Article Text
Abstract
Objectives To assess the prevalence of coronary heart disease (CHD) and chronic hypertension among patients with rheumatoid arthritis (RA) at the time of diagnosis, in comparison with age-specific and sex-specific non-RA subjects. Furthermore, the impacts of age at the onset of RA, as well as gender and the presence of rheumatoid factor (RF) on the risk of these comorbidities, were evaluated.
Methods A cohort of 7209 RA patients diagnosed between January 2004 and December 2007 was identified, based on a Finnish nationwide register on special reimbursements for medication costs. The presence of CHD and chronic hypertension antedating the diagnosis of RA was identified from the same register. The prevalence of the cardiovascular comorbidities was compared with the general Finnish population, and a standardised rate ratio (SRR) for both these cardiovascular diseases was calculated.
Results The risk of having CHD at RA diagnosis was slightly elevated, the SRR being 1.10 (95% CI 1.01 to 1.20). Younger age at the onset of RA seemed to be related with higher SRR for CHD. In a subset analysis, an increased prevalence of hypertension (SRR 1.19, 95% CI 1.10 to 1.30) and CHD (SRR 1.15, 95% CI 1.00 to 1.32) was apparent only among the RF negative RA cases.
Conclusions The SRR for CHD is augmented in RA patients already at disease onset, and more pronouncedly in early onset RA. The findings highlight the importance of early prevention of atherosclerosis, regardless of RF status.
- Rheumatoid Arthritis
- Cardiovascular Disease
- Rheumatoid Factor
- Epidemiology
Statistics from Altmetric.com
Introduction
Rheumatoid arthritis (RA) is a chronic inflammatory disease which is associated with increased morbidity and mortality from cardiovascular disease.1 ,2 This is partly due to traditional cardiovascular risk factors, as well as systemic inflammation and autoimmune processes involved in RA pathogenesis.3 ,4 The link between RA and cardiovascular disease may also involve similar predisposing factors,4 and multiple gene polymorphisms have been shown to associate with an increased cardiovascular risk in RA.5–7
There are many temporal aspects of RA-related cardiovascular risk that have not yet been clarified. Initial data is controversial on whether coronary heart disease (CHD) is more prevalent than expected at RA onset.8 ,9 CHD in established RA may occur at a younger age than in the general population,10 and the rate ratio for cardiovascular events and death is highest in young RA patients, albeit the rate difference increases with age.11 There are also implications that a young age at the onset of RA and rheumatoid factor (RF) positivity are associated with increased risk of cardiovascular death over the course of years.12 None of the previous studies have assessed if the cardiovascular risk antedating RA diagnosis is affected by age at the onset of RA.
The major aim of this national registry study was to evaluate the prevalence of CHD and chronic hypertension among RA patients at the onset of the disease, compared with age-specific and sex-specific non-RA subjects. Other objectives were to determine whether age at the onset of RA, gender or the presence of RF exert any impact on the risk of presenting these diseases.
Methods
In Finland, all permanent residents are covered under the National Health Insurance, and patients with certain chronic and severe diseases are entitled to special reimbursement of defined medications. When prescribing these medications to these patients, the treating physician routinely, irrespective of the place of residence and the socioeconomic status of the family, makes a certificate for application of the special reimbursement. With regard to RA, a patient must file an examination-based medical certificate issued by a rheumatologist or physician with experience in rheumatology, in order to establish entitlement. The certificate must describe the proper diagnostic procedures and ICD-10 code (M05 for RF positive RA and M06 for RF negative RA, no other rheumatic diseases were included) as well as a treatment plan in accordance with good clinical practice. The certificates are checked and approved by a medical examiner physician at the Social Insurance Institution before the reimbursement is granted. The reimbursement system covers any antirheumatic medication needed to control the disease activity of each patient, also during remission.
Patients with CHD are similarly entitled to special reimbursement for medication costs. The medical certificate by their physician must show that the objective criteria of CHD are fulfilled, including one of the ICD-10 codes: I20-I22, I24.0 or I25. Chronic hypertension is the most common disease entitling to special reimbursement of medications in Finland, covering more than 10% of the population. To establish entitlement, the patient must have been treated for at least 6 months. The certificate must include an ICD-10 code of I10-I13, I15 or I27.0, and describe either clearly elevated blood pressures or the presence of an end-organ complication. Thus, patients with only mild hypertension are excluded.
The Social Insurance Institution gathers the special reimbursement decisions in a nationwide register. From that register, we collected all the patients with codes M05 and M06 who were granted special reimbursement for RA medications for the first time between 1 January 2004 and 31 December 2007. The prior existence of CHD and chronic hypertension was collected, with the requirement that the cardiovascular disease medication reimbursement decision had to antedate the RA medication reimbursement.
We only used officially archived register data, which is available without any patient-specific consent. Permission for the usage of the data was obtained from the administrator of each register, that is, the Social Insurance Institution and Statistics Finland. Every patient was given an identification number, according to which the data was analysed anonymously.
Statistical analysis
Standardised rate ratios (SRR) for CHD and chronic hypertension with 95% CI were calculated assuming a Poisson distribution, and adjusted by using the age-specific and sex-specific Finnish population (from Statistics Finland) as a reference.
Results
Characteristics of the cohort
We identified 7209 RA patients over 20 years of age who were granted the special reimbursement of medications for RA between 1 January 2004 and 31 December 2007. Of them, 4887 (68%) were women (64% RF+) with a mean age of 56.6 (SD 15.0) years and 2322 (32%) were men (66% RF+) whose mean age was 58.2 (SD 13.5) years.
Coronary heart disease (CHD)
A total of 533 patients, 355 (67%) of which were RF positive, had a validated diagnosis of CHD at the diagnosis of RA, corresponding to a slightly elevated CHD risk compared with non-RA subjects (table 1). The SRR was fairly similar in RF positive and RF negative patients, but only that of the RF negatives reached statistical significance (table 2). Figure 1 illustrates the SRRs stratified by gender and age at the onset of RA. The patients with earlier onset of RA showed a trend towards a higher SRR for CHD than patients who were older at the time of RA diagnosis, although the elevated SRRs were mostly no longer significant in the age subgroups. No gender differences with respect to CHD among RA patients compared with the general population could be detected.
Chronic hypertension
There was no excess prevalence of chronic hypertension at the time of RA diagnosis in the RA population as a whole or in the RF positive subgroup. The risk of having chronic hypertension prior to RA was slightly elevated only among the RF negative patients, the SRR being 1.19 (95% CI 1.10 to 1.30) (tables 1 and 2).
Discussion
Coronary heart disease (CHD)
A large body of evidence implies that subjects with RA are at increased risk of CHD early in the disease process.4 ,9 ,13 Nevertheless, it remains to be ascertained whether the risk of CHD is already present before the onset of the RA symptoms, or whether it appears rapidly after the first indications of the disease.13 ,14 Maradit–Kremers and coworkers reported in a population-based cohort study, that patients with RA were at an increased risk of CHD during the 2-year period before the fulfilment of the 1987 American College of Rheumatology (ACR) criteria for RA.9 Holmqvist and colleagues, however, did not find any evidence of an increased risk of CHD before the onset of RA symptoms in two large Swedish cohorts,8 ,14 whereas, a subsequent investigation disclosed an increased myocardial infarction risk after 1 year since the fulfilment of ACR criteria for RA.14
Our results amplify these findings with the observed slightly but significantly increased SRR for CHD among RA patients prior to the date of RA diagnosis. As this date resides some time after the onset of RA symptoms, the premise of a rapid increase in the clinical manifestation of CHD following the onset of RA symptoms, if nothing else, is not ruled out. Alternatively, RA being a complex, multifactorial and polygenic disease, the current observation may support the role of shared risk factors partly accounting for the increased risk of cardiovascular disease in newly diagnosed RA. For instance, gene polymorphisms located both in7 ,15 and outside5 ,6 ,16 the major histocompatibility complex (MHC) region contributing to approximately half the genetic susceptibility for RA17 have been considered to modulate the risk of endothelial dysfunction,6 ,15 ,16 and to increase the risk of cardiovascular disease in RA.5–7
Identification of RA patients in the present study relied on the combination of a rheumatologist's clinical diagnosis, commencement of antirheumatic medication, and the allowance of special reimbursement for that medication. The reimbursement register does not contain data on the fulfilment of any classification criteria for RA. Some of the patients may not yet have fulfilled the ACR criteria, and our cohort may include patients with earlier RA than previous studies. This suggestion is supported by a Finnish study that also made use of the Social Insurance Institution register, as well as reported a slightly higher incidence of RA over the years 2000–2007, by comparison with earlier studies using the ACR criteria.18 ,19 Our findings, thereby, shift the time scale of increased CHD risk somewhat closer to the first symptoms of RA than that observed in previous investigations.
Interesting observations have been made about the impact of age on the increased cardiovascular risk of RA patients. Data from the Rochester Epidemiology Project demonstrated a decline in the rate ratio, but an increase in the absolute rate difference of cardiovascular disease among RA patients with increasing age.11 The finding was recently repeated in a large Danish nationwide cohort study using myocardial infarction as the end-point.20 However, these studies considered the impact of age in general in established RA populations, whereas, we examined the impact of age at the onset of RA on the rate ratio of CHD among those who subsequently developed RA.
The specific effect of age at the onset of RA has been examined with regard to the course and outcome of arthritis itself,21–23 but not as extensively when considering the associated increased cardiovascular disease risk. Recently, Naz and colleagues examined patients with inflammatory polyarthritis and found the highest rates of cardiovascular mortality among the RF positive patients who were younger than 55 years at symptom onset.12 Also, a decade ago, a large Swedish study on 46 917 RA patients showed that the excess risk of death from CHD in established RA was especially high with early onset of disease.24
In our study population, the rate ratio of CHD antedating RA diagnosis seemed higher among patients who developed RA earlier in life, compared with those who were older at the index date (figure 1), but there was probably a lack of power to prove this trend as statistically significant. Speculatively, it is possible that earlier-onset RA is more pathogenic with regard to CHD than later-onset RA. On the other hand, the concept of competing risks is obviously of prime importance: when a particular risk such as age increases, the contribution of another risk factor to the overall risk loses its significance.11 The absolute excess risk of CHD is a separate question, and might be constant, or even increased, in patients with late-onset RA compared with early onset RA.11
While both subgroups divided by the RF status had slight numerical increases in the risk of CHD, only those without RF reached significance despite being the smaller pool. On the contrary, a recent Finnish study conducted among an RA cohort similar to ours showed that the presence of RF remained a risk factor for overall mortality over the years 2000–2008,25 and the RA-related cardiovascular risk has often, but not always, been more profound with positive RF.4 Our current findings do not constitute a contradiction with previous literature, as we looked at the cardiovascular risk before RA while the earlier studies focussed on established RA cohorts.25 RF positivity is associated with a more severe RA and a poorer prognosis,26 and it might be that it is this more fiery chronic inflammation of the RF positive patients that cumulatively deteriorates their cardiovascular status in comparison with the RF negative ones over the course of years. Holmqvist and colleagues, who studied the myocardial infarction risk also at the time of RA diagnosis, did not find RF positivity to be a risk stratifier.14
Gender did not appreciably modify the rate ratio of CHD (figure 1). This finding is not unique;11 ,14 ,20 ,27 the cardiovascular burden seems to be equally elevated in both men and women.
Smoking, non-steroidal anti-inflammatory drugs (NSAIDs) and obesity could interfere with the relation of CHD to RA as well as RF status or age at the onset of RA. Unfortunately, we do not have data on these variables. Smoking is traditionally associated with RF positive RA28 and, therefore, is not likely to contribute to increased CHD risk among the RF negative subset. Hypothetically, smoking could contribute more to the development of early onset RA, and thereby explain some of the observed excess prevalence of CHD among patients who were younger at the time of RA diagnosis. Another potential confounder in this study is the use of NSAIDs for joint pain before the diagnosis of RA, as they all have a somewhat unfavourable cardiovascular risk profile.29 However, the use of any other RA medication, such as corticosteroids, does not induce bias since the patients get their prescriptions and start using antirheumatic medications synchronously with the certificate of entitlement to special reimbursement for those medications, indicating the end of our observation period.
Chronic hypertension
Despite the bulk of previous investigations, the prevalence of hypertension in RA patients, compared with the general population, is still a matter of debate.30–33 A recent meta-analysis of 15 case-control studies examining the prevalence of cardiovascular risk factors among RA patients and controls reported a similar prevalence of hypertension in both groups.34 In a subset analysis, we found a small increase in the risk of chronic hypertension among the RF negative subgroup. The result is in line with our finding regarding CHD. The pathophysiologic background of the present finding remains elusive.
Study limitations
We were likely to underestimate the prevalence of cardiovascular disease among both RA patients and the general population, because milder forms of the studied diseases do not necessarily entitle patients to special reimbursement of medications. What was noteworthy was that a significant part of hypertensive patients were likely to fall out of our reach. Theoretically, the possible overlap in the medications for various cardiovascular diseases may enable a patient with two or more cardiovascular diseases to get their medication reimbursements with only one reimbursement diagnosis. However, the imperfect coverage of patients suffering from cardiovascular disease is likely to concern RA patients and the reference population equally. The date of diagnosis is likely to reside a varying period of time after the clinical onset of the disease, since the traditional hallmarks of RA develop over time.35 In addition, the lack of access to patients’ medical records hindered us from accounting for many important confounders, such as smoking, and from surveying more specific data on the patients’ diseases than the diagnosis, limiting the strength of inference.
As a conclusion, the prevalence of CHD is augmented in RA patients already at disease diagnosis, and more pronouncedly in early onset RA. The findings highlight the importance of the early prevention of atherosclerosis in RA regardless of RF status.
References
Footnotes
-
Handling editor Tore K Kvien
-
Contributors The study was designed by all the authors. AMK and TVMN did the literature search. Data collection was performed by KP, LJV and HK, the last of whom conducted the statistical analysis. Interpretation of findings lay on all members of the research group. The first manuscript was written by AMK under the guidance of TVMN, and the document was finalised by all members of the research group. All the authors approved the manuscript as submitted.
-
Funding The Onni and Hilja Tuovinen Foundation, the Research Fund of the Päijät-Häme Central Hospital.
-
Competing interests None.
-
Provenance and peer review Not commissioned; externally peer reviewed.