Unveiling endometriosis hidden comorbidities using a data-driven approach: a retrospective matched cohort study

Overview of data-driven analysis The multi-event endometriosis cohort consisted of 11,848 individuals of which 8299 have been matched to non-endometriosis counterparts. Table1 provides an overview of the characteristics of the endometriosis and non- endometriosis cohorts. While the median age of participants in both cohorts 1KI Research Institute, Kfar Malal, Israel. 2Department of Obstetrics and Gynecology, Sheba Medical Center, Ramat Gan, Israel. 3Faculty of Health Sciences, Ben Gurion University of the Negev, Beer Sheva, Israel. 4Maccabi Health Services South District, Beer Sheva, Israel. e-mail: [email protected] npj Women's Health | (2025)3:30 1 1234567890():,; 1234567890():,; is 34 years, endometriosis patients had signi ficantly more annual visits compared to non-endometriosis individuals (median of 10 [IQR: 7, 15] versus 6 [4, 10], respectively). This gap, likely a result of the ongoing medical attention required to monitor and treat a chronic and progressive condition, attests to the substantial impact of endometriosis on healthcare resource utilization. We leveraged SNOMED-CT hierarchy (see Methods for further details) to identify the most speci fic (lowest-level) conditions that differ meaningfully between cohorts, as th ese often offer the most actionable insights for further investigation. Our data-driven analysis compared the prevalence of 4850 conditions and c ondition groups recorded for endo- metriosis and control, non-endometriosis individuals. Of these conditions, 973 (20.1%) have been signi ficantly ( q-value 0.1) more prevalent in the endometriosis cohort than in the non- endometriosis one. The full results are available in Supplementary Data 1 and Data 2. Here, we focus on four examples that support the validity of our analysis by reproducing previously reported associations: migraines,fibro- myalgia, allergic disorders, and reflux; and four examples that showcase its breadth and highlight associations that, to the best of our knowledge, have not yet been reported in prior research: sinusitis, laryngitis, herpes virus infection, and sciatica—the latter being a known rare symptom. For each example, we present the prevalence ofthe corresponding condition and its hierarchical descendants in the endo metriosis and non-endometriosis cohorts and assign measures of statistical significance and clinical meaning to their observed difference. Representative endometriosis comorbidities First, we examined the prevalence of headache disorders, a known comorbidity of endometriosis, as shown in Supplementary Table S1 and Fig. S1. Our analysis revealed a signi ficant difference between the endometriosis and non-endometriosis cohorts, with 29% and 17% experiencing headache disorders, respectively. Delving deeper, we found that vascular headaches, speci fically migraines 8,23,26–29, were pre- dominant, affecting 24% of the endometriosis cohort and 13% of the non-endometriosis cohort. Similarly, and as has been previously demonstrated, the prevalence of fibromyalgia 7,13,15 is signific a n t l yg r e a t e ri no u re n d o m e t r i o s i sc o h o r tt h a ni n the non-endometriosis one (3.7% vs 1.6%, respectively; Supplementary Table S2 and Fig. S2). As a result, the ancestor condition group chronic pain is diagnosed in the endometriosis cohort more often than in the non- endometriosis cohort (5.9% vs 1.9%, respectively). Table 2,F i g .1, and Supplementary Fig. S3 compare the prevalence of allergic disorders 7,11–16,21,22,24,30,31 and sinusitis (i.e., previously reported and unreported conditions, respectivel y) and their descendant conditions. Allergic disorders are signi ficantly more prevalent in the endometriosis cohort (24% versus 18% in the non-end ometriosis cohort), with allergic rhinitis contributing most to this difference (22% versus 15%, corre- spondingly). For sinusitis, the endometriosis cohort has a prevalence of 32%, compared to 20% in the non-endometriosis cohort, with two meaningful contributing descendant comorbidities: chronic sinusitis (25% versus 16%) and acute sinusitis (3.8% versus 1.8%). Table 3 and Supplementary Figs. S4-S5 present the top comorbidities finding of esophagus and disorder of the larynx, as well as their descendants. For the former condition group, as previously reported 28,29,31–33,t h ed i f f e r - ences between the endometriosis and non-endometriosis cohorts are 20% vs 12%, respectively. The main contributors to these differences are esophageal reflux (12% vs 8%), esophagitis (6.3% versus 3.2%), and gastroesophageal reflux disease (4.5% vs 2.3%). The condition group disorder of the larynx was diagnosed in 9.6% of individuals in the endometriosis cohort and 5.9% in the non-endometriosis cohort. The main contributor to this difference is acute laryngitis, with 8.2% vs 5%, respectively. Next, we present the prevalence o f conditions grouped under viral disease in Table 4 and Supplementary Fig. S6, to explore the connection between endometriosis and the immune system. Our analysis revealed a striking difference between the endometriosis and non-endometriosis cohorts, with 55% and 42% experiencing viral infections, respectively. Delving deeper, the top categories co ntributing to viral infections were herpesvirus infection (23% vs 17%), viral infection of the skin (15% vs 11%), viral respiratory tract infection (14%vs 11%), viral infection of the digestive tract (12% vs 8.2%), and diseases due to papillomaviridae (18% vs 13%). To emphasize the significance of these findings, it is important to note that 12 other conditions (at the same level of hierarchy) were found to be not meaningful clinically. Finally, we compare the prevalence of the condition group neuro- pathy of lower limb and its descendant conditions in the matched cohorts (Supplementary Table S3 and Fig. S7). This group has been diagnosed in 12% of endometriosis patients, but only 7.7% of non-endometriosis individuals. The condition contributing most to this difference is sciatica with rates of 11% and 7.1% in the endometriosis and non-endometriosis, respectively. Sensitivity analysis As sensitivity analysis, we expanded the target cohort to include all females with one or more endometriosis-related event and identi fied a matching cohort (by sex and year of birth); the characteristics of these single-event cohorts are shown in Supplementary Table S4. We then compared the prevalence of 7129 conditions and condition groups recorded in these cohorts; results for the conditions discussed above appear in Supplementary Tables S5-S10, and full results can be found in Supplementary Data 1 and Data 3. Overall, 603 (8.5%) conditions have been signi ficantly (q-value 0.1) more prevalent in the endometriosis cohort than in the non-endometriosis one and, reassuringly, these condi- tions overlap with the ones identified in the multi-event analysis. We note that, presumably, the smaller number (and percentage) of clinically meaningful conditions identified in the single-event analysis attests to the lower disease severity in the corresponding patients, compared to the multi- event cohort. Finally, we note that our analyses identi fied hundreds of conditions (522 or 10.8% and 1036 or 14.5% in the mul ti- and single-event analysis, respectively) that are more prevalent in the non-endometriosis cohort, but only “delivery normal” was both statistically significant and (borderline) meaningful. Table 1 | Characteristics of the endometriosis and non- endometriosis cohorts in the IMRD-THIN database Endometriosis Non- Endometriosis P-valueb N 8299 Age [years]a 34 (28, 39) 34 (28, 39) – Baseline period [years] a 4.2 (2.2, 7.7) 4.5 (2.4, 7.9) <0.001 Follow-up period [years] a 6.6 (3.3, 11.2) 5.6 (2.3, 10.6) <0.001 Average annual visit count a 10 (7, 15) 6 (4, 10) <0.001 Country of residence England 5846 (70%) 5814 (70%) 0.6 Scotland 977 (12%) 1107 (13%) 0.002 Wales 936 (11%) 895 (11%) 0.3 Northern Ireland 540 (6.5%) 483 (5.8%) 0.066 Townsend deprivation index Deprived c 2199 (27%) 2538 (32%) <0.001 Missing 295 383 aMedian (interquartile range, IQR). bWilcoxon rank sum test. cTownsend deprivation index was dichotomized into a binary variable, with values ≥ 4 considered deprived. Baseline and follow-up periods measure the time span of data available for individuals before and after the index date, respectively; average annual visit count indicates the number of date-distinct medical visit each individual had, divided by their data time span https://doi.org/10.1038/s44294-025-00073-z Article npj Women's Health | (2025)3:30 2

Our study utilized a data-driven, hie rarchical approach to analyze com- munity healthcare data, identifying comorbidities more prevalent in indi- viduals with endometriosis comparedt oam a t c h e dc o n t r o lc o h o r t .W e found hundreds of statistically signi ficant and clinically meaningful comorbidities, some of wh ich corroborate existing literature, thereby affirming the validity of our methodology, while others are novelfindings. Notably, our analysis revealed a significant association between endome- triosis and migraines8,23,26–29 (SMD = 0.3), aligning with previous studies that suggest overlapping factors such as chronic pain, hormonal fluctuations, Table 2 | Comparison of the prevalence of allergic disorders and sinusitis conditions in the endometriosis and matched control cohorts Condition Endometriosis Non-Endometriosis q-value SMD Allergic disorders Allergic disorder 2027 (24%) 1479 (18%) <0.001 0.2 –Allergic disorder caused by substance 1397 (17%) 960 (12%) <0.001 0.2 ––Allergic rhinitis due to pollen 1396 (17%) 959 (12%) <0.001 0.2 –IgE-mediated allergic disorder 1804 (22%) 1264 (15%) <0.001 0.2 ––Atopic IgE-mediated allergic disorder 1804 (22%) 1264 (15%) <0.001 0.2 –––Allergic rhinitis 1804 (22%) 1264 (15%) <0.001 0.2 –Allergic disorder of respiratory system 1812 (22%) 1286 (15%) <0.001 0.2 ––Allergic disorder of respiratory tract 1812 (22%) 1286 (15%) <0.001 0.2 –––Allergic asthma 52 (0.6%) 39 (0.5%) 0.3 0.021 –Non-IgE-mediated allergic disorder 44 (0.5%) 25 (0.3%) 0.05 0.036 ––Allergic contact dermatitis 44 (0.5%) 24 (0.3%) 0.037 0.038 –Allergic disorder of skin 94 (1.1%) 74 (0.9%) 0.2 0.024 ––Allergic urticaria 51 (0.6%) 50 (0.6%) >0.9 0.002 –Allergic conjunctivitis 304 (3.7%) 233 (2.8%) 0.006 0.048 ––Chronic allergic conjunctivitis 11 (0.1%) 11 (0.1%) >0.9 0 ––Atopic conjunctivitis 292 (3.5%) 218 (2.6%) 0.003 0.052 –––Acute atopic conjunctivitis 32 (0.4%) 30 (0.4%) 0.9 0.004 Sinusitis Sinusitis 2652 (32%) 1670 (20%) <0.001 0.3 –Chronic sinusitis 319 (3.8%) 152 (1.8%) <0.001 0.12 –Acute sinusitis 2096 (25%) 1288 (16%) <0.001 0.2 ––Acute frontal sinusitis 26 (0.3%) 18 (0.2%) 0.3 0.019 ––Acute maxillary sinusitis 66 (0.8%) 34 (0.4%) 0.004 0.05 ––Acute rhinosinusitis 20 (0.2%) 10 (0.1%) 0.13 0.028 –Rhinosinusitis 20 (0.2%) 10 (0.1%) 0.13 0.028 –Maxillary sinusitis 95 (1.1%) 49 (0.6%) <0.001 0.06 –Frontal sinusitis 47 (0.6%) 26 (0.3%) 0.034 0.038 Hyphen length preceding each condition name indicates its SNOMED-CT hierarchical depth (relative to the top condition); descendants are shown up to depth 3. Rows are ordered by depth first search; meaningful differences (SMD > 0.1) are shown in bold face. See Fig. 1 and Supplementary Fig. S3 for a graphical presentation. 0 0.1 0.2 0.3 0.4 0.5 SMD Allergic urticariac Allergic rhinitis tis pollendue to polleollen Allergic rhinitisn is Acute atopic ctivitisconjun Chronic allergicer conjunctivitiscti Allergic contactct dermatitisitis Allergic asthmah Allergic disorder caused by substancey bstance Atopic IgE-mediatedA allergic disorderd IgE-mediatedIg allergic disordersorde Allergic disorderor of respiratory systemory Non-IgE-mediatedIg allergic disorderc Allergic disorder of skindisor Atopic conjunctivitisun Allergic disorder Allergic disorderd of respiratory tract Allergic conjunctivitisco Fig. 1 | SNOMED-CT hierarchical presentation of allergic disorders and its descendant conditions, as shown in Table 2. Nodes are colored by standardized mean difference (SMD); statistically signi ficant and clinically meaningful conditions (or condition groups) are highlighted in boldface. https://doi.org/10.1038/s44294-025-00073-z Article npj Women's Health | (2025)3:30 3 and inflammatory processes may contribute to this correlation. Addition- ally, we confirmed a higher prevalence of fibromyalgia34 in endometriosis patients7,13,15 (SMD = 0.13), suggesting a complex interplay between these chronic pain conditions, potentially linked by mechanisms like central sensitization. Alternatively, it may indicate misdiagnosis of fibromyalgia when endometriosis is the underlying cause. Our research suggests a higher prevalence of allergic disorders in individuals with endometriosis 7,11,16,21,22,24,30,31, potentially involving pro- inflammatory cytokines in the peritonealfluid, which stimulate histamine- releasing factors, or activated mast cells in endometriosis lesions, con- tributing to hypersensitivity reactions 35. Intriguingly, our study also unveiled a higher prevalence of sinusitis among individuals with endometriosis (SMD = 0.3). This finding aligns with a previous survey-based research, indicating a link between endome- triosis and sinusitis symptoms 36 and can potentially be explained through the association between endometriosi s and allergies: allergic symptoms often lead to nasal mucosal edema, i ncreased secretions, and impaired mucociliary function, potentially resulting in sinus obstruction which, in turn, may lead to stagnant debris that becomes infected. From an immu- nologic point of view, the presence of eosinophils that are more prevalent in allergic rhinitis flares can also cause chronic inflammation of the mucosa even without bacteria, and cause sinusitis37. High levels of IgE antibodies, a hallmark of allergic reactions, have also been associated with severe chronic sinusitis38. Another comorbidity identified in our study was gastroesophageal reflux disease (GERD), which showed a notably higher prevalence among individuals with endometriosis (SMD = 0.13), con firming a previously recognized comorbidity28,29,31–33. A potential underlying mechanism sug- gests that estrogen may induce relaxation of the lower esophageal sphincter and thereby exacerbate gastroesophageal reflux symptoms39.A d d i t i o n a l l y , immune system dysregulation and in flammatory mediators, which play critical roles in both endometriosis and GERD, may contribute to this association 40. We also uncovered a significant increase in the prevalence of laryngitis and acute laryngitis (SMD = 0.13), an association that has not been pre- viously studied. This observation ma y be related to the established link between endometriosis and GERD. Research has shown that GERD and laryngopharyngeal reflux are closely connected conditions 41.D i g e s t i v e enzymes, such as activated pepsin, can in flame the laryngopharyngeal mucosa, resulting in edema, erythema, and laryngeal damage 42,43.A d d i - tionally, acid in the lower esophagus may cause repeated neural stimulation of the vagus nerve, which subsequently activates mast cells. The degranu- lation of mast cells can lead to laryngeal in flammation and associated symptoms 44. Table 3 | Comparison of the prevalence of finding of esophagus and disorders of the larynx in the endometriosis and matched control cohorts Condition Endometriosis Non-Endometriosis q-value SMD Finding of esophagus Finding of esophagus 1696 (20%) 1027 (12%) <0.001 0.2 –Disorder of esophagus 936 (11%) 490 (5.9%) <0.001 0.2 ––Candidiasis of the esophagus 108 (1.3%) 66 (0.8%) 0.004 0.05 –––Candidiasis of mouth and esophagus 108 (1.3%) 66 (0.8%) 0.004 0.05 ––Esophagitis 520 (6.3%) 264 (3.2%) <0.001 0.15 –––Gastro-esophageal reflux dis w/ esophagitis 399 (4.8%) 195 (2.3%) <0.001 0.13 ––Gastroesophageal reflux disease 376 (4.5%) 188 (2.3%) <0.001 0.13 ––Laceration of esophagus 21 (0.3%) 11 (0.1%) 0.14 0.027 –––Serosal tear of esophagus 21 (0.3%) 11 (0.1%) 0.14 0.027 ––Esophageal injury 21 (0.3%) 11 (0.1%) 0.14 0.027 –Pain in esophagus 618 (7.4%) 430 (5.2%) <0.001 0.093 ––Heartburn 618 (7.4%) 428 (5.2%) <0.001 0.094 –Finding of esophageal function 1034 (12%) 660 (8.0%) <0.001 0.15 ––Esophageal reflux finding 1034 (12%) 660 (8.0%) <0.001 0.15 –––Gastric reflux 182 (2.2%) 115 (1.4%) <0.001 0.061 –––Acid reflux 281 (3.4%) 138 (1.7%) <0.001 0.11 –Lesion of esophagus 41 (0.5%) 19 (0.2%) 0.013 0.044 –Swallowing symptoms 32 (0.4%) 24 (0.3%) 0.4 0.017 Disorder of the larynx Disorder of the larynx 795 (9.6%) 492 (5.9%) <0.001 0.14 –Laryngitis 754 (9.1%) 468 (5.6%) <0.001 0.13 ––Acute laryngitis 679 (8.2%) 412 (5.0%) <0.001 0.13 –––Acute laryngotracheitis 35 (0.4%) 20 (0.2%) 0.088 0.031 ––Infective laryngitis 58 (0.7%) 40 (0.5%) 0.13 0.028 –––Croup 44 (0.5%) 31 (0.4%) 0.2 0.023 ––Laryngotracheitis 103 (1.2%) 67 (0.8%) 0.015 0.043 –––Laryngotracheobronchitis 52 (0.6%) 34 (0.4%) 0.1 0.03 –Infection of larynx 58 (0.7%) 40 (0.5%) 0.13 0.028 See Table 2 for more info; meaningful differences (SMD > 0.1) are shown in bold face and borderline clinically meaningful (0.08 < SMD ≤ 0.1) entities are shown in italics. See Figs. S4 and S5 for a graphical presentation. https://doi.org/10.1038/s44294-025-00073-z Article npj Women's Health | (2025)3:30 4 Previous studies observed transient vocal changes such as loss of high tones, pitch uncertainty, and submucous hemorrhages in singers before and during menstruation, correlati ng these symptoms with hormonal fluctuations45. Given that endometriosis is also driven by hormonal changes, it is plausible that this condition could exacerbate menstrual-related vocal cord alterations, increasing the suscep tibility to laryngitis or vocal dys- function during critical phases of the menstrual cycle. Additionally, we discovered a signi ficant increase in various viral infections within the endometriosis cohort. These associations raise important questions about the role of the immune system in the patho- genesis of endometriosis. Given that e n d o m e t r i o s i si sc h a r a c t e r i z e db ya n altered immune response, the presence of these viral infections may further complicate the immune landscape, potentially exacerbating symptoms or influencing disease progression. Hormones, particularly estrogen, play a key role in modulating the immune system 46,47,i n fluencing immune responses by promoting the proliferation of immune cells and modulating cytokines production. This hormonal influence can lead to an altered immune environment, which may contribute to the pathogenesis of endometriosis. Hormonal fluc- tuations throughout the menstrual cycle can further impact the immune system, potentially exacerbating inflammation and pain in affected indi- viduals. A deeper understanding of the intricate relationship between hormones, immune function, viral infections and endometriosis could yield valuable insights into the disease ’s underlying mechanisms. Future research should focus on elucidating the immunological factors involved and exploring how viral infections may interact with the disease ’s inflammatory process. A final example is sciatic pain, which clinically manifests as limiting pain in the hip and gluteal region, radiating down the lower leg and foot, and may include tingling and numbness within the affected nerve ’s dermatome 48. Nerve involvement in endometriosis is uncommon, with the sacral plexus and sciatic nerve being the most frequently affected nerves49. Deep lesions within the nerve roots of the pelvis are rare, estimated to occur in approximately 0.1% of cases 50. Our study reveals a much higher prevalence of sciatica at 11%, suggest ing additional pathophysiological mechanisms beyond direct neural involvement. Cross-organ sensitization offers one possible mechanism, character- ized by the propagation of noxious i nputs from a diseased organ to an adjacent, healthy organ through sensory projections into dorsal root ganglia and subsequent spinal cord integra tion. This phenomenon may lead to sensations of pain or discomfort in unaffected organs, anatomically distant from the origin of the pathology51,52. Alternatively, prolonged exposure to noxious stimuli may trigger neural adaptations which eventually manifest in instability and a hypertonic contractile state within the muscles53. Several studies have demonstrated hypertonicity and spasms in the pelvic floor muscles, lower limb muscles, and the piriformis muscle among individuals with endometriosis 54–56. These muscular changes, particularly the hyperto- nicity and spasms of the piriformis muscle, are implicated in the onset of sciatica symptoms57,58. The coexistence of endometriosis with other conditions arises from pathophysiological mechanisms, ass u g g e s t e di nt h i ss t u d y ,a sw e l la sa complex interplay of environmental f actors and genetic predisposition. Complementary genetic studies 20,23,28,33,40 suggest biological links and potential shared mechanisms between endometriosis and various comor- bidities, including migraine, cancer, and depression. Exploring the diverse mechanisms underlying endometrio sis and its comorbidities not only deepens our understanding of endometriosis but may also pave the way for new therapeutic strategies, by repurposing existing drugs for related comorbidities 23,59,60 or developing novel targeted therapies, ensuring that treatment approaches address the root causes of comorbidities rather than just managing symptoms. Our study leverages a novel approach and a large routinely collected dataset to test links between endometriosis and all data-reported conditions. It has, however, several limitations.First, the community data resource we analyzed has incomplete informati on on surgeries and imaging, thus endometriosis diagnosis could not be ascertained. Note, however, that endometriosis misclassification is expected to reduce the difference between cases and controls. Second, conditions recorded during a typical multi-year Table 4 | Comparison of the prevalence of Viral diseases in the endometriosis and matched control cohorts; shown are conditions with SMD > 0.08 (meaningful and borderline meaningful) Condition Endometriosis Non-Endometriosis q-value SMD Viral disease Viral disease 4571 (55%) 3525 (42%) <0.001 0.3 –Herpesvirus infection 1922 (23%) 1381 (17%) <0.001 0.2 ––Disease due to Alphaherpesvirinae 1707 (21%) 1234 (15%) <0.001 0.15 –––Herpes simplex 814 (9.8%) 573 (6.9%) <0.001 0.11 –––Varicella-zoster virus infection 1055 (13%) 734 (8.8%) <0.001 0.12 –Viral infection of skin 1247 (15%) 936 (11%) <0.001 0.11 ––Verruca vulgaris of skin of lower extremity 695 (8.4%) 517 (6.2%) <0.001 0.083 –––Verruca plantaris 695 (8.4%) 517 (6.2%) <0.001 0.083 –Viral respiratory infection 1195 (14%) 873 (11%) <0.001 0.12 ––Viral upper respiratory tract infection 869 (10%) 677 (8.2%) <0.001 0.08 ––Influenza 385 (4.6%) 235 (2.8%) <0.001 0.1 –Viral infection of the digestive tract 976 (12%) 684 (8.2%) <0.001 0.12 ––Viral enteritis 348 (4.2%) 216 (2.6%) <0.001 0.088 –––Viral gastroenteritis 342 (4.1%) 207 (2.5%) <0.001 0.091 ––Viral gastritis 342 (4.1%) 207 (2.5%) <0.001 0.091 –Disease due to Papillomaviridae 1470 (18%) 1103 (13%) <0.001 0.12 ––Disease due to Papilloma virus 1470 (18%) 1103 (13%) <0.001 0.12 –––Human Papilloma virus infection 1470 (18%) 1103 (13%) <0.001 0.12 –Disease due to Orthomyxoviridae 388 (4.7%) 237 (2.9%) 0.1) are shown in bold face, borderline clinically meaningful entities are shown in italics. See Fig. S6 f or a graphical presentation. https://doi.org/10.1038/s44294-025-00073-z Article npj Women's Health | (2025)3:30 5 journey to endometriosis diagnosis m ay be recognized, in retrospect, as inaccurate, and as information is not typically removed from healthcare records, these spurious reports may inflate ourfindings. The long time-span of patient data alleviates, to some extent, this concern. Finally, the closer and more frequent monitoring of indivi duals with endometriosis by their healthcare providers, compared to th eir non-endometriosis counterparts (see Table 1 and S1), may lead to surveillance bias. Our analysis primarily focused on testable and treatable conditions, that are likely recorded for all suffering individuals. Future work will focus on speci fic comorbidities identified in this study and attempt to correct for these potential biases, e.g., by considering the relative timing of endometriosis and each comorbidity diagnosis. This study significantly advances our understanding of the complex relationships between endometriosis a nd its comorbidities, highlighting their contribution to the overall health burden of individuals with endo- metriosis. To the best of our knowledge, this is the first data-driven inves- tigation that uses a large datase t to explore these associations comprehensively, rather than focusing on specificp r e d efined conditions, based on existing knowledge. We focused on several comorbidities—known and novel —and explored potential mechanisms explaining the links between endometriosis and these comorbidities. Future studies are required to further investigate the hun dreds of comorbidities identi fied in this research, particularly to elucidate the underlying mechanisms linking these conditions and endometriosis. Importantly, our research emphasizes that endometriosis is a multi-system disease capable of affecting distant organ systems, potentially triggering a cascade of events that influence other areas of the body. This underscores the need for a comprehensive, multi- disciplinary approach to the diagn osis, management, and treatment of endometriosis.

Study design A retrospective, matched cohort study using routinely collected healthcare data, standardized to the Observati onal Medical Outcomes Partnership (OMOP) Common Data Model (CDM) 61. The study adheres to the STROBE (Strengthening the Reportingof Observational Studies in Epide- miology) guidelines62. Data sources IQVIA™ Medical Research Data (IMRD), formerly known as The Health Improvement Network (THIN), contains longitudinal non-identi fied patient electronic health records (EHR) collected from clinical systems of general practitioners (GP) in the U K (incorporating data from THIN, a Cegedim database). IMRD-THIN contains records of over 13 million patients, including demographic information, prescribed medication, vital signs, diagnoses, lab tests, and additional information such as lifestyle fac- tors, BMI, and vaccinations as recorded in GP practices. It has been pre- viously shown to be representative of the UK population, in terms of demographics and condition prevalence 63. Study population The endometriosis cohort includes females aged 14–50 years who had at least two endometriosis-related events—that is, a diagnosis of endometriosis or an endometriosis-related procedure (e.g., laparoscopic laser destruction of endometriosis)—in separate dates. Individuals in the endometriosis cohort were matched 1:1 to counterpar ts with no endometriosis-related event, by birth year and sex. The index date of each matched pair is the date of the first endometriosis-related event. Eligible pairs were required to have at least 1 year of prior observation on the index date. Additionally, as sensitivity analysis, we repeated the comorbidity prevalence comparison in a less stringent endometriosis cohort, requiring a single related event. Comorbidities We took a data-driven approach and analyzed the prevalence of all con- ditions recorded for individuals in th e cohorts of interest (rather than focusing on a pre-specified, limited set of comorbidities). Moreover, we used the hierarchical structure of Systematized Nomenclature of Medicine Clinical Terms (SNOMED-CT), a standard OMOP vocabulary 61,t og r o u p together conditions into broader categories64 (for example, Fig. 1 demon- strates the hierarchical structure of allergies disorders and its descendants). We then used OHDSI’s FeatureExtraction package65 to extract a feature for each medical condition (or condition group) recorded in the dataset, indi- cating whether individuals had a reco rd of the corresponding condition. Conditions with less than ten individuals, in either the endometriosis or the matched cohorts, were excluded from the analysis. Statistical analysis We computed the prevalence of each condition (or condition group) in the endometriosis and the matched non-endometriosis cohorts and assessed its statistical significance using Pearson’s Chi-squared test, corrected for multiple testing with Benjamini and Hochberg’s false discovery rate method 66 (q-value; values smaller than 0.05 were considered significant). The magnitude of the differences was reported using standardized mean difference (SMD), which compares the difference in prevalence, in units of the pooled standard deviation 67 (values greater than 0.1 were considered meaningful67,a n d0 . 0 8t o 0.1—borderline meaningful). We depict the results as hierarchical graphs, with nodes representing conditions (or condition groups) and edges con- necting ancestor (upper) and descendant (lower) entities. Data availability All comorbidity prevalence data are available in Supplementary Data 1. The data that support the findings of this study are available to license from IQVIA, at https://www.iqvia.com/solutions/real-world-evidence/real- world-data-and-insights. Received: 15 December 2024; Accepted: 14 April 2025; Published online: 13 May 2025

1. Parente Barbosa, C., Bentes De Souza, A. M., Bianco, B. & Christofolini, D. M. The effect of hormones on endometriosis development. Minerva Ginecol. 63, 375–386 (2011). 2. Wei, Y., Liang, Y., Lin, H., Dai, Y. & Yao, S. Autonomic nervous system and inflammation interaction in endometriosis-associated pain. J. Neuroinflammation 17, 80 (2020). 3. Taylor, H. S., Kotlyar, A. M. & Flores, V. A. Endometriosis is a chronic systemic disease: clinical challenges and novel innovations. Lancet 397, 839–852 (2021). 4. Had field, R., Mardon, H., Barlow, D. & Kennedy, S. Delay in the diagnosis of endometriosis: a survey of women from the USA and the UK. Hum. Reprod. 11, 878–880 (1996). 5. Simoens, S. et al. The burden of endometriosis: costs and quality of life of women with endometriosis and treated in referral centres.Hum. Reprod. 27, 1292–1299 (2012). 6. Surrey, E., Soliman, A. M., Trenz, H., Blauer-Peterson, C. & Sluis, A. Impact of endometriosis diagnostic delays on healthcare resource utilization and costs. Adv. Ther. 37, 1087–1099 (2020). 7. Sinaii, N., Cleary, S. D., Ballweg, M. L., Nieman, L. K. & Stratton, P. High rates of autoimmune and endocrine disorders, fibromyalgia, chronic fatigue syndrome and atopic diseases among women with endometriosis: a survey analysis.Hum. Reprod.17, 2715–2724 (2002). 8. Yang, M. H. et al. Women with endometriosis are more likely to suffer from migraines: a population-based study.PLoS One7, e33941 (2012). 9. Laganà, A. S. et al. Anxiety and depression in patients with endometriosis: impact and management challenges. Int J. Women’s. Health 9, 323–330 (2017). 10. Ramin-Wright, A. et al. Fatigue —a symptom in endometriosis. Hum. Reprod. 33, 1459–1465 (2018). 11. Kvaskoff, M. et al. Endometriosis: a high-risk population for major chronic diseases?. Hum. Reprod. Update 21, 500–516 (2015). https://doi.org/10.1038/s44294-025-00073-z Article npj Women's Health | (2025)3:30 6 12. Parazzini, F., Esposito, G., Tozzi, L., Noli, S. & Bianchi, S. Epidemiology of endometriosis and its comorbidities. Eur. J. Obstet. Gynecol. Reprod. Biol. 209,3 –7 (2017). 13. Surrey, E. S. et al. Risk of developing comorbidities among women with endometriosis: a retrospective matched cohort study. J. Women’s. Health 27, 1114–1123 (2018). 14. Vargas, E., Aghajanova, L., Gemzell-Danielsson, K., Altmäe, S. & Esteban, F. J. Cross-disorder analysis of endometriosis and its comorbid diseases reveals shared genes and molecular pathways and proposes putative biomarkers of endometriosis. Reprod. Biomed. Online 40, 305–318 (2020). 15. Shafrir, A. L. et al. Co-occurrence of immune-mediated conditions and endometriosis among adolescents and adult women. Am. J. Reprod. Immunol. 86, e13404 (2021). 16. Yoshii, E., Yamana, H., Ono, S., Matsui, H. & Yasunaga, H. Association between allergic or autoimmune diseases and incidence of endometriosis: a nested case-control study using a health insurance claims database. Am. J. Reprod. Immunol. 86, e13486 (2021). 17. Shigesi, N. et al. The association between endometriosis and autoimmune diseases: a systematic review and meta-analysis. Hum. Reprod. Update 25, 486–503 (2019). 18. Sun, Y. H., Leong, P. Y., Huang, J. Y. & Wei, J. C. C. Increased risk of being diagnosed with endometriosis in patients with Systemic lupus erythematosus: a population-based cohort study in Taiwan. Sci. Rep. 12, 13336 (2022). 19. Ferrari-Souza, J. P. et al. Endometriosis and systemic lupus erythematosus: systematic review and meta-analysis. Reprod. Sci. 30, 997–1005 (2023). 20. Vazgiourakis, V. M. et al. Association of endometriosis with cardiovascular disease: genetic aspects (review). Int J. Mol. Med. 51, 29 (2023). 21. Matalliotakis, I., Cakmak, H., Matalliotakis, M., Kappou, D. & Arici, A. High rate of allergies among women with endometriosis. J. Obstet. Gynaecol. 32, 291–293 (2012). 22. Chen, S. & Liu, J. Associations between endometriosis and allergy- related diseases as well as different specific immunoglobulin E allergy: a cross-sectional study.J. Obstet. Gynaecol. Res.49, 665–674 (2023). 23. Fattori, V. et al. Nociceptor-to-macrophage communication through CGRP/RAMP1 signaling drives endometriosis-associated pain and lesion growth in mice.Sci. Transl. Med.16, eadk8230–eadk8243 (2024). 24. Bonavina, G. & Taylor, H. S. Endometriosis-associated infertility: from pathophysiology to tailored treatment. Front Endocrinol.13, 1020827 (2022). 25. Bhurke, A. V. et al. Clinical characteristics and surgical management of endometriosis-associated infertility: a multicenter prospective cohort study. Int J. Gynaecol. Obstet. 159,8 6 –96 (2022). 26. Tietjen, G. E. et al. Endometriosis is associated with prevalence of comorbid conditions in migraine. Headache 47, 1069–1078 (2007). 27. Sarria-Santamera, A. et al. A novel classi fication of endometriosis based on clusters of comorbidities. Biomedicines 11, 2448 (2023). 28. McGrath, I. M. et al. Insights from Mendelian randomization and genetic correlation analyses into the relationship between endometriosis and its comorbidities. Hum. Reprod. Update 29, 655–674 (2023). 29. McGrath, I. M. et al. Genomic charactarisation of the overlap of endometriosis with 76 comorbidities identifies pleiotropic and causal mechanisms underlying disease risk. Hum. Genet. 142, 1345–1360 (2023). 30. Lamb, K. & Nichols, T. R. Endometriosis: a comparison of associated disease histories. Am. J. Prev. Med. 2, 324–329 (1986). 31. Tore, U. et al. Diagnosis of endometriosis based on comorbidities: a machine learning approach. Biomedicines 11, 3015 (2023). 32. Ek, M. et al. Gastrointestinal symptoms among endometriosis patients -a case cohort study. BMC Women’s. Health 15, 59 (2015). 33. Yang, F. et al. Evidence of shared genetic factors in the etiology of gastrointestinal disorders and endometriosis and clinical implications for disease management. Cell Rep. Med. 4, 101250 (2023). 34. Häuser, W. et al. Fibromyalgia. Nat. Rev. Dis. Prim. 1, 15022 (2015). 35. Sugamata, M., Ihara, T. & Uchiide, I. Increase of activated mast cells in human endometriosis. Am. J. Reprod. Immunol. 53, 120–125 (2005). 36. Gete, D. G., Doust, J., Mortlock, S., Montgomery, G. & Mishra, G. D. Associations between endometriosis and common symptoms: findings from the Australian Longitudinal Study on Women ’s Health. Am. J. Obstet. Gynecol. 229, 536.e1–536.e20 (2023). 37. Feng, C. H., Miller, M. D. & Simon, R. A. The united allergic airway: connections between allergic rhinitis, asthma, and chronic sinusitis. Am. J. Rhinol. Allergy 26, 187–190 (2012). 38. Newman, L. J., Platts-Mills, T. A., Phillips, C. D., Hazen, K. C. & Gross, C. W. Chronic sinusitis. Relationship of computed tomographic findings to allergy, asthma, and eosinophilia. JAMA 271, 363–367 (1994). 39. Nordenstedt, H. et al. Postmenopausal hormone therapy as a risk factor for gastroesophageal re flux symptoms among female twins. Gastroenterology 134, 921–928 (2008). 40. Adewuyi, E. O. et al. Genetic analysis of endometriosis and depression identifies shared loci and implicates causal links with gastric mucosa abnormality. Hum. Genet. 140, 529–552 (2021). 41. Wu, Y. et al. The relationship between gastroesophageal re flux disease and laryngopharyngeal re flux based on pH monitoring. Ear Nose Throat J. 100, 249–253 (2021). 42. Koufman, J. A. The otolaryngologic manifestations of gastroesophageal reflux disease (GERD): a clinical investigation of 225 patients using ambulatory 24-hour pH monitoring and an experimental investigation of the role of acid and pepsin in the development of laryngeal injury. Laryngoscope 101,1 –78 (1991). 43. Dworkin, J. P. Laryngitis: types, causes, and treatments. Otolaryngol. Clin. North Am. 41, 419–436 (2008). 44. Theodoropoulos, D. S. et al. Prevalence of upper respiratory symptoms in patients with symptomatic gastroesophageal re flux disease. Am. J. Respir. Crit. Care Med. 164,7 2–76 (2001). 45. Brodnitz, F. S. Hormones and the human voice. Bull. N. Y Acad. Med. 47, 183–191 (1971). 46. Chakraborty, B. et al. Estrogen receptor signaling in the immune system. Endocr. Rev. 44, 117–141 (2023). 47. Sciarra, F. et al. Gender-speci fic impact of sex hormones on the immune system. Int. J. Mol. Sci. 24, 6302 (2023). 48. Thiel P., Kobylianskii A., McGrattan M., Lemos N. Entrapped by pain: the diagnosis and management of endometriosis affecting somatic nerves. Best Pract. Res. Clin. Obstet. Gynaecol . https://doi.org/10. 1016/j.bpobgyn.2024.102502 (2024). 49. Lukac, S. et al. Extragenital endometriosis in the differential diagnosis of non-gynecological diseases.Dtsch. Ärzteblatt Int.109,3 6 1–367 (2022). 50. Zamurovic, M. et al. Isolated deep in filtrating endometriosis of the sciatic nerve: a case report and overview of the literature.Medicina 59, 2161–2169 (2023). 51. Song, S. Y. et al. Endometriosis-related chronic pelvic pain. Biomedicines 11, 2868 (2023). 52. Qiao, L. Y. & Tiwari, N. Spinal neuron-glia-immune interaction in cross-organ sensitization. Am. J. Physiol. Gastrointest. Liver Physiol. 319, G748–G760 (2020). 53. Butrick, C. W. Pathophysiology of pelvic floor hypertonic disorders. Obstet. Gynecol. Clin. North Am. 36, 699–705 (2009). 54. Fraga, M. V., Oliveira Brito, L. G., Yela, D. A., Mira, T. A. & Benetti- Pinto, C. L. Pelvic floor muscle dysfunctions in women with deep infiltrative endometriosis: an underestimated association. Int J. Clin. Pract. 75, e14350–e14355 (2021). 55. Da Silva, J. P. et al. Sensory and muscular functions of the pelvic floor in women with endometriosis –cross-sectional study. Arch. Gynecol. Obstet. 308, 163–170 (2023). https://doi.org/10.1038/s44294-025-00073-z Article npj Women's Health | (2025)3:30 7 56. Dos Bispo, A. P. S. et al. Assessment of pelvic floor muscles in women with deep endometriosis.Arch. Gynecol. Obstet.294, 519–523 (2016). 57. Filler, A. G. et al. Sciatica of nondisc origin and piriformis syndrome: diagnosis by magnetic resonance neurography and interventional magnetic resonance imaging with outcome study of resulting treatment. J. Neurosurg. Spine 2,9 9–115 (2005). 58. Fishman, L. M., Anderson, C. & Rosner, B. BOTOX and Physical Therapy in the Treatment of Piriformis Syndrome. Am. J. Phys. Med. Rehabilit. 81, 936–942 (2002). 59. Ruiz, A. et al. Effect of hydroxychloroquine and characterization of autophagy in a mouse model of endometriosis. Cell Death Dis. 7, e2059–e2076 (2016). 60. Chen, F. Y. et al. Hydroxychloroquine might reduce risk of incident endometriosis in patients with systemic lupus erythematosus: a retrospective population-based cohort study. Lupus 30, 1609–1616 (2021). 61. OHDSI. The Book of OHDSI: Observational Health Data Sciences and Informatics. OHDSI; 2019. https://books.google.co.il/books?id= JxpnzQEACAAJ 62. Elm, E. et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J. Clin. Epidemiol. 61, 344–349 (2008). 63. Blak, B. T., Thompson, M., Dattani, H. & Bourke, A. Generalisability of The Health Improvement Network (THIN) database: demographics, chronic disease prevalence and mortality rates. Inf. Prim. Care. 19, 251–255 (2011). 64. Willett, D. L. et al. SNOMED CT concept hierarchies for sharing definitions of clinical conditions using electronic health record data. Appl Clin. Inform. 9, 667–682 (2018). 65. Schuemie, M. et al. FeatureExtraction: Generating Features for a Cohort. https://github.com/OHDSI/FeatureExtraction (2024). 66. Benjamini, Y. & Hochberg, Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J. R. Stat. Soc. Ser. B57, 289–300 (1995). 67. Austin, P. C. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivar. Behav. Res. 46, 399–424 (2011). Author contributions T.Z. and C.Y. developed the methodology, analyzed, and interpreted the data. M.E. and V.K.M. provided the clinical interpretation of the results. All authors were major contributors to writing the manuscript. All authors read and approved the final manuscript. Competing interests The authors declare no competing interests. Ethical approval Use of IQVIA Medical Research Data (IMRD) was approved by the NHS London–Southeast Research Ethics Committee (REC reference: 18/LO/ 0441); in accordance with this approval, the study protocol was reviewed and approved by an independent Scientific Review Committee (SRC) of IQVIA Inc. (reference number: 23SRC021). Patients’ written informed consent was waived by the London–Southeast Research Ethics Committee as the study used data from anonymized electronic health records. Additional information Supplementary informationThe online version contains supplementary material available at https://doi.org/10.1038/s44294-025-00073-z. Correspondenceand requests for materials should be addressed to Tamar Zelovich. Reprints and permissions informationis available at http://www.nature.com/reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modi fied the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party

in this article are included in the article ’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by- nc-nd/4.0/ . © The Author(s) 2025, corrected publication 2025 https://doi.org/10.1038/s44294-025-00073-z Article npj Women's Health | (2025)3:30 8