The clinical manifestations of hypotonic hyponatremia result from cerebral edema with the degree, rapidity of development, and duration of hyponatremia determining severity (5, 61). Brain cells undergo intracellular volume adaptation to hyponatremia. This adaptation determines the chronicity of hyponatremia. Within up to 7 h of hyponatremia, brain cells, mainly astrocytes which express aquaporin-4 in abundance (62), lose water to the cerebrospinal fluid through hydrostatic forces (63) and lose electrolytes, such as potassium, sodium, and chloride (5, 64). Subsequently, further intracellular brain cell volume reduction occurs through the loss of organic osmolytes, which is completed by 48 h (5). Hyponatremia lasting <48 h is considered acute, while hyponatremia of 48 h or longer is considered chronic.

Severe clinical manifestations of hyponatremia include seizures, coma, hypoxia secondary to noncardiogenic pulmonary edema and/or hypercapnic respiratory failure (65), and death from cerebral herniation; moderate manifestations include lethargy, disorientation, and confusion; and mild manifestations include fatigue, nausea, and headache (5). Even mild hyponatremia (130–135 mmol/L) is associated with attention deficits, which may require directed testing to be detected, gait disturbances, osteoporosis, and a high risk of fractures (5, 66–68). In patients with advanced CKD, the neurological manifestations of uremia can be confused with the manifestations of dysnatremias, and changes in serum urea concentration affect the treatment of dysnatremias. A paucity of studies exists regarding clinical manifestations related to dysnatremias in patients with CKD. One study reported depressed mental function in patients on peritoneal dialysis with hyponatremia (52).

Certain conditions increase the risk for hyponatremia in patients with CKD not on dialysis, and those on hemodialysis or peritoneal dialysis. These conditions include women gender (34, 36, 41, 44, 48, 54), race other than African American (34, 40, 42, 44, 54), low body weight or body mass index (34, 35, 40, 42–45, 49), diabetes mellitus (35, 40, 41, 43–45, 50, 54), and low serum albumin (10, 34, 35, 40–44, 46–50, 52, 54). In both hemodialysis and peritoneal dialysis patients, hyponatremia was found to be associated with a low residual renal function (43, 44, 47, 49, 51, 54) and excessive weight, most probably fluid, gains (40–42, 44, 45, 47, 48). Hyponatremia in CKD populations increases the risk for several adverse outcomes, such as hospitalization for infections (69), protein-energy malnutrition (70) and impaired cognitive function (71) in hemodialysis patients, and poor peritonitis outcomes (72) plus the higher incidence of new cardiovascular events in peritoneal dialysis patients (73).

Hyponatremia represents a risk for all-cause mortality in the general population (5) and in patients with CKD not on dialysis (10, 34–36), in whom it represents also a risk factor for the mortality from cardiovascular disease or malignancies (36). In hemodialysis patients, hyponatremia was found to be a risk factor for all-cause mortality (40, 42–47), for all-cause mortality only among patients with diabetes mellitus (41), and cardiovascular mortality (40). In peritoneal dialysis patients, hyponatremia was found to be a risk factor for all-cause mortality (49, 54, 72, 74). One study found no association between hyponatremia and 2-year mortality in peritoneal dialysis patients (50). The pathophysiological mechanisms by which hyponatremia increases the risk for mortality in patients with CKD are not well-understood. One study concluded that mortality in hyponatremic peritoneal dialysis patients is most probably due to co-morbidities (48).

Hospitalized patients with COVID-19 have a high prevalence of acute kidney injury with increased death risk (75) and dysnatremias (76–79). The syndrome of inappropriate antidiuretic hormone secretion (SIADH) was identified as the cause of hyponatremia in two cases of COVID-19 (80, 81). Both hyponatremia and hypernatremia represent mortality risks in patients with COVID-19 (75–79).

Acute hypernatremia causes lesions in the brain, such as cell shrinkage, petechial and subarachnoid hemorrhages, hematomas, subdural fluid collections, vascular congestion, and venous thrombosis (82). Children with hypernatremia develop irritability, restlessness, muscular twitching, hyperreflexia, and seizures. Elderly patients with hypernatremia rarely develop seizures, but manifest lethargy, delirium, and coma. Less frequent manifestations of hypernatremia include fever, nausea, and vomiting. Alert hypernatremic patients have intense thirst (82). Higher [Na] values are associated with the progression of established CKD independently of other risk factors (83).

Conditions presenting a risk for hypernatremia in patients with CKD not on dialysis include men's gender, older age, heart failure, low estimated glomerular filtration rate, and high levels of body mass index, systolic blood pressure, and serum albumin (34). Conditions predisposing patients on hemodialysis or peritoneal dialysis to hypernatremia will need further investigation.

Hypernatremia increases the risk for mortality in the general population (82) and in patients with CKD not on dialysis (10, 34–36, 38). In hemodialysis patients, hypernatremia was observed to be a risk factor for all-cause mortality (44) and mortality risk for causes other than cardiovascular disease or malignancy in a second study (46). Whether hypernatremia is associated with mortality in peritoneal dialysis patients has not been studied.

Guidelines (84, 85) and other reports (5, 19, 86, 87) address treatment of hyponatremia in the general population. Figure 1, based on Musso and Bargman's report (88), presents the steps of management of hyponatremia in all patients, such as those with CKD. The last step, evaluation of extracellular volume (ECV) is critical in determining the mechanism and guiding the management of hyponatremia in the general population (5, 19, 86, 87) and all categories of patients with CKD.

Table 3 summarizes the approaches to the treatment of hyponatremia in the general population (5). Infusion of hypertonic saline and restriction of fluid intake represent key general measures for treating hyponatremia in all patients with CKD, such as patients with anuria, while administration of loop diuretics or oral urea are effective means only in patients with significant renal function. Oral urea administration as the sole treatment of hyponatremia is safe and effective (89). Urea administration produces osmotic diuresis and increased free water clearance in the general population (90). Whether urea administration is also safe and effective in the early stages of CKD is not known currently. Even mild hyponatremia should be treated. Rondon-Berrios and Berl (91) proposed treating mild chronic hyponatremia with water intake limitation based on the electrolyte-free water clearance and oral administration of compounds increasing urinary water excretion such as sodium salts, urea, loop diuretics, and vasopressin inhibitors.

We will address the correction of hyponatremia by saline infusion. In patients with CKD, the importance of an accurate estimate of TBW—in the calculation of the volume of infused saline, as well as in the overall management of these patients, acquires even greater importance than in patients with preserved renal function.

The aims of correcting hyponatremia by saline infusion include prevention of brain herniation, improvement of symptoms, and prevention of osmotic demyelination (5). The risk of osmotic demyelination from the rapid rise in [Na] is higher in chronic than in acute hyponatremia. During the rapid correction of chronic hyponatremia, organic osmolyte accumulation in the intracellular compartment of brain cells occurs slowly leading to shrinkage of their intracellular volume. These changes are thought to be linked to the development of osmotic myelinolysis (92). The European guidelines recommend rapid correction of hyponatremia with severe symptoms, regardless of whether the hyponatremia is acute or chronic, by the infusion of hypertonic saline, frequent measurements of [Na] in the first hour of treatment, and evaluation of the symptoms after the increase of [Na] by 5 mmol/L; if symptoms improve, the guidelines recommend stopping the hypertonic saline and limiting the rise in [Na] to a total of 10 mmol/L within the first 24 h and then managing the condition that caused the hyponatremia and increasing [Na] by 8 mmol/L every 24 h thereafter until [Na] reaches 130 mmol/L: if symptoms do not improve, the guidelines suggest continuing the infusion of hypertonic saline with an increase in [Na] by 1 mmol/L hourly until a total rise of 10 mmol/L or [Na] reaches 130 mmol/L (85). Tandukar et al. (93) summarized the findings of 19 publications reporting 21 patients who developed osmotic demyelination syndrome after the correction of hyponatremia by ≤10 mmol/L per 24 h. Sterns and coauthors recommend a change in [Na] during treatment of symptomatic chronic hyponatremia up to 6 mmol/L in the first 6 h, unless symptoms persist when even higher rates of increase in [Na] are indicated, 6–8 mmol/L in the first 24 h, 12–14 mmol/L in 48 h, and 14–16 mmol/L in 72 h (94). Overcorrection should be prevented rigorously (95) and corrected promptly with hypotonic infusions if it occurs. Acute hyponatremia may be treated faster. Whether the same rates of correction of hyponatremia should be applied to patients with early stages of CKD and those treated by dialysis will need to be investigated. In the absence of new information, it would be prudent to apply these guidelines along with clinical judgment to patients with hyponatremic CKD.

Changes in [Na] resulting from the change in body sodium content are determined by the total change in body sodium and by TBW (6). Consequently, TBW, which changes in the same direction and by the same magnitude as ECV in isotonic ECV disturbances, is used when calculating the volume of saline-infused to correct hyponatremia. The Adrogué–Madias formula, which has been used extensively in the management of hyponatremias, calculates the rise in [Na] after infusion of 1 L of saline hypertonic to the serum as follows (86): (2)[Na]Final-[Na]Initial=[Na]Infusate-[Na]InitialTBWInitial+1 Where [Na]_Initial is the pre-infusion and [Na]_Final is the post-infusion [Na], [Na]_Infusate is the sodium concentration in the infused saline, and TBW_Initial is the pre-infusion TBW. Determination of the volume of infused saline (V_Infused) required to produce the desired rise in [Na] using this formula requires one more calculation. For example, in a patient with [Na]_Initial = 110 mmol/L and TBW_Initial = 40 L, the Adrogué–Madias formula calculates that infusion of 1 L of 513 mmol/L (3%) saline will result in a 9.8 mmol/L rise in [Na]. If the desired rise in [Na] is 6 mmol/L, V_Infused = 1 × 6/9.8 = 0.6 L. A formula published subsequently, and based on the same principles as the Adrogué–Madias formula, calculates directly V_Infused, as follows (19): (3)VInfused=TBWInitial×[Na]Final-[Na]Initial[Na]Infusate-[Na]Final In the above example, Equation(3) calculates also a V_Infused of 0.6 L.

One potential source of variance between [Na]_Final values predicted by Equations (2) or (3) and observed relates to the fact that these formulas do not include urinary excretions of water, sodium, and potassium. The urinary excretion may vary considerably during treatment between various categories of hyponatremia (19). Renal excretion of water and monovalent cations during treatment may be less of an issue in patients with advanced CKD.

Another source of variance may be due to an inaccurate estimate of TBW_Initial (19). Hyponatremias in both the general and CKD populations can develop in the settings of hypovolemia, euvolemia, or hypervolemia (5, 19). Several conditions and pathogenic mechanisms are responsible for the development of hyponatremia in patients with normal, low, or high ECV, and consequently abnormal TBW_Initial values. The importance of the estimate of TBW_Initial entered in formulas 2 or 3 is greater in patients with CKD, because of the limited capacity of their kidney to correct errors. Whether ECV and TBW are normal, i.e., whether patients are at their dry weight, is critical in advanced CKD not only for the evaluation of the mechanism and for the management of dysnatremias, but also other clinical reasons, e.g., prevention of cardiovascular complications of hypervolemia or management of hypertension.

Several methods have been proposed for estimating TBW in hyponatremic patients. The older method calculates TBW as an arbitrary fraction of body weight, the difference between women and men, and the difference between older and younger individuals in each gender (86). This method may provide estimates with error because it does not account for body composition at dry weight, in particular about the body fat content, or changes in TBW accompanying the hyponatremia. In addition, the use of sex differences poses a challenge among transgender patients.

Another method for estimating TBW is based on the anthropometric formulas, which represent statistical regression analyses in populations with euvolemia. These statistical analyses compared the TBW measurements by isotopic dilution methods and gender, age, body weight, height, and, in some formulas, the race of the studied subjects. The formulas provided by these studies appropriately account for weight change secondary to a change in body fat, but not for weight change caused by a change in body water. Water gain increases body weight and the fraction of TBW/body weight, while the gain in body fat increases body weight and TBW but decreases the fraction of TBW/body weight (96). The anthropometric formulas always calculate a decrease in the fraction TBW/body weight as weight increases (96). In patients with disorders of TBW, TBW should be estimated as the TBW at dry weight calculated by an anthropometric formula plus or minus the difference between actual and dry weights (19, 96, 97). The drawbacks of this method are that anthropometric formulas, which have large margins of error, and that dry weight is not known in many cases and even when known it may have been miscalculated.

Various other methods, which have been applied for evaluating ECV and TBW in dialysis patients, have limitations (98, 99). ECV associated with optimal perfusion of vital organs, i.e., with the effective arterial blood volume, may vary from the normal value in disease states, e.g., in congestive heart failure (100). Measurement of TBW and ECV by bioimpedance and simultaneous performance of lung ultrasonography to evaluate the extravascular lung ECV is considered a promising method for estimating the dry weight (98, 101, 102). Careful clinical evaluation of the fluid status of patients on hemodialysis is imperative (103, 104). In the future, bioimpedance may become the method of choice for the TBW value used to calculate the volume of the saline infusate in hyponatremic CKD patients. The management of ECV in these patients is best done by combining clinical evaluation, bioimpedance, and lung ultrasonography.

Another source of imprecision in Equations 2 and 3 is that they do not account for exchanges of sodium between the extracellular compartment and the osmotically inactive sodium stored in polyanionic proteoglycans, mainly glycosaminoglycan found in skin, cartilage, and other tissues when [Na] is changing (7). In a study of infusion of hypertonic saline in normal individuals, the rise in [Na] at the end of infusion was almost identical to the rise predicted by the Adrogué–Madias formula, but [Na] decreased 4 h later (105); this decrease in [Na] was not explained by the urinary losses of sodium, potassium, and water. The authors of this study proposed that it was due to the uptake of sodium by proteoglycans. For all these reasons, monitoring of [Na] during and after saline infusion for correction of hyponatremia is critical in patients with or without CKD (19). Measuring urine volume and monovalent cation concentrations in urine will not be needed in every case, but will be useful in some cases with early stages of CKD, foremost in cases of hypovolemic hyponatremia in which infusion of hypertonic saline not only will raise [Na] but also at some point will remove the volume stimulus for vasopressin release, which will result in the production of large volumes of dilute urine. Monitoring urine volume and measuring sodium and potassium concentrations in this urine volume when treatment started and after urine volume starts increasing may provide important therapeutic information.

Hypernatremia may also develop in the setting of hypovolemia, euvolemia, or hypervolemia (82, 106). The proper management of hypernatremia requires the identification of the underlying cause and careful correction (107). The prevention of cerebral edema during treatment is of paramount importance. If hypernatremia is chronic (≥48 h) or of unknown duration, correction of [Na] should not exceed 0.5–1.0 mmol/L per hour or 8–10 mmol/L in the first 24 h, to prevent cerebral edema, permanent neurologic damage, or death (107). Asymptomatic chronic hypernatremia may be corrected over 48 h or longer. More rapid correction (up to 1 mmol/L per hour, up to 8–15 mmol/L over the first 8 h) may be appropriate if the onset of hypernatremia is acute (<48 h), e.g., in accidental sodium loading, and unstable patients (107). In these settings, rapid correction improves the prognosis without increasing the risk of cerebral edema (107). The volume of hypotonic saline needed to produce the desired decrease in [Na] can be computed by modification of Equation 3 as follows (82): (4)VInfused=TBWInitial×[Na]Initial-[Na]Final[Na]Final-[Na]Infusate Hypernatremia can be corrected by infusion intravenously of sterile water (108). When infusion of sterile water is used, Equation 4 should be modified as follows (82): (5)VInfused=TBWInitial×[Na]Initial-[Na]Final[Na]Final Monitoring of [Na] is also imperative during treatment and for a period of at least 24 h after the treatment of hypernatremia with the infusion of hypotonic fluid.

Dangoisse et al. (109) and Rosner and Connor (110) summarized the principles of management of dysnatremias by continuous renal replacement therapy (CRRT): this management requires customizing either the CRRT circuit or the dialysis solution. The CRRT circuit in continuous venovenous hemofiltration or hemodiafiltration can be customized by computing the sodium concentration at the end of the circuit and then adding a post-filter replacement volume of hypertonic or hypotonic fluid calculated to bring the sodium concentration of the circuit to the desired level. The CRRT dialysis solutions for treating severe dysnatremias are customized by adding to the dialysis solution an appropriate volume of hypertonic saline for hypernatremia or sterile water for hyponatremia calculated by a nomogram to bring the sodium concentration of the dialysis fluid to the desired level. The drawbacks of this approach are that the concentrations of other important ingredients of the dialysis fluid, e.g., potassium, calcium, and magnesium are decreased and there are increased chances of error (110).

Dysnatremias are treated by hemodialysis by changing the sodium concentration of the dialysis fluid. The treatment of dysnatremias by peritoneal dialysis will be presented in the peritoneal dialysis section. Attention is required for the prevention of errors during the treatment of severe dysnatremia by dialytic methods (110–113). These procedures require performance in intensive care units by well-trained personnel. During the procedure, the patient should be clinically monitored, and [Na] should be determined frequently, e.g., every 1–2 h. Clinical monitoring should continue and blood chemistries and concentration of sodium in the dialysis fluid should be repeated after the procedure. Point-of-care chemistry measurements are required. Careful application of multidisciplinary protocols for managing dysnatremias by dialytic methods may improve patient outcomes (114).

Dialytic methods employed in the treatment of hyponatremia include CRRT methods, intermittent hemodialysis, and peritoneal dialysis. Table 4 contains reports of management of hyponatremia by dialytic methods (115–130). One issue that requires further study in hyponatremia treated by hemodialysis is the effect of a decrease in plasma urea concentration during the procedure. Experimental studies have demonstrated that a urea concentration gradient between the brain intracellular and extracellular compartments at the end of a hemodialysis session causes osmotic entry of fluid into the brain cells and can contribute to the dialysis disequilibrium syndrome: following hemodialysis, urea concentration in the brain cells of azotemic rats was higher than the corresponding plasma concentration and the brain cells exhibited water gain (131). In addition, the accumulation of organic osmolytes, in particular myoinositol and taurine, was higher in brain cells of azotemic than non-azotemic rats 24 h after correction of chronic hyponatremia (132).

On the basis of rapid correction of severe hyponatremia by hemodialysis with no adverse consequences in one patient, it was suggested that azotemia protects from such consequences when hyponatremia is treated by hemodialysis (133). However, osmotic demyelination syndrome developed after the correction of [Na] from 100 to 121 mmol/L over 2.5 h during the first hemodialysis of a patient with pre-dialysis blood urea of 36.4 mmol/L or 218 mg/dl (134). Sirota and Berl (135) suggested that the risk of osmotic demyelination from rapid correction of hyponatremia by hemodialysis may be low in patients who have been on a regular chronic hemodialysis schedule but is substantially higher in patients starting hemodialysis. The treatment of hyponatremia by hemodialysis using the correction rates of [Na] recommended by existing guidelines for the general population should be combined with a slow rate of reduction of blood urea level by the use of shorter dialysis times, lower blood or dialysis fluid flow rates, less efficient dialyzers, or combinations of these measures. More frequent hemodialysis sessions can compensate for the slow removal of azotemic substances.

The methodology used for treating hyponatremia by peritoneal dialysis requires either the addition to the peritoneal dialysis solution of a volume of hypotonic saline calculated to bring the sodium concentration of this solution to the desired level (130), or performance of hourly exchanges of dialysis solution containing 4.25% dextrose (9). In this last method, free water transfers from the blood compartment to the peritoneal cavity during the early phase of a peritoneal dialysis exchange when using hypertonic dialysis fluid because of sodium sieving (136). One azotemic patient who was treated for the first time with peritoneal dialysis for a [Na] of profound 101 mmol/L developed osmotic myelinolysis with a fatal outcome (137). [Na] of this patient was 126 mmol/L on day 2 and 138 mmol/L on day 3.

Dialytic methods for treating hypernatremia include CRRT (138–142), intermittent hemodialysis (124, 143, 144), and peritoneal dialysis (25, 130, 145). Management of hypernatremia by CRRT is done by customizing the dialysis fluid (139, 141, 142), or infusing hypertonic sodium solutions when there is an urgent need for the very slow rate of reduction in [Na], e.g., in patients with hypernatremia and cerebral edema (140). The rate of reduction of [Na] in hypernatremic patients treated with CRRT should be monitored. In a retrospective study, the hourly rate of reduction in [Na] > 1 mmol/L and dependency on vasopressors were shown to be risk factors for mortality (146). In contrast, a prospective study found no relation between the rate of change in [Na] during CRRT and mortality in dysnatremic patients (147).

Hypernatremia is corrected by hemodialysis by modifying the sodium concentration in the hemodialysis fluid (124, 127, 144). The rate of correction of hypernatremia should be the same as in the section of treatment in patients with CKD not on dialysis. For this reason, the sodium concentration in the dialysis fluid should not differ from [Na] greatly and dialysis should be slow as detailed in the section on the treatment of hyponatremia by hemodialysis. A proposed novel method of producing dialysis fluid allows a larger range of sodium concentration in the dialysis fluid without altering the concentrations of other important ingredients, e.g., potassium and calcium, in this fluid (127). Correction of hypernatremia by peritoneal dialysis is done by lowering the sodium concentration of the dialysis fluid (25) or by using the commercial dialysis solution in documented acute cases in which hypernatremia can be treated at a fast rate (145).