Rheumatology 2012;51:2215–2223

RHEUMATOLOGY doi:10.1093/rheumatology/kes213
Advance Access publication 25 August 2012

Original article
Increase in ventricular–arterial stiffness in patients
with psoriatic arthritis
Qing Shang1, Lai-Shan Tam2, John E. Sanderson1, Jing-Ping Sun1, Edmund
Kwok-Ming Li2 and Cheuk-Man Yu1

Abstract
Objectives. Ventricular and arterial stiffness is an accepted cause of myocardial diastolic dysfunction.
The aim of this study is to determine whether there is increased ventricular and arterial stiffness in patients

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

with PsA and any relationship with disease-related risk factors.
Methods. Seventy-three patients with PsA were divided into two subgroups based on the absence or
presence of hypertension and/or left ventricular (LV) hypertrophy. Fifty healthy controls were enrolled for
comparison. All participants underwent non-invasive assessments including conventional echocardiog-
raphy with tissue Doppler imaging and pulse wave analysis. Ventricular stiffness was measured by ven-
tricular end-systolic and diastolic elastance, whereas arterial stiffness was measured by total arterial
compliance and aortic augmentation index.
Results. There was significantly increased ventricular and arterial stiffness in patients with PsA (P < 0.001),
even in those without hypertension and/or LV hypertrophy. Based on the cut-off points derived from the
controls, 38.4% of PsA patients had increased LV stiffness including 31.5% in diastole and 17.8% in
systole, and 15.1% had increased arterial stiffness. Multivariable logistic regression analysis showed that
long PsA disease duration (>10 years) (odds ratio = 6.55, P = 0.001) was an independent risk factor for
increased LV diastolic elastance after adjusting for age, gender and hypertension.
Conclusion. Patients with PsA may have increased ventricular and arterial stiffness even without evidence
of LV remodelling, and those with long disease duration may be at a higher risk. Therefore, prolonged

CLINICAL
SCIENCE
inflammatory burden may be an important cause of early cardiovascular disease in patients with PsA.
Key words: stiffness, ventricular remodelling, PsA.

Introduction increased death [2]. However, the mechanism of CVD has

not been fully elucidated in these patients. Recently, ven-
PsA is an autoimmune inflammatory disease, which is tricular and arterial stiffness has been identified as an im-
associated with an increased prevalence of hypertension, portant pathophysiological determinant of CVD [3] and
ischaemic heart disease, congestive heart failure, periph- has been studied in different populations including the
eral vascular disease, cerebrovascular disease and type II elderly, patients with cardiovascular risk factors, hyper-
diabetes [1]. A mortality study showed that cardiovascular tension, diabetes mellitus, end-stage renal disease and
diseases (CVD) are the leading cause of death in patients obstructive sleep apnoea [4–9]. Nevertheless, no pub-
with PsA compared with controls with a 1.3-fold risk of lished study has focused on the early detection of ven-
tricular and arterial stiffness in patients with PsA.
1
Division of Cardiology and 2Divsion of Rheumatology, Department of With tissue Doppler imaging subclinical myocardial dis-
Medicine and Therapeutics, Institute of Vascular Medicine, The
Chinese University of Hong Kong, Prince of Wales Hospital, Hong
ease, not detectable by conventional echocardiography,
Kong, China. can be recognized [10–13]. In addition, ventricular and
Submitted 13 December 2012; revised version accepted arterial stiffness can be derived from echocardiographic
29 June 2012. measurements [14]. Furthermore, non-invasive pulse
Correspondence to: Cheuk-Man Yu, Division of Cardiology, wave analysis (PWA) has recently gained recognition as
Department of Medicine and Therapeutics, The Chinese University of
Hong Kong, 9/F Clinical Sciences Building, Prince of Wales Hospital, a measure of subclinical vascular disease, using applana-
Shatin, NT, Hong Kong, China. E-mail: cmyu@cuhk.edu.hk tion tonometry, which records a peripheral arterial

! The Author 2012. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
Qing Shang et al.

waveform. A derived central aortic waveform can be Sign and symptoms of psoriasis
obtained by applying a generalized transfer function. The Skin abnormality and nail lesions were examined, and the
aortic augmentation index provided by PWA can indirectly psoriasis area and severity index (PASI) was calculated to
evaluate arterial stiffness. Using these non-invasive tech- evaluate the severity of psoriasis [17]. Radiographs (spine,
niques, we sought to test the hypothesis that patients with pelvis, feet and hands) were reviewed for the presence of
PsA might have a high propensity to experience increased erosion at the time of the study.
ventricular and arterial stiffness independent from stand-
ard risk factors and which may be associated with some Laboratory
disease-related risk factors. ESR was measured using the Westergren method. The
high-sensitivity CRP level was measured using an immu-
Methods noturbidimetric assay performed with Olympus OSR6185
(Olympus Diagnostics, Lismeehan, County Clare, Ireland).
Study population Apolipoproteins A and B were tested by automated ana-
Ninety-five patients were screened, and 73 of them fulfill- lyzer (Cobas-Mira Plus, Hoffman-La Roche Diagnostics,
ing the classification of PsA criteria [15] were recruited Mannheim, Germany), using a turbidimetric assay.
consecutively from the rheumatology clinic of a university Plasminogen activator inhibitor-1 was measured by
ELISA, using commercially available kits (Diagnostica

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

affiliated teaching hospital. Exclusion criteria included
pregnancy, hypothyroidism, clinically significant renal dis- Stago, Freres Chausson, France). Fibrinogen was mea-
ease (serum creatinine level 5 270 mmol/l), diabetes melli- sured using a modified clot-rate assay.
tus, those with history of angina, stable coronary artery
disease, previous acute coronary syndromes, coronary Echocardiography
revascularization, bundle-branch block, second-degree Conventional echocardiography
or higher atrioventricular block, atrial fibrillation, valvular
A comprehensive echocardiography with Doppler studies
stenosis or at least moderate valvular regurgitation, valvu-
was performed by using Vivid 7 systems with a 3.5-MHz
lar replacement or repair or mitral annular calcification.
probe (GE Medical Systems, Milwaukee, WI, USA).
Patients were divided into two subgroups based on the
Standard echocardiographic assessments included the
absence or presence of hypertension (defined as a sys-
measurements of LV dimensions, wall thickness, LV
tolic blood pressure 5140 mm Hg or a diastolic blood
mass index, ejection fraction and colour Doppler imaging
pressure 590 mm Hg or the use of antihypertensive
of all valves were done according to the recommenda-
agents) and/or left ventricular hypertrophy (LVH) evi-
tions of American Society of Echocardiography [18].
denced by echocardiography. Fifty healthy control sub-
LVH was considered present if the LV mass index to
jects without a history of overt CVD were recruited from
body surface exceeded 95 g/m2 in women and 115 g/m2
the community through advertisement or from other
in men. Relative wall thickness (RWT) was calculated
clinics. The control group was matched to the patient sub-
by the formula (2  LV end-diastolic posterior wall thick-
group without hypertension or LVH for age, gender, BMI
ness/LV end-diastolic internal dimension) to classify
and blood pressure. The Ethics Committee of our institu-
LV remodelling as normal geometry (no LVH
tion (Ethics Committee of The Chinese University of Hong
and RWT 40.42), concentric remodelling (no LVH
Kong and the Hong Kong Hospital Authority) approved the
but RWT > 0.42), eccentric hypertrophy (LVH with
study that was conducted in compliance with the
RWT 40.42) or concentric hypertrophy (LVH with
Declaration of Helsinki (2000) of the World Medical
RWT >0.42) [18]. The LV ejection fraction was evaluated
Association. All participants provided written informed
by modified biplane Simpson method.
consent.
Tissue Doppler imaging and subclinical LV dysfunction
PsA clinical assessment
Color tissue Doppler images in at least three consecutive
Disease patterns beats were acquired from apical 4-chamber, 2-chamber
PsA patients with predominant peripheral arthritis were and long-axis with a highest possible frame rate (>100
included in the category of peripheral arthritis, and those frames/s), and off-line analysis programs (EchoPac-PC
with predominant inflammatory arthritis of the back were 108.1.5, Horten, Norway) included peak systolic (Sm)
included in the category of spondyloarthritis. and peak early diastolic (Em) myocardial velocities of 12
LV segments. Real-time pulsed-wave tissue Doppler velo-
Disease activity and severity cities were recorded from the septal and lateral sites of
Pain, physician’s and patient’s global assessments, was the mitral annulus in the apical 4-chamber view for meas-
assessed using a 10-point visual analogue scale (0 = ex- urements of early peak (E0 ) diastolic velocity. In our labora-
cellent well-being; 10 = feeling extremely unwell). Physical tory, the intraobserver and interobserver variability for
examination recorded the number of tender and swollen tissue Doppler velocity data were 3% and 5%, respect-
joints using the 68 tender/66 swollen joint count, the pres- ively, as previously reported [19]. Subclinical LV diastolic
ence of dactylitis and the number of permanently dysfunction was identified by lateral E0 <11.5 cm/s and/or
deformed joints. Disease activity was assessed using the ratio of the mitral inflow early diastolic filling velocity (E)
the disease activity score in 28 joints (DAS28) [16]. to lateral E0 (E/E0 ) >10 [20–22]; subclinical LV systolic

2216 www.rheumatology.oxfordjournals.org
 Ventricular–arterial stiffness in PsA

dysfunction was defined by mean Sm from only six basal between echocardiographic and clinical variables. Curve
LV segments <4.4 cm/s [10]. Subclinical LV dysfunction estimation is a curve-fitting program that can be used to
included subclinical LV diastolic and/or systolic compare and select suitable curve relationships (such as
dysfunction. linear, logarithmic, inverse, compound and so on) to illus-
trate the relationships between variables, which are illu-
Non-invasive vascular assessment strated in figures with r values. As for logistic regression,
all potential risk factors were included into univariable lo-
Vascular ultrasound
gistic regression analysis based on t-test results, then
Carotid intima–media thickness (IMT) was scanned by
those with P < 0.10 in univariable logistic regression ana-
using B-mode ultrasound with a 10-MHz linear vascular
lysis were selected for the multivariable logistic regression
probe (Vivid 7; GE Vingmed Ultrasound, Horten, Norway).
analysis. The cut-off points of LV and arterial stiffness
Minimal gain was adjusted to visualize the lumen–intimal
were derived from control group by using mean +2 S.D.,
and medial–adventitial interfaces defining IMT in the far
because 95% of all the data values in control group can
wall. Digital images of three cardiac cycles were saved
be found between the mean +2 S.D. and the mean 2 S.D.
with ECG signals. The IMT was measured in the distal
All tests were 2-tailed and P-values <0.05 were con-
1 cm of common carotid artery, carotid bifurcation and
sidered statistically significant.
the initial 1 cm of internal carotid artery by using dedicated

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

software (Carotid Analyzer; Medical Imaging Applications,
Coralville, IA, USA). The cardiologist was blinded to all Results
clinical information when carrying out online and offline
The mean age of these 73 PsA patients at diagnosis was
analysis. The mean IMT was recorded, not including pla-
34 years, and the median disease duration of PsA was
ques, for the following calculation. Plaque was defined as
10.7 (IQR: 6.3–18.0) years. Fifty-seven (78.1%) patients
a localized thickening >1.2 mm that did not uniformly in-
had concurrent psoriasis, 47 (64.4%) had nail lesions
volve the whole artery.
and 34 (46.6%) had erosions on radiography at the time
PWA of assessment. Sixty-one (83.6%) patients had peripheral
The examination was performed as recommended [23]. arthritis and 12 (16.4%) had spondyloarthritis. Based on
PWA was performed using the SphygmoCor device the DAS28 score, 45 (61.6%) patients had active disease.
(SCOR 2000 v.7.01; AtCor. Medical Pty Ltd, Sydney, More than half of the patients had been on NSAIDs and
Australia) with a tonometer probe at the right radial DMARDs (Table 1). Forty-three patients did not have
artery. The central aortic arterial pulse wave was trans- hypertension or LVH, whereas 30 patients had either
ferred from the peripheral arterial pulse wave, and the hypertension (n = 27, 37.0%) or LVH (n = 7, 9.6%).
aortic augmentation index was standardized to a heart
rate of 75 bpm to minimize the effect of heart rate auto- Stiffness in patients with PsA
matically [24]. The standardized aortic augmentation Compared with controls, patients with PsA demonstrated
index was used for the main analysis. an increase in ventricular and arterial stiffness and an early
impairment of LV longitudinal systolic and diastolic func-
Definitions of ventricular and arterial stiffness tion (P < 0.001), even in those without hypertension or
Ventricular stiffness was evaluated by LV end-systolic ela- LVH (Table 2). Based on the cut-off points deduced
stance (0.9  systolic blood pressure/LV end-systolic from controls (mean +2 S.D.), 28 (38.4%) PsA patients
volume) and diastolic elastance [(lateral E/E0 )/LV stroke had increased LV stiffness including 23 (31.5%) in diastole
volume] [14]. Arterial stiffness was estimated by effective (diastolic elastance >0.17) and 13 (17.8%) in systole
arterial elastance (0.9  systolic blood pressure/LV stroke (end-systolic elastance >5.6), and 11 (15.1%) had
volume), total arterial compliance (LV stroke volume/pulse increased arterial stiffness (arterial elastance >3.1).
pressure) and aortic augmentation index [14, 25, 26]. Altogether, 31 (42.5%) patients had increased ventricular
Vascular ventricular coupling was assessed by the and/or arterial stiffness, and 20 (27.4%) had only
ratio of effective arterial elastance to LV end-systolic ela- increased LV stiffness. Furthermore, patients without
stance [14]. hypertension or LVH presented a higher prevalence of
increased ventricular and/or arterial stiffness (27.9% vs.
Statistical analysis 8.0%, P = 0.011) and LV stiffness (20.9% vs. 6.0%,
SPSS 17.0 (SPSS, Inc., Chicago, IL, USA) was used for P = 0.032) compared with matched controls.
the analyses. Results are expressed as mean (S.D.) for
normally distributed data and the median [interquartile
Relationship between disease duration and stiffness
range (IQR)] for non-normally distributed data. Fig. 1 showed the positive correlation between LV dia-
Comparisons among groups were assessed using un- stolic elastance and PsA disease duration (r = 0.472,
paired t-test or analysis of variance test for continuous P < 0.001). Table 3 compared the remodelling and
variables and 2 test for categorical variables. PsA-related parameters between patients with normal
Mann–Whitney U test was used for continuous variables and increased LV diastolic stiffness (identified by diastolic
that were skewed. Curve estimation and logistic regres- elastance >0.17). There was a higher proportion of
sion analysis were performed to assess the relationship women, longer PsA disease duration and a higher

www.rheumatology.oxfordjournals.org 2217
Qing Shang et al.

TABLE 1 Disease-related characteristics in patients (Fig. 1). However, these correlations were not significant
with PsA after adjustment for age, gender and hypertension.

Parameter PsA, n = 73 Discussion

Age at psoriasis diagnosis, years 34 (15) This study has demonstrated that there is increased ven-
Age at PsA diagnosis, years 39 (12) tricular and arterial stiffness in PsA patients, even in those
Psoriasis duration, median (IQR) 10.7 (6.3–18.0) without hypertension and LVH, and this is significantly
PsA duration, years; median (IQR) 8.2 (2.1–11.4) associated with disease duration.
Disease pattern, n (%)
Peripheral 61 (83.6) Stiffness and remodelling in patients with PsA
Spondylarthritis 12 (16.4)
It has been observed that normal age-related vascular
Disease activity, n (%)
Inactive (DAS28 4 3.2) 28 (38.4)
stiffening was accompanied by comparable changes
Moderately active (3.2 < DAS28 4 5.1) 29 (39.7) in the LV stiffness and diastolic compliance [14].
Very active (DAS28 > 5.1) 16 (21.9) Furthermore, this combination concomitance of ventricu-
PASI, median (IQR) 2.4 (1.0–8.0) lar and arterial stiffening seems to be common in patients
Physician’s global (VAS 0–10), 2.0 (0–3.0) with hypertension and heart failure with preserved ejection

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

median (IQR) fraction [27, 28]. Similar changes were observed in our
Patient’s global (VAS 0–10) 4.5 (2.3) current study. Does cardiovascular involvement in PsA
Number of damaged joints, 1.0 (0–4.0)
median (IQR)
share the same mechanisms as the ageing process or
Erosion, n (%) 17 (39.5) hypertension? In our study, PsA patients without hyper-
HAQ, median (IQR) 0.4 (0–0.9) tension and/or LVH still had significantly increased LV
ESR, mm/h, median (IQR) 22.0 (10.0–37.0) elastance and decreased total arterial compliance when
hs-CRP, median (IQR) 4.2 (1.6–12.5) compared with the matched control group. In addition, the
Apo A-1, mg/dl 151.0 (32.0) prevalence of abnormal LV stiffness was significantly
Apo B, mg/dl 83.2 (15.9) higher than that in age, BMI and blood pressure matched
Fibrinogen, mg/l 4.1 (0.8) controls, whereas vascular ventricular coupling was not
PAI-1, median (IQR) ng/ml 24.0 (15.6–36.2)
different. In a recent study, Costa et al. [29] compared
White cell count, 109/l 6.8 (2.3)
Platelet count, 109/l 305.0 (69.9)
the arterial stiffness (identified by aortic pulse wave vel-
Serum creatinine, mmol/l 72.0 (15.1) ocity) between 20 PsA patients without known cardiovas-
Current NSAIDs, n (%) 47 (64.4) cular risk factors and 20 age, weight and height matched
Current DMARDs, n (%) 41 (56.2) healthy controls, and an increased arterial stiffness was
Corticosteroid ever, n (%) 6 (8.2) found in patient group after adjusting for age, weight,
Current anti-hypertensive therapy, n (%) 16 (21.9) height, heart rate and central mean pressure. Therefore,
CCB 8 (11.0) there might be other risk factors contributing to cardio-
ACEI/ARB 2 (2.7) vascular stiffness besides the traditional risk factors
b-Blocker 8 (11.0) (hypertension and LVH) and conventional mechanisms
(age and vascular ventricular coupling). Furthermore,
VAS: visual analogue scale; Apo A-1: apolipoprotein A-1; these changes of ventricular stiffness may have an import-
Apo B: apolipoprotein B; PAI-1: plasminogen activator inhi-
ant detrimental effect on LV mechanics especially on
bitor-1; hs-CRP: high-sensitivity CRP; CCB: calcium-channel
exercise and lead to significant exercise limitation and
blocker; ACEI: angiotensin-converting enzyme inhibitor;
ARB: angiotensin receptor blocker. Data are expressed as breathlessness [30].
mean (S.D.) unless specified otherwise. Ventricular remodelling generally includes cellular pro-
liferation, hyperplasia, necrosis or apoptosis and intersti-
tial fibrosis, which may cause hypertrophy, dilation and
stiffening and finally result in cardiovascular dysfunction.
prevalence of hypertension and remodelling (P < 0.10) in Pieretti et al. found increased LV mass and prevalence of
patients with increased LV diastolic stiffness. Multivariable LVH (16.9%) in patients with SLE compared with age and
logistic regression analysis showed that PsA disease dur- gender matched controls and disease-related inflamma-
ation was an independent explanatory variable of LV dia- tory markers related to LV mass index, which suggested
stolic stiffness [odds ratio (OR) = 1.12, P = 0.012], and long inflammation-related arterial stiffening may be the pos-
disease duration (>10 years) contributed more to LV dia- sible reason [31]. Such ventricular remodelling was also
stolic stiffness (OR = 6.55, P = 0.001) after adjusting for found in our study; however, the prevalence of LVH (9.6%)
age, gender and hypertension (Table 4). and concentric remodelling (11.0%) was relative low.
Similar analysis was repeated between PsA disease Furthermore, Gonzalez-Juanatey et al. investigated 50
duration and other stiffness parameters, showing positive PsA patients without clinically evident CVD or classic ath-
correlation with LV end-systolic elastance (r = 0.238, erosclerosis risk factors and found no significant differ-
P = 0.047) and natural logarithm of aortic augmentation ence of LV dimension and mass compared with
index (r = 0.259, P = 0.036) and a negative correlation matched controls [32]. In addition, Fukuhara et al. [33]
with total arterial compliance (r = 0.267, P = 0.025) and Eliakim-Raz et al. [34] ever reported a potential

2218 www.rheumatology.oxfordjournals.org
 Ventricular–arterial stiffness in PsA

TABLE 2 Comparisons of ventricular and arterial stiffness and remodelling between patients and control groups

Controls, PsA without PsA with

Parameter n = 50 HT or LVH, n = 43 HT or LVH, n = 30 ANOVA, P

Age, years 44 (14) 43 (12) 53 (10)*,y 0.001

Gender, female/male 22/28 20/23 16/14 0.717
BMI, kg/m2 22.8 (3.3) 24.2 (5.4) 26.6 (3.7)*,z 0.005
Systolic BP, mm Hg 120 (12) 119 (10) 138 (17)*,y <0.001
Diastolic BP, mm Hg 74 (9) 74 (7) 82 (11)*,y <0.001
Stiffness
LV end-systolic elastance, mm Hg/ml 3.86 (0.87) 4.31 (1.15) 4.70 (1.32)* 0.007
LV diastolic elastance 0.11 (0.03) 0.13 (0.04)** 0.20 (0.08)*,y <0.001
Effective arterial elastance, mm Hg/ml 2.22 (0.44) 2.32 (0.45) 2.78 (0.88)*,z <0.001
Effective arterial elastance to LV 0.51 (0.15) 0.56 (0.14) 0.60 (0.13)** 0.046
end-systolic elastance ratio
Total arterial compliance, ml/mm Hg 1.26 (0.37) 1.10 (0.31)** 0.92 (0.37)*,z 0.001
Aortic augmentation index, % 14.7 (8.8) 22.0 (15.5)** 29.1 (15.3)*,z 0.001

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

Remodelling
Interventricular septum thickness, cm 0.82 (0.12) 0.87 (0.16) 0.94 (0.22)*,z 0.008
LV end-diastolic diameter, cm 4.68 (0.32) 4.63 (0.45) 4.77 (0.34) 0.274
Posterior wall thickness, cm 0.79 (0.12) 0.81 (0.12) 0.89 (0.15)*,y 0.003
LV mass index, g/m2 74.6 (12.7) 77.8 (15.5) 88.5 (21.1)*,z 0.001
RWT 0.34 (0.05) 0.35 (0.06) 0.37 (0.06)** 0.047
LV remodelling, n (%)
Normal geometry 46 (92.0) 38 (88.4) 20 (66.7)* <0.001
Concentric remodelling 4 (8.0) 5 (11.6) 3 (10.0)*
Eccentric LVH 0 0 4 (13.3)*
Concentric LVH 0 0 3 (10.0)*
Mean of IMTCCA, mm 0.62 (0.08) 0.69 (0.16)** 0.75 (0.13)* 0.002
Max of IMTCCA, mm 0.65 (0.09) 0.73 (0.17)** 0.82 (0.15)*,z <0.001
Cardiac function
Left ventricle
LV ejection fraction, % 65.1 (6.0) 64.4 (5.2) 63.3 (5.7) 0.410
Mean Sm of LV 12 segments, cm/s 5.8 (1.0) 4.5 (1.0)* 4.2 (1.1)* <0.001
Mean Em of LV 12 segments, cm/s 7.8 (1.9) 7.0 (2.0)** 5.5 (1.6)*,y <0.001
E/E0 ratio 5.8 (1.1) 6.1 (1.5) 9.3 (2.8)*,y <0.001
Subclinical LV dysfunction, n (%)
Normal 41 (82.0) 24 (55.8) 5 (16.7) <0.001
Only diastolic dysfunction 9 (18.0) 9 (20.9) 14 (46.7)
Only systolic dysfunction 0 4 (9.3) 0
Both 0 6 (14.0) 11 (36.7)
Right ventricle
PASP, mm Hg 15.2 (9.7) 22.6 (10.1)** 27.3 (13.5)** <0.001
S0 at tricuspid annulus, cm/s 13.8 (1.7) 12.8 (2.4) 11.7 (1.5)** 0.001
E0 at tricuspid annulus, cm/s 14.4 (2.7) 11.8 (3.2)* 9.7 (2.0)*,z <0.001

ANOVA: analysis of variance; HT: hypertension; BP: blood pressure; IMTCCA: IMT thickness of common carotid artery; PASP:
pulmonary arterial systolic pressure; S0 : peak systolic annulus velocity. Data are expressed as mean (S.D.) unless specified
otherwise. *P < 0.005 and **P < 0.05, compared with controls; yP < 0.05 and zP < 0.05, compared with PsA without HT or LVH.

relationship between PsA/psoriasis and dilated cardiomy- accelerating necrosis and apoptosis. In general, inflam-
opathy. All the above studies suggest that there is a di- mation may lead to so-called functional stiffening of the
versity of remodelling in autoimmune inflammatory large arteries because of reduced nitric oxide bioavailabil-
diseases and that LVH may be a late manifestation of ity and increased activity of opposing mediators such as
cardiac involvement in PsA. endothelin-1 [36, 37], and structural stiffening of the larger
vessels because of smooth muscle cell proliferation and
PsA disease duration: a surrogate long-term increased synthesis of structural proteins including colla-
inflammatory burden and a risk factor for stiffness gen [38]. Recent studies have found evidence of CVD in
Many studies have verified that inflammation may accel- SLE without any clinically obvious CVD, such as LVH,
erate cardiovascular damage by its involvement in athero- subclinical myocardial dysfunction and increased arterial
sclerosis, contributing to cardiac fibrosis [35], and stiffness, which were associated with disease duration,

www.rheumatology.oxfordjournals.org 2219
Qing Shang et al.

FIG. 1 Relationship between stiffness and PsA disease duration.

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

PsA duration was positively correlated with LV diastolic elastance and negatively correlated with total arterial compliance.

TABLE 3 Comparison between patients with normal and increased LV diastolic stiffness

LV diastolic stiffness
Parameter
Normal, n = 50 Increased, n = 23 P

Age, years 46 (13) 50 (10) 0.140

Gender, male/female 30/20 7/16 0.019
HT, n (%) 14 (28.0) 13 (56.5) 0.019
LV ejection fraction, % 64.1 (5.4) 63.6 (5.6) 0.712
PASP, mm Hg 23.4 (11.0) 27.1 (13.3) 0.219
S0 at tricuspid annulus, cm/s 12.5 (2.0) 11.9 (2.2) 0.290
E0 at tricuspid annulus, cm/s 11.2 (2.9) 10.3 (3.0) 0.205
Remodelling
LV mass index, g/m2 83.1 (19.9) 80.6 (15.0) 0.597
RWT 0.37 (0.07) 0.35 (0.05) 0.190
LV remodelling, n (%)
Normal geometry 39 (78.0) 19 (82.6) 0.054
Concentric remodelling 8 (16.0) 0
LVH 3 (6.0) 4 (17.4)
Mean of IMTCCA, mm 0.71 (0.15) 0.72 (0.13) 0.877
Max of IMTCCA, mm 0.76 (0.17) 0.78 (0.15) 0.716
PsA-related parameters
Age at psoriasis diagnosis, years 35 (16) 33 (12) 0.864
Age at PsA diagnosis, years 38 (13) 39 (10) 0.853
Psoriasis duration, median (IQR), years 4.9 (1.2–9.3) 10.8 (9.4–13.4) 0.003
PsA duration, median (IQR), years 8.4 (4.0–15.3) 14.6 (10.1–25.3) 0.001
PsA duration >10 years, n (%) 11 (22.0) 15 (65.2) <0.001
Disease pattern, n (%)
Peripheral 44 (88.0) 17 (73.9) 0.131
Spondylarthritis 6 (12.0) 6 (26.1)
Disease activity, n (%)
Inactive (DAS28 4 3.2) 19 (38.0) 9 (39.1) 0.803
Moderately active (3.2 < DAS28 4 5.1) 19 (38.0) 10 (43.5)
Very active (DAS28 > 5.1) 12 (24.0) 4 (17.4)
PASI, median (IQR) 1.8 (0.8–7.7) 3.3 (1.8–8.8) 0.173
Apo A-1, mg/dl 153.9 (33.3) 146.1 (31.7) 0.441
Apo B, mg/dl 84.1 (17.5) 81.6 (13.1) 0.662
ESR, median (IQR) mm/h 16.0 (9.0–37.0) 26.0 (14.5–44.0) 0.296
hs-CRP, median (IQR) mg/l 4.0 (1.6–12.5) 4.3 (1.3–15.4) 1.000
Medical therapy, n (%)
Current NSAIDs 31 (62.0) 16 (69.6) 0.531
Current DMARDs 27 (54.0) 14 (60.9) 0.583
Current anti-hypertensive therapy 9 (18.0) 7 (30.4) 0.233

Data are expressed as mean (S.D.) unless specified otherwise. Apo A-1: apolipoprotein A-1; Apo B:
apolipoprotein B; HT: hypertension; PASP: pulmonary arterial systolic pressure.

2220 www.rheumatology.oxfordjournals.org
 Ventricular–arterial stiffness in PsA

severity and/or activity [31, 39, 40]. Similar findings were for gender, hypertension and LV remodelling. In addition,
also reported in RA [41]. Moreover, the disease-modifying patients with long disease duration (>10 years) had a
anti-rheumatic drugs could improve inflammation- higher risk of LV stiffening. This is similar to results from
associated arterial stiffness [42]. All these studies not the study by Costa et al. [29], where a positive relationship
only confirmed the role of inflammation in the pathogen- was found between arterial stiffness and PsA duration
esis of CVD but also demonstrated that patients with after adjusting for age, weight, height, heart rate and cen-
autoimmune inflammatory diseases are vulnerable to de- tral mean pressure. Although there was no correlation be-
veloping CVD early. tween stiffness and classic inflammatory markers, such as
PsA with long disease duration is characterized by a ESR and high-sensitive CRP in either Costa et al. [29] or
remitting and relapsing disease course, that is to say, our studies, the potential relationship between ventricular
long disease duration can be considered as a surrogate stiffening and inflammatory burden remains highly likely.
of long-term inflammatory burden. This was borne out in Laboratory results just report the inflammatory status at
our current study as disease duration positively correlated one time point, which could not reflect the totality of the
with ventricular and arterial elastance and negatively with inflammatory burden since the onset of the disease. Fig. 2
total arterial compliance. Also, disease duration was an summarized a possible relationship between inflammatory
independent factor of LV diastolic stiffening after adjusting burden and cardiovascular remodelling, stiffening and

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

dysfunction in patients with PsA. Finally, although there
was a significant difference in the duration of psoriasis
TABLE 4 Risk factors of increased LV diastolic stiffness between patients with normal and increased LV diastolic
stiffness, it was not included into the multi-regression ana-
lysis because psoriasis and PsA disease duration belong
Logistic regression analysis
Parameter to the same category, with both indirectly reflecting
Univariate Multivariate long-term inflammatory burdens.
OR (95% CI) OR (95% CI) In this current study, female gender was the other inde-
pendent risk factor for LV diastolic stiffness. A commu-
Age, years 1.03 (0.99, 1.08) –
nity-based study that investigated >2000 subjects also
Female gender 3.43 (1.20, 9.80) 3.12 (1.01, 9.57)
HT 3.34 (1.19, 9.36) – reported that female gender was associated with
LV remodelling 1.18 (0.55, 2.52) – increases in vascular and ventricular systolic and diastolic
PsA duration 1.12 (1.03, 1.22) 1.12 (1.02, 1.22) stiffness even in the absence of CVD [14]. The cause of
PsA duration 6.65 (2.24, 19.74) 6.55 (2.11, 20.32) gender differences in ventricular and arterial stiffening is
> 10 years unclear, although several studies suggested the effects

FIG. 2 Cardiovascular remodelling, stiffening and dysfunction in patients with PsA.

Here, the possible relationship between inflammatory burden and cardiovascular remodelling, stiffening and dysfunction
in patients with PsA is shown. In the early stage of cardiovascular involvement, long-term inflammatory burden may
contribute to increased cardiovascular stiffness and remodelling, which subsequently results in subclinical myocardial
dysfunction.

www.rheumatology.oxfordjournals.org 2221
Qing Shang et al.

of body size and vasculature length on arterial stiffness pathogenesis of heart failure with preserved ejection
[43, 44]. fraction. J Am Coll Cardiol 2009;54:410–18.
The obvious limitation of this study is the relatively the 10 Yu CM, Lin H, Yang H et al. Progression of systolic
small sample size and no intervention to improve those abnormalities in patients with ‘isolated’ diastolic heart
with increased cardiovascular stiffness, but the results failure and diastolic dysfunction. Circulation 2002;105:
seem clear even with this sample size. In conclusion, 1195–201.
this is the first case–control study to assess ventricular 11 Sanderson JE. Heart failure with a normal ejection fraction.
and arterial stiffness in PsA, and an increase in cardiovas- Heart 2007;93:155–8.
cular stiffness was found in patients with PsA, even in 12 Nishikage T, Nakai H, Lang RM, Takeuchi M. Subclinical
those without hypertension and LVH; disease duration left ventricular longitudinal systolic dysfunction in hyper-
as a long-term inflammatory surrogate may be an inde- tension with no evidence of heart failure. Circ J 2008;72:
pendent risk factor. 189–94.
13 Shang Q, Tam LS, Yip GW et al. High prevalence of sub-
clinical left ventricular dysfunction in patients with psori-
Rheumatology key messages
atic arthritis. J Rheumatol 2011;38:1363–70.
. Ventricular and arterial stiffness is high in patients 14 Redfield MM, Jacobsen SJ, Borlaug BA, Rodeheffer RJ,
with PsA.

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

Kass DA. Age- and gender-related ventricular-vascular
. PsA itself contributes to cardiac involvement. stiffening: a community-based study. Circulation 2005;
112:2254–62.
Disclosure statement: The authors have declared no 15 Taylor W, Gladman D, Helliwell P et al. Classification cri-
conflicts of interest. teria for psoriatic arthritis: development of new criteria
from a large international study. Arthritis Rheum 2006;54:
2665–73.
References 16 van der Heijde DM, van’t Hof MA, van Riel PL et al. Validity
of single variables and composite indices for measuring
1 Han C, Robinson DW Jr, Hackett MV et al. Cardiovascular disease activity in rheumatoid arthritis. Ann Rheum Dis
disease and risk factors in patients with rheumatoid arth- 1992;51:177–81.
ritis, psoriatic arthritis, and ankylosing spondylitis. J
17 Fredriksson T, Pettersson U. Severe psoriasis—oral
Rheumatol 2006;33:2167–72.
therapy with a new retinoid. Dermatologica 1978;157:
2 Wong K, Gladman DD, Husted J, Long JA, Farewell VT. 238–44.
Mortality studies in psoriatic arthritis: results from a single
18 Lang RM, Bierig M, Devereux RB et al. Recommendations
outpatient clinic. I. Causes and risk of death. Arthritis
for chamber quantification: a report from the American
Rheum 1997;40:1868–72.
Society of Echocardiography’s Guidelines and Standards
3 Kang S, Fan HM, Li J et al. Relationship of arterial stiffness Committee and the Chamber Quantification Writing
and early mild diastolic heart failure in general middle and Group, developed in conjunction with the European
aged population. Eur Heart J 2010;31:2799–807. Association of Echocardiography, a branch of the
4 Abhayaratna WP, Barnes ME, O’Rourke MF et al. Relation European Society of Cardiology. J Am Soc Echocardiogr
of arterial stiffness to left ventricular diastolic function and 2005;18:1440–63.
cardiovascular risk prediction in patients > or =65 years of 19 Yu CM, Fung JW, Zhang Q et al. Tissue Doppler imaging is
age. Am J Cardiol 2006;98:1387–92. superior to strain rate imaging and postsystolic shortening
5 Mizuguchi Y, Oishi Y, Tanaka H et al. Arterial stiffness is on the prediction of reverse remodeling in both ischemic
associated with left ventricular diastolic function in pa- and nonischemic heart failure after cardiac resynchroni-
tients with cardiovascular risk factors: early detection with zation therapy. Circulation 2004;110:66–73.
the use of cardio-ankle vascular index and ultrasonic 20 Nagueh SF, Middleton KJ, Kopelen HA, Zoghbi WA,
strain imaging. J Card Fail 2007;13:744–51. Quinones MA. Doppler tissue imaging: a noninvasive
6 Eren M, Gorgulu S, Uslu N et al. Relation between aortic technique for evaluation of left ventricular relaxation and
stiffness and left ventricular diastolic function in patients estimation of filling pressures. J Am Coll Cardiol 1997;30:
with hypertension, diabetes, or both. Heart 2004;90: 1527–33.
37–43. 21 Ommen SR, Nishimura RA, Appleton CP et al. Clinical
7 Patrianakos AP, Parthenakis FI, Karakitsos D et al. utility of Doppler echocardiography and tissue Doppler
Relation of proximal aorta stiffness to left ventricular imaging in the estimation of left ventricular filling pres-
diastolic function in patients with end-stage renal disease. sures: a comparative simultaneous
J Am Soc Echocardiogr 2007;20:314–23. Doppler-catheterization study. Circulation 2000;102:
8 Tavil Y, Kanbay A, Sen N et al. The relationship between 1788–94.
aortic stiffness and cardiac function in patients with ob- 22 Nagueh SF, Appleton CP, Gillebert TC et al.
structive sleep apnea, independently from systemic Recommendations for the evaluation of left ventricular
hypertension. J Am Soc Echocardiogr 2007;20:366–72. diastolic function by echocardiography. J Am Soc
9 Borlaug BA, Lam CS, Roger VL, Rodeheffer RJ, Echocardiogr 2009;22:107–33.
Redfield MM. Contractility and ventricular systolic stiffen- 23 Van Bortel LM, Duprez D, Starmans-Kool MJ et al.
ing in hypertensive heart disease insights into the Clinical applications of arterial stiffness,

2222 www.rheumatology.oxfordjournals.org
 Ventricular–arterial stiffness in PsA

Task Force III: recommendations for user procedures. 34 Eliakim-Raz N, Shuvy M, Lotan C, Planer D. Psoriasis and
Am J Hypertens 2002;15:445–52. dilated cardiomyopathy: coincidence or associated dis-
24 Wilkinson IB, MacCallum H, Flint L et al. The influence of eases? Cardiology 2008;111:202–6.
heart rate on augmentation index and central arterial 35 Kania G, Blyszczuk P, Eriksson U. Mechanisms of cardiac
pressure in humans. J Physiol 2000;525(Pt 1):263–70. fibrosis in inflammatory heart disease. Trends Cardiovasc
25 DeLoach SS, Townsend RR. Vascular stiffness: its meas- Med 2009;19:247–52.
urement and significance for epidemiologic and outcome 36 Wilkinson IB, Qasem A, McEniery CM et al. Nitric oxide
studies. Clin J Am Soc Nephrol 2008;3:184–92. regulates local arterial distensibility in vivo. Circulation
26 Borlaug BA, Melenovsky V, Redfield MM et al. Impact of 2002;105:213–7.
arterial load and loading sequence on left ventricular 37 McEniery CM, Qasem A, Schmitt M et al. Endothelin-1
tissue velocities in humans. J Am Coll Cardiol 2007;50: regulates arterial pulse wave velocity in vivo. J Am Coll
1570–7. Cardiol 2003;42:1975–81.
27 Varagic J, Susic D, Frohlich E. Heart, aging, and hyper-
38 Dzau VJ, Braun-Dullaeus RC, Sedding DG. Vascular pro-
tension. Curr Opin Cardiol 2001;16:336–41.
liferation and atherosclerosis: new perspectives and
28 Borlaug BA, Olson TP, Lam CS et al. Global cardio- therapeutic strategies. Nat Med 2002;8:1249–56.
vascular reserve dysfunction in heart failure with pre-
39 Shang Q, Tam LS, Li EK, Yip GW, Yu CM. Increased ar-
served ejection fraction. J Am Coll Cardiol 2010;56:

Downloaded from http://rheumatology.oxfordjournals.org/ at Oakland University on April 6, 2015

terial stiffness correlated with disease activity in systemic
845–54.
lupus erythematosus. Lupus 2008;17:1096–102.
29 Costa L, Caso F, D’Elia L et al. Psoriatic arthritis is asso-
40 Yip GW, Shang Q, Tam LS et al. Disease chronicity
ciated with increased arterial stiffness in the absence of
and activity predict subclinical left ventricular systolic
known cardiovascular risk factors: a case control study.
dysfunction in patients with systemic lupus erythemato-
Clin Rheumatol 2012;31:711–5.
sus. Heart 2009;95:980–7.
30 Tan YT, Wenzelburger F, Lee E et al. Abnormal left ven-
41 Klocke R, Cockcroft JR, Taylor GJ, Hall IR, Blake DR.
tricular function occurs on exercise in well-treated hyper-
Arterial stiffness and central blood pressure, as deter-
tensive subjects with normal resting echocardiography.
Heart 2010;96:948–55. mined by pulse wave analysis, in rheumatoid arthritis. Ann
Rheum Dis 2003;62:414–8.
31 Pieretti J, Roman MJ, Devereux RB et al. Systemic lupus
erythematosus predicts increased left ventricular mass. 42 Galarraga B, Khan F, Kumar P, Pullar T, Belch JJ.
Circulation 2007;116:419–26. Etanercept improves inflammation-associated arterial
stiffness in rheumatoid arthritis. Rheumatology 2009;48:
32 Gonzalez-Juanatey C, migo-Diaz E, Miranda-Filloy JA 1418–23.
et al. Lack of echocardiographic and Doppler abnormal-
ities in psoriatic arthritis patients without clinically evident 43 Gatzka CD, Kingwell BA, Cameron JD et al. Gender
cardiovascular disease or classic atherosclerosis risk differences in the timing of arterial wave reflection beyond
factors. Semin Arthritis Rheum 2006;35:333–9. differences in body height. J Hypertens 2001;19:
2197–203.
33 Fukuhara K, Urano Y, Akaike M, Ahsan K, Arase S.
Psoriatic arthritis associated with dilated cardiomyop- 44 Shim CY, Park S, Choi D et al. Sex differences in central
athy and Takayasu’s arteritis. Br J Dermatol 1998;138: hemodynamics and their relationship to left ventricular
329–33. diastolic function. J Am Coll Cardiol 2011;57:1226–33.

www.rheumatology.oxfordjournals.org 2223