Irisin, a novel myokine is an independent predictor for sarcopenia and carotid atherosclerosis in dialysis patients

Atherosclerosis 242 (2015) 476e482

Contents lists available at ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

Irisin, a novel myokine is an independent predictor for sarcopenia and carotid atherosclerosis in dialysis patients Mi Jung Lee a, 1, Sul A Lee a, 1, Bo Young Nam b, Sungha Park a, c, Sang-Hak Lee a, c, Han Jak Ryu a, Young Eun Kwon a, Yung Ly Kim a, Kyoung Sook Park a, Hyung Jung Oh a, Jung Tak Park a, Seung Hyeok Han a, Dong-Ryeol Ryu d, Shin-Wook Kang a, b, Tae-Hyun Yoo a, b, * a

Department of Internal Medicine, Yonsei University, College of Medicine, Seoul, Republic of Korea Severance Biomedical Science Institute, Brain Korea 21 PLUS, Yonsei University, Seoul, Republic of Korea c Cardiovascular Genome Center, Yonsei University Health System, Seoul, Republic of Korea d Department of Internal Medicine, School of Medicine, Ewha Womans University, Seoul, Republic of Korea b

a r t i c l e i n f o

a b s t r a c t s

Article history: Received 20 February 2015 Received in revised form 7 July 2015 Accepted 6 August 2015 Available online 13 August 2015

Objective: In end-stage renal disease, deleterious effect of sarcopenia on cardiovascular disease has been explained mainly by chronic inflammation. However, evidence emerged that skeletal muscles mediate their protective effect against sarcopenia by secreting myokines. Therefore, we sought to investigate the effect of irisin, a recently introduced myokine, on the association between sarcopenia and cardiovascular disease in peritoneal dialysis (PD) patients. Methods: Serum irisin concentrations were assessed by enzyme-linked immunosorbent assay in 102 prevalent PD patients and 35 age- and sex-matched controls. To determine sarcopenia and cardiovascular disease, anthropometric indices including mid-arm muscle circumference (MAMC) and carotid intima-media thickness (cIMT) were measured. Results: Serum irisin concentrations were significantly lower in PD patients than in controls (184.2 ± 88.0 vs. 457.2 ± 105.5 ng/mL, P < 0.001). In PD patients, univariate linear regression analysis showed that serum irisin was positively correlated with MAMC and thigh circumference, but negatively correlated with residual renal function and cIMT. Multivariate analysis revealed that MAMC (per 1 cm increase, B ¼ 8.78, 95% confidence interval [CI] ¼ 0.77e16.79, P ¼ 0.03) had an independent association with serum irisin. In addition, serum irisin was a significant independent predictor for carotid atherosclerosis even after adjustment for high-sensitivity C-reactive protein in PD patients (per 1 g/mL increase, odds ratio ¼ 0.990, 95% CI ¼ 0.982e0.997, P ¼ 0.007). Conclusion: This study demonstrated that serum irisin was significantly associated with sarcopenia and carotid atherosclerosis in PD patients. Additional studies to provide a confirmation and examine possible mechanisms are warranted. © 2015 Elsevier Ireland Ltd. All rights reserved.

Keywords: Carotid atherosclerosis Irisin Myokine Peritoneal dialysis Sarcopenia

1. Introduction Cardiovascular disease is the leading cause of death in end-stage renal disease (ESRD) patients [1]. Protein-energy wasting (PEW) is regarded as an important nontraditional risk factor for

* Corresponding author. Department of Internal Medicine, College of Medicine, Severance Biomedical Science Institute, Brain Korea 21 PLUS, Yonsei University, 134 Shinchon-Dong, Seodaemun-Gu, Seoul 120-752, Republic of Korea. E-mail address: [email protected] (T.-H. Yoo). 1 Both authors contributed to this manuscript equally. http://dx.doi.org/10.1016/j.atherosclerosis.2015.08.002 0021-9150/© 2015 Elsevier Ireland Ltd. All rights reserved.

development of cardiovascular disease in this population [2]. Sarcopenia, reduction in muscle mass, is frequently observed in PEW and is prevalent in chronic kidney disease (CKD) patients [3,4]. In ESRD patients, sarcopenia is significantly associated with greater mortality [4]. Several possible explanations have been proposed for the protective effect of muscle against sarcopenia in ESRD such as association with chronic inflammation [5], nutritional status [6], uremia [2], or insulin resistance [7]. In addition to these factors, skeletal muscles are suggested to mediate their protective effects via secretion of proteins that exert specific endocrine effects [8]. Indeed, skeletal muscles produce and release cytokines or other

M.J. Lee et al. / Atherosclerosis 242 (2015) 476e482

peptides that are classified as myokines [9]. Myokines have been known to stimulate muscle growth and hypertrophy, increase fat oxidation, enhance insulin sensitivity, and induce antiinflammatory activity [8]. Irisin, a novel myokine, has been introduced to drive brown-fat-like conversion of white adipose tissue and stimulate thermogenic genes including uncoupling protein 1 (UCP1) [10]. In mouse models, increased plasma irisin reduced weight gain, glucose intolerance, and insulin resistance under a high-fat diet [10]. Although the exact characteristics of irisin are controversial [11,12], irisin is proposed to be linked, at least in part, to the beneficial effects of skeletal muscle on energy homeostasis and glucose metabolism [10,11]. To date, few studies have investigated irisin in CKD patients [13e15]. Previous studies in CKD patients indicated that irisin concentrations were significantly lower in CKD patients than in control patients [13e15]. Since irisin is secreted from myocytes, it can be speculated that sarcopenia would lead to a decrease in irisin. Therefore, we primarily investigated the association between irisin and sarcopenia. In addition, sarcopenia was an independent risk factor for cardiovascular morbidity and mortality [3,4]. In this regard, we hypothesized that irisin had significant association with sarcopenia and cardiovascular disease. To address these issues, we determined serum irisin concentrations, cardiometabolic risk factors, anthropometric indices, and carotid atherosclerosis by ultrasonography in 102 peritoneal dialysis (PD) patients and 35 healthy controls. To the best of our knowledge, this is the first study to explore the pathophysiologic role of irisin on cardiovascular disease in patients with renal insufficiency. 2. Methods 2.1. Subjects The study population for this study is comprised of participants in a prospective cohort, which included prevalent PD patients in Yonsei University Health System (YUHS), designed to investigate the effect of body composition on cardiovascular risk and mortality in PD patients. In our previous report [16], we included 88 prevalent PD patients between February 2010 and July 2010. Due to limitation of relatively small subject numbers, we maintained and extended this cohort. We excluded patients who were <20 years of age or >80 years and those who were on PD for less than 3 months. Patients were eligible if they had no history of malignancy or chronic inflammatory disease such as systemic lupus erythematosus, and had no overt infection or cardiovascular hospitalization during the 3 months prior to study enrollment. Patients were excluded if they had a history of kidney transplantation or hemodialysis (HD) prior to PD. A total of 102 prevalent PD patients were finally analyzed (Supplementary Figure 1). Thirty-five age- and gender-matched healthy controls were selected from the Cardiovascular Genome Center Registry from YUHS. This study was carried out in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of the YUHS Clinical Trial Center. We obtained informed written consent from all participants. 2.2. Clinical and biochemical data collection Demographic and clinical data collected at study enrollment were: age, gender, diabetes mellitus, previous history of cardiovascular disease, smoking status, physical activity, and medication. Cardiovascular disease was defined as a history of coronary, cerebrovascular, or peripheral vascular disease. Coronary artery disease was defined as a history of angioplasty, coronary artery bypass graft, myocardial infarction or angina. Cerebrovascular disease was

477

defined as previous transient ischemic attack, stroke, or carotid endarterectomy. Peripheral vascular disease was defined as a history of claudication, ischemic limb loss and/or ulceration, or performance of a peripheral revascularization procedure. Blood was taken after a 12-h overnight fast, and the following laboratory data were measured in our hospital laboratory at the time of study enrollment: hemoglobin, glucose, blood urea nitrogen, creatinine, albumin, triglyceride, total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, calcium (Ca), phosphorus (P), and intact parathyroid hormone concentrations. High-sensitivity C-reactive protein (hs-CRP) concentrations were determined by a latex-enhanced immunonephelometric method using a BNII analyzer (Dade Behring, Newark, DE, USA). Residual renal function (RRF) was calculated as the average clearance of urea and creatinine from 24-h urine collection in PD patients. In controls, estimated glomerular filtration rate (eGFR) was calculated by the four-variable Modification of Diet in Renal Disease Study formula. To reflect actual patient conditions, usual overnight peritoneal dialysate volume and glucose concentrations were not changed for this study. Kt/V urea was determined from the total loss of urea nitrogen in spent dialysate using PD Adequest 2.0 for Windows software (Baxter Healthcare, Deerfield, IL, USA). 2.3. Measurement of serum irisin and plasma interleukin-6 At study enrollment, 10 mL of whole blood was taken using serum separation tube (SST) and ethylenediaminetetraacetic acid (EDTA) tube. The aliquots of serum and plasma were stored at a deep freezer (70  C). Serum irisin concentrations were measured from stored serum samples by colorimetric enzyme-linked immunosorbent assay (ELISA) (AG-45A-0046EK-KI01, Adipogen, San Diego, CA, USA). Interassay and intraassay coefficients of variation were less than 15% and less than 10% for irisin. Plasma interleukin-6 (IL-6) was determined by ELISA (Human IL-6 Quantikine ELISA Kit, R&D Systems, Inc., Minneapolis, MN, USA) in 76 PD patients. 2.4. Anthropometric measurement Anthropometric indices were measured in the morning after complete emptying of overnight dialysate at the time of study enrollment. Patients were weighed in light clothing, and height was measured with no shoes. Body mass index (BMI) was calculated as weight/height2 (kg/m2). Triceps skinfold thickness was measured with a conventional skinfold caliper using standard techniques [17]. The mid-arm circumference and thigh circumference was measured with a plastic tape. These anthropometric measurements were obtained by a single skilled nurse in our PD unit and the mean value of the triplicate measurements was used in the analysis. Midarm muscle circumference (MAMC) was calculated as MAMC (cm) ¼ mid-arm circumference (cm)  p  triceps skinfold thickness (cm) [4]. 2.5. Carotid atherosclerosis assessment Carotid intima-media thickness (cIMT) was assessed using a Prosound a10 (Aloka, Tokyo, Japan) by a single trained medical doctor who was blinded to patients' clinical and biochemical data at the time of study enrollment. The cIMT was measured in the prone position with head extended and turned to the opposite direction. Common carotid arteries, carotid bulb, and internal carotid arteries were examined by two different longitudinal projections. The cIMT was measured as the distance between the leading edges of the lumen interface and the media-adventitia interface at 10 mm

478

M.J. Lee et al. / Atherosclerosis 242 (2015) 476e482

proximal to the bifurcation in a plaque-free wall of the right and left common carotid arteries. The cIMT value was expressed as an average of maximal cIMT from both common carotid arteries. Atherosclerotic plaque was defined as an isolated protruding lesion with a thickness greater by at least 50% than neighboring sites. To confirm the intra-reader variability, randomly selected 60 images were reexamined by the same reader. The median intra-class correlation coefficient for cIMT was 0.89 (95% confidence interval [CI] ¼ 0.69e0.99). Carotid atherosclerosis was defined as increased cIMT (>1.0 mm) or presence of a plaque in accordance with other studies [16,18e23].

analysis was performed to identify variables strongly associated with irisin in PD patients (Table 2). In a muscle representatives-only model (model 1), MAMC (per 1 cm increase, B ¼ 9.17, 95% CI ¼ 1.12e17.21, P ¼ 0.03) was independently associated with irisin. Moreover, MAMC (per 1 cm increase, B ¼ 8.78, 95% CI ¼ 0.77e16.79, P ¼ 0.03) had an independent association with serum irisin concentration in fully adjusted model (model 3). However, RRF was not independently associated with irisin in an age- and sex-adjusted model (model 2) or a fully adjusted model (model 3). 3.3. Irisin as an independent parameter associated with carotid atherosclerosis in PD patients

2.6. Statistical analysis Continuous variables were expressed as mean ± standard deviation or median (interquartile range) and categorical variables as a number (percentage). Normality of distribution was ascertained by normal QeQ plots (Supplementary Figure 2) and the ShapiroeWilk test. To compare baseline characteristics, Student's t tests or ManneWhitney U-tests were used for continuous variables and chi square tests were used for categorical variables. Linear regression analysis determined significant factors associated with serum irisin concentrations in PD patients. Significant variables in univariate analysis were categorized into two groups and included in multivariate analysis: (1) variables representing muscle mass (MAMC and thigh circumference) and (2) RRF. Patients were divided into two groups according to the presence of carotid atherosclerosis. Logistic regression models were constructed to confirm the independent association of irisin with carotid atherosclerosis. We performed incremental levels of multivariate adjustment, which included these variables that were significant in univariate analysis: (1) demographic variables of age, sex, diabetes mellitus, previous history of cardiovascular disease, and BMI; (2) biochemical parameters of HDL cholesterol and Ca  P product and variables associated with dialysis of PD duration and RRF; and (3) hs-CRP to evaluate the effect of irisin, independently of inflammation. Linear regression analysis was also performed including cIMT as a continuous variable. In addition, we tested whether irisin had additive value to clinical variables for predicting carotid atherosclerosis by comparing c-statistics corresponding to area under the curve (AUC) of the receiver operating characteristic (ROC) curve. Statistical analysis was performed using SPSS for Windows version 18.0 (SPSS Inc., Chicago, IL, USA) and Stata version 11.0 (StataCorp., College station, TX, USA). A P value less than 0.05 was considered statistically significant. 3. Results

Among 102 PD patients, 37 patients (36.3%) had carotid atherosclerosis. Baseline characteristics according to the presence of carotid atherosclerosis are in Table 3. The mean age, BMI, and hsCRP were significantly higher in patients with carotid atherosclerosis, while the mean concentrations of irisin, HDL cholesterol, Ca  P product, and RRF were lower in the carotid atherosclerosis group. More patients with carotid atherosclerosis had a previous history of cardiovascular disease and diabetes mellitus. However, no significant differences were observed in smoking status, use of lipid-lowering therapy, systolic blood pressure, MAMC, or LDL cholesterol between the two groups. In univariate logistic regression analysis, age, diabetes mellitus, previous history of cardiovascular disease, BMI, Ca  P product, HDL cholesterol, log hs-CRP, and serum irisin concentrations (OR [odds ratio] ¼ 0.995, 95% CI ¼ 0.990e0.999) were significantly associated with carotid atherosclerosis. After adjustment for demographic parameters, biochemical parameters, and variables associated with dialysis, serum irisin exhibited an independent relationship with carotid atherosclerosis (model 2 in Table 4). When hs-CRP was included in multivariate analysis, irisin remained an independent parameter associated with carotid atherosclerosis in PD patients (model 3 in Table 4). Similar result was observed after adjustment for IL-6 (model 2 in Table 4 þ IL-6, per 1 ng/mL increase, OR ¼ 0.988, 95% CI ¼ 0.977e0.999, P ¼ 0.034). Furthermore, the independent association of irisin with carotid atherosclerosis was still consistent using different cutoff value [24] (cIMT >0.7 mm, per 1 ng/mL increase, OR ¼ 0.992, 95% CI ¼ 0.985e0.999, P ¼ 0.04) or continuous variable (per 1 ng/mL increase, B ¼ 0.002, 95% CI ¼ 0.003 to 0.001, P ¼ 0.001) (Supplementary Table 2). In addition, we determined the AUC for predicting carotid atherosclerosis from regression models (Fig. 1). The AUC of model 1, which included baseline demographics only, was 0.812 (95% CI ¼ 0.727e0.898). A full model that included serum irisin showed the highest model discrimination (model 3 in Fig. 1, AUC ¼ 0.900, 95% CI ¼ 0.843e0.957, P ¼ 0.04).

3.1. Baseline characteristics of subjects 4. Discussion Baseline characteristics of PD patients and controls are in Table 1. The mean age was 54.1 ± 11.6 years in PD patients and 54.1 ± 11.5 years in controls. Sixty patients (58.8%) were men among PD patients and 20 patients (57.1%) were men among controls. Compared with controls, PD patients showed significantly lower serum irisin concentrations (PD vs. controls, 184.2 ± 88.0 vs. 457.2 ± 105.5 ng/mL, P < 0.001) (Supplementary Figure 3). 3.2. Association of clinical and biochemical variables with serum irisin in PD patients In PD patients, serum irisin was positively associated with MAMC, thigh circumference, and lean mass index by bioelectrical impedance analysis, but negatively associated with RRF and the cIMT (Supplementary Table 1). Multivariate linear regression

In the present study, irisin, a novel myokine, was significantly lower in PD patients compared with controls. Moreover, we demonstrated for the first time that irisin was an independent parameter associated with sarcopenia and carotid atherosclerosis in PD patients. Of note, irisin remained an independent parameter associated with carotid atherosclerosis even after adjustment for hs-CRP, suggesting that irisin had a significant association with cardiovascular disease, at least in part, independently of inflammation in PD patients. Mounting evidence indicates that skeletal muscle is a secretory organ that produces and releases cytokines or other peptides which exert specific endocrine effects, myokines [8,9]. Irisin is a recently introduced myokine that drives brown-fat-like conversion of white adipose tissue [10]. Circulating irisin has been proposed to mediate

M.J. Lee et al. / Atherosclerosis 242 (2015) 476e482

479

Table 1 Baseline characteristics of subjects.

Age (years) Male, n (%) Diabetes mellitus, n (%) Cardiovascular disease, n (%) Smokers, n (%) Physical activity, n (%) Sedentary Regular exercise Lipid-lowering therapy, n (%) Antihypertensive therapy, n (%) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Body mass index (kg/m2) Mid-arm circumference (cm) Triceps skinfold thickness (cm) MAMC (cm) Thigh circumference (cm) Hemoglobin (g/L) Glucose (mmol/L) Blood urea nitrogen (mmol/L) Creatinine (mmol/L) Albumin (g/L) Triglyceride (mmol/L) Total cholesterol (mmol/L) LDL cholesterol (mmol/L) HDL cholesterol (mmol/L) Ca  P product (mmol2/L2) iPTH (ng/L) hs-CRP (mg/L) Interleukin-6 (pg/mL)a Irisin (ng/mL) cIMT (mm) eGFR (mL/min/1.73 m2) RRF (mL/min/1.73 m2) Duration of PD (months) Total Kt/V urea (per week) nPCR (g/kg/day)

PD (n ¼ 102)

Controls (n ¼ 35)

P

54.1 ± 11.6 60 (58.8) 28 (27.5) 15 (14.7) 39 (38.2)

54.1 ± 11.5 20 (57.1) 0 (0.0) 0 (0.0) 14 (40.0)

0.82 0.9 0.02 0.01 0.84 <0.001

88 (86.3) 14 (13.7) 39 (38.2) 99 (97.1) 133.9 ± 22.2 76.6 ± 12.3 22.9 ± 2.8 28.4 ± 2.9 1.5 ± 0.7 23.8 ± 2.7 49.8 ± 5.1 105 ± 13 5.3 ± 0.2 20.1 ± 5.8 928 ± 115 35 ± 5 1.5 ± 0.9 4.4 ± 1.0 2.4 ± 0.7 1.1 ± 0.3 3.5 ± 0.1 165.7 (74.3e323.1) 1.03 (0.60e2.13) 7.1 (4.0e11.4) 184.2 ± 88.0 0.85 ± 0.46 e 3.6 ± 3.7 27.5 (12.0e92.5) 2.1 ± 0.4 0.98 ± 0.19

16 (45.7) 19 (54.3) 0 (0.0) 0 (0.0) 127.7 ± 15.7 83.6 ± 11.9 25.4 ± 4.2 29.3 ± 4.3 1.4 ± 0.6 24.1 ± 4.2 e 142 ± 16 5.3 ± 0.7 5.1 ± 1.3 80 ± 9 44 ± 3 1.4 ± 0.9 5.0 ± 1.0 3.1 ± 0.9 1.3 ± 0.3 2.7 ± 0.4 e 0.85 (0.40e1.65) e 457.2 ± 105.5 0.70 ± 0.16 86.3 ± 14.5 e e e e

<0.001 <0.001 0.17 0.01 <0.001 0.22 0.09 0.9 e <0.001 0.028 <0.001 <0.001 <0.001 0.9 <0.001 <0.001 <0.001 <0.001 e 0.04 e <0.001 0.04 e e e e e

Note: Data are expressed as mean ± standard deviation, median (interquartile range), or number of patients (percent). ManneWhitney U-tests were used for continuous variables and chi square tests were used for categorical variables. Abbreviations: Ca, calcium; cIMT, carotid intima-media thickness; eGFR, estimated glomerular filtration rate; hs-CRP, high-sensitivity C-reactive protein; iPTH, intact parathyroid hormone; Kt/V urea, fractional urea clearance; MAMC, mid-arm muscle circumference; nPCR, normalized protein catabolic rate; P, phosphorus; PD, peritoneal dialysis; RRF, residual renal function. a Interleukin-6 was measured in 76 PD patients.

the beneficial effects of physical activity on metabolism including the browning of white adipose tissue and increasing energy expenditure by upregulating UCP1, resulting in reduced weight gain, glucose intolerance, and insulin resistance [10]. Recently, the pathophysiologic role of irisin has been actively investigated in humans, but the nature of irisin remains largely unknown [13e15,25e29]. Although CKD leads to metabolic abnormalities such as impairment of glucose and protein metabolism, increased oxidative stress, and inflammation [30], the role of irisin in CKD

patients is not fully understood because of limited data [13e15]. In previous studies, circulating irisin was commonly reported to be lower in CKD patients compared to the general population [13e15]. In our study, serum irisin was found to be significantly lower in PD patients than controls in accordance with previous studies [13e15]. So far, the underlying mechanism for decreased irisin in CKD patients is not fully elucidated. Previous studies proposed that uremia can be a possible candidate. Wen et al. [14] showed that indoxyl sulfate, a well-known uremic toxin, decreased FNDC5 expression in

Table 2 Independent association of clinical and biochemical variables with serum irisin in PD patients. Model 1a

Age (per 1 year increase) Female (vs. male) MAMC (per 1 cm increase) Thigh circumference (per 1 cm increase) RRF (per 1 mL/min/1.73 m2 increase)

Model 2b

Model 3c

B (95% CI)

P

B (95% CI)

P

B (95% CI)

P

0.50 (1.97e0.97) 7.92 (30.12e45.96) 9.17 (1.12e17.21) 1.85 (2.11e5.82) e

0.50 0.68 0.03 0.36

0.56 (2.12e1.01) 13.72 (48.80e21.36) e e 5.27 (11.52e0.99)

0.48 0.48

0.16 (1.69e1.36) 6.25 (31.61e44.10) 8.78 (0.77e16.79) 1.87 (2.07e5.80) 4.59 (10.60e1.41)

0.83 0.74 0.03 0.35 0.13

0.1

Note: Multivariate linear regression analyses were performed. Abbreviations: MAMC, mid-arm muscle circumference; PD, peritoneal dialysis; RRF, residual renal function; CI, confidence interval. a Model 1: adjusted for age, sex, MAMC and thigh circumference (adjusted R2 ¼ 0.358, P ¼ 0.009). b Model 2: adjusted for age, sex, and RRF (adjusted R2 ¼ 0.225, P ¼ 0.163). c Model 3: adjusted for age, sex, MAMC, thigh circumference, and RRF (adjusted R2 ¼ 0.385, P ¼ 0.008).

480

M.J. Lee et al. / Atherosclerosis 242 (2015) 476e482

Table 3 Baseline characteristics of PD patients with and without carotid atherosclerosis. Characteristics

Age (years) Male, n (%) Diabetes mellitus, n (%) Cardiovascular disease, n (%) Duration of PD (months) Smokers, n (%) Regular exercise, n (%) Lipid-lowering therapy, n (%) Total Kt/V urea (per week) RRF (mL/min/1.73 m2) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Body mass index (kg/m2) Mid-arm circumference (cm) Triceps skinfold thickness (cm) MAMC (cm) Thigh circumference (cm) nPCR (g/kg/day) Hemoglobin (g/L) Glucose (mmol/L) Blood urea nitrogen (mmol/L) Creatinine (mmol/L) Albumin (g/L) Triglyceride (mmol/L) Total cholesterol (mmol/L) LDL cholesterol (mmol/L) HDL cholesterol (mmol/L) Ca  P product (mmol2/L2) iPTH (ng/L) hs-CRP (mg/L) Irisin (ng/mL) cIMT (mm)

With carotid atherosclerosis

Without carotid atherosclerosis

(n ¼ 37)

(n ¼ 65)

60.4 ± 10.2 25 (67.6) 18 (48.6) 10 (27.0) 21.0 (12.0e54.5) 16 (43.2) 2 (5.4) 15 (40.5) 2.1 ± 0.4 3.0 ± 3.4 134.1 ± 20.3 74.5 ± 11.3 24.0 ± 2.9 29.1 ± 3.0 1.6 ± 0.7 23.6 ± 2.6 49.7 ± 6.0 0.96 ± 0.19 107 ± 12 5.4 ± 0.2 19.5 ± 5.8 893 ± 345 34 ± 4 1.5 ± 0.9 4.4 ± 1.1 2.4 ± 0.8 1.0 ± 0.3 3.1 ± 0.9 146.3 (79.2e271.1) 1.8 (0.78e5.11) 160.2 ± 95.2 1.22 ± 0.56

50.6 ± 11.0 35 (53.8) 10 (15.4) 5 (7.7) 30.0 (12.0e96.0) 23 (35.4) 12 (18.5) 24 (36.9) 2.1 ± 0.4 4.6 ± 4.2 133.8 ± 23.3 77.8 ± 12.8 22.3 ± 2.5 28.0 ± 2.9 1.4 ± 0.6 24.2 ± 3.0 50.0 ± 4.5 0.99 ± 0.19 103 ± 13 5.3 ± 0.2 20.5 ± 5.8 972 ± 354 35 ± 5 1.4 ± 0.9 4.4 ± 0.9 2.3 ± 0.7 1.1 ± 0.3 3.7 ± 0.9 174.0 (73.5e358.7) 0.85 (0.55e1.37) 197.8 ± 81.2 0.64 ± 0.20

P

<0.001 0.21 <0.001 0.02 0.28 0.53 0.24 0.83 0.9 0.03 0.9 0.20 0.003 0.07 0.21 0.30 0.78 0.50 0.11 0.68 0.42 0.07 0.34 0.56 0.9 0.34 0.004 0.004 0.12 0.003 0.03 <0.001

Note: Data are expressed as mean ± standard deviation, median (interquartile range), or number of patients (percent). Student's t tests or ManneWhitney U-tests were used for continuous variables and chi-square tests were used for categorical variables. Abbreviations: Ca, calcium; cIMT, carotid intima-media thickness; hs-CRP, high-sensitivity C-reactive protein; iPTH, intact parathyroid hormone; Kt/V urea, fractional urea clearance; MAMC, mid-arm muscle circumference; nPCR, normalized protein catabolic rate; P, phosphorus; PD, peritoneal dialysis; RRF, residual renal function.

skeletal muscle cells and irisin concentrations in cell culture medium. In the study of Ebert et al. [15], serum irisin concentrations were significantly associated with GFR. In the present study, eGFR was positively associated with serum irisin concentrations in controls (data not shown). On the contrary, RRF was negatively associated with irisin in PD patients. When patients were divided into higher and lower RRF group, total Kt/V urea per week (the sum of renal Kt/V and dialysis Kt/V) was greater in the higher RRF group than the lower RRF group (2.3 ± 0.5 vs. 1.9 ± 0.3, P < 0.001). Because we enrolled clinically stable patients who were undergoing adequate dialysis, we presumed that the degree of uremia might

Table 4 Association of serum irisin with carotid atherosclerosis in PD patients. Irisin (per 1 ng/mL increase)

Crude Model 1a Model 2b Model 3c

OR

95% CI

P

0.995 0.992 0.989 0.990

0.990e0.999 0.986e0.998 0.982e0.996 0.982e0.997

0.04 0.01 0.003 0.007

Note: Multivariate logistic regression analyses were performed. Abbreviations: BMI, body mass index; Ca, calcium; hs-CRP, high-sensitivity C-reactive protein; MAMC, mid-arm muscle circumference; P, phosphorus; PD, peritoneal dialysis; RRF, residual renal function; OR, odds ratio; CI, confidence interval. a Model 1: adjusted for age, sex, diabetes mellitus, previous history of cardiovascular disease, and BMI. b Model 2: model 1 þ biochemical parameters (HDL cholesterol and Ca  P product) þ variables associated with dialysis (dialysis duration and RRF). c Model 3: model 2 þ log hs-CRP.

Fig. 1. Receiver operating characteristic (ROC) curves for multivariate regression models to predict carotid atherosclerosis. Area under the curve (AUC) of ROC curve was compared. Irisin had additive value to clinical variables in the prediction of carotid atherosclerosis in PD patients (P ¼ 0.04). Model 1: adjusted for age, sex, diabetes mellitus, and previous cardiovascular disease, and BMI; Model 2: model 1 þ biochemical parameters (HDL cholesterol and Ca  P product) þ variables associated with dialysis (dialysis duration and residual renal function) þ inflammatory marker (log hs-CRP); Model 3: model 2 þ serum irisin. Abbreviations: BMI, body mass index; Ca, calcium; HDL, high-density lipoprotein; hs-CRP, high-sensitivity C-reactive protein; MAMC, mid-arm muscle circumference; P, phosphorus; PD, peritoneal dialysis.

M.J. Lee et al. / Atherosclerosis 242 (2015) 476e482

not have been substantially different among our PD participants. In this regard, differences in the metabolism of irisin via kidney, peritoneal membrane, or PD effluent fluid could be the reason for this discrepant finding. Ebert et al. [15] mentioned a preliminary result regarding decreased irisin levels after HD as compared to predialysis concentrations, suggesting that irisin is at least in part dialyzable. The effect of dialysis might explain the negative association between RRF and serum irisin levels. However, to date, there has been no data investigating irisin levels in PD patients and the metabolism of irisin in these patients. This issue needs to be explored in future study. In addition, we hypothesized that serum irisin levels were mainly affected by muscle mass, because irisin is released from skeletal muscle. Indeed, serum irisin levels were independently associated with MAMC, a surrogate of sarcopenia. Furthermore, lean mass index had a positive correlation with serum irisin concentrations, supporting the significant association between irisin and muscle. We speculated that muscle wasting and weakness, which are prevalent in PD patients [31], resulted in decreased irisin levels. However, there was no statistical difference in MAMC between controls and PD patients. Since we cannot include sufficient number of control subjects, type II error cannot be excluded when comparing MAMC due to limited subject numbers. Unfortunately, other objective data to estimate muscle mass were not available in controls. Nevertheless, MAMC is a validated marker for diagnosis of sarcopenia in dialysis patients [3,4]. Therefore, MAMC values in controls are unlikely to alter our findings. Taken together, decreased irisin levels can be explained, at least in part, by uremia, the effect of dialysis, and sarcopenia in PD patients. Meanwhile, studies on the association between irisin and metabolic risk factors showed conflicting results in CKD patients. Irisin was positively associated with HDL cholesterol in a study by Wen et al. [14], but only HOMA-IR was an independent factor in a study of Ebert et al. [15]. In our study, serum irisin did not show a significant association with facets of metabolic syndrome including blood pressure, fasting glucose, or lipid profiles in PD patients. Even though the exact reason is not clear, the disparity among studies can be partly explained by different study populations and sizes. Differences in renal replacement therapy modality (nondialysis, HD, or PD) and proportion of lipid-lowering therapy user and diabetes mellitus might contribute to discrepancies. Different assay methods for measuring irisin might be another possible reason for discrepancies. Taken together, further studies with larger, homogenous study populations and a standardized assay method are required to determine the association between irisin and metabolic risk factors in CKD patients. Another main finding of the study was that decreased irisin concentrations were significantly associated with carotid atherosclerosis in PD patients. Since sarcopenia is an established cardiovascular risk factor in PD patients, we inferred that irisin might have a significant association with sarcopenia and cardiovascular disease. Malnutrition and chronic inflammation have been proposed to explain the significant association between sarcopenia and cardiovascular disease [4e6]. Lower muscle mass might reflect poor nutritional status [6] and higher degrees of unopposed inflammation [5]. In addition to these factors, a myokine could be a contributor. In this study, irisin remained significantly associated with carotid atherosclerosis even after adjustment for hs-CRP or IL6, suggesting that the effect of irisin on cardiovascular disease was independent of inflammation. However, in contrast to our study, Sesti et al. [32] demonstrated that irisin was positively associated with cIMT in 192 White nondiabetic general population. We speculate that the discrepant finding regarding the association of irisin with carotid atherosclerosis may be due to different characteristics of study population. Since our subjects were prevalent PD patients with median dialysis duration of 27.5 months, differences

481

in metabolism of irisin can alter irisin concentrations. Moreover, the mean BMI was 29.6 ± 6.1 kg/m2 in a study of Sesti et al., so most patients were overweight or obese. In their study, irisin was positively associated with BMI and postprandial glucose levels. Based on these results, they suggested increased irisin could be compensatory increase to overcome underlying irisin resistance in response to deterioration of insulin sensitivity. In our study, the mean BMI was 22.9 ± 2.8 kg/m2 and irisin was not significantly associated with BMI, abnormal lipid profile, or diabetes mellitus (Supplementary Table 1). Although it is unknown whether our PD patients have irisin resistance or not, different body composition and metabolic status can be another possible explanations for discrepancy. The mechanism by which irisin mediates the deleterious effect of sarcopenia on cardiovascular disease is still unknown. A possible explanation is interplay with other myokines, adipocytokines, or classical cytokines [33]. Shan et al. [34] demonstrated an important interplay between irisin and myostatin, another myokine. In the Shan et al. experiments, myostatin knockout mice exhibited increased muscle mass and browning of white adipose tissue through activating PGC1a-FNDC5 pathway [34]. Moreover, Lee et al. [35] showed that cold exposure increased circulating irisin and fibroblast growth factor 21 (FGF21), a recently identified brown adipokine, resulting in upregulation of human adipocyte brown fat gene/protein expression and hermogenesis. Furthermore, combined FNDC5/FGF21 exogenous treatment showed additive effects. These results suggest irisin mediates muscle-adipose crosstalk in concert with FGF21. However, no data are available on the irisin receptor system and downstream signaling pathway. Hence, further studies should explore the molecular mechanism to elucidate direct irisin effects. Furthermore, future study regarding interaction between irisin and other myokines/adipocytokines might be helpful to allow better understanding of irisin. This study has several limitations. First, there is a lack of definite cutoff value for carotid atherosclerosis in PD patients. We defined carotid atherosclerosis as increased cIMT and presence of plaque to reflect focal features of carotid arteries in accordance with other studies including the general population [20,23], patients with coronary artery disease [21], non-dialysis-dependent CKD [22], HD [18], as well as PD [16,19]. Furthermore, different cutoff value (cIMT > 0.7 mm) [24] and continuous variables were also tested. Second, because of the cross-sectional study design, we did not affirm preceding relationship between irisin and cardiovascular disease. Unfortunately, we did not perform follow-up carotid artery ultrasound and data of changes in anthropometric indices or irisin levels were not available. A study with follow-up carotid ultrasound and measurement of anthropometric indices and serum irisin is necessary to clarify the association of irisin, sarcopenia, and cardiovascular disease. Therefore, the impact of irisin on future cardiovascular outcomes should be clarified in the future study. In the present study, there was no significant difference in irisin concentrations according to regular exercise. However, concentrations of irisin increased significantly after endurance exercise training in both mice and humans [36]. The effect of the intervention using exercise training on irisin, sarcopenia, and cardiovascular disease might be also worth investigating in PD patients. Third, although not only muscle mass but also muscle strength is an important predictor of clinical outcomes in dialysis patients [37], this study was not included data of muscle strength such as handgrip strength. Finally, since the study participants were all Korean prevalent PD patients, our results might not be generalizable to other populations. Protective role of irisin should be considered cautiously. Despite these limitations, we believe that this study provides useful information on the pathophysiological role of irisin in PD patients.

482

M.J. Lee et al. / Atherosclerosis 242 (2015) 476e482

In conclusion, the present study demonstrates for the first time that irisin is significantly associated with sarcopenia and carotid atherosclerosis in PD patients. Furthermore, significant association between irisin and carotid atherosclerosis was consistent after adjustment for an inflammatory biomarker, suggesting that irisin had a significant association with cardiovascular disease, at least in part, independently of inflammation. To provide a confirmation and examine possible mechanisms, additional studies are warranted. Source of funding This work was supported by the Brain Korea 21 Project for Medical Science, Yonsei University, by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 2014-050098), and by a grant of the Korea Healthcare Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (H14C2003).

[14] [15]

[16]

[17] [18]

[19]

[20]

[21]

Conflict of interest [22]

The authors have no conflicts of interest to declare. Appendix A. Supplementary data

[23]

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.atherosclerosis.2015.08.002.

[24]

References

[25] [26]

[1] R.N. Foley, P.S. Parfrey, M.J. Sarnak, Clinical epidemiology of cardiovascular disease in chronic renal disease, Am. J. Kidney Dis. 32 (1998) S112eS119. [2] C.P. Kovesdy, K. Kalantar-Zadeh, Why is protein-energy wasting associated with mortality in chronic kidney disease? Semin. Nephrol. 29 (2009) 3e14. [3] S. Beddhu, L.M. Pappas, N. Ramkumar, et al., Effects of body size and body composition on survival in hemodialysis patients, J. Am. Soc. Nephrol. 14 (2003) 2366e2372. [4] N. Noori, J.D. Kopple, C.P. Kovesdy, et al., Mid-arm muscle circumference and quality of life and survival in maintenance hemodialysis patients, Clin. J. Am. Soc. Nephrol. 5 (2010) 2258e2268. [5] H. Honda, A.R. Qureshi, J. Axelsson, et al., Obese sarcopenia in patients with end-stage renal disease is associated with inflammation and increased mortality, Am. J. Clin. Nutr. 86 (2007) 633e638. [6] T.A. Ikizler, R.L. Wingard, J. Harvell, et al., Association of morbidity with markers of nutrition and inflammation in chronic hemodialysis patients: a prospective study, Kidney Int. 55 (1999) 1945e1951. [7] K. Shinohara, T. Shoji, M. Emoto, et al., Insulin resistance as an independent predictor of cardiovascular mortality in patients with end-stage renal disease, J. Am. Soc. Nephrol. 13 (2002) 1894e1900. [8] B.K. Pedersen, M.A. Febbraio, Muscles, exercise and obesity: skeletal muscle as a secretory organ, Nat. Rev. Endocrinol. 8 (2012) 457e465. [9] B.K. Pedersen, A. Steensberg, C. Fischer, et al., Searching for the exercise factor: is IL-6 a candidate? J. Muscle Res. Cell Motil. 24 (2003) 113e119. [10] P. Bostrom, J. Wu, M.P. Jedrychowski, et al., A PGC1-alpha-dependent myokine that drives brown-fat-like development of white fat and thermogenesis, Nature 481 (2012) 463e468. [11] F. Villarroya, Irisin, turning up the heat, Cell Metab. 15 (2012) 277e278. [12] F. Sanchis-Gomar, R. Alis, H. Pareja-Galeano, et al., Inconsistency in circulating irisin levels: what is really happening? Horm. Metab. Res. 46 (2014) 591e596. [13] C. Moraes, V.O. Leal, S.M. Marinho, et al., Resistance exercise training does not

[27]

[28] [29] [30] [31] [32]

[33] [34]

[35]

[36] [37]

affect plasma irisin levels of hemodialysis patients, Horm. Metab. Res. 45 (2013) 900e904. M.S. Wen, C.Y. Wang, S.L. Lin, et al., Decrease in irisin in patients with chronic kidney disease, PLoS One 8 (2013) e64025. T. Ebert, D. Focke, D. Petroff, et al., Serum levels of the myokine irisin in relation to metabolic and renal function, Eur. J. Endocrinol. 170 (2014) 501e506. M.J. Lee, D.H. Shin, S.J. Kim, et al., Visceral fat thickness is associated with carotid atherosclerosis in peritoneal dialysis patients, Obesity (Silver Spring) 20 (2012) 1301e1307. E.E. Nelson, C.D. Hong, A.L. Pesce, et al., Anthropometric norms for the dialysis population, Am. J. Kidney Dis. 16 (1990) 32e37. A. Papagianni, M. Kalovoulos, D. Kirmizis, et al., Carotid atherosclerosis is associated with inflammation and endothelial cell adhesion molecules in chronic haemodialysis patients, Nephrol. Dial. Transpl. 18 (2003) 113e119. T. Stompor, A. Krasniak, W. Sulowicz, et al., Changes in common carotid artery intima-media thickness over 1 year in patients on peritoneal dialysis, Nephrol. Dial. Transpl. 20 (2005) 404e412. S.K. Kim, S.W. Park, S.H. Kim, et al., Visceral fat amount is associated with carotid atherosclerosis even in type 2 diabetic men with a normal waist circumference, Int. J. Obes. (Lond) 33 (2009) 131e135. M. Hirano, T. Nakamura, Y. Kitta, et al., Assessment of carotid plaque echolucency in addition to plaque size increases the predictive value of carotid ultrasound for coronary events in patients with coronary artery disease and mild carotid atherosclerosis, Atherosclerosis 211 (2010) 451e455. M.M. Lemos, A.D. Jancikic, F.M. Sanches, et al., Intima-media thickness is associated with inflammation and traditional cardiovascular risk factors in non-dialysis-dependent patients with chronic kidney disease, Nephron Clin. Pract. 115 (2010) c189ec194. A. Simon, J.L. Megnien, G. Chironi, The value of carotid intima-media thickness for predicting cardiovascular risk, Arterioscler. Thromb. Vasc. Biol. 30 (2010) 182e185. M.I. Yilmaz, A.R. Qureshi, J.J. Carrero, et al., Predictors of carotid artery intimamedia thickness in chronic kidney disease and kidney transplant patients without overt cardiovascular disease, Am. J. Nephrol. 31 (2010) 214e221. Y.K. Choi, M.K. Kim, K.H. Bae, et al., Serum irisin levels in new-onset type 2 diabetes, Diabetes Res. Clin. Pract. 100 (2013) 96e101. J.M. Moreno-Navarrete, F. Ortega, M. Serrano, et al., Irisin is expressed and produced by human muscle and adipose tissue in association with obesity and insulin resistance, J. Clin. Endocrinol. Metab. 98 (2013) E769eE778. K.H. Park, L. Zaichenko, M. Brinkoetter, et al., Circulating irisin in relation to insulin resistance and the metabolic syndrome, J. Clin. Endocrinol. Metab. 98 (2013) 4899e4907. S.A. Polyzos, J. Kountouras, A.D. Anastasilakis, et al., Irisin in patients with nonalcoholic fatty liver disease, Metabolism 63 (2014) 207e217. B. Yan, X. Shi, H. Zhang, et al., Association of serum irisin with metabolic syndrome in obese Chinese adults, PLoS One 9 (2014) e94235. B. Stengel, Chronic kidney disease and cancer: a troubling connection, J. Nephrol. 23 (2010) 253e262. N. Clyne, M. Esbjornsson, E. Jansson, et al., Effects of renal failure on skeletal muscle, Nephron 63 (1993) 395e399. G. Sesti, F. Andreozzi, T.V. Fiorentino, et al., High circulating irisin levels are associated with insulin resistance and vascular atherosclerosis in a cohort of nondiabetic adult subjects, Acta Diabetol. 51 (2014) 705e713. S.A. Polyzos, J. Kountouras, K. Shields, et al., Irisin: a renaissance in metabolism? Metabolism 62 (2013) 1037e1044. T. Shan, X. Liang, P. Bi, et al., Myostatin knockout drives browning of white adipose tissue through activating the AMPK-PGC1alpha-Fndc5 pathway in muscle, FASEB J. 27 (2013) 1981e1989. P. Lee, J.D. Linderman, S. Smith, et al., Irisin and FGF21 are cold-induced endocrine activators of brown fat function in humans, Cell Metab. 19 (2014) 302e309. F. Sanchis-Gomar, C. Perez-Quilis, Irisinemia: a novel concept to coin in clinical medicine? Ann. Nutr. Metab. 63 (2013) 60e61. N. Isoyama, A.R. Qureshi, C.M. Avesani, et al., Comparative associations of muscle mass and muscle strength with mortality in dialysis patients, Clin. J. Am. Soc. Nephrol. 9 (2014) 1720e1728.