Gender-specific associations of skeletal muscle mass and arterial stiffness among peritoneal dialysis patients

Decreased skeletal muscle has been identified as a relevant factor for arterial stiffness but has not been thoroughly investigated in peritoneal dialysis (PD) patients. The aim of this study was to investigate the relationship between skeletal muscle and arterial stiffness in PD patients. A cross-sectional study of 658 prevalent PD patients with a mean brachial-ankle pulse wave velocity (baPWV) of 1714 (±501) cm/s and mean skeletal muscle mass of 26.6 (±5.4) kg was performed. Skeletal muscle mass level was significantly higher in males than in females. When examining skeletal muscle mass as a continuous variable, skeletal muscle mass was significantly associated with baPWV in fully adjusted linear regression models in total patients [standardized coefficients (β), −0.181; 95% confidence interval (95% CI), −0.276 to −0.056; P = 0.003] or female patients (β, −0.119; 95% CI, −0.350 to −0.015; P = 0.03) but not in male patients (β, −0.117; 95% CI, −0.300 to 0.011; P = 0.07). Furthermore, in females, a significant association between the middle or highest tertile of skeletal muscle mass and baPWV was found in fully adjusted models (β, −0.123; 95% CI, −0.204 to −0.008; P = 0.03; β, −0.140; 95% CI, −0.228 to −0.016; P = 0.02, respectively). In conclusion, decreased skeletal muscle mass was independently associated with increased baPWV in PD patients, and this association was significant in females but not in males.

patients, it was demonstrated that visceral fat level was an independent predictor of PWV in a cross-sectional study 14 . However, the relationship between skeletal muscle mass and arterial stiffness in PD patients remains unknown. More recently, two studies both reported that low skeletal muscle mass level was an independent risk factor for mortality in PD patients 15,16 , which underlined the importance of skeletal muscle mass.
Physiologically, males have significantly higher skeletal muscle mass levels than females. A previous study including healthy middle-aged to elderly persons found that thigh muscle mass was significantly associated with baPWV in men but not in women 17 . Therefore, we hypothesized that there would be a gender-specific association between skeletal muscle mass and baPWV, after which we performed subgroup analyses to assess whether this association differed in male and female PD patients.

Results
Patient characteristics. In total, 805 continuous ambulatory PD (CAPD) patients underwent a vascular profiler test at our PD center from January 1, 2014, to October 31, 2016, of whom 107 patients were on PD less than 3 months, 5 patients were younger than 18 years, 7 patients were transferred from failed renal transplantation, 15 patients were transferred from permanent hemodialysis, and 13 patients were catheterized in another hospital. The remaining 658 patients were enrolled in this study (Fig. 1). The mean (±SD) age was 46.7 ± 13.9 years, the median PD vintage was 32 months (interquartile range 11 to 57 months), 56.4% of patients were men, and 15.8% of patients were diabetic. The primary cause of ESRD was chronic glomerulonephritis (68.4%), followed by diabetic nephropathy (12.8%) and hypertension (8.4%). The average baPWV value was 1714 cm/s, and the mean (±SD) skeletal muscle mass was 26.6 ± 5.4 kg ( Table 1). All patients received CAPD treatment. Conventional PD solutions (Dianeal 1.5%, 2.5% or 4.25% dextrose; Baxter Healthcare, Guangzhou, China), Y sets, and twin bag systems were used in all PD patients. Physiologic calcium peritoneal dialysate (Ca 2+ concentration = 1.25 mmol/L) was prescribed to all patients in our PD center. The management of the PD patients was implemented based on the guidelines from the International Society for Peritoneal Dialysis and patient characteristics in our PD center 1,18 . As expected, the distribution of skeletal muscle mass significantly differed between genders, with men showing higher skeletal muscle mass than women (see Supplementary Fig. S1, Table 1). In addition, men showed decreasing levels of skeletal muscle mass with age, whereas women did not. Nevertheless, at each 10-year age group, men displayed higher skeletal muscle mass than did women (see Supplementary  Fig. S2). Male patients also had higher high sensitivity C-reactive protein (HsCRP), serum phosphorus, and serum uric acid but lower visceral fat area, calcium, triglyceride, total, low-density lipoprotein, and high-density lipoprotein cholesterol levels compared to female patients. Men also more frequently took calcium channel blockers and phosphate binders compared to women. There was no significant difference between genders in age, body mass index, comorbidity score, PD vintage, and baPWV values ( Table 1).

Correlation of skeletal muscle mass with baPWV. Correlation analyses in all patients indi-
cated that baPWV level was positively correlated with age, diabetes, comorbidity score, PD vintage, neutrophil-to-lymphocyte ratio (N/L), HsCRP, alkaline phosphatase, total cholesterol, low-density lipoprotein cholesterol, triglyceride, blood pressure, and visceral fat area, while negatively correlated with serum albumin, serum prealbumin, and skeletal muscle mass (P < 0.05) ( Table 2). Both male and female patients showed a significant correlation between skeletal muscle mass and baPWV.
To explain the gender difference, sensitivity analyses were performed. We conducted multiple linear regression analyses in each tertile of patients separately. The results showed that skeletal muscle mass was significantly    Supplementary Table S1). Further analysis indicated that 93.3% of patients in the lowest tertile were women.

Discussion
In this cross-sectional study, we disclosed a significant and independent association between decreased skeletal muscle mass and increased baPWV in PD patients. This association was significant in females but not in males, which may be explained by the higher skeletal muscle mass level in males compared to females. The arterial system of ESRD patients undergoes remodeling that is characterized by dilation and arterial intima-media hypertrophy. The alterations of intrinsic properties of arterial wall materials will increase arterial stiffness 19 . Aortic PWV has been recommended as the gold-standard measure of arterial stiffness 20 . Aortic PWV is a powerful independent predictor of all-cause mortality and cardiovascular events in ESRD, hypertensive subjects, the elderly, diabetic subjects and the general population 21 . Previous studies have identified some risk factors associated with aortic PWV in PD patients, such as malnutrition 22 , peritoneal transport status 23 , residual renal function 24 , and inflammation 25 . In the present study, we first showed that skeletal muscle mass was an independent determinant of baPWV in PD patients. This result echoes previous studies in various populations 10-13 and hemodialysis patients 26 . Patients with chronic kidney disease are potentially at greater risk of muscle wasting because of increased urinary protein losses and reduced dietary protein intake 27 , and PD patients may also have protein and amino acid losses in the dialysate, especially in higher peritoneal transporters 28,29 . Recent studies have shown that muscle loss was prevalent in PD patients 30 , while shortage of skeletal muscle was an independent risk factor for mortality in PD patients 15,16 . Skeletal muscle may be viewed as an endocrine organ that releases numerous factors with the potential to influence vascular tone. Greater lean soft tissue mass attenuates age-related increases in arterial stiffness 31,32 .
Of note, gender-specific associations were found between skeletal muscle mass and baPWV in our study. By sensitivity analysis, we believe that lower skeletal muscle mass levels were generally reached by women compared with men, which may explain this gender difference. The gender-specific tertile analysis may suggest a threshold effect in males and females. There are several potential mechanisms why skeletal muscle mass decline may be associated with increased arterial stiffness in PD patients. First, since skeletal muscle mass was well correlated with serum prealbumin, low skeletal muscle mass may reflect a poor nutritional status, which is a predefined risk factor of arterial stiffness 33 . Second, skeletal muscle is the main site for insulin-mediated glucose disposal, and low  skeletal muscle mass may be associated with insulin resistance 34 . Third, decreased skeletal muscle mass will result in low physical activity, which may lead to the promotion of arteriosclerosis 35 .
There are several limitations in the present study. First, this is a single-center study, and thus center-specific effects cannot be excluded. Second, given the cross-sectional nature of our study, we established associations but not causal relationships. Third, skeletal muscle mass of PD patients in this study was measured using BIA, which is considered to be not "ideal" but a feasible method for body composition assessment and is widely used in dialysis patients 36,37 . The usefulness of BIA in measuring body composition was cross-validated against dual-energy X-ray absorptiometry 38,39 , which is regarded as the most practical means of obtaining an accurate assessment of fat-free mass and fat mass in dialysis patients 40 . Finally, the potential co-pathogenesis behind skeletal muscle decline and arterial stiffness was not fully investigated. Thus, the effect of residual confounding cannot be eliminated completely.
In conclusion, our data indicated that skeletal muscle decline was independently associated with increased baPWV in PD patients. Further prospective studies will be needed to investigate the predicative role of skeletal muscle mass in cardiovascular events and mortality in PD patients.

Methods
Study population and data collection. We studied all CAPD patients who performed a vascular profiler test at the PD center of The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China from January 1, 2014 to October 31, 2016. Inclusion criteria were age ≥18years at the time of vascular profiler test and survival for at least 90 days from the first PD therapy. The patients who were catheterized in other hospitals, transferred from permanent hemodialysis (≥3 months) or failed renal transplantation were excluded in this study. The study was conducted in compliance with the ethical principles of the Helsinki Declaration and approved by the Human Ethics Committees of Sun Yat-sen University. Written informed consent was obtained from all participants.
Demographic data included age, gender, body mass index (BMI), primary cause of ESRD and presence of diabetes and CVD. Clinical and biochemical data included PD vintage, Kt/V, residual urine volume, measured glomerular filtration rate (mGFR) calculated from 24 h urine collections and indexed for the body surface area, N/L, hemoglobin, serum albumin, prealbumin, HsCRP, serum lipid profile, serum calcium, phosphorus, intact parathyroid hormone (iPTH), alkaline phosphatase, and uric acid. Medication information was also collected. The comorbidity score was determined according to the Charlson Comorbidity Index, which is one of the most commonly used comorbidity models 41 . Body composition measurement. Body composition was measured by multi-frequency bioelectrical impedance analysis (BIA) InBody 720 (Biospace, Seoul, Korea). InBody 720 uses state-of-the-art technology and an 8-point tactile electrode system that measures the total and segmental impedance and phase angle of alternating electric current at six different frequencies (1 kHz, 5 kHz, 50 kHz, 250 kHz, 500 kHz, and 1000 kHz). The operation of InBody 720 was described in detail in our previous publication 42 . All the body composition data were performed in the instrument using internal software.
Brachial-ankle PWV measurement. The values of baPWV were measured using a noninvasive automatic vascular profiler device BP-203RPE III (Omron, Japan) as previously described 43,44 . Briefly, following 10 minutes of rest in the supine position, four oscillometric cuffs were applied to bilateral brachia and ankles, and electrocardiograph electrodes were placed on bilateral wrists. The transmission distance from the brachium to ankle was calculated according to height of the patient. The path length from the suprasternal notch to the brachium (Lb) was obtained using the following equation: Lb = 0.2195 × height (cm) − 2.0734. The path length from the suprasternal notch to the ankle (La) was obtained using the following equation: La = 0.8129 × height (cm) + 12.328. The baPWV was calculated according to the following formula: baPWV = (La − Lb)/Tba, where Tba was the time interval between the initial increase in brachial and ankle waveforms. The validation of this automatic device and its reproducibility have been previously published 45 . After obtaining bilateral baPWV values, the average value was used for analysis. Heart rate and systolic and diastolic blood pressure were also obtained by the same device. The average values of systolic and diastolic blood pressure of bilateral brachia were used for analysis. Statistical analysis. We compared demographic, clinical and laboratory parameters, body composition data, vascular profiler data, and medications information between male and female PD patients. The results were expressed as frequencies and percentages for categorical variables, means and standard deviations for normally distributed continuous variables, and medians and interquartile ranges for continuous variables not normally distributed. Chi-squared test, independent-samples t-test, and Mann-Whitney U test were used to test for differences in categorical or continuous factors between groups. The correlations between baPWV and other variables were assessed by Pearson correlation test for normally distributed variables or Spearman rank correlation test for non-parametric variables. The associations between skeletal muscle mass and baPWV were examined by multiple linear regression models in total, male, and female patients. Moreover, the interaction between sex and skeletal muscle mass was examined by performing a formal test of interaction. Variables with P < 0.05 in the univariate analysis were picked into multivariate adjusted model according to following principles: presenting clinical relevance and avoiding multicollinearity. Since previous studies have demonstrated that heart rate is an important confounder of PWV 46 , we forced heart rate into models. These covariates of malnutrition, peritoneal transport status, residual renal function, and inflammation were chosen based on their statistical associations with baPWV in univariate model (P < 0.05). Thus, the final full model included age, gender, BMI, heart rate, diabetes, CVD, PD vintage, mean arterial pressure, residual urine volume, HsCRP, serum albumin, total cholesterol, and iPTH. To further assess the different associations with baPWV at different skeletal muscle mass levels, gender-specific tertiles of skeletal muscle mass were also built using the lowest gender-specific skeletal muscle mass tertile as the reference group. The complete case analysis was used to handle missing data. The results were expressed as β and 95% CI. All analyses in the present study were conducted using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). A value of P < 0.05 was considered statistically significant.