Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Nov 1.
Published in final edited form as: Pediatr Nephrol. 2022 Feb 24:10.1007/s00467-022-05478-5. doi: 10.1007/s00467-022-05478-5

Association of early hyponatremia and the development of acute kidney injury in critically ill children

Cassandra L Formeck 1,2, Nalyn Siripong 3, Emily L Joyce 2,4, Juan C Ayus 5, John A Kellum 2,6, Michael L Moritz 1
PMCID: PMC9399308  NIHMSID: NIHMS1784140  PMID: 35211792

Abstract

Background

Hyponatremia is an independent prognostic factor for mortality; however, the reason for this remains unclear. An observed relationship between hyponatremia and the development of acute kidney injury (AKI) has been reported in certain disease states, but hyponatremia has not been evaluated as a predictor of AKI in critically ill patients or children.

Methods

This is a single-center retrospective cohort study of critically ill children admitted to a tertiary care center. We performed regression analysis to assess the association between hyponatremia at ICU admission and the development of new or worsening stage 2 or 3 (severe) AKI on days 2–3 following ICU admission.

Results

Among the 5057 children included in the study, early hyponatremia was present in 13.3% of children. Severe AKI occurred in 9.2% of children with hyponatremia compared to 4.5% of children with normonatremia. Following covariate adjustment, hyponatremia at ICU admission was associated with a 75% increase in the odds of developing severe AKI when compared to critically ill children with normonatremia (aOR 1.75, 95% CI 1.28–2.39). Evaluating sodium levels continuously, for every 1 mEq/L decrease in serum sodium level, there was a 0.05% increase in the odds of developing severe AKI (aOR 1.05, 95% CI 1.02–1.08). Hyponatremic children who developed severe AKI had a higher frequency of kidney replacement therapy, AKI or acute kidney disease at hospital discharge, and hospital mortality when compared to those without.

Conclusions

Hyponatremia at ICU admission is associated with the development of new or worsening AKI in critically ill children.

Keywords: Hyponatremia, Acute kidney injury, Pediatric, Pediatric intensive care unit, Critical care outcomes

Introduction

Hyponatremia is a common electrolyte abnormality observed in critically ill children, occurring in up to 67% of children admitted to the pediatric intensive care unit (ICU) [1–5]. Hyponatremia in hospitalized children and adults has been shown to be an independent prognostic factor for risk-adjusted mortality [6–11]; however, the reason for this remains unclear. Neurologic complications from hyponatremic encephalopathy are well reported [12, 13], though there is an emerging literature that even mild and seemingly asymptomatic hyponatremia can impair multi-organ function, including non-cardiogenic pulmonary edema [14, 15], cardiac dysfunction [16], and immune dysregulation [17].

Previous studies have described an observed association between hyponatremia and the development of acute kidney injury (AKI) in select cohorts of patients, including patients with liver disease [18], acute myocardial infarction [19], heart failure [19], human immunodeficiency virus (HIV) [20], leptospirosis [21], and malignancy [22], and in a large heterogeneous cohort of hospitalized adults [6]. In hospitalized adults, the presence of pre-existing hyponatremia and the subsequent development of AKI had a synergistic effect on the risk of short-term hospital mortality [6]. However, hyponatremia has not been evaluated as a risk factor for the development of AKI in critically ill patients or children. Renal tubular epithelial cells are exposed to numerous stressors in critically ill patients including osmotic cellular stress, which is primarily driven by changes in sodium concentration. Hypo-osmolar environments could contribute to AKI either by altering renal tubular function or by impairing cellular function systemically.

We hypothesized that hyponatremia would be associated with a higher incidence of AKI in critically ill children, and that critically ill children with hyponatremia who developed AKI would have higher rates of inhospital mortality when compared to normonatremic children with or without subsequent AKI. Children represent a good cohort to assess for this type of association as they have a lower mortality rate and fewer comorbidities when compared to critically ill adults. Our objective was to evaluate the association between hyponatremia at ICU admission and the subsequent development of AKI in a large heterogeneous cohort of critically ill children and to assess the rates of inhospital mortality and other clinically significant outcome measures among children with and without hyponatremia and subsequent AKI.

Materials and methods

Study design and patient population

We conducted a single-center retrospective cohort study assessing the risk of developing stage 2 or 3 AKI in pediatric critically ill patients after exposure to hyponatremia. We used the Pediatric High-Density Intensive Care database (Peds HiDenIC), which includes all patients aged ≥ 60 days admitted to the pediatric or cardiac ICUs at the UPMC Children’s Hospital of Pittsburgh between 2010 and 2014. Patients were included if they had a sodium measurement obtained within 6 h prior to or following ICU admission. Patients were excluded if they had age greater than or equal to 18 years, hypernatremia (sodium measurement > 145 mEq/L) at ICU admission, insufficient data to categorize AKI status, stage 3 AKI in the first 24 h of ICU admission, a serum glucose measurement > 300 mg/dL in the first 24 h of ICU admission, history of stage 5 chronic kidney disease (CKD), and kidney failure or history of a kidney transplant. There was a waiver of informed consent for this study. The study was approved by the Institutional Review Board at the University of Pittsburgh.

Definitions and measurements

The primary exposure, hyponatremia at ICU admission, was defined as a serum sodium measurement of < 135 mEq/L 6 h prior to or following ICU admission. Normonatremia was defined as a serum sodium measurement of 135–145 mEq/L. Hyponatremia was also modeled both as a continuous variable and as a categorical variable (by severity) based on the lowest serum sodium measurement in the 6-h window prior to or following ICU admission. Mild hyponatremia was defined as a serum sodium 130–134 mEq/L, moderate hyponatremia as a serum sodium of 125–129 mEq/L, and severe hyponatremia as a serum sodium < 125 mEq/L. Sodium levels during the study period were measured by indirect potentiometry utilizing two glass sodium electrodes (SYNCHRON Systems, Bristol, PA).

The primary outcome was the development of stage 2 or 3 AKI, or an increase in AKI stage from 2 to 3, within the 2 days following the day of ICU admission (i.e., days 2–3). In accordance with the Kidney Disease: Improving Global Outcomes (KDIGO) criteria, stage 2 AKI was defined by (1) an increase in serum creatinine by 2–2.9 times the baseline serum creatinine or (2) a decrease in urine output to < 0.5 mL/kg/h for ≥ 12 h, and stage 3 AKI was defined by (1) an increase in serum creatinine by 3 times the baseline serum creatinine, (2) an increase in serum creatinine to ≥ 4 mg/dL, (3) initiation of kidney replacement therapy, (4) a decrease in estimated glomerular filtration rate (eGFR) to < 35 mL/min/1.73 m2, (5) a decrease in urine output to < 0.3 mL/kg/h for ≥ 24 h, or (6) anuria for ≥ 12 h [23, 24]. The secondary outcomes were the development of stage 2 or 3 AKI, or an increase in AKI stage from 2 to 3, within the 7 days following the day of ICU admission (i.e., days 2–8), need for kidney replacement therapy (KRT) during the hospitalization, ICU and hospital length of stay, the presence of AKI or acute kidney disease (AKD) at hospital discharge, and hospital mortality. AKD was defined by AKI stage 1 or greater for ≥ 7 days after onset of AKI.

Additional covariates included in the analysis were variables that have been associated with hyponatremia and the development of AKI including age, history of heart failure, history of liver failure, history of seizures/epilepsy, history of non-kidney solid organ transplant, history of malignancy, exposure to nephrotoxic medications, vasopressor use, mechanical ventilation, suspected bacterial sepsis, thrombocytopenia, hypoalbuminemia, severe anemia, major surgery or cardiac bypass surgery prior to ICU admission, and electronic Pediatric Index of Mortality 2 (ePIM2) score [25]. Suspected bacterial sepsis was defined as the collection of blood cultures within the first 24 h after ICU admission, along with the initiation of antibiotics 24 h prior to or after collection of blood cultures. Severe anemia was defined as a hemoglobin level less than 7 g/dL in children under the age of 5 years and a hemoglobin level less than 8 g/dL in children 5–18 years. The ePIM2 score was measured within the first 24 h of ICU admission.

Statistical analysis

Continuous variables are presented as means (standard deviation) when normally distributed or as medians (interquartile range (IQR)) if not. Categorical data are presented as frequencies (with percentages). We described demographic and clinical characteristics for patients with hyponatremia compared to those with normonatremia (Table 1). Statistical comparisons of continuous variables between the two groups were done using t-test for variables with a normal distribution and Wilcoxon rank-sum test for variables with a skewed distribution. Comparisons of categorical variables were performed using chi-square test or Fisher’s exact test, where appropriate. To assess for trend in the frequency of stage 2 or 3 AKI by severity of hyponatremia, we performed a nonparametric test for trend across ordered groups developed by Cuzick (1985).

Table 1.

Patient characteristics for eligible study patients

Characteristics Normonatremia N = 4385 (86.7%) Hyponatremia N = 672 (13.3%) All N = 5057
Age, median (IQR), years 4.9 (1.3–11.9) 5.2 (1.3–12.1) 4.9 (1.3–11.9)
Males no. (%) 2485 (56.7) 398 (59.2) 2883 (57.0)
Race no. (%)
 Caucasian 3456 (78.8) 518 (77.1) 3974 (78.6)
 African-American 692 (15.8) 113 (16.8) 805 (15.9)
 Other 237 (5.4) 41 (6.1) 278 (5.5)
Mechanical ventilation no. (%)a 1468 (33.5) 262 (39.0) 1730 (34.2)
Vasopressor use no. (%)a 926 (21.1) 148 (22.0) 1074 (21.2)
Suspected sepsis no. (%)a 564 (12.9) 114 (17.0) 678 (13.4)
History of heart failure no. (%) 142 (3.2) 37 (5.5) 179 (3.5)
History of liver failure no. (%) 73 (1.7) 24 (3.6) 91 (1.9)
History of non-kidney solid organ transplant no. (%) 158 (3.6) 21 (3.1) 179 (3.5)
History of malignancy no. (%) 226 (5.2) 42 (6.3) 268 (5.3)
History of seizures/epilepsy no. (%) 660 (15.1) 115 (17.1) 775 (15.3)
Surgery prior to ICU admission no. (%) 1882 (42.9) 236 (35.1) 2118 (41.9)
Cardiac bypass prior to ICU admission no. (%) 381 (8.7) 16 (2.4) 397 (7.9)
ePIM2 score likelihood of mortality, median (IQR)a, % 0.86 (0.24–2.6) 1.0 (0.78–3.7) 0.90 (0.26–2.9)
Open in a new tab

ePIM2, electronic Pediatric Index of Mortality 2 score; ICU, intensive care unit; IQR, interquartile range.

a

Measured within 24 h following ICU admission

We conducted a logistic regression analysis to assess the association between hyponatremia at ICU admission and our primary outcome (the development of stage 2 or 3 AKI, or an increase in AKI stage from 2 to 3, within the 2 days following the day of ICU admission) after adjusting for confounders. Potential confounders for the model were identified as all covariates associated with both the exposure and outcome variables on univariate analysis. All confounders were included in the initial model. To identify a more parsimonious model, several variables were removed based on their independent association with the outcome. After removing each variable, we confirmed that we did not introduce significant confounding in our assessment of the association between hyponatremia and AKI and confirmed that model fit was not compromised. The same analysis was repeated for our secondary outcome, the development of stage 2 or 3 AKI, or an increase in AKI stage from 2 to 3, within the 7 days following the day of ICU admission (i.e., days 2–8).

We further compared patient outcomes (need for KRT, AKI or AKD at hospital discharge, ICU and hospital length of stay, and hospital mortality) for children in the following four groups based on their exposure and primary outcomes status: normonatremia and no stage 2/3 AKI, hyponatremia and no stage 2/3 AKI, normonatremia and stage 2/3 AKI, and hyponatremia and stage 2/3 AKI. Statistical comparisons of hospital and ICU length of stay between the four groups were using Kruskal–Wallis rank test. Comparisons of categorical variables were performed using contingency table chi-square test. Statistical significance was set at a p-value of less than or equal to 0.05. Statistical analyses were performed using Stata software, version SE 14.2 (StataCorp, TX, USA).

Results

From our query of the Peds HiDenIC database, 8733 patients were admitted at UPMC Children’s Hospital of Pittsburgh between 2010 and 2014 and were at least 60-day old at admission (Fig. 1). Of these, 7374 had a sodium measurement within 6 h prior to or following ICU admission. An additional 2317 patients were excluded because they had age greater than or equal to 18 years (N = 553); hypernatremia (N = 113); insufficient information to categorize AKI status (N = 468); stage 3 AKI in the first 24 h of ICU admission (N = 76); a serum glucose measurement > 300 mg/dL in the first 24 h of ICU admission (N = 1087); a history of a kidney transplant (N = 18); or a history of stage 5 CKD or kidney failure (N = 2).

Fig. 1.

Fig. 1

Open in a new tab

Flow diagram of study cohort with exclusions. AKI, acute kidney injury; CKD, chronic kidney disease; ICU, intensive care unit. aMeasured within 6 h prior to or after ICU admission. bMeasured within 24 h of ICU admission

A total of 5057 children were included in the final analysis, of which 13.3% had hyponatremia at ICU admission (Fig. 1). Of the 672 patients with hyponatremia, 547 (81.4%) had mild hyponatremia, 77 (11.5%) had moderate hyponatremia, and 48 (7.1%) had severe hyponatremia (Fig. 1). There were no significant differences in age, sex, or race between children with or without hyponatremia (Table 1). Children with hyponatremia had a higher frequency of mechanical ventilation, suspected sepsis, history of heart failure, history of liver failure, and higher ePIM2 score likelihood of mortality (Table 1).

Development of stage 2 or 3 AKI on days 2–3 following ICU admission

When assessing the development of AKI, 5.1% (N = 258) of children developed stage 2 or 3 AKI, or an increase in AKI stage from 2 to 3, on days 2–3 following ICU admission. Stage 2 or 3 AKI developed in 9.2% (N = 62) of children with hyponatremia compared to 4.5% (N = 196) of children with normonatremia (Table 2). Of the 258 children who developed stage 2 or 3 AKI, maximum AKI stage was consistent between serum creatinine and urine output measurements or determined by need for KRT in 206 (79.8%) children. Of the remaining 52 children, maximum AKI stage was determined by serum creatinine measurement in 16 (6.2%) children and by urine output in 36 (14%) children. For the 36 children with maximum AKI stage determined by urine output criteria, 4 had hyponatremia and 32 had normonatremia at ICU admission.

Table 2.

Frequency of stage 2 or 3 AKI among critically ill children with normonatremia and hyponatremia at ICU admission

Sodium level at ICU admission Stage 1 or no AKI N = 4799 Stage 2 or 3
AKI N = 258
Normonatremia (serum sodium 135–145 mEq/L) 4189 (95.5) 196 (4.5)
Mild hyponatremia (serum sodium 130–134 mEq/L) 498 (91) 49 (9.0)
Moderate hyponatremia (serum sodium 125–129 mEq/L) 69 (89.6) 8 (10.4)
Severe hyponatremia (serum sodium < 125 mEq/L) 43 (89.6) 5 (10.4)
Open in a new tab

AKI, acute kidney injury; ICU, intensive care unit

On univariate analysis, hyponatremia at ICU admission was associated with the development of stage 2 or 3 AKI (OR 2.17 [95% CI 1.16–2.93], P < 0.001) (Table S1). Increasing severity of hyponatremia was associated with increasing odds of developing stage 2 or 3 AKI: mild hyponatremia OR 2.10 (95% CI 1.52–2.91, P < 0.001), moderate hyponatremia OR 2.48 (95% CI 1.18–5.22, P = 0.02), and severe hyponatremia OR 2.49 (95% CI 0.97–6.34, P = 0.06). The test for trend was significant at P < 0.001, consistent with our hypothesis of a dose–response relationship between hyponatremia severity and odds of AKI.

Following covariate adjustment, hyponatremia at ICU admission was associated with increased odds of developing stage 2 or 3 AKI (aOR 1.75 [95% CI 1.28–2.39], P < 0.001) when compared to children with normonatremia. Additional variables associated with development of stage 2 or 3 AKI and included in the final multivariable model were patient age, mechanical ventilation, hypoalbuminemia, thrombocytopenia, severe anemia, history of liver failure, and history of seizures/epilepsy. Looking at hyponatremia severity, mild hyponatremia was associated with the development of stage 2 or 3 AKI (aOR 1.71 [95% CI 1.22–2.41], P = 0.002) compared to normonatremia. When comparing children with moderate hyponatremia and severe hyponatremia to those with normonatremia, the adjusted odds for subsequent stage 2 or 3 AKI were 1.91 (95% CI 0.89–4.11, P = 0.1) and 1.92 (95% CI 0.73–5.06, P = 0.2), respectively. When combining children with moderate and severe hyponatremia (sodium level < 130 mEq/L), the adjusted odds of developing stage 2 or 3 AKI when compared to children with normonatremia were 1.92 (95% CI 1.04–3.52, P = 0.036). Modeling serum sodium level as a continuous variable, results of the multivariable regression model show that for every 1 mEq/L decrease in serum sodium level, there was a 0.05% increase in the odds of developing stage 2 or 3 AKI (aOR 1.05, 95% CI 1.02–1.08, P = 0.001) (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Adjusted probability of developing stage 2 or 3 AKI among critically ill children by sodium level at ICU admission. AKI, acute kidney injury; ICU, intensive care unit

Due to linked phenomena between reduced urine output and hyponatremia among critically ill children with increased antidiuretic hormone (ADH) secretion, a sensitivity analysis was performed to assess the association between early hyponatremia and subsequent development of stage 2 or 3 AKI if maximum stage of AKI was not defined by urine output. When children with maximum stage of AKI defined by urine output criteria (N = 36) were (1) excluded or (2) assumed to have stage 1 or no AKI, hyponatremia remained associated with the development of AKI with adjusted odds of 1.98 (95% CI 1.42–2.74) and 1.99 (95% CI 1.44–2.76), respectively.

Need for kidney replacement therapy, ICU length of stay, hospital length of stay, AKI or AKD at hospital discharge, and hospital mortality

To assess how the combination of hyponatremia and AKI impact patient outcomes, patients were divided into four distinct categories based on hyponatremia and AKI status: 4189 children (82.8%) had normonatremia and no stage 2/3 AKI, 610 children (12.1%) had hyponatremia and no stage 2/3 AKI, 196 children (3.9%) had normonatremia and stage 2/3 AKI, and 62 children (1.2%) had hyponatremia and stage 2/3 AKI. All patients had complete data on need for KRT, ICU length of stay, hospital length of stay, and hospital mortality. Data on AKI/AKD stage at hospital discharge was available on 4974 (98.4%) of children. Children with a combination of hyponatremia and stage 2 or 3 AKI had a higher frequency of KRT, longer length of ICU and hospital stay, higher frequency of AKI/AKD at hospital discharge, and higher frequency of hospital mortality when compared to children in the remaining subgroups (Fig. 3, Table S2).

Fig. 3.

Fig. 3

Open in a new tab

Secondary outcomes among critically ill children by primary exposure and primary outcome status. AKI, acute kidney injury; AKD, acute kidney disease; ICU, intensive care unit; KRT, kidney replacement therapy

Development of stage 2 or 3 AKI on days 2–8 following ICU admission

Looking at the development of stage 2 or 3 AKI on days 2–8 following ICU admission, 459 children had insufficient data to categorize AKI status. Following application of the remaining inclusion and exclusion criteria, a total of 5065 children were included in this analysis. Of these children, 13.3% (N = 674) had hyponatremia at ICU admission. In this cohort, 11.3% (N = 76) of children with hyponatremia developed stage 2 or 3 AKI on days 2–8 following ICU admission compared to 6.9% (N = 303) of children with normonatremia. The overall rate of stage 2 or 3 AKI was 7.5%. Hyponatremia was associated with the development of stage 2 or 3 AKI on unadjusted analysis (OR 1.71 [95% CI 1.31–2.22], P < 0.001). On adjusted analysis, children with hyponatremia had a 33% increase in the odds of developing stage 2 or 3 AKI compared to those with normonatremia (aOR 1.33 [95% CI 0.99–1.78], P = 0.05). Additional variables associated with development of stage 2 or 3 AKI on days 2–8 following ICU admission included patient age, vasopressor use, mechanical ventilation, hypoalbuminemia, thrombocytopenia, and history of liver failure.

Discussion

To our knowledge, this is the first study to assess the association between hyponatremia and the subsequent development of AKI in a large heterogeneous cohort of critically ill children. In our study cohort, new or worsening stage 2 or 3 AKI developed in 9.2% of children with hyponatremia compared to 4.5% of children with normonatremia. These findings are similar to those reported in a study of hospitalized adults, which showed a significantly higher incidence rate of AKI in patients presenting with hyponatremia compared to those with normonatremia (12.8% vs. 4.4%) [6]. In our study, hyponatremia at ICU admission was associated with a 75% increase in the odds of developing stage 2 or 3 AKI on days 2–3 following ICU admission, when compared to critically ill children with normonatremia on adjusted analysis. This association followed a dose response, with an increase in the adjusted odds of developing AKI with an increase in the severity of hyponatremia. While moderate and severe hyponatremia were not individually associated with the development of stage 2/3 AKI on adjusted analysis, when combining children with moderate and severe hyponatremia, a sodium level < 130 mEq/L was predictive of subsequent development of stage 2/3 AKI when compared to normonatremia. These results are likely reflective of the lower number of children with moderate and severe hyponatremia (N = 125) in our cohort. The strength of the association between hyponatremia, measured at ICU admission, and the development of AKI was strongest in the first 72 h, and attenuated over the following week.

Whether hyponatremia is a causal factor in the development of AKI, an early marker of kidney injury, or a manifestation of comorbid conditions leading to AKI remains unclear. While the cause of hyponatremia and its relationship with AKI development was unable to be determined based on available data for our cohort, we were able to assess the association between hyponatremia and the occurrence of stage 2/3 AKI among children with AKI defined by both urine output and serum creatinine criteria and by serum creatinine criteria alone. In both analyses, the association between hyponatremia and occurrence of AKI remained unchanged, which supports that the association identified is not a reflection of elevated ADH secretion in the setting of critical illness. In our cohort, hyponatremia remained a significant predictor for the development of AKI following adjustment for comorbid conditions and exposures commonly associated with hyponatremia and kidney injury including edematous states such as heart and liver failure, along with postoperative status and mechanical ventilation. This further supports that hyponatremia may be an early indicator of kidney dysfunction or potentially an independent risk factor for the development of AKI. Renal medullary and tubular epithelial cell injury secondary to hyperosmotic stress has been well studied [26–28]; however, few studies have assessed the impact of hypoosmotic stress on the health of renal medullary and tubular epithelial cells. Hypo-osmolar conditions have been shown to enhance injury-induced degeneration of neurons following axonal injury [29], and low extracellular sodium concentrations cause an accumulation of oxygen-free radicals in osteoclasts leading to oxidative stress response and increased susceptibility to cellular injury [30]. Experimental models of renal tubular epithelial cells have also shown that hypotonic stress induces disruption of the tubular epithelial tight-junction barrier through claudin-mediated pathways [31]; whether this response is damaging or involved in the maintenance of kidney function following cell injury remains unknown. Based on this evidence, hyponatremia may cause direct injury to renal epithelial cells, leading to AKI or an increased susceptibility to AKI when occurring in the presence of other nephrotoxic insults.

While these data suggest that hypoosmotic stress may lead to direct kidney injury, hyponatremia as an early indicator of AKI may be equally as important. Despite the utility of the pRIFLE, AKIN, and KDIGO criteria for the diagnosis of AKI, elevation in serum creatinine level is often a late marker of kidney injury. The use of urinary biomarkers for the early identification of AKI has been validated in select pediatric populations, including children undergoing cardiopulmonary bypass surgery; however, many urinary biomarkers have demonstrated lower discriminatory ability to predict AKI in more heterogenous cohorts of critically ill children [32–34]. To address this, clinical prediction tools such as the renal angina index (RAI) have been developed to help identify patients at high risk for developing AKI following ICU admission [35]. Presently, these prediction tools do not include sodium measurements. Given the findings of our study, additional work is needed to assess if inclusions of hyponatremia to the RAI, or other AKI prediction tools, will further improve our ability to identify critically ill children at high risk for severe AKI at the time of ICU admission.

When assessing other hospital outcomes in our cohort, critically ill children with a combination of hyponatremia who subsequently developed stage 2 or 3 AKI had a higher frequency of KRT, longer hospital and ICU length of stay, higher frequency of AKI or AKD at hospital discharge, and higher frequency of hospital mortality when compared to those without. In addition, the presence of hyponatremia and the subsequent development of stage 2 or 3 AKI appeared to have a multiplicative effect, rather than an additive effect, on the subsequent risk for KRT, frequency of AKI or AKD at hospital discharge, and the incidence of hospital mortality. These results were especially striking for hospital mortality and need for KRT. These findings further support the clinical significance of hyponatremia at ICU admission and the need for additional research to understand if hyponatremia is a modifiable risk factor in the development of AKI among critically ill children.

Our study had several important strengths. The Peds HiDenIC cohort includes a diverse population of general and cardiac ICU patients, which allowed us to assess the association between hyponatremia and the subsequent development of AKI in a large heterogeneous cohort of critically ill children. Although this is a single-center study, overall rates of severe AKI in the Peds HiDenIC cohort are similar to those reported in other pediatric intensive care studies, such as AWARE [36] (11.1% vs. 11.6%) supporting the generalizability of the results. In addition, we were able to define AKI by complete KDIGO criteria, including both serum creatinine and urine output into our determination of AKI stage. This point is important, as many studies define AKI by serum creatinine alone; defining AKI by function of urine output and serum creatinine in combination has been shown to be more predictive of short- and long-term clinical outcomes when compared to the use of urine output or serum creatinine in isolation [37, 38]. Lastly, we were able to identify a dose–response relationship between hyponatremia and the subsequent development of AKI, which helps to support a true association between the two variables.

There are also several important limitations to our study. Although we were able to establish temporal priority of our exposure variable, as a retrospective study, we could not establish if hyponatremia is casually related to the development of AKI. Similarly, retrospective interrogation of our data set did not allow us to identify causes of hyponatremia in our cohort. In addition, because many children lacked laboratory data prior to ICU admission, we were unable to determine the duration of hyponatremia and, therefore, were unable to assess the influence of acute versus chronic hyponatremia on the subsequent risk for AKI. Regarding risk factors for AKI, while children admitted to the cardiac ICU are included in the Peds HiDenIC database and exposure to cardiac bypass surgery prior to ICU admission is known, the database currently does not include data on cardiac bypass time, cross-clamp duration, primary cardiac diagnosis, or Pediatric Index of Cardiac Surgical Intensive Care Mortality Risk (PICSIM) score; therefore, these variables could not be included in our multivariable analysis. Regarding evaluation of hospital outcomes, due to the small number of patients who required KRT and experienced hospital mortality, adjusted analyses for these secondary outcomes could not be performed. Indication for initiation of KRT was also unknown, and decision to initiate KRT may have been influenced by the presence of electrolyte derangements among children with hyponatremia and AKI. Lastly, although children in our final cohort included a heterogeneous mixture of pediatric ICU patients, all subjects in the study came from a single academic medical center providing tertiary care. Accordingly, associated comorbidities and severity of illness of the patient population may not be representative of critically ill children across the country.

In conclusion, hyponatremia at ICU admission is associated with the subsequent development of AKI in critically ill children. This association followed a dose response, with an increase in the adjusted odds of developing AKI with an increase in the severity of hyponatremia. Clinicians should be aware of the increased risk for AKI in children presenting to the ICU with hyponatremia and should employ measures to mitigate the risk for new or worsening AKI in this population. Such measures may include limiting exposure to nephrotoxic medications and avoidance of hypovolemia and volume overload. The combination of hyponatremia and stage 2 or 3 AKI appears to have a multiplicative effect on the need for KRT, the presence of AKI or AKD at hospital discharge, and hospital mortality. Further investigations utilizing bench research methods and prospective clinical studies to assess if hyponatremia is a modifiable risk factor in the development of AKI may help us to further understand the relationship between hyponatremia and AKI in critically ill patients.

Supplementary Material

Supplemental Material

Acknowledgements

We would like to thank the Biostatistical and Data Management Core within the CRISMA (Clinical Research Investigation and Systems Modeling of Acute Illness) Center for their help in variable development as well as data acquisition, management, and storage.

Funding

This study was supported by the National Institutes of Health (T32 DK 91202 and UL1-TR-001857).

Footnotes

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00467-022-05478-5.

Code availability Code for the statistical analysis can be provided upon request.

Conflict of interest The authors declare no competing interests.

Reprints Reprints will not be ordered.

Availability of data and material

Data is not publicly available.

References

  • 1.Al-Sofyani KA (2019) Prevalence and clinical significance of hyponatremia in pediatric intensive care. J Pediatr Intensive Care 8:130–137. 10.1055/s-0038-1676635PMC6687457 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Au AK, Ray PE, McBryde KD, Newman KD et al. (2008) Incidence of postoperative hyponatremia and complications in critically-ill children treated with hypotonic and normotonic solutions. J Pediatr 152:33–38. 10.1016/j.jpeds.2007.08.040 [DOI] [PubMed] [Google Scholar]
  • 3.Kaufman J, Phadke D, Tong S, Eshelman J et al. (2017) Clinical associations of early dysnatremias in critically ill neonates and infants undergoing cardiac surgery. Pediatr Cardiol 38:149–154. 10.1007/s00246-016-1495-3 [DOI] [PubMed] [Google Scholar]
  • 4.Sachdev A, Pandharikar N, Gupta D, Gupta N et al. (2017) Hospital-acquired hyponatremia in pediatric intensive care unit. Indian J Crit Care Med 21:599–603. 10.4103/ijccm.IJCCM_131_17PMC5613613 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Elliman MG, Vongxay O, Soumphonphakdy B, Gray A (2019) Hyponatraemia in a Lao paediatric intensive care unit: prevalence, associations and intravenous fluid use. J Paediatr Child Health 55:695–700. 10.1111/jpc.14278 [DOI] [PubMed] [Google Scholar]
  • 6.Lee SW, Baek SH, Ahn SY, Na KY et al. (2016) The effects of pre-existing hyponatremia and subsequent-developing acute kidney injury on in-hospital mortality: a retrospective cohort study. PLoS ONE 11:e0162990. 10.1371/journal.pone.0162990PMC5021268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Funk GC, Lindner G, Druml W, Metnitz B et al. (2010) Incidence and prognosis of dysnatremias present on ICU admission. Intensive Care Med 36:304–311. 10.1007/s00134-009-1692-0 [DOI] [PubMed] [Google Scholar]
  • 8.Holland-Bill L, Christiansen CF, Heide-Jorgensen U, Ulrichsen SP et al. (2015) Hyponatremia and mortality risk: a Danish cohort study of 279 508 acutely hospitalized patients. Eur J Endocrinol 173:71–81. 10.1530/eje-15-0111 [DOI] [PubMed] [Google Scholar]
  • 9.Mohan S, Gu S, Parikh A, Radhakrishnan J (2013) Prevalence of hyponatremia and association with mortality: results from NHANES. Am J Med 126:1127–37.e1. 10.1016/j.amjmed.2013.07.021PMC3933395 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Price JF, Kantor PF, Shaddy RE, Rossano JW et al. (2016) Incidence, severity, and association with adverse outcome of hyponatremia in children hospitalized with heart failure. Am J Cardiol 118:1006–1010. 10.1016/j.amjcard.2016.07.014 [DOI] [PubMed] [Google Scholar]
  • 11.Guarner J, Hochman J, Kurbatova E, Mullins R (2011) Study of outcomes associated with hyponatremia and hypernatremia in children. Pediatr Dev Pathol 14:117–123. 10.2350/10-06-0858-oa.1 [DOI] [PubMed] [Google Scholar]
  • 12.Ayus JC, Wheeler JM, Arieff AI (1992) Postoperative hyponatremic encephalopathy in menstruant women. Ann Intern Med 117:891–897. 10.7326/0003-4819-117-11-891 [DOI] [PubMed] [Google Scholar]
  • 13.Arieff AI, Ayus JC, Fraser CL (1992) Hyponatraemia and death or permanent brain damage in healthy children. BMJ 304:1218–1222. 10.1136/bmj.304.6836.1218PMC1881802 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Ayus JC, Varon J, Arieff AI (2000) Hyponatremia, cerebral edema, and noncardiogenic pulmonary edema in marathon runners. Ann Intern Med 132:711–714. 10.7326/0003-4819-132-9-200005020-00005 [DOI] [PubMed] [Google Scholar]
  • 15.Ayus JC, Arieff AI (1995) Pulmonary complications of hyponatremic encephalopathy. Noncardiogenic pulmonary edema and hypercapnic respiratory failure. Chest 107:517–521. 10.1378/chest.107.2.517 [DOI] [PubMed] [Google Scholar]
  • 16.Portales-Castillo I, Sterns RH (2019) Allostasis and the clinical manifestations of mild to moderate chronic hyponatremia: no good adaptation goes unpunished. Am J Kidney Dis 73:391–399. 10.1053/j.ajkd.2018.10.004 [DOI] [PubMed] [Google Scholar]
  • 17.Swart RM, Hoorn EJ, Betjes MG, Zietse R (2011) Hyponatremia and inflammation: the emerging role of interleukin-6 in osmoregulation. Nephron Physiol 118:45–51. 10.1159/000322238 [DOI] [PubMed] [Google Scholar]
  • 18.Hackworth WA, Heuman DM, Sanyal AJ, Fisher RA et al. (2009) Effect of hyponatraemia on outcomes following orthotopic liver transplantation. Liver Int 29:1071–1077. 10.1111/j.1478-3231.2009.01982.x [DOI] [PubMed] [Google Scholar]
  • 19.Aronson D, Darawsha W, Promyslovsky M, Kaplan M et al. (2014) Hyponatraemia predicts the acute (type 1) cardio-renal syndrome. Eur J Heart Fail 16:49–55. 10.1093/eurjhf/hft123 [DOI] [PubMed] [Google Scholar]
  • 20.Liborio AB, Silva GB Jr, Silva CG, Lima Filho FJ et al. (2012) Hyponatremia, acute kidney injury, and mortality in HIV-related toxoplasmic encephalitis. Braz J Infect Dis 16:558–563. 10.1016/j.bjid.2012.08.015 [DOI] [PubMed] [Google Scholar]
  • 21.Daher EF, Silva GB Jr, Karbage NN, Carvalho PC Jr et al. (2009) Predictors of oliguric acute kidney injury in leptospirosis. A retrospective study on 196 consecutive patients. Nephron Clin Pract 112:c25–30. 10.1159/000210571 [DOI] [PubMed] [Google Scholar]
  • 22.Salahudeen AK, Doshi SM, Pawar T, Nowshad G et al. (2013) Incidence rate, clinical correlates, and outcomes of AKI in patients admitted to a comprehensive cancer center. Clin J Am Soc Nephrol 8:347–354. 10.2215/CJN.03530412PMC3586962 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Joyce EL, DeAlmeida DR, Fuhrman DY, Priyanka P et al. (2019) eResearch in acute kidney injury: a primer for electronic health record research. Nephrol Dial Transplant 34:401–407. 10.1093/ndt/gfy052PMC6399481 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kidney Disease (2012) Improving global outcomes (KDIGO) acute kidney injury work group. KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney inter., Suppl 2:1–138 [Google Scholar]
  • 25.Joyce EL, Crana CM, Yabes J, Kellum JA (2020) Validation of an electronic Pediatric Index of Mortality 2 score in a mixed quaternary PICU. Pediatr Crit Care Med 21:e572–e575. 10.1097/PCC.0000000000002347PMC7417868 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Fogo AB, Lusco MA, Najafian B, Alpers CE (2017) AJKD atlas of renal pathology: osmotic tubular injury. Am J Kidney Dis 69:e11–e12. 10.1053/j.ajkd.2016.12.003 [DOI] [PubMed] [Google Scholar]
  • 27.Dmitrieva NI, Burg MB (2007) Osmotic stress and DNA damage. Methods Enzymol 428:241–252. 10.1016/s0076-6879(07)28013-9 [DOI] [PubMed] [Google Scholar]
  • 28.Burg MB (2002) Response of renal inner medullary epithelial cells to osmotic stress. Comp Biochem Physiol A Mol Integr Physiol 133:661–666. 10.1016/s1095-6433(02)00203-9 [DOI] [PubMed] [Google Scholar]
  • 29.Dohanics J, Hoffman GE, Verbalis JG (1996) Chronic hyponatremia reduces survival of magnocellular vasopressin and oxytocin neurons after axonal injury. J Neurosci 16:2373–2380 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Barsony J, Sugimura Y, Verbalis JG (2011) Osteoclast response to low extracellular sodium and the mechanism of hyponatremia-induced bone loss. J Biol Chem 286:10864–10875. 10.1074/jbc.M110.155002PMC3060537 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Fujii N, Matsuo Y, Matsunaga T, Endo S et al. (2016) Hypotonic stress-induced down-regulation of claudin-1 and −2 mediated by dephosphorylation and clathrin-dependent endocytosis in renal tubular epithelial cells. J Biol Chem 291:24787–24799. 10.1074/jbc.M116.728196PMC5114426 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Wheeler DS, Devarajan P, Ma Q, Harmon K et al. (2008) Serum neutrophil gelatinase-associated lipocalin (NGAL) as a marker of acute kidney injury in critically ill children with septic shock. Crit Care Med 36:1297–1303. 10.1097/CCM.0b013e318169245aPMC2757115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Herrero-Morin JD, Malaga S, Fernandez N, Rey C et al. (2007) Cystatin C and beta2-microglobulin: markers of glomerular filtration in critically ill children. Crit Care 11:R59. 10.1186/cc5923PMC2206414 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Washburn KK, Zappitelli M, Arikan AA, Loftis L et al. (2008) Urinary interleukin-18 is an acute kidney injury biomarker in critically ill children. Nephrol Dial Transplant 23:566–572. 10.1093/ndt/gfm638 [DOI] [PubMed] [Google Scholar]
  • 35.Basu RK, Zappitelli M, Brunner L, Wang Y et al. (2014) Derivation and validation of the renal angina index to improve the prediction of acute kidney injury in critically ill children. Kidney Int 85:659–667. 10.1038/ki.2013.349PMC4659420 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kaddourah A, Basu RK, Bagshaw SM, Goldstein SL et al. (2017) Epidemiology of acute kidney injury in critically ill children and young adults. N Engl J Med 376:11–20. 10.1056/NEJMoa1611391PMC5322803 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Kellum JA, Sileanu FE, Murugan R, Lucko N et al. (2015) Classifying AKI by urine output versus serum creatinine level. J Am Soc Nephrol 26:2231–2238. 10.1681/ASN.2014070724PMC4552117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Kaddourah A, Basu RK, Goldstein SL, Sutherland SM et al. (2019) Oliguria and acute kidney injury in critically ill children: implications for diagnosis and outcomes. Pediatr Crit Care Med 20:332–339. 10.1097/PCC.0000000000001866 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
NIHMS1784140-supplement-Supplemental_Material.docx (20.7KB, docx)

Data Availability Statement

Data is not publicly available.

RESOURCES