right side of the participant’s body was acquired and total body values were determined utilizing a mirroring protocol.^{17, 19} Appendicular lean mass, which represents the sum of the lean tissue in the arms and legs, was also collected from DXA.²⁰ Appendicular lean mass index (ALMI) was subsequently calculated as appendicular lean mass over height squared (kg/m²).

CT Acquisition

All CT-acquired outcome measures were collected at baseline (N=21) and six months (N=20) with optional 12-month follow-up (N=12). For CT-acquired measures (trunk muscle cross-sectional area, trunk muscle attenuation, mid-thigh muscle cross-sectional area, mid-thigh muscle attenuation), helical scans of the bilateral hips (superior acetabulum to mid-femur) and lumbar spine (L1-L5) were collected on a single CT scanner (Siemens SOMATOM Definition Flash CT, Siemens Healthineers, Erlangen, Germany). Scans were collected using a standard abdominal reconstruction filter at 120 kV, 350 mA, and a slice thickness of 0.625 mm (hips) and 2 mm (spine).¹⁷

Trunk and Mid-Thigh Muscle Segmentation

Muscle segmentation was performed on axial CT images using Mimics software (v.23; Materialise, Leuven, Belgium) by a single reader (LEF). After excluding visceral contents, semi-automated radiodensity thresholding was used to isolate trunk skeletal muscle (Hounsfield unit (HU) range: −29 to 150) at the L3 vertebral level. The trunk at L3 was chosen as it is a reliable surrogate for total body skeletal muscle.²¹ Trunk muscles were segmented on three adjacent mid-vertebral axial slices of the L3 vertebra, and included the following muscles: psoas, quadratus lumborum, erector spinae, and those of the abdominal wall (external and internal obliques, rectus abdominus, transversus abdominus).²² Trunk muscle cross-sectional area (CSA) measurements were calculated by summing muscle pixels in the segmentation and multiplying by pixel surface area. The mean of three segmented slices was calculated to represent mean trunk CSA. Trunk skeletal muscle index (SMI) was calculated as CSA divided by the participant’s height squared²³. Mean muscle attenuation was reported as the average radiodensity (HU) within the total trunk skeletal muscle.²⁴ Segmentation at the mid-thigh was likewise performed on the CT images. Mid-thigh was identified at the midpoint between the lesser trochanter and the intercondylar fossa of the femur. The right and left thigh were segmented separately. If participants exceeded the CT field of view, segmentation was performed on only one leg within the field of view. Mid-thigh skeletal muscle cross-sectional area and mean thigh muscle attenuation (HU) were collected. Measurements are reported as the average of the left and right thighs (or the single thigh if only one leg was in the field of view).

Statistics

Baseline characteristics were presented overall and by group using means and standard deviations (±SD) for continuous variables and counts and percentages (%) for discrete variables. Lean mass treatment effect outcome measures include six- (n=21; primary timepoint) and 12-month (n=14; secondary timepoint) change in DXA-acquired total body lean mass and appendicular lean mass, as well as CT-derived trunk and mid-thigh cross-sectional area and muscle attenuation (n=20; primary timepoint; n=12; secondary timepoint. All analyses were conducted using an intent to treat approach; participants were analyzed according to the group to which they were randomized regardless of adherence. Group treatment effects were estimated using linear mixed effects models fit with indicators for treatment group assignment and visit code, adjusted for baseline values of the outcome as a covariate. Visit-specific treatment effects were estimated from this model using contrast statements at 6 and 12 months. Sensitivity analyses further adjusted for total fat mass change.

Results:

Baseline Participant Characteristics (Full Randomized Sample)

Baseline participant characteristics and body composition variables are provided for all randomized participants, overall and by treatment group [risedronate, n=11, placebo, n=13] in Table 1. Three individuals in the risedronate group dropped out of the trial within the first 6 months, leaving 21 total participants with complete baseline and 6-month data. A diagram detailing participant flow through the study, with details about attrition has been published.^{17, 25} Participants were 56±7 years (mean±SD), and predominantly Caucasian (79%), premenopausal (63%) women (83%). Participants had class III obesity on average (Body mass index [BMI] 44.7±6.3 kg/m²) with the risedronate group having a significantly higher BMI compared to the placebo group (48.1±7.2 kg/m² vs 41.9±3.8 kg/m²). Baseline descriptive variables were similar for each group, with no significant differences noted. Adverse events experienced in the intervention and placebo group are detailed in the study design and feasibility report.¹⁷ In short, five adverse events were reported within six months of intervention, one of which was nausea related to the intervention, and four of which were not related to the intervention.

Table 1.

Baseline characteristics of study sample, overall and by treatment group.

	All		Risedronate		Placebo
Baseline Variable	N	Mean ± SD or N (%)	N	Mean ± SD or N (%)	N	Mean ± SD or N (%)
Age (years)	24	55.7 ± 6.7	11	53.8 ± 7.7	13	57.3 ± 5.7
Female Gender	24	20 (83.3)	11	9 (81.8)	13	11 (84.6)
Postmenopausal status	24	15 (62.5)	11	6 (54.5)	13	9 (69.2)
Black	24	5 (20.8)	11	3 (27.3)	13	2 (15.4)
Education, high school or less	24	4 (16.7)	11	3 (27.3)	13	1 (7.7)
Education, some college	24	12 (50.0)	11	4 (36.4)	13	8 (61.5)
Education, college+	24	8 (33.3)	11	4 (36.4)	13	4 (30.8)
Weight (kg)	24	122.1 ± 22.6	11	132.9 ± 25.3	13	113.0 ± 15.7
Excess Weight (kg)	24	54.0 ± 19.0	11	63.9 ± 21.5	13	45.7 ± 11.8
BMI (kg/m2)	24	44.7 ± 6.3	11	48.1 ± 7.2	13	41.9 ± 3.8
Clinical Bone Categorization, Normal	24	21 (87.5)	11	8 (72.7)	13	13 (100.0)
Clinical Bone Categorization, Osteopenic	24	3 (12.5)	11	3 (27.3)	13	0 (0.0)
Body mass (kg)	24	122.1 ± 22.6	11	132.9 ± 25.3	13	113.0 ± 15.7
Body Mass Index (kg/m2)	24	44.7 ± 6.3	11	48.1 ± 7.2	13	41.9 ± 3.8
DXA-Acquired Measures
Total Fat Mass (kg)	24	61.4 ± 13.1	11	68.6 ± 14.2	13	55.3 ± 8.4
Percent Fat Mass (%)	24	50.3 ± 5.0	11	51.7 ± 5.8	13	49.0 ± 4.0
Total Lean Mass (kg)	24	56.4 ± 11.5	11	59.7 ± 14.1	13	53.7 ± 8.2
Appendicular Lean Mass (kg)	24	25.7 ± 5.6	11	26.4 ± 6.9	13	25.2 ± 4.5
Appendicular Lean Mass Index (kg/m2)	24	9.4 ± 1.4	11	9.4 ± 1.7	13	9.3 ± 1.1
Lean-to-Fat Ratio	24	0.9 ± 0.2	11	0.9 ± 0.2	13	1.0 ± 0.1
CT-Acquired Measures
Total Trunk Muscle Cross Sectional Area (cm2)	21	139.1 ± 23.2	9	144.0 ± 25.1	12	135.5 ± 22.1
Skeletal Muscle Index (cm2/m2)	21	54.4 ± 9.1	9	56.3 ± 9.8	12	52.9 ± 8.6
Total Trunk Muscle Attenuation (HU)	21	40.3 ± 6.4	9	43.1 ± 6.9	12	38.2 ± 5.4
Mid-Thigh Muscle Cross Sectional Area (cm2)	20	111.6 ± 25.1	9	110.7 ± 33.8	11	112.3 ± 16.6
Mid-Thigh Muscle Attenuation (HU)	20	37.9 ± 4.7	9	37.4 ± 6.4	11	38.4 ± 3.2
HU: Hounsfield Units

Six-Month Treatment Effects on DXA-and CT-Acquired Body Composition

Raw aggregate means (SD) and model-adjusted changes in DXA- and CT-derived body composition measures over the initial six months are presented in Table 2. All participants lost a significant amount of body mass in the six months following bariatric surgery, with marginally greater losses experienced in the placebo [−20.9 kg (−23.7, −18.1)] versus risedronate [−16.3 kg (−20.0, −12.5)] group; p=.057. Similarly, the placebo group lost more total fat mass [−17.3 kg (−20.1, −14.6)] than the risedronate group [−12.4 kg (−16.2, −8.6)]; p=.048; however, total body lean mass losses were similar [risedronate: −3.2 kg (−4.5, −1.9); placebo: −2.9 kg (−3.9, −1.9)]. Appendicular lean mass loss was halved in the risedronate group [−1.2 kg (−2.3, −0.1)] versus placebo [−2.1 kg (−3.0, −1.2)], though this finding was not statistically significant (p=.20). Changes in total body fat and lean masses resulted in a more favorable lean-to-fat ratio for both groups at the six-month time point [placebo: 1.3 (1.2, 1.4) vs risedronate: 1.2 (1.0, 1.3); p=.054], although this finding was modest. CT scans at the mid-thigh echoed the potential appendicular lean mass relationship observed with DXA, with a signal suggesting less thigh skeletal muscle CSA loss in the risedronate group [−5.5 cm² (−11.0, −0.1)] compared to the placebo group [−12.4 cm² (−17.1, −7.7)]; p=.067 (Figure 1). Trunk CSA, skeletal muscle index, and muscle attenuation each followed a similar trend as the mid-thigh CT, with no significant differences observed between groups.

Table 2.

Six Month Treatment Effects on Body Composition Metrics.

Outcome Measures	Risedronate			Placebo
	n	Six Month	Six Month Δ	n	Six Month	Six Month Δ	p
Body mass (kg)	8	99.3 (95.6, 103.1)	−16.3 (−20.0, −12.5)	13	94.7 (91.9, 97.5)	−20.9 (−23.7, −18.1)	.057
BMI (kg/m2)	8	37.2 (35.8, 38.7)	−6.1 (−7.5, −4.6)	13	35.7 (34.7, 36.7)	−7.6 (−8.6, −6.6)	.095
DXA-Acquired Measures
Total Fat Mass (kg)	8	45.6 (41.8, 49.4)	−12.4 (−16.2, −8.6)	13	40.6 (37.9, 43.4)	−17.3 (−20.1, −14.6)	.048
Percent Fat Mass (%)	8	46.2 (43.1, 49.2)	−3.9 (−7.0, −0.8)	13	42.1 (39.8, 44.5)	−7.9 (−10.3, −5.6)	.045
Total Lean Mass (kg)	8	50.4 (49.1, 51.8)	−3.2 (−4.5, −1.9)	13	50.7 (49.7, 51.7)	−2.9 (−3.9, −1.9)	.745
Appendicular Lean Mass (kg)	8	23.1 (22.0, 24.2)	−1.2 (−2.3, −0.1)	13	22.2 (21.4, 23.1)	−2.1 (−3.0, −1.2)	.202
Appendicular Lean Mass Index (kg/m2)	8	8.6 (8.2, 9.0)	−0.4 (−0.8, −0.0)	13	8.3 (8.0, 8.6)	−0.8 (−1.1, −0.4)	.212
Lean-to-Fat Ratio	8	1.2 (1.0, 1.3)	0.2 (0.1, 0.3)	13	1.3 (1.2, 1.4)	0.4 (0.3, 0.5)	.054
CT Acquired Measures
Total Trunk Muscle CSA (cm2)	8	126.9 (121.1, 132.6)	−10.2 (−15.9, −4.4)	12	120.6 (115.7, 125.4)	−16.5 (−21.3, −11.6)	.098
Skeletal Muscle Index (cm2/m2)	8	49.6 (47.3, 51.8)	−4.0 (−6.2, −1.7)	12	47.1 (45.2, 49.0)	−6.4 (−8.3, −4.5)	.098
Total Trunk Muscle Attenuation (HU)	8	35.5 (33.2, 37.9)	−3.6 (−5.9, −1.3)	12	36.2 (34.4, 38.1)	−2.9 (−4.7, −1.0)	.633
Mid-Thigh Muscle CSA (cm2)	8	99.4 (93.9, 104.9)	−5.5 (−11.0, 0.1)	12	92.5 (87.8, 97.2)	−12.4 (−17.1, −7.7)	.067
Mid-Thigh Muscle Attenuation (HU)	8	37.5 (35.0, 39.9)	−0.3 (−2.8, 2.2)	12	38.2 (36.1, 40.3)	0.4 (−1.6, 2.5)	.638

Group treatment effect and difference estimates from a mixed model using contrast statements at 6 months adjusted for baseline values of each outcome

BMI: Body Mass Index; CSA: Cross-Sectional Area; HU: Hounsfield Units

Figure 1.

Six and 12 Month Treatment Effects on Select Lean Mass Variables

12-Month Treatment Effects on DXA and CT-Acquired Body Composition

Secondary intervention effects on the subset of participants with DXA and CT scans at 12-month follow-up are presented in Table 3. Notably, change in total body mass plateaued across both groups from six to 12-months, with a similar degree of total mass loss observed from baseline to 12-months (risedronate: 11.4% versus placebo: 16.4%). Greater fat mass loss in the placebo versus risedronate group persisted at 12 months, although results were no longer significant. The magnitude of total body lean mass loss was halved in the risedronate [−1.9 kg (−3.6, −0.1)] versus placebo [−3.5 kg (−4.8, −2.1)] group (p=.141), with striking findings observed at the appendicular skeleton (Figure 1). Specifically, the risedronate group maintained their baseline appendicular lean mass [−0.5 kg (−1.5, 0.6)], while the placebo group experienced significant loss [−2.9 kg (−3.6, −2.1)]; p=.001. Correspondingly, this difference was echoed in the appendicular lean mass index (ALMI) with risedronate maintaining ALMI [−0.1 kg/m² (−0.5, 0.3)] compared to the placebo group [−1.0 kg/m² (−1.3, −0.7)]. No significant differences in CT-acquired body composition metrics of the trunk or the mid-thigh were observed between groups at 12-months; however, trends revealed less lean mass loss at the mid-thigh in the risedronate group (−1.9%) compared to the placebo group (−9.1%); p=.095 (Figure 1). Likewise, the risedronate group experienced a loss of 5.7% of trunk CSA, compared to the placebo group, who experienced nearly double the loss (−10.6%); p=.111.

Table 3:

12 Month Treatment Effects on Body Composition Metrics.

Outcome Measures	Risedronate			Placebo
	n	12-Month	12-Month Δ	n	12-Month	12-Month Δ	p
Body mass (kg)	5	101.7 (95.4, 108.0)	−13.9 (−20.2, −7.6)	9	95.6 (90.9, 100.3)	−20.0 (−24.6, −15.3)	.121
BMI (kg/m2)	5	38.4 (35.9, 40.8)	−4.9 (−7.4, −2.5)	9	36.1 (34.3, 37.9)	−7.2 (−9.0, −5.4)	.130
DXA-Acquired Measures
Total Fat Mass (kg)	5	47.0 (41.4, 52.6)	−11.0 (−16.6, −5.4)	9	42.1 (38.0, 46.2)	−15.9 (−20.0, −11.7)	.163
Percent Fat Mass (%)	5	46.7 (42.7, 50.7)	−3.4 (−7.4, 0.6)	9	43.2 (40.1, 46.2)	−6.9 (−10.0, −3.9)	.162
Total Lean Mass (kg)	5	51.8 (50.0, 53.5)	−1.9 (−3.6, −0.1)	9	50.2 (48.8, 51.5)	−3.5 (−4.8, −2.1)	.141
Appendicular Lean Mass (kg)	5	23.9 (22.8, 24.9)	−0.5 (−1.5, 0.6)	9	21.5 (20.7, 22.3)	−2.9 (−3.6, −2.1)	.001
Appendicular Lean Mass Index (kg/m2)	5	9.0 (8.6, 9.4)	−0.1 (−0.5, 0.3)	9	8.0 (7.7, 8.3)	−1.0 (−1.3, −0.7)	.001
Lean-to-Fat Ratio	5	1.2 (1.0, 1.4)	0.2 (0.0, 0.4)	9	1.3 (1.1, 1.4)	0.3 (0.2, 0.5)	.342
CT Acquired Measures
Total Trunk Muscle CSA (cm2)	4	129.2 (122.4, 135.9)	−7.9 (−14.6, −1.1)	8	122.2 (116.9, 127.6)	−14.8 (−20.1, −9.5)	.111
Skeletal Muscle Index (cm2/m2)	4	50.5 (47.8, 53.1)	−3.1 (−5.7, −0.4)	8	47.7 (45.7, 49.8)	−5.8 (−7.9, −3.7)	.111
Total Trunk Muscle Attenuation (HU)	4	34.9 (31.9, 37.9)	−4.2 (−7.2, −1.2)	8	37.0 (34.8, 39.3)	−2.1 (−4.4, 0.1)	.259
Mid-Thigh Muscle CSA (cm2)	4	102.8 (95.8, 109.8)	−2.1 (−9.0, 4.9)	8	95.0 (89.2, 100.9)	−9.8 (−15.6, −4.0)	.095
Mid-Thigh Muscle Attenuation (HU)	4	36.4 (34.2, 38.7)	−1.3 (−3.6, 1.0)	8	38.4 (36.5, 40.3)	0.6 (−1.3, 2.5)	.183

Group treatment effect and difference estimates from a mixed model using contrast statements at 12 months adjusted for baseline values of each outcome

BMI: Body Mass Index; CSA: Cross-Sectional Area; HU: Hounsfield Units

Discussion:

The main objective of this secondary analysis of the WE RISE pilot RCT was to evaluate the impact of six months of risedronate use (versus placebo) on several indices of lean mass following SG. Preliminary treatment effect estimates suggest a signal for lean mass preservation following six months of risedronate use, particularly at the thigh, with robust findings across both DXA and CT modalities. While the sample size was limited, appendicular lean mass sparing was also maintained throughout the 12-month postoperative period. This signal is intriguing and worth studying in the context of a definitively powered trial, with far-reaching clinical applications, not only for persons experiencing rapid weight-loss, but also in other lean mass wasting conditions.

Although SG offers many clinical benefits for the treatment of severe obesity, rapid loss of lean mass (and associated reductions in functional capacity and metabolic rate) is an increasingly recognized surgical risk.^{4–6, 26} Indeed, limited observational data demonstrate that persons undergoing SG can lose over 10% of their lean mass in the six months following surgery,^{4, 27, 28} with considerable lean mass losses maintained five years postoperatively.^{5, 29} In agreement, our sample lost 5.8–6.1% of their total lean mass within the first six months, and 3.5–6.5% by 12-months, postoperatively. Compared to aging skeletal mass, which decreases by 3–8% per decade,³⁰ the lean mass losses experienced by these patients signal a significant acceleration in lean mass loss following SG, comparable to 10-years of progression via aging.

Encouragingly, risedronate significantly mitigated loss of appendicular lean mass by the 12-month timepoint, with lean mass sparing potential noted at the six-month timepoint. These findings support data from murine models of clinical pathology, where bisphosphonates mitigate lean mass loss associated with a variety of wasting conditions, including cancer cachexia, immobility, and muscular dystrophy.^11–14 Although, there is observational research describing this premise in humans,^{15, 31} ours is the first study to describe the potential use of bisphosphonates for lean mass preservation in the context of surgical weight loss. A potential mechanism for the lean mass sparing effect of bisphosphonates is derived from the bone-muscle crosstalk literature.³² Bisphosphonates work as inhibitors of calcification and bone resorption and our work may indicate that pharmacologic preservation of bone mass may benefit lean mass via bone-muscle crosstalk. If this is the case, future fully powered studies could significantly impact the field of sarcopenia research, positively influencing aging populations, and other populations experiencing combined bone and muscle loss.

This pilot RCT represents a novel analysis of the impacts of oral risedronate following SG on body composition measures. Our analysis is strengthened by the use of DXA and CT for body composition quantification, while a majority of body composition literature following SG relies on bioimpedance analysis.^{9, 29, 33} One author (LEF) performed all CT segmentations, thus eliminating inter-user variability. While we had a high retention rate at the six-month timepoint (n=21/24), we were limited by the small sample size and differential group drop out, particularly at the 12-month exploratory timepoint.

The use of DXA presents limitations given the size of our subject population, as DXA precision declines with increasing BMI.³⁴ However, standardized scanning procedures and established mirroring protocols were used when acquiring DXA outcomes to reduce error. Although bioimaging measures are sophisticated for quantification of lean mass, our findings should not be applied to all measures of lean mass. An alternative would be to directly measure total body lean mass using D₃-Creatine dilution, a method that can be employed easily and quantify muscle changes over time.^{35, 36}

In conclusion, these results add to an emerging body of literature suggesting lean mass preservation with bisphosphonate use. In addition to differences in volumetric changes observed between the risedronate and placebo groups, our results demonstrate a preference for lean mass sparing at the appendicular lean mass sites. Others have proposed that osteokines, which are released during periods of high bone-turnover, may impact lean mass by influencing myogenesis.¹⁴ Future studies should investigate these biologic signaling and inflammatory factors to further elucidate potential mechanisms for lean mass maintenance.

As our results stem from a secondary analysis of a pilot study, they should not be considered definitive. Given the well-described biological effects of bisphosphonates as inhibitors of bone resorption, it is possible that pharmacological preservation of bone mass may benefit muscle mass via bone factors targeting muscle.³⁷ Thus, we implore future clinical studies to definitively test whether bisphosphonate use can offset surgical weight loss associated lean mass loss, along with basic studies to elucidate underlying mechanisms of action. As SG provides considerable health benefit as a treatment option for the millions of Americans living with severe obesity, identification of therapies to mitigate associated musculoskeletal harm has the potential for significant public health impact while also exerting influence on the emerging field of bone-muscle crosstalk.