Rheumatology International

, Volume 32, Issue 11, pp 3455–3461

The major determinants of arterial stiffness in Korean patients with rheumatoid arthritis are age and systolic blood pressure, not disease-related factors

Authors

  • Young-Sam Kim
    • Hanyang University Hospital for Rheumatic Diseases
  • Yoon-Kyoung Sung
    • Hanyang University Hospital for Rheumatic Diseases
  • Chan-Bum Choi
    • Hanyang University Hospital for Rheumatic Diseases
  • Wan-Sik Uhm
    • Hanyang University Hospital for Rheumatic Diseases
  • Tae-Hwan Kim
    • Hanyang University Hospital for Rheumatic Diseases
  • Jin-Ho Shin
    • Division of Cardiology, Department of Internal MedicineHanyang University Hospital
    • Hanyang University Hospital for Rheumatic Diseases
Original Article

DOI: 10.1007/s00296-011-2198-y

Cite this article as:
Kim, Y., Sung, Y., Choi, C. et al. Rheumatol Int (2012) 32: 3455. doi:10.1007/s00296-011-2198-y

Abstract

Patients with rheumatoid arthritis are at increased risk of cardiovascular morbidity and mortality. This study was undertaken to investigate the prevalence of peripheral arterial disease, and to identify factors, especially those related to rheumatoid arthritis, influencing arterial stiffness in Korean patients with rheumatoid arthritis. A total of 262 patients with rheumatoid arthritis managed in a tertiary clinic were included. Ankle–brachial index and brachial–ankle pulse wave velocity were measured. Rheumatoid arthritis-related factors were determined, as well as the traditional cardiovascular risk factors. The prevalence of peripheral arterial disease was only 1.5%. Mean pulse wave velocity was 1,559 ± 354 cm/s. Age, body mass index, blood pressure, lipid profile, and glucose, not rheumatoid arthritis-related factors such as disease duration, seropositivity and disease activity, were significantly correlated with pulse wave velocity. Moreover, stepwise multiple regression analysis revealed that only age over 65 (OR = 9.1, 95% CI 4.3–19.1, P < 0.001), systolic blood pressure over 140 mmHg (OR = 15.7, 95% CI 7.4–33.1, P < 0.001), and corticosteroid use (OR = 2.1, 95% CI 1.03–4.3, P = 0.04) were independent risk factors for high pulse wave velocity. The prevalence of peripheral arterial disease in Korean patients with rheumatoid arthritis is very low. Among the many factors related to arterial stiffness, only old age, high systolic blood pressure, and, to a certain extent, corticosteroid use appear to be major determinants, especially in clinical setting with relatively well controlled patients with rheumatoid arthritis.

Keywords

Rheumatoid arthritisArterial stiffnessAnkle–brachial indexPeripheral arterial diseasePulse wave velocity

Introduction

Rheumatoid arthritis (RA) is a systemic immune and inflammatory disease associated with accelerated atherosclerosis and excess cardiovascular (CV) morbidity and mortality [1]. Some studies suggest that immune dysregulation and systemic inflammation play important roles in the accelerated atherosclerosis of RA [2].

Ankle–brachial index (ABI), the ratio of resting ankle to brachial systolic blood pressure, is widely used to screen for peripheral arterial disease (PAD) [3]. Although the prevalence of a low ABI in the Korean population is lower than in the West and in Japanese men, ABI has been associated with age, smoking status, waist circumference, hypertension, and blood pressure [4]. ABI in patients with RA tends to be either abnormally low or abnormally high, and the degree of abnormality is more severe in patients with more deformed joints [5].

In RA patients, increased arterial stiffness has been reported, which is higher as compared with healthy control and comparable to patients with known traditional CV risk [6] and may increase CV mortality [7]. Furthermore, RA is thought to be equivalent to coronary heart disease [8], and the comparative studies of accelerated atherosclerosis in rheumatic diseases, such as rheumatoid arthritis as well as systemic lupus erythematous, and familial Mediterranean fever have the rationale in terms of long-standing inflammatory pathogenesis [9].

Pulse wave velocity (PWV) is a noninvasive measure of arterial stiffness, and an independent marker of CV morbidity and mortality [10] for assessing atherosclerotic changes. The relationship between RA-related factors, as well as traditional CV risk factors, and the predictive value of vascular stiffness measured by PWV has not previously been established in Korean patients with RA.

The objectives of this study were therefore to examine the prevalence of peripheral arterial disease and to investigate RA-related risk factors other than the traditional CV risk factors, influencing PWV in Korean patients with RA.

Materials and methods

Subjects

A total of 262 patients, 212 of them women, with a mean age of 56.7 (range 27–85), attending outpatient clinics at Hanyang University Hospital for Rheumatic Disease were enrolled in this study. All the patients met the American College of Rheumatology 1987 revised criteria for RA [11]. The Institutional Review Board of Hanyang University Medical Center approved the protocol of this study, and written informed consent was obtained from all participants.

Anthropometric parameters (age, height, weight, and body mass index [BMI]) were measured during the physical examinations. Traditional CV risk factors studied were as follows: smoking (classified as never, former, or current), lipid profile (total cholesterol, HDL cholesterol, LDL cholesterol, and triglyceride), CV disease (CVD) history (coronary heart disease, cerebrovascular disease), diabetes mellitus [DM], hypertension, and history of medication for hypertension and DM. RA-related factors consisted of disease duration, disease activity parameters (swollen joint count of DAS28 [number 0–28], evaluator’s global assessment of activity [visual analog scale: 0–10], and Steinbrocker’s functional capacity [grade 1–4], ESR, and CRP), seropositivity for rheumatoid factor (RF) and anti-cyclic citrullinated protein/peptide antibody (ACPA), current disease-modifying antirheumatic drugs (DMARD) or corticosteroid usage for 6 months before the study, and extra-articular complications (interstitial lung disease [ILD], and rheumatoid nodule).

Systolic and diastolic blood pressures were obtained manually on the right arm with the subjects in supine position after a 5-min rest. Hypertension was defined as a systolic blood pressure of over 140 mmHg and/or a diastolic blood pressure of over 90 mmHg and/or current use of antihypertensive drugs. For each patient, fasting blood samples were collected for the assessment of serum glucose, creatinine, lipid level, CRP, ESR, RF, and ACPA.

Measurement of ABI and PWV

After the subject had rested in a supine position for more than 5 min, brachial–ankle PWV (baPWV) was measured using an automatic waveform analyzer (VP-2000, Colin Co., Ltd., Komaki, Japan) that simultaneously records blood pressure, ABI, baPWV, the electrocardiogram, and heart sounds. Blood pressure was measured simultaneously in both arms and legs, followed by automatic calculation of ABI on left and right sides. Pulse waves were also recorded automatically by sensors in the cuffs; the transmission times and distances between the cuffs of arms and legs were recorded, and the baPWV outputted. The higher of the baPWVs in the left and right arm was used for the analysis of risk factors, although there was a strong positive correlation between left and right baPWV (r = 0.963, P < 0.05). Intra-observer intra-class correlation coefficient (ICC) of PWV measurement was assessed by repeated measurement in 18 healthy persons without arthritis (ICC of left arm = 0.902, 95% CI 0.759–0.962, ICC of right arm = 0.782, 95% CI 0.507–0.912).

Statistical analysis

Statistical analyses were performed using SPSS software (ver. 18, SPSS Inc., Chicago, IL, USA) and SigmaPlot ver. 11.2 (Systat Software, Inc., San Jose, CA, USA). baPWV values for comparing subgroups were expressed as means ± standard deviations (M ± SD), and significance was determined either by an independent t-test or one-way analysis of variance. To test for correlations between continuous variables, we used Spearman’s rank correlation coefficient. Thereafter, stepwise forward multiple logistic regression analysis was performed to identify the independent variables for baPWV. P values less than 0.05 were considered statistically significant.

Results

Clinical characteristics

Anthropometric measurements (age, height, weight, and BMI), clinical characteristics, traditional cardiovascular risk factors, and RA-related factors of the RA patients (212 women and 59 men, age: 56.7 ± 11.2 years) are shown in Table 1. According to swollen joint count, functional capacity, ESR, and CRP, disease activity of the RA patients studied was mild to moderate. Average dose in RA patients taking corticosteroid (n = 149) was 3.9 mg/day of prednisolone or its equivalent. Differences of baPWV according to subgroups of these factors are also presented.
Table 1

Socio-demographic and clinical characteristics of participants, and differences in pulse wave velocity according to subgroups of variables

Variable

M ± SD

n

 

n

%

baPWV (m/s)

P

M

SD

Age (years)

56.7 ± 11.2

262

≥65

75

28.6

1,859.7

358.3

<0.001

   

<65

187

71.4

1,438.4

271.5

 

Sex (M, F)

 

262

M

50

19.1

1,706.3

365.8

0.001

   

F

212

80.9

1,524.2

342.9

 

BMI (kg/m2)

23.2 ± 3.3

261

≥25

63

24.1

1,659.1

395.7

0.009

   

<25

198

75.9

1,525.9

334.9

 

SBP (mmHg)

124 ± 19

262

≥140

78

29.8

1,911.6

349.9

<0.001

   

<140

184

70.2

1,409.5

227.5

 

DBP (mmHg)

75.4 ± 10.3

262

≥90

53

20.2

1,879.7

377.0

<0.001

   

<90

209

79.8

1,477.6

298.1

 

CHOL (mg/dl)

194 ± 36

262

≥200

107

40.8

1,641.3

335.0

0.002

   

<200

155

59.2

1,502.1

356.7

 

LDL (mg/dl)

109 ± 30

257

≥130

63

24.5

1,662.4

364.2

0.005

   

<130

194

74.5

1,522.2

334.3

 

Medical history

 Hypertension

 

262

(−)

176

67.2

1,479.6

328.6

<0.001

   

(+)

86

32.8

1,721.4

350.5

 

 Diabetes mellitus

 

262

(−)

244

93.1

1,549.9

351.3

0.127

   

(+)

18

6.9

1,682.0

377.6

 

 Coronary artery disease

 

262

(−)

254

96.9

1,558.4

354.4

0.891

   

(+)

8

3.1

1,575.9

365.1

 

 Cerebrovascular disease

 

262

(−)

257

98.1

1,558.8

353.6

0.951

   

(+)

5

1.9

1,568.6

419.5

 

 Smoking

 

261

Non-

217

83.1

1,540.0

352.7

0.123

   

Ex-

22

8.4

1,681.3

377.4

 
   

Current

22

8.4

1,632.4

330.4

 

Characteristics associated with rheumatoid arthritis

 Disease duration (month)

78.5 ± 11.7

262

≥60

189

72.1

1,581.6

353.0

0.097

   

<60

73

27.9

1,500.5

352.4

 

 ESR (mm/h)

30.6 ± 24.0

262

≥20

154

58.8

1,544.3

360.3

0.425

   

<20

108

41.2

1,579.9

345.5

 

 CRP (mg/dl)

0.94 ± 1.62

262

≥0.3

131

50.0

1,576.7

378.0

0.419

   

<0.3

131

50.0

1,541.2

328.7

 

 SJC (0–28)

2.2 ± 2.8

262

≥1

165

63.0

1,538.0

358.7

0.211

   

<1

97

27.0

1,594.7

344.9

 

 FC (1–4)

 

262

1

160

61.1

1,543.6

339.5

0.659

   

2

97

37

1,581.0

372.9

 
   

≥3

5

1.9

1,622.8

483.7

 

 Interstitial lung disease

 

261

(−)

250

95.8

1,545.8

348.3

0.016

   

(+)

11

4.2

1,807.0

384.0

 

 Rheumatoid factor

 

262

(−)

44

16.8

1,640.3

424.8

0.095

   

(+)

218

83.2

1,542.6

336.7

 

 ACPA

 

234

(−)

36

15.4

1,601.0

421.6

0.278

   

(+)

198

84.6

1,533.6

326.3

 

Present medication

 Methotrexate

 

262

(−)

42

16

1,677.8

398.1

0.017

   

(+)

220

84

1,536.3

341.3

 

 Corticosteroid

 

262

(−)

113

43.1

1,501.0

307.8

0.017

   

(+)

149

56.9

1,603.0

380.5

 

baPWV brachial–ankle pulse wave velocity, n: number, M mean, SD standard deviation, BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, CHOL total cholesterol, LDL low density lipoprotein cholesterol, − absence, + presence, smoking, Non- non-smoker, Ex- ex-smoker, SJC swollen joint count, FC Steinbrocker’s functional capacity, ACPA anti-cyclic citrullinated protein/peptide antibody

Prevalence of PAD in RA

The mean ABI values of left and right sides were 1.10 ± 0.08 and 1.12 ± 0.08, respectively. The prevalence of abnormal ABI of <0.9, indicating mild-to-moderate PAD, was 1.5% (n = 4, all women, left side: 2, right side: 2). However, their ABI values were not far off from the threshold (0.82, 0.87, 0.9, and 0.9, respectively). Moreover, none of the patients complained of symptoms of peripheral arterial disease. A high ABI indicating incompressibility was observed in only three patients.

Difference in PWV according to characteristics

As shown in Table 1, baPWV was significantly higher in older patients and in men, as well as those with high BMI, hypertension and its history, and hyperlipidemia. Past histories of DM, coronary artery disease, cerebrovascular disease, and smoking were not associated with high baPWV. RA-related factors such as disease duration, seropositivity (RF, ACPA), and disease activity indices (swollen joint count, evaluator’s global assessment, functional capacity, ESR, and CRP) were also not associated with high baPWV. However, the RA-related complication, ILD as seen on chest radiographs, and medication such as MTX and corticosteroid were associated with high baPWV.

Associating factors for high PWV

To evaluate the association between baPWV and continuous variables, we used Spearman’s correlation analysis (Table 2). baPWV was significantly correlated with systolic blood pressure (r = 0.739, P < 0.001), age (r = 0.661, P < 0.001), diastolic blood pressure (r = 0.595, P < 0.001), triglyceride (r = 0.322, P < 0.001), total cholesterol (r = 0.314, P < 0.001), LDL (r = 0.294, P < 0.001), and hemoglobin (r = 0.234, P < 0.001). However, RA-related variables such as swollen joint count, evaluator’s global assessment, CRP, and ESR did not show significant correlations.
Table 2

Correlation between patients’ characteristics and pulse wave velocity (n = 262)

 

Spearman’s rho

P value

Age (years)

0.661

<0.001

Body mass index (kg/m2)

0.184

0.003

Systolic blood pressure (mmHg)

0.739

<0.001

Diastolic blood pressure (mmHg)

0.595

<0.001

Swollen joint count (0–28)

−0.110

0.075

Evaluator’s global assessment (0–10)

0.099

0.111

CRP (mg/dl)

0.062

0.314

ESR (mm/h)

−0.017

0.788

Total cholesterol (mg/dl)

0.314

<0.001

HDL (mg/dl)

0.016

0.804

LDL (mg/dl)

0.294

<0.001

Triglyceride (mg/dl)

0.322

<0.001

Glucose (mg/dl)

0.195

0.002

Hemoglobin (g/dl)

0.234

<0.001

We divided the RA patients into two groups (baPWV ≤ 1,635 cm/s, n = 89 vs. baPWV > 1,635 cm/s, n = 173) for the socio-demographic and clinical parameters shown in Table 1. This value was originally suggested as the cutoff value for multiple coronary artery occlusive disease in Korean type 2 DM patients [12]. The univariate analysis revealed significant differences between the lower baPWV group and the high baPWV group in regard to age over 65 years (P < 0.001), total cholesterol over 200 mg/dl (P = 0.011), systolic and diastolic blood pressure (P < 0.001), hypertension history (P < 0.001), RF positivity (P = 0.037), and the use of glucocorticoid in any dose (P = 0.053; Table 3).
Table 3

Bivariable and multivariable logistic regression analyses of the impacts of various factors on high pulse wave velocity in Korean patients with rheumatoid arthritis

Variables

Unadjusted

Adjusted

OR (95% CI)

P

OR (95% CI)

P

Age (≥65 years)

10.1 (5.5–18.7)

<0.001

9.1 (4.3–19.1)*

<0.001

Sex (male)

1.7 (0.9–3.1)

0.098

  

BMI (≥25)

1.8 (1.01–3.2)

0.048

  

Disease duration (≥5 years)

1.4 (0.8–2.4)

0.245

  

Total cholesterol (≥200 mg/dl)

2.0 (1.2–3.3)

0.011

  

LDL (≥130 mg/dl)

1.3 (0.2–8.2)

0.795

  

Systolic BP (≥140 mmHg)

15.9 (8.3–30.5)

<0.001

15.7 (7.4–33.1)*

<0.001

Diastolic BP (≥90 mmHg)

6.2 (3.2–12.0)

<0.001

  

HTN

3.8 (2.2–6.7)

<0.001

  

DM

2.1 (0.8–5.4)

0.144

  

Smoking (ever)

1.4 (0.6–2.5)

0.487

  

RF positive

0.5 (0.3–0.96)

0.037

  

ACAP positive

0.8 (0.4–1.7)

0.530

  

Glucocorticoid use

 Any dose

1.7 (0.99–2.9)

0.053

2.1 (1.03–4.3)*

0.04

 PDS ≥ 5 mg

1.4 (0.8–2.5)

0.238

  
   

R2

0.564

OR odds ratio, CI confidence interval, BMI body mass index, BP blood pressure, HTN hypertension, DM diabetes mellitus, RF rheumatoid factor, ACPA anti-cyclic citrullinated protein/peptide antibody, PDS prednisolone equivalent dose

* Pulse wave velocity as a dependent variable was categorized into two groups by 1,635 cm/s, statistical significance with P < 0.05

Variables were entered into a forward stepwise logistic regression analysis; R2, Nagelkerke’s contribution

Evaluation by forward logistic regression analysis of the influence on high baPWV of age, sex, BMI, smoking history, duration of RA, history of medication with MTX and corticosteroid, and cholesterol level demonstrated that the model was significant (<0.01), and the Nagelkerke’s R2 value was 0.563. It revealed that only age over 65 (OR = 9.1, 95% CI 4.3–19.1, P < 0.001), high systolic blood pressure over 140 mmHg (OR = 15.7, 95% CI 7.4–33.1, P < 0.001), and, to a lesser extent, corticosteroid use (OR = 2.1, 95% CI 1.03–4.3, P = 0.04) independently influence baPWV in Korean RA patients (Table 3).

Discussion

RA as a model of systemic inflammatory disorder driving atherosclerosis has been recently reviewed [13]. It was concluded that inflammatory processes contribute to the increased CV morbidity and mortality in RA via inflammatory and/or non-inflammatory mechanisms including increased arterial stiffness and the CV toxicity of antirheumatic drugs. In our study, we investigated the prevalence of PAD. We also sought to identify RA-related risk factors for CV other than the traditional risk factors in Korean patients with RA managed at a tertiary academic clinic for rheumatic diseases.

The prevalence (1.5%) of low ABI (<0.9) indicating PAD in Korean patients with RA was similar to that in a population of middle-aged and elderly Koreans [4]. This prevalence was lower than that reported previously in Korean patients with type 2 DM [14], as well as in Western patients with RA [5]. This is compatible with an epidemiologic study comparing RA versus DM as a risk factor for CVD, which showed that the prevalence of PAD in RA was lower than that in type 2 DM, although the prevalence of CVD, including coronary artery disease and cerebrovascular disease, in RA was comparable to that in type 2 DM [15].

baPWV in our RA patients was 1,525 ± 345 on the right side and 1,528 ± 341 cm/s on the left; these values are higher than those obtained in Korean adults attending a health promotion center and cardiology outpatient clinics measured with same device [16].

In a univariate model, baPWV in RA patients did not show any significant correlation with RA-related factors, such as RA duration, seropositivity, and disease activity variables, except for ILD complication and medication with MTX and corticosteroid. This outcome is in part consistent with previous reports [1719]. However, in contrast to some reports [17, 19], baPWV was higher in the corticosteroid group of RA patients and lower in the MTX group in our study.

In a correlation model, disease activity variables, such as swollen joint count, evaluator’s global assessment, CRP, and ESR, did not show any correlation with baPWV in RA patients. This result differs from a previous report that PWV was correlated with log CRP and improved with antitumor necrosis factor therapy consisting of etanercept and infliximab [20, 21]. In fact, CRP is now generally considered a kind of biomarker of CVD without any causative role [22].

In a multivariate model, baPWV in RA patients showed a significant and positive correlation only with old age and high systolic blood pressure, as previously reported in RA [18] and in other diseases such as systemic lupus erythematosus [23]. However, RA disease-related factors, such as disease duration, disease activity, and seropositivity, were not significantly correlated with baPWV. This is consistent in part with previous results [1719]. Corticosteroid use was also correlated with baPWV, although less significant than age or systolic blood pressure. However, it seemed to be a very meaningful finding as exposure to low-dose corticosteroid in RA showed a trend of increasing major CV events [24].

The limitations of our study were as follows: first, the study group was selected from patients who were supposed to be tested for blood on a regular basis, so there might be a selection bias. Second, the disease activity of the RA patients in this study was relatively mild, as shown in Table 1, as the patients were strictly managed in the tertiary clinic setting. Therefore, potential effects of RA-related factors on baPWV might have been masked. Thirdly, we were not able to evaluate the cumulative inflammatory burden due to the cross-sectional setting. Fourthly, ILD was evaluated only with chest radiographs or patients’ statement on the results of their chest X-rays at health promotion center.

In conclusion, the prevalence of PAD was very low in Korean patients with RA. For RA patients managed in a tertiary clinic setting, RA-related risk factors except corticosteroid use are not associated with increased CV risk. However, old age and high systolic blood pressure as traditional risk factors for CVD are undoubtedly important in RA patients in terms of the prevention of CVD. Assessing arterial stiffness by measuring PWV, and strategies for preventing CVD in RA, should improve the management of patients with RA, although we failed to demonstrate that RA-related factors were independent risk factors for CVD in RA.

Acknowledgments

We would like to thank Kim HK, Research Nurse, for collection of clinical data. This work was supported by a grant from Korea Institute of Medicine.

Conflict of interest

There are no conflicts of interest related to the manuscript.

Copyright information

© Springer-Verlag 2011