This chapter should be cited as follows:
Kvaskoff M, Glob. libr. women's med.,
ISSN: 1756-2228; DOI 10.3843/GLOWM.417713

The Continuous Textbook of Women’s Medicine Series – Gynecology Module

Volume 3

Endometriosis

Volume Editors: Professsor Andrew Horne, University of Edinburgh, UK
Dr Lucy Whitaker, University of Edinburgh, UK

Chapter

Endometriosis Comorbidities

First published: July 2024

Study Assessment Option

By completing 4 multiple-choice questions (randomly selected) after studying this chapter readers can qualify for Continuing Professional Development awards from FIGO plus a Study Completion Certificate from GLOWM
See end of chapter for details

INTRODUCTION

Endometriosis has been associated with several other chronic diseases in recent decades, most commonly cancer, autoimmune diseases, cardiovascular diseases, asthma or atopic diseases, and other chronic pain conditions such as migraine.¹ Describing the links between endometriosis and its comorbidities, and understanding the mechanisms underlying them, is critical – first, because it may enhance our knowledge of the causes or consequences of endometriosis, which is currently limited; second, because given the occurrence of endometriosis in about 10% of women from menarche to menopause, this knowledge may have significant public health impact through the development of targeted prevention and early detection for these chronic diseases.² In this chapter, we summarize the available epidemiological evidence on the links between endometriosis and several chronic diseases, discuss hypotheses for potential mechanisms, possible sources of bias and underlying methodological complexities, and provide practice recommendations for clinical care.

EPIDEMIOLOGICAL EVIDENCE ON THE LINKS BETWEEN ENDOMETRIOSIS AND OTHER CHRONIC DISEASES

Endometriosis and cancer

Endometriosis has been suggested to share similar features with malignant tumors, such as abnormal tissue growth, local and distant invasion, resistance to apoptosis, or dysfunction of target organs,³ and several studies investigated the links between endometriosis and cancer. A systematic review and meta-analysis of epidemiological studies on these links was published in 2021 and included 38 studies, of which 19 were case-control and 19 were cohort studies.⁴ While the analysis showed a modest and non-statistically significant association between endometriosis and overall cancer risk [summary relative risk (SRR) = 1.07, 95% confidence interval (CI) = 0.98–1.16], heterogeneity among studies was substantial (I² = 88%, P < 0.0001), i.e. the high level of variation in study methodology and characteristics may have led to an invalid summary estimate. Regarding specific cancer sites, most studies included in the meta-analysis investigated gynecological cancers (ovarian cancer: n = 24 studies; breast cancer: n = 20 studies; endometrial cancer: n = 17 studies), while research on other sites were more limited (n = 2–7 studies).

Ovarian cancer

Overall associations

Ovarian cancer has been the most studied comorbidity in relation to endometriosis. The above-mentioned meta-analysis reported a positive association between endometriosis and this cancer, with an SRR of 1.93 (95% CI = 1.68–2.22).⁴ However, there was significant evidence of publication bias, i.e. positive and statistically significant results were more likely to be published than null or negative findings. This may bias the results towards an overestimation of the relation between endometriosis and ovarian cancer, which is important to consider in its interpretation. Moreover, the level of heterogeneity across studies was high (I² = 78%, P < 0.0001) and incompletely explained by the quality of the studies with a high level of heterogeneity remaining in subgroup analyses by study quality according to the ROBINS-I tool.

Heterogeneity by ovarian cancer histotype or/and endometriosis macro-phenotype

Further subgroup analyses revealed that heterogeneity was explained away by ovarian cancer histotype: associations were stronger for the clear-cell (SRR = 3.44, 95% CI = 2.82–4.20) and endometrioid (SRR = 2.33, 95% CI = 1.82–2.98) histotypes, and there was a modest association also with serous tumors (SRR = 1.17, 95% CI = 1.03–1.32), which was restricted to low-grade serous tumors (SRR = 2.33, 95% CI = 1.64–3.31, n = 2 studies; high-grade tumors: SRR = 1.08, 95% CI = 0.88–1.32, n = 3 studies; P for heterogeneity < 0.0001). Although no publication bias was detected in this subanalysis, the authors stated that this bias cannot be ruled out given the small number of studies providing histotype-specific estimates.

Among the four studies performing subgroup analyses according to the macro-phenotype of endometriosis, three focused on endometrioma only, irrespectively of other subtypes. Based on these four studies, endometrioma was positively associated with ovarian cancer risk, with an SRR of 5.41 (95% CI = 2.25–13.00). Only one study reported the association between endometriosis and ovarian cancer by both macro-phenotype of endometriosis and histotype of ovarian cancer. In this retrospective cohort study, endometrioma was associated with the clear cell [standardized incidence ratio (SIR) = 10.1], endometrioid (SIR = 4.7), and serous (SIR = 1.62) histotypes; similar associations were reported for superficial peritoneal endometriosis, but with a lower magnitude of effect (SIRs of 2.67, 2.03, and 1.32, respectively); while there was no association between deep endometriosis and ovarian cancer risk (SIRs of 0.00, 3.35 and 1.41, respectively, although power was limited given the small number of ovarian cancer cases with deep endometriosis).

Endometrioma – a precancerous lesion?

The association between endometrioma and ovarian cancer risk should be interpreted with caution based on several points. First, only four studies investigated this link in the meta-analysis, with a high level of heterogeneity among them (I² = 82%, P = 0.001) and a high level of imprecision given the wide confidence intervals. Two studies included selected samples (subfertile women⁵ or women with endometriosis only)⁶ and none of these studies was able to analyze the association with endometrioma exclusively. Only one study provided estimates for each of the endometriosis macro-phenotypes – endometrioma (SRR = 2.56, 95% CI = 1.98–3.27), superficial peritoneal (SRR = 1.32, 95% CI = 0.99–1.72), and deep endometriosis (SRR = 1.41, 95% CI = 0.29–4.10)⁷ – and although the association was stronger for endometrioma, the subtype groups were not mutually exclusive, making findings difficult to interpret. In addition, exclusive subgroups should be considered not only for endometrioma, but also for the superficial peritoneal and deep endometriosis macro-phenotypes, since these subtypes are also associated with a cancer-prone hyper-inflammatory environment⁸ and some ovarian cancers have a non-ovarian origin.⁹ In order to have a clearer understanding of the subtype-specific associations of endometriosis with ovarian cancer risk, future studies should provide results among both exclusive and non-exclusive macro-phenotypic subgroups of endometriosis.

Another important point related to endometrioma is that its diagnosis is more straightforward than for other macro-phenotypes of endometriosis;¹⁰ thus, studies may be more likely to include cases involving endometrioma, which would yield a diagnostic bias. In addition, endometrioma is more likely to be visualized among ovarian cancer patients, which again induces a bias when the temporality of the association is not taken into account. Thus, while endometrioma has been proposed to be the main macro-phenotype of endometriosis that could lead to malignant transformation,^11,12 the methodologic points highlighted above show that many uncertainties remain.

Moreover, it has been proposed that endometrioma should be considered a precancerous lesion; it has been posited as a precursor for clear-cell and endometrioid ovarian cancer.¹³ However, although endometrioma shares similar molecular alterations with ovarian cancer,¹⁴ the prevalence of cancer-driver mutations has been shown to be identical in deep endometriosis and endometrioma (a quarter to a third of patients),¹⁵ and that in the eutopic endometrium of healthy women has been suggested to be even higher – over 50%.¹² Thus, the data available to date are insufficient to conclude on the precancerous potential of endometrioma, and more research is requested to have a clear understanding of these associations.

Breast cancer

The meta-analysis yielded a statistically significant but very modest association between endometriosis and breast cancer, based on 20 studies (SRR = 1.04, 95% CI = 1.00–1.09). Only two studies generated estimates by breast cancer subtype: the first study, an American prospective cohort, showed no association with breast cancer overall, however there was a positive association restricted to estrogen receptor-positive (ER+)/progesterone receptor-negative (PR-) tumors [hazard ratio (HR) = 1.90, 95% CI = 1.44–2.50; P for heterogeneity = 0.001].¹⁶ There was no difference according to menopausal status or type of menopause among post-menopausal women. The second study in Denmark reported no differences according to breast cancer subtype.¹⁷ No previous study has reported estimates of the association between endometriosis and breast cancer by endometriosis macro-phenotype. Breast cancer is known as a heterogeneous tumor for which risk factors may differ according to subtype;^18,19 given these differences, it is essential that future research reports associations with endometriosis within breast cancer subgroups.

Endometrial cancer

In the meta-analysis, the SRR was of 1.23 (95% CI = 0.97–1.57) for the association between endometriosis and endometrial cancer, which was close to statistical significance, based on 17 studies. When investigating the high heterogeneity level among studies (I² = 81%, P < 0.0001), however, the authors observed that among studies using a prospective cohort design, which ensures rigorous temporality (i.e. endometriosis precedes endometrial cancer diagnosis), the association was null and the heterogeneity among studies disappeared (SRR = 0.99, 95% CI = 0.72–1.37, I² = 0%, P = 0.51), while the association remained among retrospective cohort studies (SRR = 1.40, 95% CI = 1.00–1.96, I² = 87%, P < 0.0001). These stark differences according to the temporality associated with each design suggest a potential diagnostic bias in the association between endometriosis and endometrial cancer: women undergoing evaluation for endometrial cancer are more likely to be detected for endometriosis, which may artificially lead to a higher co-occurrence of both diseases. Of note, the SRR varied greatly according to other methodologic characteristics of the studies. These differences call for further research to better characterize the link between endometriosis and this cancer.

Very few studies explored the association according to endometrial cancer subtype. One reported a higher association for type I (SIR = 1.54, 95% CI = 1.20–1.96) compared with type II tumors (SIR = 1.06, 95% CI = 0.28–2.71),¹⁷ and one showed a higher association with uterine sarcomas [relative risk (RR) = 2.72] than for indolent types (RR = 1.14).²⁰ Only one study performed subgroup analyses by macro-phenotype of endometriosis and reported a higher risk of endometrial cancer associated with adenomyosis [HR = 4.38, 95% CI = 1.22–15.72], but no statistically significant association with endometrioma (HR = 3.23, 95% CI = 0.54–19.27). While limited statistical power did not permit examination of endometrioma exclusively, the association was strengthened in the exclusive adenomyosis group, although confidence intervals were wide (HR = 5.13, 95% CI = 1.36–19.40).

Skin cancer

The meta-analysis of associations between endometriosis and cutaneous melanoma was based on seven studies, with a resulting SRR estimate of 1.17 (95% CI = 0.97–1.41).⁴ There was a moderate level of heterogeneity among studies (I² = 51%, P = 0.05), which was substantially reduced in subgroup analyses according to ROBINS-I risk of bias, with a SRR of 1.71 (95% CI = 1.24–2.36) in studies with low or moderate risk of bias compared with 1.08 (95% CI = 0.87–1.26) in those with critical or serious risk of bias (P for heterogeneity = 0.07). In contrast to findings on endometrial cancer, the substantial increase in magnitude of effect and statistical significance in higher-quality studies suggests that for cutaneous melanoma, biases may be more likely to mask a true association. Of note, the sole study reporting estimates by histotype of melanoma reported no statistically significant differences.²¹

A meta-analysis of the two studies^21,22 evaluating associations between endometriosis and basal-cell carcinoma resulted in a SRR of 1.18 (95% CI = 1.11–1.25) with no evidence of heterogeneity (I² = 0%, P = 0.89).⁴

Thyroid cancer

The meta-analysis included a total of five studies on the associations between endometriosis and thyroid cancer, resulting in a statistically significant SRR of 1.39 (95% CI = 1.24–1.57) with no evidence of heterogeneity among studies (I² = 0%, P = 0.69).⁴

Colorectal cancer

The five studies that were included in the meta-analysis of associations between endometriosis and colorectal cancer yielded a SRR of 1.00 (95% CI = 0.87–1.16), with a low level of heterogeneity among studies (I² = 40%, P = 0.16).⁴ However, there was a positive association among studies with low or moderate risk of bias (SRR = 2.29, 95% CI = 1.00–5.26), suggesting a potential impact of bias in the overall results here also, although confidence intervals were wide.

Cervical cancer

The meta-analysis included four studies on the links between endometriosis and cervical cancer, resulting in a robust and statistically significant SRR of 0.68 (95% CI = 0.56–0.82) with no heterogeneity among studies (I² = 0%, P = 0.76).⁴ However, it should be noted that all studies were conducted in Europe and were based on self-reported endometriosis. This 32% lower risk of cervical cancer associated with endometriosis is consistent with the results from two previous meta-analyses reporting reduced risks of 33%²³ and 22%.²⁴ However, caution should be used in the interpretation of this inverse association: while the association is unlikely to reflect causality, it implies probable diagnostic and treatment biases linked to a higher exposure to the healthcare system among women with endometriosis.²⁵ Due to this closer contact with the medical system (and gynecologists, in particular), endometriosis patients are indeed more likely to be routinely screened for cervical hyperplasia, and to receive treatment if positive than women without endometriosis or gynecological conditions. This highlights the importance of considering the impact of access to care on the associations between endometriosis and cancer risk in general. Another potential pathway that could explain this inverse association is the impact of dyspareunia and chronic pelvic pain on the frequency of sexual intercourse, which could lead to a lower prevalence of human papillomavirus infection among women with endometriosis.

Other cancers

Endometriosis was assessed in relation to several other cancer types, with <4 studies included in the meta-analysis.⁴ There were small associations between endometriosis and the risks of lymphatic and hematopoietic cancers (SRR = 1.09, 95% CI = 1.00–1.19, n = 2 studies), non-Hodgkin lymphoma (SRR = 1.18, 95% CI = 1.00–1.41, n = 3 studies), and brain cancer (SRR = 1.18, 95% CI = 1.02–1.36, n = 2 studies). However, the number of studies was low and mostly used a retrospective design, thereby limiting the ability to adjust for potential confounding factors. More research will thus need to be performed to inform our knowledge of the associations between endometriosis and risk of other cancers.

Informing patients on their cancer risk

Women living with endometriosis may be concerned about their long-term cancer risk and need to be accurately informed about this risk. For clinicians, the reported links between endometriosis and some types of cancer raise a number of practical questions on the long-term management of endometriosis patients. For these reasons, it is critical to translate the results expressed in relative risks into absolute risks for easier interpretation and balanced perspective.

The cancer types with the most robust positive associations with endometriosis are ovarian, breast, and thyroid cancers. While the absolute risk of developing ovarian cancer in the general female population is 1.3% (1.3 out of 100 women will develop ovarian cancer in their lifetime),²⁶ the translation of the SRR reported in the meta-analysis (SRR = 1.93)⁴ translates into an absolute risk of 2.5% among women with endometriosis (+1.2% only), which remains very low.²⁷ Although this number is small, the true absolute risk is likely even smaller given the publication bias that likely over-estimated this association. Based on the meta-analytic results for other cancers, the absolute risk of developing breast cancer in a woman’s lifetime increases from 12.8% in the general female population²⁶ to 13.3% in women with endometriosis (+0.5%).⁴ For thyroid cancer, this risk increases from 1.3% in the general female population to 1.8% in women with endometriosis (+0.5%). Thus, as recommended in the most recent update of the European Society of Human Reproduction and Embryology (ESHRE) guideline on the management of endometriosis,²⁸ clinicians should reassure women with endometriosis with regards to their cancer risk, which remains close to that of the general female population. They may address their concern to reduce their risk by recommending general cancer prevention measures for the general population (avoiding smoking, maintaining a healthy weight, exercising regularly, having a balanced diet with high intakes of fruits and vegetables and low intake of alcohol, and using sun protection).

Should women with endometriosis be monitored for early detection of cancer?

The reported increased risks of ovarian (93%), breast (4%), and thyroid cancer (39%) among women with endometriosis could lead clinicians to direct patients to more frequent cancer screening. However, as detailed above, the translation of these relative risks into absolute risks equate to only up to 1.2% increases in absolute risks of cancer compared with the general population. These very small differences suggest that cancer screening recommendations should not differ in women with endometriosis compared with those in the general female population. Thus, increased screening is only recommended in women with known cancer risk factors (i.e. family history of cancer or germline mutation predisposing to cancer risk). The absolute risk of ovarian cancer in these women may indeed be up to 50% compared to the 1.3% in the general population. Bilateral salpingo-oophorectomy (BSO) may be recommended in some of these high-risk women to provide additional ovarian cancer risk reduction;²⁹ however, BSO is not recommended in women with endometriosis in the absence of other ovarian cancer risk factors, given the significant harms associated with this procedure with regards to the long-term health consequences in women (before menopause: cardiovascular disease, depression, arthritis, asthma, chronic obstructive pulmonary disease, osteoporosis; after menopause: cardiovascular disease, anxiety, sexual dysfunction, fractures, neurologic disorders, cognitive impairment).²⁸

While cancer guidelines recommend regular screening for breast, cervical, and colorectal cancers in the general population based on evidence of significant benefit,³⁰ regular screening for ovarian cancer through serum CA-125 or transvaginal ultrasound is not recommended in the general population. Randomized-controlled trials have indeed shown not only no benefit of screening on early ovarian cancer detection or mortality reduction, but also significant harms (e.g. unnecessary surgery, surgical complications, infections, and cardiovascular or pulmonary complications) among those receiving false-positive test results.^31,32 Of course, long-term management decisions must vary according to endometriosis patients’ medical history, characteristics, other risk factors, and patient preferences after receiving full information on the current evidence and benefit/risk ratio.⁴ One area of uncertainty on this topic is monitoring for detection of cancer among asymptomatic endometriosis patients and among postmenopausal women with endometrioma.²⁸ Further investigation is requested on the malignant transformation of endometriosis in order to precisely guide monitoring for malignancy.

Endometriosis and autoimmune diseases

Immune system abnormalities have been consistently shown among women with endometriosis, such as altered immune surveillance, depressed cell-mediated immunity (high T-, B-, and NK cell counts with decreased activity), and higher humoral immune response (higher levels of IgG, IgA, IgM, and anti-endometrial antibodies).³³ A systematic review and meta-analysis published in 2019 summarized the literature on the associations between endometriosis and autoimmune diseases.³⁴ While the review included 26 studies in the qualitative analysis and 14 in the quantitative analysis, only a small number of studies explored each autoimmune disease outcome in relation to endometriosis, and the quality of studies evaluated through the GRADE criteria was generally low: only five studies could provide high-quality evidence, and among them, four supported associations with systemic lupus erythematosus (SLE), Sjögren syndrome (SS), rheumatoid arthritis (RA), celiac disease (CeD), multiple sclerosis (MS), or inflammatory bowel diseases (IBDs).

The meta-analysis included six studies on the associations between endometriosis and SLE: one cross-sectional study, three case-control studies, and two cohort studies.³⁴ The relative risk for this association was of 1.74 (95% CI = 1.10–2.77) among cohort studies, with no heterogeneity detected. Based on five studies, the meta-analysis yielded positive associations between endometriosis and SS, with a RR of 1.60 in the sole cohort study included (95% CI = 1.30–2.00).³⁵ Five studies were included also in the meta-analysis for RA; the sole prospective cohort study reported a RR of 1.41 (95% CI = 1.05–1.89).³⁶ Celiac disease was associated with endometriosis based on three studies, with one cohort study reporting a RR of 1.39 (95% CI = 1.14–1.70).³⁷ Only two studies on multiple sclerosis were included in the systematic review, and the sole retrospective cohort study reported a SIR of 1.2 (95% CI = 1.1–1.5).³⁵ Regarding IBD, the review included one retrospective cohort (n = 37,661 endometriosis cases and n = 320 IBD cases)³⁸ and one case-control study (n = 19,005 endometriosis cases and n = 14,279 IBD cases).³⁹ Both reported positive associations (SIR = 1.5, 95% CI = 1.4–1.7 in the former; OR = 2.06, 95% CI = 1.83–2.31 in the latter), with similar magnitude of effect for Crohn's disease and for ulcerative colitis. The meta-analysis also showed positive associations between endometriosis and autoimmune disease of the thyroid based on three case-control studies (OR = 1.31 to 1.36).³⁴

Endometriosis and cardiovascular diseases

Cardiovascular disease risk may be affected by several characteristics of women’s health or female conditions, such as pregnancy-related disorders, polycystic ovarian syndrome, premature menopause, or hysterectomy.⁴⁰ Several epidemiological studies investigated the links between endometriosis and cardiovascular disease risk, and a systematic review and meta-analysis published in 2023 evaluated associations with ischemic heart disease (including myocardial infarction) and cerebrovascular disease (including ischemic or hemorrhagic stroke).⁴⁰ Based on a total of six cohort studies (two prospective – both based on the Nurses’ Health Study II, and four retrospective), the HR for the association between endometriosis and ischemic heart disease was of 1.50 (95% CI = 1.36–1.66), and that for the association with cerebrovascular disease was of 1.20 (95% CI = 1.11–1.30).

Endometriosis and asthma/atopic diseases

Endometriosis is known to be associated with aberrant immune response and heightened inflammatory milieu, and women with endometriosis have been reported to have a higher susceptibility to allergic manifestations and atopic diseases such as asthma than women without endometriosis.⁴¹ The first reports date from the 1980s where two small US case-control studies suggested higher rates of eczema, hay fever, food sensitivities and allergic reactions among women with endometriosis compared with controls.^42,43 Then, a survey of over 3500 patient members of the Endometriosis Association in 2002 reported a higher proportion of self-reported allergies (three-fold) and asthma (two-fold) as compared with the general US female population.⁴⁴ While an Italian hospital-based case-control study found no differences in terms of asthma prevalence between endometriosis cases and surgical controls,⁴⁵ endometriosis cases were more likely than controls to report allergies, asthma, allergic rhinitis, medication allergy or family history of allergy in another hospital-based case-control study from the US.⁴⁶ Moreover, another Italian case-control study⁴⁷ described a higher proportion of allergies in women with endometriosis and reported that carriers of the C allele of the acid phosphatase locus 1 (ACP1) polymorphism (suggested to have a role in allergic manifestations) were at higher risk of endometriosis than those carrying other genotypes. Of note, this gene was not reported to be associated with endometriosis in genome-wide scans.⁴⁸

Several other studies suggested a higher prevalence of allergic manifestations among women with endometriosis more recently. In a cross-sectional analysis of a US longitudinal study of adolescent endometriosis cases and controls, participants with allergies (OR = 1.76, 95% CI = 1.32–2.36) or asthma (OR = 1.35, 95% CI = 0.97–1.88) had higher odds of co-occurring endometriosis.⁴⁹ A Japanese study using insurance claims data involving over 30,000 endometriosis cases and close to 121,000 controls reported a positive association between type 1 allergy and endometriosis (incidence rate ratio = 1.10, 95% CI = 1.06–1.13).⁵⁰ A Taiwanese study based on national insurance claims compared over 7300 women with asthma and close to 29,000 women without and yielded a HR of 1.50 (95% CI = 1.33–1.70) for the association with endometriosis.⁵¹ A Finnish study using national register data with follow-up of almost 4000 women under age 50 reported positive associations between endometriosis and self-reported asthma (OR = 1.51, 95% CI = 1.10–2.15), atopic and allergic eczema (OR = 1.39, 95% CI = 1.07–1.81), eye allergy symptoms (OR = 1.54, 95% CI = 1.61–2.04), and emphysema or chronic bronchitis (OR = 1.72, 95% CI = 1.01–2.95).⁵² Finally, a small Chinese cross-sectional study reported higher odds of allergy, hay fever, and dog-specific IgE among women with endometriosis compared with those without.⁵³

Endometriosis and chronic overlapping pain conditions

Chronic pain may lead to central sensitization – a persistently high-activity state of the nervous system that worsens sensitivity to pain⁵⁴ – which in turn may predispose patients to the development of additional chronic pain conditions.⁵⁵ Several studies have suggested an association between endometriosis and several chronic overlapping pain conditions, such as migraine, interstitial cystitis, or fibromyalgia. A 2020 meta-analysis on the links between endometriosis and migraine yielded a SRR of 1.56 (95% CI = 1.21–1.90) based on nine studies, with a low degree of heterogeneity (I² = 28.7%) and no indication of publication bias.⁵⁶ When restricted to cohort studies (n = 3), the SRR decreased to 1.36, but with a narrower confidence interval (95% CI = 1.27–1.46). The Newcastle-Ottawa tool indicated high quality for all included studies.

Among the studies that assessed the links between endometriosis and interstitial cystitis (or bladder pain syndrome), several clinical investigations reported co-occurrence of both conditions, with high prevalence rates among women with chronic pelvic pain^{57,58,59,60,61} – the co-occurrence of interstitial cystitis and endometriosis being named the ‘evil twins’ syndrome.^58,59,62 A 2013 systematic review of nine studies among women with chronic pelvic pain reported a mean prevalence of 61% for bladder pain syndrome, of 70% for endometriosis, and of 48% for both conditions.⁶² Two large cohort studies confirmed these associations more recently: a US retrospective matched cohort study (n~27,000 endometriosis cases and 108,000 controls) based on insurance claims reported that endometriosis was associated with a close-to-doubled risk of interstitial cystitis,⁶³ and a retrospective cohort study (n~9200 endometriosis cases and 27,000 controls) using the Taiwanese national insurance database reported that the risk of developing interstitial cystitis over a 3-year period among cases compared to non-cases was 4.43 times higher (95% CI = 2.13–9.23); the HR was 3.74 (95% CI = 1.76–7.94) after adjustment for other comorbidities.⁶⁴

In 2002, the cross-sectional study among members of the US Endometriosis Association reported higher rates of fibromyalgia compared to the general female population (5.9% vs. 3.4%).⁴⁴ A few other studies explored this link more recently. A retrospective cohort study based on the Swedish national register on healthcare consultations in 2007–2016 yielded an IRR of 2.83 (95% CI = 1.96–4.08) for the association between endometriosis and later diagnosis of fibromyalgia.⁶⁵ Interestingly, a small Spanish case-control study reported significantly higher rates of fibromyalgia assessed by LFESSQ-4 among women with deep endometriosis (39%), than among women with superficial endometriosis or endometrioma (16%) or among controls (8%).⁶⁶ A cross-sectional study of the computerized database of a large healthcare plan in Israel (~6650 endometriosis cases, 775,000 non-cases) yielded a crude OR of 1.92 for the association between endometriosis and fibromyalgia.⁶⁷ Finally, a cross-sectional analysis within the Women’s Health Study: From Adolescence to Adulthood reported that female adolescents and young adults with chronic fatigue syndrome and/or fibromyalgia had increased odds of co-occurring endometriosis (OR = 5.81, 95% CI = 1.89–17.90).⁴⁹

Three meta-analyses were published on the associations between endometriosis and irritable bowel syndrome (IBS) between 2020 and 2022.^68,69,70 The most recent included 11 studies, which yielded a pooled OR of 2.97 (95% CI = 2.17–4.06) and reported a high level of heterogeneity (I² = 91%) and evidence of publication bias; heterogeneity was higher among studies not reporting the criteria used to diagnose IBS, while the sole study using Rome IV criteria yielded the association with the highest magnitude of effect (OR = 5.26, 95% CI = 2.13–13.0).⁷⁰ The summary estimate was similar in one of the meta-analyses (SRR = 2.39, 95% CI = 1.83–3.11),⁶⁸ although it was higher in the other one (SRR = 3.26, 95% CI = 1.97–5.39) due to the inclusion of a reduced set of studies.⁶⁹ One meta-analysis reported a pooled prevalence of IBS in women with endometriosis of 23.4% (n = 6 studies; 95% CI = 9.7–37.2%).⁶⁸ A higher prevalence of IBS was reported among women with minimal to mild endometriosis (65%) compared to moderate to severe endometriosis (55%).^68,69

Endometriosis and other diseases

This review is not comprehensive as endometriosis has been associated with many other conditions in the literature. Some of the diseases that were described as co-occurring with endometriosis include pregnancy complications,⁷¹ depression/anxiety and eating disorders,^72,73,74 uterine fibroids,^63,75 long Covid,⁷⁶ or endometritis.⁷⁷ Many other conditions that are currently under-explored in relation to endometriosis would be important to cover in the field in order to have a more complete picture of the comorbidities associated with endometriosis.

METHODOLOGICAL CONSIDERATIONS IN THE STUDY OF COMORBIDITIES ASSOCIATED WITH ENDOMETRIOSIS

Previous work^1,4 has identified a number of methodologic complexities that need to be carefully considered in the study of the associations between endometriosis and the risk of other diseases.

Temporality

To discuss the impact of endometriosis as a potential risk factor for other chronic diseases, analysis need to ensure that endometriosis actually precedes the diagnosis of the comorbidity. To ensure temporality of the association, the use of a prospective cohort design is recommended to enable time-varying analysis, with a sufficiently long duration of follow-up to allow for the development of the studied comorbidity following endometriosis diagnosis. However, retrospective designs may also ensure this temporal order by restricting the definition of endometriosis diagnosed at least 1 year before the diagnosis of the comorbidity. The findings on the links between endometriosis and endometrial cancer showed that considering temporality can dramatically modify the conclusions on an association.⁴

Misclassification and population sampling

Misclassification of endometriosis is a likely bias given the difficulties associated with diagnosis, possibility of asymptomatic disease, and diagnostic delays.⁷⁸ In the absence of a non-invasive diagnostic tool, diagnostic biases are likely to be driven by the characteristics of the women who are able to achieve diagnosis and their symptoms. There is misclassification potential for the comorbidities under study also (with a low likelihood for cancer, usually confirmed through pathology reports). This double imprecision, and the potential overlapping age-specific incidence curves between endometriosis and some of its comorbidities, make it difficult to assess the temporality of the relationship (endometriosis leads to a higher risk of the comorbidity, is a consequence of the comorbidity, or co-occurs with the comorbidity). This misclassification has the potential to either induce spurious associations due to systematic error, or to attenuate the associations under study. Associations are driven towards the null also in studies basing the comparison group on a selected sample (e.g. infertile patients or women who have undergone hysterectomy), since potential underlying pathology in these women may also be associated with risk of the comorbidity.⁷⁹ It is thus important to use a population-based sample and to interpret the results considering this methodologic issue inherent to endometriosis research.

Confounding and mediation

Studies evaluating the association between endometriosis and other diseases should assess if these associations are driven by common risk factors, whether these factors precede endometriosis and the comorbidity under study (confounders) or are along the causal pathway between endometriosis and the comorbidity (mediators). Few of the studies reviewed above were able to adjust for potential confounders, and even fewer explored potential mediators of the associations.

Study robustness and study heterogeneity

The assessed quality of evidence in the reviewed meta-analyses showed that generally high proportions of the included studies were at risk of bias, and in the cancer meta-analysis, statistically significant heterogeneity was observed for a large portion of cancer types.⁴ Part of this heterogeneity was driven by the methodologic limitations of many studies, potentially leading to inaccurate or invalid estimates. Heterogeneity among studies also pertained to differences in the characteristics of the studied populations, which may lead to true differences in associations.

Endometriosis and disease heterogeneity

As reviewed above, extremely few studies were able to study associations between endometriosis and comorbidity risk by macro-phenotype of endometriosis, and by comorbidity subtype where relevant (e.g. cancer). In the cancer meta-analysis, while subgroup analysis confirmed that the association with ovarian cancer was restricted to specific histotypes (clear cell and endometrioid), only four studies were able to provide estimates according to the macro-phenotype of endometriosis, and only one by both type of endometriosis and histotype of ovarian cancer, none of which provided estimates for each type exclusively.⁴ This is particularly important however, given our lack of knowledge on endometriosis heterogeneity and the potential of this knowledge to lead to breakthroughs in our understanding of the disease. Beyond the study of associations between endometriosis and its comorbidities, the lack of routine standardized reporting of endometriosis characteristics in medical reports currently hampers analysis to this level of detail. The record and use of such data is critical in endometriosis research and has a high potential to provide new insights into the etiology of the disease.⁸

Potential mechanisms

While the exact pathophysiology underlying the associations between endometriosis and its comorbidities is unknown, several mechanisms may be hypothesized.

First, associations between endometriosis and other diseases may be driven by shared risk factors (e.g. genetic susceptibility, patient characteristics, or environmental exposures). With regards to genetics, while cross-disease genetic overlap studies reported loci common to endometriosis and ovarian cancer,^80,81 endometrial cancer,⁸² cardiovascular diseases,⁸³ or other pain and inflammatory conditions,⁸⁴ it should be mentioned that these studies did not examine variations in associations according to the method of endometriosis assessment or consider the potential impact of diagnostic bias. Other factors (such as demographics, menstrual characteristics, anthropometry, lifestyle, or exposure to environmental toxicants) may act as confounders of the associations between endometriosis and cancer; however, too few studies were able to adjust for such factors or to examine potential mediating effects.

Beyond overlapping risk factors raising the risk of other diseases among women with endometriosis through various pathways, it is possible that endometriosis is causally associated with these comorbidities. In this regard, mediation analysis is likely to provide insights into the pathways involved in these associations, which may include infertility, stress, anxiety and depression, or lifestyle changes associated with endometriosis symptoms (e.g. decreased physical activity). Mediators may also include treatment of endometriosis lesions and associated symptoms (medication use, surgical procedures including lesion ablation/excision, hysterectomy, bilateral salpingo-oophorectomy), which have also been associated with the risk of several chronic diseases.

Through mediators or through direct pathways, the presence of endometriosis may induce systemic changes that create a milieu prone to the development of other chronic diseases, such as chronic inflammation, aberrant hormonal milieu, or aberrant immune response.⁸⁵ Further research is warranted to enhance our understanding of the pathways underlying the associations between endometriosis and its comorbidities, to describe these associations according to the characteristics of endometriosis lesions and of the comorbidities where possible, and to elucidate why associations differ according to these characteristics when they do. The -omics technologies are likely to help increase our knowledge and understanding of these associations in the future.

Finally, reported associations between endometriosis and other diseases may be driven by methodological bias, as discussed above. To reduce risk of bias and ensure a high-quality level of response to these research questions, it is important to consider validated diagnoses of endometriosis and its comorbidities as much as possible to reduce misclassification, to use a population-based sample and a prospective design (or a retrospective design ensuring temporality of the association) with a long duration of follow-up, to take into account potential confounders and mediators of the associations, to investigate endometriosis and disease heterogeneity and provide estimates within disease subgroups, and to use harmonized endometriosis data to allow for international comparisons (WERF-EPHect tools^86,87,88,89).

CONCLUSION

In conclusion, comorbidities research in endometriosis has suggested associations between endometriosis and several other chronic diseases. Meta-analyses were reviewed on the associations with cancer, autoimmune diseases, cardiovascular diseases, migraine, and irritable bowel syndrome; associations with other reviewed comorbidities were investigated among fewer studies. Based on the current evidence, endometriosis has been associated with a 93% increased risk of ovarian cancer [particularly the clear-cell (244%) and endometrioid (133%) histotypes], a very small (4%) increase in breast cancer risk, a 39% increase in thyroid cancer risk, and a 32% decreased risk of cervical cancer (most likely reflecting diagnostic and treatment bias). Based on very few studies, endometriosis has also been associated with an increased risk of several autoimmune diseases: systemic lupus erythematosus (74%), Sjögren syndrome (60%), rheumatoid arthritis (41%), celiac disease (39%), multiple sclerosis (20%), and inflammatory bowel diseases (50%). Evidence on cardiovascular diseases suggests a 50% increased risk of ischemic heart disease and a 20% increased risk of cerebrovascular disease. One meta-analysis suggested a 56% increased risk of migraine, and the meta-analyses on irritable bowel syndrome suggested a two- to three-fold increased risk.

Next to these meta-analytic findings, the literature suggests associations between endometriosis and asthma and several allergic manifestations, chronic overlapping pain conditions such as interstitial cystitis, fibromyalgia or irritable bowel syndrome, and many other conditions for which the associations with endometriosis were not reviewed here, including pregnancy complications, depression and anxiety, long Covid, uterine fibroids, or endometritis.

It should be noted that a high proportion of studies investigating these associations had methodological flaws, making the evidence available to date generally poorly reliable, which highlights the importance of performing research with high-standards to produce high-quality evidence in this area. The confirmation of these associations in carefully conducted studies will inform our knowledge on the etiology and heterogeneity of endometriosis and enable the development of specific guidelines for clinical care for patients with endometriosis and/or other conditions.

To ensure a high quality of response to this or any other research questions on endometriosis, future studies should collect and analyze data according to the recommendations of the World Endometriosis Research Foundation Endometriosis Phenome and Biobanking Harmonization Project (WERF-EPHect), which will ultimately allow for a high quality of endometriosis data globally, international collaboration, and cross-country comparisons.^90,91,92

PRACTICE RECOMMENDATIONS

Online Study Assessment Option
All readers who are qualified doctors or allied medical professionals can now automatically receive 2 Continuing Professional Development credits from FIGO plus a Study Completion Certificate from GLOWM for successfully answering 4 multiple choice questions (randomly selected) based on the study of this chapter.
Medical students can receive the Study Completion Certificate only.

 

(To find out more about FIGO’s Continuing Professional Development awards programme CLICK HERE)

 

I wish to proceed with Study Assessment for this chapter