Background: Hyponatremia and delirium are frequent problems in older hospitalized patients. Although confusional states are considered to be a possible complication of hyponatremia, there has been no systematic study to date of the precise prevalence of delirium among patients with hyponatremia and its effect on long-term outcomes.
Methods: In a 13-month period in 2009/2010, all patients with a serum sodium level less than or equal to 130 mmol/L (the hyponatremia group) in a cohort of hospitalized older patients were studied and compared to a normonatremic control group of patients who were matched for age, sex, and diagnosis group. The prevalence of delirium was determined by two-stage examination. In-hospital mortality, mortality six months after initial examination, and functional status were prospectively analyzed.
Results: 179 patients were identified whose serum sodium level was less than or equal to 130 mmol/L (7.9% of all treated patients), of whom 141 were included in the hyponatremia group. The mean age of the participants was 83 (range, 63–102), and 84% were women. Patients with hyponatremia suffered more often from delirium (22.7% versus 8.5%; p = 0.002) and had a higher in-hospital mortality (10.6% versus 2.1%; p = 0.005). The mortality six months after initial examination was 31.9% versus 22.7% (p = 0.080). 59.7% of patients in the hyponatremia group and 49% in the control group (p = 0.146) needed a higher level of chronic care after discharge than they had needed before the hospitalization.
Conclusion: Hyponatremia in hospitalized older patients is associated with a higher likelihood of delirium and an elevated in-hospital mortality.
With a prevalence of 15–30%, hyponatremia is the most common electrolyte abnormality in hospital inpatients (1–3). Geriatric patients are a high-risk group because of their changed physiology, multimorbidities, and polypharmacy (4). In older persons, hyponatremia is usually mildly to moderately pronounced, with a serum sodium (SNa) concentration of 130–134 mmol/L and 125–129 mmol/L, respectively, as well as chronic—that is, the hyponatremia developed within a time period of more than 48 hours (5). Hyponatremia has a negative effect on a multitude of clinical parameters, among these several functional aspects that are reflected in the basic geriatric assessment (6, 7), and on cognitive performance. Even a mildly to moderately lowered sodium concentration is associated with cognitive impairment (8), which may improve again once the sodium concentration has been balanced (9). In contrast to acute hyponatremia, which results in cerebral edema (10), the mechanisms that lead to cognitive impairment in chronic hyponatremia remain unclear (4). The few studies on the effect of chronic hyponatremia on cognition described in an undifferentiated way cognitive impairments that had been recorded by means of screening instruments for dementia (6, 8) or an attention test (9). None of the studies documented whether participants were subject to delirium. The prevalence of delirium in older patients in general medical or geriatric wards is 29–64% (11). It constitutes a severe complication of the inpatient stay, with mortality 1.5 times higher in the first year after hospital admission (11), poorer long-term outcomes in terms of cognition (12), a reduction in functional performance, and a rate of institutionalization of 33.4% in patients with delirium, compared with 10.7% in patients without delirium (13).
Our study was the first to investigate the association of hyponatremia with delirium in a cohort of geriatric hospital inpatients compared with a carefully matched control group without hyponatremia. Subsequently we prospectively described the effects of hyponatremia and delirium on mortality and independent living in follow-up examinations during the inpatient stay and after six months.
We planned a prospective observational study with paired controls and a follow-up data collection six months later over the telephone. Heidelberg University’s ethics committee approved the study protocol (application number S-224/2009). The study was registered with the German Clinical Trials Register (Deutsches Register Klinischer Studien, DRKS), registration number DRKS00004280.
Setting and participants
Study participants were consecutively recruited at AGAPLESION Bethanien Hospital Heidelberg, the university’s geriatric center, in the acute geriatric ward and the inpatient geriatric rehabilitation ward from May 2009 to June 2010.
On each working day, the laboratory variables of all patients with hyponatremia (SNa < 135 mmol/L) that had been measured on that day were recorded.
Hyponatremia group (HNG): All patients with moderately severe or severe hyponatremia (SNa ≤ 130 mmol/l) were eligible for inclusion in the study. We excluded patients whom we had not been able to examine within 24 hours or who refused participation in the study. If patients had cognitive impairment, their legal representatives agreed to participation; where necessary, where necessary, guardianship was instituted by local authorities. Each patient was included in the study only once. The treatment of hyponatremia was the responsibility of the treating physicians. The study did not affect the frequency of laboratory tests nor the treatment of hyponatremia.
Control group (CG): Each patient in the hyponatremia group was matched with a suitable patient with a serum sodium concentration that was consistently in the normal range. To this end, we selected for every case the patient who next received hospital treatment, who was comparable in terms of sex, age group (65–80 years or ≥ 80), and main diagnosis, and who gave written consent. The main diagnoses were combined into groups according to the codes of the International Statistical Classification of Diseases and Related Health Problems (ICD-10), following the chapter structure of the ICD-10. ICD-subgroups with disease symptoms such as cirrhosis of the liver, in whom hyponatremia typically occurs more often, were documented as independent groups (eTable).
We used the routine data shown in Table 1 to check the comparability of the groups. All drugs were recorded as possible risk factors.
Cross-sectional study: delirium examination (T1)
All patients in both groups were examined for delirium within 24 hours after inclusion in the study (T1 examination). This was done in two stages: the mini mental state examination (MMSE; 15) confirmed cognitive impairment, and the confusion assessment method (CAM; 19, 20) confirmed delirium. In order to do justice to the potential fluctuation of the symptoms of delirium, we searched the documentation kept by the nursing staff and doctors for indications of delirium. Where this was suspected, we questioned nurses and doctors accordingly. In the second step, medical specialists ascertained the diagnosis on the basis of the criteria of the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). The screening examinations were not blinded regarding the patients’ group allocation, but the subsequent investigations were. In a small number of cases however, the investigators were also the responsible treating physicians, so that complete blinding was not guaranteed. In order to distinguish delirium and dementia, or delirium in dementia, family members were surveyed in a standardized manner about the cognitive status five years earlier and 14 days earlier, by using Jorm’s (14) Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) questionnaire.
After seven days, the same standardized two-step diagnostic test for delirium was undertaken (T2 examination). Hospital mortality was determined after discharge.
A follow-up study took place six months after inclusion into the study. The long term effects of hyponatremia and delirium on mortality (primary analysis), housing situation, and level of care (‘Pflegestufe’, a four-step grading system provided by the German Social Security system) (secondary analysis) were documented by means of a questionnaire-supported, semistructured telephone survey of patients and/or their relatives. If it was not possible to reach these people then we interviewed their general practitioners. The investigation was blinded with regard to group allocation.
Sample size calculation: We calculated the sample size for the primary endpoint rate of delirium. The risk of delirium in patients with hyponatremia was estimated on the basis of a previously conducted pilot study, in which 10 out of 27 patients had delirium. In a more conservative calculation we assumed a rate of 35% for the sample size calculation. For the control group we used as our basis a study of delirium conducted in 2003/2004 in our hospital, in which 12 out of 50 patients with dementia had delirium. As dementia increases the risk of delirium, we estimated a rate of 20%. On the assumption of an α error of 5% and a statistical power of 80%, a sample of 109 cases per group was required in order to prove statistically significant differences. In order to be able to compensate for a dropout rate of 20% we attempted to recruit 137 patients per group.
Analysis of results: For categorical variables we presented absolute and relative frequencies. When comparing categorical variables between groups we used the McNemar test for paired samples, and for continuous variables we used the Wilcoxon test. We used the χ2 test to compare categorical variables of non-paired subgroups. We used the t test for dependent variables to test the development over time of the sodium concentrations. Patients who had died in hospital were categorized as deceased in the follow-up examination. We assigned a higher level of care to deceased patients in order to avoid survivorship bias. The secondary endpoints (mortality, institutionalization, level of care) were evaluated in a purely descriptive way. The significance level was defined as P = 0.05. We used the software packages SPSS (Statistical Package for the Social Sciences, version 21) and R (23) for our analyses.
In the study period, a total of 2267 patients (excluding patients with repeated admissions received inpatient treatment in the AGAPLESION Bethanien Hospital Heidelberg. The prevalence of serum sodium concentrations <135 mmol/L was 20.7% (470 out of 2267), seen in 22.7% (356 of 1566) of women and in 16.3% (114 of 701) of men. A serum sodium concentration of ≤ 130 mmol/L was seen in 179 patients; the prevalence was 7.9% (9.5% in women and 4.4% in men). Each group included 141 patients (118 women, 23 men) (Figure 1). The distribution of main diagnoses is shown in the eTable.
The groups did not differ with regard to social situation, independence, comorbidities, and mobility (Table 1). Because of work flow planning, we conducted the T1 examination in the control group at a later date during their inpatient stay (a median of 8 days later). According to family members, patients in the hyponatremia group had experienced cognitive deterioration significantly more often, in the preceding five years as well as in the preceding 14 days. Patients in the hyponatremia group took thiazide diuretics significantly more often and loop diuretics significantly less often. The difference in the consumption of other drugs did not reach statistical significance.
A comparison of baseline data of patients with or without delirium showed the following results for those with delirium:
Progression of hyponatremia
In 79.4% of cases, patients had hyponatremia at the time of their admission to the hospital. The average sodium concentration rose from 127.0 ± 3.9 mmol/L (T1) to 133.0 ± 4.6 mmol/L (T2) (P <0.001) within a week.
Occurrence of delirium
Delirium was significantly more common in the hyponatremia group (22.7 % versus 8.5 %, P = 0.002), and the same was true for the subgroup with moderate hyponatremia (20.7 %, P = 0.005). Seven days later, differences between the groups did no longer reach significance (Table 2). In the hyponatremia group, the delirium went into remission in 12 of 28 patients, in the control group in only 2 of 11 patients.
In-hospital mortality was five times higher in patients in the hyponatremia group than in those in the control group. After six months, the outcomes seen in the hyponatremia group were worse (mortality 32% versus 23%, P=0.080; mortality or institutionalization 46% versus 35%, P=0.067), especially for the subgroup without delirium (45 % versus 32 %, P = 0.036) (Table 3, Figure 2).
In the total population, delirium was associated with a 2.5-fold increase in in-hospital mortality. Mortality after 6 months rose in patients with delirium in the total population as well as in the control group (Table 3, Figure 2).
An increased need for care, determined on the basis of the level of care, in both groups was associated with delirium. Hyponatremia, however, did not seem to have any relevant effect on the level of care (Table 3).
The present study of geriatric patients showed a strong association of hyponatremia—even moderate hyponatremia—with delirium.
The prevalence of hyponatremia (SNa concentration <135 mmol/L in 20.7% and 131–118 mmol/L in 6.0%) was higher than in the geriatric cohort studied by Gosch et al. (6) (16.7% and 4.5%). The prevalence of milder forms was similar to the values in a review (5) for geriatric wards (SNa concentration <135 mmol/L: 20.7% versus 22.2%; SNa concentration <130 mmol/L: 4.7% versus 7.5%). The prevalence of severe hyponatremia was much lower (SNa concentration <125 mmol/L: 0.9% versus 4.5%). As described earlier (25), the prevalence in women was higher than in men (9.5% versus 4.4%).
The prevalence of delirium was clearly higher in the hyponatremia group than in the control group. Earlier studies, mostly retrospective observational studies, showed that attention and cognitive performance are impaired in hyponatremia, which was interpreted as a causal association (4, 6, 8, 9). Notably, however, hyponatremia and delirium have important risk factors in common, especially age and underlying conditions. Conclusions about causality are therefore difficult to confirm on the basis of retrospective or uncontrolled studies. Furthermore, monocausal associations seem fundamentally unlikely in a multimorbid patient population. In the present study, the researchers were able to rule out important confounding variables because of comparisons with a matched control group. This adds support to the hypothesis that hyponatremia, even when less pronounced, may be regarded as a relevant cofactor in the pathogenesis of delirium. However, the fact that the hyponatremia group included a higher proportion of patients in whom relatives reported having observed cognitive deterioration before hospital admission illustrates the complexity of the association between hyponatremia and delirium. Factors that cause cognitive deterioration may also facilitate hyponatremia.
In-hospital mortality was five times higher in the hyponatremia group, even without a diagnosis of delirium. This means that in our study, mortality was higher than in a recent study (7) in patients who were younger (median 71 years) and in whom hospital mortality was 4%. In patients without delirium, the long-term study found a poorer prognosis for hyponatremia. Earlier studies identified hyponatremia as an independent risk factor (25, 26). The comparison with a matched control group with comparable comorbidities in our study supports the hypothesis that hyponatremia directly contributes to mortality.
Our study confirms that the administration of thiazide diuretics is a relevant risk factor for hyponatremia in older patients (27, 28). The fact that guidelines still recommend thiazide diuretics as the first-line medication for the treatment of arterial hypertension should be subject to critical revision in the light of these data.
An earlier study identified delirium as an independent risk factor for mortality and admission to institutionalized care (30). Accordingly, in the control group, delirium was associated with an increase in long-term mortality from 20% to 50% compared with patients without delirium. In the total cohort, mortality after delirium was slightly higher—at 43.2%—than in other geriatric cohorts with death rates of up to 40% (30).
Limitations and strengths
Our data cannot describe the causes of hyponatremia or the volume status with any certainty, but this is a fundamental difficulty in the geriatric population (4). Screening for delirium was not undertaken every day but only twice, at a week’s interval. Furthermore, the patients in the control group were examined at a later time during their inpatient stay. We therefore cannot rule out the possibility that the delirium had already subsided. Thorough review of the documentation and discussion with the staff treating the patients did, however, enable us to include indications of delirium at other time points. As the precise onset of hyponatremia cannot be defined in most patients, no conclusions can be drawn regarding the sequence in which hyponatremia and delirium occurred. In order to ensure as much objectivity in the diagnostic evaluation of delirium, standardized approaches were used in the first stage, and in the second stage, the researchers were not informed about patients’ sodium status. Complete blinding was, however, impossible in the clinical setting, which may have affected the assessment/evaluation in spite of the careful approach. The follow-up examination was completely blinded. The power analysis related to the prevalence of delirium depending on hyponatremia. The study was not adequately powered for the purposes of subgroup comparison regarding long term mortality.
The study had the following strengths:
Our results allow us to draw conclusions about the treatment of geriatric patients and further research in this area. Hyponatremia is common in older patients and is a risk factor for delirium even when mild. Furthermore, our study indicates that hyponatremia increases in-hospital mortality even in the absence of delirium and leads to poor long-term outcomes. The study did not show any increases in long-term mortality. Thiazide diuretics seem to constitute a crucial, avoidable risk factor. In view of its high prevalence, the absence of an algorithm for the diagnostic evaluation of hyponatremia deserves criticism, as does the fact that the existing guidelines are still based largely on expert recommendations, owing to lacking evidence (31).
Conflict of interests statement
The authors declare that no conflict of interest exists.
Manuscript received on 17 March 2016, revised version accepted on
11 August 2016.
Translated from the original German by Birte Twisselmann, PhD.
Dr. med. Tania Zieschang
Geriatrisches Zentrum an der Universität Heidelberg
Agaplesion Bethanien Krankenhaus Heidelberg
Rohrbacher Str. 149
69126 Heidelberg, Germany
|1.||Correia L, Ferreira R, Correia I, et al.: Severe hyponatremia in older patients at admission in an internal medicine department. Arch Gerontol Geriatr 2014; 59: 642–7 CrossRef MEDLINE|
|2.||Upadhyay A, Jaber BL, Madias NE: Incidence and prevalence of hyponatremia. Am J Med 2006; 119: S30–5 CrossRef MEDLINE|
|3.||Waikar SS, Mount DB, Curhan GS: Mortality after hospitalization with mild, moderate, and severe hyponatremia. Am J Med 2009; 122: 857–65 CrossRef MEDLINE PubMed Central|
|4.||Soiza RL, Cumming K, Clarke JM, Wood KM, Phyo KM: Hyponatremia: Special considerations in older patients. J Clin Med 2014; 3: 944–58 CrossRef MEDLINE PubMed Central|
|5.||Mannesse CK, Vondeling AM, Van Marum RJ, Van Solinge WW, Egberts TC, Jansen PA: Prevalence of hyponatremia on geriatric wards compared to other settings over four decades: a systematic review. Ageing Res Rev 2013; 12: 165–73 CrossRef MEDLINE|
|6.||Gosch M, Joosten-Gstrein B, Heppner HJ, Lechleitner M: Hyponatremia in geriatric inhospital patients: effects on results of a comprehensive geriatric assessment. Gerontology 2012; 58: 430–40 CrossRef MEDLINE|
|7.||Nigro N, Winzeler B, Suter-Widmer I, et al.: Symptoms and characteristics of individuals with profound hyponatremia: a prospective multicenter observational study. J Am Geriatr Soc 2015; 63: 470–5 CrossRef MEDLINE|
|8.||Gunathilake R, Oldmeadow C, McEvoy M, et al.: Mild hyponatremia is associated with impaired cognition and falls in community-dwelling older persons. J Am Geriatr Soc 2013; 61: 1838–9 CrossRef MEDLINE|
|9.||Renneboog B, Musch W, Vandemergel X, Manto MU, Decaux G: Mild chronic hyponatremia is associated with falls, unsteadiness, and attention deficits. Am J Med 2006; 119: 71.e1–8 CrossRef MEDLINE|
|10.||Ayus JC, Arieff AI: Pathogenesis and prevention of hyponatremic encephalopathy. Endocrinol Metab Clin N Am 1993; 22: 425–46 MEDLINE|
|11.||Inouye SK, Westendorp RGJ, Saczynski JS: Delirium in elderly people. Lancet 2014; 383: 911–22 CrossRef|
|12.||Gross AL, Jones RN, Habtemariam DA, et al.: Delirium and long-term cognitive trajectory among persons with dementia. Arch Intern Med 2012; 156: 848–56 CrossRef|
|13.||Witlox J, Eurelings LS, de Jonghe JF, Kalisvaart KJ, Eikelenboom P, van Gool WA: Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementia: a meta-analysis. JAMA 2010; 304 : 443–51 CrossRef MEDLINE|
|14.||Jorm AF: The informant questionnaire on cognitive decline in the elderly (IQCODE): a review. Int Psychogeriatr 2004; 16: 275–93 CrossRef|
|15.||Folstein MF, Folstein SE, McHugh PR: “Mini-mental state”. A practical method for grading the cognitive status of patients for the clinician. J Psychiatr Res 1975; 12: 189–98 CrossRef|
|16.||Mahoney FI, Barthel DW: Functional evaluation: the Barthel Index. Md State Med J 1965; 14: 61–5 MEDLINE|
|17.||Tinetti ME: Performance-oriented assessment of mobility problems in elderly patients. J Am Geriatr Soc 1986; 34: 119–26 CrossRef|
|18.||Parmelee PA, Thuras PD, Katz IR, Lawton MP: Validation of the Cumulative Illness Rating Scale in a geriatric residential population. J Am Geriatr Soc 1995; 43: 130–7 CrossRef|
|19.|| Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI: Clarifying confusion: the confusion assessment method. A new|
method for detection of delirium. Ann Intern Med 1990; 113: 941–8 CrossRef MEDLINE
|20.||Hestermann U, Backenstrass M, Gekle I, et al.: Validation of a German version of the Confusion Assessment Method for delirium detection in a sample of acute geriatric patients with a high prevalence of dementia. Psychopathology 2009; 42: 270–6 CrossRef MEDLINE|
|21.|| Diegert C, Diegert KV: Note on inversion of casagrande-pike-smith approximate sample-size formula for Fisher-Irwin Test on 2x2|
tables. Biometrics 1981; 37: 595.
|22.||Thomas C, Hestermann U, Walther S, et al.: Prolonged activation EEG differentiates dementia with and without delirium in frail elderly patients. J Neurol Neurosurg Psychiatr 2008; 79: 119–25 CrossRef MEDLINE|
|23.||R Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing: Vienna 2015. www.R-project.org (last accessed on 26 July 2016)|
|24.|| Shapiro DS, Sonnenblick M, Galperin I, Melkonyan L, Munter G:|
Severe hyponatremia in elderly hospitalized patients: prevalence, aetiology and outcome. Intern Med J 2010; 40: 574–80 CrossRef MEDLINE
|25.||Hoorn EJ, Zietse R: Hyponatremia and mortality: moving beyond associations. Am J Kidney Dis 2013; 62: 139–49 CrossRef MEDLINE|
|26.||Holland-Bill L, Christiansen CF, Heide-Jørgensen U, et al.: Hyponatremia and mortality risk: a Danish cohort study of 279508 acutely hospitalized patients. Eur J Endocrinol 2015; 173: 71–81 CrossRef MEDLINE|
|27.||Diaconu CC, Balaceanu A, Bartos D: Diuretics, fist-line antihypertensive agents: are they always safe in the elderly? Rom J Intern Med 2014; 52: 87–90 MEDLINE|
|28.||Rodenburg EM, Hoorn EJ, Ruiter R, et al.: Thiazide-associated hyponatremia: a population-based study. Am J Kidney Dis 2013; 62: 67–72 CrossRef MEDLINE|
|29.||ESH/ESC Task Force for the Management of Arterial Hypertension: 2013 Practice guidelines for the management of arterial hypertension of the European Society of Hypertension (ESH) and the European Society of Cardiology (ESC): ESH/ESC Task Force for the Management of Arterial Hypertension. J Hypertens 2013; 31: 1925–38 CrossRef MEDLINE|
|30.||Pikälä KH, Laurila JV, Strandberg TE, Tilvis RS: Prognostic significance of delirium in frail older people. Dement Geriatr Cogn Disord 2015; 19: 158–63 CrossRef MEDLINE|
|31.||Burst V, Hensen J: Hyponatriämie – häufig und komplex. Dtsch Med Wochenschr 2015; 140: 608–11 CrossRef MEDLINE|