Visual Abstract
Keywords: sodium, survival, hospitalized patients, hyponatremia, hypernatremia, hospital mortality, cohort studies, hospitalization, hospitals
Abstract
Background and objectives
This study aimed to investigate the association between in-hospital trajectories of serum sodium and risk of in-hospital and 1-year mortality in patients in hospital.
Design, setting, participants, & measurements
This is a single-center cohort study. All adult patients who were hospitalized from years 2011 through 2013 who had available admission serum sodium and at least three serum sodium measurements during hospitalization were included. The trend of serum sodium during hospitalization was analyzed using group-based trajectory modeling; the five main trajectories were grouped as follows: (1) stable normonatremia, (2) uncorrected hyponatremia, (3) borderline high serum sodium, (4) corrected hyponatremia, and (5) fluctuating serum sodium. The outcome of interest was in-hospital mortality and 1-year mortality. Stable normonatremia was used as the reference group for outcome comparison.
Results
A total of 43,539 patients were analyzed. Of these, 47% had stable normonatremia, 15% had uncorrected hyponatremia, 31% had borderline high serum sodium, 3% had corrected hyponatremia, and 5% had fluctuating serum sodium trajectory. In adjusted analysis, there was a higher in-hospital mortality among those with uncorrected hyponatremia (odds ratio [OR], 1.33; 95% CI, 1.06 to 1.67), borderline high serum sodium (OR, 1.66; 95% CI, 1.38 to 2.00), corrected hyponatremia (OR, 1.50; 95% CI, 1.02 to 2.20), and fluctuating serum sodium (OR, 4.61; 95% CI, 3.61 to 5.88), compared with those with the normonatremia trajectory. One-year mortality was higher among those with uncorrected hyponatremia (hazard ratio [HR], 1.28; 95% CI, 1.19 to 1.38), borderline high serum sodium (HR, 1.18; 95% CI, 1.11 to 1.26), corrected hyponatremia (HR, 1.24; 95% CI, 1.08 to 1.42), and fluctuating serum sodium (HR, 2.10; 95% CI, 1.89 to 2.33) compared with those with the normonatremia trajectory.
Conclusions
More than half of patients who had been hospitalized had an abnormal serum sodium trajectory during hospitalization. This study demonstrated that not only the absolute serum sodium levels but also their in-hospital trajectories were significantly associated with in-hospital and 1-year mortality. The highest in-hospital and 1-year mortality risk was associated with the fluctuating serum sodium trajectory.
Podcast
This article contains a podcast at https://www.asn-online.org/media/podcast/CJASN/2020_03_25_CJN.12281019.mp3
Introduction
Typically, the osmoregulatory system tightly regulates serum sodium concentration at 135–142 mEq/L to prevent hypotonic or hypertonic stress (1). Based on recent evidence, any degree of deviation in serum sodium outside of the normal range is associated with higher mortality (2). Previous studies suggested that a serum sodium concentration of 138–142 mEq/L is the best physiologically and is associated with the lowest risk of death (3–6).
Dysnatremia is one of the most common electrolyte abnormalities among patients who are hospitalized (7). Because serum sodium concentration affects the conformations of proteins and enzymes, impulse transmissions, and excitation of nerves and muscles, variations in sodium levels—even within a narrow range—might result in physiologic derangements of many organs including heart and brain (8). Dysnatremia, even borderline dysnatremia, on admission to hospital was shown to be associated with a higher risk of in-hospital mortality (9,10). Furthermore, previous studies demonstrated that fluctuation in serum sodium during hospitalization was associated with higher mortality (6,11,12). Although single serum sodium levels have been shown to be associated with patient outcomes, its variations or direction of changes could also provide additional information regarding outcomes of patients who are hospitalized. It is known that the rapid rate of serum sodium disturbances and their correction could lead to neurologic sequelae, but there is no evidence regarding the relationship of the trajectory of serum sodium with outcomes.
We hypothesized that different in-hospital serum sodium trajectories were associated with different in-hospital and 1-year mortality outcomes. Looking at serum sodium trends in addition to a point-specific serum sodium level might better predict mortality. Therefore, the aim of this study is to investigate the association between in-hospital serum sodium trajectories and in-hospital and 1-year mortality rates by conducting a retrospective cohort study among adult patients who had been admitted between January 2011 and December 2013.
Materials and Methods
Study Population
This is a single-center, retrospective, observational study conducted at a tertiary referral hospital. All adult patients admitted in Mayo Clinic Hospital (Rochester, MN) from January 2011 through December 2013, who had available admission serum sodium and at least three serum sodium measurements during hospitalization were included. For patients with multiple hospitalizations during this period, only the first hospital admission was analyzed. For patients who were hospitalized for >30 days, only the first 30 days of in-hospital serum sodium measurements were analyzed. The Mayo Clinic Institutional Review Board approved this study and exempted the need for informed consent due to the minimal-risk nature of the study. All patients provided authorization for the use of their information for research purposes.
Data Collection
Clinical characteristics, demographic information, and laboratory data were collected using automated retrieval from the institutional electronic medical record system. The main predictor of interest was the in-hospital serum sodium trajectory. In-hospital serum sodium trajectory was created based on serum sodium values during the hospital stay. Serum sodium values were not available every single day because they were not always measured daily. However, we did not impute serum sodium on any days that serum sodium was missing. For those who had more than one serum sodium measurement on the same calendar day, the mean serum sodium level was calculated to represent the serum sodium value for the index day. Apart from serum sodium, there were no missing data for other variables.
The outcomes of interest were in-hospital and 1-year mortality rates. The survival status was obtained from the institutional registry and the Social Security Death Index database. Principal diagnoses were grouped based on admission International Classification of Diseases, Ninth Edition codes. eGFR was calculated based on age, sex, race, and serum creatinine, using the CKD Epidemiology Collaboration equation (13). The Charlson Comorbidity Index was calculated to assess the comorbidities at the time of admission. These comorbidities include AIDS, cerebrovascular disease, chronic pulmonary disease, congestive heart failure, connective tissue disease, dementia, diabetes mellitus with and without end-organ damage, hemiplegia, leukemia, lymphoma, malignant tumor, metastatic solid cancer, mild liver disease, moderate-severe liver disease, moderate-severe kidney disease, myocardial infarction, peptic ulcer, and peripheral vascular disease (14,15). The comorbidities were collected using a previously validated data abstraction (16).
Statistical Analyses
Continuous variables were summarized as mean±SD unless otherwise specified. Categoric variables were summarized as numbers with percentages. Continuous and categoric variables were compared among different in-hospital serum sodium trajectories, using ANOVA and the chi-squared test, respectively.
Serum sodium trajectories were created by group-based trajectory modeling to group longitudinal measurements into inter-related subgroups. Group-based trajectory modeling considers the patterns of change for measures across multiple time points and identifies distinctive trajectories (17). It predicts the trajectory for each group, estimates the probability for each individual of group membership, and assigns them to the group based on their highest probabilities which were summarized by a finite set of different polynomial functions of time. The group-based trajectory model was determined using maximum likelihood estimation. Maximizing the likelihood estimation was performed using a general quasi-Newton procedure to handle missing data and provide asymptomatically unbiased parameter estimates (18,19). Group-based trajectory modeling was fitted using linear, quadratic, and cubic polynomials with censored normal distribution for serum sodium measurements. The model with the highest number of fitting categories was selected based on the Bayesian information criterion and the average posterior probability of each trajectory group membership, as well as a preference for a useful parsimonious model which fitted the data well (20). As a result, serum sodium trajectories were categorized into five main trajectories: (1) stable normonatremia, (2) uncorrected hyponatremia, (3) borderline high serum sodium, (4) corrected hyponatremia, and (5) fluctuating serum sodium. All trajectories are depicted in Figure 1.
Figure 1.
In-hospital sodium trajectories.
Multivariable logistic regression analysis was performed to assess the independent association between in-hospital serum sodium trajectories and in-hospital mortality, using stable normonatremia as the reference group. Model 1 was unadjusted whereas in model 2 we adjusted for age, sex, race, eGFR, Charlson Comorbidity Index, principal diagnosis, coronary artery disease, diabetes mellitus, congestive heart failure, peripheral vascular disease, stroke, chronic obstructive pulmonary disease, cirrhosis, length of hospital stay, and numbers of sodium measurements. Model 3 was adjusted for all variables in model 2, plus admission serum sodium. Survival after hospital admission was estimated using the Kaplan–Meier plot and compared using the log-rank test. Patients were followed until death or 1-year after hospital admission. Patients who were lost to follow-up or whose vital status was not known were censored at the date of their last follow-up visit. Cox proportional hazard analysis was performed to assess the association between in-hospital serum sodium trajectories and 1-year mortality, adjusting for previously described variables. A two-tailed P value of <0.05 was considered statistically significant. All analyses were performed using STATA version 14.1 statistical software (StataCorp LLC, College Station, TX).
Results
Clinical Characteristics
A total of 43,539 patients were studied. Using group-based trajectory modeling, all patients were categorized into five in-hospital serum sodium trajectories. The clinical characteristics of patients based on in-hospital serum sodium trajectories are shown in Table 1. The mean age±SD of patients was 64±17 years, 54% were male, and 93% were white. The mean eGFR±SD was 72±30 ml/min per 1.73 m2 and the average admission serum sodium was 138±4 mEq/L. The median (interquartile range) length of stay in hospital and number of serum sodium measurement during hospitalization were 6 (5) days, and five (four) times, respectively. Of all patients, 47% had a stable normonatremia trajectory, 15% had an uncorrected hyponatremia trajectory, 31% had a borderline high serum sodium trajectory, 3% had a corrected hyponatremia trajectory, and 5% had a fluctuating serum sodium trajectory.
Table 1.
Baseline clinical characteristics
| Variables | All | Serum Sodium Trajectories | ||||
|---|---|---|---|---|---|---|
| Stable Normonatremia | Uncorrected Hyponatremia | Borderline Serum High Sodium | Corrected Hyponatremia | Fluctuating Serum Sodium | ||
| N | 43,539 | 21,916 | 5755 | 12,903 | 1344 | 1621 |
| Age (yr) | 64±17 | 62±17 | 66±16 | 65±17 | 70±15 | 69±16 |
| Male | 23,685 (54) | 12,374 (57) | 3207 (56) | 6608 (51) | 613 (46) | 883 (55) |
| White | 40,535 (93) | 20,323 (93) | 5330 (93) | 12,086 (94) | 1265 (94) | 1531 (94) |
| eGFR (ml/min per 1.73 m2) | 72±30 | 75±30 | 70±30 | 71±29 | 71±29 | 60±30 |
| Principal diagnosis | ||||||
| Cardiovascular | 12,614 (29) | 6377 (29) | 1489 (26) | 4025 (31) | 268 (20) | 455 (28) |
| Hematology/oncology | 6207 (14) | 3299 (15) | 928 (16) | 1621 (13) | 180 (13) | 179 (11) |
| Infectious disease | 1392 (4) | 845 (4) | 358 (6) | 548 (4) | 65 (4.8) | 116 (7) |
| Endocrine/metabolic | 1237 (3) | 537 (3) | 195 (3) | 303 (2) | 150 (11) | 52 (3) |
| Respiratory | 2118 (5) | 904 (4) | 317 (6) | 651 (5) | 89 (7) | 157 (10) |
| Gastrointestinal | 4433 (10) | 2087 (10) | 567 (10) | 1408 (11) | 149 (11) | 222 (14) |
| Injury/poisoning | 6562 (15) | 3394 (16) | 820 (14) | 1941 (15) | 185 (14) | 222 (14) |
| Other | 7002 (16) | 3757 (17) | 883 (15) | 1996 (16) | 213 (16) | 153 (9) |
| Charlson Comorbidity Index | 2.1±2.5 | 2.0±2.5 | 2.5±2.7 | 2.0±2.4 | 2.6±2.7 | 2.3±2.4 |
| Comorbidities | ||||||
| Coronary artery disease | 3960 (9) | 1932 (9) | 537 (9) | 1190 (9) | 129 (10) | 172 (11) |
| Congestive heart failure | 4185 (10) | 1855 (9) | 650 (11) | 1281 (10) | 154 (12) | 245 (15) |
| Peripheral vascular disease | 1790 (4) | 830(4) | 298 (5) | 503 (4) | 63 (5) | 96 (6) |
| Stroke | 3834 (9) | 1738 (8) | 530 (9) | 1220 (10) | 150 (11) | 196 (12) |
| Diabetes mellitus | 10,146 (23) | 5065 (23) | 1551 (27) | 1437 (11) | 309 (23) | 425 (26) |
| COPD | 43,539 (11) | 2058 (9) | 661 (12) | 1437 (11) | 192 (14) | 263 (16) |
| Cirrhosis | 1429 (3) | 634 (3) | 323 (6) | 286 (2) | 130 (10) | 56 (4) |
| Admission serum sodium (mEq/L) | 138±4 | 138±3 | 133±4 | 140±3 | 126±6 | 142±5 |
| Lowest serum sodium (mEq/L) | 135±4 | 135±2 | 130±3 | 138±3 | 124±4 | 138±4 |
| Highest serum sodium (mEq/L) | 141±4 | 141±2 | 133±4 | 144±3 | 138±3 | 149±5 |
| Number of serum sodium measurement, median (IQR) | 5 (4) | 5 (4) | 5 (4) | 5 (4) | 5 (4) | 7 (8) |
| Hospital length of stay (d), median (IQR) | 6 (5) | 6 (5) | 7 (5) | 6 (4) | 6 (6) | 8 (11) |
Continuous data are presented as mean±SD unless otherwise indicated; categoric data are presented as count (%). COPD, chronic obstructive pulmonary disease; IQR, interquartile range.
Serum Sodium Trajectories and In-Hospital Mortality
Among 43,539 patients, 808 (2%) died in the hospital. The in-hospital mortality rate was 1% in stable normonatremia, 2% in uncorrected hyponatremia, 2% in borderline high serum sodium, 4% in corrected hyponatremia, and 8% in fluctuating serum sodium. After adjusting for potential confounders in model 2, significantly higher odds of in-hospital mortality was noted among uncorrected hyponatremia (odds ratio [OR], 1.58; 95% CI, 1.27 to 1.96), borderline high serum sodium (OR, 1.49; 95% CI, 1.24 to 1.79), corrected hyponatremia (OR, 2.49; 95% CI, 1.81 to 3.44), and fluctuating serum sodium (OR, 3.78; 95% CI, 3.00 to 4.76) compared with the stable normonatremia group. However, after additionally adjusting for admission serum sodium in model 3, significantly higher odds of in-hospital mortality were noted among those with uncorrected hyponatremia (OR, 1.33; 95% CI, 1.06 to 1.67), borderline high serum sodium (OR, 1.66; 95% CI, 1.38 to 2.00), corrected hyponatremia (OR, 1.50; 95% CI, 1.02 to 2.20), and fluctuating serum sodium (OR, 4.61; 95% CI, 3.61 to 5.88) (Table 2).
Table 2.
The association between serum sodium trajectories in the hospital and in-hospital mortality
| Trajectory of Sodium | In-Hospital Mortality | Model 1a | Model 2b | Model 3c | |||
|---|---|---|---|---|---|---|---|
| OR (95% CI) | P Value | Adjusted OR (95% CI) | P Value | Adjusted OR (95% CI) | P Value | ||
| Stable normonatremia | 249 (1) | 1 (reference) | — | 1 (reference) | — | 1 (reference) | — |
| Uncorrected hyponatremia | 130 (2) | 0.70 (0.48 to 0.91) | <0.001 | 1.58 (1.27 to 1.96) | <0.001 | 1.33 (1.06 to 1.67) | 0.01 |
| Borderline high serum sodium | 245 (2) | 0.52 (0.34 to 0.70) | <0.001 | 1.49 (1.24 to 1.79) | <0.001 | 1.66 (1.38 to 2.00) | <0.001 |
| Corrected hyponatremia | 49 (4) | 1.19 (0.88 to 1.50) | <0.001 | 2.49 (1.81 to 3.44) | <0.001 | 1.50 (1.02 to 2.20) | 0.04 |
| Fluctuating serum sodium | 135 (8) | 2.07 (1.85 to 2.28) | <0.001 | 3.78 (3.00 to 4.76) | <0.001 | 4.61 (3.61 to 5.88) | <0.001 |
OR, odds ratio.
Model 1: unadjusted.
Model 2: adjusted for age, sex, race, GFR, Charlson Comorbidity Index, coronary artery disease, diabetes mellitus, congestive heart failure, peripheral vascular disease, stroke, chronic obstructive pulmonary disease, cirrhosis, principal diagnosis, length of hospital stay, and number of sodium measurement.
Model 3: adjusted for all variables in model 2 and baseline sodium.
Serum Sodium Trajectories and 1-Year Mortality
Based on Kaplan–Meier plots, 1-year mortality after hospital admission was 12% in stable normonatremia, 21% in uncorrected hyponatremia, 13% in borderline high serum sodium, 27% in corrected hyponatremia, and 30% in fluctuating serum sodium (Figure 2). In the adjusted analysis in model 2, significantly higher risk of 1-year mortality was found in those with uncorrected hyponatremia (hazard ratio [HR], 1.47; 95% CI, 1.37 to 1.57), borderline high serum sodium (HR, 1.08; 95% CI, 1.01 to 1.14), corrected hyponatremia (HR, 1.87; 95% CI, 1.68 to 2.10), and fluctuating serum sodium (HR, 1.80; 95% CI, 1.63 to 1.99) compared with the stable normonatremia group. After additionally adjusting for the admission serum sodium in model 3, significantly higher risk of 1-year mortality was found in those with uncorrected hyponatremia (HR, 1.28; 95% CI, 1.19 to 1.38), borderline high serum sodium (HR, 1.18; 95% CI, 1.11 to 1.26), corrected hyponatremia (HR, 1.24; 95% CI, 1.08 to 1.42), and fluctuating serum sodium (HR, 2.10; 95% CI, 1.89 to 2.33) compared with the stable normonatremia group (Table 3).
Figure 2.
Kaplan–Meier survival estimates of 1-year mortality among each serum sodium trajectory.
Table 3.
The association between serum sodium trajectories in the hospital and 1-yr mortality
| Trajectory of Sodium | 1-yr Mortality | Model 1a | Model 2b | Model 3c | |||
|---|---|---|---|---|---|---|---|
| HR (95% CI) | P Value | Adjusted HR (95% CI) | P Value | Adjusted HR (95% CI) | P Value | ||
| Stable normonatremia | 2587 (12) | 1 (reference) | — | 1 (reference) | — | 1 (reference) | — |
| Uncorrected hyponatremia | 1187 (21) | 1.81 (1.69 to 1.94) | <0.001 | 1.47 (1.37 to 1.57) | <0.001 | 1.28 (1.19 to 1.38) | <0.001 |
| Borderline high serum sodium | 1690 (13) | 1.13 (1.07 to 1.21) | <0.001 | 1.08 (1.01 to 1.14) | 0.02 | 1.18 (1.11 to 1.26) | <0.001 |
| Corrected hyponatremia | 356 (27) | 2.44 (2.19 to 2.73) | <0.001 | 1.87 (1.68 to 2.10) | <0.001 | 1.24 (1.08 to 1.42) | 0.003 |
| Fluctuating serum sodium | 486 (30) | 2.93 (2.66 to 3.23) | <0.001 | 1.80 (1.63 to 1.99) | <0.001 | 2.10 (1.89 to 2.33) | <0.001 |
HR, hazard ratio.
Model 1: unadjusted.
Model 2: adjusted for age, sex, race, GFR, Charlson Comorbidity Index, coronary artery disease, diabetes mellitus, congestive heart failure, peripheral vascular disease, stroke, chronic obstructive pulmonary disease, cirrhosis, principal diagnosis, length of hospital stay, and number of sodium measurement.
Model 3: adjusted for all variables in model 2 and baseline sodium.
Discussion
Our retrospective cohort study demonstrated that—compared with stable normonatremia—sodium trajectories described as uncorrected hyponatremia, borderline high serum sodium, corrected hyponatremia, and fluctuating serum sodium were associated with significantly higher risks of in-hospital and 1-year mortality. Of these, the fluctuating serum sodium trajectory represented the highest risk trajectory for both in-hospital and 1-year mortality. These associations persisted after adjustment for admission serum sodium.
It is well documented that point-specific hyponatremia and hypernatremia are significantly associated with in-hospital and 1-year mortality. Moreover, a serum sodium change of >6 mEq/L during hospitalization was associated with higher mortality (6). Although our study supports these findings, unlike other studies that were limited by generalizability such as the study by Ng et al. (21) which examined the association between trajectories of hyponatremia and mortality after acute pulmonary embolism, our study captured all-comers to a large medical center who had a minimum of three serum sodium measurements. Furthermore, our analysis uniquely modeled the serum sodium trajectory for each patient. To the best of our knowledge, it is the first study of its kind to accomplish this. Our study underscores the value of evaluating the trajectory of the serum sodium rather than a static value.
Hypernatremia represents hypertonicity, which potentially damages the cytoskeleton and DNA, whereas hyponatremia usually indicates hypotonicity, which leads to cellular swelling and membrane rupture (22). Previous literature has suggested that correcting hyponatremia is associated with a lower risk in overall mortality and better neurocognitive functioning, particularly among the elderly (23–27). However, our study found that both corrected and uncorrected hyponatremia trajectories had similar in-hospital mortality (P=0.52) and 1-year mortality (P=0.64). The insignificant difference in mortality among both groups could potentially be explained by a mild degree of hyponatremia in uncorrected hyponatremia or an inadequate sample size. Furthermore, future studies are needed to assess if severity and duration of hyponatremia, among those in the uncorrected hyponatremia group, and the rate of sodium correction, among those in the corrected hyponatremia group, play an important role in the insignificant difference in mortality among these two groups.
Interestingly, we found that the magnitude of the association between the fluctuating serum sodium trajectory and mortality was significantly higher than that of other trajectories. The reason for this is not clear, but there are several hypotheses. We postulated that the fluctuating serum sodium creates osmotic stress that affects cellular functioning and viability via dysregulation of the interaction between protein phosphatase 6 (PP6) and apoptosis signal-regulating kinase 3 (ASK3), and the cell volume recovery system. Cell volume regulation is a vital system for cellular homeostasis and osmotic stress is a major threat to cells. PP6 inactivates ASK3 through a hyperosmolar-dependent interaction and ASK3 bidirectionally adjusts cell size under both hypo-osmotic and hyperosmotic stress (28,29). However, repetitive osmotic stress can perturb cell function and cause PP6-ASK3 dysregulation, resulting in DNA damage, cell cycle arrest, and apoptosis (22,30). Alternatively, fluctuating serum sodium also reflects impairment in the osmoregulation required to maintain balance in serum sodium, which might be responsible for the higher risk of mortality. Moreover, critical illness and fluid management might remain as residual confounders. Although we comprehensively adjusted for potential confounders, we could not completely exclude the possibility that patients with fluctuating serum sodium might have represented a more critically ill population that required fluid management or administration of hypertonic or hypotonic solutions, which could lead to abrupt and profound changes in serum sodium. Future studies are required to examine whether the association of fluctuating serum sodium trajectory with mortality is mainly influenced by the characteristics of the fluctuating trajectory group or iatrogenic causes related to administration of fluid management. In addition, further investigations are needed to identify optimal management and strategies to decrease mortality risk among patients with a fluctuating serum sodium trajectory during hospitalization.
The limitations of our study are worth mentioning. First, regarding group-based trajectory modeling, we assumed that missing serum sodium values were missing at random. However, imputations of missing data were not performed to avoid biasing the study result. Second, we did not have data on the type of fluid replacement or management strategies for serum sodium correction, nor did we have information about blood glucose or lipid levels that can affect the serum sodium concentration. The lack of these data limits our understanding of situations causing serum sodium change or fluctuation during hospitalization. Third, the effect of serum sodium trajectories on other patient outcomes—in particular, neurologic outcomes such as osmotic demyelination syndrome (ODS), cerebral edema, or seizure—was not assessed in this study. By using the International Classification of Diseases, Ninth Edition diagnosis code followed by medical record review, only one patient with clinical and radiologic findings consistent with osmotic demyelination syndrome was identified in our entire cohort. This patient with ODS was in the corrected hyponatremia trajectory group. Moreover, the incidence of ODS was low at the Mayo Clinic (31). Future studies should investigate the association between serum sodium trajectories and the risk of adverse neurologic outcomes. Finally, our study was a retrospective, observational investigation. Therefore, the association between serum sodium trajectories and in-hospital mortality or 1-year mortality does not prove causality and we cannot infer that there were no residual confounders, even after adjusting for several known confounders. Nevertheless, several strengths in our study should be highlighted. We conducted the study on a large sample of 43,539 patients, allowing us to adjust for several confounders with a lower risk of overfitting. Additionally, using group-based trajectory modeling resulted in a linear, quadratic, or cubic model that allowed us to visualize the changes in a particular trend over time, captured nonlinearity, and provided more accuracy compared with a multiple linear regression model.
In conclusion, more than half of patients who had been hospitalized had an abnormal serum sodium trajectory during hospitalization. Our study demonstrated that not only the absolute serum sodium levels but also their in-hospital trajectories of serum sodium levels were significantly associated with in-hospital and 1-year mortality rates. The risk associated with death was highest when serum sodium fluctuated considerably during hospitalization.
Disclosures
Dr. Cheungpasitporn, Dr. Chewcharat, Dr. Kashani, Dr. Mao, Dr. Thirunavukkarasu, and Dr. Thongprayoon have nothing to disclose.
Funding
None.
Acknowledgments
All authors had access to the data and a role in writing the manuscript.
This work was performed at Mayo Clinic in Rochester, Minnesota.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
References
- 1.Sterns RH: Disorders of plasma sodium--causes, consequences, and correction. N Engl J Med 372: 55–65, 2015 [DOI] [PubMed] [Google Scholar]
- 2.Corona G, Giuliani C, Parenti G, Norello D, Verbalis JG, Forti G, Maggi M, Peri A: Moderate hyponatremia is associated with increased risk of mortality: Evidence from a meta-analysis. PLoS One 8: e80451, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tsipotis E, Price LL, Jaber BL, Madias NE: Hospital-associated hypernatremia spectrum and clinical outcomes in an unselected cohort. Am J Med 131: 72–82.e1, 2018 [DOI] [PubMed] [Google Scholar]
- 4.Wald R, Jaber BL, Price LL, Upadhyay A, Madias NE: Impact of hospital-associated hyponatremia on selected outcomes. Arch Intern Med 170: 294–302, 2010 [DOI] [PubMed] [Google Scholar]
- 5.Kumar S, Berl T: Sodium. Lancet 352: 220–228, 1998 [DOI] [PubMed] [Google Scholar]
- 6.Thongprayoon C, Cheungpasitporn W, Yap JQ, Qian Q: Increased mortality risk associated with serum sodium variations and borderline hypo- and hypernatremia in hospitalized adults [published online ahead of print June 20, 2019]. Nephrol Dial Transplant doi: 10.1093/ndt/gfz098 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rose BD: New approach to disturbances in the plasma sodium concentration. Am J Med 81: 1033–1040, 1986 [DOI] [PubMed] [Google Scholar]
- 8.Waikar SS, Mount DB, Curhan GC: Mortality after hospitalization with mild, moderate, and severe hyponatremia. Am J Med 122: 857–865, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Girardeau Y, Jannot A-S, Chatellier G, Saint-Jean O: Association between borderline dysnatremia and mortality insight into a new data mining approach. BMC Med Inform Decis Mak 17: 152, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hu J, Wang Y, Geng X, Chen R, Zhang P, Lin J, Teng J, Zhang X, Ding X: Dysnatremia is an independent indicator of mortality in hospitalized patients. Med Sci Monit 23: 2408–2425, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Marshall DC, Salciccioli JD, Goodson RJ, Pimentel MA, Sun KY, Celi LA, Shalhoub J: The association between sodium fluctuations and mortality in surgical patients requiring intensive care. J Crit Care 40: 63–68, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sen S, Tran N, Chan B, Palmieri TL, Greenhalgh DG, Cho K: Sodium variability is associated with increased mortality in severe burn injury. Burns Trauma 5: 34, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration): A new equation to estimate glomerular filtration rate [published correction appears in Ann Intern Med 155: 408, 2011]. Ann Intern Med 150: 604–612, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Charlson M, Szatrowski TP, Peterson J, Gold J: Validation of a combined comorbidity index. J Clin Epidemiol 47: 1245–1251, 1994 [DOI] [PubMed] [Google Scholar]
- 15.Charlson ME, Pompei P, Ales KL, MacKenzie CR: A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 40: 373–383, 1987 [DOI] [PubMed] [Google Scholar]
- 16.Singh B, Singh A, Ahmed A, Wilson GA, Pickering BW, Herasevich V, Gajic O, Li G: Derivation and validation of automated electronic search strategies to extract Charlson comorbidities from electronic medical records. Mayo Clin Proc 87: 817–824, 2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Nagin D: Group-Based Modeling of Development, Cambridge, MA, Harvard University Press, 2005 [Google Scholar]
- 18.Nagin DS, Jones BL, Passos VL, Tremblay RE: Group-based multi-trajectory modeling. Stat Methods Med Res 27: 2015–2023, 2018 [DOI] [PubMed] [Google Scholar]
- 19.Jones BL, Nagin DS: A note on a stata plugin for estimating group-based trajectory models. Sociol Methods Res 42: 608–613, 2013 [Google Scholar]
- 20.Shearer DM, Thomson WM, Broadbent JM, McLean R, Poulton R, Mann J: High-risk glycated hemoglobin trajectories established by mid-20s: Findings from a birth cohort study. BMJ Open Diabetes Res Care 4: e000243, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ng ACC, Chow V, Yong ASC, Chung T, Kritharides L: Fluctuation of serum sodium and its impact on short and long-term mortality following acute pulmonary embolism. PLoS One 8: e61966, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Burg MB, Ferraris JD, Dmitrieva NI: Cellular response to hyperosmotic stresses. Physiol Rev 87: 1441–1474, 2007 [DOI] [PubMed] [Google Scholar]
- 23.Corona G, Giuliani C, Verbalis JG, Forti G, Maggi M, Peri A: Hyponatremia improvement is associated with a reduced risk of mortality: Evidence from a meta-analysis. PLoS One 10: e0124105, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Brinkkoetter PT, Grundmann F, Ghassabeh PJ, Becker I, Johnsen M, Suaréz V, Schulz RJ, Streichert T, Burst V: Impact of resolution of hyponatremia on neurocognitive and motor performance in geriatric patients. Sci Rep 9: 12526, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Refardt J, Kling B, Krausert K, Fassnacht M, von Felten S, Christ-Crain M, Fenske W: Impact of chronic hyponatremia on neurocognitive and neuromuscular function. Eur J Clin Invest 48: e13022, 2018 [DOI] [PubMed] [Google Scholar]
- 26.Vandergheynst F, Gombeir Y, Bellante F, Perrotta G, Remiche G, Mélot C, Mavroudakis N, Decaux G: Impact of hyponatremia on nerve conduction and muscle strength. Eur J Clin Invest 46: 328–333, 2016 [DOI] [PubMed] [Google Scholar]
- 27.Ahluwalia V, Heuman DM, Feldman G, Wade JB, Thacker LR, Gavis E, Gilles H, Unser A, White MB, Bajaj JS: Correction of hyponatraemia improves cognition, quality of life, and brain oedema in cirrhosis. J Hepatol 62: 75–82, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Watanabe K, Umeda T, Niwa K, Naguro I, Ichijo H: A PP6-ASK3 module coordinates the bidirectional cell volume regulation under osmotic stress. Cell Reports 22: 2809–2817, 2018 [DOI] [PubMed] [Google Scholar]
- 29.Dai G, Yu H, Kruse M, Traynor-Kaplan A, Hille B: Osmoregulatory inositol transporter SMIT1 modulates electrical activity by adjusting PI(4,5)P2 levels. Proc Natl Acad Sci U S A 113: E3290–E3299, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ohama T: The multiple functions of protein phosphatase 6. Biochim Biophys Acta Mol Cell Res 1866: 74–82, 2019 [DOI] [PubMed] [Google Scholar]
- 31.Geoghegan P, Harrison AM, Thongprayoon C, Kashyap R, Ahmed A, Dong Y, Rabinstein AA, Kashani KB, Gajic O: Sodium correction practice and clinical outcomes in profound hyponatremia. Mayo Clin Proc 90: 1348–1355, 2015 [DOI] [PubMed] [Google Scholar]