Abstract
Behçet's disease (BD) is a vasculitic and inflammatory disease causing endothelial dysfunction. Irisin is a metabolic hormone related to insulin resistance and endothelial functions. In this study, we investigated the relationship between irisin and carotid intima-media thickness (cIMT), which is a marker of atherosclerosis in patients with BD. 48 patients with BD and 50 healthy individuals were enrolled in the study. Disease severity was evaluated by BD current activity form. Irisin, glucose, insulin, C reactive protein, erythrocyte sedimentation rate and lipid panel were examined in all patients. Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) was used to calculate insulin resistance. A simple and inexpensive cIMT test was used as indicator of atherosclerosis. cIMT was 0.62 (0.45–1.05) mm in the patients, while it was 0.38 (0.25–0.65) mm in the control group (p<0.001). Irisin value was found to be 197.3 (24.8–834.2) ng/mL in the control group, while it was 85.4 (4.7–471.1) ng/mL in the patient group (p=0.007). There was a negative correlation between irisin level and cIMT (r=−0.511, p<0.001) and HOMA-IR (r=−0.371, p=0.009). Decreased irisin levels (OR 0.996, 95% CI 0.992 to 1.000, p=0.041), male gender (OR 7.634, 95% CI 1.415 to 41.191, p=0.018), and HOMA-IR (OR 2.596, 95% CI 1.451 to 4.643, p=0.001) are independent risk factors for cIMT in patients with BD. We detected a very strong relationship between cIMT, which is an indicator of subclinical atherosclerosis, and decreased irisin levels in patients with BD. BD is characterized by chronic inflammation, and low serum irisin levels in BD may be related to atherosclerosis.
Significance of this study
What is already known about this subject?
Behçet's disease is a vasculitic disease characterized by inflammation.
Patients with Behçet's disease frequently experience cardiovascular ailments.
In various studies, it has been stated that carotid intima-media thickness is a predictive marker for cardiovascular disease development.
Irisin is a popular myokine that is associated with increased insulin resistance and endothelial dysfunction.
What are the new findings?
The serum irisin level of the patients with Behçet's disease was significantly lower when compared to the irisin level of the control group.
Serum irisin levels were related to insulin resistance in patients with Behçet's disease.
Serum irisin levels have strong negative association with carotid intima-media thickness in patients with Behçet's disease.
Low serum irisin levels may be independent risk factor for carotid intima-media thickness.
How might these results change the focus of research or clinical practice?
Whether or not patients will acquire atherosclerotic heart disease or be insulin resistance can be estimated by following serum irisin levels in those with Behçet's disease. Additionally, in patients with rheumatological diseases, irisin levels can be a predictive marker for atherosclerotic heart disease, and our study can guide further studies performed on these diseases.
Introduction
Behçet's disease (BD) is a chronic, relapsing and inflammatory disease in which cardiovascular involvement has been estimated to range between 7% and 46%.1 Main histopathological features of BD are characterized by acute systemic inflammation and chronic systemic vasculitis associated with endothelial cell dysfunction.2 Systemic inflammation seen in chronic inflammatory disorders contributes to cardiovascular disease (CVD) via proven mechanisms: accelerated atherosclerosis, insulin resistance (IR), hyperglycemia, hypercoagulability, hypercholesterolemia and platelet dysfunction.3 Observational studies have supported the information that patients with pre-existing chronic inflammatory diseases have a dramatically increased risk for CVD at younger ages, which is related to the fact that endothelial dysfunction is considered a common initial lesion in the development of atherosclerosis.3
Endothelial dysfunction, a well-recognized index of subclinical vascular atherosclerosis that is measured by carotid intima-media thickness (cIMT) of the common carotid artery on ultrasonography, is the earliest event in vascular complications of BD, and contributes significantly to the initiation and progression of vascular injuries in different regions of the body, causing metabolic disease complications.4 Recent studies have shown that an increased susceptibility to IR was related to chronic inflammation, endothelial dysfunction and metabolic abnormalities in patients with BD.5 6
IR is a pathological condition characterized by a decrease in insulin activity regulating blood glucose levels, and it occurs as a response to a complex interaction between metabolic and inflammatory mediators ensuring body energy balance.7 It has been reported in recent studies that irisin had the potential to become a therapeutic target for endothelial dysfunction and metabolic disorders.8 Irisin, as a novel hormone-like myokine that plays a pivotal role in energy expenditure and metabolic regulation, is mainly secreted by the heart, skeletal muscle, liver, kidneys, nerves and skin.9 Previous studies revealed the relationship between circulating irisin levels, endothelial dysfunctions and subclinical atherosclerosis in non-diabetic adult patients.10 In another recent study, it was demonstrated that serum irisin level was significantly correlated with carotid atherosclerosis in patients receiving dialysis.11 Interestingly, there are studies specifying that there is either negative or positive relationship between irisin levels and metabolic syndrome/HOMA-IR.12 13 There is a contradiction in this regard.
Until today, no studies have been performed to directly examine the relationship between circulating irisin levels, IR and subclinical atherosclerosis in BD after adjusting for potential confounders. Mounting evidence for irisin may contribute to the exploration of novel and effective therapeutic targets or therapeutic strategies. Therefore, the aim of present study was to evaluate whether circulating irisin was related to endothelial dysfunction and IR in patients with BD.
Material and methods
Study design and patient selection
This prospective and observational study was conducted in the Rheumatology and Cardiology outpatient clinics of our hospital. Forty-eight female patients with BD and age-body mass index (BMI), muscle mass, and exercise condition (all individuals were under sedentary position), matched with 50 healthy female subjects as the control group, were enrolled in the study. The patients who had been diagnosed with BD according to International BD Study Group criteria were enrolled in the study (International study group for BD. Evaluation of diagnostic (‘classification’) criteria in BD disease towards internationally agreed criteria).14 Disease severity in the patients with BD was evaluated by using BD current activity form.15 Patients with concomitant systemic diseases such as diabetes mellitus, chronic obstructive lung disease, coronary artery disease, cancer, thyroid function disorder, hematological disorders, acute or chronic liver and renal diseases, acute or chronic infections, and a history of smoking and alcohol consumption, were excluded. Patients with BD who were included in this study were using only colchicine as a drug. We excluded patients with BD who were using drugs that can affect cIMT (such as steroids or statins). Healthy subjects with no regular use of medications, history of neither smoking nor alcohol consumption, and those who did not have a known disease, were involved as control group. The study was approved by the Ethical Committee of Our University Medical Faculty and the study was performed according to the Declaration of Helsinki. It was explained to all participants that cIMT measurement and serum analyses would not cost them any money, and blood samples collected for cIMT measurement and serum analyses would harm neither the patients nor those in the control group. After explaining these details, written consent of participants who agreed to join the study was obtained.
Ultrasonographic examination method
Radiological imaging
Following ultrasonographic detection of right and left arteries of patient and control groups, in supine position, carotid intima media thickness values were measured from right and left arteria carotis communis. cIMT measurement was carefully performed using a high-resolution ultrasound machine (Logiq S6; General Electric, Milwaukee, Wisconsin, USA) with a 12-MHz mechanical sector transducer device by one radiologist who was blinded to the results of the blood tests. Three measurements were performed in systolic phase on three different points on the posterior wall of the artery, 2 cm proximal to the bifurcation of the arteria carotis communis. cIMT was calculated by using the average of these three measurements.
Biochemical analysis
Venous blood samples were obtained from all participants after 10–12 h fasting. Fasting plasma glucose (FPG), creatinine, alanine aminotransferase, total cholesterol (TC), triglyceride (TG), and high-density lipoprotein cholesterol (HDL-C) levels were measured using a Synchron LX20 system (Beckman Coulter, USA) and Beckman reagents. HDL-C levels were detected using a direct enzymatic method without precipitation. Low-density lipoprotein cholesterol levels were computed using the Friedewald formula. Hemoglobin level was determined using a Coulter Max M mode device. C reactive protein (CRP) levels were analyzed by an autoanalyser Synchron LX System (Beckman Coulter, Fullerton, California, USA). Erythrocyte sedimentation rate was determined by iSed (Alcor Scientific). Thyroid stimulating hormone (TSH) levels (0.4–4 mU/L) were determined using commercially available kits (DPC Diagnostics Products Corporation, Los Angeles, California, USA). Insulin resistance score, Homeostasis Model Assessment+Insulin Resistance (HOMA+IR), was calculated by FPG (mmol/L)×fasting serum insulin (mU/mL)/22.5 formula and 2.7 was accepted as cut-off value.
Irisin measurements
Venous blood samples were collected from all patients in the morning at 08:00 after a long-term rest. If blood samples for irisin are stored for a long period of time (eg, at −80°C for 6 months), its levels will strongly effect;16 therefore, serum samples were stored for less than 3 months, after being evaluated immediately after collection. The concentration of irisin was assessed by using the ELISA method. We used a commercially available human irisin ELISA kit (Cusabio, PR China). The procedure applied for the ELISA method was consistent with the instructions provided by the manufacturer. Absorbance was measured at a wavelength of 450 ηm using an ELISA reader. Irisin levels were presented as ng/mL. The smallest detectable dose for irisin assay was 0.78 ng/mL. Intra-assay coefficient of variation (CV) was <8% and inter-assay CV was <10%.
Statistical analysis
SPSS V.18 was used for statistical analysis. The results were presented as mean±SD, median and minimum–maximum. Kolmogorov Smirnov test was used to determine if the groups were homogeneously distributed. Normally and non-normally distributed data were analyzed by Independent T test and Mann Whitney U test, respectively. Spearman correlation analysis test was used for correlation analysis. Multiple logistic regression analysis was implemented to detect the independent relationship between cIMT and irisin levels, and other parameters such as CRP, IR and BMI. p Values <0.05 were accepted as significant.
Results
Baseline characteristics
The patient group's age (36.7±11.5 years) and BMI (26.1±4.2 kg/m2) values were similar to the control group's age (36.4±12.0 years) and BMI (25.6±4.2 kg/m2) values. In the patients with BD, disease duration was 7.8±6.2 years and disease activity index was found to be 3.0 (1.0–7.0). cIMT of the patient group was significantly higher than that of control group—(0.62 (0.45–1.05) mm vs 0.38 (0.25–0.65) (p<0.001). Table 1 shows sociodemographic features of the patient and control groups.
Sociodemographic characteristics of the patient and control groups
BMI, body mass index; cIMT, carotid intima-media thickness, CNS, central nervous system.
The serum irisin level of the patients with BD (85.4 (4.7–471.1) ng/mL) was significantly lower when compared to the irisin level of the control group (197.3 (24.8–834.2) ng/mL, p=0.007) (figure 1). In the patients with BD, serum glucose (93.0 (72.0–122.0) mg/dL), insulin (13.4±7.4), CRP (2.6 (0.5–43.0) mg/dL) and HOMA-IR (2.7 (0.8–7.5)) values were significantly higher than glucose (88.0 (67.0–109.0) mg/dL, p=0.005), insulin (9.6±4.7, p=0.003), CRP (1.2 (0.2–5.0) mg/dL, p<0.001), and HOMA-IR (1.7 (0.7–5.2), p=0.002) levels of the control group. All biochemical results of the patients with BD and control group patients are presented in table 2.
Biochemical parameters of the patient and control groups
Mean±SD.
ALT, alanine aminotransferase; CRP, C reactive protein; ESR, erythrocyte sedimentation rate; Hb, hemoglobin; HDL-C, high-density Lipoprotein cholesterol; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; TSH, thyroid-stimulating hormone; WCC, white cell counts.
Serum irisin levels in BD and control groups. BD, Behçet's disease.
When serum irisin (ng/mL) levels of the groups were compared according to activity score, no significant difference was detected between activity scores 1, 2, 3, 4, 5 and higher, in terms of serum irisin level (45.8 (20.3 to 207.91); 23.5 (4.7 to 418.5); 167.8 (22.0 to 471.1), 51.4 (0.6 to 317.9); 67.0 (11.0 to 441.0), respectively, p>0.05 in each).
Correlation analysis
There was a negative correlation between serum irisin level and cIMT (r=−0.511, p<0.001) (figure 2), insulin (r=−0.332, p=0.021), HOMA-IR (r=−0.371, p=0.009) and creatinine (r=−0.394, p=0.006). A positive correlation was detected between cIMT and insulin (r=0.434, p=0.002), HOMA-IR (r=0.527, p<0.001), age (r=0.436, p=0.002) and creatinine (r=0.419, p=0.003) level. Table 3 shows all correlation results for irisin and cIMT. In addition, a strong positive correlation was present between HOMA-IR and BMI (r=0.362, p=0.012).
Correlation analysis of other factors in Behçet's disease patients with cIMT
ALT, alanine aminotransferase; CRP, C reactive protein; ESR, erythrocyte sedimentation rate; Hb, hemoglobin; HDL-C, high-density lipoprotein cholesterol; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride; TSH, thyroid-stimulating hormone; WBC, white cell counts.
Correlation between serum irisin and carotid intima-media thickness.
Regression analysis
Logistic regression analysis was performed using a cIMT-dependent variable. According to results of the logistic regression analysis, decreased serum irisin level (OR 0.996, 95% CI 0.992 to 1.000, p=0.041), male gender (OR 7.634, 95% CI 1.415 to 41.191, p=0.018), and HOMA-IR (OR 2.596, 95% CI 1.451 to 4.643, p=0.001) were independent risk factors of cIMT in patients with BD. Results of logistic regression analysis can be seen in table 4.
Multiple logistic regression analysis for cIMT in patients with Behçet's disease
BMI, body mass index; cIMT, carotid intima-media thickness; CRP, C reactive protein; HDL-C, high-density lipoprotein cholesterol; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; LDL-C, low-density lipoprotein cholesterol; TC, total cholesterol; TG, triglyceride.
Subgroup analysis
In subgroup analysis, cIMT of the patients with BD was higher in males when compared with that in females (p=0.001), while serum irisin level was lower in males than in females (p=0.023). cIMT was higher and irisin level was very low in patients with BD with IR when compared to patients with BD without IR. In subgroup analysis of healthy control subjects with regard to gender and IR, serum irisin levels were found to be similar in both genders, and among the patients with and without IR. Results of subgroup analysis are shown in table 5.
Subgroup analysis according to HOMA-IR, BMI and gender
Median (range).
BMI, body mass index; cIMT, carotid intima-media thickness; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance.
Furthermore, when subgroup analysis was performed according to HOMA-IR values, groups were as follows: group 1 (BMI <25 kg/m2 plus HOMA-IR <2.7 in BD), group 2 (BMI <25 kg/m2 plus HOMA-IR ≥2.7 in BD), group 3 (BMI ≥25 kg/m2 plus HOMA-IR <2.7 in BD), group 4 (BMI ≥25 kg/m2 plus HOMA-IR ≥2.7 in BD), group 5 (BMI <25 kg/m2 plus HOMA-IR <2.7 in control), group 6 (BMI <25 kg/m2 plus HOMA-IR ≥2.7 in control), group 7 (BMI ≥25 kg/m2 plus HOMA-IR <2.7 in control), group 8 (BMI ≥25 kg/m2 plus HOMA-IR ≥2.7 in control). There was a significant difference between group 4 and group 1 (p=0.025), group 5 (p=0.003) and group 7 (p=0.002), group 2 and group 5 (p=0.009), and group 7 (p=0.007). There was no significant difference between the other group pairs not mentioned above (all p>0.05). Analysis results of all subgroups can be seen in table 5.
Discussion
Irisin is an exercise-induced myokine that is released from the liver, kidney, heart, skeletal muscles and skin. It has an active role in energy metabolism.9 17 A low serum irisin level is related to many adverse conditions. Several studies have reported that, in patients with prediabetes and type II diabetes, irisin levels were lower than normal, and this was found to be related to hyperglycemia and hyperlipidemia.18 19 Low serum irisin levels play an important role in insulin sensitivity and decreased insulin secretion, development of IR and glucose-fatty acid metabolism disorders.20 Serum irisin levels were reported to be lower in diabetic patients with macrovascular complications compared to non-diabetic patients without macrovascular complication.21 Hyperglycemia, advanced-glycosylated end products, and IR accelerate the atherosclerotic process by causing direct endothelial dysfunction, augmenting cytokine release from endothelium, increasing lipid peroxidation and causing oxidative damage.22 Hyperglycemia, IR and chronic inflammation cause endothelial dysfunction and atherosclerosis in patients with BD.23 24 In our study, cIMT, serum CRP, glucose and HOMA-IR levels of the patients with BD were significantly higher when compared to those in the control group, while the serum irisin level was found to be very much lower in the BD group than in the healthy control group. There was positive correlation between age, serum creatinine, insulin, HOMA-IR and cIMT, while a negative correlation was detected between cIMT and serum irisin level. In addition, we found a negative correlation between HOMA-IR and irisin level. Our regression analysis results showed that decreased irisin level was a strong and predictive for HOMA-IR and cIMT. In the literature, Lee et al11 detected significantly lower levels of serum irisin in patients receiving peritoneal dialysis when compared to those of healthy individuals, and found a strong correlation between serum irisin level and cIMT. However, it was stated in their study that irisin level might be associated with reduced muscle mass. We hypothesized that, since low irisin level may strongly related to IR and hyperglycemia, low serum irisin level should obviously be related to accelerated atherosclerosis. Therefore, we might find a strong association between low irisin levels and cIMT. Unlike our study and other studies showing that a low level of serum irisin is related to IR, some studies state that increased irisin levels are associated with IR and cIMT.12 25 It was reported that an increased irisin level prevented endothelial dysfunction in diabetic patients by decreasing the formation of peroxynitrite and superoxide radicals.8 Sesti et al25 reported that a high serum irisin level was related to cIMT; however, the patient group in their study consisted completely of healthy subjects with no cardiac disease and no diabetes.
The subgroup analysis can provide information about the relationship between low irisin levels and IR. In our study, serum irisin level was lower in obese and overweight patients with BD with IR, and in lean patients with BD with IR, compared to obese and overweight BD patients without IR, and to lean patients with BD without IR. In addition, serum irisin level was very low in obese and overweight controls with IR, and in lean controls with IR when compared to obese and overweight controls without IR, and lean controls without IR. We found low serum irisin levels in both, patients with BD and healthy individuals; particularly in patients with IR other than obesity. Irisin is mainly secreted from myocytes and it may not be a good indicator for the amount of the BMI for muscle tissue since BMI will be high in individuals with high fat tissue even though they have a low amount of muscle tissue. According to our hypothesis, there is a strong relationship between IR and low irisin levels, and decreased irisin levels may lead to atherosclerosis via increasing IR. This could be the case because we found that IR was a strong independent risk factor for cIMT according to the regression analysis results.
Male gender is a risk factor for disease severity and atherosclerotic cardiac diseases in BD.26 Many studies have revealed no correlation between genders, age and serum irisin level.27 On the other hand, Anastasilakis et al28 reported that serum irisin level was significantly lower in lean males than in lean females. In regression analysis of our study, male gender was found to be an independent risk factor for cIMT. We also found that the serum irisin level was lower but cIMT was higher in male patients with BD when compared to female patients. Some studies reported a positive correlation between LDL, TC and irisin,20 while the findings of other studies have not supported this. In the study by Wen et al,29 a high level of correlation was detected between low serum irisin levels and low HDL level in patients with chronic renal failure. In our study, we did not find any correlation between lipid panel and serum irisin level.
In contrast to a few studies,30 Liu et al31 found a positive correlation between glomerular filtration rate and low serum irisin level in patients with type II diabetes, and the authors reported that progressive decrease in muscle mass and increased level of uremic toxins might decrease irisin release in chronic renal failure. In our study, we also detected a negative correlation between serum irisin level and creatinine level. Moreover, there was a strong positive correlation between creatinine level and cIMT. Renal involvement is known to be common in patients with BD, and glomerulonephritis, amyloidosis, renal vascular involvement and interstitial nephritis may develop.32 In patients with BD, a low irisin level might increase renal damage by causing hyperglycemia and IR.
Unlike our study, some studies, interestingly, showed cardioprotective effects of low serum irisin levels. Some studies have reported that a low serum irisin level plays a protective role in acute coronary events such as myocardial infarction by decelerating energy metabolism of myocardial tissue, and decreasing tissue oxygenation and ATP expenditure.33 Kuloglu et al34 reported that high irisin levels cause ATP loss during uncoupled biochemical reactions. However, low irisin levels are known to be cardioprotective. In another study, it was reported that the level of irisin was lower during the first 2 days of myocardial infarction and it increased after the 72nd hour of infarction.35
Limitations of the study
First, we had a relatively small sample size. Second, disease activity indexes of patients with BD were comparatively low and it is possible that the correlation between the high disease activity index, low number of patients and serum irisin levels may have influenced our results. Third, duration of the BD disease and active arthritis patient numbers were, relatively, not high. An increment in disease duration can affect cIMT. Furthermore, duration of the disease and arthritis affected the exercise status of patients, and thus irisin levels could be influenced. There should be further broader studies regarding this issue.
Conclusion
Irisin is an adipocytokine that is related to energy metabolism, insulin sensitivity and release. We found that serum irisin levels in patients with BD had a negative relationship with cIMT, which is a well-known indicator for IR and subclinical atherosclerosis. It is possible that low irisin levels in patients with BD can be associated with atherosclerosis. Since our study is the first to evaluate serum irisin and cIMT in patients with BD, further studies are required in this field.