Study population
Patients who had received hemodialysis treatment for at least 3 months were enrolled from the Blood Purification Center of Dalian Municipal Central Hospital. Patients with any of the following were not eligible for the study: an acute systemic infection; a cardiovascular event in the previous 3 months, including coronary artery disease, myocardial ischemia, cerebrovascular disease and peripheral artery disease; a malignancy; severe hypoalbuminemia (< 2.6 g/dl); < 18 years of age; or declined participation (Fig. 1). Patients were treated 3 times per week (4 hours per session) with a standard bicarbonate dialysate. The blood flow was 200–300 ml/min and the dialysate flow was 500 ml/min. The enrollment for the prospective HD patient cohort was begun in December 2014. The duration of follow-up was 6 years and ended on 31 December 2020. The study was approved by the Ethical Committee of Dalian Municipal Central Hospital. All participants provided written informed consent.
Laboratory measurements
Blood was sampled immediately using 5 ml separating gel accelerator tube and 5 ml EDTA anticoagulation tube before the midweek dialysis treatment by the slow flow/stop pump technique. Blood samples were processed (centrifugal, 3500 rpm, 5 minute) within 30 minutes of sampling to obtain plasma, while serum was allowed to clot for 30 minutes at room temperature prior centrifugation (3500 rpm, 5 minute) and stored at − 80 °C until used in assays. Hemoglobin (Hb) was measured using sodium dodecyl lauryl sulfate. Blood urea nitrogen (BUN), creatinine (Cr), albumin (Alb), alanine aminotransferase (ALT), alkaline phosphatase (ALP), potassium (K), sodium (Na), calcium (Ca), phosphorus (P), and chlorine (Cl) were assayed in an on-site biochemistry laboratory using standard autoanalyzer techniques (Siemens, Tarrytown, New York, Germany). The adequacy of dialysis was calculated by measuring urea clearance (Kt/V) using the standard method [18], as follows: Kt/V = −ln(R-0.008 × t) + (4–3.5 × R) × UF/W, where R is the post−/pre-plasma BUN ratio, t is the dialysis session length (in h), UF is the ultrafiltrate volume (in l), and W is the post-dialysis weight (in kg).
Serum tIS was analyzed and determined by high-performance liquid chromatography (HPLC) [19]. Serum samples were deproteinized by the addition of three parts methanol-to-one part serum. The analyses were performed using an Agilent Technologies 1200 Series HPLC (Agilent, US). Serum tIS was detected at 280 nm and appeared at 5.79 min. The limits of detection of this assay were 0.132 mg/L for tIS. Calibration curves were constructed by plotting the peak areas versus the concentrations of analyte with average R2 values of 0.999 ± 0.001. Intra-day and inter-day coefficients of variation were 0.06 and 0.07 for tIS.
Outcome evaluation
During the follow-up period, the primary outcome for our analysis was all-cause mortality. The secondary outcome was cardiovascular mortality. We adjudicated mortality using information recorded on TSS version 2.0 (Therapy Support Suite, Baden Humboldt, German) and hospital records on BS-EAP (version 5.5; B-soft Enterprise Application Portal, HangZhou, China). Cardiovascular mortality included deaths due to coronary events, sudden cardiac death, heart failure, myocardial ischemia, arrhythmias, and cerebrovascular accidents. Causes for death were reviewed by one independent physician, who was blinded to the tIS levels. For each participant, the time-to-event was calculated as the time from the date of entry into the study until the date of the first studied event (mortality), the date of disenrolling from the study, the date of kidney transplantation, or the study completion date, whichever came first.
Statistical analysis
The normality of all continuous variables was evaluated using the Shapiro-Wilk statistic. The results of continuous variables are expressed as the mean ± standard deviation (SD) or median [quartile1- quartile3], and intergroup comparisons were analyzed using t-tests for normally distributed data or the Mann-Whitney U tests for non-normally distributed data. Categorical variables are expressed as the count with percentage, and differences between the two groups were examined using chi-square tests.
Optimal cut-off points for tIS were determined using X-tile software (version 3.6.1; Yale University School of Medicine, New Haven, CT, USA) [20]. Specifically, the tIS cut-off point was derived from the minimum P values from log-rank 2 statistics for the categorical tIS concentration in terms of survival [21]. Survival curves were generated using the Kaplan-Meier method, and differences between the curves were analyzed using the log-rank test. Colinearity among predictors was examined using variance inflation factors.
Univariate and multivariate Cox proportional hazard regression models were performed to calculate hazards ratios (HRs) and the corresponding 95% confidence intervals (CIs) based on the optimal cut-off point of tIS and tIS as continuous variable for all-cause and CVD mortality. We used the Schoenfeld residual test to verify the assumption of proportional hazards in the Cox analysis, and no violations were found (all P > 0.05). Statistically significant covariates in the univariate model were included in the multivariate model (P < 0.05) with a enter conditional method of analysis, including age, diabetic nephropathy, glomerulus nephritis, hypertension benign renal arteriosclerosis, albumin, urea nitrogen, creatinine, sodium, phosphorus, and chlorine. Restricted cubic spline regression [22] with five knots at the 5th, 35th, 50th, 65th, and 95th centiles were used to estimate the dose-response association between tIS and mortality. Statistical significance was set at a P < 0.05 and based on a two-sided test. All analyses were carried out using SAS (version 9.4; SAS Institute, Inc., Cary, NC, USA).