Dysnatremia in Relation to Frailty and Age in Communitydwelling Adults in the National Health and Nutrition Examination Survey

Transcription

1 Journals of Gerontology: Medical Sciences cite as: J Gerontol A Biol Sci Med Sci, 2017, Vol. 72, No. 3, doi: /gerona/glw114 Advance Access publication June 29, 2016 Research Article Dysnatremia in Relation to Frailty and Age in Communitydwelling Adults in the National Health and Nutrition Examination Survey Amanda J. Miller, 1 Olga Theou, 2 Miranda McMillan, 2 Susan E. Howlett, 2,3 Karthik K. Tennankore, 1 and Kenneth Rockwood 2 1 Department of Nephrology, 2 Division of Geriatric Medicine, and 3 Department of Pharmacology, Dalhousie University, Halifax, Nova Scotia, Canada. Address correspondence to Kenneth Rockwood, MD, Division of Geriatric Medicine, Dalhousie University, 5955 Veterans Memorial Lane, Veterans Memorial Building, Room 1421, Halifax, NS B3H 2E1, Canada. kenneth.rockwood@dal.ca Received February 19, 2016; Accepted June 3, 2016 Decision Editor: Stephen Kritchevsky, PhD Abstract Background: Frailty represents an age-related state of increased risk of adverse health outcomes, reflecting some combination of increased damage and compromised repair processes. Our objectives were to establish whether frailty is associated with dysnatremia (a deviation of serum sodium from normal values), to determine whether frailty explains the previously established association between age and dysnatremia and to assay the impact of each on mortality. Methods: The relationship between age, frailty, and dysnatremia was investigated across the adult life course in 8,911 respondents from the and cross-sectional National Health and Nutrition Examination Survey, on whom both laboratory and mortality data were available. A frailty index (FI) was calculated for each respondent and related to dysnatremia (serum sodium values outside a mmol/l reference range). Results: Dysnatremia was significantly related to both age and frailty; as the degree of frailty increased, so did the proportion with dysnatremia, for example, from 4.1% in those with FI less than 0.10, to 12.4% in those with FI 0.40 or more; p less than.001. Adjusted for frailty, the relationship between age and dysnatremia was no longer significant. In the age- and sex-adjusted Cox models, both frailty (hazard ratio [HR]: 1.05; 95% confidence interval [CI]: for every 0.01 increase in FI) and dysnatremia (HR: 1.85; 95% CI: ) were significant predictors of mortality; when hyponatremia was separated from hypernatremia in the Cox models, hypernatremia wasn t significant, but only 41 participants were identified as hypernatremic. Conclusion: Increasing frailty is associated with dysnatremia and confounds the association between age and dysnatremia. Keywords: Frail elderly adults Frailty index Sodium NHANES The prevalence of dysnatremia (a deviation of serum sodium from normal values) increases with age (1 4). Both hypo- and hypernatremia are associated with adverse outcomes in older adults. Even mild aberrations in serum sodium are associated with significantly worse results both on standardized cognitive testing (5,6) and in mortality risk compared with normonatremic controls in age-adjusted models (1,7). Many factors can lead to sodium abnormalities including comorbidities, multiple medications, and age-related alterations in renal and hormonal processes (2 4,8). Age-related illnesses are typically seen as part of a broader accumulation of health deficits in a range of organ systems which have been used to characterize frailty. Frailty represents a multiply determined state of increased risk of adverse outcomes, compared with others of the same age (9). Frailty is viewed both as a phenotype (10,11) and as a risk state (12). The latter, operationalized as a frailty index (FI) based on the ratio of actual to potential deficits, allows frailty to be graded (13 16) and has been used as an objective measure of frailty in both the general population and in clinical studies (17 19). Increasing frailty represents an accumulation of deficits The Author Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please journals.permissions@oup.com. 376

2 Journals of Gerontology: MEDICAL SCIENCES, 2017, Vol. 72, No with loss of the adaptive capacity required to maintain normal function in the face of physiologic stressors (20,21). Across the life course, frailty increases with age (22,23) and is independently associated with worse health outcomes, higher institutionalization rates, and accelerated mortality (11,21,24). If frailty implies a breakdown of macroscopic and subcellular adaptive processes (25), it is possible that progressive frailty might be associated with impairments in sodium homeostasis, an integrated system which ordinarily is complex and tightly regulated. The National Health and Nutrition Examination Survey (NHANES) has demonstrated that the prevalence of dysnatremia increases with age (12), as does the prevalence of frailty (13,16,26,27). How frailty relates to dysnatremia has not been explored. Here, we evaluated the association between frailty and dysnatremia in community-dwelling adults in NHANES and examined how these each related to mortality risk. We hypothesized that the degree of frailty was associated with the occurrence of dysnatremia and that frailty would confound the previously established association between age and dysnatremia. Methods We analyzed data from respondents to the and cross-sectional NHANES data sets who also had laboratory data collected. NHANES evaluated community-dwelling individuals in the United States. All ages were surveyed, with oversampling of people aged 60 and older to produce a representative and reliable national sample. The study was approved by the National Center for Health Statistics Research Ethics Review Board. All respondents provided written consent. A summary of the data acquisition and survey recruitment process is available through the Centre for Disease Control and Prevention (28). Frailty Index We defined frailty using the deficit accumulation approach in which, following standard procedures (14), we constructed a 46-item FI by combining comorbidities, signs, symptoms, and laboratory abnormalities. Details on this FI have been described elsewhere (13,29). Each participant s FI score was calculated by dividing the total number of deficits by the 46 potential problems evaluated, resulting in a score between 0 and 1, where FI = 0 is full health and higher scores represent worse frailty. For example, here a person with 23 deficits in the 46 items has an FI = 23/46 = 0.50 (14). For each respondent, an FI was only calculated for those who had data for 80% or more of FI items. Dysnatremia Dysnatremia was defined as respondents with serum sodium outside the reference interval (ie, either hypo- or hypernatremic). Over its course, the NHANES laboratory parameters for defining a normal sodium level have varied, most recently using a reference interval of mmol/l (7). Following that, here we described serum sodium less than 136 mmol/l as hyponatremic and serum sodium more than 144 mmol/l as hypernatremic. Even so, hyponatremia has also been defined as a serum sodium of less than 135 mmol/l (5) and hypernatremia as a serum sodium of greater than 145 mmol/l (30). To evaluate the impact of the reference range, we conducted sensitivity analyses, using reference ranges , , and mmol/l. Mortality Mortality information was retrieved from a linked mortality file provided by the National Center for Health Statistics. The follow-up data monitored the survey participants from the date of survey participation ( ) through to December 31, Using National Death Index submission records, only 10 people (0.1% of the sample) were lost to follow-up prior to ascertainment of mortality status. Statistical Analysis Baseline characteristics were expressed as either mean ± standard deviation or as a percent of the sample. The proportion of people with hyponatremia and hypernatremia and the proportion with normal sodium levels were plotted within each frailty group as stack-bar diagrams. For stratified graphs, FI groups were as follows: FI < 0.1; 0.1 FI < 0.2; 0.2 FI < 0.3; 0.3 FI < 0.4; and FI 0.4. The effect of frailty and age on dysnatremia was assessed using logistic regression analysis, adjusted for sex. We also plotted the proportion of people with dysnatremia by age and frailty groups and assessed this relationship using chi-square trend testing. Predictive validity of dysnatremia and frailty on mortality was tested using Cox regression analysis, controlling for sex and age. The relationship of dysnatremia to mortality was assessed using Kaplan Meier survival curves and the survival distributions were compared using the Log-rank test. All statistical analyses used codes developed in RStudio (version , ) and SPSS (IBM SPSS Statistics, version ). Graphs were created in RStudio. Statistical significance was set at p less than.05. Results Of 10,020 individuals aged 20+ years, 8,917 individuals also had laboratory sodium data, of whom 8,911 were included in these analyses; six respondents were excluded due to missing FI data. The mean sodium for our sample was 139 ± 2 mmol/l (median 139, minimum 99, and maximum 153) and the mean FI was 0.14 ± 0.12 (median 0.10, minimum 0, and maximum 0.77). Overall, the prevalence of dysnatremia was 6% (95% confidence interval [CI]: ); 5.6% (95% CI: , n = 496) hyponatremic and 0.5% (95% CI: , n = 41) hypernatremic. In those with hyponatremia, mean and median sodium values were and mmol/l, respectively. In those with hypernatremia, mean and median sodium values were and mmol/l, respectively. Association Between Age, Frailty, and Dysnatremia Dysnatremia increased with advancing age from 5.4% (95% CI: ) in those younger than 60 years to 9.3% (95% CI: ) in those 75 years or older; p less than.001. An increase in the proportion of patients with dysnatremia was also observed as the degree of frailty increased, for example, from 4.1% (95% CI: ) in those with FI less than 0.10, to 12.4% (95% CI: ) in those with FI 0.40 or more; p less than.001. This reflects a higher prevalence of both hypo- and hypernatremia with advancing frailty (Figure 1). Logistic regression models (adjusted for sex) showed that both FI and age were significantly associated with dysnatremia; however, when FI and age were included in the same model, FI remained significant while age became insignificant (Table 1). Multinomial logistic regression analyses showed that age was not significantly associated with hyponatremia (adjusted for sex) and when FI and age were included in the same model, age was protective (Supplementary Appendix 1). The prevalence of dysnatremia increased with advancing

3 378 Journals of Gerontology: MEDICAL SCIENCES, 2017, Vol. 72, No. 3 frailty for all age groups (Figure 2 and Supplementary Appendix 2). Furthermore, when stratifying by frailty level, there was no association between age and dysnatremia within each stratum (Figure 2; p values for chi-square trend testing ranged from ) Association of Frailty and Dysnatremia With Mortality In age- and sex-adjusted Cox regression models, the FI and dysnatremia were significantly associated with mortality (Table 2). Figure 1. Proportion of people with normonatremia, hyponatremia, and hypernatremia by frailty index level. Figure 2. Proportion of people with dysnatremia by age, stratified by frailty index level. When included in the same model, the hazard ratio (HR) for the FI was similar (HR: 1.05; 95% CI: for every 0.01 increase in FI) to the model when only FI was included, whereas the HR for dysnatremia was attenuated in the combined model (HR: 1.56; 95% CI: ) versus in the independent model (HR: 1.85; 95% CI: ) (Table 2). The interaction of FI with dysnatremia was significant, showing that the FI could be an effect modifier for the relationship of dysnatremia with mortality (Table 2; Model 4). When hyponatremia was separated from hypernatremia in the Cox regression models, findings were similar for hyponatremia but not for hypernatremia; however, only 41 participants were identified as hypernatremic (Supplementary Appendix 3). People with dysnatremia had worse survival than did those with normal sodium levels. After we stratified the analysis by frailty group, the difference was statistically significant in all frailty groups except those with an FI score greater than 0.4 (p =.936). Mortality among the normonatremic, hyponatremic, and hypernatremic cohorts was significantly different (p <.001), with hypernatremia conferring the lowest survival probability (Figure 3). Table 1. Relationship Between Age, Frailty, and Dysnatremia (logistic regression analyses) Outcome: Dysnatremia OR 95% CI p Value Wald Model 1 FI (per 0.01) < Sex (females) < Model 2 Age (per year) < Sex (females) < Model 3 FI (per 0.01) < Age (per year) Sex (females) < Model 4 FI (per 0.01) < Age (per year) Sex (females) < FI*Age Note: CI = confidence interval; FI = frailty index; OR = odds ratio. Table 2. Relationship Between Frailty, Dysnatremia, and Mortality (Cox regression analyses) Outcome: Mortality HR 95% CI p Value Wald Model 1 FI (per 0.01) < Age (per year) < Sex (females) < Model 2 Dysnatremia < Age (per year) <.001 1,055.1 Sex (females) < Model 3 FI (per 0.01) < Dysnatremia < Age (per year) < Sex (females) < Model 4 FI (per 0.01) < Dysnatremia < Age (per year) < Sex (females) < FI*Dysnatremia Figure 3. The association between normonatremia, hyponatremia, hypernatremia, and mortality. All survival probabilities are statistically significant from each other. Notes: CI = confidence interval; FI = frailty index; HR = hazard ratio. Dysnatremia was defined as respondents with serum sodium outside the reference interval (ie, either hypo- or hypernatremic). When hyponatremia was separated from hypernatremia in the Cox regression models, findings were similar for hyponatremia but not for hypernatremia; however, only 41 participants were identified as hypernatremic (Supplementary Appendix 3).

4 Journals of Gerontology: MEDICAL SCIENCES, 2017, Vol. 72, No Impact of Sodium Reference Range There were 254 people with a sodium value of 135 mmol/l. Using a sodium cut point of less than 135 mmol/l in the sensitivity analysis reduced the number of people with hyponatremia to 242. The main findings remained similar with this reference range except that in the logistic regression analysis, adjusted for frailty and sex, age was now significantly (p =.044) associated with dysnatremia (odds ratio: 1.008, 95% CI: ). Survival was not different between hyponatremic and hypernatremic people. A second sensitivity analysis was conducted to evaluate the impact of using a sodium cut point of more than 145 mmol/l to define hypernatremia instead of the more than 144 mmol/l cut point used in the primary analysis. Of the 41 hypernatremic people in the original analysis, 35 had a value of 145 mmol/l; thus with a sodium cut point of more than 145 mmol/l, the number of people with hypernatremia was reduced to 6. The main findings were unchanged; however, in this small population with hypernatremia, survival was no longer significantly worse compared with hyponatremic people. The findings when using the mmol/l reference range were similar to those from this second sensitivity analysis. Discussion This study demonstrates an association between frailty and dysnatremia and confirms a relationship between age and sodium abnormalities. Furthermore, it shows that frailty confounds the previously established association between age and dysnatremia. Additionally, in our study both FI and dysnatremia were associated with mortality. The prevalence of hypo- and hypernatremia demonstrated in our study (5.6 and 0.5%, respectively) is comparable with that reported in earlier studies of community-dwelling individuals (hyponatremia % and hypernatremia %) (2,7). Although the mean values for hypo- and hypernatremia in our study only deviated minimally from the normal range (133.7 and mmol/l, respectively), it is known that even very small aberrations in serum sodium are associated with adverse health outcomes. In hyponatremia, for example, serum sodium between 130 and 135 mmol/l has been associated with impaired cognition, falls, hip fracture, and death in otherwise well ambulatory populations (31 33). Earlier literature has also demonstrated that even mild hyponatremia is associated with increased mortality in both outpatient (7) and hospital settings (34). Our data must be interpreted with caution. The NHANES data set of community-dwelling adults has nearly 9,000 respondents with complete data. Although we have follow-up data, the data on frailty and dysnatremia are only available at baseline, making it difficult to establish causality between the two. In addition, although the NHANES data have been used in a number of earlier studies (7,13,16) and include a highly regimented data collection protocol, we did not have access to medication records. Some medication classes affect the evolution of dysnatremia (35), and we were unable to correct for this variable in our analysis. The standard reference range for normonatremia varies in the literature, a fact which could potentially affect the generalizability of our results. For our sensitivity analyses, we showed that variations in the sodium reference range ( , , or mmol/l) did not significantly alter our results. Increasing the upper limit to 145 mmol/l reduced the number of cases with hypernatremia from 41 to 6, but did not change the trends observed, although with such small numbers, survival was no longer significantly worse compared with hyponatremic people. Furthermore, the FI used here was established using self-reported comorbidities, which carries a risk of recall bias and inaccuracies. Even so, the use of comorbidity data in NHANES has been well established in previous studies (7,13,16). Lastly, as NHANES is a study of community-dwelling adults, institutionalized and hospitalized people (who might be expected to reflect advanced, or end-stage frailty) were excluded. Therefore, the full spectrum of frailty is not represented. Frailty represents a multifactorial state of increased risk of adverse outcomes compared with others of the same age (9). Although multimorbidity and physiologic impairments are more common with advancing age, the health status among individuals of the same chronological age varies considerably (15). The prior perception that increasing age impacts sodium balance likely represents the increased prevalence of frailty in older populations, with as many as 30% of people older than 75 years considered frail (2,20). The maintenance of sodium and water balance is complex and includes an interplay between hormonal mediators (eg, antidiuretic hormone and aldosterone), renal function, total body water composition, thirst, and access to fluids (36). This process has sufficient redundancy that the body typically compensates for mild derangements. In frail patients, in addition to an accumulation of clinically visible deficits, subclinical deficits also are present (37) impairing compensatory processes that allow adaptation (31). Indeed, dysregulation of the salt and water homeostasis may be a special case of a more widespread loss of adaptive capacity that goes with frailty. Here hyponatremia was more common than hypernatremia. Low serum sodium may be a marker for early dysregulation of salt and water homeostasis, whereas abnormally high serum sodium may potentially represent a progressive impairment in compensatory mechanisms with a complete loss of capacity to adapt to physiologic stressors. Hypernatremia occurs when there is a breakdown in sodium regulatory pathways resulting in increased sodium relative to body water that is not rectified by a compensatory intake of dilute fluids. In patients with very advanced frailty (not included in our analysis by virtue of the study population), impaired mobility or cognition may impede their access to free water. Therefore, it is not surprising that our sample displayed an increased prevalence of hypernatremia among the frailest individuals (38). In further support of our hypothesis, studies have shown a greater prevalence of dysnatremia in nursing home populations compared with community-dwelling individuals, with a dramatic increase in the prevalence of both hypo- (18%) and hypernatremia (22%) observed in most studies (8,39) compared with community settings (4.3% hyponatremia and 0.72% hypernatremia) (2). The higher prevalence of hypernatremia in these exceedingly frail populations strongly supports the view that hypernatremia represents the end result in the progressive decompensation of sodium homeostasis in advanced frailty. This is further supported by the fact that our study demonstrates hypernatremia to be associated with a higher mortality rate than those individuals with hyponatremia, with normal sodium homeostasis conferring the best survival rate. Conclusion This study demonstrates that a higher level of frailty is associated with dysnatremia, irrespective of age, and that the previously established relationship between age and dysnatremia is lost when stratifying by frailty status. In addition, our results suggest that the presence of dysnatremia and advancing frailty are both associated with a higher mortality, again independent of age. Collectively, these observations support the hypothesis that increasing frailty is linked

5 380 Journals of Gerontology: MEDICAL SCIENCES, 2017, Vol. 72, No. 3 to a progressive breakdown in salt and water homeostasis, which results in gross and measurable abnormalities in sodium status. Further studies are required to determine if this relationship persists at higher degrees of frailty (such as with institutionalized populations) and to tease apart the association between advanced frailty and hyper- versus hyponatremia. Whether evaluation and treatment of dysnatremia can add to other interventions now being studied that can reduce the level of frailty (40) is an important consideration that will motivate future inquiries. Supplementary Material Supplementary material can be found at: oxfordjournals.org/ Funding S.E.H. s work is supported by the Canadian Institutes of Health Research (MOP ). K.R. s work is supported by the Canadian Institutes of Health Research (MOP ) and the Fountain Innovation Fund of the Queen Elizabeth II Health Sciences Foundation. Conflict of Interest The authors report no declarations of interest. References 1. Bourdel-Marchasson I, Laksir H, Puget E. Interpreting routine biochemistry in those aged over 65 years: a time for change. Maturitas. 2010;66: doi: /j.maturitas Hawkins RC. Age and gender as risk factors for hyponatremia and hypernatremia. Clin Chim Acta. 2003;337: Madias NE, Adrogue HJ. Hypo hypernatraemia: disorders of water balance. In: Davidson AM, Cameron JS, Grünfeld J-P, et al., eds. Oxford Textbook of Clinical Nephrology. 3rd ed. Oxford, UK: Oxford University Press; 2005: Chassagne P, Druesne L, Capet C, Ménard JF, Bercoff E. Clinical presentation of hypernatremia in elderly patients: a case control study. J Am Geriatr Soc. 2006;54: 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: doi: / Bruce JM, Harrington CJ, Foster S, Westervelt HJ. Common blood laboratory values are associated with cognition among older inpatients referred for neuropsychological testing. Clin Neuropsychol. 2009;23: doi: / Mohan S, Gu S, Parikh A, Radhakrishnan J. Prevalence of hyponatremia and association with mortality: results from NHANES. Am J Med. 2013;126: e1. doi: /j.amjmed Rosner MH. Hyponatreamia in the elderly: etiologies, implications and therapy. Aging Health. 2011;7(5): Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet. 2013;381: doi: /s (12) Theou O, Cann L, Blodgett J, Wallace LM, Brothers TD, Rockwood K. Modifications to the frailty phenotype criteria: systematic review of the current literature and investigation of 262 frailty phenotypes in the Survey of Health, Ageing, and Retirement in Europe. Ageing Res Rev. 2015;21: doi: /j.arr Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56(3):M146 M de Vries NM, Staal JB, van Ravensberg CD, Hobbelen JS, Olde Rikkert MG, Nijhuis-van der Sanden MW. Outcome instruments to measure frailty: a systematic review. Ageing Res Rev. 2011;10: doi: /j.arr Blodgett J, Theou O, Kirkland S, Andreou P, Rockwood K. Frailty in NHANES: comparing the frailty index and phenotype. Arch Gerontol Geriatr. 2015;60: doi: /j.archger Searle SD, Mitnitski A, Gahbauer EA, Gill TM, Rockwood K. A standard procedure for creating a frailty index. BMC Geriatr. 2008;8:24. doi: / Mitnitski AB, Graham JE, Mogilner AJ, Rockwood K. Frailty, fitness and late-life mortality in relation to chronological and biological age. BMC Geriatr. 2002;2: Goggins WB, Woo J, Sham A, Ho SC. Frailty index as a measure of biological age in a Chinese population. J Gerontol A Biol Sci Med Sci. 2005;60: Armstrong JJ, Mitnitski A, Launer LJ, White LR, Rockwood K. Frailty in the Honolulu-Asia Aging Study: deficit accumulation in a male cohort followed to 90% mortality. J Gerontol A Biol Sci Med Sci. 2015;70: doi: /gerona/glu Kelaiditi E, Andrieu S, Cantet C, Vellas B, Cesari M; ICTUS/DSA Group. Frailty index and incident mortality, hospitalization, and institutionalization in Alzheimer s disease: data from the ICTUS Study. J Gerontol A Biol Sci Med Sci. 2016;71: doi: /gerona/glv Zeng A, Song X, Dong J, et al. Mortality in relation to frailty in patients admitted to a specialized geriatric intensive care unit. J Gerontol A Biol Sci Med Sci. 2015;70(12): doi: /gerona/glv Strandberg TE, Pitkala KH, Tilvis RS. Frailty in older people. Eur Geriatr Med. 2011;2(6): Rockwood K, Mitnitski A. Frailty defined by deficit accumulation and geriatric medicine defined by frailty. Clin Geriatr Med. 2011;27: doi: /j.cger Rockwood K, Song X, Mitnitski A. Changes in relative fitness and frailty across the adult lifespan: evidence from the Canadian National Population Health Survey. CMAJ. 2011;183:E487 E494. doi: / cmaj Kennedy CC, Ioannidis G, Rockwood K, et al. A Frailty Index predicts 10-year fracture risk in adults age 25 years and older: results from the Canadian Multicentre Osteoporosis Study (CaMos). Osteoporos Int. 2014;25: doi: /s Topinková E. Aging, disability and frailty. Ann Nutr Metab. 2008;52(suppl 1):6 11. doi: / Mitnitski A, Song X, Rockwood K. Assessing biological aging: the origin of deficit accumulation. Biogerontology. 2013;14: doi: / s Smit E, Winters-Stone KM, Loprinzi PD, Tang AM, Crespo CJ. Lower nutritional status and higher food insufficiency in frail older US adults. Br J Nutr. 2013;110: doi: /s x 27. Wilhelm-Leen ER, Hall YN, Tamura MK, Chertow GM. Frailty and chronic kidney disease: the Third National Health and Nutrition Evaluation Survey. Am J Med. 2009;122: e2. doi: /j. amjmed Centers for Disease Control and Prevention, National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Data , Public Data General Release File Documentation Blodgett J, Theou O, Kirkland S, Andreou P, Rockwood K. The association between sedentary behaviour, moderate-vigorous physical activity and frailty in NHANES cohorts. Maturitas. 2015;80: doi: /j. maturitas Turgutalp K, Özhan O, Gök Oğuz E, et al. Community-acquired hypernatremia in elderly and very elderly patients admitted to the hospital: clinical characteristics and outcomes. Med Sci Monit. 2012;18:CR729 CR 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 71.e Gankam-Kengne F, Ayers C, Khera A, de Lemos J, Maalouf NM. Mild hyponatremia is associated with an increased risk of death in an ambulatory setting. Kidney Int. 2013;83: doi: /ki

6 Journals of Gerontology: MEDICAL SCIENCES, 2017, Vol. 72, No Hoorn EJ, Rivadeneira F, van Meurs JB, et al. Mild hyponatremia as a risk factor for fractures: the Rotterdam study. J Bone Miner Res. 2011;26(8): doi: /jbmr Conway R, Byrne D, O Riordan D, Silke B. Hyponatraemia in emergency medical admissions outcomes and costs. J Clin Med. 2014;3(4): doi: /jcm Liamis G, Milionis H, Elisaf M. A review of drug-induced hyponatremia. Am J Kidney Dis. 2008;52: doi: /j.ajkd Harring TR, Deal NS, Kuo DC. Disorders of sodium and water balance. Emerg Med Clin North Am. 2014;32: doi: /j. emc Howlett SE, Rockwood K. New horizons in frailty: ageing and the deficitscaling problem. Age Ageing. 2013;42: doi: /ageing/aft El-Sharkawy AM, Watson P, Neal KR, et al. Hydration and outcome in older patients admitted to hospital (The HOOP prospective cohort study). Age Ageing. 2015;44: doi: /ageing/afv Frangeskou M, Lopez-Valcarcel B, Serra-Majem L. Dehydration in the elderly: a review focused on economic burden. J Nutr Health Aging. 2015;19: doi: /s Ng TP, Feng L, Nyunt MS, et al. Nutritional, physical, cognitive, and combination interventions and frailty reversal among older adults: a randomized controlled trial. Am J Med. 2015;128: e1. doi: /j.amjmed