Increased Risk of Rheumatoid Arthritis Among Patients With Endometriosis

A Nationwide Population-Based Cohort Study

Yu-Hao Xue; Liang-Tian You; Hsin-Fu Ting; Yu-Wen Chen; Zi-Yun Sheng; Yi-Dong Xie; Yu-Hsun Wang; Jeng-Yuan Chiou; James Cheng-Chung Wei

Disclosures

Rheumatology. 2021;60(7):3326-3333.

Abstract and Introduction

Abstract

Objectives: Autoimmunity may play a role in endometriosis. The association between endometriosis and RA remains unknown. This study was conducted to identify any evidence for this relationship.

Methods: This 13-year, nationwide, population-based, retrospective cohort study analysed the risk of RA in a cohort of individuals with endometriosis. We investigated the incidence of RA among patients with endometriosis using data from the Longitudinal Health Insurance Database 2000, which is maintained by the Taiwan National Health Research Institutes. We used propensity scores to match comorbidities in the two cohorts. Kaplan–Meier analysis and Cox proportional hazard model were employed to analyse the association between endometriosis and RA among patients with different potential risks.

Results: Patients with endometriosis [adjusted hazard ratio (HR) 1.75, 95% CI 1.27, 2.41], aged ≥45 years (adjusted HR 1.50, 95% CI 1.06–2.13) and with autoimmune disease (adjusted HR 6.99, 95% CI 2.84–17.21) had a significantly higher risk of RA. The analyses also showed that when stratified by age, comorbidities and medication use, the risk of RA in patients with endometriosis was also higher than in those without endometriosis.

Conclusions: This 14-year, nationwide, population-based retrospective cohort study revealed that patients with endometriosis have a higher risk of RA. In the clinical management of patients with RA, rheumatologists should be especially mindful of the possibility of underlying endometriosis.

Introduction

Endometriosis is a common gynaecological disease, that is characterized by abnormal growth of endometrial cells outside the uterus. It can result in discomfort, and in the most severe cases, even the use of an analgesic does not adequately reduce the pain. The most widely accepted hypothesis for the pathophysiology of endometriosis is the retrograde menstruation hypothesis, whereby endometrial fragments reaching the pelvis via transtubal retrograde flow implant onto the peritoneum and abdominal organs, proliferate, and cause chronic inflammation with formation of adhesions.^[1–3] It has also been associated with infertility, although the mechanism is still not fully understood. In a systemic review, the first-degree relatives of women with severe endometriosis were found to have a six times greater risk for endometriosis than that of unaffected women.^[2,4] The prevalence of the symptoms of endometriosis in women experiencing pain, infertility or both is as high as 35–50%.^[2] The prevalence of endometriosis increases dramatically to as high as 25–50% in women with infertility, and 30–50% of women with endometriosis are infertile.^[5] Moreover, women with endometriosis are at higher risk of developing ovarian cancer compared with the general female population, and they also may be at increased risk of breast cancer and other cancers, as well as autoimmune and atopic disorders.^[3] In some patients with endometriosis, there is also a chronic local inflammatory process and presence of autoantibodies.^[6] Furthermore, endometriosis shares similarities with several autoimmune diseases, such as elevation in levels of cytokines, decreased apoptosis and cell-mediated abnormalities.^[7]

A prospective cohort study conducted as part of the Nurses' Health Study II showed that endometriosis was significantly associated with subsequent SLE and RA.^[8] However, no population-based studies have been conducted in an Asian country. Thus, we conducted a long-term nationwide cohort study by analysing data from Taiwan's Longitudinal Health Insurance Database 2000 (LHID2000) to assess the association between endometriosis and the subsequent development of RA.

1 2 3 4 5

Next Section

Abstract and Introduction
Methods
Results
Discussion
Conclusion
References
Sidebar

Table 1. Demographic characteristics of the endometriosis and non-endometriosis groups
	Before PSM				P-value	After PSM				P-value
	Endometriosis (N = 14 464)		Non-endometriosis (N = 86 784)			Endometriosis (N = 14 463)		Non-endometriosis (N = 14 463)
	n	%	n	%		n	%	n	%
Age (years)					1					0.110
20–44	10 880	75.2	65 280	75.2		10 879	75.2	10 761	74.4
≥45	3584	24.8	21 504	24.8		3584	24.8	3702	25.6
Mean (S.D.)	38.7 (8.7)		38.7 (8.7)		1	38.7 (8.7)		38.9 (8.7)		0.276
Hypertension	730	5.0	3053	3.5	<0.001	730	5.0	746	5.2	0.669
Hyperlipidaemia	293	2.0	1188	1.4	<0.001	292	2.0	287	2.0	0.834
Chronic liver disease	294	2.0	1036	1.2	<0.001	293	2.0	306	2.1	0.591
COPD	157	1.1	587	0.7	<0.001	157	1.1	133	0.9	0.157
Diabetes	318	2.2	1433	1.7	<0.001	318	2.2	320	2.2	0.936
Coronary artery disease	155	1.1	497	0.6	<0.001	154	1.1	142	1.0	0.483
Cerebrovascular disease	71	0.5	294	0.3	0.005	71	0.5	64	0.4	0.546
Autoimmune disease	74	0.5	243	0.3	<0.001	73	0.5	69	0.5	0.736
Corticosteroids	8141	56.3	35 439	40.8	<0.001	8140	56.3	8167	56.5	0.749
NSAIDs	11 009	76.1	48 865	56.3	<0.001	11 008	76.1	11 023	76.2	0.836
Hormonal medications	6165	42.6	10 477	12.1	<0.001	6164	42.6	6162	42.6	0.981

Bold font represents statistical significance (P < 0.05). PSM: propensity score-matched; COPD: chronic obstructive pulmonary disease.

Table 2. Cox proportional hazard model of RA risk in women with and without endometriosis
	No. of RA	Person-years	Incidence density (per 1000 person-years)	Crude HR	95% CI	Adjusted HR^a	95% CI
Endometriosis
No	62	118 863	0.5	1		1
Yes	95	105 888	0.9	1.75	1.27, 2.41	1.75	1.27, 2.41
Age (years)
20–44	105	171 275	0.6	1		1
≥45	52	53 476	1.0	1.60	1.15, 2.23	1.50	1.06, 2.13
Hypertension	9	10 422	0.9	1.26	0.64, 2.47	0.75	0.35, 1.58
Hyperlipidaemia	6	3844	1.6	2.32	1.03, 5.25	1.66	0.67, 4.07
Chronic liver disease	6	4578	1.3	1.91	0.84, 4.31	1.56	0.68, 3.60
COPD	4	2229	1.8	2.60	0.97, 7.03	2.07	0.75, 5.68
Diabetes	4	4501	0.9	1.29	0.48, 3.48	0.77	0.27, 2.21
Coronary artery disease	5	2260	2.2	3.22	1.32, 7.85	2.44	0.92, 6.51
Cerebrovascular disease	2	942	2.1	3.10	0.77, 12.52	2.39	0.57, 10.02
Autoimmune disease	5	984	5.1	7.55	3.10, 18.41	6.99	2.84, 17.21
CS	100	144 036	0.7	0.94	0.68, 1.31	0.77	0.55, 1.09
NSAIDs	141	187 871	0.8	1.63	0.97, 2.73	1.68	0.98, 2.88
Hormonal medications	78	107 782	0.7	1.04	0.76, 1.43	1.00	0.73, 1.38

^aBold font represents statistical significance (P < 0.05). Adjusted for age, hypertension, hyperlipidaemia, chronic liver disease, COPD, diabetes, coronary artery disease, cerebrovascular disease, autoimmune disease, CS, NSAIDs and hormonal medications. COPD: chronic obstructive pulmonary disease.

Table 3. Subgroup analysis of RA risk in women with and without endometriosis using a Cox proportional hazard model
	Endometriosis		Non-endometriosis		HR	95% CI
	N	No. of RA	N	No. of RA	HR	95% CI
Age (years)
20–44	10 879	65	10 761	40	1.85	1.24, 2.74
≥45	3584	30	3702	22	1.59	0.92, 2.76
						P for interaction = 0.735
Hypertension
No	13 733	92	13 717	56	1.87	1.34, 2.61
Yes	730	3	746	6	0.57	0.14, 2.30
						P for interaction = 0.114
Hyperlipidaemia
No	14 171	91	14 176	60	1.73	1.25, 2.40
Yes	292	4	287	2	2.16	0.39, 11.82
						P for interaction = 0.742
Chronic liver disease
No	14 170	92	14 157	59	1.78	1.29, 2.47
Yes	293	3	306	3	1.08	0.22, 5.38
						P for interaction = 0.579
COPD
No	14 306	95	14 330	58	1.87	1.35, 2.59
Yes	157	0	133	4	NA	NA
Diabetes
No	14 145	92	14 143	61	1.72	1.24, 2.38
Yes	318	3	320	1	3.46	0.36, 33.37
						P for interaction = 0.546
Coronary artery disease
No	14 309	93	14 321	59	1.80	1.30, 2.50
Yes	154	2	142	3	0.63	0.11, 3.79
						P for interaction = 0.323
Cerebrovascular disease
No	14 392	94	14 399	61	1.76	1.27, 2.43
Yes	71	1	64	1	1.13	0.07, 18.17
						P for interaction = 0.713
Autoimmune disease
No	14 390	93	14 394	59	1.80	1.30, 2.50
Yes	73	2	69	3	0.69	0.12, 4.15
						P for interaction = 0.341
CS
No	6323	33	6296	24	1.57	0.93, 2.65
Yes	8140	62	8167	38	1.82	1.22, 2.73
						P for interaction = 0.854
NSAIDs
No	3455	9	3440	7	1.38	0.51, 3.72
Yes	11 008	86	11 023	55	1.76	1.26, 2.47
						P for interaction = 0.871
Hormonal medications
No	8299	48	8301	31	1.76	1.12, 2.76
Yes	6164	47	6162	31	1.73	1.10, 2.72
						P for interaction = 0.894

Bold font represents statistical significance (P < 0.05). Hazard ratios (HR) was estimated by univariate Cox proportional hazard model. COPD: chronic obstructive pulmonary disease; NA: not applicable.