Ozempic Bone Loss: What the Label and Trials Show (2026)

Does Ozempic Cause Bone Loss? The Short Answer

The evidence is thin and mixed. Our search of the Ozempic injection label on October 29, 2026 found no fracture or bone-density wording. In the one placebo-controlled semaglutide trial that scanned bone, hip and spine density ended lower than placebo as weight fell, a secondary result. None of the semaglutide trials here was sized to detect fractures.

Two brands matter here. Ozempic is the type 2 diabetes brand: an injection dosed from 0.25 to 2 mg weekly, and, since FDA approval on January 30, 2026, also a daily Ozempic tablet. Wegovy is the weight-reduction brand, with 2.4 mg the recommended weekly maintenance dose and 7.2 mg the maximum for weight reduction in adults. The studies below span both ranges, so each result carries its dose and population.

What the Ozempic and Wegovy Labels Say About Bone and Fractures

We searched the full text of each label on DailyMed on October 29, 2026, including its Medication Guide and Instructions for Use, for the terms fracture, osteoporosis, bone mineral, BMD, bone density, bone loss and falls. Each label is a bundle of documents with its own version, so the table gives the version we read.

The Wegovy Label's Fracture Paragraph (Sections 6.1 and 8.5)

Section 6.1, Clinical Trials Experience, lists adverse reactions seen in trials. Under the heading Fractures it reads:

"In the CV outcomes trial in adults, more fractures of the hip and pelvis were reported on WEGOVY than on placebo in female patients: 1% (24/2,448) vs. 0.2% (5/2,424), and in patients ages 75 years and older: 2.4% (17/703) vs. 0.6% (4/663), respectively. In a clinical trial in adults with MASH, fractures occurred in 4.4% of WEGOVY-treated patients (2.6 cases per 100 patient years) compared to 3.3% of placebo-treated patients (2 cases per 100 patient years). Fractures were reported in both males and females with a median age of 61 years (range, 44 to 75)."

Section 8.5, Geriatric Use, repeats the older-adult finding: "In the CV outcomes trial, patients aged 75 years and older reported more hip and pelvis fractures in the WEGOVY injection-treated patients than placebo-treated patients."

Each percentage belongs to its own group. The 1% versus 0.2% covers women in the cardiovascular outcomes trial described in the Wegovy label, adults with established cardiovascular disease and overweight or obesity. The 2.4% versus 0.6% covers people 75 and older in that same trial. The 4.4% versus 3.3% comes from a separate trial in adults with MASH, a fatty liver disease.

The passage is not a boxed warning and not a Section 5 warning. It states no mechanism and gives no bone density figure. It also belongs to Wegovy, at Wegovy's doses, in Wegovy's trial populations, so none of its percentages transfers to Ozempic.

What Our Search of the Ozempic Label Found

Our search of the full text of the Ozempic injection label on October 29, 2026 found no fracture, osteoporosis or bone-density wording. The only bone mentions were animal findings on fetal skeletal development and a rat bone-marrow genetic test. The same search of the combined Ozempic tablets and Rybelsus label gave the same result. For comparison, the same search of the Mounjaro label (version 40, effective August 27, 2026) and the Zepbound label (version 40, effective August 28, 2026) also found no fracture, osteoporosis or bone-density wording; their only bone-related mentions were the same kind of rat bone-marrow test and animal findings on fetal skeletal development.

Why a Label Sentence Is Not a Risk Estimate

Label adverse-reaction counts come from trials built to answer other questions: heart outcomes in one case, liver disease in the other. And a missing sentence doesn't show safety: a label records what was reported, not every effect a drug might have.

What Bone Density Scans Measure, and What DXA Lean Mass Is Not

The trials below use DXA, a low-dose X-ray scan. It reports two things this topic keeps mixing up. Bone mineral density, in grams per square centimetre, is read at sites such as the hip and lumbar spine. Body composition is reported separately, split into fat mass, bone mineral and lean soft tissue.

So "lean mass" on a DXA report means lean soft tissue only. It excludes bone mineral, and it is not the same as muscle. A trial reporting lean mass loss is not reporting bone loss, and the reverse holds too. The semaglutide muscle loss page covers the lean tissue side.

Three outcomes recur below, and they are not interchangeable. Bone turnover markers in blood show how fast bone is being formed or broken down. Bone density shows how much mineral sits at a site. Fractures are what actually harm people. A change in the first two does not by itself show a change in the third.

What the Ozempic Bone Loss Trials Measured

The Semaglutide Trial: Hansen 2024

The 2026 systematic review by Lai and colleagues found one randomized trial of semaglutide with bone density as an outcome in adults with overweight or obesity. That trial, by Hansen and colleagues (eClinicalMedicine, 2024), enrolled 64 adults aged 40 to 85 in Denmark at increased fracture risk: low bone density at the hip or spine, a low-energy fracture in the past 3 years, or both. Of the 64, 55 (86%) were postmenopausal women. People with type 1 or type 2 diabetes were excluded, and mean BMI was about 27 to 28.

Semaglutide injection was stepped up from 0.25 mg to 0.5 mg to 1.0 mg weekly, a dose within the Ozempic range and below the 2.4 mg Wegovy maintenance dose, and continued for 52 weeks. Everyone took at least 800 mg of calcium and 20 micrograms of vitamin D daily.

The primary outcome was P-PINP, a blood marker of bone formation. It showed no difference: estimated treatment difference 3.8 micrograms per litre (95% CI -5.6 to 13.3; p=0.418). The bone density results were secondary outcomes.

The authors suggested the rise in bone breakdown may be explained by the weight loss that came with treatment. The trial was sized to detect a change in the bone-formation marker, not fractures, and its secondary outcomes were not adjusted for multiple testing. Its funders were the Region of Southern Denmark, the Novo Nordisk Foundation and the Gangsted Foundation, and Novo Nordisk supplied the drug and placebo.

A Different Drug: Liraglutide Plus Exercise (Jensen 2024)

This trial used liraglutide 3.0 mg daily, not semaglutide. Jensen and colleagues (JAMA Network Open, 2024) put 195 adults with obesity and no diabetes on an 8-week, 800-calorie diet, then randomized them for 52 weeks to exercise, liraglutide, both, or placebo. With exercise added to liraglutide, hip and spine density were unchanged compared with placebo. Liraglutide alone ended lower than exercise alone at the hip (-0.013 g/cm², P=.03) and spine (-0.016 g/cm², P=.04).

What Reviews Conclude, and Why They Disagree

Lai and colleagues (Archives of Osteoporosis, published October 2026) reviewed 8 randomized trials, 493 people with overweight or obesity, without pooling them. Semaglutide had one trial, Hansen. In their words: "Certainty of evidence was very low for every agent." Their conclusion: "The randomized evidence is too sparse and too clinically heterogeneous to determine the skeletal effects of GLP-1 RA therapy in adults with overweight or obesity." They noted that the two trials with the largest weight loss, the semaglutide trial and the liraglutide 3.0 mg trial, both reported lower spine and hip density.

Beaudart and colleagues (Drugs, October 2026) pooled 60 studies covering 1,250,717 people: 46 randomized trials, 13 real-world studies and a pharmacovigilance study, across many GLP-1 drugs and populations, searched to March 2025. Using the most adjusted estimates, they reported "No effect on bone outcomes (i.e., bone mineral density [all sites] and fractures [all sites])". They rated the certainty of the bone density evidence as high for the lumbar spine and moderate for the total hip, femoral neck and whole body. A pooled result across drugs and populations does not isolate semaglutide at weight-loss doses.

Zhang and colleagues (Bone, 2025) pooled 44 randomized trials in 47,823 people with type 2 diabetes and found fewer fractures with GLP-1 drugs than with placebo or other diabetes drugs (RR 0.77, 95% CI 0.61 to 0.96), seen only beyond 78 weeks of treatment, from 292 fractures in total. Those trials were run in diabetes for other endpoints, so they say little about non-diabetic weight-loss use. At the small end, Dinkla and colleagues' 20-person pilot (semaglutide 1.0 mg, 20 weeks, mean age 72.7) found no significant differences, while noting consistently lower density in the semaglutide group. Twenty people can't settle it either way.

Video: endocrinologist Susan Brian, MD, explains why bone turnover markers, bone density and fractures are separate outcomes and why weight loss complicates the trials. It is one clinician's reading. Her conclusion is close to the Drugs meta-analysis and more reassuring than the Archives of Osteoporosis review, and she cites a 26-trial review we could not find in PubMed. Numbers spoken in the video are not used on this page.

Ozempic Bone Loss: The Drug, the Weight Loss, or Both?

Weight loss without any drug lowers hip density. Zibellini and colleagues (Journal of Bone and Mineral Research, 2015) pooled 41 published diet studies in adults with overweight or obesity. Total hip density fell 0.010 to 0.015 g/cm² at 6, 12 and 24 months, with no significant change at the lumbar spine or whole body except a small whole-body drop at 6 months. The authors judged the decrease small next to the metabolic benefits of losing excess weight.

The Hansen trial can't separate drug from weight loss. The placebo group didn't lose the same amount of weight, so the gap in bone density reflects both at once. The authors' own suggestion, that weight loss may explain the rise in breakdown, is an interpretation, not something the design could test.

A 2026 cohort points the same way without proving it. Among people on semaglutide or tirzepatide in Liu and colleagues' study, more weight lost went with more bone lost at the total hip (r=0.32) and femoral neck (r=0.17). A correlation in an observational group shows that the two moved together, not that one drove the other.

What Cohort Studies Found (Association Only)

Cohort studies follow people who chose, or were prescribed, a drug. They're larger than trials but can't rule out that users differ from non-users in ways that affect bone. Everything in this section is association, not cause.

• Huang and colleagues, 2026 (Diabetes, Obesity and Metabolism, TriNetX health records). In people with obesity and type 2 diabetes followed for 3 years, starting semaglutide was associated with fewer major osteoporotic fractures than empagliflozin (HR 0.69), glipizide (HR 0.72) or usual care (HR 0.84). In people with obesity but without diabetes, followed for 2 years, there was no significant association in any comparison.
• Hamad and colleagues, 2026 (JAMA Network Open, TriNetX). In adults aged 50 to 90 with type 2 diabetes, starting any GLP-1 drug rather than a DPP-4 inhibitor was associated with fewer fragility fractures (HR 0.79, 95% CI 0.76 to 0.83). In a separate sensitivity analysis that compared GLP-1 starters with people who had never taken a GLP-1 drug, split by diabetes status, the lower risk held with diabetes at 3 years (HR 0.91). Without diabetes there was no significant difference at 1 year (HR 1.02), and at 3 years GLP-1 starters had more fragility fractures (HR 1.13, 95% CI 1.04 to 1.23). This covers the drug class, not semaglutide alone.
• Liu and colleagues, 2026 (Journal of Clinical Endocrinology and Metabolism). At one medical center, 255 users of semaglutide or tirzepatide, pooled together, were matched with 255 non-users; 92% were women, mean age 64. Over a median 17 months, both groups lost hip and femoral neck density by similar amounts. Without diabetes, users lost more at the total hip per year (-1% vs -0.6%).
• Chen and colleagues, 2026 (Osteoporosis International, TriNetX). In adults 65 and older with type 2 diabetes and overweight or obesity, starting semaglutide rather than a DPP-4 inhibitor was associated with fewer femoral fractures over one year (0.3% vs 0.5%; HR 0.488, 95% CI 0.367 to 0.649).
• Wu and colleagues, 2026 (Obesity, TriNetX). In adults 50 and older with obesity and without diabetes, users of liraglutide, semaglutide or tirzepatide had fewer new osteoporosis diagnoses (RR 0.48, 95% CI 0.43 to 0.52) and fewer major osteoporotic fractures (RR 0.15, 95% CI 0.11 to 0.22) than users of other obesity medicines such as orlistat or phentermine. That is the opposite direction from Hamad's result without diabetes, which used a different comparison group. The authors note that GLP-1 users were followed for a shorter time (median 632 vs 1,309 days), that fractures were identified by body site rather than confirmed as low-trauma, and that residual confounding cannot be excluded.

The studies differ in diabetes status, comparator (another diabetes drug, another obesity drug, another GLP-1 drug, usual care or no GLP-1 drug) and length of follow-up. Huang, Hamad, Wu and Chen above, and Hsu (tirzepatide, below), all used the TriNetX network, so they are not independent confirmations of each other.

Tirzepatide and Bone: What Exists

Less, and it points both ways. Our PubMed search found no tirzepatide-specific randomized trial with bone density as an outcome, which is a search result, not proof that none exists. Hsu and colleagues (Diabetes Research and Clinical Practice, 2025) matched 66,329 people per group with type 2 diabetes or obesity in TriNetX and followed them for 14 months. Tirzepatide was associated with more new osteoporosis diagnoses or fragility fractures than other GLP-1 drugs (HR 1.44, 95% CI 1.22 to 1.69). That composite includes diagnosis codes, so differences in testing could play a part, and the study does not show cause.

Two other TriNetX cohorts went the other way. In adults 65 and older with type 2 diabetes, tirzepatide was associated with fewer femoral fractures over one year than DPP-4 inhibitors (HR 0.45, 95% CI 0.28 to 0.73; Chen and colleagues, 2026). In adults 50 and older with obesity and without diabetes, tirzepatide users had fewer new osteoporosis diagnoses than users of other obesity medicines (RR 0.34, 95% CI 0.27 to 0.43; Wu and colleagues, 2026). A post hoc analysis of Lilly's phase 3 tirzepatide trials, written mostly by Lilly employees, reported no clinically meaningful difference from placebo in falls or fractures recorded as adverse events in adults 65 and older (Alfaris and colleagues, 2026). None of these shows cause.

The same label search of Mounjaro and Zepbound found no fracture, osteoporosis or bone-density wording. For body composition on tirzepatide, see tirzepatide muscle loss; for its labelled side effects, see tirzepatide side effects.

What Medical Societies Say About Protecting Bone on a GLP-1

We do not give personal medical advice. What follows is what two named bodies state, in their own words, and this page does not recommend starting, stopping or changing any medicine.

The 2025 joint advisory from the American College of Lifestyle Medicine, the American Society for Nutrition, the Obesity Medicine Association and The Obesity Society is an expert-group appraisal, not a systematic review or graded guideline. It states:

• "GLP-1s should be prescribed together with a structured exercise program, aiming for regular strength training at least three times weekly plus at least 150 min of moderate-intensity aerobic exercise weekly to preserve muscle and bone mass."
• "Sufficient dietary protein should be a priority to help preserve muscle mass and bone density, particularly in combination with a structured strength training program."
• "Dietary supplements can be proactively considered for at-risk nutrients, such as vitamin D, calcium, B12, or a multivitamin-mineral tablet, at appropriate doses and tailored to each person's needs."

The advisory also states that substantial weight reduction (14% or more) that is rapid (over 3 to 4 months) is associated with significant bone loss, and that more moderate, slower reduction may better preserve bone mass.

For women around and after menopause, the Spanish Menopause Society's 2026 position statement says incretin-based therapy "should be considered within a cardiometabolic and functional framework that incorporates visceral adiposity, skeletal muscle health, fracture risk, cardiovascular prevention, individualized hormone therapy decision-making, resistance exercise, adequate protein intake, and structured monitoring of functional and metabolic outcomes."

One limit: the only randomized test of exercise for bone on a GLP-1 drug that we found is the Jensen trial, which used liraglutide, not semaglutide.

What We Do Not Know Yet

• No semaglutide trial covered here had fractures as an endpoint, and none studied people using it for weight loss with bone as the main question.
• None of the trials here compared semaglutide with an equal amount of weight loss by another route, so the drug and weight effects stay tangled.
• We found no trial at 2.4 mg or 7.2 mg with bone density as the primary outcome (a search result, not proof that none exists).
• Follow-up is short: one year in the main trial, and one to 3 years in the cohorts.
• Whether changes of 2 to 3% in density translate into more fractures is not shown by any study here.
• The evidence is moving. The newest review cited here was published October 23, 2026, and new cohort studies keep appearing. We last searched PubMed on October 29, 2026.