Abstract
Background
High interdialytic weight gain (IDWG) and ultrafiltration rate have been associated with increased mortality in patients undergoing haemodialysis. Cardiac biomarkers reflecting strain, including N-terminal pro-B-type natriuretic peptide (NT-proBNP) and high-sensitivity troponin T (TnT), are known to increase during haemodialysis sessions. Previous studies suggest that these intradialytic increases may be related to ultrafiltration rate, with a proposed threshold of 0.6 IDWG per hour. This study therefore aimed to prospectively evaluate whether reducing ultrafiltration rate attenuates the haemodialysis-associated increase in these cardiac biomarkers.
Method
In this multicentre, prospective study prevalent haemodialysis patients with a relative IDWG > 2.5% of body weight and ultrafiltration rate ≥ 0.72 IDWG per hour underwent one standard and one prolonged haemodialysis session with a ≥ 20% reduction in ultrafiltration rate, targeting an ultrafiltration rate ≤ 0.6 IDWG per hour. NT-proBNP and TnT were analysed directly before dialysis, at 180 minutes, and at the end of each haemodialysis session.
Results
Of 238 screened patients, 65 met the inclusion criteria and none of the exclusion criteria. Of these, 56 were enrolled and 41 included in the final analysis, meeting the pre-specified sample size required to detect a clinically relevant difference. NT-proBNP and TnT increased significantly during both dialysis sessions (p < 0.001). However, no difference was observed in ∆NT-proBNP (p = 0.967) or ∆TnT (p = 0.823) between treatments with different ultrafiltration rates. Achieving an ultrafiltration rate ≤ 0.6 IDWG per hour required a mean treatment duration of 423 ± 139 minutes (min 246, max 775). Among the 12 patients (29.3%) who achieved the targeted ultrafiltration rate of ≤0.6 IDWG per hour, changes in NT-proBNP (p = 0.062) and TnT (p = 0.114) did not differ significantly between sessions with different ultrafiltration rates.
Conclusions
In stable, prevalent haemodialysis patients who tolerate their ordinary treatment, reducing the ultrafiltration rate by at least 20% did not attenuate the increase in NT-proBNP and TnT during a single haemodialysis session. In these patients, achieving an ultrafiltration rate ≤ 0.6 IDWG/h generally required treatment durations that are not feasible in routine in-centre haemodialysis.
Trial registration
Registered at clinicaltrials.gov (NCT06153888) November 6 2023 (Retrospectively registered).
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12882-026-05364-4.
Keywords: Haemodialysis, Heart, NT-proBNP, Troponin T, Ultrafiltration rate
Introduction
The prevalence of cardiovascular disease (CVD) rises as kidney function declines and remains a leading cause of morbidity and mortality in patients receiving haemodialysis [1–3]. These patients have a substantial burden of traditional, but also uraemia-specific, cardiovascular risk factors, such as anaemia, fluid overload, retention of uraemic toxins, and disturbances in calcium-phosphate balance, all of which promote CVD [4].
Haemodialysis itself contributes to structural and functional changes in the heart. Repeated episodes of intradialytic hypotension, reduced myocardial blood flow, and myocardial stunning can, over time, lead to permanent myocardial injury [5]. Such injury is thought to contribute to the development of myocardial fibrosis [6] and the increase in left ventricular mass index (LVMI) observed in many individuals undergoing long-term haemodialysis [7].
Overhydration is associated with impaired left ventricular diastolic function [8], increased LVMI [9], and mortality [10]. High inter-dialytic weight gain (IDWG) is associated with an increased risk of cardiac events [11] and death [12]. To manage fluid overload, ultrafiltration is required during haemodialysis. High IDWG often necessitates higher ultrafiltration rates, which have been linked to an increased mortality in several large retrospective cohort studies [13–17]. The risk appears to rise already from ultrafiltration rates as low as 6 ml/h/kg [16]. Imaging studies using magnetic resonance imaging (MRI) or echocardiograms suggest that the acute cardiac effects of haemodialysis and the extent of the functional and structural damages seen during a haemodialysis session correlated with increased ultrafiltration rate [18, 19].
N-terminal-pro-B-type natriuretic peptide (NT-proBNP) and high-sensitivity cardiac Troponin T (TnT) are well-established markers of cardiac strain and injury [20] and have been linked to increased mortality [21–23]. NT-proBNP release is induced by stretching of the myocardial wall, often as a response to volume overload [24] whereas TnT mainly is released following myocardial cell damage [25]. Levels of both markers increase during haemodialysis, even after correction for haemoconcentration [26], particularly in patients with IDWG over 2.5% of body weight [27]. However, the increase in cardiac markers during haemodialysis is masked when using high flux dialysers, making it necessary to use low flux dialysers for study purposes [26]. In a study using high flux dialysers, a lower reduction during haemodialysis was a risk factor for mortality [28]. Even in healthy individuals, small, acute increases in TnT affects mortality risk [29]. A previous study found a positive correlation between the ultrafiltration rate and the increase in NT-proBNP during a single haemodialysis session, defining a breaking point of the ultrafiltration rate relative to body weight exceeding 0.6 IDWG/h [30].
Previous studies examining the effects of different ultrafiltration rates have been largely observational and retrospective in their nature, limiting the ability to isolate the impact of ultrafiltration rate from other patient-related factors. Using a prospective, crossover design in which patients serve as their own controls, minimizes these confounders.
The aim of this study was to determine whether reducing the ultrafiltration rate, by longer dialysis, attenuates the increase in NT-proBNP and TnT seen during a haemodialysis session.
Materials and methods
The study was a multicentre, prospective, self-controlled crossover study (NCT06153888 registered at clinicaltrials.gov November 6 2023) conducted at five haemodialysis centres in Sweden. Chronic haemodialysis patients were enrolled between January 2023 and October 2025. Ethical approval was obtained from the Swedish Ethical Review Authority (2012–42-31 M, revised Dnr 2023–06921-02) and complied with the Declaration of Helsinki. All participants provided written informed consent. The study is reported in accordance with the CONSORT guidelines.
Patient selection
All patients at the participating sites were screened. In this study, IDWG was defined as interdialytic weight gain relative to target weight. Inclusion criteria were IDWG of more than 2.5% and a relative ultrafiltration rate over 0.72 IDWG/h. Eligible patients were to perform one standard dialysis and one prolonged haemodialysis session with a ≥ 20% reduction in ultrafiltration rate, targeting an ultrafiltration rate ≤ 0.6 IDWG per hour.
Patients were excluded if they were under 18 years of age, had cognitive impairment preventing them from providing written, informed consent, used single-needle-dialysis, had medical reasons why they could not undergo dialysis treatment with a low flux dialyser or were clinically unstable (terminally ill, suffering from acute infection or anything else making it unlikely for them to keep their regular dialysis schedule without significant changes during the study period).
Variables
NT-proBNP and TnT were used as markers for cardiac strain. Background variables included age, sex, renal diagnosis, dialysis vintage, comorbidities, ordinary dialysis regimen, haemodialysis access, weight, and height. Pulse and blood pressure were recorded during treatment. Clotting of the dialyser was recorded and graded as clear, some degree of residual clotting, or coagulated. Adverse events were defined as symptomatic intradialytic hypotension (patient complaints of symptoms associated with hypotension in the presence of drop of blood pressure or hypotension needing intervention), intradialytic muscle cramps, and complete coagulation resulting in interruption of treatment.
From echocardiographs, left ventricular ejection fraction (EF), left ventricular dimensions, diastolic function, and aortic- and mitral valve dysfunctions were recorded. From electrocardiograms (ECG), QRS-duration was recorded.
Outcomes
The main outcome was the difference of the change (∆) in NT-proBNP- and TnT-levels during haemodialysis sessions with ordinary ultrafiltration rate compared with reduced ultrafiltration rate.
Secondary outcomes included correlations between the primary outcomes and LVMI, EF, valve dysfunctions, QRS duration, and comorbidities as well as adverse events during haemodialysis.
Samples size calculation
Data from Goto et al. [30] was used for power calculation. Based on clinical experience, an effect size of 0.6 was deemed clinically relevant for the difference in change of NT-proBNP during a haemodialysis session. Using two-tailed paired t-test, this would require 40 patients to achieve a power of 95% at 0.05 significance level.
Study design
Each patient underwent two haemodialysis sessions, the same day of the week, in two consecutive weeks. Whether the study dialyses were performed after the long or the short interval (mid-week) depended on the day the patient was most likely to reach their target weight and still fulfilled the inclusion criteria. For a more precise calculation of dialysis session duration, it was preferred that the standard session was performed during week one and the prolonged session during week two. However, if this was not possible because of scheduling difficulties, the standard session could instead be performed one week after the prolonged session. One haemodialysis session was performed with the ordinary dialysis prescription, and one session with prolonged time and reduced ultrafiltration rate. To minimise confounding, the subsequent session within the study was planned one week later with change of only the treatment time. All other aspects of the dialysis prescription (dialyser, blood flow, dialysate flow, temperature, and dialysate composition) were kept unchanged. Between the two sessions, target weight and blood pressure medications were not altered.
In the case of prolonged session, the principal investigator at each site was free to increase the dose of anticoagulation as deemed appropriate to reduce the risk of clotting.
All haemodialysis sessions were performed with low flux dialysers (FX10® or FX8®, Fresenius Medical Care GmbH, Bad Homburg, Germany).
Target weight was adjusted before the first study session to ensure accurate target weight. This was done according to clinical practice at the participating sites.
Determination of treatment time
Standard dialysis time was based on routine prescription. The study dialysis was prolonged to achieve a ≥ 20% reduction in ultrafiltration rate compared with the standard dialysis. If possible, an ultrafiltration rate of ≤0.6 IDWG/h was targeted. This approach was based on findings from a previous study indicating that the increase in NT-proBNP differed between patients with ultrafiltration rates above versus below this threshold [30]. In all cases, the selected duration approximated the target as closely as feasible within operational constraints. The calculation formula is provided in Supplementary Item 1.
Data collection
Blood samples for analysis of NT-proBNP, TnT, urea and haematocrit were drawn before treatment start, after 180 minutes, and five minutes before the end of treatment. The blood samples were immediately sent to the local laboratory for analysis. NT-proBNP levels above the upper limit of quantitation were diluted until quantifiable measurements could be made. Cardiac biomarkers taken during or at the end of haemodialysis were corrected for haemoconcentration using Schneditz formula for plasma components [31].
Pulse and blood pressure were taken using the automatic cuff linked to the dialysis machine at the same time points as the blood samples. The attending nurse clinically evaluated the pulse as regular or irregular. Parameters regarding the haemodialysis treatment, including adverse events, were recorded during treatment by the attending nurse.
Background data were collected by the principal investigator at each unit.
Echo- and electrocardiograms were considered for this study when they were performed under stable conditions within the last year. LVMI was calculated using the formula of Teichholz [32] for left ventricular mass (LVM) and indexed to body surface area.
Statistical analysis
The effects of different ultrafiltration rates were assessed according to treatment group, standard vs. prolonged dialysis session. The analysis regarding cardiac markers during or after session was based on values adjusted for haemoconcentration.
Data were reported as mean and standard deviation (SD), median and interquartile range (IQR) or frequencies (%) as appropriate. Statistical analysis was performed using Student’s paired t-test for comparison between treatments or Wilcoxon signed rank test for not normally distributed data. Spearman´s rank correlation coefficient was calculated for non-normally distributed data. Linear mixed models were fitted to log-transformed NT-proBNP and TnT values, with treatment (short vs. long session), timepoint (baseline, 180 min, end), and their interaction as fixed effects, and a random intercept for patients. Fixed effects were tested using Type III tests with Satterthwaite degrees of freedom. Significance level was set to p < 0.05.
All statistical analyses were made using IBM SPSS Statistics for Windows version 30.0.0.0 (IBM Corp., Armonk, NY, USA).
Results
A total of 238 patients were screened, of whom 65 patients provided informed consent and were initially enrolled. At one site, nine patients were lost prior to recruitment. During the study, an additional 15 patients were excluded, leaving 41 patients for the final analysis. The study flow-chart is presented in Fig. 1.
Fig. 1.
Study flow
Baseline characteristics are presented in Table 1. All patients achieved a reduction of the ultrafiltration rate of ≥20% compared to the standard session. For one patient, NT-proBNP at the end of the treatment was missing. One site (n = 8) had no TnT available for analysis. There was a significant increase in NT-proBNP and TnT during both sessions (p < 0.001 for NT-proBNP and TnT during both sessions). Levels of cardiac markers and change during sessions are presented in Table 2. When comparing sessions with ordinary and prolonged treatment time, the difference in ∆NT-proBNP was 37.3 (IQR 1240) ng/l and the difference in ∆TnT was −0.4 (IQR 12.7) ng/l (p = 0.967 for NT-proBNP and p = 0.823 for TnT). When comparing measurements at 180 minutes and end of treatment there was a continued increase in NT-proBNP during both sessions (p < 0.001 for ordinary treatment and p = 0.002 for prolonged treatment). There was no difference in ∆TnT when comparing measurements at 180 minutes and end of treatment (p = 0.114 for ordinary treatment and p = 0.228 for prolonged treatment). Details are presented in Table 2.
Table 1.
Baseline characteristics
| Age (years) | 63.5 | SD 14.8 |
|---|---|---|
| Vintage (months) | 32.9 | IQR 44.5 |
| Height (cm) | 170.9 | SD 9.5 |
| Weight (kg) | 74.5 | SD 13.8 |
| BMI (kg/m2) | 25.7 | SD 5.4 |
| Sex | ||
| Men | 28 | 68.3% |
| Women | 13 | 31.7% |
| Primary kidney diagnosis | ||
| Diabetes nephropathy | 14 | 34.1% |
| Glomerulonephritis | 10 | 24.4% |
| Other | 7 | 17.1% |
| Unknown | 3 | 7.3% |
| Polycystic kidney disease | 3 | 7.3% |
| Tubulointerstitial nephritis | 2 | 4.9% |
| Nephrosclerosis | 2 | 4.9% |
| Comorbidities | ||
| Previous myocardial infarction | 7 | 17.1% |
| Persistent or permanent atrial fibrillation | 6 | 14.6% |
| Pacemaker | 2 | 4.9% |
| Previous stroke | 10 | 24.4% |
| Diabetes mellitus | 18 | 43.9% |
| Type 1 | 8 | 19.5% |
| Type 2 | 10 | 24.4% |
| Haemodialysis access** | ||
| Central venous catheter | 28 | 68.3% |
| Forearm arteriovenous fistula | 10 | 24.4% |
| Upper arm arteriovenous fistula | 3 | 7.3% |
| Arteriovenous graft | 2 | 4.8% |
| Ordinary treatment regimen | ||
| Haemodiafiltration | 28 | 68.3% |
| Haemodialysis, high flux dialyser | 12 | 29.3% |
| Haemodialysis, medium cut-off dialyser | 1 | 2.4% |
| Treatments per week | ||
| One | 1 | 2.4% |
| Two | 1 | 2.4% |
| Three | 32 | 78.0% |
| Four | 7 | 17.1% |
| Time per treatment (minutes) | 246 | SD 23.0 |
Data presented as mean (SD), median (IQR) or number (%)
** 2 patients had simultaneous use of catheter and functioning arteriovenous fistula
Table 2.
Differences between treatments
| Standard session | Prolonged session | p | |||
|---|---|---|---|---|---|
| Median | IQR | Median | IQR | ||
| NT-proBNP before treatment (baseline) (ng/l) | 5850.0 | 20950.0 | 4370.0 | 22467.0 | 0.789 |
| ∆NT-proBNP start-180 min of treatment (ng/l)* | 784.2 | 4022.3 | 1141.1 | 3202.2 | 0.936 |
| ∆NT-proBNP start-end of treatment (ng/l)* | 2244.0 | 6342.0 | 1468.9 | 3961.7 | 0.967 |
| Tn T before treatment (baseline) (ng/l) | 70.0 | 44.0 | 69.0 | 49.0 | 0.322 |
| ∆TnT start-180 min of treatment (ng/l)* | 5.4 | 8.4 | 5.6 | 11.8 | 0.348 |
| ∆TnT start-end of treatment (ng/l)* | 5.0 | 8.6 | 4.4 | 9.1 | 0.823 |
| Mean | SD | Mean | SD | p | |
| Interdialytic weight gain (%) | 4.3 | 1.3 | 4.1 | 1.3 | 0.342 |
| Expected ultrafiltration (l) | 3.1 | 0.9 | 3.2 | 0.8 | 0.842 |
| Delivered treatment time (min) | 247 | 25 | 334 | 37 | <0.001 |
| Ultrafiltration rate (IDWG/h) | 0.97 | 0.18 | 0.75 | 0.21 | <0.001 |
| Kt/V | 1.46 | 0.27 | 1.90 | 0.50 | <0.001 |
| Total dose of low molecular heparin (IE) | 4506.2 | 1829.7 | 6032.5 | 2614.8 | <0.001 |
| Haematocrit before treatment (%) | 34.2 | 3.7 | 34.0 | 4.0 | 0.652 |
| Haematocrit after 180 min of treatment (%) | 35.6 | 4.0 | 34.4 | 4.0 | <0.001 |
| Haematocrit at end of treatment (%) | 36.2 | 4.0 | 36.0 | 4.1 | 0.352 |
| Systolic blood pressure before treatment (mmHg) | 149.0 | 21.9 | 146.6 | 20.0 | 0.181 |
| Systolic blood pressure after 180 min of treatment (mmHg) | 128.3 | 19.9 | 127.6 | 18.4 | 0.905 |
| Systolic blood pressure at end of treatment (mmHg) | 128.8 | 17.0 | 131.3 | 21.0 | 0.521 |
| Diastolic blood pressure before treatment (mmHg) | 75.7 | 20.0 | 76.0 | 13.5 | 0.971 |
| Diastolic blood pressure after 180 min of treatment (mmHg) | 70.0 | 12.1 | 69.6 | 13.6 | 0.924 |
| Diastolic blood pressure at end of treatment (mmHg) | 72.0 | 13.8 | 70.0 | 13.0 | 0.294 |
| Pulse before treatment (bpm) | 69.6 | 11.5 | 69.5 | 9.6 | 0.951 |
| Pulse after 180 min of treatment (bpm) | 68.6 | 10.9 | 68.2 | 8.1 | 0.750 |
| Pulse at end of treatment (bpm) | 68.8 | 10.6 | 71.1 | 15.6 | 0.353 |
| n/N | % | n/N | % | p | |
| Symptomatic hypotension | 2/33 | 6.1 | 4/33 | 12.1 | 0.500 |
| Cramps | 6/33 | 18.2 | 6/33 | 18.2 | 1 |
| Dialyser appearance after treatment | |||||
| Clear | 4/41 | 9.8 | 4/41 | 9.8 | 1 |
| Some degree of clotting | 37/41 | 90.2 | 37/41 | 90.2 | 1 |
| Coagulation | 0 | 0 | 0 | 0 | 1 |
Comparison between haemodialysis sessions with standard treatment time and prolonged treatment time. NT-proBNP: N-terminal proB-Type natriuretic peptide. TnT: High sensitivity Troponin T. *Values at 180 minutes and end of treatment are adjusted for haemoconcentration
In the linear mixed models, the treatment × timepoint interaction was not statistically significant (F (2,198) = 0.001, p = 0.99 for NT-proBNP, F(2,160) = 0.57, p = 0.56 for TnT), indicating no evidence that the trajectory of cardiac biomarkers over the session differed between short and long dialysis. There was a significant main effect of timepoint (F(2,198) = 49.1, p < 0.001 for NT-proBNP, F(2,160) = 15.0, p < 0.001 for TnT), reflecting a biomarker increase from baseline to end of session in both groups, but no significant main effect of treatment (F(1,198) = 0.44, p = 0.50 for NT-proBNP and F(1,160) = 2.23, p = 0.137 for TnT). The 95% CI derived from estimated marginal means for the interaction between ultrafiltration rate and time of measurement is presented in Supplement Table 1 (NT-proBNP) and Supplement Table 2 (TnT).
Unadjusted values are presented in Supplement Table 3. Individual values are presented in Supplement Figure 1 (NT-proBNP) and Supplement Figure 2 (TnT).
To achieve an ultrafiltration rate below 0.6 IDWG/h, a mean time of 423 ± 139 (min 246, max 775) minutes was needed. The maximum available time slot for a treatment was limited to 6–6.5 hours due to practical constrains in the haemodialysis units. A total of 12 patients (29.3%) reached an ultrafiltration rate below or equal to 0.6 IDWG/h. When analysing only these patients, there was no difference in the change of cardiac biomarkers (p = 0.062 for ∆NT-proBNP and p = 0.114 for ∆TnT). Among the 12 patients, only one demonstrated a clinically meaningful attenuation in the rise of NT-proBNP (>25 percentage points) during the prolonged session relative to the standard session.
One patient performed the prolonged session before the standard session.
In addition to clinical examination target weight was determined with the guidance of blood volume monitoring in one (2.4%) patient, NT-proBNP in 13 (31.7%) patients and echocardiography in one (2.4%) patient. Chest X-ray or lung ultrasound were not used. Bioimpedance measurement was used in 32 (78%) patients. In these patients the mean difference between target weight and normohydrated weight according to bioimpedance measurements was 1.5 ± 3.8 kg.
Echocardiographs were available for 19 patients (46.3%), of whom 14 (34.1%) had information on left ventricular measurements. Only a small number of patients exhibited more than mild valve dysfunctions. Details are presented in Supplement Table 4. Due to the small numbers in subgroups, no statistical testing was performed. However, individual data did not reveal any clear pattern suggesting subgroup differences.
An ECG was available in 40 patients of whom 10 had a QRS duration over 120 ms. Comparisons between patients with QRS < 120 ms vs ≥ 120 ms showed no difference in ∆NT-proBNP (p = 0.092) or ∆TnT (p = 0.445). Among patients with QRS ≥ 120 ms, there was no significant difference in biomarker changes between sessions with ordinary vs reduced ultrafiltration rate (p = 0.575 for ∆NT-proBNP, p = 0.779 for ∆TnT).
There was a positive correlation between age and ∆NT-proBNP during standard (R = 0.437, p = 0.005) and prolonged (R = 0.371, p = 0.018) sessions. Age also correlated with baseline levels of both cardiac markers (R = 0.477, p = 0.002 for NT-proBNP and R = 0.396, p = 0.023 for TnT). No correlation was observed between LVMI and ∆NT-proBNP (R = 0.253, p = 0.405 for ordinary treatment, R = 0.220, p = 0.471 for prolonged treatment). Detailed correlation data are presented in Supplement Table 5.
The number of adverse events was low during both treatments (Table 2).
Discussion
In this crossover study, we investigated whether reducing the ultrafiltration rate modifies haemodialysis-induced increase in cardiac biomarkers. We found no difference in the changes of NT-proBNP and TnT between treatments with higher vs lower ultrafiltration rates. Consistent with previous observations, both NT-proBNP and TnT increased during haemodialysis sessions.
To our knowledge, this is the first prospective study to intentionally modify ultrafiltration rate in haemodialysis patients. Acute cardiac effects of haemodialysis, including myocardial stunning and the development of regional wall motion abnormalities, have been previously demonstrated [33]. Ultrafiltration has been shown to reduce cardiac output and increases systemic vascular resistance even in patients without previous cardiovascular disease [34]. Observational studies suggest a correlation between ultrafiltration rate and both functional cardiac changes [18, 19] and mortality [13–16].
Ultrafiltration rate is closely linked to IDWG, as accumulated fluid must be removed within a limited timeframe. Since most previous studies are observational, it remains unclear whether the adverse outcomes are directly attributable to high ultrafiltration rates or reflect volume overload. This is supported by previous findings indicating that persistent overhydration over time rather than dialysis frequency is associated with cardiac structural changes [9].
Patient characteristics may modify the haemodynamic impact of ultrafiltration. Previous data have suggested that age [35] and body size [36] may influence vulnerability. In the present study, only age was significantly correlated with ∆NT-proBNP. Although higher ultrafiltration rates have been linked to worse outcomes in patients with cardiac enlargement [37] we found no evidence that baseline NT-proBNP or TnT modified biomarker responses. There was no correlation between change in cardiac biomarkers and LVMI, however the number of available echocardiographs were limited.
A clear separation in ultrafiltration rates between treatments was achieved with target levels comparable to previously reported thresholds (6 ml/h/kg) [16]. Notably, half of the screened patients did not meet the inclusion criteria due to low IDWG or ultrafiltration rate during ordinary treatment, suggesting that high ultrafiltration rates are already avoided in clinical practice. Dialysis prescriptions were individualised across centres, with treatment frequency ranging from one to four sessions per week. Adverse events were infrequent, likely reflecting clinical optimisation of ultrafiltration rates to rates well tolerated by each patient. Intradialytic hypotension is a risk factor for mortality [38] and reducing ultrafiltration rates to avoid hypotension is probably beneficial, if this effects biomarker response is however unclear.
Only one third of participants reached the predefined ultrafiltration rate (≤0.6 IDWG/h) during prolonged sessions. In this subgroup, a tendency towards a smaller increase in NT-proBNP was noted, although this was clinically meaningful only for one patient. Determining ultrafiltration rate in dialysis patients needs to be individualised. A prolongation of dialysis time might be beneficial for some patients. However, for most patients, the treatment duration required to reach an ultrafiltration rate below 0.6 IDWG/h was not feasible in standard in-centre haemodialysis.
Previous singular centre data suggest improved survival with long dialysis sessions, although this may reflect centre-specific factors, including lower ultrafiltration rate and lower blood pump flow [39]. A randomized controlled trial has shown that that increased dialysis frequency may reduce LVMI [40], although this effect was not shown with nocturnal dialysis [41]. Increased dialysis intensity has also been associated with reduction in regional wall motion abnormalities [18]. These treatment modalities were not evaluated in our study.
The adverse effects of haemodialysis are likely multifactorial. Prolonged treatment duration leads to a prolonged exposure to the extracorporeal circuit, which may offset potential benefits of reducing ultrafiltration rate through a prolonged blood-membrane interaction inducing complement activation and coagulation [42], and exposure to microbubbles [43]. In our study no differences in cardiac marker trajectory were observed at 180 minutes between treatment arms. NT-proBNP continued to increase beyond 180 minutes during both sessions.
In clinical practice, reducing the ultrafiltration rate is most effectively achieved by limiting IDWG. However, this remains difficult and requires a comprehensive, patient-centred approach [44].
Limitations
This study has several limitations. The dialysis population is highly heterogeneous; however, a crossover design was used to minimise confounding. While we reached our prespecified sample size, the ultrafiltration rate target was only achieved in few patients. While the absence of differences in marker response suggests the main findings are robust, the study was underpowered for subgroup analyses.
Since lower ultrafiltration rate with unchanged IDWG can only be achieved by prolonging treatment time, this intervention cannot fully isolate the physiological effects of ultrafiltration rate itself. However, there were no differences in change of cardiac biomarkers at 180 minutes. Since both cardiac biomarkers are dialysable with high flux dialysers and even more so with haemodiafiltration it was necessary to use low flux dialysers for this study. To evaluate treatments using haemodiafiltration, other markers or imaging techniques would be required.
Cardiac effects were assessed using biomarkers alone to assess cardiac strain. Although NT-proBNP and TnT are easily available and prognostically relevant [45], they may not capture all aspects of dialysis-induced cardiac stress. A multimodal approach incorporating imaging (echocardiographs or magnetic resonance imaging) and continuous ECG monitoring would have strengthened the analysis.
The observed increase in TnT during haemodialysis was well below the threshold for myocardial infarction. However, TnT may increase in response to myocardial stress or transient ischemia [25, 46] and peak levels may occur after the dialysis session. Therefore, peak TnT levels may not have been captured. Nevertheless, intradialytic increases in TnT have been associated with impaired myocardial function [47]. In contrast, NT-proBNP is released more rapidly, and is expected to peak during or shortly after dialysis [48].
Echocardiographic data was only available for half of the study population and were collected retrospectively. Prospective and standardized echocardiographic assessment would have provided more robust mechanistic insights and is suggested for future studies.
Conclusion
In stable, prevalent haemodialysis patients who tolerate their ordinary treatment, reducing the ultrafiltration rate by at least 20% did not attenuate the increase in NT-proBNP and TnT during a single haemodialysis session. In these patients, achieving an ultrafiltration rate ≤ 0.6 IDWG/h generally required treatment durations that are not feasible in routine in-centre haemodialysis.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
We thank all participants for their time and contribution to this clinical trial. Special thanks to research nurses Kerstin Marttala and Christina Gull as well as all staff at the participating dialysis units for their assistance in conducting the study.
Abbreviations
- CVD
Cardiovascular disease
- LVM
Left ventricular mass
- LVMI
Left ventricular mass index
- EF
Ejection fraction
- NT-proBNP
N-terminal pro-B-type natriuretic peptide
- TnT
hs cardiac Troponin T
- IDWG
Interdialytic weight gain
Author contributions
Initiation and Conceptualization: BS. Planning and development of study protocol: EL, MO, EB. Data collection and execution of study on site: Linköping: FU, AI Skövde: AW, BS Uppsala: NT, HF, MKS Umeå: EL, MO Östersund: JF. Data curation: EL. Statistical analysis: EL. Writing original draft: EL and MO. Writing, reviewing and editing: EL, FU, NT, AW, AI, JF, EB, HF, AF, MKS, BS, MO. All authors have read and agreed on the final version of the manuscript.
Funding
Open access funding provided by Umea University. This study was funded by region Västerbotten’s Research & Learning Fund (ALF), VISARE NORR (Northern Counties’ Councils, Sweden), The Kidney Foundation (Stiftelsen för njursjuka), The Swedish Kidney Foundation (Njurfonden), The AstraZeneca scholarship in collaboration with the Swedish Kidney Association (Svensk Njurmedicinsk Förening - SNF).
Data availability
Due to its sensitive nature, data is not publicly available. Data may be available from the corresponding author upon reasonable request and after ethical review.
Declarations
Ethics approval and consent to participate
Ethical approval was obtained from the Swedish Ethical Review Authority (2012-42-31 M, revised Dnr 2023-06921-02) and complied with the Declaration of Helsinki. All participants provided written informed consent.
Consent for publication
Participants have consented to publication of aggregated data.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Stel VS, Boenink R, Astley ME, Boerstra BA, Radunovic D, Skrunes R, et al. A comparison of the epidemiology of kidney replacement therapy between Europe and the United States: 2021 data of the ERA registry and the USRDS. Nephrol Dial Transpl. 2024;39(10):1593–603. 10.1093/ndt/gfae040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Boerstra BA, Boenink R, Astley ME, Bonthuis M, Abd ElHafeez S, Arribas Monzon F, et al. The ERA registry annual report 2021: a summary. Clin Kidney J. 2024;17(2):sfad281. 10.1093/ckj/sfad281. [DOI] [PMC free article] [PubMed]
- 3.Minciunescu A, Genovese L, DeFilippi C. Cardiovascular alterations and structural changes in the setting of chronic kidney disease: a review of cardiorenal syndrome type 4. SN Compr Clin Med. 2022;5(1):15. 10.1007/s42399-022-01347-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ronco C, Haapio M, House AA, Anavekar N, Bellomo R. Cardiorenal syndrome. J Am Coll Cardiol. 2008;52(19):1527–39. 10.1016/j.jacc.2008.07.051. [DOI] [PubMed] [Google Scholar]
- 5.Burton JO, Jefferies HJ, Selby NM, McIntyre CW. Hemodialysis-induced repetitive myocardial injury results in global and segmental reduction in systolic cardiac function. Clin J Am Soc Nephrol. 2009;4(12):1925–31. 10.2215/CJN.04470709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gross ML, Ritz E. Non-coronary heart disease in dialysis patients: hypertrophy and fibrosis in the cardiomyopathy of Uremia—beyond coronary heart disease. Semin Dial. 2008;21(4):308–18. 10.1111/j.1525-139X.2008.00454.x. [DOI] [PubMed] [Google Scholar]
- 7.Nguyen NN, Van Duong P, Ngoc Mai TH, Vo NH, Luong DK, Ngo TH. Left ventricular Mass index in end-Stage renal disease patients during hemodialysis and continuous ambulatory peritoneal dialysis. Int J Clin practiceInt J Clin Pract. 2023;2023:1–7. 10.1155/2023/8816478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Han BG, Pak D, Kim JS, Sohn Y. The moderating effect of fluid overload on the relationship between the augmentation index and left ventricular diastolic function in patients with CKD. Sci Rep. 2024;14(1):480. 10.1038/s41598-023-50746-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Elsayed E, Farag YMK, Ravi KS, Chertow GM, Mc Causland FR. Association of changes in vector length with changes in left ventricular Mass among patients on maintenance hemodialysis: a secondary analysis of the frequent hemodialysis network daily trial. Kidney360. 2024;5(6):870–76. 10.34067/KID.0000000000000443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Christa M, Smyth B, Mayne KJ, Stuard S, Canaud B, Genser B, et al. Increased risk of cardiac death with cumulative exposure to fluid overload and dialysate sodium ≤138 mmol/l in hemodialysis patients. Clin Kidney J. 2025;18(10):sfaf259. 10.1093/ckj/sfaf259. [DOI] [PMC free article] [PubMed]
- 11.Holmberg B, Stegmayr BG. Cardiovascular conditions in hemodialysis patients may be worsened by extensive interdialytic weight gain. Hemodial Int. 2009;13(1):27–31. 10.1111/j.1542-4758.2009.00335.x. [DOI] [PubMed] [Google Scholar]
- 12.Westenbrink BD, Hovinga TK, Kloppenburg WD, Veeger NJ, Janssen WM. B-type natriuretic peptide and interdialytic fluid retention are independent and incremental predictors of mortality in hemodialysis patients. CN. 2011;76(11):373–79. 10.5414/CN106858. [DOI] [PubMed] [Google Scholar]
- 13.Saran R, Bragg-Gresham JL, Levin NW, Twardowski ZJ, Wizemann V, Saito A, et al. Longer treatment time and slower ultrafiltration in hemodialysis: associations with reduced mortality in the DOPPS. Kidney Int. 2006;69(7):1222–28. 10.1038/sj.ki.5000186. [DOI] [PubMed] [Google Scholar]
- 14.Movilli E, Gaggia P, Zubani R, Camerini C, Vizzardi V, Parrinello G, et al. Association between high ultrafiltration rates and mortality in uraemic patients on regular haemodialysis. A 5-year prospective observational multicentre study. Nephrol Dial Transpl. 2007;22(12):3547–52. 10.1093/ndt/gfm466. [DOI] [PubMed] [Google Scholar]
- 15.Flythe JE, Kimmel SE, Brunelli SM. Rapid fluid removal during dialysis is associated with cardiovascular morbidity and mortality. Kidney Int. 2011;79(2):250–57. 10.1038/ki.2010.383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Assimon MM, Wenger JB, Wang L, Flythe JE. Ultrafiltration rate and mortality in maintenance hemodialysis patients. Am J Kidney Dis. 2016;68(6):911–22. 10.1053/j.ajkd.2016.06.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Navarrete JE, Rajabalan A, Cobb J, Lea JP. Proportion of hemodialysis treatments with high ultrafiltration rate and the association with mortality. Kidney360. 2022;3, 3(8):1359–66. 10.34067/KID.0001322022. [DOI] [PMC free article] [PubMed]
- 18.Jefferies HJ, Virk B, Schiller B, Moran J, McIntyre CW. Frequent hemodialysis schedules are associated with reduced levels of dialysis-induced cardiac injury (myocardial stunning). Clin J Am Soc Nephrol. 2011;6(6):1326–32. 10.2215/CJN.05200610. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Burton JO, Jefferies HJ, Selby NM, McIntyre CW. Hemodialysis-induced cardiac injury: determinants and associated outcomes. Clin J Am Soc Nephrol. 2009;4(5):914–20. 10.2215/CJN.03900808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mahmood U, Johnson DW, Fahim MA. Cardiac biomarkers in dialysis. Aims Genet. 2017;4(1):1–20. 10.3934/genet.2017.1.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yamashita K, Mizuiri S, Nishizawa Y, Shigemoto K, Doi S, Masaki T. Addition of novel biomarkers for predicting all-cause and cardiovascular mortality in prevalent hemodialysis patients. Ther Apher Dial. 2018;22(1):31–39. 10.1111/1744-9987.12593. [DOI] [PubMed] [Google Scholar]
- 22.Kawagoe C, Sato Y, Toida T, Nakagawa H, Yamashita Y, Fukuda A, et al. N-terminal-pro-B-type-natriuretic peptide associated with 2-year mortality from both cardiovascular and non-cardiovascular origins in prevalent chronic hemodialysis patients. Ren Fail. 2018;40(1):127–34. 10.1080/0886022X.2018.1437047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Eriguchi M, Tsuruya K, Lopes M, Bieber B, McCullough K, Pecoits-Filho R, et al. Routinely measured cardiac troponin I and N-terminal pro-B-type natriuretic peptide as predictors of mortality in haemodialysis patients. ESC Heart Fail. 2022;9(2):1138–51. 10.1002/ehf2.13784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Daniels LB, Maisel AS. Natriuretic peptides. J Am Coll Cardiol. 2007;50(25):2357–68. 10.1016/j.jacc.2007.09.021. [DOI] [PubMed] [Google Scholar]
- 25.Chaulin AM. Cardiac troponins metabolism: from biochemical mechanisms to clinical practice (literature review). IJMS. 2021;22(20):10928. 10.3390/ijms222010928. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Laveborn E, Lindmark K, Skagerlind M, Stegmayr B. NT-proBNP and troponin T levels differ after haemodialysis with a low versus high flux membrane. Int J Artif Organs. 2015;38(2):69–75. 10.5301/ijao.5000387. [DOI] [PubMed] [Google Scholar]
- 27.Goto J, Forsberg U, Jonsson P, Matsuda K, Nilsson B, Nilsson Ekdahl K, et al. Interdialytic weight gain of less than 2.5% seems to limit cardiac damage during hemodialysis. Int J Artif Organs. 2021;44(8):539–50. 10.1177/0391398820981385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Pimenta J, Sampaio F, Martins P, Carvalho B, Rocha-Gonçalves F, Ferreira A, et al. Aminoterminal B-type natriuretic peptide (NT-proBNP) in end-stage renal failure patients on regular hemodialysis: does it have diagnostic and prognostic implications? Nephron Clin Pract. 2009;111(3):c182–8. 10.1159/000199458. [DOI] [PubMed]
- 29.Neumann JT, Twerenbold R, Ojeda F, Sorensen NA, Chapman AR, Shah ASV, et al. Application of high-sensitivity troponin in suspected myocardial infarction. N Engl J Med. 2019;380(26):2529–40. 10.1056/NEJMoa1803377. [DOI] [PubMed] [Google Scholar]
- 30.Goto J, Ott M, Stegmayr B. Myocardial markers are highly altered by higher rates of fluid removal during hemodialysis. Hemodial Int. 2024;28(1):17–23. 10.1111/hdi.13124. [DOI] [PubMed] [Google Scholar]
- 31.Schneditz D, Putz-Bankuti C, Ribitsch W, Schilcher G. Correction of plasma concentrations for effects of hemoconcentration or hemodilution. Asaio J. 2012;58(2):160–62. 10.1097/MAT.0b013e318243660f. [DOI] [PubMed] [Google Scholar]
- 32.Sveric KM, Cansız B, Winkler A, Ulbrich S, Ende G, Heidrich F, et al. Accuracy of Devereux and Teichholz formulas for left ventricular mass calculation in different geometric patterns: comparison with cardiac magnetic resonance imaging. Sci Rep. 2023;13(1):14089. 10.1038/s41598-023-41020-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Selby NM, McIntyre CW. The acute cardiac effects of dialysis. Semin Dial. 2007;20(3):220–28. 10.1111/j.1525-139X.2007.00281.x. [DOI] [PubMed] [Google Scholar]
- 34.Bos WJW, Bruin S, van Olden RW, Keur I, Wesseling KH, Westerhof N, et al. Cardiac and hemodynamic effects of hemodialysis and ultrafiltration. Am J Kidney Dis. 2000;35(5):819–26. 10.1016/S0272-6386(00)70250-2. [DOI] [PubMed] [Google Scholar]
- 35.Fernandez-Prado R, Pena-Esparragoza JK, Santos-Sanchez-Rey B, Pereira M, Avello A, Goma-Garces E, et al. Ultrafiltration rate adjusted to body weight and mortality in hemodialysis patients. Nefrologí(Engl Ed). 2021;41(4):426–35. 10.1016/j.nefroe.2021.10.005. [DOI] [PubMed] [Google Scholar]
- 36.Raimann JG, Wang Y, Mermelstein A, Kotanko P, Daugirdas JT. Ultrafiltration rate thresholds associated with increased mortality risk in Hemodialysis, unscaled or scaled to body size. Kidney Int Rep. 2022;7(7):1585–93. 10.1016/j.ekir.2022.04.079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yang LJ, Chao YL, Kuo IC, Niu SW, Hung CC, Chiu YW, et al. High ultrafiltration rate is associated with increased all-cause mortality in incident hemodialysis patients with a high Cardiothoracic ratio. J Pers Med. 2022;12(12):2059. 10.3390/jpm12122059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Kanbay M, Ertuglu LA, Afsar B, Ozdogan E, Siriopol D, Covic A, et al. An update review of intradialytic hypotension: concept, risk factors, clinical implications and management. Clin Kidney J. 2020;13(6):981–93. 10.1093/ckj/sfaa078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Charra B, Terrat JC, Vanel T, Chazot C, Jean G, Hurot JM, et al. Long thrice weekly hemodialysis: the Tassin experience. Int J Artif Organs. 2004;27(4):265–83. 10.1177/039139880402700403. [DOI] [PubMed] [Google Scholar]
- 40.Group FHNT, Chertow GM, Levin NW, Beck GJ, Depner TA, Eggers PW, et al. In-center hemodialysis six times per week versus three times per week. N Engl J Med. 2010;363(24):2287–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Rocco MV, Lockridge RS Jr, Beck GJ, Eggers PW, Gassman JJ, Greene T, et al. The effects of frequent nocturnal home hemodialysis: the frequent hemodialysis network nocturnal trial. Kidney Int. 2011;80(10):1080–91. 10.1038/ki.2011.213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Avila E, Sepulveda RA, Retamal J, Hachim D. Biocompatibility in hemodialysis: artificial membrane and human blood interactions. BMC Nephrol. 2025;26(1):482. 10.1186/s12882-025-04401-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Forsberg U, Jonsson P, Stegmayr B. The Emboless® venous chamber efficiently reduces air bubbles: a randomized study of chronic hemodialysis patients. Clin Kidney J. 2024;17(11):sfae323. 10.1093/ckj/sfae323. [DOI] [PMC free article] [PubMed]
- 44.Bossola M, Mariani I, Strizzi CT, Piccinni CP, Di Stasio E. How to limit interdialytic weight gain in patients on maintenance hemodialysis: state of the art and perspectives. J Clin Med. 2025;14(6):1846. 10.3390/jcm14061846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Yang WL, Fahim M, Johnson DW. Pathophysiology and significance of natriuretic peptides in patients with end-stage kidney disease. Clin Biochem. 2020;83:1–11. 10.1016/j.clinbiochem.2020.05.013. [DOI] [PubMed] [Google Scholar]
- 46.Arnadottir Á, Pedersen S, Bo Hasselbalch R, Goetze JP, Friis-Hansen LJ, Bloch-Munster AM, et al. Temporal release of high-sensitivity cardiac troponin T and I and copeptin after brief induced coronary artery balloon occlusion in humans. Circulation. 2021;143(11):1095–104. 10.1161/CIRCULATIONAHA.120.046574. [DOI] [PubMed] [Google Scholar]
- 47.Unlu S, Sahinarslan A, Sezenoz B, Uludag OM, Gokalp G, Seckin Ö, et al. High-sensitive troponin T increase after hemodialysis is associated with left ventricular global longitudinal strain and ultrafiltration rate. Cardiol J. 2020;27(4):376–83. 10.5603/CJ.a2018.0118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Liebetrau C, Gaede L, Dorr O, Troidl C, Voss S, Hoffmann J, et al. Release kinetics of N-terminal pro-B-type natriuretic peptide in a clinical model of acute myocardial infarction. Clinica (Rome) Acta. 2014;429:34–37. 10.1016/j.cca.2013.11.017. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Due to its sensitive nature, data is not publicly available. Data may be available from the corresponding author upon reasonable request and after ethical review.