Serum Osmolarity and Vasopressin Concentration in Acute Heart Failure—Influence on Clinical Course and Outcome

Neurohormone activation plays an important role in Acute Heart Failure (AHF) pathophysiology. Serum osmolarity can affect this activation causing vasopressin excretion. The role of serum osmolarity and vasopressin concentration and its interaction remain still unexplored in AHF. The objective of our study was to evaluate the relationship of serum osmolarity with clinical parameters, vasopressin concentration, in-hospital course, and outcomes in AHF patients. The study group consisted of 338 AHF patients (male (76.3%), mean age of 68 ± 13 years) with serum osmolarity calculated by the equation: 1.86 × sodium [mmol/L] + (glucose [mg/dL]/18) + (urea [mg/dL]/2.8) + 9 and divided into osmolarity quartiles marked as: low: <287 mOsm/L, intermediate low: 287–294 mOsm/L, intermediate high: 295–304 mOsm/L, and high: >304 mOsm/L. There was an increasing age gradient in the groups and patients differed in the occurrence of comorbidities and baseline clinical and laboratory parameters. Importantly, analysis revealed that vasopressin presented a linear correlation with osmolarity (r = −0.221, p = 0.003) and its concentration decreased with quartiles (61.6 [44.0–81.0] vs. 57.8 [50.0–77.3] vs. 52.7 [43.1–69.2] vs. 45.0 [30.7–60.7] pg/mL, respectively, p = 0.034). This association across quartiles was observed among de novo AHF (63.6 [55.3–94.5] vs. 58.0 [50.7–78.6] vs. 52.0 [46.0–58.0] vs. 38.0 [27.0–57.0] pg/mL, respectively, p = 0.022) and was not statistically significant in patients with acute decompensated heart failure (ADHF) (59.5 [37.4–80.0] vs. 52.0 [38.0–74.5] vs. 57.0 [38.0–79.0] vs. 50.0 [33.0–84.0] pg/mL, respectively, p = 0.849). The worsening of renal function episodes were more frequent in quartiles with higher osmolarity (4 vs. 2 vs. 13 vs. 11%, respectively, p = 0.018) and patients that belonged to the quartiles with low and high osmolarity were characterized more often by incidence of worsening heart failure (20 vs. 9 vs. 10 vs. 22%, respectively, p = 0.032). There was also a U-shape distribution in relation to one-year mortality (31 vs. 19 vs. 23 vs. 37%, respectively, p = 0.022). In conclusion, there was an association of serum osmolarity with clinical status and both in-hospital and out-of-hospital outcomes. Moreover, the linear dependence between vasopressin concentration and serum osmolarity in the AHF population was identified and was driven mainly by patients with de novo AHF which suggests different pathophysiological paths in ADHF and AHF de novo.


Introduction
Acute heart failure (AHF) is a life-threatening entity with 25-30% one-year mortality and a very high risk of rehospitalization [1]. One of its clinical manifestations is congestion, which occurs in 80% of cases on admission to the hospital. One of the principal reasons for AHF episodes is the inability to control euvolemia and proper water-electrolyte balance [2][3][4]. Its pathogenesis contributes to several factors, such as a natriuretic response, volume displacement, and neurohormonal activity [4][5][6]. Vasopressin is an important neurohormonal factor connecting kidney function and the body's water accumulation. It is called the antidiuretic hormone because, in addition to activities such as vasoconstriction it causes water reabsorption by acting in the renal collecting tubule [7][8][9]. Its antidiuretic effect may result in fluid retention and deterioration of HF [10].
Osmolarity is an essential factor in water-electrolyte homeostasis maintenance. It depends on several well-known factors such as sodium, urea, and glucose [11]. The serum osmolarity reference range is relatively narrow despite its broader concentration norms. This means that osmolyte concentration changes should be immediately compensated, especially by free-water volume. Therefore, serum osmolarity may, to some degree, reflect total body fluid volume [7]. Proper osmolarity level maintenance mechanisms are, among others, the action of vasopressin and the Renin-Angiotensin-Aldosterone System (RAAS), which leads to an increase in the body's neurohormonal activity-one of the well-known AHF pathways [12][13][14]. The most important organ for their effective action is the kidney. The pathophysiological interplay between the heart and kidney in AHF is called cardiorenal syndrome and resulting in renal dysfunction [15][16][17]. On the other hand, the kidneys are the key organs in the decongestion process, indirectly expressed by proper quantitative and qualitative urine excretion, which is dependent on vasopressin as a result of its tightly intertwined pathophysiology [4,9,18]. However, despite these facts, vasopressin and osmolarity are not commonly used markers in routine clinical practice. In contrast to RAAS, the role of serum osmolarity and vasopressin in the prognosis of AHF is still unclear.
This study aimed to evaluate the differences between highlighted serum osmolarity quartiles, especially in clinical features, laboratory parameters, kidney function, urine composition, and vasopressin concentration. We also wanted to establish the link between osmolarity, vasopressin, and outcomes in the AHF group, including de novo AHF and the acute decompensated acute heart failure (ADHF) subpopulations.

Material and Methods
This is a single-center, observational study. The study population included patients admitted with AHF to the Centre of Heart Diseases, 4th Military Hospital, Wroclaw, Poland. All participants were enrolled in the AHF registry carried out in 2010-2012 and 2016-2017. The inclusion criteria were as follows: AHF as the primary cause of hospitalization (according to the European Society of Cardiology HF Guidelines [19,20]), adults (age ≥ 18 years old), and a written informed consent provided by the patient. Exclusion criteria were cardiogenic shock, diagnosis of acute coronary syndrome, known severe liver disease, end-stage renal failure requiring renal replacement therapy, and evidence of infection. The study was approved by the local ethics committee (the Ethics Committee of Wroclaw Medical University) and was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice.

Study Design
After admission to the hospital, detailed information related to participants' demographic medical history, previous treatment, and co-morbidities was recorded. Clinical and laboratory examinations were carried out at four timepoints: admission, on the first and second day of hospitalization, and discharge. Participants were subjected to a 365-day follow-up.

Laboratory Parameters
Laboratory parameters such as N-terminal pro-B-type natriuretic peptide (NT-proBNP), troponin I, peripheral blood morphology, renal function parameters (urea, creatinine), electrolytes (sodium, potassium), liver function parameters (bilirubin, aspartate aminotransferase (AST), alanine aminotransferase (ALT), albumin), C-reactive protein (CRP), lactates in capillary blood, and urine spot samples (sodium, creatinine, urea) were performed in a hospital laboratory in standard laboratory tests during patient hospitalization. Vasopressin concentration was measured post hoc from frozen samples using the ELISA test.

Serum Osmolarity Determination
Baseline serum osmolarity was calculated from serum sodium, serum glucose, and urea using the equation:

Subsection
The study population was divided into four equal quartiles based on a baseline serum osmolarity and marked as: The paper presents all comparisons as low to high osmolarity quartiles, respectively. Study endpoints were defined as: • one-year all-cause mortality, • worsening of HF (WHF), • worsening of renal function (WRF).
The worsening of HF was defined as no improvement or deterioration of the patient's condition during hospitalization [21,22]. The WRF was defined as an increase > 0.3 mg/dL of creatinine concentration within 48 h after admission [23].

Statistical Analysis
Continuous variables with normal distribution were presented by mean ± standard deviation (SD). Skewed variables were defined as a median [lower and higher quartile]. Categorized values were presented as numbers and/or percentages. Shapiro-Wilk tests were performed to check the normality of the distribution. The Levene test was used to evaluate homogeneity of variance.
Analysis of variance (ANOVA), Kruskal-Wallis tests with post-hoc tests were used to test for differences between quartiles in quantitative variables and the Chi-squared test was utilized for qualitative variables. The Mann-Whitney U-test, t-test, and Chi-squared test were applied to compare variables between two groups when appropriate (HF de novo and Acute Decompensated Heart Failure). Correlations between variables were assessed by the Pearson linear correlation test and Spearman correlation test. Kaplan-Meier curves were constructed to estimate the survival probability. The Cox Proportional Hazard Model was used to evaluate the prognostic significance of variables and its hazard ratio to one year mortality. The multivariable regression model was made to establish independent factors that influenced osmolarity. Those p values < 0.05 were considered statistically significant. Analysis was performed with STATISTICA 12 software (StatSoft Polska Sp. z o.o., Kraków, Poland).

Kidney Function Parameters and Components of Serum Osmolarity by Quartiles
Kidney function parameters such as urea and creatinine increased gradually across osmolarity quartiles (37

Urine Composition by Osmolarity Quartiles
The lowest spot urine sodium was observed in the group with the lowest serum osmolarity ( No correlation was found between urine sodium, creatinine, and osmolarity or vasopressin concentration.

Discussion
The issue of osmolarity is of great interest in the context of heart failure. Our study provides insight into the understanding of serum osmolarity and its neurohormonal (vasopressin) control in AHF pathophysiology. First, the relationship between serum osmolarity and one-year mortality and WHF in the study population was not linear. The highest risk of mortality and worsening HF was observed in patients in the low and the high quartile of serum osmolarity. The risk, therefore, had a U-shape trajectory. It is essential to acknowledge that patients with low serum osmolarity, despite being relatively young, exhibited characteristics of more severe HF. These included: lower blood pressure and lower serum sodium concentration as well as: high NTproBNP, elevated bilirubin, and low albumin concomitant with the highest vasopressin concentration [24,25]. On the other hand, the high osmolarity group had a different profile. They were the oldest, with the greatest number of comorbidities and renal dysfunction probably being contributors to poor outcomes in high osmolarity. The results show two distinct phenotypes of AHF patients, both with poor outcomes but most likely resulting from diverse mechanisms. The research of Nakagawa et al. gives interesting results, where it was observed that the highest serum osmolarity is associated with the highest mortality. It was, however, a study conducted in a group of patients with heart failure with a preserved left ventricle ejection fraction (HFpEF). Therefore, the characteristics of the study population were also specific. The female gender was dominant and the mean age was higher than in our study group. However, the common feature was higher comorbidity: chronic kidney disease, diabetes, and arterial hypertension in patients with high osmolarity, which were probably some of the essential factors accounting for a worse prognosis [14].
The low osmolarity group exhibited high sodium avidity, with low urine sodium excretion and high urine creatinine concentration. On the other hand, the high serum osmolarity quartile had high sodium excretion but low eGFR. This further confirms WRF incidents were more frequent in patients with higher osmolarity (in the intermediate high and high osmolarity quartile) (4% vs. 2% vs. 13% vs. 11%; p = 0.018, respectively) ( Figure 1, Table 4).

Discussion
The issue of osmolarity is of great interest in the context of heart failure. Our study provides insight into the understanding of serum osmolarity and its neurohormonal (vasopressin) control in AHF pathophysiology. First, the relationship between serum osmolarity and one-year mortality and WHF in the study population was not linear. The highest risk of mortality and worsening HF was observed in patients in the low and the high quartile of serum osmolarity. The risk, therefore, had a U-shape trajectory. It is essential to acknowledge that patients with low serum osmolarity, despite being relatively young, exhibited characteristics of more severe HF. These included: lower blood pressure and lower serum sodium concentration as well as: high NTproBNP, elevated bilirubin, and low albumin concomitant with the highest vasopressin concentration [24,25]. On the other hand, the high osmolarity group had a different profile. They were the oldest, with the greatest number of comorbidities and renal dysfunction probably being contributors to poor outcomes in high osmolarity. The results show two distinct phenotypes of AHF patients, both with poor outcomes but most likely resulting from diverse mechanisms. The research of Nakagawa et al. gives interesting results, where it was observed that the highest serum osmolarity is associated with the highest mortality. It was, however, a study conducted in a group of patients with heart failure with a preserved left ventricle ejection fraction (HFpEF). Therefore, the characteristics of the study population were also specific.
The female gender was dominant and the mean age was higher than in our study group. However, the common feature was higher comorbidity: chronic kidney disease, diabetes, and arterial hypertension in patients with high osmolarity, which were probably some of the essential factors accounting for a worse prognosis [14].
The low osmolarity group exhibited high sodium avidity, with low urine sodium excretion and high urine creatinine concentration. On the other hand, the high serum osmolarity quartile had high sodium excretion but low eGFR. This further confirms glomerular and tubular dysfunction dissociation in AHF, which may be associated with a worse decongestion process and poor diuretic response leading to HF worsening during hospitalization [26][27][28]. Additionally, both extreme profiles had elevated serum lactate, which was already shown to be a strong prognosticator in AHF [29][30][31].
There was a significant correlation between serum osmolarity and vasopressin in the population. However, the latter was not confirmed in the subgroup analysis of the de novo and ADHF patients. Interestingly, there was no gradient of vasopressin across the serum osmolarity quartiles in patients with ADHF. All quartiles had the same vasopressin level (~55 pg/mL). These findings might indicate that patients with acutely decompensated HF had relatively elevated vasopressin independent from serum osmolarity (so-called non-osmotic release of the hormone), which may have an influence for worse diuretic response and in-hospital course. On the other hand, there was a strong correlation between vasopressin and serum osmolarity in patients with de novo AHF. This concept refers to the gradual activation and dysfunction of several pathophysiological mechanisms [28].
Vasopressin is responsible for the regulation of serum osmolarity. Acting in the collecting tubule and the distal part of the ascending tubule, it induces expression of aquaporin, thereby increasing osmotic reabsorption of water. Moreover, it activates sodium transporters in the ascending tubule of the nephron, allowing iso-osmolar transport of sodium ions from the lumen of the tubule to the renal interstitium [7]. Through V2 receptors, it increases the activity of urea transporters, facilitating its reabsorption. The aforementioned mechanisms result in water reabsorption and urine concentration. Thus, the serum osmolarity seems to be an important factor in AHF pathophysiology.
These results contribute to a better understanding of the pathophysiology of AHF. This is extremely important in more effectively identifying groups with a high risk of a worse clinical course in patients with AHF already at the initial stage of inpatient treatment. We strongly believe that in the future they will contribute to the development of new diagnostic and therapeutic strategies aimed at improving the prognosis of patients with AHF.

Limitations
This is a retrospective single-center study with analysis and division of patients performed post hoc. The number of participants was relatively low. There is also limited information regarding diuresis, which may be a key to a better understanding of the process. Osmolarity was calculated from a mathematical formula based on the concentration of osmolytes. The method for precisely measuring osmolality is osmometry. There is a difference between osmolality and osmolarity; however, it is small, so it allows the use of mathematical formula to make an estimation [11]. Given the variability of the factors from which the osmolarity was calculated, it seems reasonable to conduct a similar study with its measured values.

Conclusions
Serum osmolarity is associated with both short-term and long-term outcomes in AHF patients (however, this relationship is not linear). Consequently, its implementation could potentially facilitate risk stratification of AHF patients. Further research is required to accurately assess the predictive role of this parameter.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/biomedicines10082034/s1, Table S1: Univariate and multivariate analysis to establish factors influencing serum osmolarity; Figure S1: Relationship between serum osmolarity and Hazard Ratio; Figure