Comparison of Overhydration with Ambulatory Hypertension in Pediatric Chronic Peritoneal Dialysis: A Prospective Observational Study

Introduction

Hydration status is a major determinant of cardiovascular morbidity and mortality in children on dialysis.^1,2 Establishing an optimal target weight is a critical component of the dialysis prescription, as over- or underestimation of the target weight could have deleterious consequences. In the pediatric population, target weight has traditionally been ascertained based on blood pressure (BP) control and the presence of clinical signs of overhydration (OH).³

In children on chronic peritoneal dialysis (CPD), including automated peritoneal dialysis (APD) and chronic ambulatory peritoneal dialysis (CAPD), OH is encountered in 30–45%.^4,5 This may not reflect the true prevalence as adequate assessment of OH in children is hampered by the absence of reliable, accurate, and objective tools as well as the lack of high-quality evidence supporting newer techniques. Adjusting CPD prescription based on clinical assessment of hydration status, despite being the most frequently used method, is subjective, inconsistent, and has significant limitations.⁶

Gold standard dilution methods for measuring body fluid volumes are invasive, cumbersome, expensive, and impractical in routine clinical practice. Lung ultrasound (LU) and bioimpedance analysis (BIA) are objective tools that have been shown to be superior to clinical evaluation and comparable to gold standard methods for assessment of hydration status in adult patients on dialysis.^7,8 The main difference between these tools lies in the fluid compartment being analyzed. LU primarily detects extravascular lung volume, while BIA provides information regarding intracellular fluid (ICF), extracellular fluid (ECF), and total body water (TBW).⁷ Bioimpedance spectroscopy (BIS), has similar efficacy as BIA for hydration status evaluation in patients with kidney failure.⁴

The measurements of TBW and ECW by BIA are based on predictive equations derived from healthy adult Caucasians and confounded by tissue hydration, limiting their use in children. These shortcomings of BIA are overcome by bioimpedance vector analysis (BIVA), a variant of BIA, which is a qualitative and semiquantitative graphical representation of resistance (R) and reactance (Xc), with an added advantage of combined tracking of hydration and nutritional status.⁹ These graphs play an integral role in adjusting target weight and have a proven efficacy in reducing mortality in adult patients on HD and CPD.¹⁰

There is limited evidence for the utility of these tools to detect subclinical OH in children on CPD and also a lack of validated reference BIVA graphs for children.¹¹ Uncontrolled hypertension (UH), hypertension requiring >2 antihypertensive medications, associated with cardiovascular morbidity and mortality, is reported in 75–90% of children on CPD.¹² Ambulatory blood pressure monitoring (ABPM) is the gold standard for detecting and monitoring hypertension in children on CPD.¹³

Management of hypertension in children on dialysis usually begins with addressing volume status. Although hypertension is considered as one of the universal clinical indicators of OH, there is ambiguity regarding the relationship between OH and hypertension in children on dialysis.⁴

It is imperative to identify bedside tools to assess OH in children on CPD and establish the relationship between hypertension and OH, ensuring appropriate modification to dialysis prescriptions and antihypertensive medications. The objectives of this study were to assess the burden of OH and UH, and to determine the association of OH (detected by BIVA or LU) with UH (by ABPM) in children on CAPD at baseline and short-term follow-up within three months.

Materials and Methods

A prospective observational study was conducted in children aged 5–18 years on CAPD on regular follow-up at the CKD clinic in St Johns Medical College Hospital, a tertiary pediatric nephrology center, from March to September 2021.

Clinically stable children on CAPD for more than one month were included, while those with clinical OH (defined below), metallic implants like nails, stents, or prostheses, amputees, or non-compliant with assessment and follow-up were excluded from the study. The study was approved by the Institutional Ethics Board (IEC 85/2021) and informed consent and assent (when applicable) were taken at enrollment.

Eligible children on CAPD were recruited and subjected to the assessment of OH by BIVA (OH-BIVA), OH by LU (OH-LU), and 24-hour ABPM monitoring. These measurements were repeated at one follow-up visit within 12 weeks of enrollment. The study protocol is depicted in Figure 1. The standard CAPD prescription, as decided by the treating team, consisted of 1.5% or 2.5% glucose-based bags with a dwell volume of 800–1200mL/m² and an average duration of 12–24 hours. Treatment changes were based on the clinician’s judgment as per the standard of care. Analysis was undertaken in only those children who completed the follow-up measurements.

Close

Figure 1:

Study protocol. ABPM: Ambulatory blood pressure monitoring, BIVA: Bioimpedance vector analysis, CAPD: Chronic ambulatory peritoneal dialysis, LU: Lung ultrasound, OH: Overhydration.

Export to PPT

Results

Twenty-five children were screened over six months, of whom six had clinical OH and were excluded. Nineteen children who were clinically stable were enrolled. Among these children, 15 [mean (±SD) age 10.3 ± 2.1 years, eight (53%) girls] completed the follow-up measurements and were included for analysis. The median (IQR) weight at baseline and follow-up was 21.4 (18.2, 23.2) kg and 22.15 (18.5,24) kg, respectively. The cause of kidney failure was of glomerular etiology in eight (53%). Preserved RKF was seen in seven (47%). The mean (±SD) duration on CAPD was 40.3 ± 29.1 months. Other clinical and sociodemographic details are shown in Table 1.

Burden of OH and hypertension at baseline and follow-up

In the study cohort, OH-BIVA was noted in three (20%), while five (33.3%) had OH-LU (one [6.7%] had OH-BIVA and OH-LU). Eight (58%) children in the cohort had significant clinical hypertension. UH was seen in 11 (73%). The proportion of children with OH-BIVA, OH-LU, significant clinical hypertension, and UH at baseline and follow-up is shown in Figure 2. The median (IQR) duration of follow-up was eight (6,12) weeks. During the follow-up assessment, clinical OH was seen in one (6.7%). OH-BIVA was seen in two (13%) and OH-LU was noted in three (20%) (one [6.7%] had OH-BIVA and OH-LU). The BIVA plots of the study population at baseline and follow-up are shown in Figure 3. Of the three children (two girls, one boy) with OH-BIVA at baseline, two continued to remain overhydrated while one was noted to have a reduction in OH. The tracking of hydration status is depicted in Figure 4. Significant clinical hypertension was seen in seven (47%) while UH was confirmed in nine (60%).

Close

Figure 2:

Burden of overhydration (OH) and uncontrolled ambulatory hypertension (UH) at baseline and follow-up. The proportion of children with OH by lung ultrasound (OH-LU) and bioimpedance vector analysis (OH-BIVA) and UH at baseline (blue bars) and follow-up (orange bars) are shown.

Export to PPT

Close

Figure 3:

Bioimpedance vector analysis (BIVA) graphs of the study population at baseline and follow-up. The bioimpedance measurements of resistance (R) and reactance (Xc) indexed to height (h) and standardized to bivariate z-scores (zR/h and zXc/h) of the study population at baseline (purple diamond) and follow-up (green triangle) plotted on the gender-specific reference graphs (upper panel: boys, lower panel: girls). The vector in the left lower quadrant beyond the 75th tolerance ellipse was defined as overhydration (OH) by BIVA.

Export to PPT

Close

Figure 4:

Tracking overhydration status by the bioimpedance vector analysis (BIVA). The figure shows the BIVA reference graph with the four quadrants: overhydration (OH), decreased body cell mass, dehydration, and increased body cell mass. Standardized bivariate resistance indexed to height (zR/h) is plotted on the x-axis and standardized bivariate reactance indexed to height (zXc/h) on the y-axis. The vector in the left lower quadrant beyond the 75th tolerance ellipse was defined as OH by BIVA. The figure depicts the change in hydration status among the three patients who were OH by BIVA at baseline (diamond). The hydration status at follow-up (within three months) is represented by a triangle. Patient 1 (blue) had an improvement in hydration status at follow-up (lengthening of vector moved within the 75th tolerance ellipse), while patient 2 (red) and patient 3 (grey) continued to remain overhydrated (beyond the 75th tolerance ellipse). No significant vector displacement at follow-up was noted (T2: 1.5).

Export to PPT

Modification to treatment was seen in nine (60%), based on the clinician’s judgment, of whom all had a change in CAPD prescription (change in the concentration of CAPD fluid from 1.5% to 2.5% and shortened cycles in five, addition of icodextrin in four) while only one had a change in antihypertension medications (increased dose of enalapril).

In those who had a modification to treatment made, six (67%) showed improvement in OH status while only one (11%) demonstrated well-controlled ambulatory hypertension. In the overall cohort at follow-up, the reduction in OH-BIVA was 33.3% and OH-LU was 40%, while reduction in UH was in 18% of children.

Discussion

Identifying bedside tools to detect subclinical OH and studying the association of OH with UH is critical in children on dialysis. This prospective study in children on CAPD showed that OH and UH were frequently encountered. There was no significant association between OH-BIVA or OH-LU with UH at baseline and short-term follow-up (three months).

In children on CPD, OH is a hidden and pervasive complication and appropriate volume control is one of the key goals of therapy.^4-6 The key component of an effective dialysis prescription is defining the target weight, which is a challenging task as clinical assessment of OH is often inaccurate and unreliable.^5,22 There is strong evidence to support the use of BIA and LU for assessment of hydration status in adult dialysis patients.^23-27 The newer bioimpedance device, BIS, has been studied in children on dialysis and demonstrated valuable in detecting OH.^14,28,29 However, BIVA is more cost effective and easily available compared to BIS. The other advantages of BIVA include the ability to pictorially track hydration and nutrition status, perform measurements with minimal training and negligible interobserver difference.^9,16,30 In adults on maintenance dialysis, the role of BIVA to guide prescriptions to reduce OH has been validated, including in the Indian population.^31,32 There is limited data available on BIVA in children on dialysis. The lack of specific reference graphs may also hinder the correct interpretation of BIVA in children. In previous studies from our center, BIVA aided in targeting dry weight in children on HD and reflected OH better than the clinical assessment in children on CAPD.^33,34

Subclinical OH is known to improve patient’s quality of life and long-term cardiovascular outcomes.^2,35 In this study, both BIVA and LU detected subclinical OH. In addition, the creation of more robust reference graphs helped in assessing and tracking hydration status, highlighting the utility of BIVA as a valuable bedside tool to target dry weight in children on HD and CPD. The discrepancies in the prevalence of OH identified by BIVA and LU could be attributed to the difference in body compartments analyzed.^5,6 Thus, the information provided by combining these tools should be considered complementary to clinical signs in the assessment of hydration status.

There was a lack of association between OH and UH in children on CAPD. UH is associated with left ventricular hypertrophy and cardiac dysfunction, which contributes to poor patient outcomes.³⁶ The commonly practiced “volume first strategy” of managing hypertension may not be beneficial in all, as BP measurements are not solely determined by intravascular volume. Targeting dry weight based on BP control also carries the risk of aggressive and unjustified ultrafiltration, leading to intradialytic hypotension, loss of RKF, and myocardial stunning.⁴ In this study, despite improvement in OH at follow-up, a majority continued to have UH, suggesting that other causes of hypertension need to be explored. The multifactorial etiology of hypertension in children on dialysis include volume-independent factors such as vascular stiffness, congestive heart failure, underlying kidney disease, sympathetic overactivity, and the involvement of renin-angiotensin-aldosterone system.

Similar findings of lack of association between OH and UH in dialysis patients have been reported earlier. In a study of 639 adult APD patients, van Biesen et al. noted that 30–40% had systolic hypertension, despite normo-or underhydration.³⁷ Only a few studies have investigated the association between OH and hypertension in the pediatric dialysis population. In a study by Zaloszyc et al. of 23 children on chronic HD, hypertension (clinic BPs >95th percentile) was seen in 45.3% with normal hydration (by BIS) and BP normalized in only 50% of them following dialysis.¹ Allinovi et al. reported that among13 children on dialysis (eight on APD), hypertension (clinic BPs >95th centile) diagnosed in 36% was not associated with OH (BIS and LU).⁶ Eng et al. observed that 25% of high BP measurements (clinic SBP >90th percentile) were not accompanied by OH (BIS) in their study of 20 children on dialysis (ten on APD).⁴ More recent studies have also shown similar results.^14,38

Our study has notable strengths. The creation of unique gender-specific reference BIVA graphs from healthy controls is one of the highlights. This is also the only study reported in children on dialysis where the gold standard technique of ABPM was used to study the association between hypertension and OH. Interobserver bias was avoided as only a single trained researcher performed all measurements of BIVA and LU. Limitations of the study included small sample size, the nonuniform follow-up due to the pandemic phase, absence of echocardiographic assessment, and lack of analysis of other factors affecting ambulatory hypertension like salt intake and type of antihypertensive medication.

In conclusion, this prospective study in children with kidney failure on CAPD indicates that subclinical OH and UH are frequently detected. However, UH is not associated with OH-BIVA or OH-LU. Besides OH, other etiological factors like excess dietary sodium intake, vascular stiffness, and renin-angiotensin-aldosterone system overactivation for UH need to be explored in these children to ensure optimal management.