# Bone Density on a GLP-1: The Trial That Compared It Against Exercise

By Nora Bissett, Pricing Editor (https://glpwatchdog.org/authors/nora-bissett) | Published: 2026-08-21 | Updated: 2026-10-27

Canonical page: https://glpwatchdog.org/research/glp1-bone-density-and-exercise

> A four-arm randomized trial found liraglutide alone lost hip and spine bone density relative to exercise, while the combined arm had the greatest weight loss and kept bone density level with placebo.

## Key points

- Randomized 4-arm trial, 195 adults with obesity and no diabetes, 52 weeks after an 800 kcal/day diet.
- Liraglutide alone decreased BMD at hip (-0.013 g/cm2) and spine (-0.016) relative to exercise.
- Losing the most weight of any arm (16.88 kg), the combination group still held BMD statistically level with placebo.
- A meta-analysis of 25 RCTs in type 2 diabetes found the opposite: GLP-1 RAs superior on spine and hip BMD, with no significant increase in fracture risk.

One randomized trial put the question directly: after losing weight, does it matter *how* you keep it off, as far as your bones are concerned? Adults with obesity spent eight weeks on a very low calorie diet, then a year on exercise, liraglutide, both, or placebo. Liraglutide alone lost bone mineral density at the hip and spine relative to exercise. The **combination** held bone density level with placebo while producing the largest weight loss of the four groups.[1]

## The four-arm trial

One hundred and ninety-five adults aged 18 to 65 with a BMI between 32 and 43, none of them diabetic, completed an eight-week 800 kcal/day diet and were then randomized for 52 weeks to a moderate-to-vigorous exercise program, liraglutide 3.0 mg daily, both, or placebo. Bone mineral density at hip, lumbar spine and distal forearm was measured by DXA from before the diet to the end of treatment.[1]

*Total estimated mean weight loss across the study, by arm (n = 195).[1]*

| Arm | Weight loss | 95% CI |
| --- | --- | --- |
| Placebo | 7.03 kg | 4.25–9.80 |
| Exercise alone | 11.19 kg | 8.40–13.99 |
| Liraglutide alone | 13.74 kg | 11.04–16.44 |
| **Combination** | **16.88 kg** | 14.23–19.54 |

Now the bone results, which do not follow the weight results. Against placebo, the combination group’s bone mineral density was *unchanged* at the hip (−0.006 g/cm², P = .24) and lumbar spine (−0.010 g/cm², P = .20), even though that group lost more than twice as much weight. Against exercise, liraglutide alone *decreased* bone density at the hip (−0.013 g/cm², P = .03) and spine (−0.016 g/cm², P = .04).[1]

> The combination arm lost more than twice the weight of placebo, and its hip and spine bone density was unchanged against placebo.

> This is a **predefined secondary analysis** of a trial whose primary purpose was elsewhere. Predefined is much better than post-hoc. Still, 195 people across four arms is roughly 49 per arm, which is small for a bone-density endpoint. Treat it as the best available randomized answer rather than a definitive one.

## And a meta-analysis that points the other way

A 2025 systematic review pooled 25 randomized trials in people with *type 2 diabetes* and found the opposite direction. GLP-1 receptor agonists came out superior to control on lumbar spine BMD (+0.07 g/cm²), hip neck BMD (+0.05) and total hip BMD (+0.06), all highly significant, with no significant increase in fracture risk (RR 0.80, 95% CI 0.47–1.36).[2]

Both results stand, because they measure different things. The trial studied people **without** diabetes and compared the drug against **exercise**. The meta-analysis studied people **with** diabetes and compared against **control**. Diabetes itself damages bone quality, and improving glucose control plausibly helps it; in someone without diabetes there is no such deficit to repair, and what remains is the bone loss that accompanies losing weight.

*Why two good sources disagree.*

|   | The randomized trial[1] | The meta-analysis[2] |
| --- | --- | --- |
| Population | Obesity, **no diabetes** | **Type 2 diabetes** |
| Comparator | Exercise, and placebo | Control |
| Finding | Liraglutide alone **lost** hip and spine BMD vs exercise | GLP-1 RAs **superior** on spine, hip neck and total hip BMD |
| Fracture | Not the endpoint | No significant increase (RR 0.80) |

## What follows from it

- The trial's practical message concerns the exercise: the arm combining both held bone density while losing the most weight.
- Exercise is the intervention with randomized support here. In this trial that meant vigorous interval cycling plus circuit training that included strength work.[1] The muscle literature points the same way, toward resistance training.
- Fractures were not the trial's endpoint, and no fragility fractures were reported in any group.[1] The meta-analysis did test fracture risk and found no significant increase.[2]
- If you have type 2 diabetes, the pooled evidence in your population points toward benefit rather than harm.
- Bone density is one input to fracture risk, not the whole of it. Whether you need a DXA scan is a clinical judgment, not one this article can make.

Our companion page on [bone density and fracture risk](https://glpwatchdog.org/research/glp1-bone-density-and-fracture-risk) covers the fracture question in its own right, and [what the body-composition studies measured](https://glpwatchdog.org/research/glp1-hair-loss-and-muscle-loss) sets out the parallel finding for muscle. There, too, resistance training is the only intervention with randomized backing.

## Frequently asked questions

### Do GLP-1 drugs weaken your bones?

It depends on the population and what you compare against. In a randomized trial of adults with obesity and no diabetes, liraglutide alone lost bone mineral density at hip and spine relative to exercise. In a meta-analysis of 25 trials in people with type 2 diabetes, GLP-1 receptor agonists were superior to control on spine and hip bone density.

### Does exercise protect bone during GLP-1 weight loss?

In this trial it appears to. The group taking liraglutide plus exercise lost the most weight of any arm (16.88 kg), and its bone mineral density was statistically unchanged against placebo at both hip and lumbar spine.

### Do these drugs increase fracture risk?

The meta-analysis of 25 randomized trials found no significant association with increased fracture risk, at a relative risk of 0.80 with a confidence interval spanning 0.47 to 1.36.

### Why do the two studies disagree?

They studied different people against different comparators. The trial enrolled adults without diabetes and compared the drug against exercise; the meta-analysis pooled trials in type 2 diabetes against control. Diabetes itself harms bone quality, so improving glucose control may help bone in that group while offering nothing to someone without the condition.

## References

1. Jensen SBK, Sørensen V, Sandsdal RM, et al. Bone Health After Exercise Alone, GLP-1 Receptor Agonist Treatment, or Combination Treatment: A Secondary Analysis of a Randomized Clinical Trial. JAMA Network Open. 2024. PMID: 38916894. https://pubmed.ncbi.nlm.nih.gov/38916894/
2. Tan Y, Liu S, Tang Q. Effect of GLP-1 receptor agonists on bone mineral density, bone metabolism markers, and fracture risk in type 2 diabetes: a systematic review and meta-analysis. Acta Diabetologica. 2025. PMID: 39985672. https://pubmed.ncbi.nlm.nih.gov/39985672/

---

Cite as: GLP Watchdog, Nora Bissett, "Bone Density on a GLP-1: The Trial That Compared It Against Exercise", published 2026-08-21, updated 2026-10-27. https://glpwatchdog.org/research/glp1-bone-density-and-exercise

Canonical HTML page: https://glpwatchdog.org/research/glp1-bone-density-and-exercise

Medical disclaimer: Nothing here is medical advice. We report on this subject; we do not treat anyone. Whether to begin a medication, supplement, or treatment, whether to adjust it, and whether to stop are questions for a clinician licensed to answer them, who knows your history.