Pre-dialysis fluid status, pre-dialysis systolic blood pressure and outcome in prevalent haemodialysis patients: results of an international cohort study on behalf of the MONDO initiative

Pre-dialysis fluid status, pre-dialysis systolic blood pressure and outcome in prevalent... Abstract Background Pre-dialysis fluid overload (FO) associates with mortality and causes elevated pre-dialysis systolic blood pressure (pre-SBP). However, low pre-SBP is associated with increased mortality in haemodialysis patients. The objective of this study was to investigate the interaction between pre-dialysis fluid status (FS) and pre-SBP in association with mortality. Methods We included all patients from the international Monitoring Dialysis Outcome Initiative (MONDO) database with a pre-dialysis multifrequency bioimpedance spectroscopy measurement in the year 2011. We used all parameters available during a 90-day baseline period. All-cause mortality was recorded during 1-year follow-up. Associations with outcome were assessed with Cox models and a smoothing spline Cox analysis. Results We included 8883 patients. In patients with pre-dialysis FO (>+1.1 to +2.5 L), pre-SBP <110 mmHg was associated with an increased risk of death {hazard ratio (HR) 1.52 [95% confidence interval (CI) 1.06–2.17]}. An increased risk of death was also associated with pre-dialysis fluid depletion (FD; <−1.1 L) combined with a pre-SBP <140 mmHg. In normovolemic (NV) patients, low pre-SBP <110 mmHg was associated with better survival [HR 0.46 (95% CI 0.23–0.91)]. Also, post-dialysis FD associated with a survival benefit. Results were similar when inflammation was present. Only high ultrafiltration rate could not explain the higher mortality rates observed. Conclusion The relation between pre-SBP and outcome is dependent on pre-dialysis FS. Low pre-SBP appears to be disadvantageous in patients with FO or FD, but not in NV patients. Post-dialysis FD was found to associate with improved survival. Therefore, we suggest interpreting pre-SBP levels in the context of FS and not as an isolated marker. fluid overload, haemodialysis, pre-dialysis blood pressure, pre-dialysis fluid status, survival INTRODUCTION In chronic haemodialysis (HD) patients, a consistent association between extracellular fluid overload (FO), assessed by multifrequency bioimpedance (MF-BIS), and an increased risk of mortality has been observed [1–6]. This relation is even apparent at mild levels of FO [7]. Recently, we showed that pre-dialysis fluid depletion (FD) was also associated with an increased mortality [8]. In addition, a recent study showed that cumulative FO over a 1-year period was strongly related to outcome [6]. FO has been associated with an increase in systolic blood pressure (SBP) [9–11], whereas correction of FO usually leads to improvement of hypertension [5, 12]. However, with FO assessed by MF-BIS, not all hypertensive patients were found to be fluid overloaded, nor did all patients with FO have an elevated SBP [1, 3, 6, 13]. Several observational studies have observed an increased mortality risk with high (SBP >140–180 mmHg), but especially with low (SBP <110–130 mmHg) pre- and post-dialysis SBP levels [11, 14]. This is in contrast to observations in the general population, where recent results of the Systolic Blood Pressure Intervention Trail (SPRINT) trail showed that an SBP target <120 mmHg is beneficial in patients with increased cardiovascular risk but without diabetes and end-stage kidney disease [15]. The relation between low SBP and increased mortality risk in dialysis patients has been coined as an epidemiological paradox [16, 17]. However, given its strong relation with both outcome as well as SBP levels per se, we hypothesized that fluid status (FS) is an important modifier of the association between mortality and pre-dialysis SBP (pre-SBP) in HD patients (Figure 1). For instance, the fact that low or normal pre-SBP may be physiological when associated with normovolemia (NV), but pathological when associated with FO, as the normal cardiovascular response to FO (i.e. an increase in cardiac output and/or an increase in systemic vascular resistance) may be impaired. A recent study indeed showed that the highest risk of mortality was observed in patients with FO with low (<130 mmHg) pre- SBP [6]. Other potentially relevant factors in this context include systemic inflammation, which was shown to be a strong amplifier of the FO–mortality relationship, as well as ultrafiltration rate, which bears a strong relation with both outcome as well as pre-dialysis FO [18]. FIGURE 1: View largeDownload slide Associations of fluid status and blood pressure with death. *not investigated in this study. UF-rate, ultrafiltration rate. FIGURE 1: View largeDownload slide Associations of fluid status and blood pressure with death. *not investigated in this study. UF-rate, ultrafiltration rate. The primary objective of this study was to investigate whether the relation between pre-SBP and mortality differs in HD patients with pre-dialysis FD, FO or NV, assessed by pre-dialysis MF-BIS, with traditional and novel statistical methods. Next to this, we aimed to investigate if this association remains stable when inflammation is present and when accounting for higher Ultrafiltration rate (UFR). Additionally, the relation between post-dialysis SBP and FS with outcome was assessed. MATERIALS AND METHODS Patient cohort selection and stratification For this observational cohort study, we used a European subset of the MONDO initiative database [19]. In these European clinics MF-BIS measurements were introduced earlier and were already part of daily routine since 2006; therefore, we chose to use the observations in this study cohort. We included all patients with at least one MF-BIS measurement and at least one C-reactive protein (CRP) measurement in 2011, as we wanted to treat the presence of inflammation as a confounder and CRP measurements were not routinely performed in all participating centres. The baseline period was defined as the 90 days prior to the date of the last MF-BIS measurement of 2011. All recorded laboratory and clinical parameters, derived from digital patients’ records were used to calculate average values for the baseline period. The baseline period was followed by a 1-year follow-up period during which all-cause mortality was recorded. Patients gave informed consent for the use of their data in anonymized form in epidemiological studies. The New England Institutional Review Board (IRB) has reviewed the claim of exemption for the study and has determined that this research activity is exempt from IRB review (see Supplementary Material for complete statement). Definitions and stratifications of FS and blood pressure Blood pressure was measured prior to the start of dialysis in the clinic in seated position using an oscillometric method. For the analyses, we stratified patients based on pre-SBP into three groups as reported in earlier studies: low pre-SBP <110 mmHg [16, 17, 20], normotensive (NT) pre-SBP 110–140 mmHg and high pre-SBP >140 mmHg [1, 9]. Body composition was assessed by whole-body MF-BIS measurements [Body Composition Monitor (BCM)®; Fresenius Medical Care, Bad Homburg, Germany], which expresses extracellular FO as a separate compartment called overhydration (OH) [4]. Post-dialysis FS was calculated from pre-dialysis FS minus ultrafiltration volume. According to their FS pre-dialysis, patients were stratified in the following groups: NV, with an OH compartment between −1.1 L and +1.1 L; FD OH <−1.1 L; and two levels of extracellular FO, FO (OH >+1.1 L to +2.5 L) and FO (OH >+2.5 L). NV limits were based on the clinical normogram, validated relative to an age- and gender-matched healthy patient population, used by the BCM device [3]. We used the lean tissue index (LTI) and fat tissue index (FTI) percentiles (<10th percentile; 10–90th percentile; >90th percentile) relative to an age- and gender-matched healthy population reported by the BCM device as markers for nutritional status [21]. Stratification of confounders To correct for geographic differences we stratified the patients into four European regions (north, east, south and west) based on the United Nations geographic scheme [22]. Inflammation was defined as a CRP level >6.0 mg/L. UFR was stratified based on the existing literature in <10, 10–13 and >13 mL/kg body weight/h [18, 23]. Statistical analysis Cox proportional hazard models with different levels of adjustment were used to assess the association between SBP and FS combined with outcome. In Model-I adjustment was made for: gender, age, vintage, geographic region, access type (catheter versus fistula) and the presence of the comorbidities diabetes mellitus, congestive heart failure, cerebrovascular disease, peripheral arterial disease and cancer. In the more advanced model, Model-II, we adjusted for: Model-I + LTI (<10th percentile versus >10th percentile), CRP (log transformed), albumin (g/dL), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (<4 h, ≥4 h), dialysate sodium (mEq/L), serum sodium (mEq/L) and UFR (mL/kg/h). Second, we assessed SBP and FS as continuous variables in a smoothing spline analysis of variance (ANOVA) Cox proportional hazard model. Smoothing spline ANOVA analysis allows modelling the joint effects of SBP and FS by treating them as continuous variables, thereby avoiding the need to create somewhat arbitrary categories. The results of this analysis are shown in an elevation plot, which can be read like a geographic topographic map [21]. We report point estimates and 95% confidence intervals (CIs) of hazard ratios (HRs). Censoring events were transfer to other clinics, transplantation, dialysis modality change or end of the study on 31 December 2012. A P-value of <0.05 was considered statistically significant. The analyses were performed with SAS version 9.3 (Cary, NC, USA) and SPSS statistics version 23.0 (IBM Corp., Armonk, NY, USA). RESULTS Study population In total, 8883 patients from the European MONDO Initiative Database (Figure 2) were included. The majority was male (57.2%), the median dialysis vintage was 3.6 (1.6–6.9) years and 18.6% of them were diagnosed with diabetes mellitus (Table 1, Supplementary data, Table S1 for all subgroup details). Table 1 Patients characteristics of the total cohort (n = 8883) Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 a Median (25th and 75th percentile). Table 1 Patients characteristics of the total cohort (n = 8883) Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 a Median (25th and 75th percentile). FIGURE 2: View largeDownload slide Study flow chart. FIGURE 2: View largeDownload slide Study flow chart. Pre-dialysis FS and blood pressure When pre-dialysis FS was analysed according to patients’ pre-SBP, no clear association between these two parameters was observed (Figure 3). In NT patients both pre-dialysis FO (> + 1.1 L to +2.5 L) [HR 1.57 (95% CI 1.21–2.04)] and FD (<–1.1 L) [HR 1.95 (95% CI 1.08–3.51)] were associated with increased mortality risk (Figure 3), when compared with patients with NV and NT. In NV patients a low pre-SBP (<110 mmHg) was associated with a reduced mortality risk [HR 0.52 (95% CI 0.27–1.00)] in the extensively adjusted model (Figure 3, Supplementary data, Table S2). FIGURE 3: View largeDownload slide The association of the combination of pre-SBP (mmHg) and pre-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n=1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). FIGURE 3: View largeDownload slide The association of the combination of pre-SBP (mmHg) and pre-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n=1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). Post-dialysis FS and blood pressure In patients stratified based on post-dialysis FS and SBP, a survival benefit was observed for patients with post-dialysis FD (<−1.1 L) and a SBP <140 mmHg [HR 0.64 (95% CI 0.45–0.90)] and for patients with SBP between 110 and 140 mmHg [HR 0.50 (95% CI 0.29–0.86)] (Figure 4, Supplementary data, Table S3). Patients within this subgroup most frequently had NV pre-dialysis (n = 1842, 69.0%) or moderate FO (n = 556, 20.8%). Furthermore, the same trend, an increased risk of death with increasing FO and decreasing SBP, as was also observed in the pre-dialysis analysis (Figure 4). FIGURE 4: View largeDownload slide The association of the combination of post-dialysis SBP (mmHg) and post-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a post-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS post-dialysis. Patients were stratified into different groups (horizontal and vertical lines) based on post-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and post-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L). FIGURE 4: View largeDownload slide The association of the combination of post-dialysis SBP (mmHg) and post-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a post-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS post-dialysis. Patients were stratified into different groups (horizontal and vertical lines) based on post-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and post-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L). Pre-dialysis FS subgroup analyses To further assess the effects of both pre-dialysis FS and pre- SBP separately, a subset analysis was performed with stratification on the basis of pre-dialysis FS with NT patients as reference group within the different subgroups (Figure 5, Supplementary data, Table S4). In the subgroup of FO patients, an increased mortality risk was observed in patients with a low pre-SBP [FO (>+1.1 to +2.5 L): HR 1.65 (95% CI 1.18–2.30) and FO (>+2.5 L): HR 1.50 (95% CI 1.04–2.14) in the basic model]. However, after adjustment for additional nutritional and inflammatory parameters (extended model) the associations lost statistical significance. In the subgroup of NV patients, we confirmed the association between low pre-SBP and a survival benefit [HR 0.46 (95% CI 0.23–0.91), in the extended model]. In the small subgroup of FD patients, no statistically significant effects of SBP were noted, but a tendency toward a lower mortality in the high SBP (>140 mmHg) was observed. FIGURE 5: View largeDownload slide Forest plot with the results of a Cox analysis. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n = 1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). FIGURE 5: View largeDownload slide Forest plot with the results of a Cox analysis. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n = 1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). Subset and sensitivity analyses FO and blood pressure pre-dialysis as continuous variables When both parameters were assessed as continuous variables to avoid loss of information due to categorization in different subgroups, by a smoothing spline analysis, a comparable trend was observed (Figure 6). In patients with FO the increased mortality risk with lower SBP was also observed, however, the aforementioned minimal survival advantage of higher SBP in both FD and FO patients disappeared (Figure 6). FIGURE 6: View largeDownload slide Risk of death across levels of pre-dialysis FS (L) and SBP (mmHg). Contour plot of the estimated probability of death in the next year as a joint function of pre-SBP and OH for female patients with age (at median values) and diabetes mellitus (as present) fixed. FIGURE 6: View largeDownload slide Risk of death across levels of pre-dialysis FS (L) and SBP (mmHg). Contour plot of the estimated probability of death in the next year as a joint function of pre-SBP and OH for female patients with age (at median values) and diabetes mellitus (as present) fixed. Effects of inflammation Due to the strong association between FS and inflammation on one hand and both of these parameters separately and combined with outcome, the association between pre-dialysis FS and SBP combined with outcome was analysed in patients stratified by inflammation status separately (Supplementary data, Figures S1A and S1B, Table S5). The effects of altered FS and SBP appeared to be more pronounced in inflamed patients. In the inflamed cohort, NV patients showed improved survival with low SBP [HR 0.37 (95% CI 0.15–0.93)], whereas in patients with altered pre-dialysis FS the presence of low pre-SBP (<110 mmHg) was associated with increased mortality risks: FD HR 3.15 (95% CI 1.26–7.91), moderate FO HR 2.26 (95% CI 1.44–3.56) and severe FO HR 3.28 (95% CI 2.05–5.24). Effects of ultrafiltration rate No clear association was found between pre-dialysis FS and UFR (Supplementary data, Figure S2). Of the patients with moderate or severe FO, only 3.4% of them had a UFR >13 mL/kg/h. When the association of UFR with outcome was analysed within the different FS subgroups, no significant associations with mortality were found (Supplementary data, Figure S2, Table S6). DISCUSSION In this study in prevalent HD patients we showed that the relation between pre-SBP and outcome is, at least partly, modified by pre-dialysis FS. In patients with FO, low SBP was associated with an increased mortality risk, compared with both NT patients with FO and NT patients with NV. In contrast, low SBP was associated with a survival benefit in patients with NV. Patients with post-dialysis FD and a post-dialysis SBP <140 mmHg also had an improved outcome. All effects were more pronounced when inflammation was present. The higher mortality rates in patients with FO could not be exclusively explained by higher UFR. Pre-dialysis FS and low pre-dialysis blood pressure (<110 mmHg) In patients with pre-dialysis FO, across all levels of pre-SBP, an increased risk of mortality was observed. The persisted association of FO with mortality is in line with several previous studies [4, 24]. The association of low pre-SBP (<110 mmHg) with adverse outcome in HD patients, without taking pre-dialysis FS into account, has been shown in several studies [9, 14, 16, 24]. It is often assumed that low pre-SBP reflects cardiac failure [10], and is, therefore, associated with a greatly increased mortality risk. On the contrary, previous studies rejected the hypothesis that the low SBP in HD patients identifies patients with cardiac failure, as they found preserved ejection fractions of the patients in the low SBP group in their study [25]. In the most recent study on this subject in an even larger patients cohort (39 566 incident patients using the Fresenius Medical Care dialysis network), Zoccali et al. also showed that cumulative FO, in combination with low pre-SBP (<130 mmHg) was associated with the highest risk of mortality [6]. In their study, FO was defined as an OH/extracellular water ratio of >15% in men or >13% in woman, approximating an absolute value of 2.5 L. Although SBP was taken into account in the analysis, the main topic of the study of Zoccali et al. was primarily addressing the effects of FO in a time-dependent fashion [6]. Our study focussed, in prevalent patients, on the interaction between SBP and abnormalities of FS including both different levels of FO as well as FD, taking potentially relevant confounders such as inflammation and UFR into account in separate analyses. Therefore, although the topic of both studies is related, they address the problem from quite different perspectives. A study by Chazot et al. is the only one where the combined effects of SBP and FS on mortality were analysed. The authors found that FO was a statistically stronger risk factor for mortality than high SBP [1]. Patients in this study were stratified in subgroups above and below 2.5 L FO with no further distinction in patients with NV or FD. Hypertension was defined as a pre-SBP >140 mmHg. The reference group of this study consisted of selected patients of the Tassin centre. In this centre, where survival rates are high due to excellent control of FS and sodium restriction, many insights into physiological patterns have been generated. However, in this particular study patients with a history of hypertension and antihypertensive medication were excluded. Our study adds several novel findings. We found an increased mortality risk in patients with moderate FO (>+1.1 to 2.5 L) even with normal (110–140 mmHg) and low (<110 mmHg) SBP, compared with the reference group of NT and NV patients. This appears to be a relevant addition to the literature, as the group of patients with pre-dialysis FO between +1.1 L and +2.5 L represents a large segment (38.8% in this study cohort), not always easily identified because of the absence of high SBP (58.4% of the NV cohort in this study) [26]. In contrast to the study of Chazot et al., in our study no a priori selection criteria were applied. The results of our study might, therefore, be influenced by the concomitant use of antihypertensive medication, which reflects general practice, but may interfere with the interpretation of physiological mechanisms. Within the subgroup of patients with FO >+2.5 L, the effects of low pre-SBP (<110 mmHg) were no longer statistically significant after extensive adjustment for nutritional, dialysis treatment-related and inflammatory markers. Besides the fact that FO has a strong association with increased mortality [3, 4, 8, 27], it also has strong correlations with markers of malnutrition [28] and inflammation [29], and has been associated with increased atherosclerosis and cardiac remodelling, eventually leading to cardiac failure [10]. Based on the present results, the relation between FO, SBP and outcome is likely complex, and also partly dependent on additional comorbid factors. However, as also shown in the study of Zoccali et al., FO per se appears to be a strong independent marker of outcome over the whole range of SBP levels [6]. In direct contrast to the patients with FO, a survival benefit in NV patients with a low pre-SBP (<110 mmHg) was observed. This is in some agreement with the recent findings of the SPRINT study, where in non-dialysis patients aiming for an SBP target <120 mmHg was associated with improved survival [15]. As such, in NV patients the U-shaped association between pre-SBP and outcome does not seem to exist. It is likely that patients with pre-dialysis NV represent a subgroup of patients in relative good health and relative good cardiovascular status. A prospective single-centre study, where fluid removal was recorded by MF-BIS measurements, found pre-SBP to decrease after improvement of FS after initiation of dialysis, which was also associated with an improved survival [30]. Pre-dialysis FS and high pre-SBP (>140 mmHg) Interestingly, we did not find an association between elevated SBP and increased 1-year mortality in the NV subgroup. Elevated pre-SBP in the NV cohort may be dependent on other factors, such as activation of the renin-angiotensin or sympathetic nervous system hyperreactivity, endothelins and prostaglandins [26]. This type of (dialysis resistant and volume independent) hypertension is often difficult to treat and most of the time is already apparent years before initiation of dialysis. Due to the relative short follow-up time of this study we cannot exclude an effect of SBP on survival by longer follow-up. In the subgroup of FO patients we were not able to conclusively demonstrate the survival benefit of high pre-SBP, as was found in several previous studies not taking pre-dialysis FS into account [14, 31], but a tendency toward this direction was observed. In a recent study by Bansal et al., where data of 1705 patients of the Chronic Renal Insufficiency Cohort Study (CRIC) study were used, a discrepancy was found between outpatients and in-centre SBP measurements, where a U-shaped relationship of SBP and mortality was found by in-centre pre-dialysis measurements associating higher SBP with a survival benefit [25]. Based on these results the authors suggested that the ability to achieve a high pre-SBP in response to interdialytic weight gain is a relative sign of health. In patients with pre-dialysis FD, in which we previously showed an increased mortality, this risk was predominantly observed in patients with low or normal pre-SBP (<140 mmHg), but not in the subgroup of patients with high pre-SBP, compared with the NV and NT reference population. However, in the subgroup analysis of the FD patients, the effects of the different SBP groups were not significant, possibly due to the relatively small number of patients. We hypothesize that in patients with pre-dialysis FD and lower SBP levels, the change of creating perfusion-dependent injury to a wide range of vulnerable vascular beds, as described by McIntyre [32], is more likely to occur due to the already present insufficient FS. Interestingly however, post-dialysis FD was not associated with an increased mortality risk, which is likely due to the fact that the majority of patients had NV and FS pre-dialysis. The impact of inflammation and UFR When patients were stratified based on the presence or absence of inflammation, the effects of altered FS were more pronounced in inflamed patients, but the general trend was not altered, likely due to the addition of an independent risk factor. Also, increased UFRs could not fully account for the increased mortality rates in the fluid overloaded patients, as a subanalysis within these FS groups showed hardly any effect of UFR. Limitations Despite the large and diverse HD cohort with patients from several European countries, our study has some limitations. First, we have no documentation of antihypertensive medications. Second, based on earlier studies in-centre pre-SBP measurements are possibly not an ideal reflection of SBP in HD patients [25, 33]; unfortunately neither interdialytic nor home BP measurements were available in our database, as we have no detailed documentation on the exact measurement procedures. In this retrospective cohort study, there are various unmeasured confounders not accounted for, the most important being residual kidney function which could have led to a more neutral FS in the NV patients subgroup and also partly explain the better survival observed. However, the median dialysis vintage of 3.6 years, with an equal distribution between the different FS subgroups, makes it unlikely that a substantial fraction of patients had residual renal function [34]. In an extensive record study performed by Robinson et al., a persistent relationship between SBP and outcome was found even after extensive correction for unmeasured confounders based on dialysis facility level [9]. In addition, we were able to correct extensively for nutritional and inflammatory parameters. But due to the observational design of this study, we were not able to completely eliminate residual confounding. Most importantly, echocardiographic results are not available and cardiac failure is likely underreported in our database. CONCLUSION This study shows that the relation between pre-SBP and outcome is at least partly dependent on pre-dialysis FS. Low SBP appears to be disadvantageous in FO and FD patients, but not in NV patients, where low SBP is actually associated with a survival benefit. Post-dialysis FD (<−1.1 L) and an SBP <140 mmHg was associated with a survival benefit. The effects of FS and pre-SBP are more pronounced in patients with inflammation. The higher mortality in patients with FO cannot be exclusively attributed to higher UFRs. Therefore, we suggest that BP levels should be interpreted in the context of FS and not as an isolated risk marker. Further, interventional studies should assess whether correction of pre-dialysis FO, as well as normalization of pre-dialysis FD in patients with low SBP is safe and associated with a survival benefit. Until such data are available, we suggest using this study mainly to gain insight in pathophysiological mechanisms contributing to the understanding of the often complex relationships between risk factors in HD patients. SUPPLEMENTARY DATA Supplementary data are available at ndt online. CONFLICT OF INTEREST STATEMENT B.C., P.C., S.S., M.M. and L.U. are employees of Fresenius Medical Care and hold shares in the company. P.K. holds stocks in Fresenius Medical Care. M.M. is advisory for Baxter, and a principal investigator for Bayer, Boheringer and ABBVIe clinical ongoing trials. J.K. reports grants from FMC outside the submitted work. J.R. is an employee of the Renal Research Institute. REFERENCES 1 Chazot C , Wabel P , Chamney P et al. Importance of normohydration for the long-term survival of haemodialysis patients . Nephrol Dial Transplant 2012 ; 27 : 2404 – 2410 Google Scholar CrossRef Search ADS PubMed 2 Hur E , Usta M , Toz H et al. Effect of fluid management guided by bioimpedance spectroscopy on cardiovascular parameters in hemodialysis patients: a randomized controlled trial . Am J Kidney Dis 2013 ; 61 : 957 – 965 Google Scholar CrossRef Search ADS PubMed 3 Wabel P , Moissl U , Chamney P et al. Towards improved cardiovascular management: the necessity of combining blood pressure and fluid overload . Nephrol Dial Transplant 2008 ; 23 : 2965 – 2971 Google Scholar CrossRef Search ADS PubMed 4 Wizemann V , Wabel P , Chamney P et al. The mortality risk of overhydration in haemodialysis patients . Nephrol Dial Transplant 2009 ; 24 : 1574 – 1579 Google Scholar CrossRef Search ADS PubMed 5 Onofriescu M , Hogas S , Voroneanu L et al. Bioimpedance-guided fluid management in maintenance hemodialysis: a pilot randomized controlled trial . Am J Kidney Dis 2014 ; 1 – 8 6 Zoccali C , Moissl U , Chazot C et al. Chronic fluid overload and mortality in ESRD . J Am Soc Nephrol 2017 ; 28 : 2491 – 2497 Google Scholar CrossRef Search ADS PubMed 7 Dekker MJE , Marcelli D , Canaud BJ et al. Impact of fluid status and inflammation and their interaction on survival: a study in an international hemodialysis patient cohort . Kidney Int 2017 ; 91 : 1214 – 1223 Google Scholar CrossRef Search ADS PubMed 8 Dekker MJE , Marcelli D , Canaud B et al. Unraveling the relationship between mortality, hyponatremia, inflammation and malnutrition in hemodialysis patients: results from the international MONDO initiative . Eur J Clin Nutr 2016 ; 70 : 779 – 784 Google Scholar CrossRef Search ADS PubMed 9 Robinson BM , Tong L , Zhang J et al. Blood pressure levels and mortality risk among hemodialysis patients in the dialysis outcomes and practice patterns study . Kidney Int 2012 ; 82 : 570 – 580 Google Scholar CrossRef Search ADS PubMed 10 Foley RN , Parfrey PS , Harnett JD et al. Impact of hypertension on cardiomyopathy, morbidity and mortality in end-stage renal disease . Kidney Int 1996 ; 49 : 1379 – 1385 Google Scholar CrossRef Search ADS PubMed 11 Nongnuch A , Campbell N , Stern E et al. Increased postdialysis systolic blood pressure is associated with extracellular overhydration in hemodialysis outpatients . Kidney Int 2014 ; 87 : 1 – 6 12 Heerspink HLJ , Ninomiya T , Zoungas S et al. Effect of lowering blood pressure on cardiovascular events and mortality in patients on dialysis: a systematic review and meta-analysis of randomised controlled trials . Lancet 2009 ; 373 : 1009 – 1015 Google Scholar CrossRef Search ADS PubMed 13 Odudu A , McIntyre C. Volume is not the only key to hypertension control in dialysis patients . Nephron Clin Pract 2012 ; 120 : c173 – c177 Google Scholar CrossRef Search ADS PubMed 14 Li Z , Lacson E , Lowrie EG et al. The epidemiology of systolic blood pressure and death risk in hemodialysis patients . Am J Kidney Dis 2006 ; 48 : 606 – 615 Google Scholar CrossRef Search ADS PubMed 15 SPRINT Research Group , Wright JT , Williamson JD et al. A randomized trial of intensive versus standard blood-pressure control . N Engl J Med 2015 ; 373 : 2103 – 2116 Google Scholar CrossRef Search ADS PubMed 16 Zager PG , Nikolic J , Brown RH et al. ‘U’ curve association of blood pressure and mortality in hemodialysis patients . Kidney Int 1998 ; 54 : 561 – 569 Google Scholar CrossRef Search ADS PubMed 17 Kalantar-Zadeh K , Kilpatrick RD , McAllister CJ et al. Reverse epidemiology of hypertension and cardiovascular death in the hemodialysis population . Hypertension 2005 ; 45 : 811 Google Scholar CrossRef Search ADS PubMed 18 Flythe JE , Assimon MM , Wenger JB et al. Ultrafiltration rates and the quality incentive program: proposed measure definitions and their potential dialysis facility implications . Clin J Am Soc Nephrol 2016 ; 11 : 1422 Google Scholar CrossRef Search ADS PubMed 19 Usvyat LA , Haviv YS , Etter M et al. The MONitoring Dialysis Outcomes (MONDO) initiative . Blood Purif 2013 ; 35 : 37 – 48 Google Scholar CrossRef Search ADS PubMed 20 Port FK , Hulbert-Shearon TE , Wolfe RA et al. Predialysis blood pressure and mortality risk in a national sample of maintenance hemodialysis patients . Am J Kidney Dis 1999 ; 33 : 507 – 517 Google Scholar CrossRef Search ADS PubMed 21 Marcelli D , Usvyat L. A , Kotanko P et al. Body composition and survival in dialysis patients: results from an international cohort study . Clin J Am Soc Nephrol 2015 ; 10 : 1192 – 1199 Google Scholar CrossRef Search ADS PubMed 22 United Nations Statistics Division- Standard Country and Area Codes Classifications (M49) [cited 23 January 2016] http://unstats.un.org/unsd/methods/m49/m49regin.htm#europe 23 Saran R , Bragg-Gresham JL , Levin NW et al. Longer treatment time and slower ultrafiltration in hemodialysis: associations with reduced mortality in the DOPPS . Kidney Int 2006 ; 69 : 1222 – 1228 Google Scholar CrossRef Search ADS PubMed 24 Kalantar-Zadeh K , Kilpatrick RD , Kopple JD. Reverse epidemiology of blood pressure in dialysis patients . Kidney Int 2005 ; 67 : 2067 , author reply 2067–2068 Google Scholar CrossRef Search ADS PubMed 25 Bansal N , McCulloch CE , Rahman M et al. Blood pressure and risk of all-cause mortality in advanced chronic kidney disease and hemodialysis: the chronic renal insufficiency cohort study . Hypertension 2015 ; 65 : 93 – 100 Google Scholar CrossRef Search ADS PubMed 26 Jaeger JQ , Mehta RL. Assessment of dry weight in hemodialysis: an overview . J Am Soc Nephrol 1999 ; 10 : 392 – 403 Google Scholar PubMed 27 Kalantar-Zadeh K , Regidor DL , Kovesdy CP et al. Fluid retention is associated with cardiovascular mortality in patients undergoing long-term hemodialysis . Circulation 2009 ; 119 : 671 – 679 Google Scholar CrossRef Search ADS PubMed 28 Agarwal R. Body mass index-mortality paradox in hemodialysis: can it be explained by blood pressure? Hypertension 2011 ; 58 : 1014 – 1020 Google Scholar CrossRef Search ADS PubMed 29 McIntyre CW , Harrison LEA , Eldehni MT et al. Circulating endotoxemia: a novel factor in systemic inflammation and cardiovascular disease in chronic kidney disease . Clin J Am Soc Nephrol 2011 ; 6 : 133 – 141 Google Scholar CrossRef Search ADS PubMed 30 Chazot C , Vo-Van C , Deleaval P et al. Predialysis systolic blood pressure evolution in incident hemodialysis patients: effects of the dry weight method and prognostic value . Blood Purif 2012 ; 33 : 275 – 283 Google Scholar CrossRef Search ADS PubMed 31 Tentori F , Hunt WC , Rohrscheib M et al. Which targets in clinical practice guidelines are associated with improved survival in a large dialysis organization? J Am Soc Nephrol 2007 ; 18 : 2377 – 2384 Google Scholar CrossRef Search ADS PubMed 32 McIntyre CW. Recurrent circulatory stress: the dark side of dialysis . Semin Dial 2010 ; 23 : 449 – 451 Google Scholar CrossRef Search ADS PubMed 33 Agarwal R. Volume-associated ambulatory blood pressure patterns in hemodialysis patients . Hypertension 2009 ; 54 : 241 – 247 Google Scholar CrossRef Search ADS PubMed 34 Termorshuizen F , Dekker FW , van Manen JG et al. Relative contribution of residual renal function and different measures of adequacy to survival in hemodialysis patients: an analysis of the Netherlands Cooperative Study on the Adequacy of Dialysis (NECOSAD)-2 . J Am Soc Nephrol 2004 ; 15 : 1061 – 1070 Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nephrology Dialysis Transplantation Oxford University Press

Pre-dialysis fluid status, pre-dialysis systolic blood pressure and outcome in prevalent haemodialysis patients: results of an international cohort study on behalf of the MONDO initiative

Loading next page...
 
/lp/ou_press/pre-dialysis-fluid-status-pre-dialysis-systolic-blood-pressure-and-qkWlxKDMMC
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
ISSN
0931-0509
eISSN
1460-2385
D.O.I.
10.1093/ndt/gfy095
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Pre-dialysis fluid overload (FO) associates with mortality and causes elevated pre-dialysis systolic blood pressure (pre-SBP). However, low pre-SBP is associated with increased mortality in haemodialysis patients. The objective of this study was to investigate the interaction between pre-dialysis fluid status (FS) and pre-SBP in association with mortality. Methods We included all patients from the international Monitoring Dialysis Outcome Initiative (MONDO) database with a pre-dialysis multifrequency bioimpedance spectroscopy measurement in the year 2011. We used all parameters available during a 90-day baseline period. All-cause mortality was recorded during 1-year follow-up. Associations with outcome were assessed with Cox models and a smoothing spline Cox analysis. Results We included 8883 patients. In patients with pre-dialysis FO (>+1.1 to +2.5 L), pre-SBP <110 mmHg was associated with an increased risk of death {hazard ratio (HR) 1.52 [95% confidence interval (CI) 1.06–2.17]}. An increased risk of death was also associated with pre-dialysis fluid depletion (FD; <−1.1 L) combined with a pre-SBP <140 mmHg. In normovolemic (NV) patients, low pre-SBP <110 mmHg was associated with better survival [HR 0.46 (95% CI 0.23–0.91)]. Also, post-dialysis FD associated with a survival benefit. Results were similar when inflammation was present. Only high ultrafiltration rate could not explain the higher mortality rates observed. Conclusion The relation between pre-SBP and outcome is dependent on pre-dialysis FS. Low pre-SBP appears to be disadvantageous in patients with FO or FD, but not in NV patients. Post-dialysis FD was found to associate with improved survival. Therefore, we suggest interpreting pre-SBP levels in the context of FS and not as an isolated marker. fluid overload, haemodialysis, pre-dialysis blood pressure, pre-dialysis fluid status, survival INTRODUCTION In chronic haemodialysis (HD) patients, a consistent association between extracellular fluid overload (FO), assessed by multifrequency bioimpedance (MF-BIS), and an increased risk of mortality has been observed [1–6]. This relation is even apparent at mild levels of FO [7]. Recently, we showed that pre-dialysis fluid depletion (FD) was also associated with an increased mortality [8]. In addition, a recent study showed that cumulative FO over a 1-year period was strongly related to outcome [6]. FO has been associated with an increase in systolic blood pressure (SBP) [9–11], whereas correction of FO usually leads to improvement of hypertension [5, 12]. However, with FO assessed by MF-BIS, not all hypertensive patients were found to be fluid overloaded, nor did all patients with FO have an elevated SBP [1, 3, 6, 13]. Several observational studies have observed an increased mortality risk with high (SBP >140–180 mmHg), but especially with low (SBP <110–130 mmHg) pre- and post-dialysis SBP levels [11, 14]. This is in contrast to observations in the general population, where recent results of the Systolic Blood Pressure Intervention Trail (SPRINT) trail showed that an SBP target <120 mmHg is beneficial in patients with increased cardiovascular risk but without diabetes and end-stage kidney disease [15]. The relation between low SBP and increased mortality risk in dialysis patients has been coined as an epidemiological paradox [16, 17]. However, given its strong relation with both outcome as well as SBP levels per se, we hypothesized that fluid status (FS) is an important modifier of the association between mortality and pre-dialysis SBP (pre-SBP) in HD patients (Figure 1). For instance, the fact that low or normal pre-SBP may be physiological when associated with normovolemia (NV), but pathological when associated with FO, as the normal cardiovascular response to FO (i.e. an increase in cardiac output and/or an increase in systemic vascular resistance) may be impaired. A recent study indeed showed that the highest risk of mortality was observed in patients with FO with low (<130 mmHg) pre- SBP [6]. Other potentially relevant factors in this context include systemic inflammation, which was shown to be a strong amplifier of the FO–mortality relationship, as well as ultrafiltration rate, which bears a strong relation with both outcome as well as pre-dialysis FO [18]. FIGURE 1: View largeDownload slide Associations of fluid status and blood pressure with death. *not investigated in this study. UF-rate, ultrafiltration rate. FIGURE 1: View largeDownload slide Associations of fluid status and blood pressure with death. *not investigated in this study. UF-rate, ultrafiltration rate. The primary objective of this study was to investigate whether the relation between pre-SBP and mortality differs in HD patients with pre-dialysis FD, FO or NV, assessed by pre-dialysis MF-BIS, with traditional and novel statistical methods. Next to this, we aimed to investigate if this association remains stable when inflammation is present and when accounting for higher Ultrafiltration rate (UFR). Additionally, the relation between post-dialysis SBP and FS with outcome was assessed. MATERIALS AND METHODS Patient cohort selection and stratification For this observational cohort study, we used a European subset of the MONDO initiative database [19]. In these European clinics MF-BIS measurements were introduced earlier and were already part of daily routine since 2006; therefore, we chose to use the observations in this study cohort. We included all patients with at least one MF-BIS measurement and at least one C-reactive protein (CRP) measurement in 2011, as we wanted to treat the presence of inflammation as a confounder and CRP measurements were not routinely performed in all participating centres. The baseline period was defined as the 90 days prior to the date of the last MF-BIS measurement of 2011. All recorded laboratory and clinical parameters, derived from digital patients’ records were used to calculate average values for the baseline period. The baseline period was followed by a 1-year follow-up period during which all-cause mortality was recorded. Patients gave informed consent for the use of their data in anonymized form in epidemiological studies. The New England Institutional Review Board (IRB) has reviewed the claim of exemption for the study and has determined that this research activity is exempt from IRB review (see Supplementary Material for complete statement). Definitions and stratifications of FS and blood pressure Blood pressure was measured prior to the start of dialysis in the clinic in seated position using an oscillometric method. For the analyses, we stratified patients based on pre-SBP into three groups as reported in earlier studies: low pre-SBP <110 mmHg [16, 17, 20], normotensive (NT) pre-SBP 110–140 mmHg and high pre-SBP >140 mmHg [1, 9]. Body composition was assessed by whole-body MF-BIS measurements [Body Composition Monitor (BCM)®; Fresenius Medical Care, Bad Homburg, Germany], which expresses extracellular FO as a separate compartment called overhydration (OH) [4]. Post-dialysis FS was calculated from pre-dialysis FS minus ultrafiltration volume. According to their FS pre-dialysis, patients were stratified in the following groups: NV, with an OH compartment between −1.1 L and +1.1 L; FD OH <−1.1 L; and two levels of extracellular FO, FO (OH >+1.1 L to +2.5 L) and FO (OH >+2.5 L). NV limits were based on the clinical normogram, validated relative to an age- and gender-matched healthy patient population, used by the BCM device [3]. We used the lean tissue index (LTI) and fat tissue index (FTI) percentiles (<10th percentile; 10–90th percentile; >90th percentile) relative to an age- and gender-matched healthy population reported by the BCM device as markers for nutritional status [21]. Stratification of confounders To correct for geographic differences we stratified the patients into four European regions (north, east, south and west) based on the United Nations geographic scheme [22]. Inflammation was defined as a CRP level >6.0 mg/L. UFR was stratified based on the existing literature in <10, 10–13 and >13 mL/kg body weight/h [18, 23]. Statistical analysis Cox proportional hazard models with different levels of adjustment were used to assess the association between SBP and FS combined with outcome. In Model-I adjustment was made for: gender, age, vintage, geographic region, access type (catheter versus fistula) and the presence of the comorbidities diabetes mellitus, congestive heart failure, cerebrovascular disease, peripheral arterial disease and cancer. In the more advanced model, Model-II, we adjusted for: Model-I + LTI (<10th percentile versus >10th percentile), CRP (log transformed), albumin (g/dL), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (<4 h, ≥4 h), dialysate sodium (mEq/L), serum sodium (mEq/L) and UFR (mL/kg/h). Second, we assessed SBP and FS as continuous variables in a smoothing spline analysis of variance (ANOVA) Cox proportional hazard model. Smoothing spline ANOVA analysis allows modelling the joint effects of SBP and FS by treating them as continuous variables, thereby avoiding the need to create somewhat arbitrary categories. The results of this analysis are shown in an elevation plot, which can be read like a geographic topographic map [21]. We report point estimates and 95% confidence intervals (CIs) of hazard ratios (HRs). Censoring events were transfer to other clinics, transplantation, dialysis modality change or end of the study on 31 December 2012. A P-value of <0.05 was considered statistically significant. The analyses were performed with SAS version 9.3 (Cary, NC, USA) and SPSS statistics version 23.0 (IBM Corp., Armonk, NY, USA). RESULTS Study population In total, 8883 patients from the European MONDO Initiative Database (Figure 2) were included. The majority was male (57.2%), the median dialysis vintage was 3.6 (1.6–6.9) years and 18.6% of them were diagnosed with diabetes mellitus (Table 1, Supplementary data, Table S1 for all subgroup details). Table 1 Patients characteristics of the total cohort (n = 8883) Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 a Median (25th and 75th percentile). Table 1 Patients characteristics of the total cohort (n = 8883) Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 Patient characteristics Mean/median SD/25th–75th percentile Male (%) 57.2 Age (years) 63 14.8 Vintage (years)a 3.60 1.63–6.94 Comorbidities (%)  Diabetes mellitus 18.6  Congestive heart failure 0.9  Cardiovascular disease 2.2  Peripheral vascular disease 0.8  Cancer 3.5 Dialysis characteristics  Access catheter (%) 18.6  Dialysate sodium (mEq/L) 139 1.85  Ultrafiltration rate (mL/kg/h) 7.36 2.85  Pre-SBP (mmHg) 139 19.9  Interdialytic weight gain (kg) 2.10 0.84  Treatment time (min) 247 19.72 Body composition  Pre-dialysis weight (kg) 72.69 16.01  LTI <10th percentile (%) 45.5  Pre-dialysis OH (L)a 1.65 0.78–2.62  Body mass index (kg/m2) 25.96 5.23 Laboratory parameters  Albumin (g/dL) 3.85 0.39  Sodium (mEq/L) 138 3.31  C-Reactive protein (mg/L) 5.7 2.4–13.0 a Median (25th and 75th percentile). FIGURE 2: View largeDownload slide Study flow chart. FIGURE 2: View largeDownload slide Study flow chart. Pre-dialysis FS and blood pressure When pre-dialysis FS was analysed according to patients’ pre-SBP, no clear association between these two parameters was observed (Figure 3). In NT patients both pre-dialysis FO (> + 1.1 L to +2.5 L) [HR 1.57 (95% CI 1.21–2.04)] and FD (<–1.1 L) [HR 1.95 (95% CI 1.08–3.51)] were associated with increased mortality risk (Figure 3), when compared with patients with NV and NT. In NV patients a low pre-SBP (<110 mmHg) was associated with a reduced mortality risk [HR 0.52 (95% CI 0.27–1.00)] in the extensively adjusted model (Figure 3, Supplementary data, Table S2). FIGURE 3: View largeDownload slide The association of the combination of pre-SBP (mmHg) and pre-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n=1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). FIGURE 3: View largeDownload slide The association of the combination of pre-SBP (mmHg) and pre-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n=1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). Post-dialysis FS and blood pressure In patients stratified based on post-dialysis FS and SBP, a survival benefit was observed for patients with post-dialysis FD (<−1.1 L) and a SBP <140 mmHg [HR 0.64 (95% CI 0.45–0.90)] and for patients with SBP between 110 and 140 mmHg [HR 0.50 (95% CI 0.29–0.86)] (Figure 4, Supplementary data, Table S3). Patients within this subgroup most frequently had NV pre-dialysis (n = 1842, 69.0%) or moderate FO (n = 556, 20.8%). Furthermore, the same trend, an increased risk of death with increasing FO and decreasing SBP, as was also observed in the pre-dialysis analysis (Figure 4). FIGURE 4: View largeDownload slide The association of the combination of post-dialysis SBP (mmHg) and post-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a post-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS post-dialysis. Patients were stratified into different groups (horizontal and vertical lines) based on post-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and post-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L). FIGURE 4: View largeDownload slide The association of the combination of post-dialysis SBP (mmHg) and post-dialysis FS (L) with mortality. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a post-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS post-dialysis. Patients were stratified into different groups (horizontal and vertical lines) based on post-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and post-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L). Pre-dialysis FS subgroup analyses To further assess the effects of both pre-dialysis FS and pre- SBP separately, a subset analysis was performed with stratification on the basis of pre-dialysis FS with NT patients as reference group within the different subgroups (Figure 5, Supplementary data, Table S4). In the subgroup of FO patients, an increased mortality risk was observed in patients with a low pre-SBP [FO (>+1.1 to +2.5 L): HR 1.65 (95% CI 1.18–2.30) and FO (>+2.5 L): HR 1.50 (95% CI 1.04–2.14) in the basic model]. However, after adjustment for additional nutritional and inflammatory parameters (extended model) the associations lost statistical significance. In the subgroup of NV patients, we confirmed the association between low pre-SBP and a survival benefit [HR 0.46 (95% CI 0.23–0.91), in the extended model]. In the small subgroup of FD patients, no statistically significant effects of SBP were noted, but a tendency toward a lower mortality in the high SBP (>140 mmHg) was observed. FIGURE 5: View largeDownload slide Forest plot with the results of a Cox analysis. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n = 1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). FIGURE 5: View largeDownload slide Forest plot with the results of a Cox analysis. HRs of a Cox proportional hazards model adjusted for gender, age, dialysis vintage, region (categorical), access type (catheter versus fistula) albumin (g/dL), C-reactive protein (ln) (mg/L), LTI (categorical as <10th percentile versus >10th percentile), body mass index (kg/m2), interdialytic weight gain (kg), treatment time (categorized >240 and <240 min), dialysate sodium (mEq/L), serum sodium (mEq/L), ultrafiltration rate (mL/kg/h) and the comorbidities congestive heart failure, diabetes mellitus, cardiovascular disease, peripheral vascular disease and cancer. The reference group consists of patients with a pre-dialysis SBP between 110 and 140 mmHg and NV (between −1.1 L and +1.1 L) FS pre-dialysis (n = 1362). Patients were stratified into different groups (horizontal and vertical lines) based on pre-dialysis SBP (<110 mmHg, 110–140 mmHg, >140 mmHg) and pre-dialysis FS by MF-BIS (<−1.1 L, −1.1 to +1.1 L, >+1.1 L to +2.5 L, >+2.5 L) (see Supplementary data, Table S1 for subgroup characteristics). Subset and sensitivity analyses FO and blood pressure pre-dialysis as continuous variables When both parameters were assessed as continuous variables to avoid loss of information due to categorization in different subgroups, by a smoothing spline analysis, a comparable trend was observed (Figure 6). In patients with FO the increased mortality risk with lower SBP was also observed, however, the aforementioned minimal survival advantage of higher SBP in both FD and FO patients disappeared (Figure 6). FIGURE 6: View largeDownload slide Risk of death across levels of pre-dialysis FS (L) and SBP (mmHg). Contour plot of the estimated probability of death in the next year as a joint function of pre-SBP and OH for female patients with age (at median values) and diabetes mellitus (as present) fixed. FIGURE 6: View largeDownload slide Risk of death across levels of pre-dialysis FS (L) and SBP (mmHg). Contour plot of the estimated probability of death in the next year as a joint function of pre-SBP and OH for female patients with age (at median values) and diabetes mellitus (as present) fixed. Effects of inflammation Due to the strong association between FS and inflammation on one hand and both of these parameters separately and combined with outcome, the association between pre-dialysis FS and SBP combined with outcome was analysed in patients stratified by inflammation status separately (Supplementary data, Figures S1A and S1B, Table S5). The effects of altered FS and SBP appeared to be more pronounced in inflamed patients. In the inflamed cohort, NV patients showed improved survival with low SBP [HR 0.37 (95% CI 0.15–0.93)], whereas in patients with altered pre-dialysis FS the presence of low pre-SBP (<110 mmHg) was associated with increased mortality risks: FD HR 3.15 (95% CI 1.26–7.91), moderate FO HR 2.26 (95% CI 1.44–3.56) and severe FO HR 3.28 (95% CI 2.05–5.24). Effects of ultrafiltration rate No clear association was found between pre-dialysis FS and UFR (Supplementary data, Figure S2). Of the patients with moderate or severe FO, only 3.4% of them had a UFR >13 mL/kg/h. When the association of UFR with outcome was analysed within the different FS subgroups, no significant associations with mortality were found (Supplementary data, Figure S2, Table S6). DISCUSSION In this study in prevalent HD patients we showed that the relation between pre-SBP and outcome is, at least partly, modified by pre-dialysis FS. In patients with FO, low SBP was associated with an increased mortality risk, compared with both NT patients with FO and NT patients with NV. In contrast, low SBP was associated with a survival benefit in patients with NV. Patients with post-dialysis FD and a post-dialysis SBP <140 mmHg also had an improved outcome. All effects were more pronounced when inflammation was present. The higher mortality rates in patients with FO could not be exclusively explained by higher UFR. Pre-dialysis FS and low pre-dialysis blood pressure (<110 mmHg) In patients with pre-dialysis FO, across all levels of pre-SBP, an increased risk of mortality was observed. The persisted association of FO with mortality is in line with several previous studies [4, 24]. The association of low pre-SBP (<110 mmHg) with adverse outcome in HD patients, without taking pre-dialysis FS into account, has been shown in several studies [9, 14, 16, 24]. It is often assumed that low pre-SBP reflects cardiac failure [10], and is, therefore, associated with a greatly increased mortality risk. On the contrary, previous studies rejected the hypothesis that the low SBP in HD patients identifies patients with cardiac failure, as they found preserved ejection fractions of the patients in the low SBP group in their study [25]. In the most recent study on this subject in an even larger patients cohort (39 566 incident patients using the Fresenius Medical Care dialysis network), Zoccali et al. also showed that cumulative FO, in combination with low pre-SBP (<130 mmHg) was associated with the highest risk of mortality [6]. In their study, FO was defined as an OH/extracellular water ratio of >15% in men or >13% in woman, approximating an absolute value of 2.5 L. Although SBP was taken into account in the analysis, the main topic of the study of Zoccali et al. was primarily addressing the effects of FO in a time-dependent fashion [6]. Our study focussed, in prevalent patients, on the interaction between SBP and abnormalities of FS including both different levels of FO as well as FD, taking potentially relevant confounders such as inflammation and UFR into account in separate analyses. Therefore, although the topic of both studies is related, they address the problem from quite different perspectives. A study by Chazot et al. is the only one where the combined effects of SBP and FS on mortality were analysed. The authors found that FO was a statistically stronger risk factor for mortality than high SBP [1]. Patients in this study were stratified in subgroups above and below 2.5 L FO with no further distinction in patients with NV or FD. Hypertension was defined as a pre-SBP >140 mmHg. The reference group of this study consisted of selected patients of the Tassin centre. In this centre, where survival rates are high due to excellent control of FS and sodium restriction, many insights into physiological patterns have been generated. However, in this particular study patients with a history of hypertension and antihypertensive medication were excluded. Our study adds several novel findings. We found an increased mortality risk in patients with moderate FO (>+1.1 to 2.5 L) even with normal (110–140 mmHg) and low (<110 mmHg) SBP, compared with the reference group of NT and NV patients. This appears to be a relevant addition to the literature, as the group of patients with pre-dialysis FO between +1.1 L and +2.5 L represents a large segment (38.8% in this study cohort), not always easily identified because of the absence of high SBP (58.4% of the NV cohort in this study) [26]. In contrast to the study of Chazot et al., in our study no a priori selection criteria were applied. The results of our study might, therefore, be influenced by the concomitant use of antihypertensive medication, which reflects general practice, but may interfere with the interpretation of physiological mechanisms. Within the subgroup of patients with FO >+2.5 L, the effects of low pre-SBP (<110 mmHg) were no longer statistically significant after extensive adjustment for nutritional, dialysis treatment-related and inflammatory markers. Besides the fact that FO has a strong association with increased mortality [3, 4, 8, 27], it also has strong correlations with markers of malnutrition [28] and inflammation [29], and has been associated with increased atherosclerosis and cardiac remodelling, eventually leading to cardiac failure [10]. Based on the present results, the relation between FO, SBP and outcome is likely complex, and also partly dependent on additional comorbid factors. However, as also shown in the study of Zoccali et al., FO per se appears to be a strong independent marker of outcome over the whole range of SBP levels [6]. In direct contrast to the patients with FO, a survival benefit in NV patients with a low pre-SBP (<110 mmHg) was observed. This is in some agreement with the recent findings of the SPRINT study, where in non-dialysis patients aiming for an SBP target <120 mmHg was associated with improved survival [15]. As such, in NV patients the U-shaped association between pre-SBP and outcome does not seem to exist. It is likely that patients with pre-dialysis NV represent a subgroup of patients in relative good health and relative good cardiovascular status. A prospective single-centre study, where fluid removal was recorded by MF-BIS measurements, found pre-SBP to decrease after improvement of FS after initiation of dialysis, which was also associated with an improved survival [30]. Pre-dialysis FS and high pre-SBP (>140 mmHg) Interestingly, we did not find an association between elevated SBP and increased 1-year mortality in the NV subgroup. Elevated pre-SBP in the NV cohort may be dependent on other factors, such as activation of the renin-angiotensin or sympathetic nervous system hyperreactivity, endothelins and prostaglandins [26]. This type of (dialysis resistant and volume independent) hypertension is often difficult to treat and most of the time is already apparent years before initiation of dialysis. Due to the relative short follow-up time of this study we cannot exclude an effect of SBP on survival by longer follow-up. In the subgroup of FO patients we were not able to conclusively demonstrate the survival benefit of high pre-SBP, as was found in several previous studies not taking pre-dialysis FS into account [14, 31], but a tendency toward this direction was observed. In a recent study by Bansal et al., where data of 1705 patients of the Chronic Renal Insufficiency Cohort Study (CRIC) study were used, a discrepancy was found between outpatients and in-centre SBP measurements, where a U-shaped relationship of SBP and mortality was found by in-centre pre-dialysis measurements associating higher SBP with a survival benefit [25]. Based on these results the authors suggested that the ability to achieve a high pre-SBP in response to interdialytic weight gain is a relative sign of health. In patients with pre-dialysis FD, in which we previously showed an increased mortality, this risk was predominantly observed in patients with low or normal pre-SBP (<140 mmHg), but not in the subgroup of patients with high pre-SBP, compared with the NV and NT reference population. However, in the subgroup analysis of the FD patients, the effects of the different SBP groups were not significant, possibly due to the relatively small number of patients. We hypothesize that in patients with pre-dialysis FD and lower SBP levels, the change of creating perfusion-dependent injury to a wide range of vulnerable vascular beds, as described by McIntyre [32], is more likely to occur due to the already present insufficient FS. Interestingly however, post-dialysis FD was not associated with an increased mortality risk, which is likely due to the fact that the majority of patients had NV and FS pre-dialysis. The impact of inflammation and UFR When patients were stratified based on the presence or absence of inflammation, the effects of altered FS were more pronounced in inflamed patients, but the general trend was not altered, likely due to the addition of an independent risk factor. Also, increased UFRs could not fully account for the increased mortality rates in the fluid overloaded patients, as a subanalysis within these FS groups showed hardly any effect of UFR. Limitations Despite the large and diverse HD cohort with patients from several European countries, our study has some limitations. First, we have no documentation of antihypertensive medications. Second, based on earlier studies in-centre pre-SBP measurements are possibly not an ideal reflection of SBP in HD patients [25, 33]; unfortunately neither interdialytic nor home BP measurements were available in our database, as we have no detailed documentation on the exact measurement procedures. In this retrospective cohort study, there are various unmeasured confounders not accounted for, the most important being residual kidney function which could have led to a more neutral FS in the NV patients subgroup and also partly explain the better survival observed. However, the median dialysis vintage of 3.6 years, with an equal distribution between the different FS subgroups, makes it unlikely that a substantial fraction of patients had residual renal function [34]. In an extensive record study performed by Robinson et al., a persistent relationship between SBP and outcome was found even after extensive correction for unmeasured confounders based on dialysis facility level [9]. In addition, we were able to correct extensively for nutritional and inflammatory parameters. But due to the observational design of this study, we were not able to completely eliminate residual confounding. Most importantly, echocardiographic results are not available and cardiac failure is likely underreported in our database. CONCLUSION This study shows that the relation between pre-SBP and outcome is at least partly dependent on pre-dialysis FS. Low SBP appears to be disadvantageous in FO and FD patients, but not in NV patients, where low SBP is actually associated with a survival benefit. Post-dialysis FD (<−1.1 L) and an SBP <140 mmHg was associated with a survival benefit. The effects of FS and pre-SBP are more pronounced in patients with inflammation. The higher mortality in patients with FO cannot be exclusively attributed to higher UFRs. Therefore, we suggest that BP levels should be interpreted in the context of FS and not as an isolated risk marker. Further, interventional studies should assess whether correction of pre-dialysis FO, as well as normalization of pre-dialysis FD in patients with low SBP is safe and associated with a survival benefit. Until such data are available, we suggest using this study mainly to gain insight in pathophysiological mechanisms contributing to the understanding of the often complex relationships between risk factors in HD patients. SUPPLEMENTARY DATA Supplementary data are available at ndt online. CONFLICT OF INTEREST STATEMENT B.C., P.C., S.S., M.M. and L.U. are employees of Fresenius Medical Care and hold shares in the company. P.K. holds stocks in Fresenius Medical Care. M.M. is advisory for Baxter, and a principal investigator for Bayer, Boheringer and ABBVIe clinical ongoing trials. J.K. reports grants from FMC outside the submitted work. J.R. is an employee of the Renal Research Institute. REFERENCES 1 Chazot C , Wabel P , Chamney P et al. Importance of normohydration for the long-term survival of haemodialysis patients . Nephrol Dial Transplant 2012 ; 27 : 2404 – 2410 Google Scholar CrossRef Search ADS PubMed 2 Hur E , Usta M , Toz H et al. Effect of fluid management guided by bioimpedance spectroscopy on cardiovascular parameters in hemodialysis patients: a randomized controlled trial . Am J Kidney Dis 2013 ; 61 : 957 – 965 Google Scholar CrossRef Search ADS PubMed 3 Wabel P , Moissl U , Chamney P et al. Towards improved cardiovascular management: the necessity of combining blood pressure and fluid overload . Nephrol Dial Transplant 2008 ; 23 : 2965 – 2971 Google Scholar CrossRef Search ADS PubMed 4 Wizemann V , Wabel P , Chamney P et al. The mortality risk of overhydration in haemodialysis patients . Nephrol Dial Transplant 2009 ; 24 : 1574 – 1579 Google Scholar CrossRef Search ADS PubMed 5 Onofriescu M , Hogas S , Voroneanu L et al. Bioimpedance-guided fluid management in maintenance hemodialysis: a pilot randomized controlled trial . Am J Kidney Dis 2014 ; 1 – 8 6 Zoccali C , Moissl U , Chazot C et al. Chronic fluid overload and mortality in ESRD . J Am Soc Nephrol 2017 ; 28 : 2491 – 2497 Google Scholar CrossRef Search ADS PubMed 7 Dekker MJE , Marcelli D , Canaud BJ et al. Impact of fluid status and inflammation and their interaction on survival: a study in an international hemodialysis patient cohort . Kidney Int 2017 ; 91 : 1214 – 1223 Google Scholar CrossRef Search ADS PubMed 8 Dekker MJE , Marcelli D , Canaud B et al. Unraveling the relationship between mortality, hyponatremia, inflammation and malnutrition in hemodialysis patients: results from the international MONDO initiative . Eur J Clin Nutr 2016 ; 70 : 779 – 784 Google Scholar CrossRef Search ADS PubMed 9 Robinson BM , Tong L , Zhang J et al. Blood pressure levels and mortality risk among hemodialysis patients in the dialysis outcomes and practice patterns study . Kidney Int 2012 ; 82 : 570 – 580 Google Scholar CrossRef Search ADS PubMed 10 Foley RN , Parfrey PS , Harnett JD et al. Impact of hypertension on cardiomyopathy, morbidity and mortality in end-stage renal disease . Kidney Int 1996 ; 49 : 1379 – 1385 Google Scholar CrossRef Search ADS PubMed 11 Nongnuch A , Campbell N , Stern E et al. Increased postdialysis systolic blood pressure is associated with extracellular overhydration in hemodialysis outpatients . Kidney Int 2014 ; 87 : 1 – 6 12 Heerspink HLJ , Ninomiya T , Zoungas S et al. Effect of lowering blood pressure on cardiovascular events and mortality in patients on dialysis: a systematic review and meta-analysis of randomised controlled trials . Lancet 2009 ; 373 : 1009 – 1015 Google Scholar CrossRef Search ADS PubMed 13 Odudu A , McIntyre C. Volume is not the only key to hypertension control in dialysis patients . Nephron Clin Pract 2012 ; 120 : c173 – c177 Google Scholar CrossRef Search ADS PubMed 14 Li Z , Lacson E , Lowrie EG et al. The epidemiology of systolic blood pressure and death risk in hemodialysis patients . Am J Kidney Dis 2006 ; 48 : 606 – 615 Google Scholar CrossRef Search ADS PubMed 15 SPRINT Research Group , Wright JT , Williamson JD et al. A randomized trial of intensive versus standard blood-pressure control . N Engl J Med 2015 ; 373 : 2103 – 2116 Google Scholar CrossRef Search ADS PubMed 16 Zager PG , Nikolic J , Brown RH et al. ‘U’ curve association of blood pressure and mortality in hemodialysis patients . Kidney Int 1998 ; 54 : 561 – 569 Google Scholar CrossRef Search ADS PubMed 17 Kalantar-Zadeh K , Kilpatrick RD , McAllister CJ et al. Reverse epidemiology of hypertension and cardiovascular death in the hemodialysis population . Hypertension 2005 ; 45 : 811 Google Scholar CrossRef Search ADS PubMed 18 Flythe JE , Assimon MM , Wenger JB et al. Ultrafiltration rates and the quality incentive program: proposed measure definitions and their potential dialysis facility implications . Clin J Am Soc Nephrol 2016 ; 11 : 1422 Google Scholar CrossRef Search ADS PubMed 19 Usvyat LA , Haviv YS , Etter M et al. The MONitoring Dialysis Outcomes (MONDO) initiative . Blood Purif 2013 ; 35 : 37 – 48 Google Scholar CrossRef Search ADS PubMed 20 Port FK , Hulbert-Shearon TE , Wolfe RA et al. Predialysis blood pressure and mortality risk in a national sample of maintenance hemodialysis patients . Am J Kidney Dis 1999 ; 33 : 507 – 517 Google Scholar CrossRef Search ADS PubMed 21 Marcelli D , Usvyat L. A , Kotanko P et al. Body composition and survival in dialysis patients: results from an international cohort study . Clin J Am Soc Nephrol 2015 ; 10 : 1192 – 1199 Google Scholar CrossRef Search ADS PubMed 22 United Nations Statistics Division- Standard Country and Area Codes Classifications (M49) [cited 23 January 2016] http://unstats.un.org/unsd/methods/m49/m49regin.htm#europe 23 Saran R , Bragg-Gresham JL , Levin NW et al. Longer treatment time and slower ultrafiltration in hemodialysis: associations with reduced mortality in the DOPPS . Kidney Int 2006 ; 69 : 1222 – 1228 Google Scholar CrossRef Search ADS PubMed 24 Kalantar-Zadeh K , Kilpatrick RD , Kopple JD. Reverse epidemiology of blood pressure in dialysis patients . Kidney Int 2005 ; 67 : 2067 , author reply 2067–2068 Google Scholar CrossRef Search ADS PubMed 25 Bansal N , McCulloch CE , Rahman M et al. Blood pressure and risk of all-cause mortality in advanced chronic kidney disease and hemodialysis: the chronic renal insufficiency cohort study . Hypertension 2015 ; 65 : 93 – 100 Google Scholar CrossRef Search ADS PubMed 26 Jaeger JQ , Mehta RL. Assessment of dry weight in hemodialysis: an overview . J Am Soc Nephrol 1999 ; 10 : 392 – 403 Google Scholar PubMed 27 Kalantar-Zadeh K , Regidor DL , Kovesdy CP et al. Fluid retention is associated with cardiovascular mortality in patients undergoing long-term hemodialysis . Circulation 2009 ; 119 : 671 – 679 Google Scholar CrossRef Search ADS PubMed 28 Agarwal R. Body mass index-mortality paradox in hemodialysis: can it be explained by blood pressure? Hypertension 2011 ; 58 : 1014 – 1020 Google Scholar CrossRef Search ADS PubMed 29 McIntyre CW , Harrison LEA , Eldehni MT et al. Circulating endotoxemia: a novel factor in systemic inflammation and cardiovascular disease in chronic kidney disease . Clin J Am Soc Nephrol 2011 ; 6 : 133 – 141 Google Scholar CrossRef Search ADS PubMed 30 Chazot C , Vo-Van C , Deleaval P et al. Predialysis systolic blood pressure evolution in incident hemodialysis patients: effects of the dry weight method and prognostic value . Blood Purif 2012 ; 33 : 275 – 283 Google Scholar CrossRef Search ADS PubMed 31 Tentori F , Hunt WC , Rohrscheib M et al. Which targets in clinical practice guidelines are associated with improved survival in a large dialysis organization? J Am Soc Nephrol 2007 ; 18 : 2377 – 2384 Google Scholar CrossRef Search ADS PubMed 32 McIntyre CW. Recurrent circulatory stress: the dark side of dialysis . Semin Dial 2010 ; 23 : 449 – 451 Google Scholar CrossRef Search ADS PubMed 33 Agarwal R. Volume-associated ambulatory blood pressure patterns in hemodialysis patients . Hypertension 2009 ; 54 : 241 – 247 Google Scholar CrossRef Search ADS PubMed 34 Termorshuizen F , Dekker FW , van Manen JG et al. Relative contribution of residual renal function and different measures of adequacy to survival in hemodialysis patients: an analysis of the Netherlands Cooperative Study on the Adequacy of Dialysis (NECOSAD)-2 . J Am Soc Nephrol 2004 ; 15 : 1061 – 1070 Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Nephrology Dialysis TransplantationOxford University Press

Published: Apr 27, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off