GLP-1 + Osteoporosis: DEXA + Bisphosphonate Stacking Evidence

Every kilogram of weight loss takes a small share of bone with it. The classic Study of Osteoporotic Fractures cohort (Ensrud 2003^[1]) quantified roughly 1% hip BMD loss per 5% body-weight loss in older women, and the published post-bariatric data are several times larger (Mitchell 2023^[5]). GLP-1 therapy sits in between — faster and larger than diet alone, slower and smaller than sleeve gastrectomy. For most adults under 60 with normal baseline DEXA, the bone change is clinically irrelevant. For postmenopausal women, men over 70, anyone with a baseline T-score below −1.0, or anyone titrating to maximum dose with target weight loss above 20%, the bone math deserves a protocol. This article walks through what the published BMD data actually say, what the bisphosphonate / denosumab / romosozumab options do, and the practical DEXA monitoring cadence we would build into a Wegovy or Zepbound program.

The honest summary

• Bone loss tracks weight loss. Ensrud 2003^[1] in the SOF cohort found roughly 1% hip BMD loss per 5% body-weight loss in older women, with both intentional and unintentional weight loss increasing subsequent hip-fracture risk.
• GLP-1 BMD signal is small but real. The one published prospective head-to-head (Iepsen 2015^[2]) randomized weight-reduced women to liraglutide vs control and found liraglutide increased bone formation markers and prevented BMD loss, while the control arm lost roughly 2.5% of total hip BMD over 52 weeks. Indirect evidence from STEP-1^[3] and SURMOUNT-1^[4] is consistent with a 0.5–2% 12-month BMD reduction in higher-risk subgroups.
• Bariatric surgery is the cautionary tale. Mitchell 2023^[5] in adolescents and young adults found roughly 5–10% total hip and femoral neck BMD loss at two years post sleeve gastrectomy — several times the GLP-1 signal at comparable weight loss. The GLP-1 picture is closer to diet than to surgery.
• Exercise is the cheapest mitigator. Villareal 2017 NEJM^[6] (the LIFE-Moves trial) randomized older obese adults to diet plus aerobic, resistance, combined, or no exercise; the combined-exercise arm preserved hip BMD while diet-only lost meaningful BMD.
• Pharmacotherapy works. HORIZON^[7] (zoledronic acid IV), FREEDOM^[8] (denosumab SQ), and ARCH^[9] (romosozumab vs alendronate) all show robust hip and vertebral fracture reduction in postmenopausal osteoporosis. The AACE 2020 guideline^[10] is the practical reference.

What we know about GLP-1 and bone density

The cleanest mechanistic study remains Iepsen 2015 in JCEM^[2]. Investigators put 37 obese women through a low-calorie diet to roughly 12% body-weight loss, then randomized them to either liraglutide 1.2 mg daily or control for 52 weeks of weight maintenance. The control arm lost another 2.5% of total hip BMD over the maintenance year. The liraglutide arm did not. Bone formation markers (P1NP) rose on liraglutide and fell on control. The trial was small and the dose is well below today’s semaglutide 2.4 mg or tirzepatide 15 mg, but the directional signal is clear: GLP-1 receptor agonism is not actively bone-toxic, and may attenuate the bone loss that otherwise accompanies a sustained caloric deficit.

The bigger modern trials (STEP-1, Wilding 2021^[3]; SURMOUNT-1, Jastreboff 2022^[4]) did not pre-specify BMD endpoints, so we are stuck with indirect evidence and substudy-level reports. The directional picture across the published data: roughly 0.5–2% 12-month BMD reduction in higher-risk subgroups (postmenopausal women, men over 70, rapid titration to maximum dose with target weight loss above 20%). For a 55-year-old with a baseline T-score of 0 losing 15% body weight on tirzepatide, that change is clinically irrelevant. For a 68-year-old postmenopausal woman with a baseline T-score of −2.0 doing the same, it crosses the line where a bisphosphonate prophylactic is worth the conversation.

The bariatric contrast: why the GLP-1 number is not the surgery number

Mitchell 2023 in JCEM^[5] followed 54 adolescents and young adults for two years after sleeve gastrectomy and measured total hip and femoral neck BMD by DEXA plus volumetric BMD by quantitative CT. The published 2-year changes were in the −5% to −10% range — several times what the GLP-1 literature suggests at comparable total weight loss. The mechanism is partly the speed of weight loss, partly nutritional malabsorption (calcium, vitamin D, protein), and partly the loss of the gastric gut-hormone milieu. The clinical implication is important: do not extrapolate post-bariatric DEXA monitoring intensity onto GLP-1 patients. Post-bariatric patients need annual DEXA. Most GLP-1 patients do not.

Who actually needs a baseline DEXA

The AACE 2020 guideline^[10] recommends baseline BMD measurement for women age 65 or older, men age 70 or older, and any adult with clinical risk factors for fracture. For GLP-1 candidates we would add a baseline DEXA for anyone meeting any of:

• Postmenopausal women age 60 or older. Baseline BMD is the lever that decides whether 1–2% GLP-1-attributable loss matters.
• Men age 70 or older.
• Prior fragility fracture at any adult age.
• Chronic glucocorticoid use (≥ 5 mg prednisone-equivalent daily for ≥ 3 months).
• Planned aggressive titration to maximum dose with target TBWL > 20%.
• Family history of hip fracture, current smoking, chronic kidney disease, hyperthyroidism, or primary hyperparathyroidism.

Patients who fail any of those criteria deserve a repeat DEXA at 12 months. If the rate of BMD loss exceeds the least-significant-change threshold for that scanner (typically about 3% at the hip, 5% at the lumbar spine), the AACE guideline supports anti-resorptive initiation regardless of T-score.

Magnitude: BMD change by intervention at 12 months

The pharmacotherapy menu

Once a patient hits the threshold for treatment (T-score ≤ −2.5, prior fragility fracture, or T-score −1.0 to −2.5 with FRAX 10-year hip fracture risk ≥ 3% or major osteoporotic fracture risk ≥ 20%), the AACE 2020 guideline^[10] provides a clear ladder.

• Oral bisphosphonates — alendronate (Fosamax, 70 mg weekly), risedronate (Actonel, 35 mg weekly or 150 mg monthly), ibandronate (Boniva, 150 mg monthly). Cheap (generic, $5–15/month), proven, but require 30–60 minutes upright + empty stomach dosing. The dosing window is the practical problem on a GLP-1, which slows gastric emptying.
• Zoledronic acid IV (Reclast, 5 mg IV annually) — HORIZON-PFT (Black 2007 NEJM^[7]) showed 70% vertebral and 41% hip fracture reduction over 3 years. Bypasses gastric emptying entirely, so the GLP-1 interaction concern disappears. Best practical option for GLP-1 patients who need anti-resorptive therapy. Cash cost $250–1,500 per infusion; covered by most insurance.
• Denosumab (Prolia, 60 mg SQ q6 months) — FREEDOM (Cummings 2009 NEJM^[8]) showed 68% vertebral and 40% hip fracture reduction. No GI absorption, so no GLP-1 interaction. The catch: discontinuation causes rapid rebound bone loss and a spike in vertebral fracture risk, so denosumab is a lifetime commitment unless transitioned to a bisphosphonate. Sticker price ~$1,500 per dose; most commercial insurance covers it.
• Romosozumab (Evenity, 210 mg SQ monthly × 12) — ARCH (Saag 2017 NEJM^[9]) showed romosozumab plus alendronate beat alendronate alone on hip and vertebral fracture. Anabolic (bone-forming) plus anti-resorptive in one molecule. Reserved for high-risk patients with prior fracture. Boxed warning for cardiovascular events; avoid if prior MI or stroke.
• Teriparatide (Forteo) and abaloparatide (Tymlos) — anabolic, daily SQ. Effective but expensive ($2,000–3,000/month) and limited to a 24-month lifetime course. Useful for very high baseline risk; usually managed by endocrinology.

The GLP-1 dosing interaction question

GLP-1 receptor agonists slow gastric emptying by 30–70 minutes in the early titration window. That matters for any oral drug whose Cmax depends on rapid absorption from an empty stomach. Oral bisphosphonates fit that pattern: the labeled dosing window is 30–60 minutes before any food, water only, upright. On a GLP-1 with delayed gastric emptying, the practical risk is a sub-therapeutic Cmax and a higher rate of esophageal irritation if the tablet lingers.

The pragmatic resolution: if a GLP-1 patient needs anti-resorptive therapy, prefer zoledronic acid IV annually or denosumab SQ every 6 months. Both bypass the GI absorption question. If oral bisphosphonate is the only option (cost, access), schedule the bisphosphonate on the day farthest from the most recent GLP-1 injection — for once-weekly semaglutide or tirzepatide, that means days 5–6 after the injection, when gastric emptying has returned closer to baseline.

Exercise: the cheapest mitigator

Villareal 2017 NEJM^[6] (LIFE-Moves) randomized 160 obese adults age 65+ to a 39-week weight-loss program with either aerobic exercise, resistance exercise, both, or no exercise. The diet-only arm lost about 3% of total hip BMD. The combined aerobic plus resistance arm preserved hip BMD while still achieving roughly 9% weight loss. The mechanism is mechanical loading: weight-bearing impact and progressive resistance both signal osteoblast activity. The translation for GLP-1 patients is unglamorous — 2–3 sessions per week of compound resistance training (squat, hinge, press, row) plus 30 minutes of weight-bearing aerobic activity most days. The same protocol that preserves muscle (see our muscle-loss prevention article) does most of the bone-preservation work too.

The practical bone-health protocol

1. Baseline DEXA for postmenopausal women age 60+, men age 70+, anyone with prior fragility fracture, chronic glucocorticoids, or planned aggressive titration with target TBWL > 20%. Order at the same visit as GLP-1 initiation.
2. Calcium 1,000–1,200 mg/day + vitamin D 800–2,000 IU/day. Check baseline 25-OH-D; repeat at 6 months. Target serum 25-OH-D ≥ 30 ng/mL.
3. Resistance training 2–3 sessions/week + weight-bearing aerobic activity most days. Compound lifts; progressive load.
4. Repeat DEXA at 12 months for any baseline-DEXA patient. Same scanner if possible (BMD comparisons across scanners are unreliable).
5. Initiate anti-resorptive therapy if any of: T-score ≤ −2.5, prior fragility fracture, T-score −1.0 to −2.5 with FRAX 10-year hip fracture risk ≥ 3% or major osteoporotic fracture risk ≥ 20%, or 12-month BMD loss exceeds the least-significant-change threshold for the scanner.
6. Prefer IV zoledronic acid or SQ denosumab over oral bisphosphonates in GLP-1 patients. Bypasses the gastric-emptying interaction question. If oral, dose on days 5–6 after the weekly GLP-1 injection.
7. Dental clearance before initiating any bisphosphonate or denosumab. Osteonecrosis of the jaw is rare but well-documented; pre-treatment dental extractions and implants should happen first.
8. Re-evaluate at 24 months on the combined protocol. Many patients stabilize and do not need ongoing anti-resorptive therapy once weight is stable.

Fracture-risk modeling: when the conversation becomes urgent

The FRAX tool (developed by the WHO Collaborating Centre, Sheffield) converts age, sex, BMI, prior fracture history, parental hip fracture, smoking, glucocorticoid use, rheumatoid arthritis, secondary osteoporosis, alcohol intake, and femoral neck BMD into a 10-year probability of major osteoporotic fracture and hip fracture. The AACE 2020 threshold^[10] for pharmacotherapy is a 10-year hip fracture risk ≥ 3% or major osteoporotic fracture risk ≥ 20%. For a postmenopausal GLP-1 candidate, running FRAX at baseline and again at 12 months — with the updated weight and BMD — surfaces the patients in whom rapid weight loss has tipped them across the treatment threshold.

Related research

• GLP-1 muscle loss prevention protocol — the protein + resistance-training half of the body-composition story, which double-counts as bone-preservation
• GLP-1 in adults 65 and older — the demographic in which bone density and fracture risk become decision-grade
• BHRT plus GLP-1 in perimenopause — estradiol’s independent BMD-protective effect during the weight-loss window

Important disclaimer. This article is educational and does not constitute medical advice. Osteoporosis pharmacotherapy decisions should be individualized by a clinician with access to baseline DEXA, FRAX, calcium and vitamin D status, and the patient’s fracture history. Bisphosphonates and denosumab carry rare but real risks of osteonecrosis of the jaw and atypical femoral fracture; romosozumab carries a boxed warning for cardiovascular events. Dental clearance before initiation is standard of care. PMIDs were verified live against the PubMed E-utilities API on 2026-05-29.

Last verified: 2026-05-29. Next review: every 12 months, or sooner if a SURMOUNT-derived or STEP-derived BMD substudy is published or if AACE or the Endocrine Society updates its osteoporosis treatment thresholds.