Incidence Rate of Cardiovascular Events in Rheumatoid Arthritis: An Observational Cohort Study in Saudi Arabia

Introduction

Rheumatoid arthritis (RA) is the most prevalent autoimmune inflammatory arthritis,¹ with an estimated prevalence rate of 460 per 100,000 population between 1980 and 2019.² In addition to affecting joints, RA induces a chronic systemic inflammatory state that impacts various tissues and organs, including the cardiovascular system.³ RA is linked to an increased incidence of cardiovascular disease (CVD), which is the primary cause of death among RA patients, resulting in 50% excess mortality compared to the general population.⁴ It is believed that RA patients are at heightened risk for sudden cardiac death due to ischemic and non-ischemic heart disease, and cardiac arrhythmias.⁵ The inflammatory nature of RA may contribute to the connection between RA and CVD, due to the increased production of pro-inflammatory substances like erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), interleukins, and tumor necrosis factors. These substances are known to promote endothelial dysfunction, atherosclerosis plaque formation, and cardiac remodeling.^6–9

Previous studies on the relationship between CVD and RA within a Saudi Arabian cohort are scarce. One study from the western region of Saudi Arabia assessed hyperlipidemia prevalence in 180 RA patients, finding that 55.1% had elevated total cholesterol and 55.2% had elevated low-density lipoprotein cholesterol.¹⁰ These rates surpass the 20–40% hyperlipidemia rates reported in the general Saudi population by Al-Kaabba et al.¹¹ Additionally, Al-Bishri et al analyzed 340 RA patients from three centers in Saudi Arabia, identifying hypertension (35.9%), diabetes (30.9%), and dyslipidemia (19.4%) as the most common comorbidities. Furthermore, 7.4% of these patients had ischemic heart disease, 3.3% had cerebrovascular disease, and 2.1% had deep venous thrombosis.¹² Another study involving 75 RA patients showed that 18.7% had diabetes, 26.7% had hypertension, and 6.7% had CVD, with a significantly higher risk of CVD found in those with highly active disease (p < 0.001).^13,14 Another study reported that among 116 RA patients in western Saudi Arabia, cardiovascular disease was the most significant cause of death, accounting for 58% of cases.¹⁴

To the best of our knowledge, there is no published data available in Saudi Arabia regarding the incidence rate of cardiovascular disease (CVD) in patients with rheumatoid arthritis (RA). It is critical to understand the epidemiology of CVD in RA to guide current and future healthcare policies. This study was conducted to estimate the incidence rate of CVD among Saudi RA patients and to identify the risk factors associated with CVD in this population.

Materials and Methods

Study Setting

This retrospective study analyzed electronic medical records of patients who attended rheumatology clinics and were prescribed disease-modifying antirheumatic drugs (DMARDs) at Dammam Medical Complex (DMC), as well as hospitalization records at Saud Al-Babtain Center for Cardiac Medicine and Surgery (SBCC) for those hospitalized with a CVD event. Additionally, patients were contacted by phone to collect further data that might have been missed due to incomplete documentation or loss of follow-up.

Dammam Medical Complex, managed by the Eastern Health Cluster, is a public hospital in Dammam with 51 clinics and centers offering a variety of medical services, including rheumatology. Specialized cardiac services are provided by Saud Al-Babtain Cardiac Center, which is part of DMC.¹⁵

The study received approval from the Institutional Review Board of Saud Al-Babtain Cardiac Center (SBCC-IRB-MC-2021-05). Since DMC lacked an independent IRB at the time of the study, approval was also obtained from the review board of King Fahad Specialist Hospital – Dammam (No. EXT0383). All participants provided verbal informed consent prior to participation, ensuring that they were not subjected to any undue pressure. The interviews were conducted with the utmost confidentiality and respect. IRB approved the verbal informed consent process and the study procedures adhered to the Declaration of Helsinki guidelines for good clinical practice.

Subjects and Procedures

In the current study, we screened all patients who visited rheumatology clinics at DMC from 2016 to 2021 and were treated with disease-modifying anti-rheumatic drugs (DMARDs). We excluded 856 subjects who received DMARDs for conditions other than rheumatoid arthritis (RA) and 9 subjects who were under 18 at the time of their RA diagnosis, from a total of 1516 screened subjects. Follow-ups were conducted via phone calls by three interviewers using a standardized questionnaire and continued until the last recorded notes in charts or the date of data collection on February 2, 2022, whichever came first. To verify cardiovascular disease (CVD) incidents, data were collected from three sources: DMC electronic records, SBCC hospitalization records, and patient self-reports, and were entered into the REDCap software. Discrepancies between records and patient-reported outcomes were resolved by prioritizing the documented chart information.

Measurements and Definitions

Baseline data collection included sociodemographic information, CVD history and risk factors, other comorbidities, RA history, and DMARD usage. Sociodemographic data captured age, gender, nationality, and smoking status. CVD risk factors included diagnosed hypertension, diabetes, family and personal history of CVD, previous coronary artery disease, stroke/transient ischemic attack, heart failure, atrial fibrillation, venous thromboembolism, treated dyslipidemia, and obesity (BMI ≥30 kg/m²).¹⁶ RA diagnosis followed the 2010 American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) criteria.¹⁷ Other disease-related data such as date of diagnosis, disease duration, and plasma levels of inflammatory markers (CRP and ESR) at diagnosis were collected. RA medications were categorized into four groups: conventional synthetic DMARDs (methotrexate and hydroxychloroquine), TNF inhibitors (adalimumab, infliximab, etanercept), non-TNF biologics (tocilizumab, abatacept, rituximab), and the JAK inhibitor tofacitinib. The study also examined exposure to oral corticosteroids and NSAIDs due to their potential relevance to RA and associated outcomes. Chronic use of NSAIDs and steroids was defined as continuous use for 6 months or more.

Outcomes

Incidence of cardiovascular events was defined as the occurrence and/or diagnosis of ischemic heart disease (myocardial infarction, unstable angina, coronary revascularization), stroke/transient ischemic attack, venous thromboembolism, heart failure, and arrhythmia/conduction abnormalities after the diagnosis of RA. Events were identified from hospitalization/procedural records and patient self-reports of CVD collected via telephone.

Statistical Analysis

Categorical variables were presented as counts and percentages, while continuous variables were presented as either mean ± standard deviation or median and interquartile range, depending on their distribution. Patient-years at risk for cardiovascular diseases were calculated for each subject from the baseline visit to the diagnosis of each outcome, the time of the last outpatient clinic visit, or the time of the telephone interview, whichever came first. Incidence rates were calculated by dividing the number of new events for each outcome by the respective patient-years at risk.

Our exploratory analysis included a univariable logistic regression to identify possible predictors of CVD or arrhythmia among RA patients. Results are expressed as odds ratios (ORs) and 95% confidence intervals (CIs). Multivariable logistic regression analysis further supplemented our analysis. A p-value < 0.05 was considered statistically significant. All statistical analyses were performed using R software.¹⁸

Results

A total of 651 subjects were eligible for inclusion, of whom 52 had hospitalization records at SBCC for cardiovascular events and 365 were contacted by phone. Figure 1 showed the flowchart of study population.

Figure 1 Participants Flowchart.

Baseline Characteristics

Of the 651 included patients, 80.5% were female, and 93% were Saudis, with an average age of 51 (SD±13) years. The most prevalent comorbid diseases were obesity (42.9%), diabetes (18.9%), and hypertension (18.7%). Only a small proportion of patients (1.7%) had established cardiovascular disease before the diagnosis of RA. Regarding smoking status, twenty-three patients were current smokers, while nine reported a previous smoking history. Concerning rheumatoid arthritis history, the median duration since RA diagnosis was 9 years, with median baseline levels of inflammatory markers at 45.4 mm/hr for erythrocyte sedimentation rate and 4.8 mg/dL for C-reactive protein. The most commonly used therapies for RA were methotrexate (91.2%) and hydroxychloroquine (87.4%). Adalimumab was the most frequently prescribed biologic drug (24.6%), while Infliximab was the least frequently used (3.2%). Regarding adjunctive medication use, 47.5% were chronic steroid users and 22.3% were chronic NSAIDs users. Table 1 shows the cohort’s baseline characteristics.

Table 1 Patients’ Baseline Characteristic (n=651)

Incidence of Cardiovascular Events

Over a median follow-up of 6 years (interquartile range: 3 to 9 years), the incidence of CVD events or arrhythmia was 54 (8.2%), including 17 (2.6%) cases of ischemic heart disease, 10 (1.53%) ischemic strokes or TIAs, 5 (0.76%) venous thromboembolisms (VTE), 17 (2.6%) arrhythmias, and 9 (1.38%) heart failure diagnoses. The crude incidence rates per 1000 person-years were 3.4 for ischemic heart disease, 2.02 for ischemic stroke or TIA, 1.02 for VTE, 3.5 for arrhythmias, and 1.84 for heart failure. Overall, the cumulative 6-year incidence rate for CVD was 11.2 events per 1000 person-years. Males had a higher incidence rate compared to females (69.9 vs 13.3 per 1000 person-years). Table 2 shows the six-years cumulative incidence and incidence rates of cardiovascular disease in Rheumatoid Arthritis Patients from the Kingdom of Saudi Arabia.

Table 2 Six-Year Cumulative Incidence and Incidence Rates of Cardiovascular Disease in RA Patients

Analyses of Risk Factors

Table 3 presents the results from the univariable multivariable logistic regression analyses of incident CVD. Statistically significant risk factors included hypertension (odds ratio [OR] 4.15, 95% CI 2.29–7.51), dyslipidemia (OR 3.67, 95% CI 1.77–7.59), diabetes (OR 2.93, 95% CI 1.57–5.46), and previous history of CVD events (OR 4.95, 95% CI 1.35–18.14). Age (OR 1.07, 95% CI 1.04–1.10) and a RA duration of 15 years or more were also significant predictors of incident CVD. After adjustment, significant predictors included male sex (adjusted OR 3.17, 95% CI 1.10–9.14), hypertension (adjusted OR 3.40, 95% CI 1.09–10.61), and a RA duration of 15 years or more (adjusted OR 64.81, 95% CI 3.68–1140.6).

Table 3 Univariable and Multivariable Logistic Regression for Cardiovascular Disease in Rheumatoid Arthritis Patients

Discussion

This study is the first to explore the incidence rates of CVD among RA patients in Saudi Arabia. The overall incidence was 11.2 per 1000 person-years over six years. Demographically, 80.5% of the participants were female, 93% were Saudis, and the mean age was 51 years. The prevalence of traditional CVD risk factors was 18.9% for diabetes, 18.7% for hypertension, 42.9% for obesity, and 7.4% for dyslipidemia. Globally, CVD incidence varies by country and population. The Global Burden of Disease study reported a global incidence of 684.33 per 100,000 population (6.8 per 1000), and in Central Asia, the age-adjusted incidence was 1100 per 100,000 population.¹⁹ Locally, a one-year study from a tertiary center in Jeddah (2019–2020) reported a lower incidence of ischemic heart diseases at 2.2 per 1000 person-years in the general population.²⁰

Several studies around the world have investigated the incidence of cardiovascular disease (CVD) among rheumatoid arthritis (RA) populations, revealing varied rates. For instance, a national Korean study reported an incidence rate of 18.2 per 1000 person-years.²¹ In Sweden, the rate was 21 per 1000 person-years,²² while a Canadian inception cohort study revealed a rate of 7.1 per 1000 person-years.²³ Similarly, a Portuguese cohort of RA women reported a rate of 7 per 1000 person-years.²⁴ In England, the incidence rate was 35.33 per 1000 person-years, using a broader definition of CVD that included twelve different presentations.²⁵ Note that all rates were standardized to per 1000 person-years for comparison. These rates indicate higher incidences in RA patients compared to the general population, which may be attributed to differences in study methodologies, inclusion criteria, CVD definitions, follow-up durations, and variations in genetic, environmental, and lifestyle risk factors.

Ischemic Heart Disease (I.e Coronary Heart Disease)

As previously mentioned, a retrospective study conducted by Albeladi et al in Jeddah, Saudi Arabia, reported the incidence of coronary artery disease (CAD) in 2019 as 2.20 per 1000 person-years.²⁰ This rate is lower than the 3.4 per 1000 person-years reported for ischemic heart diseases. Interestingly, Albeladi et al also reported a rate of 303 per 1000 person-years in 2020, a significant increase likely due to COVID-19, which is known to exacerbate CVD.²⁶ Our study, which followed up until 2021, did not account for the COVID-19 virus. Additionally, gender differences may have influenced these findings, as men, who comprised 77% of the study’s participants, have a twofold higher incidence of CAD than women.²⁷ In our study, 80.5% of participants were females, reflecting the higher prevalence of RA among women.²⁸ Cho et al reported a CAD incidence of 5.2 per 1000 person-years in males versus 1.5 in females,²¹ and Chung et al reported 3.36 per 1000 person-years in males versus 1.36 in females.²⁹ These findings suggest a two- to threefold increase in CAD among RA males compared to females, mirroring trends seen in the general population. Our study also confirmed a higher incidence of CVD or arrhythmia in males, with rates of 69.9 per 1000 person-years versus 13.3 per 1000 person-years in females.

Stroke and/or Transient Ischemic Attack

On a national level, the incidence of first-ever stroke was estimated through five cross-sectional studies across different regions of Saudi Arabia. Our findings indicated a rate of 1.53% (2.02 per 1000 person-years) for ischemic stroke or TIA in patients with RA, which is higher than previously reported in the general population: 0.0139% (13.9 per 100,000 people) over 1 year,³⁰ 0.0159% (15.9 per 100,000 people) over two years,³¹ 0.0298% (29.8 per 100,000 people) over 3 years,³² 0.0438% (43.8 per 100,000 people) over ten years,³³ and 0.0576% (57.6 per 100,000 people) over one year.³⁴ A systematic review and meta-analysis of these studies found an annual incidence of 29 strokes per 100,000 people (95% CI: 15 to 47), yielding a pooled rate of 0.029% (95% CI: 0.015 to 0.047), which is lower than in other high-income countries.³⁵

Internationally, stroke rate for women with RA, as indicated by the Nurses’ Health Study (1.12 per 1000 person-years), is not considered statistically significant compared to the general population, with a multivariable-adjusted relative risk of 1.48 (95% CI, 0.70 to 3.12).³⁶ However, our results align with other studies that explored the association between stroke and RA. For example, a study in the Korean population revealed a stroke incidence rate of 2.85 per 1000 person-years in RA patients, with a more pronounced association in females (adjusted HR: 1.27) compared to males (adjusted HR: 1.14; p = 0.037).³⁷ Conversely, our rates are lower than those reported in a Danish nationwide cohort study, which showed a stroke incidence rate of 7.6 per 1000 person-years among 718 subjects with RA.³⁸

Venous Thromboembolism

Regarding venous thromboembolism (VTE) events, a single-center cohort study among the Saudi population reported a cumulative incidence of 0.17% (1.7 per 1000 patients) over 3 years, with an overall incidence of 158 per 1000 person-years for recurrent VTE.³⁹ These findings are higher than those in the current study, reported an event rate of 0.76% (1.02 per 1000 person-years). Contrarily, a Swedish register-based cohort study found a VTE incidence of 0.71% among RA patients, compared to 0.36% in the matched general population.⁴⁰ An earlier study in Sweden observed even higher rates of VTE, with 5.9 per 1000 person-years in RA patients versus 2.8 in the general population, noting that the increased risk remained consistent up to 10 years from RA diagnosis.⁴¹ Meanwhile, a cohort of 609 RA patients showed a VTE incidence of 7.2% over 30 years via ultrasound.⁴² The relatively short median follow-up of 6 years in our study may account for the lower rate observed.

Arrhythmias

We reported an incidence rate of 3.5 per 1000 person-years with a total of 17 events (2.6%) for arrhythmias, which included various types such as supraventricular tachycardia, atrial flutter, atrial fibrillation, ventricular tachycardia, premature ventricular contraction, and left bundle branch block. The incidence rate of atrial fibrillation (AF) in rheumatoid arthritis (RA) patients was investigated through a nationwide, longitudinal, register-based cohort study in Denmark. The study found an incidence rate of 8.2 per 1000 person-years among 18,247 RA patients, compared to 6.0 per 1000 person-years in age- and sex-matched controls. Women exhibited a slightly higher relative risk than men, and AF was more prevalent in the youngest age groups.³⁸ Another smaller inception cohort of 813 RA subjects reported a cumulative incidence of AF at 18.3% over an average follow-up of 9.6 (± 6.9) years.⁴³ Furthermore, an Italian retrospective study noted an AF incidence rate of 7.01 per 1000 person-years among 21,201 RA patients, higher than the 5.98 per 1000 person-years in a non-RA cohort. The risk remained significant even after adjusting for cardiovascular comorbidities such as hypertension and diabetes, which were more prevalent in the RA cohort.⁴⁴ Despite a broader definition of arrhythmia used in our study, the resultant incidence rate was lower. However, a similar rate was reported by Kim et al in a US retrospective cohort of 20,852 RA patients from a commercial health plan, with an AF incidence rate of 4.0 (95% CI 3.4 to 4.7) per 1000 person-years over a two-year follow-up. This study did not demonstrate an increased risk of AF in the RA cohort compared to non-RA, with a hazard ratio (HR) of 1.1 (95% CI 0.9 to 1.4) adjusted for diabetes, cardiovascular disease, and medications.⁴⁵ Similarly, Jang et al did not establish an association between AF and RA, with a cumulative incidence of AF over 10 years of follow-up at 4.01 per 1000 person-years (95% CI 329–489) in RA versus 3.79 (95% CI 329–437) in matched controls.⁴⁶

Heart Failure

The study identified 9 (1.38%) cases of heart failure diagnosis, corresponding to an incidence rate of 1.84 per 1000 person-years, which is low compared to other studies. For instance, a retrospective population-based study in Minnesota including 575 RA subjects reported a heart failure (HF) incidence rate of 19.9 per 1000 person-years over a median follow-up of 11.8 years, with a higher risk among rheumatoid factor-positive patients.⁴⁷ The positive correlation between HF and elevated rheumatologic biomarkers such as ESR, CRP, rheumatoid factor, and TNF was described in several other studies.^48–50 In another prospective study evaluating the effect of anti-TNF medications on HF rates among 13,171 RA patients, the reported rate of HF was 3.9%, significantly (P <0.05) less common in anti-TNF-treated patients (3.1%) compared to the remaining patients (3.8%).⁵¹

Risk Factors

In rheumatoid arthritis (RA), accelerated atherosclerosis might be due to a higher occurrence of conventional CVD risk factors, including smoking, hypertension (HTN), diabetes mellitus (DM), dyslipidemia, and obesity.⁵² Our study recorded the prevalence of these risk factors as follows: 18.9% for diabetes, 18.7% for hypertension, 7.4% for dyslipidemia, and 49.2% for obesity. The Saudi National Heart Center reported lower prevalences in the general population of Saudi Arabia, including 12.3% for diabetes mellitus, 16% for smoking, 22% for hypertension, and 35–40% for dyslipidemia.⁵³ Our RA cohort showed higher diabetes rates, but lower rates of hypertension and dyslipidemia compared to the general population. Notably, the lower dyslipidemia rates in our RA patients align with previous findings that RA patients do not generally have higher hyperlipidemia than the general population, and their cholesterol levels are often lower when the disease is active.⁵⁴ This phenomenon, known as the “lipid paradox”, suggests that reduced lipid levels in RA patients are paradoxically linked to an increased CVD risk.⁵⁵ Regarding obesity, our findings were double those of the Arab Teens Lifestyle Study, which reported a 23% obesity rate among adults in the general Saudi population.⁵⁶ In contrast, we observed a lower prevalence of all risk factors compared to another cohort of 340 RA patients in Saudi Arabia, which reported higher rates of hypertension (35.9%), diabetes (30.9%), and dyslipidemia (19.4%) in a cross-sectional study.¹² These variations could be due to demographic, regional, and study design differences. In our study, hypertension was identified as the most significant predictor of incident CVD in the multivariable analysis.

Beyond the traditional risk factors, RA patients also face increased CVD morbidity due to disease-related factors such as systemic inflammation and vascular endothelial activation caused by immune dysregulation and inflammation, as well as treatment-related factors including corticosteroids that can accelerate atherosclerotic disease.^57–59 Additionally, RA patients with elevated erythrocyte sedimentation rates are at heightened risk for cardiovascular complications.⁶⁰ Unlike previous studies, we found no significant association between CVD incidence and inflammation measured by baseline C-reactive protein (OR 1.04, CI: 0.94 to 1.15, P-value = 0.417) or erythrocyte sedimentation rate (OR: 0.99, CI: 0.99 to 1.01, P-value = 0.417). These findings may be attributable to our cohort selection criteria, which required at least one DMARD prescription before inclusion, suggesting low systemic inflammation as all RA patients received DMARDs. Additionally, the values obtained at a single point may not adequately reflect the severity of the disease since some patients may experience flare-ups and progressions over time.

Generally, we found no significant association between the use of DMARDs and the incidence of CVD, although the odds ratios suggest a protective effect. These results align with previous studies investigating the impact of treatments on CVD. Methotrexate (MTX) was the most used DMARD in our cohort and is a highly recommended agent according to the 2015 and 2021 American College of Rheumatology Guidelines for the treatment of RA.^1,61 A 2015 analysis found that MTX use was associated with a 21% reduction in cardiovascular events compared to non-MTX use, including other DMARDs or no therapy at all.⁶² This improvement in CVD outcomes was also observed with TNF-alpha inhibitors,⁶³ abatacept,⁶⁴ and Janus kinase inhibitors.⁶⁵ Additionally, the use of low-dose glucocorticoids (GC) showed a significant association with CVD incidence in the current study, with an adjusted odds ratio (aOR) of 9.50 (95% CI: 1.348 to 66.89, P=0.024). There is conflicting evidence regarding the effect of GC on incident CVD; while a reduced CVD risk with low-dose corticosteroids was observed in a US registry-based study,⁶⁶ a more recent analysis reported an increased risk associated with GC use, even at low doses, in a study involving 87,794 adults with immune-mediated inflammatory diseases and no prior CVD over a 5-year median follow-up.⁶⁷

Study Limitations

The retrospective design of our study may have led to some subjects being lost to follow-up, which could have resulted in an underestimation of CVD incidence. Although we made every effort to ensure complete data collection, information loss is possible due to the nature of the study design. Additionally, the lack of a non-RA control group limits our ability to definitively attribute CVD incidence to RA alone. Instead, we relied on previously published literature to contextualize our results. This study was conducted in eastern Saudi Arabia, which may limit the generalizability of the findings to other regions or populations with different genetic, environmental, and lifestyle factors.

The median follow-up period of 6 years may not fully capture the long-term cardiovascular risks in patients with RA. Extended follow-up is necessary to more accurately determine the lifetime risk and the effects of RA management on cardiovascular outcomes. The impact of different RA treatment regimens on CVD risk was not thoroughly examined in the current study. As RA management evolves, it is critical to understand how various treatments affect cardiovascular outcomes, especially as treatment approaches can change over time and establishing a causal relationship was challenging with the study’s design. Additionally, the study did not deeply analyze the role of lifestyle factors like diet, physical activity, and stress, which are crucial to cardiovascular health. Our univariable analysis aimed to be exploratory and identify predictors. This approach was limited by residual confounding due to variables like socioeconomic status or medication adherence, which were either unmeasured or inadequately measured, impacting our ability to establish a clear link between RA and CVD. The retrospective nature of the study also meant that we could not assess disease activity and severity indicators, such as the disease activity score in 28 joints (DAS28-CRP or DAS28-ESR), during data collection, nor were these examined as predictors of CVD in our cohort. We did, however, obtain baseline values of inflammatory markers ESR and CRP, which did not emerge as significant predictors of CVD. Likewise, data on rheumatoid factor (RF) and anti-citrullinated peptide antibodies (ACPA) were largely absent from our dataset and thus were excluded from the analysis. Care should be taken when interpreting the results of multivariable logistic regression analysis due to potential misinterpretations known as the “Table 2 fallacy”.⁶⁸

Conclusion

To the best of our knowledge, this is the only study estimating the incidence of CVD and associated risk factors in RA patients in Saudi Arabia. Our findings indicate that male gender, longer disease duration, and comorbid hypertension are significant predictors of CVD in this group. Therefore, aggressive management of modifiable CVD risk factors, coupled with regular and early risk assessments, is recommended to reduce the risk of CVD in the RA population.