Endometriosis, Anxiety, and Atherosclerosis: A Study of Eight Million Young Hospitalized Women in the USA

Endometriosis is a condition defined by the presence and growth of estrogen-dependent endometrial-like tissue outside the uterus [1]. It is estimated that 10% of women in their reproductive age suffer from this condition [2]. Endometriosis can cause a variety of symptoms including pelvic pain, infertility [3], anxiety, or depression [4]. Some studies suggest that endometriosis is a systemic disease rather than a disease predominantly affecting the pelvis [5], and there is increasing evidence defining endometriosis as a state of chronic inflammation [6]. Prostaglandins, vascular endothelial growth factor, tumor necrosis factor, nerve growth factor, and interleukins are some of the cytokines released in endometriosis [7]. Some studies suggest a link between endometriosis and endothelial dysfunction [8].

Atherosclerosis is characterized by plaque formation in the vascular intima [9]. It involves large- and medium-sized arteries and can lead to cardiovascular conditions such as stroke, ischemic heart disease, and peripheral vascular disease [10]. Risk factors such as abnormal cholesterol level, high blood pressure, diabetes, obesity, both active and passive smoking, and conditions like psoriasis [11] increase the odds of atherosclerosis development. Both inflammation [12] and endothelial cell dysfunction [13] have a strong link to atherosclerosis. Endothelial dysfunction typically appears in lesion-prone area in the artery and alters local vascular tone and redox balance, involves the focal permeation of lipoprotein particles, trapping, and eventually plaque formation [13].

Some studies suggest that chronic stress is associated with endothelial dysfunction and atherosclerosis [14]. Similarly, several authors have advocated that there is an association between endometriosis, endothelial dysfunction, and atherosclerosis [8, 15, 16]. Our study examined the association between endometriosis, anxiety, and atherosclerosis in a large population and aimed to investigate the scale of this association in young patients. Considering that endometriosis is mainly present in women in childbearing age [1] and the fact that the risk of atherosclerosis rises with advanced age [11], we restricted our analysis only to patients below the age of 35 years old.

Using a cross-sectional retrospective study design, we analyzed data from the Healthcare Cost and Utilization Project – National Inpatient Sample (HCUP NIS) database. Our study population included women under 35 years of age who were admitted to one of the HCUP participating hospitals during the study period between 2007 and 2014. Patients who were diagnosed with cancer were excluded from the analyses.

HCUP databases are comprehensive databases that contain information on inpatient stays, emergency department visits, and ambulatory cares in the USA. These data were collected from state data organizations, hospital associations, private data organizations, and the Federal government.

The International Classification of Disease, ninth revision (ICD-9) codes were used to extract the relevant variables. We used 617x for endometriosis, 440 for atherosclerosis, 250x for diabetes mellitus, 338.2,338.4, 338.29 for chronic pain, 305.1 for smoking, 642.0–642.4, 642.10–642.14, 642.20–642.24, 401.0, 401.1, 201.9 for chronic hypertension, 278.00, 278.01, 278.02, 278.03, 278, 278.0 for obesity, 296.2x, 296.3x, 300.4x, 311x for depression, 696x for psoriasis, 714x for rheumatoid arthritis, 272x for dyslipidemia, 150x–165x, 170x–176x, 179x–184x, 188x, 189x for cancer, and 300.0x, 300.2x, 300.3x, 309.8x for anxiety disorders.

Statistical analyses were executed with SPSS version 27.0. We evaluated the prevalence of endometriosis and atherosclerosis. We applied the Kolmogorov-Smirnov one sample test to evaluate the data distribution. Chi-square test was used to compare the proportion of categorical variables between two groups. The Pearson correlation between independent variables were under 0.7, indicating that variables were independent. Logistic regression analyses were performed to evaluate the association between endometriosis and atherosclerosis. We adjusted the analyses for age, race, type of insurance, smoking, chronic hypertension, DM, dyslipidemia, obesity, PCOS, acute DVT, psoriasis, chronic pain, rheumatoid arthritis, and depression. We calculated the odds ratio (OR) and its 95% confidence interval (CI). p value <0.05 was considered significant. Our primary outcome aimed to determine whether endometriosis and anxiety increased the risk of atherosclerosis in patients under 35 years of age.

In this study, we compared 7,670 atherosclerosis patients with 8,054,084 control patients. The proportion of endometriosis and anxiety diagnoses within the non-atherosclerosis group was 0.05%. Considering α = 0.05 and OR = 3.075, the power of the study was >99%.

Of 8,061,754 patients, 7,670 had atherosclerosis, 35,639 had endometriosis, 352,432 had anxiety, while 4,118 presented with both endometriosis and anxiety. The patients under 35 years of age who were hospitalized for atherosclerotic cardiovascular disease had an upward trend during the study period. The prevalence of atherosclerosis in 2007 was 7 per 10,000 patients, while in 2014, it was 11 per 10,000 (Fig. 1a). In contrast, the prevalence of hospitalized endometriosis patients declined from 47 cases per 10,000 patients in 2007 to 35 diagnoses per 10,000 patients in 2014 (Fig. 1b).

Table 1 shows the characteristics of women diagnosed with atherosclerosis. Age (i.e., 25–34), race (i.e., both white and black), low income, Medicare insurance plan and self-pay, as well as comorbidities that included smoking, obesity, hypertension, dyslipidemia, diabetes mellitus, rheumatoid arthritis, psoriasis, chronic pain, depression, acute DVT, PCOS, endometriosis, and anxiety, were more prevalent in patients with atherosclerosis than those in the control group. Similarly, age (i.e., 25–34), race (i.e., both white and black), earning higher than low-income threshold, Medicare and private insurance plan, and self-pay, and comorbidities including smoking, obesity, thyroid disease, hypertension, dyslipidemia, diabetes mellitus, rheumatoid arthritis, psoriasis, chronic pain, depression, acute DVT, PCOS, and anxiety, were more common in patients with endometriosis when compared to control group patients (Table 2).

Table 3 shows the logistic regression model for atherosclerosis. An unadjusted analysis indicated a moderate association between endometriosis and atherosclerosis (OR = 1.740; 95% CI: 1.325–2.286). After adjusting the analysis for age categories, race, type of insurance, smoking, chronic hypertension, DM, dyslipidemia, obesity, PCOS, acute DVT, psoriasis, chronic pain, rheumatoid arthritis, and depression, the probability of having atherosclerosis was roughly 42% higher in patients with endometriosis (OR = 1.421; 95% CI: 1.058–1.910). Similarly, young patients with anxiety disorders had an increased risk of atherosclerosis (OR = 1.352; 95% CI: 1.249–1.464), and the group with both endometriosis and anxiety presented a risk of atherosclerosis about threefold greater compared to those without endometriosis (OR = 3.075; 95% CI: 1.969–4.803).

We observed a rising prevalence of atherosclerosis between 2007 and 2014 among young, hospitalized women, whereas in-patient endometriosis presented an opposite trend. We found a significant association between endometriosis and atherosclerosis, and this association was stronger in the presence of anxiety.

In a retrospective observational study, Mackinnon et al. [17] reported increasing prevalence of atherosclerotic cardiovascular disease in Ontario, between 2002 and 2018. The estimated global prevalence of peripheral artery disease revealed a similar tendency [18]. A study in Europe predicted an increase in cerebrovascular accident in some European countries, such as Lithuania and a decrease in other countries like Portugal [19] with overall growth in the prevalence of cardiovascular accident.

Population-based studies estimated the annual incidence of endometriosis ranging from 0.1% to 0.3% [20]. The lower prevalence and the declining trend of endometriosis seen in our analysis could be potentially explained by the fact that the cases reported in the HCUP NIS database were hospitalized patients, while milder endometriosis cases were treated in ambulatory clinics and thus not included. A 10-year retrospective cohort study using insurance electronic health records database between 2006 and 2015 among 332,056 eligible participants in the USA revealed that the incidence rate of endometriosis had declined during this period [21]. Another study in the Icelandic population reported similar results [20].

In agreement with our findings, Kucharska-Newton et al. [22] reported that low income, as a social determinant of health, was associated with a higher risk of cardiac events in Finland and in the USA. Different studies have demonstrated the association between anxiety and incidence [23] of coronary artery disease and its mortality [24]. Nevertheless, studies on the concurrent impact of endometriosis and anxiety on atherosclerosis remain scant. In our study, we observed a modest combined effect of anxiety and endometriosis on the development of atherosclerosis.

In a Mendelian randomized study, Zheng et al. [25] showed that endometriosis was associated with an increased risk of stroke. In a prospective study, Farland et al. [26] documented 893 strokes over a follow-up period of 2,770,152 person-years in patients with laparoscopically confirmed endometriosis (hazard ratio = 1.34; 95% CI: 1.10–1.62). Likewise, in a prospective cohort study that included 116,430 women, laparoscopically confirmed endometriosis was associated with a higher risk of coronary heart disease [27]. In another 9-year cohort study of 17,543 Taiwanese women [28], the authors concluded that endometriosis patients present with an increased risk of major adverse cardiovascular and cerebrovascular events (hazard ratio = 1.2: 95% CI: 1.05–1.29). A different population-based cohort study in the UK [29] compared 56,090 endometriosis patients with 223,669 matched controls without endometriosis. Adjusted hazard ratios for ischemic heart disease were 1.4 (95% CI: 1.22–1.61) and for cerebrovascular disease were 1.19 (95% CI: 1.04–1.36). This study suggests that even young patients with endometriosis should potentially present as targets for risk assessment and prevention of cardiovascular disease.

Different pathophysiological pathways could play a role in the associations found. Chronic inflammation appears to be the leading mechanism in the onset and progression of atherosclerosis and would concurrently play a key role in the pathogenesis of endometriosis [30]. Also, endothelial dysfunction is believed to be a key variable in the pathophysiology of atherosclerosis, which can potentially result from chronic inflammation.

Our findings support the existing body of evidence that endometriosis is a risk factor for cardiovascular disease. Though menopause increases the risk of atherosclerosis, and endometriosis is largely considered to be a disease that occurs in reproductive-aged women, endometriosis should also be considered as a risk factor for atherosclerosis in younger populations. Further studies should explore the underlying mechanisms that link endometriosis and the development of atherosclerosis. In a small number of patients, we observed the synergic effect of endometriosis and anxiety on atherosclerosis. Therefore, research on the potential role of psychologic conditions on systemic inflammatory diseases is deemed timely and important.

Screening patients for possible anxiety with brief and valid screening tools such as the Mental Health Inventory (MHI-5) or the Generalized Anxiety Disorder scale-7 (GAD-7), which is an effective tool for screening generalized anxiety disorder, could be recommended in practice. The impact of these comorbid clinical conditions on patients’ productivity and activity level should also be evaluated using a valid tool for the health provider to act promptly and more effectively to improve patient mental health and well-being.

HCUP NIS databases gather data from over one thousand hospitals in the USA and accommodate over several million patients each year. To our knowledge, our study has the largest population among other population-based studies on these research questions. Further, we evaluated young patients with atherosclerosis and endometriosis or anxiety disorders. However, the cross-sectional design of the HCUP database limits our ability to establish temporal relationships between these conditions. Additionally, certain information such as the severity of disease, laboratory findings, or medical treatment is not available in HCUP NIS databases. Moreover, even though we adjusted our analysis for important variables that can have confounding effect, residual confounding might still be a factor.

Our study suggests that endometriosis is associated with an elevated risk of atherosclerosis in young patients. These results are in agreement with the hypothesis that endometriosis is a chronic systemic inflammatory condition that can lead to endothelial dysfunction. Further studies on the potential role of psychologic conditions, such as anxiety, on systemic inflammatory diseases, are deemed of critical importance.

Ethical approval was not required for this study in accordance with local and national guidelines, as outlined in the Tri-Council Policy Statement (2010). The need for informed consent was waived by the MUHC REB Full Board Meeting, as the study was based solely on anonymous, publicly available data.

The authors report no conflict of interest.

Access to HCUP dataset and ES was supported by the Academic Enrichment Fund of the Department of Obstetrics and Gynecology of the McGill University Health Centre.

E.S., T.T., and H.N.K. designed the study. H.N.K. drafted the manuscript and extracted and managed data from the HCUP dataset. H.N.K. performed statistical analysis under the supervision of E.S. T.M. and T.T. provided input for the analysis. E.S., T.T, and T.M. critically reviewed the manuscript.

The data utilized in this study are obtained from the Healthcare Cost and Utilization Project (HCUP) NIS database. While this database is accessible through the HCUP-US website or the HCUP Central Distributor Online Reporting System, it is not hosted in a public repository with a DOI. Researchers who wish to access the HCUP NIS database can do so by submitting a request through the HCUP Central Distributor (available at https://www.hcup-us.ahrq.gov).