Co‐occurrence of immune‐mediated conditions and endometriosis among adolescents and adult women

The participants in this study included 551 females with surgically-confirmed endometriosis and 652 females without known endometriosis (control group). Among cases and controls, 71.6% were aged <26 years old with a median age for cases of 19 years (range: 12–51) and for controls of 24 (range: 7–55). The majority of participants had a BMI in the normal weight range (cases: 61.3%, controls: 66.4%) and were of white race (cases: 90.2%, controls: 71.9%), although there was a higher proportion of Black (cases: 2.6%, controls: 6.3%) and Asian (cases: 0.5%, controls: 14.0%) races among control participants ( Table 1 ). As expected due to clinical history, 87.8% of participants with endometriosis had ever taken oral contraceptives compared with 65.4% of those never diagnosed with endometriosis. A family history of endometriosis in a first-degree relative was more common among participants with endometriosis (33.2%) compared to participants without (4.8%). Among the participants with endometriosis, 96% reported that pain symptoms led to their diagnosis of endometriosis while the remaining 4% of participants reported infertility or other reasons for their diagnostic surgery. A higher proportion of cases compared to controls reported any autoimmune and/or inflammatory disease (4.5% vs. 3.4%), chronic pain and/or fatigue disorder (3.3% vs. 1.1%), allergies (40.8% vs. 26.4%), asthma (27.6% vs. 17.6%), and previous mononucleosis infection (14.2% vs. 9.4%) ( Table 2 ). Conversely, a higher proportion of control participants reported endocrine disorders (7.1%) and eczema (13.3%) compared to cases (4.0% and 9.8%, respectively). Participants with an autoimmune and/or inflammatory disease had a greater likelihood of also having surgically diagnosed endometriosis compared to participants without an autoimmune and/or inflammatory disease (OR: 1.87; 95% CI: 0.92–3.80; Table 2 ). Although limited by small sample sizes, this association appeared to be driven by an increased odds of surgically-confirmed endometriosis among participants with SLE (OR: 5.39; 95% CI: 0.54–54.2), RA (OR: 2.14; 95% CI: 0.54–8.39), and IBD (OR: 6.78; 95% CI: 1.30–35.5; Supplemental Table 1 ). Additionally, participants with fibromyalgia and/or CFS had an increased likelihood of co-occurring endometriosis (OR: 5.81; 95% CI: 1.89–17.9) compared to participants without fibromyalgia and/or CFS, with increased odds of endometriosis for both conditions analyzed separately (fibromyalgia OR: 6.18; 95% CI: 1.61–23.8 and CFS OR: 21.9; 95% CI: 3.06–156). Participants with allergies were 76% more likely to also have surgically diagnosed endometriosis compared to participants without allergies (OR: 1.76; 95% CI: 1.32–2.36). When analyzed by type of allergy, we observed a statistically significant increased odds of endometriosis for each allergy type except for animal allergies ( Supplemental Table 2 ). Additionally, we observed an increased odds of endometriosis among participants with asthma as opposed to those without (OR: 1.35; 95% CI: 0.97–1.88) and among participants with previous mononucleosis infection compared to those without (OR: 1.75; 95% CI: 1.14–2.68). Conversely, participants with eczema had a 32% decreased likelihood of endometriosis compared to participants without eczema (OR: 0.68; 95% CI: 0.44–1.04). Finally, we observed no associations for endocrine disorders and endometriosis (OR: 0.85; 95% CI: 0.46–1.58). When analyses were restricted to participants <30 years old, the association for autoimmune/inflammatory conditions was attenuated (OR: 1.42; 95% CI: 0.60–3.33) whereas the associations for CFS/fibromyalgia (OR: 6.04; 95% CI: 1.55–23.5), allergies (OR: 1.95; 95% CI: 1.38–2.76), asthma (OR: 1.29; 95% CI: 0.88–1.89), eczema (OR: 0.63; 95% CI: 0.38–1.04) and previous mononucleosis infection (OR: 1.98; 95% CI: 1.19–3.30) were either similar or strengthened (data not shown). Analysis of the cumulative number of immune-mediated conditions showed a 29% increased odds of endometriosis with each additional immune-mediated condition (OR: 1.29, 95% CI: 1.13–1.48; p-trend=0.0002; Figure 1 and Supplemental Table 3 ). Compared to participants with no immune-mediated conditions, participants with one, two, or ≥ three immune-mediated conditions had an increased likelihood of also having endometriosis, with the suggestion of a threshold at two or more immune-mediated conditions (one condition vs. none OR: 1.42, 95% CI: 1.04–1.96; two conditions vs. none OR: 2.13, 95% CI: 1.43–3.19; three or more conditions vs. none OR: 1.75, 95% CI: 1.04–2.93). When analyses were restricted to participants <30 years old, the results were very similar (p-trend<0.001; Supplemental Figure 1 ). Family history of UC or SLE were both associated with increased odds of being diagnosed with endometriosis (positive family history of UC OR: 1.65 (95% CI: 0.91–2.98) and positive family history of SLE OR: 1.67 (95% CI: 0.80–3.50; Table 3 ), however these analyses were based on small sample sizes and therefore had limited power. We observed no association between family history of CD or MS and an endometriosis diagnosis.

This cross-sectional study was conducted within an ongoing longitudinal cohort study, The Women’s Health Study: From Adolescence to Adulthood (A2A) ( 7 ). The A2A is a study of adolescents and women oversampled for those surgically diagnosed with endometriosis. Participants were enrolled from 2012 to 2018. Endometriosis cases were enrolled from Boston Children’s Hospital (BCH) and Brigham and Women’s Hospital (BWH) and were eligible if they were (i) female; (ii) aged 7–55 years; and (iii) had a surgical diagnosis of endometriosis at one of the two participating hospitals or had a prior surgical diagnosis elsewhere but were receiving follow-up treatment at one of the two hospitals. Controls were females aged 7–55 years, recruited from clinics at BCH and BWH and from the local Boston community through local advertisements, online postings, and word of mouth. The control participants had never received a surgical diagnosis of endometriosis. The study was approved by the BCH Institutional Review Board on behalf of both BCH and BWH. Informed consent was obtained, with both parental consent and participant assent for girls less than 18 years of age at enrollment. Participants completed an extensive baseline questionnaire to assess lifestyle, reproductive factors, pain levels, and medication use that expands upon the World Endometriosis Research Foundation (WERF) Endometriosis Phenome and Biobanking Harmonization Project (EPHect) standard clinical questionnaire ( 8 ). REDCap electronic data capture tools were used to manage survey data ( 9 ). Of the 1,549 participants enrolled in the A2A cohort, the analytic sample for this study was restricted to participants who completed the medical history section of the baseline questionnaire – 551 surgically-diagnosed endometriosis cases and 652 controls (70.4% and 85.1%, respective response rates). Participants were excluded if they did not complete the baseline questionnaire (n=159 cases and 78 controls) or did not complete the medical history section of the questionnaire (n=73 cases and 36 controls). Demographic characteristics did not differ substantially between cases included in the analytic sample and those who were not included. On average, control participants who did not complete the medical history section were more likely to be obese (19% vs. 10%) and ever smokers (13% vs. 7%) compared to controls who were included in the analyses. On the baseline questionnaire, participants self-reported any physician diagnosis of autoimmune and inflammatory diseases (SLE, MS, RA, IBD [Crohn’s disease (CD), ulcerative colitis (UC)], Sjogren’s syndrome, and psoriasis), endocrine disorders (hyperthyroidism, hypothyroidism, and diabetes mellitus), chronic pain and fatigue syndromes (fibromyalgia and chronic fatigue syndrome (CFS)), and atopic diseases (asthma, eczema, allergies). Participants specified the type(s) of allergies with which they were diagnosed (grass/pollen/mold, food, latex, drug, animal). Additionally, participants reported any previous mononucleosis infection. To assess cumulative immune dysfunction burden among the participants, we summed the number of immune-mediated conditions listed above, with the exception of previous mononucleosis, for each participant. Participants also reported family history of endometriosis (female relative only), CD, UC, SLE, and MS. Descriptive characteristics were collected including age (continuous), race (white, Black, Asian, other/uknown race), smoking history (never, ever), parity (nulliparous, parous), age at menarche (years), and use of oral contraceptives (never, ever). Additionally, participants reported any regular use of analgesic medications (i.e., used at least once a week for a period of three months or longer) which was categorized as never, <2 days of use per week, or ≥2 days of use per week. Body mass index (BMI) was calculated from self-reported current weight and height. For women aged ≥20 years, participants were categorized as underweight (BMI, <18.5 kg/m 2 ), normal weight (BMI, 18.5–24.9 kg/m 2 ), overweight (BMI, 25–29.9 kg/m 2 ), or obese (BMI, ≥30 kg/m 2 ) per World Health Organization criteria ( 10 ). For adolescents, the age-and gender-specific BMI Z-score was calculated and categorized as underweight (Z-score, ≤−2), normal weight (Z-score, >−2 to 2) ( 11 ). We calculated age-adjusted and multivariable-adjusted odds ratios (OR) and 95% confidence intervals (CI) using logistic regression to investigate the association between (i) each of the immune-mediated conditions and (ii) the cumulative number of immune-mediated conditions with the odds of surgically diagnosed endometriosis. In age-adjusted models, we adjusted for age at baseline (continuous). In multivariable-adjusted models, we further adjusted for age at menarche (continuous), BMI (underweight/normal-weight/overweight/obese), oral contraceptive use (never/ever), and family history of endometriosis in a first-degree relative (yes/no). Two participants missing the response to the question regarding oral contraceptive use were set to never use, and one participant missing age at menarche was set to the cohort median ( 12 ). Participants who had not started menstruating at the time of survey completion (cases=5, controls=14) were included in a missing indicator for age at menarche adjustment. The Wald statistic was calculated to test for a linear trend across the number of reported immune-mediated conditions and the odds of endometriosis (p-trend). In sensitivity analyses, we restricted the above analyses to participants aged <30 years at study enrollment. To examine the association of family history of CD, UC, SLE, or MS with the odds of endometriosis, we calculated age-adjusted and multivariable-adjusted ORs and 95% CIs using logistic regression adjusting for the same variables as above. Data were analyzed with SAS (version 9.4; SAS Institute Inc., Cary, NC). All p-values are two-sided.

We investigated the associations between immune-mediated conditions and endometriosis in a cross-sectional study nested within a longitudinal cohort at baseline. Our goal was to better define the clinical, multi-systemic burden experienced by endometriosis patients, particularly adolescents and young women. We observed that participants with an autoimmune and/or inflammatory disease, fibromyalgia and/or CFS, allergies, asthma, or previous mononucleosis infection were more likely to also have surgically diagnosed endometriosis compared to participants without these conditions, although not all of these associations were statistically significant due to small numbers. Conversely, participants with eczema were less likely to have endometriosis compared to those without eczema. We observed a significant positive trend with increasing number of immune-mediated conditions and co-occurring endometriosis, highlighting the increased clinical burden experienced by women with endometriosis. When analyses were restricted to participants aged <30 years at baseline, we saw little change in the effect estimates and those that did attenuate, i.e. autoimmune and inflammatory conditions, are normally diagnosed later in life. Although based on a small number of participants with SLE, IBD, or RA, we observed similar results to previous studies of a positive association between SLE, IBD, or RA and endometriosis ( 12 – 16 ). A recent meta-analysis reported an increased likelihood of multiple autoimmune diseases, including SLE, RA, MS, and Sjogren’s syndrome, among women with endometriosis compared to women without endometriosis ( 4 ). However, it is important to note that the majority of previous studies have compared the risk of developing an autoimmune disease among women with and without endometriosis, whereas the current study investigated the co-occurrence of these diseases. Limited research has assessed IBD and endometriosis ( 16 ). A cohort study within the Danish National Hospital Register observed an increased risk of IBD for women with endometriosis compared to women without; however, this study lacked control for confounding factors ( 17 ). Additionally, a survey of the Endometriosis Association members from the USA and Canada reported a higher prevalence of SLE, RA, MS, and Sjogren’s syndrome among endometriosis patients compared to the general female population ( 6 ). Due to the younger age range of the current study population, we were unable to calculate odds ratios for MS or Sjogren’s syndrome, suggesting that potential associations between the co-occurrence of endometriosis and MS or Sjogren’s syndrome occur later in life. The current results add to the growing literature of a possible link between endometriosis and autoimmune conditions even among a younger population. Women with endometriosis have immunologic similarities to people with autoimmune disorders including polyclonal B cell activation, abnormalities in T and B cell functions, and increased apoptosis and tissue damage ( 5 ). However, how these conditions are associated remains unclear. Future research should focus on the underlying biologic mechanisms potentially linking autoimmune disorders and endometriosis. Similar to our findings of an increased likelihood of endometriosis among women with fibromyalgia and CFS, the Endometriosis Association survey noted a prevalence of 5,897 fibromyalgia cases per 100,000 women with endometriosis compared to an estimated prevalence of 3,400 fibromyalgia cases per 100,000 women in the general USA population ( 6 ). However, the Endometriosis Association survey was conducted among women with moderate to severe endometriosis pain symptoms who were part of a membership dues-paying population and thus are not representative of all women with endometriosis. Additionally, confounding factors were not considered within the analyses. While the etiology and pathogenesis of fibromyalgia and CFS are still not clear, both conditions appear to involve a pro-inflammatory state with alterations to immune molecules including increased levels of pro-inflammatory cytokines and dysfunctional natural killer cells similar to the immune dysregulation observed in endometriosis ( 18 – 21 ). Larger studies are needed to confirm our results and to understand the biological reasons for this possible association. Consistent with our findings, previous studies have reported a greater odds of having endometriosis among participants with allergies ( 6 , 22 , 23 ). A 2014 systematic review identified four studies investigating allergies and endometriosis, with all four studies reporting an increased risk of allergic disease among women with endometriosis ( 23 ). However, none of these studies controlled for confounding factors. Conversely, conflicting results have been reported for the association between asthma and endometriosis, with most studies reporting a positive association ( 6 , 22 , 24 ), one study reporting no association ( 25 ), and one study reporting an inverse association ( 26 ). These contradicting results may be due to differences in controlling for confounding factors and differences in the population from which controls were selected. For eczema, two older case-control studies noted an increased odds of eczema among women with endometriosis compared to controls ( 27 , 28 ). Similarly, results from the Endometriosis Association survey suggested an increased prevalence of eczema among endometriosis patients with a co-occurring diagnosis of fibromyalgia or CFS compared to endometriosis patients without these conditions ( 6 ). These results contrast with our findings of a decreased odds of endometriosis among participants with eczema. These differing findings could be due to differences in age distributions of the studied populations leading to differences in comorbidity patterns. Allergies, asthma, and eczema have all been associated with a Th2-type inflammatory response to allergic stimuli, with interleukin (IL)-4 expression important for the development of the Th2 immune response ( 29 , 30 ). Women with endometriosis tend to shift from a Th1 to Th2 immune response, and elevated levels of IL-4 have been observed in endometriosis patients ( 31 – 33 ). How this immune shift may impact on the co-occurrence of endometriosis and atopic diseases, including the timing of disease onset, deserves further investigation. The association between endometriosis and mononucleosis has not been well-studied. While we observed an increased odds of endometriosis among participants with a previous mononucleosis infection, Gemmill et al (2010) observed a lower prevalence of previous mononucleosis among women with endometriosis in the Endometriosis Association survey compared to the estimated general population prevalence ( 34 ). Conversely, a small case-control study observed an increased odds of previous mononucleosis infection among endometriosis cases compared to controls ( 27 ). Prior mononucleosis infection has been associated with the development of CFS, particularly among adolescents ( 35 ); therefore, it may be that the same mechanisms that potentially link mononucleosis and CFS also link mononucleosis and endometriosis. The conflicting results in the literature indicate that further research is needed to understand these relationships. Given that hereditary associations are observed for endometriosis and autoimmune diseases ( 1 , 36 ), we sought to investigate the association between a family history of CD, UC, SLE, or MS in relation to endometriosis. Although limited by small sample sizes, our data suggested an increased risk of surgically diagnosed endometriosis among participants with a family history of UC or SLE compared to participants without. Larger studies are needed to confirm our results and to investigate whether these associations are due to shared immune profiles between autoimmune diseases and endometriosis or through a different biologic mechanism. Our study had several limitations that were largely related to the natural history of the conditions of interest. The majority of autoimmune diseases are diagnosed in middle age or later, beyond the median age of our study population. Thus, there were small numbers of participants diagnosed with several diseases of interest; however, we had a sufficient sample size to assess atopic diseases. Additionally, these analyses were cross-sectional. As such, we could not assess the causative relationship between immune-mediated conditions and endometriosis (i.e., which preceded which). There is also potential for detection bias as participants with immune-mediated conditions may have more frequent access to the healthcare system compared to those without. Further, there were differences in the prevalence of obesity and smoking status between participants included in the analyses and participants who did not complete the medical history questions. As these are hallmarks of poorer health status, it is possible that we may be missing some controls with co-morbid disease and thus have biased our results away from the null. Finally, there is the potential for misclassification of endometriosis status among our controls as some controls may have asymptomatic or yet undiagnosed disease. However, this would lead to non-differential misclassification that would bias our results towards the null. Additionally, the likely community prevalence of undiagnosed endometriosis is <2%,( 37 ) and the characteristics of this small proportion of undiagnosed cases will be diluted among the true endometriosis-free comparison women. Our study also had several strengths. These data are drawn from the largest adolescent and young adult population with surgically-confirmed endometriosis to date, the cohort of the Women’s Health Study: Adolescence to Adulthood. Furthermore, participants completed comprehensive clinical questionnaires, including demographic, personal medical, and family medical history. With this thorough information, we are able to address several key confounders in our multivariable-adjusted analysis, including BMI, family history of endometriosis, age at menarche, and prior use of oral contraceptives.

Our results demonstrate that the co-occurrence of endometriosis and conditions characterized by immune dysfunction occurs even in sample populations younger than previously studied. Unlike previous studies investigating the risk of autoimmune or inflammatory disease in women with endometriosis, we demonstrated that endometriosis was more likely in women with underlying autoimmune and/or inflammatory diseases, fibromyalgia and/or CFS, allergies, asthma, and previous mononucleosis infection, and less likely in women with eczema. Importantly, the odds of endometriosis increased with each additional co-occurring immune-mediated disease, suggesting that there may be pathophysiologic pathways in common between endometriosis and immune-mediated conditions. Clinically, these results highlight the need to consider endometriosis as a potential co-morbidity for adolescents and women with immune-mediated conditions. Future research should investigate the biologic underpinnings of these co-occurring conditions and the most efficacious treatment options for endometriosis patients with immune-mediated comorbidities

Endometriosis, characterized by the presence of endometrial-like glands and stroma outside of the uterus, affects approximately 10% of women of reproductive age ( 1 , 2 ). Endometriosis can lead to debilitating pelvic pain, diminished quality of life, increased school and work absenteeism, and substantial healthcare costs ( 1 ). Endometriosis is characterized by immune dysregulation, including increased levels of neutrophils and macrophages and reduced cytotoxicity of natural killer cells in the peritoneum, which helps the endometriotic implants to grow and survive within the peritoneal cavity ( 3 ). These immune changes and the presence of autoantibodies among some women with endometriosis has led researchers to investigate the associations between endometriosis and autoimmune/inflammatory diseases. Women with endometriosis may be more likely to have co-occurring or subsequent autoimmune and inflammatory diseases compared to women without endometriosis ( 4 – 6 ). A recent meta-analysis noted clinically and statistically significant associations between endometriosis and various autoimmune diseases including systemic lupus erythematosus (SLE), Sjogren’s syndrome, rheumatoid arthritis (RA), celiac disease, multiple sclerosis (MS), and inflammatory bowel disease (IBD) ( 4 ). However, the authors of the meta-analysis paper graded most of the studies as “low quality” in design and most of the studies included adult women only. Research on autoimmune and inflammatory conditions most frequently present among younger endometriosis patients would fill an important gap and help improve earlier detection and efficacious treatment options for this pervasive condition. The objective of this study was to investigate immune-mediated conditions including allergies, asthma, diabetes, eczema, SLE, RA, and MS in relation to endometriosis among a younger population (72% aged <26 years).

Supplemental Figure 1 : Association between the cumulative number of immune-mediated conditions and surgically-confirmed endometriosis among participants aged <30 years old. Immune-mediated conditions included systemic lupus erythematosus, multiple sclerosis, rheumatoid arthritis, Sjogren’s syndrome, psoriasis, diabetes, hypothyroidism, hyperthyroidism, fibromyalgia, chronic fatigue syndrome, allergies, asthma, and eczema. Graph shows odds ratios and 95% confidence intervals adjusted for age (continuous), BMI (underweight/normal-weight/overweight/obese), first-degree family history of endometriosis (yes/no), age at menarche (continuous), oral contraceptive use (never/ever)