Rheumatoid Arthritis is Independently Associated with Increased Left Ventricular Mass but not Reduced Ejection Fraction

Abstract

Background

Rheumatoid arthritis (RA) is a chronic inflammatory disease associated with premature atherosclerosis, vascular stiffening, and heart failure. Whether RA is associated with underlying structural and functional abnormalities of the left ventricle (LV) is poorly understood.

Methods and Results

89 patients with RA without clinical cardiovascular disease and 89 healthy matched controls underwent echocardiography, carotid ultrasonography, and radial tonometry to measure arterial stiffness. RA patients and controls were similar in body size, hypertension and diabetes status, and cholesterol. LV diastolic diameter (4.92 vs. 4.64 cm, p <0.001), mass (136.9 vs. 121.7 g, p = 0.001 or 36.5 vs. 32.9 g/m ^2.7, p = 0.01), ejection fraction (EF) (71% vs. 67%, p <0.001), and prevalence of LV hypertrophy (LVH) (18% vs. 6.7%, p = 0.023) were all higher among RA patients. In multivariate analysis, presence of RA (p = 0.004) was an independent correlate of LV mass. Furthermore, RA was independently associated with the presence of LVH (OR 4.14, [95% CI 1.24-13.80; p=0.021]). Among RA patients, age at diagnosis and disease duration were independently related to LV mass. RA patients with LVH were older and had higher systolic pressure, damage index score, C-reactive protein, homocysteine and arterial stiffness index compared to those without LVH.

Conclusion

RA is associated with increased LV mass. Disease duration is independently related to increased LV mass, suggesting a pathophysiological link between chronic inflammation and LVH. In contrast, LV systolic function is preserved in RA patients indicating that systolic dysfunction is not an intrinsic feature of RA.

INTRODUCTION

Rheumatoid arthritis (RA) is the most common systemic autoimmune disease and affects approximately 1-3% of the developed world and 1.3 million adults in the United States.¹ Although RA is a disease of joints, extra-articular manifestations occur, and premature mortality among patients with RA is frequently due to cardiovascular disease,^2,3,4,5,6,7 primarily ischemic heart disease ^6,7,8 and congestive heart failure.^{9,10,11,12,13}

Abnormalities in left ventricular (LV) structure and function have also been reported in this population. However, whether these changes develop independent of concomitant cardiovascular disease (including hypertension, valvular disease, coronary artery disease and diabetes) is unknown. LV hypertrophy predicts cardiovascular events independent of traditional risk factors^14,15,16 and therefore, if present, may also be contributing to the early cardiovascular morbidity and mortality seen in patients with RA. In addition, conflicting data exist regarding the independent effects of RA on systolic function, a clinically relevant question given the excess of heart failure reported in this population. ^11-13,17,18

Similar to our earlier findings in systemic lupus erythematosus,¹⁹ we hypothesized that LV mass may be increased in RA, possibly due to inflammation-induced arterial stiffness. In the present case-control study, we sought to determine whether LV mass is increased in RA and, if so, whether this is related to traditional risk factors for LV hypertrophy or to disease-related aspects, including arterial stiffening.

METHODS

Study Population

Patients with RA who were enrolled in the Rheumatoid Arthritis Registry at the Hospital for Special Surgery in New York were consecutively recruited from regular outpatient visits with their rheumatologists. ²⁰ All patients met the American College of Rheumatology’s classification criteria for a diagnosis of RA.²¹ Exclusion criteria included recent (within the past 3 months) pregnancy, age younger than 18 years, and serum creatinine >3.0 mg/dl or creatinine clearance ≤30 ml/min. For the present analyses, of 99 patients recruited, 9 were excluded for significant valvular disease (moderate mitral regurgitation in 7 and moderate aortic regurgitation in 2). A standardized data collection instrument was administered to all patients to elicit a history of angina, myocardial infarction or coronary revascularization, with supporting documentation sought from the medical record. Clinical evidence of coronary artery disease (coronary artery bypass surgery) was present in only one patient who was also excluded from the present analysis. Segmental wall motion abnormalities to suggest the presence of previous myocardial infarction or severe regional ischemia were absent in all study participants. The remaining 89 patients were matched to control subjects on the basis of age (within 5 years), sex, ethnicity and hypertension status. Control subjects were normotensive and hypertensive individuals without clinical coronary artery disease or significant valvular disease who underwent similar imaging protocols in NIH-funded studies at the Hypertension Center of the New York Hospital. ^22,23

Disease activity and disease-related damage were quantified by the active joint count (number of tender or swollen joints out of 68 joints bilaterally), damage index score (joints irreversibly damaged),²⁴ history of joint replacement, number of RA criteria met, and the patient’s score on the Multidimensional Health Assessment Questionnaire (MDHAQ)²⁵. The age at diagnosis and disease duration were also noted. Medication history was obtained through interview and chart review; the use of methotrexate, corticosteroids, anti-tumor necrosis factor (TNF) agents, gold and hydroxychloroquine was recorded as never, former or current.

Traditional cardiovascular risk factors were assessed in all patients and controls, including fasting serum cholesterol, smoking history, presence or absence of hypertension (defined by a blood pressure of at least 140/90 mm Hg or the use of antihypertensive medications), presence or absence of diabetes mellitus and body mass index. Control subjects with hypertension were studied after antihypertensive medications had been systematically withheld for at least 3 weeks, whereas antihypertensive medications were not withheld in hypertensive RA patients. The institutional review board approved the study protocol, and all participants gave written informed consent.

Ultrasonographic Studies

Echocardiography was performed according to standardized procedures, previously reported in detail,²⁶ by a highly-skilled research sonographer. All studies were interpreted by a single cardiologist who was blinded to the clinical characteristics of the patients and controls. LV internal dimensions and wall thicknesses were measured at end diastole and end systole according to conventions established by the American Society of Echocardiography.²⁷ LV mass was calculated using an anatomically-validated method²⁸ according to the formula: LVM=0.8 {1.04[(LVID_d+PWT_d+SWT_d)³–(LVID_d)³]}+0.6 g, where LVID is LV internal dimension, PWT is posterior wall thickness, SWT is septal wall thickness, and subscript d represents end diastole. LV mass was normalized for both body surface area (g/m²) and height (g/m^2.7), where 2.7 is an exponent (allometric signal) linearizing the relation of LV mass to height in normal individuals.²⁹ LV hypertrophy was considered present if the mass index exceeded 45 g/m^2.7 or 96 g/m² in woman or 49 g/m^2.7 or 116 g/m² in men, and LV geometry was categorized according to standard guidelines.²⁷ LV systolic function was evaluated by ejection fraction and fractional shortening.

All study participants underwent carotid ultrasonography to assess carotid atherosclerosis. Studies were performed by the same research sonographer with the use of a standardized protocol and interpreted by a single, blinded cardiologist. In brief, both the right and left common, internal, and external carotid arteries were examined in multiple projections to identify the presence of atherosclerotic plaque, defined as focal protrusion > 50% beyond the thickness of the surrounding wall.³⁰ Lumen diameters were measured from end-diastolic (minimum dimension) and peak-systolic (maximum dimension) M-mode images of the distal common carotid artery using continuous tracing of the lumen-intimal interfaces of the near and far walls.²³

Assessment of Arterial Stiffness

Pressure-diameter relations of the common carotid artery and pressure waveforms obtained by applanation tonometry of the radial artery (using a high-fidelity external pressure transducer) were used to assess arterial stiffness. Central arterial waveforms and pressures were calculated with the SphygmoCor device using a generalized transfer function (Atcor Medical, Sydney, Australia) and calibrated using the brachial mean and diastolic pressures. Applanation tonometry has been previously validated to provide accurate estimates of intra-arterial pulse pressure when compared with simultaneous invasive pressure recordings.³¹ Measurement of carotid diameters with ultrasonography (as described above) was performed immediately before applanation tonometry without intervening change in subject position or environment. Arterial stiffness was estimated with the arterial stiffness index (ß): ln(P_s/P_d)/([D_s–D_d]/D_d), where P_s and P_d are aortic systolic and diastolic pressures, respectively, and D_s and D_d are carotid systolic and diastolic diameters, respectively.

Laboratory Assessment

At the time of the study visit fasting RA patients underwent phlebotomy to obtain a routine biochemical profile, complete blood count, LP(a) lipoprotein, lipid profile, homocysteine, rheumatoid factor, and anti-cyclic citrullinated peptide (CCP) antibody. A high-sensitivity assay to determine serum levels of C-reactive protein was used. Serum interleukin-6 and tumor necrosis factor p55 receptors were measured with the use of kits (BioSource International, Carlsbad, California). Serum levels of soluble intracellular adhesion molecule-1 and vascular cell adhesion molecule-1 were measured with an enzyme-linked immunosorbent assay (Caltag, Burlingame, California, and R&D Systems, Minneapolis, Minnesota, respectively).

Statistical Methods

Continuous variables are presented as mean values with standard deviations. Comparisons between RA patients and control subjects and between patients with and without LV hypertrophy were made using Student t-tests for independent samples for continuous variables and by Х² analyses for categorical variables. Univariate relationships with LV mass index (g/m^2.7) were assessed using the Pearson correlation coefficient. Variables found to have significant bivariate relations to LV mass index were assessed for independent association using stepwise, multivariable linear regression analysis; unstandardized regression coefficients (B) with their 95% confidence intervals are reported. Two-sided p<0.05 was considered statistically significant. Multivariable logistical regression analysis was used to determine the independent predictors of LV hypertrophy.

RESULTS

Characterization of the RA Patients

The 89 patients with rheumatoid arthritis ranged in age from 20 to 80 years; mean age at diagnosis was 35±14 years and mean duration of disease was 12±10 years. Rheumatoid factor was present in 56% and anti-CCP was present in 53% of patients. 26% of patients had undergone joint replacement. Prednisone, methotrexate, or anti-TNF agents were used currently or previously by 70%, 62% and 46% of patients, respectively.

Comparison of RA Patients and Control Subjects

By design, the RA patients and controls were matched for age, sex, ethnicity and hypertension status (Table 1) and did not differ in body size, diabetes status or cholesterol level. At the time of the study, participants in the control group were more likely to be current smokers and had higher blood pressure, likely due to cessation of anti-hypertensive medications in this group. Carotid atherosclerosis was more than three times more prevalent in RA patients, consistent with our earlier findings in the larger RA population, without exclusion of patients with conditions likely to affect LV mass.²⁰

Table 1.

Characteristics of the RA Patients and Control Subjects

Variable	Patients (n=89)	Controls (n=89)	P value
Age (years)	46.7±12.7	45.7±12.6	---
Female sex (%)	98.9	98.9	---
Hypertension (%)	15.7	15.7	---
Race (% white)	71.9	69.7	---
Systolic pressure (mm Hg)	108±17	116±22	0.009
Diastolic pressure (mm Hg)	69±9	72±11	0.034
Body surface area (m2)	1.7±0.17	1.7±0.18	0.590
Body mass index (kg/m2)	24.9±5.8	25.2±4.5	0.769
Diabetes (%)	1.1	1.1	1.000
Current smoking (%)	5.6	18.0	0.011
Cholesterol (mg/dl)	207±45	210±47	0.639
Carotid atherosclerosis (%)	39.3	12.4	< 0.001

LV structure and function in the two groups are compared in Table 2. Although posterior wall thicknesses were comparable, septal thickness tended to be greater and LV end-diastolic diameter was significantly higher in the RA group. As a consequence, LV mass, mass index and prevalence of LV hypertrophy were higher in this group. None of the RA patients had concentric LV remodeling and, in 15 of 16 patients with LV hypertrophy, the geometry was eccentric. Ejection fraction and cardiac output were significantly higher in the RA population whereas heart rates were similar. In addition, left atrial diameter was significantly larger in the RA group. Results were comparable when the analyses were limited to normotensive patients and controls (75 in each group; data not shown).

Table 2.

Comparison of Left Ventricular Structure and Function in Patients with Rheumatoid Arthritis and Control Subjects

Variable	Patients	Controls	P value
Septal Thickness (cm)	0.87±0.13	0.84±0.12	0.107
Posterior wall thickness (cm)	0.76±0.12	0.76±0.12	0.945
LV end-diastolic dimension (cm)	4.92±0.38	4.64±0.42	< 0.001
LV end-systolic dimension (cm)	2.93±0.32	2.91±0.36	0.688
Relative wall thickness	0.31±0.05	0.33±0.05	0.014
LV mass (g)	136.9±34.3	121.7±34.4	0.004
LV mass index (g/m2)	79.6±17.1	71.0±15.5	0.001
LV hypertrophy (%)	15.7	6.7	0.058
LV mass index (g/m2.7)	36.5±9.8	32.9±8.26	0.010
LV hypertrophy (%)	18.0	6.7	0.023
Fractional shortening (%)	41±4	37±5	< 0.001
Ejection fraction (%)	71±5	67±6	< 0.001
Cardiac output (L/min)	5.8±1.3	4.7±1.2	< 0.001
Cardiac index (L/min/m2)	3.4±0.64	2.8±0.63	< 0.001
Heart rate (bpm)	72±10	70±11	0.370
Left atrium (cm)	3.41±0.49	3.14±0.43	< 0.001
Aortic root (cm)	2.93±0.30	2.86±0.34	0.118

In multivariate analysis (Table 3), LV mass was strongly correlated with the presence of RA, independent of the effects of age, BMI and hypertension, whereas neither carotid plaque nor cholesterol level was associated with LV mass (both p>0.20). Furthermore, the presence of RA was independently associated with LV hypertrophy (OR 4.14, [95% CI 1.24-13.80; p=0.021]) with adjustment for the above covariates.

Table 3.

Multivariate Determinants of LV Mass Index (g/m^2.7) in the Entire Population

Variable	B (95% CI)	Beta Coefficient	P value
Age (per 10 years)	1.68 (0.69-2.68)	0.229	0.001
Body mass index (kg/m2)	0.81 (0.59-1.03)	0.464	<0.001
Hypertension	6.44 (3.33-9.56)	0.259	<0.001
Disease status	3.24 (1.05-5.42)	0.177	0.004

Relation of LV Mass and Hypertrophy to Traditional Risk Factors and Disease Aspects of RA

Among RA patients (Table 4), significant correlates of LV mass index included age, blood pressure and presence of hypertension, body size, cholesterol, homocysteine, and presence of carotid atherosclerosis. Carotid stiffness tended to relate to LV mass (p = 0.055). RA disease characteristics significantly related to LV mass index included age at diagnosis and duration of disease, and damage index score. Current or prior use of immunosuppressive therapy with steroids, methotrexate, anti-TNF agents, hydroxychloroquine or gold was not related to LV mass index. Serum markers of inflammation and disease activity were likewise unrelated to LV mass index. In multivariate analysis (Table 5), BMI and hypertension, along with age at diagnosis and disease duration, were significant, independent correlates of LV mass among RA patients.

Table 4.

Univariate Correlates of Left Ventricular Mass Index (g/m^2.7) in Rheumatoid Arthritis Patients

Variable	Correlation Coefficient	P value
Age (years)	0.439	<0.001
Systolic pressure (mmHg)	0.423	<0.001
Diastolic pressure (mmHg)	0.356	0.001
Hypertension	0.421	<0.001
Body surface area (m2)	0.300	0.004
Body mass index (kg/m2)	0.576	<0.001
Cholesterol (mg/dl)	0.327	0.002
LDL cholesterol (mg/dl)	0.352	0.001
HDL cholesterol (mg/dl)	−0.142	0.200
Homocysteine (μmol/L)	0.270	0.011
Carotid atherosclerosis	0.250	0.018
Cigarette smoking	0.077	0.471
Stiffness index (beta)	0.213	0.055
Age at diagnosis (years)	0.230	0.030
Disease duration (months)	0.243	0.022
RA criteria (number met)	−0.113	0.309
Damage index score	0.241	0.023
Joint count (#)	0.142	0.186
Joint replacement (#)	0.186	0.086
MD HAQ	0.118	0.277
Anti-CCP	0.017	0.877
Methotrexate use	0.019	0.865
Steroid use	−.078	0.467
Anti-TNF agents use	−0.009	0.934
Gold use	0.151	0.159
Hydroxychloroquine use	0.128	0.235
C-reactive protein (mg/dl)	0.102	0.358
ICAM-1 (ng/ml)	0.070	0.581
VCAM-1 (ng/ml)	0.109	0.388
Interleukin-6 (pg/ml)	−0.132	0.233

Abbreviations: LDL=low-density lipoprotein, HDL=high-density lipoprotein, CCP=cyclic citrullinated peptide, TNF=tumor necrosis factor, ICAM=intracellular adhesion molecule, VCAM=vascular cell adhesion molecule.

Table 5.

Multivariate Determinants of LV Mass Index (g/m^2.7) in RA Patients

Variable	B (95% CI)	Beta Coefficient	P value
Body mass index	0.574 (0.22-0.93)	0.340	0.002
Hypertension	7.455 (1.56-13.35)	0.258	0.014
Age at diagnosis	0.237 (0.02-0.45)	0.339	0.032
Disease duration	0.024 (0.003-0.044)	0.302	0.026
Damage index	0.072 (−0.185-0.329)	0.066	0.579
Homocysteine	0.098 (−0.885-1.082)	0.021	0.843
Arterial stiffness index	−0.818 (−2.31-0.675)	−0.127	0.278
Cholesterol	0.002 (−0.45-0.050)	0.011	0.921
Carotid atherosclerosis	−0.273 (−4.69-4.146)	−0.014	0.902

RA patients with LVH (as compared to those without LVH) were older (57.4±11.6 vs. 44.3±11.8 years, p<0.001), had higher systolic blood pressure (122±22 vs. 105±14 mmHg, p<0.001), BMI (31.1±8.0 vs. 23.6±4.14 kg/m², p<0.001), and stiffness index (4.06±1.68 vs. 3.02±1.38, p=0.045), and were more likely to be hypertensive (50 vs. 8%, p<0.001), have carotid atherosclerosis (62 vs. 34%. p=0.036), and actively smoke (19 vs. 3%, p=0.039). Patients with LVH tended to have higher damage index scores, higher CRP levels, older age at onset of disease, and higher scores on the Multidimensional Health Assessment Questionnaire. Because of the small number of RA patients with LVH, multivariate analyses were not performed.

DISCUSSION

The present study documents that RA is independently related to LV mass and the presence of LV hypertrophy. The design of our study eliminates the potential confounding influences of concomitant valvular heart disease or clinical coronary artery disease and controls for hypertension status. These findings are particularly striking given the higher average blood pressure in the control subjects. Among RA patients, age at diagnosis and disease duration were independently related to LV mass. These findings suggest a direct disease-related effect of RA on LV structure, linking chronic inflammation to the development of LV hypertrophy. Similar to our earlier findings in SLE,¹⁹ we had hypothesized that LV mass might be increased in RA due to inflammation-induced vascular stiffening. Although in this group of RA patients, vascular stiffness was not an independent determinant of LV mass, the sample size was relatively small and the correlation coefficient, significant in univariate analyses, was only slightly lower than that which attained statistical significance in the larger SLE population. Since vascular stiffness is increased in RA, even in the absence of atherosclerosis,³² and vascular stiffness contributes to LV hypertrophy independent of blood pressure,³³ it is likely that our hypothesis would be confirmed in a larger study. In addition, in the absence of clinical or echocardiographic evidence of coronary or significant valvular heart disease, systolic function was found to be preserved in RA patients suggesting that systolic dysfunction is not an intrinsic feature of RA or the cause of heart failure in RA independent of other cardiovascular diseases.

To date, only a few studies have evaluated LV structure and function in RA patients and have yielded varying results. Corrao et al. performed echocardiography in 40 otherwise healthy RA patients and 40 matched controls; LV mass index was significantly higher in RA patients as was the prevalence of diastolic abnormalities (defined as elevated A/E ratio).³⁴ In contrast, Di Franco et al. found no difference in either absolute or indexed (g/m²) LV mass between 32 otherwise healthy RA patients and matched controls. However they too reported abnormalities of left ventricular filling (characterized by a reduced E/A ratio) among the RA patients.³⁵ Gonzalez-Juanatey et al. compared 47 otherwise healthy RA patients who had been on treatment for at least five years to 47 matched controls. Again, LV mass was similar in the 2 groups; however 66% of the RA patients had diastolic dysfunction (based on transmitral flow) as compared to 43% of the controls.³⁶

Only two studies have reported systolic dysfunction in RA patients. Bhatia et al. performed echocardiography in 226 RA patients and compared findings with local population estimates. While systolic dysfunction (LV ejection fraction <50%) was more prevalent in RA patients (10.2% vs. 5.3%), they were older, more likely to have hypertension and/or diabetes, and to use tobacco.¹⁷ Since these factors may adversely impact LV systolic function and were not controlled in their statistical analyses, it is possible that RA per se was not the cause of reduced ejection fraction. Wislowska et al. also reported depressed LV ejection fraction in a group of 100 RA patients compared to 100 matched controls. While they excluded patients with hypertension, myocardial infarction, rheumatic fever, or a history of diabetes, their RA population had significantly more valvular heart disease than controls (39% vs. 19%), possibly confounding the results.¹⁸

Abnormal LV filling based on tissue Doppler imaging (TDI) has been reported. Birdane et al. compared 60 otherwise healthy RA patients with normal rest electrocardiograms and systolic function to 40 healthy, matched controls. LV mass index was similar in the 2 groups, however TDI was abnormal in RA patients.³⁷ Similarly Arslan et al. compared 52 otherwise healthy RA patients with 47 healthy controls. They too found diastolic dysfunction in the RA group characterized by reduced E/A and E’/A’ ratios and prolonged deceleration time.³⁸

In the current study, diastolic filling parameters (E and A velocities, isovolumic relaxation time, and deceleration time) were available in 84 of 89 RA patients; only 6% had abnormal diastolic filling (defined by an E/A ratio<1, deceleration time >240 msec and isovolumic relaxation time >100 msec). Thus since no study has clearly demonstrated an independent relation of RA to systolic dysfunction, the increased incidence of clinical evidence of heart failure reported in RA^11-13 may either reflect data collected before the regular use of disease modifying agents and TNF-α inhibition³⁹ or diastolic heart failure as the primary etiology of clinical heart failure. In this regard, the significantly larger left atrial diameter seen in RA patients, despite similar body size, suggests that left atrial pressure might be systematically higher in RA patients and thereby predispose to symptoms of congestive heart failure and/or atrial arrhythmias.

The association of duration of disease, damage index, age at diagnosis and higher levels of homocysteine with LV mass index suggests that the duration and severity of the inflammatory state play pivotal roles in the development of LV hypertrophy in RA. Interleukin (IL)-1 and tumor necrosis factor (TNF) are pro-inflammatory cytokines known to play an integral part in the pathogenesis of RA, stimulating fibroblast production, collagen synthesis, cell proliferation and death, endothelial damage and a marked immunological response (specifically inducing endothelial adhesion molecules) in both articular and non-articular tissues. ^40,41,42 In addition, elevated cardiac and circulating levels of these same cytokines have been shown in a variety of cardiovascular diseases, including myocarditis, myocardial infarction, congestive heart failure, and LV pressure overload; furthermore they have been implicated in the development of ventricular remodeling and myocyte hypertrophy.^43,44,45,46 Increased heart weight-body weight ratio and histological LV hypertrophy developed in transgenic mice over-expressing the human IL-1 gene (originally generated as a mouse model of RA), even when over-expression was restricted to cardiomycotes.^45,46 TNF-α knockout mice who underwent aortic banding showed attenuated cardiac apoptosis, hypertrophy, inflammatory response, reparative fibrosis, and reduced level of cardiac matrix metalloproteinase-9 activity compared with wild-type (WT) mice. Moreover, increased cardiac levels of TNF-α, which is not expressed in the normal heart, were detected in the WT mice after aortic banding was performed and were related to the above-listed deleterious cardiac changes.⁴⁴ These experimental findings suggest that pro-inflammatory cytokines, which are increased in RA patients, may contribute to the adverse LV remodeling documented by our study. Although the single measurement of these circulating inflammatory markers did not correlate with LV mass in our study, an ‘area under the curve’ estimate of chronic elevation might. Disease duration (an independent predictor of LV mass in multivariate analysis) may summate the severity of chronic inflammation and chronic levels of exposure to these pro-inflammatory cytokines.

Limitations of our study include the single measurement of inflammatory markers which, as discussed above, may not accurately represent cumulative inflammatory burden. In addition, although we relied on patient report and chart review for documentation of medication history, quantitative drug exposure could not be accurately calculated and may partly explain the failure of immunosuppressive medications to attenuate the observed structural changes. In fact, other studies have reported that use of disease-modifying or immunosuppressive medications in RA patients is associated with a lower rate of cardiovascular events⁴⁷ and decrease in arterial stiffening.⁴⁸ On the other hand, the lack of relation between immunosuppressive therapy and LV mass may also imply that blockade of inflammatory mediators may not reestablish the balance necessary to remodel and restore the physical properties of the heart and/or large conduit arteries.¹⁹ In addition we did not systematically evaluate diastolic function in either the RA patients or control subjects using load-independent techniques such as tissue Doppler imaging.

CONCLUSIONS

LV mass is increased in RA patients independent of known risk factors, including hypertension, obesity, age and concomitant cardiovascular disease. The independent association of LV mass with disease duration (a marker of long standing inflammation) suggests a pathophysiological link between chronic inflammation and LV hypertrophy. In contrast, LV systolic function was preserved in this population. Insofar as LV hypertrophy is associated with excess cardiovascular morbidity and mortality,^14-16 our findings may partially explain the premature cardiovascular disease seen in this population. Preliminary data indicating that aggressive immunosuppressive therapy, including statin use,⁴⁷-⁵⁰ may mitigate adverse cardiovascular effects indicates that large-scale, prospective intervention studies are warranted to determine whether cardiovascular complications of RA, including premature atherosclerosis, arterial stiffening and LV hypertrophy, can be minimized or prevented, in addition to effectively treating the traditional joint manifestations of RA. Furthermore, given the marked increase in LV mass even in the absence of hypertension, our study findings highlight the need to aggressively detect and effectively control hypertension in RA patients.