Is endometriosis genetic? What a 2026 genome study of one million people found
A Nature Genetics GWAS of over one million women found 58 risk regions and 314 genes linked to endometriosis. Heredity, family history and what changes for training.
revns editorial · Published November 20, 2026 · 7 min read

Guare et al. in Nature Genetics (20 Nov 2026) meta-analysed over one million participants and mapped 58 genomic regions for endometriosis. Genetics matter; training rules do not flip overnight.
Yes — endometriosis has a real genetic component, but risk is spread across many DNA regions, not one inherited on/off switch. On 20 November 2026, Nature Genetics published an open-access genome-wide association study (GWAS) led by Lindsay A. Guare and colleagues, pooling more than one million participants from 14 global biobanks (about 32% non-European ancestry in the largest phenotype). The team reports 58 genomic regions tied to endometriosis risk, 27 of them not reported before, and 314 genes implicated after multi-omic follow-up. If you searched is endometriosis genetic or does endometriosis run in families, that is the headline: heritage matters in the data, yet a positive family story is neither required nor sufficient for any one person — and no clinical genetic test ships from this paper.
TL;DR
| Topic | What Guare et al. report (Nature Genetics, 20 Nov 2026) |
|---|---|
| Scale | GWAS meta-analysis across 14 biobanks; largest "wide" phenotype 1,063,657 women with 50,265 cases (31.9% non-European) |
| New geography | 27 previously unreported risk regions among 58 consolidated regions; some African- and admixed American–specific loci |
| Genes | 314 genes implicated across multi-omic analyses — research targets, not a patient gene list to memorize |
| Heritability context | Paper cites twin estimates near 47% broad-sense heritability; prior GWASs explained only ~7% of phenotypic variance |
| SNP heritability (wide phenotype) | Liability-scale estimates about 0.108 (European), 0.133 (African), 0.106 (East Asian) in single-ancestry analyses |
| Comorbidity signals | Genetic correlations with traits such as headache, insomnia and antidepressant prescriptions in Mendelian-randomization screens (associations, not exercise prescriptions) |
| Your week | Log pain and training; tell clinicians about family history; do not expect a home DNA result; movement stays symptom-led |

What did the Nature Genetics study find?
The peer-reviewed article "Multiancestry genome-wide association and integrative multiomics analyses implicate key pathways in endometriosis" appeared in Nature Genetics on 20 November 2026 (Nature Genetics). It comes from the Global Biobank Meta-analysis Initiative, with coordinating work at the Penn Medicine Biobank and contributors including the Million Veteran Program, All of Us, FinnGen and other programmes named in the methods.
Researchers built six case–control definitions from electronic health records — from a broad "wide" phenotype to narrower groups confirmed by surgery or imaging. The largest, phenotype W, included 1,063,657 women and 50,265 endometriosis cases. They ran 11 meta-analyses (multi-ancestry and single-ancestry), then layered pleiotropy, single-cell expression and transcriptome/epigenome/proteome tools.
Headline discovery numbers from the abstract and results:
- 58 genomic regions after consolidating linkage blocks — 27 previously unreported
- 3,342 unique genome-wide significant variants across analyses, rolled into 108 distinct loci and 143 independent signals in the wide meta-analysis
- 314 genes implicated when multi-omic pipelines were included — the authors frame these as potential therapeutic targets for research, not drugs to self-select
- 14 loci with at least one variant carrying posterior inclusion probability above 0.5 in fine-mapping — stronger evidence for specific causal variants at those loci
The paper also highlights ancestry-specific signal: for example, two African-ancestry–specific wide-phenotype regions and loci unique to African and admixed American participants in narrower phenotypes. That matters for equity: older GWASs skewed European; this design deliberately widens the lens.
Medical Xpress and other desks covered the work on 20 November 2026 as genetic data from over one million people (Medical Xpress summary). The actionable science lives in the journal article, not the newswire paraphrase.
Is endometriosis genetic or environmental?
Both — and that is why the question frustrates people in pain. Endometriosis is tissue similar to the womb lining growing outside the uterus; hormones and immune signalling still drive flares even when DNA raises baseline risk.
The Guare paper opens with prevalence near 10% of people of reproductive age with a uterus — consistent with estimates used in our November review on pain mechanisms and exercise. Genetics sets part of the loading; retrograde menstruation, inflammation, pain sensitization and scarring still explain why two relatives with similar variants diverge in symptoms.
Twin literature in the introduction points to ~47% broad-sense heritability, while earlier endometriosis GWASs captured only about 7% of phenotypic variance from common variants. The 2026 work closes part of that gap by scale (more cases, more ancestries, sharper phenotyping) and by multi-omics that points from SNPs toward genes and cell types — stromal, epithelial, immune and endothelial contexts in the single-cell section, with ESR1 and GREB1 among frequently dysregulated names in menstrual-cycle–stratified data.
That is mechanistic progress for labs. For you, it reinforces a familiar clinic message: family history increases suspicion; lack of family history does not clear you.
Does endometriosis run in families?
Often, yes — and this study maps why at the DNA level. Family stories show up in waiting rooms long before biobanks existed: mothers and sisters with laparoscopy scars, cousins with "bad periods" never investigated.
Biobanks capture diagnosed cases from codes and procedures. The authors report variable case rates by ancestry and phenotype strictness — for example, European ancestry showed higher case proportions than other groups on wide phenotypes, while narrower surgically confirmed groups shifted rankings. That is a reminder that health-system access and surgery rates shape who appears in a GWAS, not just biology.
Practical steps if families ask is endometriosis hereditary:
- List relatives with severe period pain, endometriosis surgery or infertility work-ups — names and approximate ages at diagnosis if known.
- Bring that list to a GP or gynaecologist with your own symptom timeline (pain pattern, bowel/bladder symptoms, bleeding).
- Avoid treating family history as fate. Risk is probabilistic; many carriers never meet diagnostic criteria in their twenties.
revns does not diagnose or genotype you. Logging cycles, pain scores and which training days were fine versus impossible helps those conversations — see how to track your period alongside workouts in the AI coach view.
What about headache, sleep and mood in the same dataset?
Active women often notice migraine, insomnia and low mood next to pelvic pain. The Guare paper’s genetic correlation and Mendelian-randomization section screens thousands of traits. Examples reported at stringent thresholds include positive genetic correlation with headache, insomnia/sleeplessness and antidepressant prescriptions, among many others.
Those numbers describe population-level overlap, not "you must have all three." They align with clinical experience that endometriosis behaves like a systemic inflammatory and pain disorder, which is why our pain-mechanisms news piece emphasized nerves and hormones beyond lesion size.
If you train with recurring headache or poor sleep, track them next to load — our guides on sleep and mood and training around your cycle stay relevant. Do not change psychiatric medication because a GWAS mentioned prescriptions; that is a prescriber conversation.

How does this compare with the earlier 2026 million-person GWAS?
Readers may also see April 2026 coverage of a separate Nature Genetics paper analysing on the order of 1.4 million women and 105,869 cases, led by Dora Koller and an international team including Yale School of Medicine and University of Barcelona (Yale School of Medicine release). That study reported 80 genomic regions (37 new) with integrated multi-omics and drug-repurposing hypotheses.
The 20 November 2026 Guare study is not a duplicate press release: it stresses multi-ancestry biobank phenotyping, procedure-confirmed subsets, fine-mapping with ancestry-specific credible sets, and 314 genes after xQTL-style follow-up. Science moves in waves — bigger maps, then sharper maps. For daily life, both papers say the same patient-level truth: research is ahead of the clinic.
Neither paper releases a validated home genetic test. Yale’s plain-language summary for the earlier work states a genetic test is not yet available for clinical use.
What should you do this week?
- If pelvic pain disrupts training monthly, book a clinician review before you overhaul your program. Average diagnostic delay in literature remains many years after symptoms — genetics will not shorten that wait by itself.
- For two cycles, log pain 0–10, bleeding heaviness, bowel/bladder symptoms, fatigue and which sessions felt fine. Bring the log to appointments.
- Keep movement symptom-led. Gentle rhythmic options you can repeat (walking, easy cycling, light mobility) are reasonable experiments while specialists lead diagnosis — consistent with the Endocrine Reviews exercise discussion, not a prescription for everyone.
- Skip consumer DNA kits for this purpose. Polygenic scores may enter research clinics years from now; today’s care still leans on history, exam and often imaging or surgery.
- Stack recovery basics. Sleep regularity, protein across your diet pattern (vegetarian or not), and stress load matter for pain tolerance even when genes load risk — see iron, energy and training for women if fatigue dominates.
When should you see a doctor?
Book routine review if period pain worsens year on year, pain persists between bleeds, sex hurts deeply, or bowel/bladder pain tracks your cycle. Seek urgent care for sudden severe pelvic pain, fainting with heavy bleeding, fever with pelvic pain, or weakness/numbness in the legs.
Mental health: if you feel hopeless or have thoughts of harming yourself, contact a clinician or crisis line immediately. In the UK, Samaritans are on 116 123 (free, 24 hours). Endometriosis is not "all in your head"; it is also not something an app treats.
Red flags that should not wait for another deload week:
- Pain that wakes you nightly despite rest
- Bowel or bladder dysfunction that is new or worsening
- Suspected pregnancy with one-sided pain and dizziness
This is general training guidance, not medical advice.
Related
- Endometriosis pain and exercise: 2026 Endocrine Reviews review
- Period pain and exercise: what helps
- Perimenopause vs menopause: symptoms and training
- Training around your menstrual cycle
Sources
- Guare LA et al. Multiancestry genome-wide association and integrative multiomics analyses implicate key pathways in endometriosis. Nature Genetics 2026
- Koller D et al. Multi-ancestry genome-wide association and integrated multi-omics analyses of endometriosis and its clinical manifestations. Nature Genetics 2026
- Yale School of Medicine news release on Koller et al., 2026
Terms in this post
revns publishes training and nutrition guidance, not medical advice. See a clinician for symptoms, injury or a medical condition. Accurate as of November 20, 2026.


