Introduction

Spondyloarthritis (SpA), one of the most prevalent chronic rheumatic diseases, is characterized by the occurrence of accelerated atherosclerosis and increased cardiovascular (CV) morbidity and mortality compared with the general population [1]. Although mechanisms of excess CV risk are not fully elucidated, chronic inflammatory process is believed to be the underlying cause contributing to all stages of atherosclerosis [2], including endothelial dysfunction, early atheroma formation, plaque instability, and arterial stiffness. Assessment of subclinical vascular changes before clinical atherosclerotic manifestations is important for CV risk stratification and optimal management as has been recommended by the European League Against Rheumatism (EULAR).

A variety of imaging modalities have been used to assess subclinical atherosclerosis and CV risk. The most widely recognized and validated marker of subclinical atherosclerosis is carotid intima-media thickness (cIMT) measured non-invasively by high resolution B-mode which has proved its reproducibility and reliability [3]. A second approach is the flow-mediated dilation (FMD) measurement which is also a validated method for the assessment of endothelial dysfunction [4]. Endothelial dysfunction is the key early event in atherogenesis, appearing long before the structural atherosclerotic changes. Recently, studies have identified epicardial adipose tissue (EAT) located between the myocardium and visceral pericardium as an active endocrine organ and a source of several pro-atherogenic and inflammatory adipokines exerting major effects on the vascular system and atherosclerosis pathogenesis [5]. It has been also documented that EAT is rather strongly associated to coronary artery diseases (CAD) than other visceral adipose tissues [6]. Echocardiographic EAT thickness measurement has then emerged as a novel marker of subclinical coronary atherosclerosis and has been proposed as a useful and reliable tool for CV risk stratification [7]. While some reports demonstrated a positive correlation between EAT thickness and cIMT in SpA patients [8, 9], the association between EAT thickness and FMD has not yet been studied in SpA.

Therefore, this study aimed to investigate the relationship between EAT thickness, FMD, and cIMT, as markers of subclinical atherosclerosis in SpA patients compared with those of the healthy population.

Materials and methods

This was a cross-sectional study, enrolling young SpA patients (aged ≤ 50 years) meeting ASAS 2009 criteria for axial and peripheral SpA. Healthy volunteers matched for age and gender were recruited as a control group. Cases of juvenile SpA, enteropathic SpA, psoriatic arthritis, or reactive arthritis were not included. In addition, patients and controls were not included if they had a previous history of congestive heart failure, CAD, cerebrovascular disease, valvulopathy or chronic kidney disease, family history of premature CAD, high blood pressure (systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg), hyperlipidemia, obesity, diabetes mellitus, active infection or connective tissue disease, chronic inflammatory bowel disease, dysthyroidism, liver disease, or nephrotic syndrome capable of affecting lipid metabolism. Subjects using alcohol or cigarettes were also not included.

Before the beginning of the study, all SpA patients and healthy controls gave written informed consent following the principles of the Declaration of Helsinki. The institutional ethics committee approved our study protocol.

Data collection

In both patients and control groups, general data (age, gender and body mass index (BMI)) were collected. Blood pressure and heart rate (HR) were measured after a 10-min rest in a sitting position.

In SpA patients, we recorded disease-related parameters: age at onset of SpA, disease duration, extra-articular features, pain visual analog scale (VAS), bath ankylosing spondylitis global-score (BASG-s), ankylosing spondylitis disease activity score (ASDAS), bath ankylosing spondylitis disease activity index (BASDAI), bath ankylosing spondylitis mobility index (BASMI), bath ankylosing spondylitis functional index (BASFI), Lequesne index and therapeutics data (non-steroidal anti-inflammatory drugs (NSAIDs), conventional synthetic DMARDs (csDMARDs), or tumor necrosis factor α inhibitors (TNFi)).

Laboratory data

Venous blood samples were taken in all participants after 12 h of fasting and were analyzed in order to determine fasting glucose, lipid profile (total cholesterol, triglyceride, HDL-C (high-density lipoprotein-cholesterol), LDL-C (low-density lipoprotein-cholesterol)), creatinine, and C-reactive protein (CRP) levels. Lipid ratios (cholesterol to HDL-C and LDL-C to HDL-C) were also calculated.

Radiographic data

For the assessment of structural radiographic damage, radiographs (cervical and lumbar spine) were scored using the modified stoke ankylosing spondylitis spinal score (mSASSS), and the bath ankylosing spondylitis radiology index (BASRI)). Lequesne’s false profile radiograph was obtained to allow diagnosis of coxitis.

Ultrasound examination

Epicardial adipose tissue thickness assessment

A transthoracic echocardiographic examination was performed on all participants by an experienced cardiologist blinded to SpA diagnosis and clinical status, using a Vivid 9-general electric GE system. Subjects were lying in left lateral decubitus position. EAT thickness was measured according to the validated method described by Iacobellis et al. [10].

EAT was identified as the echo-free space between the outer surface of the myocardium and the visceral layer of the pericardium. EAT thickness was measured in parasternal long-axis view, perpendicularly to the free wall of the right ventricle, at end-systole in three cardiac cycles. Maximum EAT thickness was measured from a point on the free wall of the right ventricle and a second point perpendicular to the aortic annulus. We considered the average value of three cardiac cycles.

Flow-mediated dilatation

FMD was measured ultrasonographically by a single experienced radiologist according to the guidelines of the American college of cardiology [11], using a Mindray Resona 7 Z ST + and a 7–10-MHz linear transducer. Measurements were performed in a quiet ambient temperature-controlled room (24 °C) after 6 h of fasting. Caffeine intake, smoking, vasoactive drugs, high-fat foods, vitamin C, and intense exercise were not allowed for 24 h before the brachial study. The right arm was kept immobilized in supination during the entire study. The brachial artery was imaged in the longitudinal plane above the antecubital fossa. A blood pressure cuff was first placed on the forearm. After brachial artery diameter was determined in end-diastole, ischemia was induced by inflating the cuff to a supra systolic level (usually at least 50 mm Hg above the systolic pressure) to obtain arterial occlusion for 5 min. Then, a measurement of the maximal post-ischemic diameter was taken 60 s after cuff deflation. FMD was calculated as the percent change of the artery diameter compared to the baseline value according to the following formula: FMD = (maximum diameter − baseline diameter)/baseline diameter × 100. The percentage of FMD reflects the arterial diameter response to increased blood flow. Lower values indicate impaired endothelial function.

Carotid intima-media thickness

A high-resolution B-mode ultrasonography with the same machine (Mindray Resona 7 ZST +) was used for cIMT measurement. During the examination, the subject lay in a dorsal decubitus position, with the neck extended and slightly turned to the opposite side. The probe was placed parallel to the common carotid artery (CCA). On a longitudinal view, we initially identified the two parallel echogenic lines separated by a hypoechogenic border corresponding to the intima-media complex. The distance between the leading edge of the first echogenic line (lumen-intima interface) of the far wall and the leading edge of the second echogenic line (upper layer of tunica adventitia) was taken as the intima-media thickness. cIMT was measured at three points on the far walls of the left and right CCA 2 cm proximal to the carotid bulb. Then, we calculated the average value of the three locations to obtain the mean cIMT on each side.

We considered that cIMT was increased between 0.7 and 1.5 mm. A spur larger than 1.5 mm inside the vessel lumen was defined as an atherosclerotic plaque [12].

Statistical analysis

We used SPSS software version 25.0 for statistical analysis. Given the lack of standardized cut-off values, we considered in our analysis the EAT thickness and FMD as continuous binary variables. Categorical variables were expressed as percentages. Skewed parameters were expressed as medians and 25th and 75th percentiles. Comparisons were performed using non-parametric tests because of non-normally distributed data: Mann Whitney U, Pearson’s chi-squared test, and Kruskal–Wallis. Spearman’s correlation coefficient and Person correlation were used to analyzing the relationship between the variables. The Cramer’s Phi and V tests were used to determine the strength of the association. p-values < 0.05 were accepted as statistically significant.

Results

A total of 94 subjects were enrolled (47 SpA patients and 47 healthy controls). Baseline clinical and biological characteristics of SpA patients and cross-matched healthy volunteers are listed in Table 1. There was no significant difference between the patients and control groups in terms of age, gender, BMI, and blood pressure. In addition, no difference was found in blood lipid profile, fasting glucose, and renal function. SpA patients had significantly higher levels of CRP (p = 0.001).

Table 1 Clinical and biological characteristics in SpA and control groups
Full size table

Disease characteristics

The median age at onset of SpA was 20 years (IQR: 18–32), and the median disease duration was 11 years (IQR: 5–16). Regarding disease activity, median BASDAI and ASDAS-CRP were 2.6 (IQR: 1.8–3.8) and 2.18 (IQR: 1.62–2.91) respectively. High disease activity according to BASDAI and ASDAS-CRP were noted in 21% and 55% of patients, respectively. CRP levels were increased in 21 patients (45%). For structural damage, median mSASSS and BASRI scores were 10 (IQR: 4–15) and 3 (IQR: 2-–), respectively. Hip involvement was noted in 53% of patients. As regard treatment, SpA patients received: NSAIDs (92%), csDMARDs (51%), and TNFi (38%). Table 2 summarizes the main SpA-related parameters.

Table 2 Clinical and biological features related to SpA
Full size table

Ultrasound assessment

SpA patients exhibited significantly thicker EAT values compared to controls: median EAT thickness was 3.1 mm (IQR: 2.5–4) vs 2.4 mm (IQR: 2–3); (p = 0.001). FMD was significantly lower in patients group (with a median of 14.6% (IQR: 9–24) vs 18.8% (IQR: 12.8–23.1%); p = 0.008) (Figs. 1 and 2). Median right, left, and mean cIMT were respectively 0.54 mm (IQR: 0.50–0.63), 0.55 mm (IQR: 0.49–0.61), and 0.55 mm (IQR: 0.48–0.62) in SpA patients and 0.45 mm (IQR: 0.42–0.50), 0.47 mm (IQR: 0.45–0.50), and 0.46 mm (IQR: 0.43–0.50) in healthy controls. Median right, left, and mean cIMT values were significantly higher in SpA patients in comparison with control subjects (p < 0.0001). Increased cIMT values (cIMT > 0.7 mm) were noted in 8 patients (17%), while no patient had atherosclerotic plaque (cIMT > 1.5 mm).

Fig. 1
figure 1

Comparaison of EAT thickness between the patients and the control groups

Full size image
Fig. 2
figure 2

Comparaison of FMD values between SpA and control groups

Full size image

The analysis of the association between these three ultrasound (US) measurements is summarized in Table 3. We demonstrated a negative correlation between EAT thickness and FMD (p = 0.026; r =  − 0.325) illustrated in Fig. 3. FMD was also negatively correlated with left cIMT (p = 0.027; r =  − 0.322). However, there was no significant association between EAT thickness and cIMT values.

Table 3 Associations between echographic parameters
Full size table
Fig. 3
figure 3

Association between EAT thickness and FMD

Full size image

Discussion

The present study confirmed firstly the increased CV burden and accelerated atherosclerosis as witnessed in all markers of subclinical atherosclerosis. The three US parameters including EAT thickness, FMD, and cIMT were significantly impaired in a cohort of young SpA patients without CV risk factors compared to age and sex-matched healthy controls.

These findings were consistent with the literature data concerning EAT thickness. To date, 8 case–control studies have recently focused on EAT thickness in SpA patients and found significantly thicker the US EAT values in comparison to the healthy population [8, 9, 13,14,15,16,17,18].

Previous reports including 2 meta-analysis demonstrated also significantly increased cIMT values in SpA patients [8, 9, 19,20,21,22] and concluded to a subsequent higher risk of subclinical atherosclerosis in line with our result. Regarding endothelial dysfunction, our results were in agreement with the updated data which showed a significant decrease in FMD in SpA patients compared to healthy subjects [23,24,25,26,27,28,29,30,31,32,33,34,35].

Interestingly, the novel finding of our study is the negative association between FMD and EAT thickness. To our knowledge, this is the first study to demonstrate a negative correlation between endothelial dysfunction and a marker of subclinical coronary atherosclerosis (EAT thickness) in SpA. Similar results were described in patients with type I diabetic patients [26, 27], and in a cohort of 90 rheumatoid arthritis (RA) patients [28]. EAT thickness was an independent factor influencing endothelial function [26, 27].

There is growing evidence about the physiologic and metabolic significance of EAT. Rather being a passive fat deposit, EAT is involved in lipid and energy homeostasis and is regarded as a functional endocrine organ that secretes several pro-atherogenic mediators accused of contributing to atherosclerosis pathogenesis [10]. Inflammatory adipokines such as tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), monocyte, chemoattractant protein-1 (MCP1), IL-1b, plasminogen activator inhibitor-1 (PAI-1), and resistin (which has a mimetic effect and is involved in insulin resistance) were found to be highly expressed by EAT [29] and promoted atherogenesis either by an endocrine effect or by direct diffusion through the vasa vasorum [10]. Previous studies have also demonstrated that EAT thickness is strongly linked to the presence and severity of CAD [30, 31]. Several studies have also shown that endothelial dysfunction especially the reduction in the bioavailability of endothelium-derived nitric oxide (NO) is the initiator and reversible step in atherosclerosis pathogenesis and is one of the best predictors of CVD risk [32]. Recent reports revealed a significant association between endothelial dysfunction and coronary atherosclerosis in patients with intermediate cardiac risk [33]. Brachial artery FMD was independently associated with the presence and extent of subclinical atherosclerosis on coronary computed tomography angiography; it was negatively associated with higher coronary artery calcium (CAC) score, segment involvement score (total number of segments including any plaque), and segment stenosis score in a recent study [33]. Consequently, we suggest that FMD as a surrogate marker of endothelial dysfunction may reflect coronary atherosclerotic burden.

Therefore, EAT thickness as an emerging cardiometabolic factor correlated with the earliest step of atherosclerosis. Adipokines that play an important role in the regulation of the arterial tonus may explain this association. It seems that being higher than the physiological limit, EAT induces a harmful cytokine phenotype. The imbalance between vasodilator adiponectin (which has an anti-atherogenic effect by reducing endothelial activation) and vasoconstrictor resistin has been associated with endothelial dysfunction [34, 35]. In vivo, resistin impaired bradykinin-induced relaxation and authors suggested it may act at a cell signaling point upstream of NO or prostaglandin production [35]. EAT has been also described to affect the endothelium by inducing cell surface expression of adhesion molecules and increasing adhesion of monocytes to endothelial cells [36]. Pro-inflammatory cytokines secreted in excess by increased EAT may also explain its association with FMD, by playing a pivotal role in the pathogenesis of endothelial dysfunction [37]. TNF-α over-expression can decrease the release of endothelial NO and cause subsequent impairment of endothelium-dependent vasodilatation [38]. Similarly, administration of IL-6 to mice induced vasoconstriction, increased vascular superoxide production, and impaired endothelium-dependent vasodilatation [39]. These findings raise the question if EAT thickness measurement could be an alternative to reflect endothelial dysfunction.

As we verified in the current study, Bodnàr et al. also demonstrated a significant negative association between cIMT and FMD in a cohort of 43 SpA patients [23].

There are some controversies about the correlation between EAT thickness and cIMT. While some researchers such as Üstün et al. found no evidence of any association [16], others confirmed that EAT thickness was positively correlated with cIMT in SpA patients [8, 9]. In our study, no significant association was obtained between EAT thickness and cIMT. The relationship between EAT and intimal infiltration has been well investigated in non-rheumatic patients, and several authors have demonstrated their strong independent association [40, 41]. EAT thickness was found to be a significant predictor of increased cIMT and the presence of carotid plaque [40, 41]. Athough its threshold value is not yet precisely defined, echocardiographic measured EAT thickness in diastole ≥ 5.0 mm was associated with a significantly higher prevalence of carotid artery plaque in patients with low Framingham risk scores or overweight and has been therefore suggested to identify individuals with a higher risk of having detectable carotid atherosclerosis [42]. Although underlying mechanisms are complex and not fully understood, this association could be explained again by the pro-inflammatory and pro-atherogenic effects of EAT. Increased EAT thickness act associated with adipocytokine imbalance (with up up-regulation of inflammatory adipokines and down-regulation of anti-inflammatory adipokines) [5], affecting coronary and also systemic arteries [43]. IL-6 secreted in excess by EAT induces stimulation of vascular smooth muscle proliferation and subsequently atherosclerotic plaques as well as endothelial cell activation [44]. Oxidative stress with high levels of reactive oxygen species in EAT, increased secretion of group IID secretory phospholipase A2 resulting in accumulation of lipids within atherosclerotic plaques, and high expression of adhesion molecules involved in the pathogenesis of atherosclerosis such as MCP-1, growth-regulated α protein and C–C motif chemokine 5 [36] contribute also to the atherogenicity of EAT [43].

This study has some limitations. First, it was a cross-sectional study; therefore, we cannot conclude to the causal relationship between SpA characteristics and subclinical atherosclerosis markers. Further longitudinal studies involving larger samples are necessary to confirm our results. Second, to date, there are no standardized cut-threshold values for FMD and EAT thickness which has been considered as continuous binary variables in our analysis. Our study group was also heterogeneous in terms of age and disease duration which can cause an interpretation bias.

Conclusion

EAT thickness, FMD, and cIMT as markers of subclinical atherosclerosis were significantly impaired in SpA patients free from CV risk factors in comparison with healthy controls supporting the evidence of inflammatory induced accelerated atherosclerosis. In addition, our study reported the first evidence of a cross-relationship between EAT thickness and endothelial dysfunction suggesting the interest of EAT thickness measurement in diagnosis of early functional atherosclerotic changes.