Drug facts

Tirzepatide and bone density

By David Chen·Editor·Reviewed September 2026

What the labels say¹

An absence is only worth reporting with the list of what was read. We searched the full text of the Zepbound prescribing information and Medication Guide, both revision 8/2026, for every term a bone finding would use:

“Fracture”, “osteoporosis”, “osteopenia”

Not present.

“Bone mineral density”, “DXA”, “absorptiometry”, “calcium”

Not present.

“Bone”

Twice, and neither is about your skeleton: once in “peptide backbone”, describing how the drug is broken down, and once in a rat bone marrow test showing the drug does not damage genes.

The Medication Guide

No mention of bone or fractures.

The one sentence the label does give on body composition, in its pharmacology section, is that tirzepatide lowers body weight “with greater fat mass loss than lean mass loss.” It says nothing about bone.

The Mounjaro label, for the same molecule in type 2 diabetes, is the same.² Revision 8/2026 contains no fracture, osteoporosis or bone density language, and its Medication Guide has none either. Mounjaro and Zepbound compared →

No phase 3 trial measured bone density⁴

The clearest statement comes from a 2026 analysis of the obesity trials written largely by the manufacturer’s employees. Of their own finding that fractures were balanced, they write:

“Interpretation of no imbalance in fractures between tirzepatide and placebo in this analysis is limited by the absence of bone mineral density assessments across the Phase 3 trials and a follow-up duration that may have been insufficient to fully characterize skeletal outcomes.”

The one trial scan people point to does not fill the gap.³ A substudy of SURMOUNT-1 put 160 participants through DXA scans at the start and at week 72. It reported fat mass and lean mass, and it defines lean mass as fat-free mass minus bone mineral content. Bone was subtracted out of the measure; the paper reports no bone density figure. What that substudy found about muscle →

Fractures and falls in the trials⁴

What the trials did record is fractures and falls as adverse events. The same analysis pooled seven phase 3 trials — SURMOUNT-1 through -5, SURMOUNT-OSA and SUMMIT — and split them by age. Every figure below is tirzepatide against placebo. One detail matters: SURMOUNT-5 compared tirzepatide with semaglutide, so its tirzepatide group is in the tirzepatide column, and its semaglutide group is in neither column.

The authors call the differences not clinically meaningful and put the higher rates in older people down to age rather than the drug. Read the counts as well as the percentages: these are single-digit numbers of events in trials that ran one to three years, some rows lean one way and some the other, and a trial of this size cannot detect a modest change in fracture risk. It is reassurance of a limited kind, and the quotation above is the authors’ own statement of the limit.

Two large studies, pointing opposite ways⁵

Both studies below drew on the same database of U.S. health records, TriNetX. Both are observational: they compare people whose doctors chose different drugs, so they can show an association, not a cause.

Against other GLP-1 drugs, more osteoporosis or fracture. A Taiwanese team matched 66,329 people starting tirzepatide with the same number starting another GLP-1 receptor agonist, all with type 2 diabetes or obesity, and followed them for 14 months. The main outcome counted either a new osteoporosis diagnosis or a fragility fracture. Tirzepatide was associated with a higher risk of it (hazard ratio 1.44, 95% CI 1.22 to 1.69) and of starting osteoporosis treatment (1.61, 1.22 to 2.12). The abstract does not say which GLP-1 drugs made up the comparison group, and because a diagnosis counts the same as a fracture, the result is not a fracture rate.

Against DPP-4 inhibitors, fewer femoral fractures.⁶ A second study took adults 65 and older with type 2 diabetes and a BMI of 25 or more, and compared tirzepatide with DPP-4 inhibitors, an older class of diabetes tablets. In 12,808 matched patients, femoral fractures over one year were 0.2% on tirzepatide against 0.4% (hazard ratio 0.452, 95% CI 0.280 to 0.729), and falls were 3.6% against 5.7% (0.664, 0.591 to 0.747). The same paper ran a separate comparison for semaglutide; only the tirzepatide figures are quoted here.

These are not contradictory so much as answers to different questions. One compares tirzepatide with drugs of its own kind, in a younger and broader group, and counts diagnoses; the other compares it with a class of tablets, in older people with diabetes, and counts broken femurs. What neither can do is say whether tirzepatide itself strengthens or weakens bone.

Where bone has been scanned⁷

One study reports bone on tirzepatide alone, and it is small. A Japanese cohort of 112 adults without diabetes, not randomized, compared 2.5 mg with 5 mg over six months and reported “similar reductions in BMI, skeletal muscle mass, and bone mass” at the two doses. The abstract gives no figure and no measuring method, and there was no untreated group to compare with. Keeping people at 2.5 mg is also off-label: the Zepbound label says that dose “is for treatment initiation and is not approved as a maintenance dosage.”⁸

Two studies measured bone density directly with DXA scans. Both pooled tirzepatide with other drugs, so neither is a tirzepatide result on its own:

• Semaglutide and tirzepatide, pooled.⁹ A single-center study in New York matched 255 people using either drug, 92% of them women with a mean age of 64, with 255 non-users. Over a median of 17 months both groups lost hip bone density to a similar degree. In people without diabetes, those on the drugs lost more at the hip (−1% against −0.6%), and more weight lost went with more bone lost.
• Liraglutide, semaglutide and tirzepatide, pooled.¹⁰ A study in Kuwait followed 70 people with type 1 diabetes and obesity for 12 months. Weight fell 6.3% and total bone density was unchanged.

In mice¹¹

Two mouse studies of tirzepatide disagree. In obese, diabetic mice, one reported that tirzepatide reduced bone mass and linked the effect to changes in gut bacteria. The other gave diabetic mice four weeks of tirzepatide or semaglutide and found ¹²a neutral effect on bone mass for both. Mouse results are a reason to study people, not a finding about people.

What none of this covers

This page reports what has and has not been measured. Whether your own bone health needs checking, and how, is a question for the prescriber who knows your history. See the medical disclaimer, and the label’s full list of reported side effects.