Pulse ·
Endometriosis hormone therapy and bone loss: what 37 studies show
A 2026 meta-analysis of 37 studies — 28 RCTs — found all commonly prescribed endometriosis hormone therapies cause measurable bone mineral density reduction. GnRH antagonists (elagolix, linzagolix, relugolix) produced the largest losses at 2.17% over six months; GnRH agonists with add-back therapy, synthetic progestins including dienogest, and depot medroxyprogesterone acetate showed smaller but consistent reductions.
Bone loss appears to stabilise after the first six months with continued treatment. Long-term fracture risk remains uncertain and must be weighed against symptom burden, treatment duration, and individual baseline bone health risk factors.
If you are managing endometriosis — or advising someone who is — and you have wondered what the cumulative effect of hormone treatment is on bone health, there is now a cleaner evidence base to work from.
A 2026 meta-analysis published in Obstetrics & Gynaecology, pooling data from 37 studies — 28 of them randomised controlled trials — found measurable bone mineral density reductions associated with all major endometriosis hormone therapies. The effect is not uniform across drug classes, and the size matters when you are weighing treatment options. But the headline is: no commonly used hormone therapy for endometriosis is neutral on bone.
The analysis covered data from Europe, Asia, and North America, with bone mineral density measured primarily at the lumbar spine — the most frequently assessed site across included studies — and outcomes tracked at six and twelve months across the main drug classes used for symptom control.
What the numbers show
GnRH antagonists — the newer class that includes elagolix, linzagolix, and relugolix — showed the largest reductions at 2.17% bone mineral density loss at six months. These medications work by suppressing oestrogen; more complete suppression correlates with more effective pain relief and more bone exposure.
GnRH agonists — an older class including leuprorelin and goserelin — showed similar oestrogen suppression but produced smaller measurable bone effects when add-back therapy was used: 1.33% reduction for leuprorelin, 0.42% for goserelin at six months. Add-back therapy — progestins, combined oestrogen-progestin, or tibolone — appeared similarly protective across approaches, which is clinically relevant for treatment planning.
Dienogest (Visanne), the synthetic progestin used widely in Australia for endometriosis, showed 0.83% bone mineral density reduction at six months and 1.91% at twelve months. Depot medroxyprogesterone acetate (Depo-Provera) showed 0.99 to 1.1% reduction at six months.
One consistent finding across the analysis: bone loss appears to stabilise with ongoing treatment, with the greatest changes in the first six months and plateauing thereafter. Long-term fracture risk remains uncertain from current evidence.
On the Australian PBS: dienogest and relugolix (Ryeqo) are currently listed; the low-dose goserelin 3.6mg formulation is being withdrawn from November 2026; leuprolide, elagolix, and linzagolix are not PBS-listed.
Both-and
The evidence does not suggest stopping effective endometriosis treatment on the basis of bone concern alone. Endometriosis causes real, sometimes severe, sometimes fertility-affecting harm. The bone effect must be weighed against that harm — and against the bone health impact of undertreated disease itself, including inflammation and the surgical procedures that occur when medical management fails.
That 2.17% bone mineral density reduction for GnRH antagonists is a real number — not alarming at six months for most patients, but meaningfully different from zero, and the population of people managing endometriosis is not a low-risk bone health population by default. Women in their 30s and 40s managing endometriosis are approaching the bone health picture that precedes menopause. That context matters when thinking about baseline and trajectory.
The finding that bone loss stabilises with ongoing use is reassuring for sustained treatment. But “stabilises” does not mean “recovers,” and the uncertainty around long-term fracture risk is genuine, not rhetorical. The meta-analysis is explicit that the skeletal impact requires contextualisation against symptom severity, treatment duration, and individual patient risk — there is no single threshold number that drives a universal recommendation.
The add-back therapy data is also practically useful. For people on GnRH agonists — leuprorelin or goserelin — the evidence now supports that add-back approaches are meaningfully bone-protective, and the choice between add-back formulations appears similarly effective. That is a conversation worth having with your GP if add-back has not been discussed.
My two cents
If you are currently on hormone therapy for endometriosis and have not had a conversation with your GP about bone health monitoring, this evidence provides a grounded reason to have it. The relevant considerations: how long you have been on treatment, at what baseline bone density, and whether other factors affect your bone health picture — dietary calcium intake, vitamin D, weight-bearing activity, smoking, corticosteroid use, other medications.
This is not cause for alarm. It is cause for a complete picture. For some people, knowing that add-back therapy reduces bone effects with GnRH agonists will change the treatment conversation. For others, understanding that dienogest’s bone effect is smaller and stabilises at six months may be reassuring for sustained use. The evidence is now specific enough to have a drug-class-level discussion rather than a general “these medications affect bone” conversation.
Bone health is one of the variables that should be tracked, not assumed neutral, in extended endometriosis treatment. General practice is a reasonable place to start that conversation, with referral for DEXA scanning where the individual risk picture warrants it.
Verdict: yes — clinically useful evidence that gives general practice specific, drug-class-level data to inform endometriosis treatment planning.