Increased coronary-artery atherosclerosis in rheumatoid arthritis: Relationship to disease duration and cardiovascular risk factors

Authors


Abstract

Objective

To compare the prevalence and severity of coronary-artery atherosclerosis in patients with early and established rheumatoid arthritis (RA) and controls.

Methods

Electron-beam computed tomography was used to measure the extent of coronary-artery calcification in 227 subjects, of whom 70 had early RA, 71 had established RA, and 86 were controls. Coronary-artery calcification calculated according to the Agatston calcium score was compared in patients and controls, and its relationship to clinical characteristics was examined. Adjusted odds ratios (ORs) were obtained with the use of proportional odds logistic regression models to determine independent associations of early and established RA and coronary-artery calcification.

Results

Calcium scores were higher in patients with established RA (median 40.2, interquartile range [IQR] 0–358.8) compared with those with early disease (median 0, IQR 0–42.6) and controls (median 0, IQR 0–19.2) (P = 0.001). Coronary-artery calcification occurred more frequently in patients with established RA (60.6%) than in patients with early RA (42.9%) and control subjects (38.4%) (P = 0.016) The OR for the likelihood of having more severe coronary-artery calcification (defined as an Agatston score >109) in patients with established disease was 3.42 (P = 0.002) after adjusting for cardiovascular risk factors. Among patients with RA, smoking (OR 1.02, P = 0.04) and an elevated erythrocyte sedimentation rate (OR 1.02, P = 0.05) were associated with more severe coronary-artery calcification after adjustment for age and sex.

Conclusion

The prevalence and severity of coronary calcification is increased in patients with established RA and is related, in part, to smoking and an increased erythrocyte sedimentation rate.

Rheumatoid arthritis (RA) is a common inflammatory disease that affects 0.5–1% of the population (1) and is associated with increased mortality rates (2–4). Epidemiologic studies have shown that this increased mortality is largely attributable to cardiovascular disease, primarily coronary heart disease (3). Considerable evidence suggests that inflammation plays a role in the pathogenesis of atherosclerosis (5–7). In several animal models that were genetically modified to produce higher levels of inflammatory mediators, atherosclerosis was more severe (8, 9), and inflammatory cells and mediators have been found in human atherosclerotic plaque (10). Furthermore, higher concentrations of markers of inflammation, such as levels of C-reactive protein (CRP) (11) and interleukin-6 (12), are predictive of the risk of coronary heart disease. Therefore, premature atherosclerosis could represent a consequence of chronic inflammation, as occurs with RA.

The results of some studies in which indirect indicators of coronary atherosclerosis, such as carotid-artery intima-media thickness, were used suggest that atherosclerosis is more prevalent in patients with RA as compared with controls (13–15). However, the prevalence of coronary-artery atherosclerosis and the mechanisms underlying the increased risk of ischemic heart disease in RA patients are not known. The presence and extent of coronary-artery calcification can be determined by computed tomography (CT) (16), and a correlation has been found between this calcification and the severity of atherosclerosis measured in pathologic specimens (17) or by angiography (18, 19). Furthermore, there is accumulating evidence that coronary-artery calcification may be predictive of the risk of both myocardial infarction and death due to coronary heart disease (20–22). Recently, we used electron-beam CT to show that the prevalence of atherosclerosis is markedly increased in patients with systemic lupus erythematosus (SLE) (23). RA and SLE are both characterized by chronic inflammation, but their pathogenesis differs. These differences may provide insight into the relationship between chronic inflammation and the pathogenesis of atherosclerosis.

In addition, structural damage to joints occurs early in the course of RA, usually within the first few years of disease (24), and thus structural changes in the vasculature of the coronary artery could also occur at an early time point. We tested the hypothesis that patients with RA have more severe coronary-artery atherosclerosis than that in control subjects of similar age, race, and sex, and that this increased severity is related to the clinical characteristics of the patients.

PATIENTS AND METHODS

Subjects.

We studied 141 patients with RA and 86 control subjects. Consecutive eligible patients older than age 18 years who met the American College of Rheumatology (formerly, the American Rheumatism Association) classification criteria for RA (25) and had a duration of disease of <5 years (early RA) or >10 years (established RA) were enrolled. Controls did not meet the classification criteria for RA or any other inflammatory disease. Control subjects were frequency matched for age, sex, and race with the entire group of RA patients so as to ensure that the control group would not differ markedly from either the early or established RA groups with respect to these variables. Recruitment continued until a minimum of 70 patients had been enrolled in each group.

Patients were obtained from a registry of patients with early RA (26), were referred by local rheumatologists, or were recruited by advertisements. Control subjects were recruited from among the patients' acquaintances, by advertisement, or from a database of volunteers maintained by the General Clinical Research Center. The study was approved by the institutional review board of Vanderbilt University Hospital, and all subjects gave their written informed consent.

Information was obtained through a structured interview, self-report questionnaires, physical examination, laboratory tests, and electron-beam CT, and the medical records of all patients were reviewed. Current and cumulative medication use was determined both from the information provided by the patients and from the medical records. A family history of coronary-artery disease was defined as a first-degree relative having a myocardial infarction or stroke before the age of 55 years in men, and before the age of 65 years in women. Height and weight were measured and body mass index was calculated. Blood pressure was determined as the average of 2 measurements obtained 5 minutes apart after subjects had rested in the supine position for at least 10 minutes. Subjects were considered to have hypertension if they were taking antihypertensive agents or if they had a systolic blood pressure of ≥140 mm Hg and/or a diastolic pressure of ≥90 mm Hg. In patients, disease activity was measured using the Disease Activity Score based on the evaluation of 28 joints (DAS28) (27). The DAS28 is a validated composite index containing a 28-joint count for tenderness, a 28-joint count for swelling, the erythrocyte sedimentation rate (ESR), and the patient's overall assessment of well-being. Ability to perform activities of daily living was measured using the modified Health Assessment Questionnaire (M-HAQ) (range 0–3, with a score of 0 representing no impairment of function) (28).

Patients were required to fast overnight prior to collection of blood for determination of a complete blood cell count as well as measurement of creatinine, total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides, lipoprotein(a) (Lp[a]), and homocysteine. In patients with RA, the level of CRP and the Westergren ESR were used, along with information from the medical records, to determine the presence or absence of rheumatoid factor.

Imaging procedures.

All subjects underwent imaging with an Imatron C-150 scanner (GE/Imatron, South San Francisco, CA). Imaging was performed with a 100-msec scanning time and a single-slice thickness of 3 mm. Forty slices were obtained during a single breath-holding period, with the area of the slices starting at the aortic arch and proceeding to the level of the diaphragm. Tomographic imaging was electrocardiographically triggered at 60% of the interval between R waves. All areas of calcification within the borders of a coronary artery with a minimal attenuation of 130 Hounsfield units (HU) were identified. A calcified coronary plaque was considered present if at least 3 contiguous pixels were detected (voxel size 1.03 mm3). If a patient had undergone prior coronary-artery stenting or bypass surgery, the stent and all metal clips were excluded from the computation of coronary-artery calcification.

The acquired images were reviewed at the core electron-beam CT laboratory on a NetraMD workstation (ScImage, Los Altos, CA). Subjects were included in this study only if complete data were available from their scans, without misregistration of slices due to artifacts of motion, respiration, or asynchronous electrocardiographic triggering. To ensure the continuity and consistency of the interpretation of scores, a single expert investigator (PR), who was unaware of the subjects' clinical status, read all the scans.

Calculation of calcium scores.

The degree of coronary-artery calcification was calculated using the method described by Agatston et al (16). Briefly, the area of each calcified plaque was multiplied by the peak radiologic attenuation inside the area (which was expressed as a coefficient of 1 for an attenuation of 130–199 HU, 2 for an attenuation of 200–299 HU, 3 for an attenuation of 300–400 HU, and 4 for an attenuation of >400 HU). The sum of the scores for all coronary-artery lesions represented the overall Agatston calcium score for each individual (graded as 0 for none, 1–109 for mild, and >109 for moderate-to-severe coronary-artery calcification).

Statistical analysis.

With the assumption that the frequency of coronary-artery calcification in control subjects would be ∼44% (29) and to achieve 85% power to detect a difference at a 2-sided significance level of 5%, the study was required to have at least 70 subjects in each of the 3 groups in order to detect a 25% relative increase in coronary-artery calcification (55% for early RA and 69% for established RA). Sample-size calculations were performed using Pearson's chi-square test for a trend of 3 proportions, with nQuery Advisor version 4.0 (Statistical Solutions, Saugus, MA). Baseline characteristics among control subjects and patients with early and established RA are expressed in median values (interquartile range [IQR]) for continuous variables and proportions for categorical variables.

Statistical analyses were performed in 2 phases. In the first phase, the prevalence of coronary-artery calcification and coronary risk factors was compared in control subjects and patients with early and established RA using Kruskal-Wallis tests for continuous variables and Fisher's exact tests or chi-square tests for categorical variables. Proportional odds logistic regression was used to test the independent relationship between disease status and severity of coronary-artery calcification. Coronary-artery calcification in all individuals having an Agatston calcium score >0 (with a cutoff point for severity based on the median score of 109) was treated as the dependent variable, and disease status (none, early disease, and established disease) was treated as the independent variable using 2 indicator variables for comparing each disease status versus none (dummy variables). Severity was assessed based on the 3 categories of the Agatston calcium score (none, mild, and moderate-to-severe coronary-artery calcification).

Multivariable analysis using proportional odds logistic regression was performed in 2 steps. Since the prevalence and extent of coronary-artery calcification in the general population is greater in men and increases with age (19), these were the first variables included in the model. We then added race, smoking, hypertension, body mass index, levels of creatinine, cholesterol, triglycerides, and Lp(a), diabetes, concentration of homocysteine, and family history of coronary disease as variables in the model.

In the second phase, we analyzed only patients with RA. Continuous variables were compared using the Mann-Whitney U test and categorical variables using Fisher's exact test. Characteristics of the patients within the 3 ordinal categories of the Agatston scores (none, mild, and moderate-to-severe) were compared using Kruskal-Wallis tests and Fisher's exact tests. Unadjusted and adjusted odds ratios (ORs) were calculated using the proportional odds logistic regression model to determine the independent associations with the 3 ordinal categories of the Agatston calcium score.

All statistical analyses were conducted using SAS software version 8.02 (SAS Institute, Cary, NC). A 2-sided significance level of 5% was used.

RESULTS

Characteristics of patient and control groups.

Patients with RA (n = 141) and control subjects (n = 86) were of similar age (median age of patients 54 years, IQR 46–64 years versus controls 52 years, IQR 44–59 years) and had a similar sex distribution (female patients 68.8% versus female controls 65.1%) and race distribution (white patients 89.4% versus white controls 86.1%). Study subjects were predominantly well-educated, with 87% of subjects having at least 12 years of education. The median duration of disease from the time of first symptom was 2 years (range 1–3) in patients with early RA and 20 years (range 14–24) in those with established disease.

The demographic characteristics and cardiovascular risk factors in the 70 patients with early RA, 71 patients with established RA, and 86 controls are shown in Table 1. A coronary-artery procedure (angioplasty, stent, or bypass graft) had been performed in 5.8% of control subjects and in 5.7% and 9.9% of patients with early and established RA, respectively. The prevalence of hypertension was 67.6% in patients with established RA, 38.6% in those with early RA, and 38.4% in control subjects (P < 0.001). Of the control subjects, 9.3% currently smoked, whereas 22.9% and 29.6% of patients with early and established RA, respectively, were current smokers (P = 0.005); exposure to smoking in pack-years was similar among the 3 groups. Serum concentrations of total, HDL, and LDL cholesterol as well as Lp(a) and triglyceride levels were similar among patients and controls (Table 1).

Table 1. Clinical characteristics of patients with rheumatoid arthritis (RA) and control subjects*
CharacteristicControl subjects (n = 86)Early RA (n = 70)Established RA (n = 71)P
  • *

    Continuous variables are presented as the median (interquartile range).

  • P values are by Kruskal-Wallis test or Fisher's 2-tailed exact test, as appropriate, for comparisons among groups.

  • Statistically significant difference between patients with established RA and control subjects.

Age, years52 (44–59)51 (43–59)58 (48–67)<0.001
Female sex, %65.161.476.10.15
White race, %86.194.384.50.14
Total pack-years of smoking0 (0–11)0 (0–22)0 (0–25)0.37
Hypertension, %38.438.667.6<0.001
Family history of coronary heart disease, %27.928.627.10.98
Body mass index, kg/m227.1 (23.4–31.6)28.4 (24.3–33.1)27.0 (24.2–31.6)0.33
Coronary-artery procedure, %5.85.79.90.65
Glucose, mg/dl89.0 (83.0–94.0)86.0 (81.0–93.0)88.0 (83.0–96.0)0.31
Creatinine, mg/dl0.8 (0.7–0.9)0.8 (0.7–0.9)0.8 (0.6–0.9)0.64
Total cholesterol, mg/dl195.0 (171.0–216.0)184.0 (161.0–210.0)189.0 (153.0–214.0)0.34
High-density lipoprotein, mg/dl45.0 (39.0–54.0)40.0 (35.0–50.0)45.5 (38.0–57.0)0.07
Low-density lipoprotein, mg/dl123.0 (105.0–145.0)114.0 (94.0–137.0)114.5 (84.0–136.0)0.10
Lipoprotein(a), mg/dl9.9 (4.0–32.3)7.2 (2.0–23.5)9.6 (4.0–28.0)0.20
Triglycerides, mg/dl106.0 (74.0–135.0)109.5 (85.0–147.0)115.0 (83.0–174.0)0.30
Current use of statins, %11.614.311.30.85
Homocysteine, μmoles/liter8.2 (7.2–9.6)9.6 (8.2–11.4)10.8 (8.2–12.5)<0.001

Coronary-artery calcification.

The prevalence of coronary-artery calcification and the median calcium scores in controls and patients with early and established RA are shown in Table 2. The median coronary-artery calcium score was 40.2 (IQR 0–358.8) in patients with established RA compared with 0 (IQR 0–42.6) in those with early disease and 0 (IQR 0–19.2) in controls (P = 0.001). Coronary-artery calcification was more prevalent in patients with established RA (60.6%) than in those with early RA (42.9%) and controls (38.4%) (overall P = 0.016, patients with early RA compared with controls P = 0.57, patients with established RA compared with controls P = 0.006) (Table 2 and Figure 1).

Table 2. Prevalence of coronary-artery calcification and Agatston calcium scores in patients with early and established rheumatoid arthritis (RA) and control subjects*
 Control subjects (n = 86)Early RA (n = 70)Established RA (n = 71)Established vs. early RA
  • *

    P values are by Kruskal-Wallis test or Fisher's 2-tailed exact test, as appropriate, for comparisons among groups (P = 0.001 for Agatston score; P = 0.016 for prevalence of coronary-artery calcification). IQR = interquartile range; OR = odds ratio; 95% CI = 95% confidence interval.

  • Proportional odds logistic regression was used to obtain unadjusted and adjusted ORs for the likelihood of having severe coronary-artery calcification (defined as an Agatston calcium score >109). Cardiovascular risk factors included age, sex, race, smoking, hypertension, body mass index, levels of creatinine, cholesterol, triglycerides, and lipoprotein(a), diabetes, homocysteine concentrations, and family history of coronary disease.

Agatston score, median (IQR)0 (0–19.2)0 (0–42.6)40.2 (0–358.8)
Prevalence of coronary-artery calcification, %38.3742.8660.56 
Agatston score subgroups, %    
 1–10924.4225.7219.72 
 >10913.9517.1440.84 
OR (95% CI)    
 Unadjusted11.21 (0.65–2.26)3.03 (1.65–5.57)2.51 (1.33–4.71)
  P 0.55<0.0010.004
 Adjusted for age and sex11.31 (0.65–2.64)2.73 (1.36–5.47)2.08 (1.01–4.3)
  P 0.450.0050.048
 Adjusted for cardiovascular risk factors11.44 (0.67–3.10)3.42 (1.55–7.53)2.32 (1.05–5.13)
  P 0.350.0020.037
Figure 1.

The frequency distribution of coronary-artery calcification (by Agatston scores) in patients with rheumatoid arthritis (RA) and healthy control subjects. Values above the bars are the no./total no. of subjects in each score subgroup.

Figure 2 illustrates the prevalence of coronary-artery calcification in controls and patients with RA according to the age categories younger than age 50 years, 50–59 years, and 60 years and older. There was a significant interaction between age and disease status (P < 0.05 for interaction). For those ages <50 years and 50–59 years, the prevalence of coronary-artery calcification was increased in patients with established RA compared with controls (both P < 0.05), but the difference was not significant among groups of subjects ages ≥60 years.

Figure 2.

The prevalence of coronary-artery calcification among patients with rheumatoid arthritis (RA) and control subjects, according to age. There was a significant interaction between age and disease status (P < 0.05 for interaction) for those under the age of 60 years.

The severity of coronary-artery calcification in patients with early RA did not differ significantly from that in control subjects (OR 1.21, 95% confidence interval [95% CI] 0.65–2.26, P = 0.55), but the likelihood of having more severe coronary-artery calcification was increased in patients with established RA compared with controls (OR 3.03, 95% CI 1.65–5.57, P < 0.001) (Table 2). A significant difference was also detected between patients with early RA and those with established RA (OR 2.51, 95% CI 1.33–4.71, P = 0.004). After adjustment for age and sex, the likelihood of having severe coronary-artery calcification remained significantly elevated in patients with established RA compared with controls (OR 2.73, 95% CI 1.36–5.47, P = 0.005) and compared with patients with early RA (OR 2.08, 95% CI 1.01–4.3, P = 0.048). These associations in patients with established RA remained significant after adjusting for the cardiovascular risk factors of age, sex, race, smoking, hypertension, body mass index, levels of creatinine, cholesterol, triglycerides, and Lp(a), diabetes, homocysteine concentrations, and family history of coronary disease (OR 3.42, 95% CI 1.55–7.53, P = 0.002). The results remained unchanged when pack-years of smoking was replaced with the dichotomous variable ever smoked (yes/no).

Relationship between disease characteristics and coronary-artery calcification among patients.

Clinical characteristics differed between patients with early RA and those with established RA (Table 3). As expected, cumulative exposure to medications for treatment of RA was greater in patients with established disease. Patients with early and established RA had similar ratings of disease activity on the DAS28, but the M-HAQ score (P = 0.02) and the ESR (P = 0.04) were higher in patients with established disease. The mean ± SD number of swollen joints was 5.0 ± 5.7 in patients with early RA and 5.2 ± 5.6 in those with established RA. Approximately 17% of patients in both groups were receiving tumor necrosis factor blockade therapy.

Table 3. Characteristics of patients with early and established rheumatoid arthritis (RA)*
CharacteristicEarly RA (n = 70)Established RA (n = 71)P
  • *

    Continuous variables are presented as the median (interquartile range). Mann-Whitney U tests (for continuous variables) and Fisher's 2-tailed exact tests (for dichotomous variable) were used for comparisons between groups. Higher scores for the 28-joint Disease Activity Score (DAS28) and the disease activity visual analog scale (VAS) indicate greater disease activity. Higher scores for the modified Health Assessment Questionnaire (M-HAQ) indicate more difficulty in performing activities of daily living. Higher scores on the pain VAS indicate a higher level of pain.

  • Data were available on 133 patients.

Rheumatoid factor, %68.777.30.27
Erythrocyte sedimentation rate, mm/hour14 (5–24)19 (7–42)0.04
C-reactive protein, mg/liter4.0 (3.0–10.0)5.0 (3.0–16.0)0.66
Number of swollen joints (range 0–28)3.0 (0.0–7.0)4.0 (0.0–9.0)0.75
Number of tender joints (range 0–28)2.0 (0.0–6.0)2.0 (0.0–7.0)0.99
Disease activity VAS score (range 0–10 cm)2.8 (1.2–5.0)3.6 (1.7–5.6)0.10
M-HAQ score (range 0–3)0.4 (0.0–0.6)0.5 (0.1–1.0)0.02
Pain VAS score (range 0–10 cm)3.1 (1.6–5.5)4.2 (2.4–5.8)0.15
DAS283.3 (2.1–3.9)3.4 (2.6–4.4)0.15
Cumulative dose of prednisone, grams1.8 (0.5–3.7)8.5 (1.8–21.0)<0.001
Cumulative dose of methotrexate, grams0.8 (0.1–1.3)3.1 (1.1–6.9)<0.001
Cumulative dose of hydroxychloroquine, grams0 (0–22.0)0 (0–292.0)0.004

The characteristics of the RA patients according to extent of coronary-artery calcification (none, mild, and moderate-to-severe) are shown in Table 4. In unadjusted comparisons, older age, male sex, higher systolic blood pressure, more total pack-years of smoking, higher homocysteine concentrations, and higher ESR were more common in RA patients with coronary-artery calcification than in those without. After adjusting for age and sex, only pack-years of smoking (OR 1.02, P = 0.04) and higher ESR (OR 1.02, P = 0.05) remained significantly associated with severe coronary-artery calcification in patients with RA. The ORs correspond to an increment of 1 pack-year of smoking or 1 mm/hour of ESR (Table 4).

Table 4. Characteristics of patients with rheumatoid arthritis according to different categories of Agatston calcium scores for coronary-artery calcification*
CharacteristicNo calcification (n = 68)Agatston score 1–109 (n = 32)Agatston score >109 (n = 41)Adjusted OR (95% CI)P
  • *

    Continuous variables are presented as the median (interquartile range). NA = not applicable.

  • Proportional odds logistic regression was used to obtain age- and sex-adjusted odds ratios (ORs) for the likelihood of having severe coronary-artery calcification (Agatston calcium score >109). 95% CI = 95% confidence interval.

Age, years48 (41–55)54 (51–64)63 (57–71)NANA
Male sex, %19.143.741.5NANA
Systolic blood pressure, mm Hg130 (115–139)131 (117–146)139 (127–158)1.00 (0.98–1.02)0.70
Diastolic blood pressure, mm Hg75 (68–80)77 (70–85)76 (69–84)1.00 (0.96–1.03)0.85
Family history of coronary disease, %25.038.724.41.16 (0.54–2.51)0.71
Total pack-years of smoking0 (0–12)14 (0–33)1 (0–30)1.02 (1.00–1.04)0.04
Body mass index, kg/m229 (24–34)27 (25–34)28 (25–30)0.96 (0.91–1.02)0.20
Creatinine, mg/dl0.8 (0.6–0.8)0.8 (0.7–1.0)0.8 (0.7–1.0)0.59 (0.12–2.87)0.51
Total cholesterol, mg/dl183 (160–202)188 (160–210)195 (164–219)1.00 (1.00–1.01)0.30
High-density lipoprotein, mg/dl44 (37–55)44 (37–54)41 (34–50)0.99 (0.96–1.01)0.32
Low-density lipoprotein, mg/dl107 (92–134)114 (95–132)117 (92–148)1.01 (1.00–1.02)0.14
Erythrocyte sedimentation rate, mm/hour15 (6–35)11 (4–30)25 (13–40)1.02 (1.00–1.04)0.05

DISCUSSION

To our knowledge this is the first controlled study of coronary-artery calcification in RA, and our results show that coronary-artery atherosclerosis, as detected by the presence of calcification on electron-beam CT of the coronary arteries, is more severe and more prevalent in patients with established RA than in those with early disease or control subjects. Patients with established RA had a median age of ∼58 years and were older than controls (median age 52 years) and patients with early RA (median age 51 years). Nevertheless, the increase in coronary-artery calcification remained significant after statistical adjustment for age and sex (OR 2.73, P = 0.005), and also remained significant after adjustment for race, smoking, hypertension, body mass index, levels of creatinine, cholesterol, triglycerides, and lp(a), diabetes, homocysteine concentrations, and family history of coronary disease (OR 3.42, P = 0.002). Thus, atherosclerosis is frequently present in patients with established RA and cannot be predicted by the presence of classic cardiovascular risk factors.

Previous studies have shown that the prevalence of subclinical atherosclerosis, detected as an increase in the intima-media thickness of the carotid artery, is higher in patients with RA than in controls (13, 14, 30). However, it has been recognized that the association between increased carotid-artery intima-media thickness and coronary-artery atherosclerosis is weak (31), and the prevalence of coronary atherosclerosis in patients with RA and its relationship to disease duration and activity have remained unclear.

Two major pathologic factors, the presence of atherosclerotic plaque and plaque rupture, are important in the pathogenesis of myocardial infarction. The role of RA in each of these processes has not been established. Our findings suggest that patients with established RA have increased atherosclerotic burden. However, calcification represents a relatively late stage of plaque development, and early coronary events may occur in the absence of calcification, especially in young patients. Recently, Maradit-Kremers et al (32) reported that patients with RA had an increased risk of myocardial infarction compared with controls without RA, and that this increased risk was present 2 years before fulfillment of the criteria for RA, suggesting that preclinical inflammation may contribute to increased risk of cardiovascular events. Thus, 2 different but related processes could be implicated. First, cardiovascular risk may be increased in patients with early RA, or even in undiagnosed RA, because inflammation leads to the formation of unstable plaque that is prone to rupture. Second, in late RA, cardiovascular risk could be increased because of increased atherosclerotic burden.

In the present study we determined the presence of coronary-artery calcification by electron-beam CT as a measure of calcified plaque in the coronary arteries. This technique provides a reproducible and quantitative method for the detection of subclinical coronary-artery atherosclerosis and yields information about the risk of future cardiovascular events in addition to that provided by other risk factors (20, 21). The finding of increased severity and higher prevalence of coronary-artery calcification indicates that coronary atherosclerosis and cardiovascular risk are increased in patients with established RA. Our data also show that the prevalence and severity of coronary-artery calcification was not increased significantly in patients with early RA. This suggests that structural coronary-artery disease, as detected by electron-beam CT, may occur later than structural joint damage, since, for example, one study demonstrated that 32% of patients with RA had bone erosions at the time of diagnosis and 70% had bone erosions after 3 years (33).

The cause of increased atherosclerosis in patients with RA is not known, and therefore comparison among patients with coronary-artery calcification of different severity levels may be informative. As occurs in the general population, traditional cardiovascular risk factors such as age, smoking, and male sex were more frequent among RA patients with coronary-artery calcification. In contrast, other traditional cardiovascular risk factors, such as LDL and HDL cholesterol levels, and other risk factors, such as homocysteine concentrations, were not associated with the severity of coronary-artery calcification in patients with RA. After statistical adjustment for age and sex, only pack-years of smoking and the ESR remained statistically significantly associated with the severity of calcification. This finding suggests that smoking and inflammation, 2 risk factors for atherosclerosis in the general population, may be particularly important risk factors in patients with RA.

Substantial evidence suggests that individuals who smoke are more likely to develop RA, and that patients with RA who smoke are more likely to have more severe disease (34). Thus, it is possible that smoking, a potent risk factor for atherosclerosis, interacts with RA to accelerate atherosclerosis to a greater degree in patients with this disease. Substantial evidence also indicates that inflammation can accelerate atherosclerosis (5, 35), and markers of inflammation such as the level of CRP are elevated in patients with coronary-artery disease. We did not measure CRP concentrations in control subjects, but the concentrations of CRP in patients with RA (mean ± SD 9.8 ± 13.1 mg/liter) were substantially higher than that found in the general population without cardiovascular disease (∼1.1 mg/liter) (11). Nevertheless, elevated concentrations of CRP in the patients with RA were not associated with coronary-artery calcification, whereas an increased ESR was associated with coronary-artery calcification. This difference may be informative, since the ESR and CRP, both acute-phase reactants, are generally correlated, as they were in this study (ρ = 0.56, P < 0.001). The fact that the ESR is particularly affected by concentrations of fibrinogen, an independent risk factor for heart disease (36), could theoretically explain the finding that the ESR, but not the CRP, was associated with coronary-artery calcification. However, it would be more informative to have serial measures of inflammation rather than a single cross-sectional measure.

Recently, we used electron-beam CT to establish that the prevalence of atherosclerosis was markedly increased in patients with SLE (23). The unadjusted OR for coronary-artery calcification in that population was ∼5, higher than the OR of ∼3 in patients with established RA. RA and SLE are both chronic inflammatory conditions, but lupus is characterized by more B cell activation and deposition of immune complexes. Both lupus and RA increase the risk of coronary-artery calcification, but understanding of why lupus appears to impart a greater risk may help to elucidate aspects of inflammation that are particularly important in the pathogenesis of atherosclerosis.

A potential interaction between age and RA that led to an increase in coronary calcium was another finding of this study. Stratified analyses by age showed that patients with established disease who were younger than age 50 years or between 50 and 59 years had increased coronary-artery calcification. Patients ages 60 years and older had a statistically nonsignificant increased risk of coronary-artery calcification. Since the latter finding could be due to limited power in a subgroup analysis, our results raise the possibility that the association between RA and coronary-artery calcification may be stronger in younger individuals.

This study has some limitations. First, with a cross-sectional design, there is a theoretical concern regarding the temporal relationship between exposure and outcomes. Longitudinal data would be ideal; however, the lack of significant association between coronary-artery calcification and early RA suggests that calcified atherosclerotic plaques are present later in the disease. In addition, another limitation of a cross-sectional design is the lack of information about risk factors, such as cholesterol and other lipid measurements over time. However, large cross-sectional studies of cardiovascular risk factors have used single fasting cholesterol measurements as a cardiovascular risk factor (37, 38). Second, generalizability of these results would require evidence from other populations. Third, electron-beam CT scans detect calcified atherosclerotic plaques as a measure of atherosclerotic burden, but do not specifically detect unstable plaques. Considering that 80% of acute ruptured plaques are calcified (39), some uncalcified, but clinically important, atherosclerotic plaque may not have been detected. Another limitation to this study is the potential for volunteer bias. This could skew the results in either a differential or a nondifferential way. However, our response rate was high, with 72.7% of the patients with RA and 85.1% of the control subjects contacted participating in the study.

In conclusion, our results indicate that coronary-artery atherosclerosis, detected as calcification of the coronary arteries on electron-beam CT, is more severe and more prevalent in patients with established RA than in patients with early disease and control subjects. Coronary atherosclerosis is associated with smoking and an increased ESR in patients with RA. This chronic inflammatory disease should be added to the list of conditions associated with an increased frequency of coronary-artery atherosclerosis. As our findings indicate, better control of inflammation and avoidance of smoking could decrease this risk in RA.

Acknowledgements

We are indebted to Carol Brannon and Elizabeth Simpson, who assisted with recruitment of subjects, and Daniel Byrne for criticism of the manuscript.

Ancillary