This cohort study assesses the association between the risk of aortic stenosis and rheumatoid arthritis.
Key Points
Question
Are people with rheumatoid arthritis (RA) at increased risk of developing aortic stenosis (AS)?
Findings
In this cohort study of 73 070 patients with RA who were matched with 639 268 patients without RA, those with RA had an increased risk of developing AS, undergoing aortic valve intervention, and AS-related death.
Meaning
Findings of this study suggest that RA is associated with a higher risk of developing AS as well as the subsequent risks of aortic valve intervention and AS-related death.
Abstract
Importance
Although an increased risk of ischemic cardiovascular disease has been associated with rheumatoid arthritis (RA), the risk of aortic stenosis (AS) is unknown.
Objective
To examine the risk of incident AS, aortic valve intervention, AS-related death, and risk factors for AS development in patients with RA.
Design, Setting, and Participants
This cohort study linked data from the Veterans Health Administration (VHA) and Centers for Medicare & Medicaid Services from 2000 to 2019. Patients with RA were matched by age, sex, and VHA enrollment year with up to 10 patients without RA. The cohort was followed until incident AS, aortic valve intervention, or death. Data were analyzed from August 23, 2022, to March 3, 2023.
Exposures
the primary exposure was the presence of RA, defined using validated RA algorithms.
Main Outcomes and Measures
Aortic stenosis was defined as a composite of inpatient or outpatient diagnoses, surgical or transcatheter aortic valve replacement, or AS-related death using diagnostic and procedural codes. Risk of AS development was assessed with multivariable Cox proportional hazards models adjusted for race, ethnicity, smoking status, body mass index, rurality, comorbidities, and health care use.
Results
The cohort included 73 070 patients with RA (64 008 [87.6%] males; mean [SD] age, 63.0 [11.9] years) matched with 639 268 patients without RA (554 182 [86.7%] males; mean [SD] age, 61.9 [11.7] years) and 16 109 composite AS outcomes that occurred over 6 223 150 person-years. The AS incidence rate was 3.97 (95% CI, 3.81-4.13) per 1000 person-years in patients with RA and 2.45 (95% CI, 2.41-2.49) per 1000 person-years in the control patients (absolute difference, 1.52 per 1000 person-years). Rheumatoid arthritis was associated with an increased risk of composite AS (adjusted hazard ratio [AHR], 1.48; 95% CI, 1.41-1.55), aortic valve intervention (AHR, 1.34; 95% CI, 1.22-1.48), and AS-related death (AHR, 1.26; 95% CI, 1.04-1.54).
Conclusions and Relevance
In this cohort study, RA was associated with a higher risk of developing AS and the subsequent risks of undergoing aortic valve intervention and suffering from AS-related death. Future studies are needed to confirm whether valvular heart disease, specifically AS, may be an overlooked cardiovascular disease complication in RA.
Introduction
Cardiovascular disease (CVD) is the most frequent cause of death and accounts for the largest proportion of excess mortality in patients with rheumatoid arthritis (RA).1,2 A substantial body of evidence supports an association between the inflammatory milieu of RA and CVD risk through oxidative stress, endothelial and lipoprotein dysfunction, and posttranslational modifications of peptides and subsequent immune responses.3 Although most studies examining these associations have focused on atherosclerotic heart disease and heart failure, preclinical studies have also reported on the development of valvular carditis in transgenic mice who develop autoantibody-mediated inflammatory arthritis similar to RA.4 Additionally, histopathologic analysis of aortic valves in patients without RA and in animal models suggests a pathogenic role of lipoprotein deposition and chronic inflammation in the development of calcific aortic stenosis (AS).5,6,7
In a recent analysis of temporal trends in RA-related cause-specific mortality risk from our research group, contemporary RA cohorts had a persistently increased risk of CVD-related mortality.8 This included a 75% increased risk of valvular heart disease–related death, the most overrepresented cause of CVD death in RA. However, valvular heart disease is rarely described as an extra-articular manifestation of RA, and investigations in RA have largely been limited to small echocardiographic studies describing the prevalence of valvular abnormalities without longitudinal outcomes.9,10,11 Aortic stenosis is the most common reason for valve replacement and the most frequent cause of valvular heart disease–related death in the US.12 Furthermore, previous case series in RA describe aortitis with adjacent aortic valvulitis and that the aortic valve was the most commonly afflicted valve in previous cross-sectional studies of patients with RA.9,11,13,14 In this study, we evaluated the risk of AS and AS-related outcomes in a large, national matched cohort of patients with RA vs patients without RA, testing the hypothesis that RA is associated with a higher risk of incident AS, aortic valve intervention, and AS-related death. Additionally, we examined risk factors for AS development in patients with RA.
Methods
Study Design and Sample
We conducted a retrospective, matched cohort study leveraging national Veterans Health Administration (VHA) and Centers for Medicare & Medicaid Services (CMS) data between January 1, 2000, and December 31, 2019. The VHA is the largest integrated health system in the US with robust administrative and electronic health record data available within the Veterans Affairs (VA) Informatics and Computing Infrastructure.15 Linkage with CMS data facilitated capture of inpatient and outpatient claims for Medicare enrollees who obtained care outside of the VHA. The VA Nebraska-Western Iowa Health Care System Institutional Review Board approved this study and waived patient informed consent because there was no prospective data collection or patient contact. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Patients with RA were identified by querying the VA Corporate Data Warehouse (CDW) using validated administrative algorithms requiring the presence of 2 or more diagnostic codes for RA, a rheumatologist diagnosis of RA, and either a positive test result for RA autoantibody (rheumatoid factor [RF] or anti–cyclic citrullinated protein [anti-CCP] antibody) or dispensation of a disease-modifying antirheumatic drug (DMARD).16 Patients with RA were subsequently matched with up to 10 patients without RA by birth year, sex, and year of enrollment into the VHA system. The cohort entry date was the date the RA algorithm was fulfilled (corresponding calendar date for patients without RA). Patients were excluded if they had 1 or more AS diagnostic code or procedural code for an aortic valve intervention prior to study entry.17 Patients were followed until the earliest of the composite AS outcome, death, or end of study period.
AS Outcome Assessment
The primary outcome was incident AS, defined as a composite of inpatient or outpatient AS diagnosis, aortic valve intervention, or AS-related death. Secondary outcomes were the individual aortic valve interventions and AS-related death outcomes. Aortic stenosis outcomes were identified in VHA and CMS data using an administrative algorithm adapted from validated diagnostic and procedural codes.18,19,20 Inpatient AS was identified as a hospital discharge diagnosis using International Classification of Diseases, 9th Revision (ICD-9) and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes. Outpatient AS diagnosis required 2 or more ICD-9 or ICD-10 codes on separate dates plus a Current Procedural Terminology code for a transthoracic echocardiogram. The ICD-9 and ICD-10 procedure codes were used to identify aortic valve interventions, including surgical aortic valve replacement (SAVR) and transcatheter aortic valve replacement (TAVR). Vital status (ie, date of death) was assessed through linkage with the National Death Index, defining AS-related death when an ICD-10 code for AS was listed as the underlying cause of death. Diagnostic and procedural codes for AS and aortic valve interventions are detailed in eTable 1 in Supplement 1.
Covariates and AS Risk Factors
Covariates were selected a priori and obtained from the CDW. Demographic characteristics, including age, sex, self-reported race and ethnicity, and urban-rural residency, were collected from VHA enrollment records. Race and ethnicity were included because these social constructs may influence the frequency of health care use and the likelihood of detecting the primary outcome. Body mass index (BMI) was categorized by World Health Organization cutoffs, using weight at the most recent VHA encounter preceding study entry and modal height, as previously described.21,22 Smoking status (current, former, or never) was identified using data from the electronic health record.23 Comorbidity burden was quantified using the Rheumatic Disease Comorbidity Index (RDCI), a validated index that includes lung disease, CVD (hypertension, myocardial infarction, stroke, or other CVD), diabetes, cancer, depression, fracture, and gastric disease.24 A higher score on the RDCI indicates greater comorbidity. These individual comorbidities were defined by 2 or more diagnostic codes on separate days prior to cohort entry. Although AS contributes to the other CVD category, patients with AS were excluded prior to RDCI assessment. Health care use was defined as the number of outpatient visits and hospitalizations in the year preceding the study entry date.
To examine AS risk factors among patients with RA, we extracted RA-related autoantibody test results (RF and anti-CCP) and erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level measurements from the CDW. Values for ESR and CRP level were considered elevated if they exceeded the upper limit of normal per the conducting laboratory. Prescription fills of conventional synthetic DMARDs (csDMARDs), biologic and targeted synthetic DMARDs (b/tsDMARDs), and glucocorticoids at any time prior to cohort entry were defined using pharmacy records within the CDW.
Statistical Analysis
Baseline characteristics were compared using descriptive statistics, and unadjusted incidence rates were estimated and stratified by RA status. Cox proportional hazards models, stratified on matched sets, were used to estimate the RA-related risk of the primary and secondary AS outcomes. Multivariable models were adjusted for race, ethnicity, BMI, smoking status, RDCI, rurality, and health care use (number of outpatient visits and hospitalizations). Race was categorized as Black, White, other (American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, or multiracial), or unknown. Sensitivity analyses were performed in an incident RA cohort who received at least 365 days of care in the VHA before an RA diagnosis code was placed or DMARD dispense occurred,17 as well as among patients 65 years or older, a Medicare-age population for which we expect VHA and CMS data to have near-complete capture of events. In addition to adjusting for health care use in primary analyses, we further examined the potential for ascertainment bias by stratifying models by quartiles of outpatient visits among patients with RA.
In secondary analyses, AS risk was assessed in models stratified by seropositivity and the presence of elevated inflammatory measures at baseline. To examine temporal trends in RA-related AS risk, analyses were stratified by RA diagnosis year (2000-2005, 2006-2011, and 2012 or later) in the incident RA cohort. Among patients with RA, we evaluated risk factors for AS incorporating the above covariates and RA-related factors including seropositivity, systemic inflammation, and DMARDs (including glucocorticoids) in a multivariable Cox proportional hazards regression model. A 2-tailed P < .05 was considered statistically significant. All analyses were performed between August 23, 2022, and March 3, 2023, using Stata, Version 17 (StataCorp LLC).
Results
Baseline Characteristics
The study cohort consisted of 73 070 patients with RA who were matched with 639 268 patients without RA. Patients were predominantly male (64 008 males [87.6%] and 9 062 females [12.4%] in the RA group and 554 182 males [86.7%] and 85 086 females [13.3%] in the control group; Table 1), with a mean (SD) age of 63.0 (11.9) years in the RA group and 61.9 (11.7) years in the control group. White race was predominant in both groups (52 180 patients [72.3%] and 394 531 patients [61.7%] in the control group), although information on race was more frequently missing among the control patients (129 462 patients [11.6%] with RA vs 8 498 [20.3%] in the control group). Patients with RA more frequently had ever-smoker status (80.8% vs 67.7%) and had a higher degree of comorbidity (mean [SD] RDCI, 1.7 [1.6] vs 1.2 [1.5]). A higher proportion of patients with RA vs without received an echocardiogram (51.0% vs 43.0%, standardized difference 0.16). Among patients with RA, 68.9% had seropositive test results for RF or anti-CCP and 47.0% had an elevated baseline ESR or CRP level. At baseline, most patients with RA had previously received a csDMARD (92.1%), whereas 18.7% had received a b/tsDMARD and 42.6% had received glucocorticoids.
Table 1. Baseline Characteristics of Patients With and Without Rheumatoid Arthritis (RA).
| Characteristic | With RA, No. (%) (n = 73 070) | Without RA, No. (%) (n = 639 268) | Standardized difference |
|---|---|---|---|
| Age, mean (SD), y | 63.0 (11.9) | 61.9 (11.7) | 0.09 |
| Sex | |||
| Female | 9062 (12.4) | 85 086 (13.3) | 0.03 |
| Male | 64 000 (87.6) | 554 182 (86.7) | |
| Race | |||
| Black | 9715 (13.3) | 98 062 (15.3) | 0.26 |
| White | 52 810 (72.3) | 394 531 (61.7) | |
| Othera | 2047 (2.8) | 17 213 (2.7) | |
| Unknown | 8498 (11.6) | 129 462 (20.3) | |
| Ethnicity | |||
| Hispanic | 3426 (4.7) | 23 504 (3.7) | 0.24 |
| Non-Hispanic | 62 617 (85.7) | 503 154 (78.7) | |
| Unknown | 7027 (9.6) | 112 610 (17.6) | |
| Smoking status | |||
| Current | 34 867 (47.7) | 258 190 (40.4) | 0.36 |
| Former | 24 164 (33.1) | 174 433 (27.3) | |
| Never | 10 620 (14.59) | 118 187 (18.5) | |
| Unknown | 3419 (4.7) | 88 458 (13.8) | |
| BMI category | |||
| <20 | 1294 (1.8) | 10 115 (1.6) | 0.38 |
| 20-24.9 | 8213 (11.2) | 68 869 (10.8) | |
| 25-29.9 | 22 669 (31.0) | 191 629 (30.0) | |
| 30-34.9 | 20 847 (28.5) | 171 215 (26.8) | |
| 35-39.9 | 11 144 (15.3) | 86 374 (13.5) | |
| ≥40 | 8576 (11.7) | 60 947 (9.5) | |
| Unknown | 327 (0.4) | 50 119 (7.8) | |
| RDCI score, mean (SD) | 1.7 (1.6) | 1.2 (1.5) | 0.34 |
| Rurality | |||
| Urban | 44 439 (60.8) | 409 352 (64.0) | 0.07 |
| Rural | 27 177 (37.2) | 217 456 (34.0) | |
| Highly rural | 1213 (1.7) | 9751 (1.5) | |
| Unknown | 241 (0.3) | 2709 (0.4) | |
| Health care use, mean (SD) | |||
| Outpatient visits in prior 12 mo | 16.9 (19.1) | 6.4 (12.4) | 0.79 |
| Hospitalizations in prior 12 mo | 0.2 (0.6) | 0.1 (0.5) | 0.16 |
| RA-related factors | |||
| Seropositivity (RF or anti-CCP)b | 42 248 (68.9) | NA | NA |
| Elevated ESR or CRP levelc | 29 053 (47.0) | NA | |
| Baseline csDMARD use | 67 288 (92.1) | NA | |
| Baseline b/tsDMARD use | 13 666 (18.7) | NA | |
| Baseline glucocorticoid use | 31 105 (42.6) | NA |
Abbreviations: anti-CCP, anti-cyclic citrullinated peptide; b/tsDMARD, biologic and targeted synthetic disease-modifying antirheumatic drug; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CRP, C-reactive protein; csDMARD, conventional synthetic disease-modifying antirheumatic drug; ESR, erythrocyte sedimentation rate; NA, not applicable; RDCI, rheumatic disease comorbidity index; RF, rheumatoid factor.
Other race includes American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, and multiracial.
There were 11 736 patients with missing RF and anti-CCP values at baseline.
There were 11 253 patients with missing ESR or CRP measurements at baseline.
AS Incidence Rates
We identified 16 109 composite AS outcomes over 6 223 150 person-years, of which 2303 occurred in patients with RA. Mean (SD) follow-up time was 7.9 (5.3) years in patients with RA and 8.8 (5.6) years in patients without RA. Most clinical AS diagnoses (79%) were identified in VHA data, whereas valvular interventions (86%) were more often identified in CMS data (eTable 2 in Supplement 1). Mean (SD) age at AS onset was similar in patients with RA (74.9 [9.7] years) and without RA (75.5 [9.1] years). Patients with RA experienced an absolute risk increase of 1.52 composite AS events per 1000 person-years (composite AS events per 1000 person-years, 3.97 [95% CI, 3.81-4.13] in patients with RA vs 2.45 [95% CI, 2.41-2.49] in control patients), which included higher rates of AS-related interventions and AS-related death (Table 2).
Table 2. Aortic Stenosis (AS) Incidence Rates in Patients With and Without Rheumatoid Arthritis (RA).
| Outcome | With RA (n = 73 070) | Without RA (n = 639 268) | ||||
|---|---|---|---|---|---|---|
| Patients, No. (%) | Person-years of follow-up | IR per 1000 person-years (95% CI) | Patients, No. (%) | Person-years of follow-up | IR per 1000 person-years (95% CI) | |
| Composite ASa | 2303 (3.2) | 580 108 | 3.97 (3.81-4.13) | 13 806 (2.2) | 5 643 042 | 2.45 (2.41-2.49) |
| Aortic valve interventionb | 601 (0.8) | 586 444 | 1.02 (0.95-1.11) | 4507 (0.7) | 5 682 498 | 0.79 (0.77-0.82) |
| SAVR | 428 (0.6) | 586 732 | 0.73 (0.66-0.80) | 3291 (0.5) | 5 684 893 | 0.58 (0.56-0.60) |
| TAVR | 178 (0.2) | 588 659 | 0.30 (0.26-0.35) | 1241 (0.2) | 5 700 887 | 0.22 (0.21-0.23) |
| AS-related death | 134 (0.2) | 808 129 | 0.17 (0.14-0.20) | 817 (0.1) | 7 254 824 | 0.11 (0.10-0.12) |
Abbreviations: IR, incidence rate; SAVR, surgical aortic valve replacement; TAVR, transcatheter aortic valve replacement.
Composite AS consists of inpatient or outpatient AS diagnosis, aortic valve intervention, and AS-related death.
Incidence rate examines first of either SAVR or TAVR.
Risk of Composite AS Outcomes
Compared with matched control patients, patients with RA had an increased risk of the composite AS outcome (hazard ratio [HR], 1.66; 95% CI, 1.59-1.74). After adjusting for potential confounders, an increased risk of incident AS persisted (adjusted HR [AHR], 1.48; 95% CI, 1.41-1.55) (Figure). Rheumatoid arthritis was also associated with a higher risk of undergoing an AS-related valvular intervention (AHR, 1.34; 95% CI, 1.22-1.48), including SAVR (AHR, 1.30; 95% CI, 1.16-1.45) and TAVR (AHR, 1.53; 95% CI, 1.26-1.85). Patients with RA had an increased risk of death related to AS (AHR, 1.26; 95% CI, 1.04-1.54) compared with patients without RA. Aortic stenosis risk was similar in sensitivity analyses restricted to patients older than 65 years (eFigure 1 in Supplement 1) and among incident cases of RA (eFigure 2 in Supplement 1), although the risk of TAVR and AS-related death did not reach statistical significance in the incident RA cohort given the smaller sample size. The risk of AS was higher in patients with RA who had an elevated ESR or CRP level at baseline (AHR, 1.56; 95% CI, 1.44-1.70) compared with those without ESR or CRP levels elevations at baseline (HR, 1.41; 95% CI, 1.30-1.53) and among patients with seronegative results for RA autoantibodies (AHR, 1.62; 95% CI, 1.47-1.79) vs those with seropositive test results (AHR, 1.44; 95% CI, 1.35-1.54), although these differences were not statistically significant. Temporal trends in risk of the composite AS outcome was similar across RA diagnosis time periods, with AHR ranging from 1.40 (95% CI, 1.20-1.63) in 2000 to 2005 to 1.51 (95% CI, 1.27-1.80) in 2012 or later (eFigure 3 in Supplement 1). In addition, the risk of the composite AS outcome defined by outpatient health care use was similar across quartiles (AHR range, 1.31 [95% CI, 1.16-1.49] to 1.57 [95% CI, 1.43-1.64]) (eFigure 4 in Supplement 1).
Figure. Risk of Aortic Stenosis, Aortic Valve Intervention, and Aortic Stenosis–Related Death in Patients With Rheumatoid Arthritis.
Forest plot illustrates the adjusted hazard ratios (AHRs) and 95% CIs for risk of incident aortic stenosis and aortic stenosis–related outcomes in patients with rheumatoid arthritis compared with matched patients without rheumatoid arthritis. Estimates are from multivariable Cox proportional hazards regression models, stratified on matched pairs and adjusted for race, ethnicity, body mass index, smoking status, comorbidity burden, and health care use. SAVR indicates surgical aortic valve replacement; TAVR, transcatheter aortic valve replacement.
AS Risk Factors in Patients With RA
Baseline characteristics of patients with RA stratified by AS development during follow-up are listed in Table 3. Age (risk quantified per 5-y increase) and BMI >30 were independently associated with a stepwise increased risk of AS (Table 4). Females and patients with Black race or Hispanic ethnicity had a lower risk of developing AS. Hypertension, stroke, and other noncoronary CVDs were associated with incident AS, whereas coronary artery disease, diabetes, and lung disease were not. Among RA-related risk factors, elevated ESR or CRP level (AHR, 1.11; 95% CI, 1.01-1.22) and baseline use of b/tsDMARDs (AHR, 1.22; 95% CI, 1.07-1.39) or glucocorticoids (AHR, 1.19; 95% CI, 1.08-1.32) were associated with a higher risk of incident AS. Seropositivity was not associated with AS development (AHR, 0.94; 95% CI, 0.86-1.04).
Table 3. Baseline Characteristics of Patients With Rheumatoid Arthritis (RA) Stratified by the Development of Aortic Stenosis (AS).
| Characteristic | With AS, No. (%) (n = 2303) | Without AS, No, % (n = 70 767) |
|---|---|---|
| Age, mean (SD), y | 68.5 (10.0) | 62.8 (11.9) |
| Sex | ||
| Female | 118 (5.1) | 8944 (12.6) |
| Male | 2185 (94.9) | 61 823 (87.4) |
| Race | ||
| Black | 153 (6.6) | 9562 (13.5) |
| White | 1877 (81.5) | 50 933 (72.0) |
| Othera | 42 (1.8) | 2005 (2.8) |
| Unknown | 231 (10.0) | 8267 (11.7) |
| Ethnicity | ||
| Hispanic | 67 (2.9) | 3359 (4.7) |
| Non-Hispanic | 2037 (88.4) | 60 580 (85.6) |
| Unknown | 199 (8.6) | 6828 (9.6) |
| Smoking status | ||
| Current | 874 (38.0) | 33 993 (48.0) |
| Former | 1029 (44.7) | 23 135 (32.7) |
| Never | 295 (12.8) | 10 325 (14.6) |
| Unknown | 105 (4.6) | 3314 (4.7) |
| BMI category | ||
| <20 | 37 (1.6) | 1257 (1.8) |
| 20-24.9 | 213 (9.2) | 8000 (11.3) |
| 25-29.9 | 728 (31.6) | 21 941 (31.0) |
| 30-34.9 | 695 (30.2) | 20 152 (28.5) |
| 35-39.9 | 328 (14.2) | 10 816 (15.3) |
| ≥40 | 299 (13.0) | 8277 (11.7) |
| Unknown | 3 (0.1) | 324 (0.5) |
| Comorbidities | ||
| RDCI score, mean (SD) | 1.7 (1.6) | 1.7 (1.6) |
| Hypertension | 1242 (53.9) | 34 244 (48.4) |
| Diabetes | 504 (21.9) | 14 247 (20.1) |
| CAD | 85 (3.7) | 2035 (2.9) |
| Stroke | 78 (3.4) | 1739 (2.5) |
| Other CVD | 711 (30.9) | 16 379 (23.1) |
| Lung disease | 302 (13.1) | 10 986 (15.5) |
| Rurality | ||
| Urban | 1377 (59.8) | 43 062 (60.9) |
| Rural | 882 (38.3) | 26 295 (37.2) |
| Highly rural | 38 (1.7) | 1175 (1.7) |
| Unknown | 6 (0.3) | 235 (0.3) |
| Health care use, mean (SD) | ||
| Outpatient visits in prior 12 mo | 17.0 (19.7) | 16.8 (19.1) |
| Hospitalizations in prior 12 mo | 0.2 (0.6) | 0.2 (0.7) |
| RA-related factors | ||
| Seropositivity (RF or anti-CCP)b | 1228 (65.3) | 41 020 (69.0) |
| Elevated ESR or CRP levelc | 866 (48.9) | 28 187 (46.9) |
| Baseline csDMARD use | 2126 (92.3) | 65 162 (92.1) |
| Baseline b/tsDMARD use | 297 (12.9) | 13 369 (18.9) |
| Baseline glucocorticoid use | 868 (37.7) | 30 237 (42.7) |
Abbreviations: anti-CCP, anti–cyclic citrullinated peptide; b/tsDMARD, biologic and targeted synthetic disease-modifying antirheumatic drug; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CAD, coronary artery disease; CRP, C-reactive protein; csDMARD, conventional synthetic disease-modifying antirheumatic drug; CVD, cardiovascular disease; ESR, erythrocyte sedimentation rate; RDCI, rheumatic disease comorbidity index; RF, rheumatoid factor.
Other race includes American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, and multiracial.
There were 11 736 patients with missing RF and anti-CCP values at baseline (422 in those who developed aortic stenosis).
There were 11 253 patients with missing ESR or CRP measurements at baseline (531 in those who developed aortic stenosis).
Table 4. Risk Factors for Incident Aortic Stenosis in Patients With Rheumatoid Arthritis.
| Risk factora | AHR (95% CI) | P value |
|---|---|---|
| Age, per 5 y | 1.42 (1.38-1.45) | <.001 |
| Sex | ||
| Female | 0.66 (0.55-0.80) | <.001 |
| Male | 1 [Reference] | NA |
| Race | ||
| Black | 0.60 (0.51-0.71) | <.001 |
| White | 1 [Reference] | NA |
| Otherb | 0.75 (0.55-1.02) | .06 |
| Unknown | 0.79 (0.66-0.94) | <.001 |
| Ethnicity | ||
| Hispanic | 0.69 (0.54-0.89) | <.01 |
| Non-Hispanic | 1 [Reference] | NA |
| Unknown | 1.05 (0.87-1.27) | .06 |
| Smoking status | ||
| Never | 1 [Reference] | NA |
| Current | 0.97 (0.84-1.11) | .63 |
| Former | 1.03 (0.90-1.17) | .69 |
| Unknown | 1.12 (0.89-1.40) | .35 |
| BMI category | ||
| <20 | 1.23 (0.86-1.74) | .25 |
| 20-24.9 | 1 [Reference] | NA |
| 25-29.9 | 1.14 (0.98-1.34) | .25 |
| 30-34.9 | 1.39 (1.19-1.63) | <.001 |
| 35-39.9 | 1.44 (1.21-1.72) | <.001 |
| ≥40 | 1.72 (1.44-2.07) | <.001 |
| Unknown | 0.55 (0.17-1.72) | .30 |
| Seropositive for RF or anti-CCP | ||
| No | 1 [Reference] | NA |
| Yes | 0.94 (0.86-1.04) | .24 |
| Unknown | 0.94 (0.82-1.07) | .35 |
| Medications at baselinec | ||
| csDMARDs | 1.04 (0.88-1.22) | .63 |
| b/tsDMARDs | 1.22 (1.07-1.39) | <.01 |
| Glucocorticoids | 1.19 (1.08-1.32) | <.01 |
| ESR or CRP level | ||
| Not elevated | 1 [Reference] | NA |
| Elevated | 1.11 (1.01-1.22) | .03 |
| Unknown | 1.08 (0.96-1.22) | .18 |
| Comorbiditiesd | ||
| Hypertension | 1.20 (1.09-1.31) | <.001 |
| Diabetes | 1.09 (0.98-1.21) | .09 |
| CAD | 1.11 (0.89-1.40) | .36 |
| Stroke | 1.32 (1.04-1.66) | .02 |
| Other CVD | 1.27 (1.15-1.41) | <.001 |
| Lung disease | 0.94 (0.82-1.06) | .32 |
Abbreviations: AHR, adjusted hazard ratio; anti-CCP, anti–cyclic citrullinated peptide; b/tsDMARDs, biologic and targeted synthetic disease-modifying antirheumatic drugs; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CAD, coronary artery disease; CRP, C-reactive protein; csDMARDs, conventional synthetic disease-modifying antirheumatic drugs; CVD, cardiovascular disease; ESR, erythrocyte sedimentation rate; NA, not applicable; RF, rheumatoid factor.
Variables included in a multivariable Cox proportional hazards regression model restricted to patients with rheumatoid arthritis. Survival outcome was composite aortic stenosis events (incident aortic stenosis, aortic valve intervention, or aortic stenosis–related death).
Other race includes American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, and multiracial.
Risk compared with not receiving the indicated drug.
Risk compared with not having the indicated comorbidity.
Discussion
In this large, national cohort study that linked VHA and CMS data to maximize capture of AS diagnoses and procedures, patients with RA had an absolute risk increase of 1.52 composite AS events per 1000 person-years compared with patients without RA. Additionally, patients with RA had a higher risk for aortic valve replacement and AS-related death. These findings support those of preclinical studies depicting the role of immune and inflammatory mediators in the development of valvular carditis.4,25,26 Together, these data support the view that valvular heart disease, specifically AS, may be a CVD complication in RA.
Several studies have reported an association between CVD risk and RA, with a meta-analysis27 estimating a 1.5-to 2-fold higher risk of incident CVD, including myocardial infarction, stroke, and heart failure, in patients with RA. Despite a prevalence in the general population similar to that of heart failure,12 AS was not evaluated in that meta-analysis.27 Small, cross-sectional studies have described echocardiographic findings of calcifications, thickening, and nodules as the most common valvular abnormalities in RA.9,10,11 However, these studies included asymptomatic younger patients (mean ages, 47 to 51 years) without a control group or longitudinal follow-up, leaving the clinical implications of these findings unclear since AS typically affects adults older than 65 years.12 In this cohort of patients with RA (mean age, 63 years) and matched control patients followed up for more than 6 000 000 person-years, patients with RA had a 48% increased risk of AS that was independent of traditional CVD risk factors.
After linking with CMS data to capture valvular interventions performed outside the VHA, RA was associated with an increased risk of any aortic valve intervention (AHR, 1.34; 95% CI, 1.22-1.48), including both SAVR (AHR, 1.30; 95% CI, 1.16-1.45) and TAVR (AHR, 1.53; 95% CI, 1.26-1.85), suggesting a more severe AS disease course that required more frequent procedural intervention. This is of particular concern since patients with RA who underwent SAVR experienced poorer postoperative outcomes in 1 study.28 Aggregate inpatient costs per patient undergoing SAVR or TAVR exceeded $60 000 at 6 to 12 months postoperatively.18,29
We found that patients with RA had an increased risk (AHR, 1.26; 95% CI, 1.04-1.54) of AS-related death, a mortality risk similar to that of ischemic heart disease in previous RA cohorts.8,27 These data support a more severe AS disease course in contrast to a smaller study that reported similar AS progression rates in patients with RA and control patients, although that study had a shorter follow-up time and did not examine clinical AS outcomes.30 The findings from the present study emphasize that valvular heart disease may be an underrecognized contributor to the persistent CVD-related mortality gap in RA,8 particularly given the lack of improvement in AS-specific risk over time.
Mechanistic associations between RA disease activity, systemic inflammation, and CVD have been reported in several previous studies.3,31 Although valvular disease is less frequently studied than atherosclerotic CVD in this context, animal models of autoantibody-induced inflammatory arthritis have found activation of valve-resident cardiac phagocytes by circulating autoantibodies, leading to valvular carditis through the recruitment of CD4+ T-cells and macrophages.25,26 In an immunohistochemical study7 of explanted aortic valves from patients without RA but with AS, macrophage and T-cell infiltrates were found, supporting the contribution of an inflammatory process to what has been traditionally considered degenerative AS. Among the RA-related factors examined in the present study, we observed a modestly increased risk for AS among patients with RA who have elevated ESR or CRP level at baseline, suggesting that more severe RA disease activity may be associated with a higher risk of developing AS. Further supporting this finding, we observed associations between the receipt of more aggressive RA treatments, including b/tsDMARDs and glucocorticoids, with incident AS. These data were in line with results of preclinical studies4,5 suggesting that higher degrees of systemic inflammation were a risk factor for AS in patients with RA and a factor in AS progression in a prior RA cohort.30 In addition to inflammation due to RA disease activity, patients with RA are disproportionately affected by multimorbidity and metabolic syndrome, which may contribute to higher systemic inflammation and poorer AS-related outcomes.32,33,34 Continued research is warranted to elucidate the inflammatory pathways mediating AS risk and severity in RA cohorts.
In addition to RA-related factors, we examined traditional CVD risk factors as risk factors for AS in patients with RA. Consistent with previously reported national health statistics,12 we found that male sex and hypertension were independently associated with AS onset in patients with RA, whereas other traditional risk factors, including smoking status, diabetes, and coronary artery disease, were not. In contrast to the so-called obesity paradox in RA in which patients with underweight have a greater risk of CVD mortality compared with patients with overweight or obesity,35,36,37,38 we found that BMI ≥30 was associated with AS and AS-related death. This finding suggests a unique interaction between obesity and AS, potentially through pathways mediated by oxidized lipoproteins.39 Further study delineating the contribution of metabolic syndrome to AS-related risk in this population is warranted.
Limitations
This study has limitations. While we observed an association between RA status and AS risk, the absolute risk differences were modest given the infrequency of AS-related events. Consistent with the VHA population, the cohort was predominantly male. This may limit the generalizability of the results because RA is more common in females.40 While misclassification of RA status is possible using administrative algorithms, the algorithms used had a high positive predictive value (>90%) and misclassification in this regard would bias results to the null. Similarly, misclassification of AS and AS-related death may have occurred using diagnostic codes. However, we adapted previously validated diagnostic and procedural codes and anticipated any misclassification to be nondifferential between patients with or without RA. We were unable to account for bicuspid aortic valves using diagnostic codes, which could in part explain the lower AS risk among Black patients with RA, given previous reports that bicuspid aortic valves are less common in Black patients.41 There was risk for residual confounding since we did not capture time-varying comorbidities and health behaviors as well as unmeasured confounders, such as activity intolerance, that might have led to a higher rate of testing and AS detection. As an AS diagnosis may be made on echocardiographic findings alone, a high-risk population receiving echocardiograms more often or for alternative indications (eg, ischemic heart disease, pericarditis, or cardiomyopathy) could have led to an overestimate of AS risk in patients with RA. To mitigate this risk, we leveraged national VHA data to account for several traditional CVD risk factors and adjusted for health care use (number of outpatient visits and hospitalizations). Furthermore, analyses stratified by frequency of health care use showed similar estimates for RA-related AS risk. Additionally, we observed an increased risk of aortic valve interventions, supporting the validity of the findings given the infrequency with which incidentally discovered AS requires procedural intervention. While the findings suggest a more severe AS disease course in patients with RA, we did not have data on the severity or progression of aortic valve dynamics. Although associations between AS development and elevated inflammatory markers and baseline glucocorticoid and b/tsDMARD use suggest that more severe RA may confer a higher risk of incident AS, we did not have longitudinal RA disease activity measures or medication courses to evaluate causal associations underlying these findings.
Conclusions
In this cohort study of VHA and CMS linked data from patients with RA who were matched with control patients, we found that patients with RA had an absolute risk increase of developing AS of 1.52 per 1000 person-years, more frequently needed aortic valve replacement, and experienced more frequent AS-related death. Future studies are needed to confirm whether valvular heart disease, specifically AS, may be an overlooked CVD complication in RA.
eTable 1. Diagnostic and Procedural Codes Used to Identify Incident Aortic Stenosis in US Veterans With and Without Rheumatoid Arthritis
eTable 2. Databases Contributing Events to the Composite Aortic Stenosis Outcome
eFigure 1. Sensitivity Analysis Examining Aortic Stenosis Risk in Rheumatoid Arthritis (N=34 012) vs. Matched Controls (N=274 300), Restricting to Patients Over 65 Years of Age
eFigure 2. Sensitivity Analysis Examining Aortic Stenosis Risk Amongst Incident Rheumatoid Arthritis Cases (N=32 369) vs Matched Controls (N=264 240)
eFigure 3. Temporal Trends in the Risk of the Composite Aortic Stenosis Outcome
eFigure 4. Aortic Stenosis Risk in Rheumatoid Arthritis vs Matched Controls, Stratified by Frequency of Outpatient Care
Data Sharing Statement
References
- 1.England BR, Sayles H, Michaud K, et al. Cause-specific mortality in male US veterans with rheumatoid arthritis. Arthritis Care Res (Hoboken). 2016;68(1):36-45. doi: 10.1002/acr.22642 [DOI] [PubMed] [Google Scholar]
- 2.Sokka T, Abelson B, Pincus T. Mortality in rheumatoid arthritis: 2008 update. Clin Exp Rheumatol. 2008;26(5)(suppl 51):S35-S61. [PubMed] [Google Scholar]
- 3.England BR, Thiele GM, Anderson DR, Mikuls TR. Increased cardiovascular risk in rheumatoid arthritis: mechanisms and implications. BMJ. 2018;361:k1036. doi: 10.1136/bmj.k1036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Breed ER, Binstadt BA. Autoimmune valvular carditis. Curr Allergy Asthma Rep. 2015;15(1):491. doi: 10.1007/s11882-014-0491-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bourgeois R, Bourgault J, Despres AA, et al. Lipoprotein proteomics and aortic valve transcriptomics identify biological pathways linking lipoprotein(a) levels to aortic stenosis. Metabolites. 2021;11(7):459. doi: 10.3390/metabo11070459 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lindman BR, Clavel MA, Mathieu P, et al. Calcific aortic stenosis. Nat Rev Dis Primers. 2016;2:16006. doi: 10.1038/nrdp.2016.6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Otto CM, Kuusisto J, Reichenbach DD, Gown AM, O’Brien KD. Characterization of the early lesion of ‘degenerative’ valvular aortic stenosis: histological and immunohistochemical studies. Circulation. 1994;90(2):844-853. doi: 10.1161/01.CIR.90.2.844 [DOI] [PubMed] [Google Scholar]
- 8.Johnson TM, Yang Y, Roul P, et al. A narrowing mortality gap: temporal trends of cause-specific mortality in a national matched cohort study in US veterans with rheumatoid arthritis. Arthritis Care Res (Hoboken). Published online November 4, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Beckhauser AP, Vallin L, Burkievcz CJ, Perreto S, Silva MB, Skare TL. Valvular involvement in patients with rheumatoid arthritis. Acta Reumatol Port. 2009;34(1):52-56. [PubMed] [Google Scholar]
- 10.Roldan CA, DeLong C, Qualls CR, Crawford MH. Characterization of valvular heart disease in rheumatoid arthritis by transesophageal echocardiography and clinical correlates. Am J Cardiol. 2007;100(3):496-502. doi: 10.1016/j.amjcard.2007.03.048 [DOI] [PubMed] [Google Scholar]
- 11.Yiu KH, Wang S, Mok MY, et al. Relationship between cardiac valvular and arterial calcification in patients with rheumatoid arthritis and systemic lupus erythematosus. J Rheumatol. 2011;38(4):621-627. doi: 10.3899/jrheum.100844 [DOI] [PubMed] [Google Scholar]
- 12.Benjamin EJ, Virani SS, Callaway CW, et al. ; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee . Heart and stroke disease statistics—2018 update: a report from the American Heart Association. Circulation. 2018;137(12):e67-e492. doi: 10.1161/CIR.0000000000000558 [DOI] [PubMed] [Google Scholar]
- 13.Gravallese EM, Corson JM, Coblyn JS, Pinkus GS, Weinblatt ME. Rheumatoid aortitis: a rarely recognized but clinically significant entity. Medicine (Baltimore). 1989;68(2):95-106. doi: 10.1097/00005792-198903000-00003 [DOI] [PubMed] [Google Scholar]
- 14.Kaneko S, Yamashita H, Sugimori Y, et al. Rheumatoid arthritis–associated aortitis: a case report and literature review. Springerplus. 2014;3:509. doi: 10.1186/2193-1801-3-509 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.VA Office of Research & Development. VA Informatics and Computing Infrastructure . Accessed August 11, 2022. https://www.research.va.gov/programs/vinci/default.cfm
- 16.Chung CP, Rohan P, Krishnaswami S, McPheeters ML. A systematic review of validated methods for identifying patients with rheumatoid arthritis using administrative or claims data. Vaccine. 2013;31(suppl 10):K41-K61. doi: 10.1016/j.vaccine.2013.03.075 [DOI] [PubMed] [Google Scholar]
- 17.Curtis JR, Xie F, Chen L, Greenberg JD, Zhang J. Evaluation of a methodologic approach to define an inception cohort of rheumatoid arthritis patients using administrative data. Arthritis Care Res (Hoboken). 2018;70(10):1541-1545. doi: 10.1002/acr.23533 [DOI] [PubMed] [Google Scholar]
- 18.Goldsweig AM, Tak HJ, Chen LW, et al. Relative costs of surgical and transcatheter aortic valve replacement and medical therapy. Circ Cardiovasc Interv. 2020;13(5):e008681. doi: 10.1161/CIRCINTERVENTIONS.119.008681 [DOI] [PubMed] [Google Scholar]
- 19.Goldsweig AM, Tak HJ, Chen LW, et al. The evolving management of aortic valve disease: 5-year trends in SAVR, TAVR, and medical therapy. Am J Cardiol. 2019;124(5):763-771. doi: 10.1016/j.amjcard.2019.05.044 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Martinsson A, Li X, Andersson C, Nilsson J, Smith JG, Sundquist K. Temporal trends in the incidence and prognosis of aortic stenosis: a nationwide study of the Swedish population. Circulation. 2015;131(11):988-994. doi: 10.1161/CIRCULATIONAHA.114.012906 [DOI] [PubMed] [Google Scholar]
- 21.England BR, Baker JF, Sayles H, et al. Body mass index, weight loss, and cause-specific mortality in rheumatoid arthritis. Arthritis Care Res (Hoboken). 2018;70(1):11-18. doi: 10.1002/acr.23258 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.World Health Organization. Body mass index (BMI). Accessed October 20, 2022. https://www.who.int/data/gho/data/themes/topics/topic-details/GHO/body-mass-index
- 23.Melzer AC, Pinsker EA, Clothier B, et al. Validating the use of Veterans Affairs tobacco health factors for assessing change in smoking status: accuracy, availability, and approach. BMC Med Res Methodol. 2018;18(1):39. doi: 10.1186/s12874-018-0501-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.England BR, Sayles H, Mikuls TR, Johnson DS, Michaud K. Validation of the rheumatic disease comorbidity index. Arthritis Care Res (Hoboken). 2015;67(6):865-872. doi: 10.1002/acr.22456 [DOI] [PubMed] [Google Scholar]
- 25.Hobday PM, Auger JL, Schuneman GR, Haasken S, Verbeek JS, Binstadt BA. Fcγ receptor III and Fcγ receptor IV on macrophages drive autoimmune valvular carditis in mice. Arthritis Rheumatol. 2014;66(4):852-862. doi: 10.1002/art.38311 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Meier LA, Auger JL, Engelson BJ, et al. CD301b/MGL2+ mononuclear phagocytes orchestrate autoimmune cardiac valve inflammation and fibrosis. Circulation. 2018;137(23):2478-2493. doi: 10.1161/CIRCULATIONAHA.117.033144 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Avina-Zubieta JA, Thomas J, Sadatsafavi M, Lehman AJ, Lacaille D. Risk of incident cardiovascular events in patients with rheumatoid arthritis: a meta-analysis of observational studies. Ann Rheum Dis. 2012;71(9):1524-1529. doi: 10.1136/annrheumdis-2011-200726 [DOI] [PubMed] [Google Scholar]
- 28.Malmberg M, Palomäki A, Sipilä JOT, Rautava P, Gunn J, Kytö V. Long-term outcomes of surgical aortic valve replacement in patients with rheumatoid arthritis. J Clin Med. 2021;10(11):2492. doi: 10.3390/jcm10112492 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Baron SJ, Ryan MP, Moore KA, Clancy SJ, Gunnarsson CL. Contemporary costs associated with transcatheter versus surgical aortic valve replacement in Medicare beneficiaries. Circ Cardiovasc Interv. 2022;15(3):e011295. doi: 10.1161/CIRCINTERVENTIONS.121.011295 [DOI] [PubMed] [Google Scholar]
- 30.Bois JP, Crowson CS, Khullar T, Achenbach SJ, Krause ML, Mankad R. Progression rate of severity of aortic stenosis in patients with rheumatoid arthritis. Echocardiography. 2017;34(10):1410-1416. doi: 10.1111/echo.13652 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Willerson JT, Ridker PM. Inflammation as a cardiovascular risk factor. Circulation. 2004;109(21)(suppl 1):II2-II10. [DOI] [PubMed] [Google Scholar]
- 32.Briand M, Lemieux I, Dumesnil JG, et al. Metabolic syndrome negatively influences disease progression and prognosis in aortic stenosis. J Am Coll Cardiol. 2006;47(11):2229-2236. doi: 10.1016/j.jacc.2005.12.073 [DOI] [PubMed] [Google Scholar]
- 33.England BR, Roul P, Yang Y, et al. Burden and trajectory of multimorbidity in rheumatoid arthritis: a matched cohort study from 2006 to 2015. Ann Rheum Dis. 2021;80(3):286-292. doi: 10.1136/annrheumdis-2020-218282 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kerekes G, Nurmohamed MT, González-Gay MA, et al. Rheumatoid arthritis and metabolic syndrome. Nat Rev Rheumatol. 2014;10(11):691-696. doi: 10.1038/nrrheum.2014.121 [DOI] [PubMed] [Google Scholar]
- 35.Coutinho T, Goel K, Corrêa de Sá D, et al. Central obesity and survival in subjects with coronary artery disease: a systematic review of the literature and collaborative analysis with individual subject data. J Am Coll Cardiol. 2011;57(19):1877-1886. doi: 10.1016/j.jacc.2010.11.058 [DOI] [PubMed] [Google Scholar]
- 36.Escalante A, Haas RW, del Rincón I. Paradoxical effect of body mass index on survival in rheumatoid arthritis: role of comorbidity and systemic inflammation. Arch Intern Med. 2005;165(14):1624-1629. doi: 10.1001/archinte.165.14.1624 [DOI] [PubMed] [Google Scholar]
- 37.Kremers HM, Nicola PJ, Crowson CS, Ballman KV, Gabriel SE. Prognostic importance of low body mass index in relation to cardiovascular mortality in rheumatoid arthritis. Arthritis Rheum. 2004;50(11):3450-3457. doi: 10.1002/art.20612 [DOI] [PubMed] [Google Scholar]
- 38.Wolfe F, Michaud K. Effect of body mass index on mortality and clinical status in rheumatoid arthritis. Arthritis Care Res (Hoboken). 2012;64(10):1471-1479. doi: 10.1002/acr.21627 [DOI] [PubMed] [Google Scholar]
- 39.Yeang C, Wilkinson MJ, Tsimikas S. Lipoprotein(a) and oxidized phospholipids in calcific aortic valve stenosis. Curr Opin Cardiol. 2016;31(4):440-450. doi: 10.1097/HCO.0000000000000300 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Finckh A, Gilbert B, Hodkinson B, et al. Global epidemiology of rheumatoid arthritis. Nat Rev Rheumatol. 2022;18(10):591-602. [DOI] [PubMed] [Google Scholar]
- 41.Chandra S, Lang RM, Nicolarsen J, et al. Bicuspid aortic valve: inter-racial difference in frequency and aortic dimensions. JACC Cardiovasc Imaging. 2012;5(10):981-989. doi: 10.1016/j.jcmg.2012.07.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Diagnostic and Procedural Codes Used to Identify Incident Aortic Stenosis in US Veterans With and Without Rheumatoid Arthritis
eTable 2. Databases Contributing Events to the Composite Aortic Stenosis Outcome
eFigure 1. Sensitivity Analysis Examining Aortic Stenosis Risk in Rheumatoid Arthritis (N=34 012) vs. Matched Controls (N=274 300), Restricting to Patients Over 65 Years of Age
eFigure 2. Sensitivity Analysis Examining Aortic Stenosis Risk Amongst Incident Rheumatoid Arthritis Cases (N=32 369) vs Matched Controls (N=264 240)
eFigure 3. Temporal Trends in the Risk of the Composite Aortic Stenosis Outcome
eFigure 4. Aortic Stenosis Risk in Rheumatoid Arthritis vs Matched Controls, Stratified by Frequency of Outpatient Care
Data Sharing Statement