Clinical Section
Gerontology 2007;53:179–183
DOI: 10.1159/000099144
Dehydration among Long-Term Care
Elderly Patients with Oropharyngeal
Dysphagia
A. Leibovitz Y. Baumoehl E. Lubart A. Yaina N. Platinovitz R. Segal
Shmuel Harofe Hospital, Geriatric Medical Center, affiliated with the Sackler Faculty of Medicine,
Tel Aviv University, Tel Aviv, Israel
Key Words
Dehydration
Oropharyngeal dysphagia
Long-term care,
elderly
Abstract
Introduction: Long-term care (LTC) residents, especially the
orally fed with dysphagia, are prone to dehydration. The clin-
ical consequences of dehydration are critical. The validity of
the common laboratory parameters of hydration status is far
from being absolute, especially so in the elderly. However,
combinations of these indices are more reliable. Objective:
Assessment of hydration status among elderly LTC residents
with oropharyngeal dysphagia. Methods: A total of 28 oral-
ly fed patients with grade-2 feeding difficulties on the func-
tional outcome swallowing scale (FOSS) and 67 naso-gastric
tube (NGT)-fed LTC residents entered the study. The com-
mon laboratory, serum and urinary tests were used as indi-
ces of hydration status. The results were considered as in-
dicative of dehydration and used as ‘markers of dehydration’,
if they were above the accepted normal values. Results: The
mean number of dehydration markers was significantly
higher in the FOSS-2 group (3.8 8 1.3 vs. 2 8 1.4, p = 0.000).
About 75% of these FOSS-2 patients had 64 dehydration
markers versus 18% of the NGT-fed group (p = 0.000). A low
urine output (!800 ml/day) was significantly more common
in the FOSS-2 group (39 vs. 12%, p = 0.002). Above normal
© 2007 S. Karger AG, Basel
0304–324X/07/0534–0179$23.50/0
Fax +41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/ger
Received: September 13, 2006
Accepted: December 4, 2006
Published online: January 30, 2007
values of blood urea nitrogen (BUN), BUN/serum creatinine
ratio (BUN/SCr), urine/serum osmolality ratio (U/SOsm), and
urine osmolality UOsm, were significantly more frequent in
the dehydration-prone FOSS-2 group. This combination of 4
indices was present in 65% of low urine output patients. In
contrast, it was present in only 36% of the higher urine out-
put patients (p = 0.01). Patients with a ‘normal’ daily urine
output (1800 ml/day) also had a significant number (2 8 1.5)
of positive indices of dehydration. Conclusions: Dehydra-
tion was found to be common among orally fed FOSS-2 LTC
patients. Surprisingly, probable dehydration, although of a
mild degree, was not a rarity among NGT-fed patients either.
The combination of 4 parameters, BUN, BUN/SCr , U/SOsm and
UOsm, offers reasonable reliability to be used as an indication
of dehydration status in daily clinical practice.
Copyright © 2007 S. Karger AG, Basel
Introduction
Dehydration, a term used to reflect several physiolog-
ical states based on the imbalance between intake and
loss of fluid and the accompanying sodium status, is very
common in the elderly. In 1991, almost 200,000 patients
older than 65 years were discharged from US hospitals
with a primary diagnosis of dehydration [1]. It is claimed
that the real incidence is probably higher because of
Dr. Arthur Leibovitz
Shmuel Harofe Hospital
Geriatric Medical Center
POB 2, Beer-Yaacov (Israel)
Tel. +972 8925 8640, Fax +972 8923 7156, E-Mail larthur@bezeqint.net
masking by other clinical conditions [2]. In a 1995 survey,
37% of emergency room elderly had hypovolemia [3].
Long-term care (LTC) residents are especially prone to
dehydration [4]. Probably, many of these patients are
chronically (and mildly) dehydrated, even without acute
medical illness [2, 4]. Considering the number of LTC
patients, in the US alone more than 2 million [5], the
enormity of the problem is highlighted.
The cardinal factors that place the elderly at particular
risk of dehydration are: changes in functional status,
mainly mobility disabilities; mental and communication
disorders; feeding difficulties (oropharyngeal dyspha-
gia), and of course medication effects [2, 6].
In addition, various physiological changes occur in the
elderly. On the whole, the ability to concentrate urine and
thus conserve water declines [7]. In particular, renin ac-
tivity and aldosterone secretion are decreased [8]; there
is relative resistance to vasopressin [9], and diminished
vasopressin response to volume stimuli [10]. Concomi-
tantly, the sensation of thirst is altered (defects in both the
intensity and threshold) leading to reduced fluid intake
[6, 11]. Early diagnosis of this treatable condition is prob-
lematic and constitutes an ongoing challenge [2, 12].
However, left untreated, mortality may exceed 50% [13,
14]. Manifestations of dehydration in the elderly differ
from those in young individuals.
Classical signs may be vague, deceptive or even absent.
On the other hand, in the elderly, nonspecific manifesta-
tions of dehydration, e.g. deterioration of cognitive sta-
tus, constipation, falls and even fever, may predominate
[2, 15, 16]. Clearly, objective tests are to be used for detec-
tion of dehydration. However, the ones that may consti-
tute reliable standards of hydration status, e.g. dual pho-
ton absorptiometry, serve mainly as research tools [17].
In daily clinical practice simple common laboratory pa-
rameters are in use. Individually the validity of each of
these indices is far from being absolute. Especially so in
the elderly (mainly due to impaired renal function). How-
ever, combinations of these indices, preferably blood
borne and urinary ones, are judged sufficiently reliable to
permit detection of, albeit not in absolute terms, dehydra-
tion [2, 18–20]. In order to assess and define aspects of
the dehydration problem, we initiated a survey among 2
subgroups of LTC patients, namely those with clear-cut
oropharyngeal dysphagia, thus at high risk of becoming
dehydrated, and supposedly well-hydrated patients on
nasogastric tube (NGT) feeding. We would like to add
that NGT feeding in Israel is still more widespread than
the use of percutaneous endoscopic gastrostomy [21].
180 Gerontology 2007;53:179–183
Materials and Methods
The study was performed in the LTC wards of a 400-bed,
teaching, university-affiliated, geriatric hospital. The research
protocol was approved by the local committee on human experi-
mentation and written informed consent was obtained from all
participants or their guardians. Data were collected during a 3-
month period. All patients with grade-2 feeding difficulties on the
functional outcome swallowing scale (FOSS) [22] and the NGT-
fed residents of these LTC wards were eligible for the study. FOSS
was used since on the same scale it grades patients with no swal-
lowing difficulties, those with various stages of dysphagia, and
also those enteral fed by tube. Grade-2 FOSS was chosen since it
represents the first significant dysphagic stage on this scale. Pa-
tients with swallowing difficulties were identified with the help of
the nursing staff, a clinical dietician and a speech therapist. Oth-
er inclusion criteria were: stable medical condition such as infec-
tious episodes or need of intravenous infusions for at least 3 weeks
prior to inclusion, and the presence of an indwelling urinary cath-
eter (in order to assess urinary output). Patients on diuretic and
steroid therapy, those with possible renal failure (creatinine 11.2),
those with glucose or lipid elevations 625% above the normal
range, electrolyte disturbances as well as the terminally ill were
excluded. Venous blood samples were drawn twice, at 1-week in-
tervals. The tests consisted of a complete blood count (CBC) and
a biochemical profile which included blood urea nitrogen (BUN),
serum creatinine (SCr), sodium (SNa), osmolality (SOsm), as well as
potassium, glucose, protein, cholesterol and triglycerides.
Urine was collected twice, concomitant with the blood tests,
for quantitative measurements and for estimation of sodium
(U Na) and creatinine (UCr) concentrations as well as for osmolal-
ity (UOsm). Given the limited value of the individual laboratory
parameters of hydration, especially so in the elderly, we used as
indices of hydration status most of the relevant common labora-
tory tests, i.e. BUN, SNa, SOsm, U Na, UOsm, BUN/SCr, UOsm/SOsm
and UCr/SCr ratios, hematocrit and hemoglobin. Based on stan-
dard guidelines for interpreting diagnostic tests [23], the results
were considered as indicative of dehydration and used as ‘markers
of dehydration’, if they matched the following levels: BUN above
normal values (120 mg/dl); BUN/SCr ratio 120; SNa 1145 mEq/l;
SOsm 1295 mosm/kg; U Na !10 mEq/l; UOsm 1500 mosm/kg; U/SCr
ratio 1 40, and U/SOsm ratio 11.2. Daily urine output of !800 ml
was considered as ‘abnormal’ [24]. Chemistry analyses were per-
formed through Olympus Automated Chemistry Analyzer –
AU400, type 401-03 (Mishima Olympus Co. Ltd., Shizuoku-ken,
Japan), using Olympus System Reagents (Olympus Diagnostica
GmbH, Ireland). Osmolality was measured by a vapor pressure
osmometer (VAPRO-Model 5520, Wescor Inc., Logan, Utah,
USA). Blood count tests were assessed by an automated hematol-
ogy analyzer (Pentra 60, ABX-Diagnostics, Montpellier, France).
The SPSS software was used for statistical processing. De-
scriptive analysis included frequencies and distributions of all
study variables. Comparison of averages and of percentages be-
tween groups were performed by means of the independent sam-
ples t test and, respectively, the
2 or Fisher’s exact tests. Cron-
bach’s  was used to test reliabilities.
To evaluate relevant clinical and laboratory factors that may
predict low (!800 ml/24 h) urine output, univariate and forward
stepwise multivariate analysis were performed for the entire study
group.
Leibovitz/Baumoehl/Lubart/Yaina/
Platinovitz/Segal
Table 1. Demographics and basal laboratory data of elderly LTC tion-prone FOSS-2 group (table 2). Cronbach’s  for this
patients – orally and NGT fed group of parameters was significant: 0.770. This combi-
nation of 4 indices was present in 65% of the low urine
NGT OF p
(n = 67) (n = 28) output patients. In contrast, it was present in only 36% of
the higher urine output participants (p = 0.01).
Age, years 79.889.9 80.087.8 NS On univariate analysis a significant correlation was
Males 16 (24%) 8 (28%) NS present between all markers of dehydration (except SOsm),
IHD, % 15 14 NS and low urine output. On multivariate analysis, only high
CHF, % 1.5 3.4 NS
Hypertension, % 30 10 0.04 urine osmolality and low UNa were significantly corre-
CVA, % 31 17 NS lated with low urine output (p = 0.001 and 0.043, respec-
Dementia, % 31 27 NS tively).
Diabetes, % 15 24 NS In 13% of the patients (all with creatinine levels well
Decubitus ulcer 16 4 NS within the ‘normal’ range), renal function was signifi-
Hb, g/dl 11.281.69 11.681.22 NS
Hct, % 3385.04 3484.09 NS cantly impaired (CCT !30 ml/min). In order to verify
Albumin, g/dl 3.180.43 3.280.34 NS whether these patients distort our results concerning de-
Serum creatinine, mg/dl 0.8280.18 0.9380.12 0.001 hydration measures, we redid the calculations after omit-
Serum creatinine, mol/l 72.5816 82.2810.6 ting this subgroup. Overall, the results did not differ sig-
Fractional excretion of Na+ 0.7780.87 0.6380.4 NS nificantly from those of the entire study group.
CCT, ml/min 53824 51825 NS

LTC = Long-term care; NGT = nasogastric tube; OF = oral feed-

ing; IHD = ischemic heart disease; CHF = chronic heart failure;
CVA = cerebrovascular accident; CCT = creatinine clearance test.
p values by t test or
2.
Results
The study groups consisted of 28 orally and 67 NGT-
fed LTC patients. Demographic and basal laboratory data
are presented in table 1. As shown, inter-group differenc-
es were few and minor.
Data concerning the indices of dehydration in the
study groups, mean as well as prevalence of above (in the
case of UNa under) normal values, are presented in ta-
ble 2.
As can be seen, except for UNa, the mean values of all
parameters were significantly worse in the FOSS-2 group.
The mean number of dehydration markers was also sig-
nificantly higher in this group (3.8 8 1.3 vs. 2 8 1.4, p !
0.001). About 75% of these FOSS-2 patients had 64 de-
hydration markers versus 18% of the NGT-fed group (p !
0.001). A low urine output (!800 ml/day) was also sig-
nificantly more common in the FOSS-2 group (39 vs.
12%, p = 0.002; table 2). The low urine output patients
had a mean of 3.6 8 1 dehydration markers versus 2 8
1.5 indices in residents with urine outputs of 1800 ml/day
(p ! 0.001).
Above normal values of BUN, BUN/SCr, U/SOsm and
UOsm were significantly more frequent in the dehydra-
Discussion
As expected, a probable state of dehydration was com-
mon in FOSS-2 patients (75%). However, the extent was
well above that anticipated. Almost 40% of these subjects
had a low daily urine output and about 75% had 64
markers of dehydration. Patients with oropharyngeal
dysphagia constitute up to 70% of the LTC population
[25]. Thus, if the results of this study were interpolated to
a wider system, the hydration status of LTC residents is
much in want. Previous studies [4, 26] report a 26–39%
prevalence of dehydration in elderly LTC residents. How-
ever, unlike our study, they do not mention the swallow-
ing condition of their patients.
Above normal values for BUN, BUN/SCr, U/SOsm and
UOsm (dehydration indicators) were significantly more
frequent in the FOSS-2 group as well as in the probably
dehydrated low urine output patients. This supports the
validity of this cluster of tests as dehydration indices.
Hence, this combination seems reliable enough to be used
in daily clinical practice as an early detection indicator
for dehydration.
Surprisingly, probable dehydration, although of a mild
degree, was not a rarity among NGT-fed patients either.
Eighteen percent had 64 markers of dehydration and 12%
a low urine output. This finding is surprising since the ac-
cepted view is that these patients are sufficiently hydrated.
Seemingly, NGT feeding does not necessarily ensure suf-
ficient hydration. Thus, NGT-fed residents also need to be
closely followed with regard to hydration status.

Dehydration in Elderly Patients with Gerontology 2007;53:179–183 181

Oropharyngeal Dysphagia
 Table 2. Dehydration indices in elderly LTC patients – orally and NGT fed

Test/abnormal value Mean p1 Abnormal, % p2

NGT (n = 67) OF (n = 28) NGT OF

BUN (>20 mg/dl or >6.7 mmol/l) 18.586.5 24.684 0.000 31 89 0.000

BUN/serum creatinine (>20) 22.685.7 2784.9 0.001 54 89 0.001
Urine/serum osmolality (>1.2) 1.180.5 1.580.4 0.000 37 82 0.000
Urine osmolality (>500 mosm/kg) 3088125 4178117 0.000 6 25 0.006
Urine/serum creatinine (>40) 54.4829.6 70833 0.026 61 79 NS
Serum osmolality (>295 mosm/kg) 279.8811.3 28587.2 0.026 7 10 NS
Urine Na+ (<10 mEq/l) 42830 54837 NS 3 7 NS
Serum Na+ (>145 mEq/l) 13484 13883 0.000 0 0
Urine volume/day, ml (<800 ml) 1,1388419 9618301 0.001 12 39 0.002

LTC = Long-term care; NGT = nasogastric feeding tube; OF = oral feeding.

p1 by t test; p2 by
2.

Another fact that emerges from our study is that while
a low urine output indicates probable dehydration, a ‘nor-
mal/high’ urine output (i.e. 1800 ml/day) does not neces-
sarily mean good hydration. These ‘normal’ urine output
patients had a mean of 2 8 1.5 markers of dehydration
and the above-mentioned combination of 4 parameters
was present in 36% of these subjects. The matter is of
some importance as in daily clinical practice all too often
one does not suspect dehydration in a resident with a
‘normal’ urine output.
Poor oral intake and the consequent free water deple-
tion lead to hypertonic dehydration. However, unexpect-
edly, none of our supposedly dehydrated patients had
hypernatremia and only a few had elevated serum osmo-
lality. Thus, the question arises as to whether these pa-
rameters are useful, i.e. sufficiently sensitive, indicators
of dehydration in the elderly.
On the other hand, the fractional excretion of sodium
was !1 in more than 80% of the participants (81% in the
NGT and 86% in the FOSS-2 group). Seemingly, it con-
stitutes an oversensitive parameter in the elderly. Hemo-
globin and hematocrit levels were not expected to reflect
hypo-hydration given the fact that most participants
were anemic.
The commonly used biochemical indices of dehydra-
tion were derived from population-based reference rang-
es (PBRRs), determined in a non-aged population, and
they show clinically relevant changes with increasing age
[26, 27]. Thus these indices are insufficiently reliable in
the geriatric population. In fact, it was shown that labora-
tory parameters of dehydration are highly variable (high
individuality) between geriatric subjects [12]. Construc-
tion of PBRRs for the aged population would much im-
prove the reliability of the common laboratory indices of
dehydration and permit a much more correct diagnosis
in the elderly patient presenting for the first time. On the
other hand, laboratory parameters of dehydration are
stable over time within geriatric subjects thus permitting
the construction of within-subject coefficients of varia-
tion, i.e. subject-specific reference ranges (SSRRs) [26].
Hence, in cases where repeated measurements are avail-
able, dehydration can be best diagnosed and fluid balance
monitored over time by applying the above-mentioned
SSRRs (instead of the similar analysis by intuition, per-
formed by many physicians, when judging serial labora-
tory results).
The frail institutionalized elderly are at increased risk
of dehydration. Most are mentally impaired, communi-
cation handicapped, totally dependent and frequently
have dysphagia [28]. Concomitantly, many facilities are
plagued by a shortage of staff, thus the consequent inabil-
ity to provide sufficient hydration [6, 29]. Early diagnosis
is rendered more difficult by the fact that intake and out-
put charts in LTC settings are frequently unreliable given
the large number of incontinent residents, and inaccura-
cies in recordings [2, 4]. Low level fluid intake constitutes
a form of neglect. In fact, dehydration has been proposed
as a quality indicator in LTC [30]. Strenuous efforts
should be made to avoid and prevent this hazardous
state.

182 Gerontology 2007;53:179–183 Leibovitz/Baumoehl/Lubart/Yaina/

Platinovitz/Segal
 References
1 Vital and Health Statistics from the Centers
for Disease Control and Prevention/Nation-
al Center for Health Statistics: Detailed Di-
agnoses and Procedures, National Hospital
Discharge Survey 1991. Washington, US De-
partment of Health and Human Services,
1994.
2 Weinberg AD, Minaker KL; Council on Sci-
entific Affairs, American Medical Associa-
tion: Dehydration – evaluation and manage-
ment in older adults. JAMA 1995; 274:
1552–1556.
3 Dauterman KW, Bennet RG, Greenough WB
3rd, Redett RJ, Gillespie JA, Applebaum G,
Schoenfeld CN: Plasma specific gravity for
identifying hypovolemia. J Diarrhoeal Dis
Res 1995;13:33–38.
4 Lavizzo-Mourey R, Johnson J, Stolley P: Risk
factors for dehydration among elderly nurs-
ing home residents. J Am Geriatr Soc 1988;
36:213–218.
5 Evans JM, Chutka DS, Fleming KC, Tangalos
EG, Vittone J, Heathman JH: Medical care of
nursing home residents. Mayo Clin Proc
1995;70:694–702.
6 Silver AJ: Aging and risks for dehydration.
Cleve Clinic J Med 1990;57:341–344.
7 Ayus JC, Arieff AI: Abnormalities of water
metabolism in the elderly. Semin Nephrol
1996;16:277–288.
8 Crane MG, Harris JJ: Effect of aging on renin
activity and aldosterone excretion. J Lab Clin
Med 1976;87:947–959.
9 Kirkland J, Lye M, Goddard F, et al: Plasma
arginine vasopressin in dehydrated elderly
patients. Clin Endocrinol 1984;20:451–456.
10 Rowe JW: Renal system; in Rowe JW, Bes-
dine RW (eds): Health and Disease in Old
Age. Boston, Little, Brown, 1982, pp 165–
183.
Dehydration in Elderly Patients with
Oropharyngeal Dysphagia
11 Andreucci VE, Russo D, Cianciaruso B, An-
dreucci M: Some sodium, potassium and wa-
ter changes in the elderly and their treat-
ment. Nephrol Dial Transplant 1996; 11
(suppl):9–17.
12 Weinberg AD, Pals JK, McGlinchey-Berroth
R, Minaker KL: Indices of dehydration
among frail nursing home patients: highly
variable but stable over time. J Am Geriatr
Soc 1994;42:1070–1073.
13 Mahowald JM, Himmelstein DU: Hyperna-
tremia in the elderly: relation to infection
and mortality. J Am Geriatr Soc 1981; 29:
177–180.
14 Warren JL, Bacon WE, Harris T, McBean
AM, Folet DJ, Phillips C: The burden and
outcomes associated with dehydration
among US elderly 1991. Am J Public Health
1994;84:1265–1269.
15 Wei JY, Gersh BJ: Heart disease in the elder-
ly. Curr Prob Cardiol 1987;12:1–65.
16 Gross CR, Lindquist RD, Wooley AC, et al:
Clinical indicators of dehydration severity in
elderly patients. J Emerg Med 1992; 10: 267–
274.
17 Sergi G, Perini P, Bussolotto M, Zurlo F, Mal-
vasi L, Carraro R, Del Prato S, Giantin V,
Enzi G: Body composition study in the el-
derly: comparison between tritium dilution
method and dual photon absorptiometry. J
Gerontol 1993;48:M244–M248.
18 Kim DE, Berlowitz DR: The limited use of
routine laboratory assessments in severely
impaired nursing home residents. JAMA
1994;274:1447–1452.
19 Weinberg AD, Pals JK, Levesque PG, Beal
LF, Cunningham TJ, Minaker KL: Dehydra-
tion and death during febrile episodes in the
nursing home. J Am Geriatr Soc 1994; 42:
968–971.
20 Thomas DR, Tariq SH, Makhdomm S, Had-
dad R, Moinuddin A: Physician misdiagno-
sis of dehydration in older adults. J Am Med
Dir Assoc 2003;4:251–254.
Gerontology 2007;53:179–183
21 Clarfield M, Paltiel A, Gindin Y: Country
profile: Israel. J Am Geriatr Soc 2000; 48:
980–984.
22 Salassa J: A Functional outcome swallowing
scale for staging oropharyngeal dysphagia.
Dig Dis 1999;17:230–234.
23 Wallach J: Interpretation of Diagnostic Tests,
ed 5. Boston, Little, Brown, 1992.
24 Mentes J: Oral hydration in older adults:
greater awareness is needed in preventing,
recognizing, and treating dehydration. Am J
Nurs 2006;106:40–49.
25 Steele CM, Greenwood C, Ens I: Mealtime
difficulties in a home for the aged: not just
dysphagia. Dysphagia 1997;12:43–50.
26 Olde Rikkert MGM, van’t Hof MA, Baaden-
huysen H, Hoefnagels WHL: Individuality
and responsiveness of biochemical indices of
dehydration in hospitalized elderly patients.
Age Ageing 1998;27:311–319.
27 Faulkner WR, Meites S (eds): Geriatric Clin-
ical Chemistry: Reference Values. Washing-
ton, AACC Press, 1993.
28 Armstrong-Esther CA, Browne KD, Arm-
strong-Esther DC, Sander L: The institu-
tionalized elderly: dry to the bone! Int J Nurs
Stud 1996;33:619–628.
29 Simmons SF, Alessi C, Schnelle JF: An inter-
vention to increase fluid intake in nursing
home residents: prompting and preference
compliance. J Am Geriatr Soc 2001;49: 926–
933.
30 Fries BE, Hawes C, Morris JN, Phillips CD,
Mor V, Park PS: Effect of the National Resi-
dent Assessment Instrument on selected
health conditions and problems. J Am Geri-
atr Soc 1997;45:994–1001.
183