Meta-analysis of risk factors for cardiovascular disease in patients with rheumatoid arthritis

Abstract

Background:

Cardiovascular disease (CVD) is a common complication of rheumatoid arthritis (RA). We aimed to explore the risk factors for cardiovascular disease in patients with rheumatoid arthritis and provide a scientific basis on effective prevention and treatments for CVD in RA patients.

Methods:

We searched for a combination of the subject words and free words involved arthritis, rheumatoid, CVD, heart disease, vascular disease, risk factors, etc. from China Knowledge Network, Wanfang, Vip, China Biomedical Literature Database, Pubmed, Embase, Web of Science, Cochrane and other databases for documents published in public in 2000-October 2022.RevMan 5.3 and Stata14.0 analysis software was used to perform a meta-analysis of case-control and cohort studies on risk factors of CVDs in patients with RA published from 2000 to 2022.The Egger method was used to determine whether there was publication bias.

Results:

Seventeen pieces of literature were included in this meta-analysis. We explored the associations between CVD and different clinical characters such as DAS28 score, rheumatoid factor, triglyceride, age, sex, C-reactive protein, total cholesterol (TC), hypertension, and D dimer. High serum levels of C-reactive protein, TC, and D-dimer, as well as hypertension, are the main risk factors for CVD in patients with RA. The OR and 95% CI of C-reactive protein were 2.06 (1.91–2.23), RR and 95% CI of TC were 1.7 (1.49–1.93), RR and 95% CI of hypertension were 3.58 (2.37–5.40), as well as OR and 95% CI of D dimer were 2.83 (1.48–5.40).Our results performed by the meta-analysis were reliable, with low publication bias existed.

Conclusion:

C-reactive protein, TC, hypertension, and D dimer are the main risk factors for CVD in patients with RA. No protective factors were found.

1. Introduction

Rheumatoid arthritis (RA) is a common autoimmune disease characterized by synovial inflammation and joint destructions, which also causes multiple system disorders. Previous studies have shown that the incidence and mortality of cardiovascular disease (CVD) in patients with RA are significantly higher than in the general population, and the researchers believe that traditional cardiovascular risk factors can not fully explain this result. Systematic inflammation is the leading cause of RA combined with CVD.^[1,2] In order to further determine the risk factors of concurrent CVD in RA, we searched the published literature to perform a meta-analysis and provided a scientific basis for effective prevention of CVD in RA patients.

2. Data and methods

We adopted a combination of the subject words and free words for literature searches. The keywords involved arthritis, rheumatoid, cardiovascular disease, heart disease, vascular disease, risk factors, etc. Following the above strategies, the 2 researchers independently searched China Knowledge Network, Wanfang, Vip, China Biomedical Literature Database, Pubmed, Embase, Web of Science, Cochrane and other databases for documents published in public in 2000-October 2022. Studies without publically published were excluded. Document tracking and manual retrieval were also applied.

Inclusion criteria: type of literature: publicly published case-control studies or cohort studies from domestic and abroad, with no less than 50 sample size per article. Subjects: the RA patients who were hospitalized or participated in the study simultaneously were divided into the case group and control group according to whether they had cardiovascular diseases or not. Diagnostic criteria: all RA patients met the American College of Rheumatology revised classification criteria for 1987/2010. Outcome measures: OR/RR/HR (95% CI) or convertible to OR/RR/HR (95% CI) was provided in the original literature. Exclusion criteria: studies other than case-control or cohort studies, sample size <50; inability to extract OR/RR/HR values and 95% CI; repeat reported literature; and failure to obtain the full text.

Literature screening and data extraction were completed by 2 researchers independently. If there are different opinions, the original data shall be checked and agreed with upon through discussion. We extracted relevant data from the identified literature through a self-designed data extraction form, including first author, year of publication, type of study design, sample size, risk factors, and OR/RR/HR (95% CI) values.

The quality of literature was evaluated by the Newcastle-Ottawa Scale, which included selecting the study population, inter-group comparability, and exposure/outcome evaluation. There were 8 items in total, with a total score of 9 points. A score of ≥6 points was regarded as a high-quality study.

Revman5.3 software and Stata14.0 software were used for meta-analysis, and the heterogeneity was tested by Q test and I². If P > .05, I² < 50%, the homogeneity among the studies was low, and we selected the fixed effect model; if P ≤ .05, I² ≥ 50%, the heterogeneity among the studies was considered, and the random effect model was applied. Furthermore, the sensitivity analysis and subgroup analysis were conducted to find the cause of the heterogeneity. The Egger method was used to determine whether there was publication bias. For example, if P > .05 of each factor, the bias was effectively controlled; otherwise, obvious bias existed.

3. Results

A total of 2471 relevant works of literature were searched in the preliminary examination of the literature search results, excluding repeated literature, systematic evaluations, reviews, and experimental studies, 2426 pieces of literature obviously inconsistent with the subject were excluded according to abstracts, titles, and keywords, and 45 studies were included in the preliminary examination. After reading the complete text, we screened the studies in strict accordance with the inclusion criteria, and 28 studies were excluded. Eventually, 17 pieces of qualified literature were included (Fig. 1).

Figure 1.

Meta-analysis literature screening process for risk factors of RA combined with CVD. CVD = cardiovascular disease, RA = rheumatoid arthritis.

Among the literature included in the meta-analysis, nine were in Chinese and 8 in English; 11 were case-control studies, and 6 were cohort studies, all published in March 2000 to 2022, with a total of 4048 cases in the case group and 17,040 in the control group. All the studies were scored according to the Newcastle-Ottawa Scale standard. The results showed that 13 studies were scored 7, and 4 were scored 6. The overall quality of included literature was high (Table 1).

Table 1.

General characteristics of the literature included in the meta-analysis.

Author	Year	Study type	Participants		Risk factors	NOS score
			Experimental group	Control group
Chunlai Zhang[3]	2012	Case control	53	126	1,2,4,28,5,7	6
Weiping Ci[4]	2016	Case control	76	60	22,11,25,36	7
Xuelian Xu[5]	2020	Case control	61	105	7,8,2,5	6
Hongbin Li[6]	2006	Case control	92	476	1,4,28,5	7
Lei Chen[7]	2019	Case control	78	122	13,.7,1,6,5,3,2	7
Tingting Lyu[8]	2016	Case control	62	60	1,5,7,15	7
Yanfang Meng[9]	2019	Case control	50	50	2,5,7	6
Yan Lin[10]	2015	Case control	49	48	14,15,10,11,13,9,20	7
Tanaka[11]	2016	Cohort study	2472	6186	20	6
Crowson[12]	2018	Cohort study	389	5249	37,20,8,23,24,11	7
Ajeganova[13]	2013	Cohort study	177	564	29,31,32,33,34,35	7
Radovits[14]	2008	Cohort study	41	181	38,37	7
Scott[15]	2012	Case control	27	282	9,22,11	7
Serelis[16]	2011	Cohort study	13	174	11,21,26,30	7
Assous[17]	2007	Cohort study	17	239	20,37,7,11,18	7
Chun Li[18]	2017	Case control	256	2013	11,9,17,25	7
Wenjun Lu[19]	2017	Case control	42	84	27,15,8,19	7

1 = DAS28 score, 2 = rheumatoid factor (RF), 3 = CCP, 4 = platelet (PLT), 5C = reactive protein, 6 = ESR, 7 = total cholesterol (TC), 8 = Glycerin (TG), 9 = hyperlipoproteinemia, 10 = lipoprotein, 11 = hypertension, 12 = diabetes mellitus, 13 = glucose, 14 = uric acid, 15 = D-dimers, 16 = RBC distribution width coefficient (RDW-CV), 17 = interstitial lung disease, 18 = osteoporosis, 19 = fracture, 20 = age, 21 = late onset age of RA, 22 = long duration of RA, 23 = ongoing smoking, 24 = previous smoking, 25 = hydroxychloroquine, 26 = leflunomide, 27 = high mobility of RA, 28 = extraarticular organ involvement, 29 = high HAQ after 2 years, 30 = CRP increase after 1 year, 31 = HAQ decrease after 1 year, 32 = DMARD in the first year, 33 = MTX in the first year, 34 = MTX after the first year, 35 = hormone after the first year, 36 = low-density lipoprotein cholesterol (LDL-C), 37 = sex, 38 = age of myocardial infarction.

This study included 9 risk factors for DAS28, rheumatoid factor (RF), triglyceride (TG), age, sex, C-reactive protein (CRP), total cholesterol (TC), hypertension, and D-dimer, all of which were heterogeneous.

DAS28, RF, TG, age, and sex risk factors were included in 3 studies, respectively, with results of obvious heterogeneity (I² > 50%). As for the study type, populations, and quality scores, no heterogeneity source was found in subgroup analysis. Due to the small quantities of pieces of literature, a descriptive analysis of the aforementioned factors was performed. The results showed that RA patients with high DAS28 scores, high RF positive rates, high TG, and elderly age have a high risk of CVD. Female RA patients have a lower risk of CVD than male patients.

The works of literature studied for the risk factors of CRP, TC, and hypertension were heterogeneous (P ≤ .05, I² ≥ 50%).

We performed the sensitivity analysis to find out the studies which caused the main heterogeneity of the results of the aforementioned risk factors and eliminated them in our meta-analysis. During the process of trying to delete several studies, the results showed that the OR value did not change significantly. The OR and 95% CI of CRP were 2.06 (1.91–2.23), RR and 95% CI of TC were 1.7 (1.49–1.93), as well as RR and 95% CI of hypertension were 3.58 (2.37–5.40). Despite slight heterogeneity in the meta-analysis of the D dimer factor (I² = 52%>50%, P > .1), the degree of heterogeneity is within an acceptable range, and the randomized effect model was conducted, with OR and 95% CI of 2.83 (1.48–5.40). We concluded that CRP, TC, hypertension, and D dimer are the main risk factors for CVD in patients with RA (Fig. 2). No protective factors were found.

Figure 2.

Meta-analysis of risk factors in the pathogenesis of RA with CVD. (A) CRP. (B) Total cholesterol. (C) Hypertension. (D) D-dimer. CRP = C-reactive protein, CVD = cardiovascular disease, RA = rheumatoid arthritis.

P values of CRP, TC, hypertension and D dimer were tested by Egger method in this study. The results were all greater than 0.05, and no significant publication bias was found.

4. Discussion

Cardiovascular disease is one of the common complications of RA, including coronary heart disease (CHD), heart failure, pericarditis, myocarditis, arrhythmia, and so on. The high incidence and elevated tendency of CHD appeared in patients with RA^[20] suggest that we should pay attention to the diagnosis and prevention of CHD in patients with RA. This study included 17 pieces of literature. The risk factors involved were DAS28 score, RF, anti-cyclic citrulline peptide antibody, platelet, C-reactive protein, ESR, TC, TG, lipoprotein, hypertension, diabetes, blood glucose, uric acid, D-dimer, coefficient of variation in red blood cell distribution, interstitial lung disease, osteoporosis, fracture, age, late onset of RA, long duration of RA disease, current smoking, previous smoking, hydroxychloroquine, leflunomide, high activity of RA, number of extraarticular organs involved, low-density lipoprotein cholesterol, sex, age of myocardial infarction, etc. However, only 9 risk factors of DAS28, RF, TG, age, sex, CRP, TC, hypertension, and D dimer were included in this meta-analysis, for few eligible studies explored other risk factors, such as ESR and uric acid.

Previous studies have shown that smoking, obesity, hypertension, hyperlipidemia, and diabetes are risk factors for CVD. Patients with RA have a higher incidence of obesity, hypertension, hyperlipidemia, and impaired glucose tolerance. These risk factors are more apparent when RA is active. However, the underlying mechanisms of RA combined with CHD were not clear yet.^[21] The risk of CHD in high cholesterol groups was 1.7 times higher than in the control group, and the risk in the hypertension group is 3.58 times higher than in the control group, which were consistent with previous reports. Triglycerides, age, sex, blood sugar and other risk factors were not included in the final meta-analysis due to the lack of literature.

RA and atherosclerosis are chronic inflammatory diseases, and inflammation plays the key role in cardiovascular risk in patients with RA.^[22] In the acute phase of RA, a large number of cytokines, such as interleukin-1, interleukin-6 and tumor necrosis factor-α, are synthesized, and the inflammatory condition is possible to change traditional risk factors such as blood lipid, induce insulin resistance, increased oxidative stress, endothelial dysfunction and increased arterial stiffness, all of which can accelerate the progress of coronary plaque.^[23–25] Elevated CRP suggests active systemic inflammation, associated with a high risk of cardiovascular and cerebrovascular events, and predicts future cardiovascular events. In RA patients with no history of vascular disease, high activity of hs-CRP and RA increased troponin.^[26] In patients with RA, plaque rupture incidence increased due to hypercoagulation of inflammation, and the risk of nonischemic heart diseases such as myocarditis, cardiac insufficiency and heart failure also increased. The results showed that CRP and D dimer were the main risk factors for RA with CVD, and the risk in the case group was 2.06 and 2.83 times higher than that in the control group, which was consistent with the above study.

Therefore, doctors should keep an eye on patients with RA complicated with CVD. On the 1 hand, they should strengthen education and advocate healthy lifestyles, including smoking cessation, mediterranean diet, reasonable exercise, blood lipid and blood pressure control, etc. On the other hand, we could decrease the risk of cardiovascular disease in RA patients based on monitoring the disease activity and targeted treatments to achieve clinical remission. It is of great significance to effectively control the traditional cardiovascular risk factors, inhibit the inflammatory response of RA, and provide active treatments in the early stage of CVD. Through these ways, the quality of life in RA patients could be increased with the mortality rate reduced.

The following deficiencies exist in this study. Firstly, the literature included were case-control and cohort studies with limited argumentation intensity; secondly, individual heterogeneity existed among studies; furthermore, some studies had inconsistent definitions of important risk factors, which are difficult to merge. Eventually, other influential factors of RA combined with CVD were reported in other literature but could not be included in this meta-analysis due to less data or other reasons, which influenced the results. In the future, it is necessary to carry out more relevant research to grasp the risk factors of RA combined with CVD more scientifically and comprehensively.

Acknowledgments

We thank all of the participating members who responded to the research.

Author contributions

Conceptualization: Qian Lyu.

Data curation: Qian Lyu, Huijie Liu.

Formal analysis: Qian Lyu, Huijie Liu.

Investigation: Jihong Wang.

Methodology: Linxiao Ma.

Project administration: Linxiao Ma.

Resources: Jihong Wang.

Software: Jihong Wang.

Writing – original draft: Qian Lyu.

Writing – review & editing: Linxiao Ma, Huijie Liu.