GLP-1 receptor agonists and obesity drugs: effect on bone

Evidence summary for clinicians. For discussion, not a prescription. Grades are a GRADE-style reading of the cited sources unless a guideline grade is quoted.

LoE (level of evidence) in the study table: I = large RCT or meta-analysis of trials; II = smaller RCT or prospective cohort; III = retrospective cohort or post hoc analysis. Most trials measure BMD or bone turnover markers (P1NP for formation, CTX for resorption), not fractures. Fracture data come mainly from people with type 2 diabetes.

What changes practice

• Pair every GLP-1 prescription with exercise. In a 52-week trial, liraglutide plus exercise kept hip and spine BMD at placebo levels while giving the largest weight loss; liraglutide alone lost more BMD than exercise alone¹.
• Prescribe resistance training (at least 2 days a week) plus aerobic exercise, and adequate protein, for everyone losing weight on these drugs^{2, 3, 4}.
• In older adults and postmenopausal women with fracture risk, expect small falls in hip and spine BMD on semaglutide⁵. Assess fracture risk before starting and treat osteoporosis on standard criteria.
• Do not tell patients these drugs protect bone. The protective signal in type 2 diabetes comes from older trial data with few fractures, and cohort data conflict^{6, 7, 8}.
• Bariatric surgery carries a clear rise in fracture risk, more after malabsorptive procedures; GLP-1 drugs have not shown this^{9, 10}.

Evidence grade: Low. Moderate for BMD effects in randomised trials; low for fracture effects; very low for tirzepatide and newer drugs.

What the guidance says (newest first)

No guideline dedicated to obesity drugs and bone was found. The statements below cover bone and muscle as part of wider obesity or nutrition advice.

OUTSIDE INDIA All statements come from the US, Europe or international panels. Not Indian data; use as a guide.

How these drugs could affect bone

• GLP-1 receptor. Mice without the receptor have cortical osteopenia, bone fragility and more bone resorption. The effect ran through thyroid calcitonin, which exendin-4 raised in mice¹⁴. This is animal data; a calcitonin route has not been shown in humans.
• GIP. Short GIP infusions in humans cut the resorption marker CTX by up to about 50 to 59%, independently of insulin and glucose, with no change in P1NP^{15, 16}. This is the rationale for hoping that GIP-containing drugs such as tirzepatide spare bone; no bone outcome trial confirms it.
• Weight loss itself. Less mechanical loading, lower intake of energy, protein and calcium, and hormonal shifts cause high-turnover bone loss after diet or bariatric surgery¹⁰.
• Lab-model benefits of GLP-1 drugs on bone are seen at drug levels much higher than those in patients¹⁰.

Evidence grade: Very low. Mechanistic and animal data; human relevance unproven.

Bone density and turnover in trials

• Liraglutide 3.0 mg after diet-induced weight loss (Denmark, n=195, 52 weeks, no diabetes). The liraglutide-plus-exercise group lost the most weight (16.9 kg) with no BMD loss against placebo. Liraglutide alone lost more BMD than exercise alone at the hip (-0.013 g/cm², P=.03) and spine (-0.016 g/cm², P=.04)¹.
• Liraglutide 1.2 mg after a 12% diet weight loss (Denmark, n=37 women, 52 weeks). Bone mineral content loss was about four times smaller than with diet alone; P1NP rose 16% and CTX did not change¹⁷.
• Semaglutide 1.0 mg in people at increased fracture risk (Denmark, n=64, 55 postmenopausal women, 52 weeks). Weight fell 6.8 kg; spine BMD fell 0.018 and total hip 0.020 g/cm² against placebo; femoral neck was unchanged. CTX rose; P1NP did not change⁵.
• Meta-analyses. In 25 trials in type 2 diabetes (2025), BMD at spine, femoral neck and hip was higher on GLP-1 drugs, with P1NP up and CTX down¹⁸. A larger 2026 meta-analysis of 60 studies (46 RCTs) found no effect on BMD in the most-adjusted models¹⁹.

OUTSIDE INDIA The BMD trials are Danish; the meta-analyses pool mainly Western and East Asian trials. Not Indian data; use as a guide.

Evidence grade: Moderate. Consistent small BMD loss with large weight loss on a GLP-1 drug alone; trial sizes small (37 to 195).

Lean mass: what is lost with the weight

• Tirzepatide (SURMOUNT-1 DXA substudy, n=160, 72 weeks): weight fell 21.3%, fat mass 33.9% and lean mass 10.9%. About three-quarters of the weight lost was fat and one quarter lean mass²⁰.
• Across 60 studies, GLP-1 drugs reduced lean mass (SMD 0.52), mainly with liraglutide and semaglutide. Heterogeneity was very high (I² 88%) and certainty low¹⁹.
• Semaglutide 2.4 mg (STEP 1) gave 14.9% weight loss at 68 weeks against 2.4% on placebo; the scale of loss is the main driver of lean and bone loss^{21, 2}.
• A 2026 review of 16 RCTs found lean mass loss in proportion to weight loss, and no trial that directly assessed bone health²².

Evidence grade: Moderate. Lean mass loss is consistent; its effect on falls and fractures is not measured.

Fracture data: trials, meta-analyses and cohorts

Randomised data lean towards neutral or slightly protective in type 2 diabetes, but fractures were rare secondary outcomes. Cohort studies conflict and are open to confounding. The signals of harm in older adults, in people without diabetes and with tirzepatide are hypotheses, not proof.

OUTSIDE INDIA The cohorts are from Israel and multinational US-led health record networks. Not Indian data; use as a guide.

Evidence grade: Low. Few events in trials; conflicting observational studies.

Other obesity drugs

• Orlistat. One small 1-year RCT (n=30) found no BMD difference from placebo, with a pattern of early negative vitamin D balance²⁵. Bone changes with orlistat were no greater than with control in a review²⁶.
• Phentermine-topiramate. Bone effect assumed negative but not well studied. Naltrexone-bupropion and setmelanotide: bone effects largely unknown²⁷.
• Dual and triple agonists and amylin analogues (the class of retatrutide and cagrilintide): only preclinical bone data, possibly positive²⁷. No human bone outcome data were found for retatrutide or cagrilintide.
• Tirzepatide. No published trial with BMD or fracture endpoints was found; the lean mass data above and one cohort signal are all that exist^{20, 24}.

Evidence grade: Very low. Little or no human bone data for any of these drugs.

Compared with diet and bariatric surgery

• Hip BMD falls by about 1 to 1.5% with diet-induced weight loss and by 8 to 11% after gastric bypass²⁶.
• Bariatric surgery raises fracture risk: RR 1.20 (1.15 to 1.26) in observational and 1.16 (1.00 to 1.33) in interventional studies, higher after malabsorptive procedures⁹. Another meta-analysis found RR 1.29, and RR 1.68 for upper limb fractures²⁸.
• Calorie restriction and surgery cause high-turnover bone loss; GLP-1 drug effects on bone so far look smaller than those of surgery^{10, 27}.

Evidence grade: Moderate. Consistent for surgery; diet-related BMD loss is small.

Older adults and postmenopausal women

• The only trial in people at increased fracture risk (mostly postmenopausal women) showed hip and spine BMD loss and higher CTX on semaglutide over one year⁵.
• In adults aged 65 or more with type 2 diabetes, one large cohort found an 11% higher fragility fracture rate on GLP-1 drugs⁷.
• In dieting obese older adults (n=160, 6 months, about 9% weight loss), hip BMD fell 1% with combined and 0.5% with resistance training, against 3% with aerobic training alone; lean mass was also better kept²⁹.
• A review of obesity drugs in older adults advises resistance training to protect muscle and bone³⁰.

Evidence grade: Low. One small trial and one cohort; no fracture trial in this group.

Protecting bone during treatment

• Exercise. WHO advice of 150 minutes of moderate aerobic activity weekly plus resistance training on at least 2 days is especially important during weight loss⁴. Combined training kept BMD in trials of liraglutide and of diet in older adults^{1, 29}.
• Protein and micronutrients. Adequate protein, micronutrients and resistance training for all patients on incretin drugs^{2, 3}. Exact protein targets rest on indirect evidence¹¹.
• Vitamin D and calcium. No routine vitamin D test is needed for obesity alone; empiric vitamin D is suggested over 75 years¹³.
• Monitoring. Baseline body composition and muscle function assessment is advised³. Interpretation (not a sourced finding): a baseline DXA is reasonable in postmenopausal women and older adults with fracture risk factors starting a GLP-1 drug.
• Antiresorptive. No trial has tested adding an antiresorptive to a GLP-1 drug. Interpretation (not a sourced finding): start one on standard osteoporosis criteria, not because of the obesity drug alone. See fracture risk.

Evidence grade: Low. Exercise data are randomised; nutrition and monitoring advice is expert consensus.

Indian context

In ICMR-INDIAB-17 (n=113,043 adults across all states), 11.4% had diabetes, 15.3% prediabetes, 28.6% generalised obesity and 39.5% abdominal obesity³¹.

NO INDIAN DATA No Indian study of GLP-1 drugs, tirzepatide or other obesity drugs with BMD, bone marker, lean mass or fracture outcomes was found.

OUTSIDE INDIA All bone and lean mass data above come from Europe, the US, Israel and multinational trials. Not Indian data; use as a guide.

Key studies, graded

Where the evidence is weak

• No trial has a fracture endpoint for any obesity drug; trial fracture data are secondary outcomes in type 2 diabetes^{6, 18}.
• BMD trials are small (37 to 195 people) and last one year or less^{17, 5, 1}.
• Tirzepatide, retatrutide, cagrilintide, setmelanotide, naltrexone-bupropion: no human bone outcome trials; the tirzepatide fracture signal is from one 14-month cohort with a composite outcome^{24, 27}.
• Cohort findings conflict (protective in type 2 diabetes, harmful in older adults and in people without diabetes)^{8, 7}.
• Nutrition and monitoring advice is consensus built on indirect evidence¹¹.
• Industry links: the semaglutide trial used drug supplied by the manufacturer⁵.
• Corrections: published corrections exist for the 2025 joint advisory, the SURMOUNT-1 DXA substudy and the Endocrine Society vitamin D guideline; figures here are from the original abstracts^{3, 20, 13}.
• India: no Indian bone data for any obesity drug.

Related evidence pages

• Fracture risk around menopause and how to reduce it
• Bone health in perimenopause: what the evidence shows
• How osteoporosis drugs work: bone builders and bone savers