In patients with rheumatoid arthritis (RA), the leading cause of mortality is accelerated atherosclerotic cardiovascular disease (CVD), with standardized mortality ratios between 1.3 and 3 12. Patients with RA have a twofold increased risk for myocardial infarction (MI), and a 10-year risk of CV events that is 60% higher than that in the general population 34. The increased CV risk in RA patients is considered secondary to both disease-specific mechanisms associated with enhanced inflammatory burden, and to traditional CV risk factors (CRFs), such as diabetes mellitus (DM)/insulin resistance, smoking, obesity, hypertension, and dyslipidemia 5678.

RA and DM share a similarly increased risk of CV events 910. However, although the frequency and severity of preclinical atherosclerosis is equal in RA and DM of similar duration, a differential impact seems to exist when comparing traditional risk factors versus systemic inflammation in both diseases. Traditional risk factors seem to play a greater role in CVD associated with DM, whereas systemic inflammation appears to play a greater role in RA 11. Preliminary guidelines exist for CV risk prevention in RA, but these are certainly far more clearly established and validated in DM 1213.

We hypothesized that traditional CRFs are not as frequently managed in patients with RA, as compared with patients with DM or with the general population (GP). We based this assumption on several observations. First, the awareness of the increased CV risk and primary prevention guidelines differs between RA and DM, as supported by the discrepancy in the strength of CV risk-prevention guidelines 1213. Second, several studies indicate potential suboptimal identification and treatment of CRFs in RA 1415. Third, as in DM, symptoms of angina and MI often go unrecognized in RA. Patients with RA are twice as likely to develop silent MIs and sudden cardiac death than is the general population 16. Finally, it has been shown that patients with chronic illnesses have unrelated conditions that are often undertreated 17.

This study analyzed how frequently traditional CRFs were identified and managed by rheumatologists compared with PCPs in patients with RA in a tertiary care center. In addition, this study assessed how frequently these traditional CRFs were identified and managed by PCPs among patients with RA, DM, and the GP.

In this cohort study performed at a tertiary-care center in the United States, we found that rheumatologists identify and manage CRFs in RA patients significantly less frequently than PCPs. This is in agreement with another study that reported that RA patients are better screened for CRFs when also followed up by a PCP rather than only by a rheumatologist. In particular, that study reported that RA patients who were followed up only by a rheumatologist, underwent lipid screening 22% of the time versus 43% to 51% when PCPs were also involved 15. These results indicate the need for improved collaboration between PCPs and rheumatologists to screen RA patients for primary CRF prevention.

In addition, our study found that abnormal BMI, BP, and lipid profiles are most frequently identified and managed by PCPs in patients with DM than in those with RA. More surprising is that abnormal BMI is more frequently identified and managed by PCPs in the GP when compared with patients with RA. This indicates that, although patients with RA have a well-established increased CV risk, equivalent to the risk observed in DM, traditional CRFs in RA are not adequately identified and managed, even to a degree comparable with that of the GP 21. This underrecognition and underassessment of CV risk in RA patients is a finding further discussed in a recently published study that involved sending questionnaires to PCPs and asking them about their awareness of and clinical practices regarding CV risk and RA 22. In that study, only 32% of PCPs identified RA as an independent risk factor for CVD, and 15% of PCPs assessed RA patients for primary prevention of CV risks. In addition, they discovered that PCPs who had received some type of education on increased CV risk in RA more frequently assessed CV risk. However, of those PCPs, only 40% performed primary prevention risk assessment. These findings highlight the suboptimal awareness of CVD in RA among primary care providers, and further support our results.

Perhaps if better established primary prevention guidelines existed regarding CRF management in RA, as proposed by EULAR, more active management of these CRFs by providers would be observed 13. Current models to predict the risk of CVD for the general population, such as the general Framingham and the Reynolds risk scores, underestimate CV risk in patients with RA, according to one study 23. Indeed, the observed CVD risk in RA patients was twofold higher in women and 65% higher in men, when compared with that predicted by the Framingham risk score. These results emphasize the need for more accurate models to predict CV risk in RA. In addition, further studies should be pursued to determine the efficacy of the available EULAR recommendations, to better formulate and establish primary prevention guidelines regarding CRF management in RA.

The baseline characteristics in our study reveal no differences in the prevalence of some of the traditional CRFs and previous CV events between RA patients and the GP, in contrast to what may be expected, given the known increased CV risk in RA and prior findings in the published literature 6724. Also, RA patients in our study were found to have significantly lower SBP and significantly higher HDL when compared with the GP. This is also in contrast with other studies that reported lower HDL levels in RA compared with the GP 725. These findings may be because the average age of patients in our study was 49 years, 8 years younger than the average age of patients in studies reporting higher prevalence of CAD, MI, and CVA, and/or lower HDL levels 325. Furthermore, White et al. 26 found a statistically significant inverse relation between HDL and CRP levels. RA patients in our study were found to have relatively well-controlled disease, with overall low CRP levels. This could potentially explain the higher-than-expected HDL levels. Also, the mere quantification of HDL in RA may not reveal true CV risk, as these patients have higher prevalence of oxidized HDL, which loses its antiinflammatory properties and may become proatherogenic 27. In addition, our study excluded patients with RA coexisting with DM to reduce confounding factors. This may have skewed our baseline characteristics, given that these patients with the highest risk were excluded.

Our study found that DM patients had significantly lower LDL levels compared with the other two groups, which may be because a significantly higher percentage of diabetes patients were taking a statin (P < 0.0001). As expected, DM patients had significantly higher BMI, SBP, and triglyceride measurements.

Limitations of our study include that rheumatologists may have not addressed certain CRFs in RA because they were aware that their patients were also being followed up by a PCP. It is possible that the same rheumatologist would have more actively identified and managed these CRFs if the patient were not seen by a PCP. Similarly, rheumatologists may not have addressed CRFs with RA patients if these patients were first examined by a PCP. It should be kept in mind that, although PCPs were shown to identify and manage CRFs better in RA compared with rheumatologists, they still did not adequately address their CRFs when compared with the GP. Therefore, even when patients are followed up by PCPs, rheumatologists should be vigilant about the detection and management of CV risk in RA patients.

A second limitation of our study is that the data were gathered retrospectively from medical charts. Hence, if the provider identified or managed a CRF but did not document it, the particular CRF would be considered as not addressed for the purposes of our study. We assume that this likely occurred equally in all three groups, so the relative discrepancies seen would still be present.

Our study has several strengths, including that we performed an extensive chart review of a large cohort of patients (n = 753). Although a prior study assessed the management of individual CRFs, to our knowledge, no other studies performed such a thorough assessment of the management of multiple CRFs in RA 15. In addition, no other studies have simultaneously assessed CRF management in patients with RA when compared with those with DM, a more-established CAD risk equivalent.

Finally, our study accounted for the overall CVD history and for individual risk-factor abnormalities in the subgroup analyses. We found that, at baseline, there was about a fivefold increase in the prevalence of a history of CVD in DM as compared with RA or GP patients. Therefore, the higher frequency of CRF identification and management found in DM patients could potentially have been due to the increased prevalence of CVD in this group. However, when excluding patients with a history of CVD, we found that PCPs still identified and managed certain CRFs most frequently in DM, followed by GP, and then RA patients. These findings persisted when looking at patients without CVD and with abnormal CRF values.

In summary, we found that despite rheumatologists' awareness of increased CV risk in RA patients, they are significantly less likely to identify and manage CRFs when compared with PCPs. In addition, the management of CRFs by general practitioners in RA patients is suboptimal, when compared with strategies implemented in the GP and DM patients. We hypothesize that health care providers spend more time focusing on rheumatologic issues rather than seemingly less-pressing primary prevention issues. However, given the well-established increased CV risk associated with RA, both rheumatologists and PCPs should collaborate and more aggressively identify and manage CRFs in patients with RA to minimize their deleterious effects on morbidity and mortality.

CAD: coronary artery disease; CRF(s): cardiovascular risk factor(s); CV: cardiovascular; CVD: cardiovascular disease; DM: type 2 diabetes mellitus; GP: general population; HDL: high-density lipoprotein; HL: hyperlipidemia; HTN: hypertension; LDL: low-density lipoprotein; PCP: primary care physician; RA: rheumatoid arthritis.

The authors declare that they have no competing interests.

SD participated in the design of the study, collected the data, and wrote the manuscript. JM participated in the design of the study and performed the statistical analysis. MK conceived of the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the manuscript for publication.