Journal of Veterinary Emergency and Critical Care 19(4) 2009, pp 347–351

Retrospective Study doi:10.1111/j.1476-4431.2009.00435.x

The prevalence and significance of hyperglycemia

in hospitalized cats
Casey C. Ray, DVM; Julie Callahan-Clark, DVM; Nicole F. Beckel, DVM, DACVECC and
Patricia C. Walters, VMD, DACVIM, DACVECC

Abstract
Objective – To report the prevalence of hyperglycemia in cats admitted to a veterinary hospital and to
determine if hyperglycemic cats had increased morbidity and mortality when compared with normoglycemic
cats.
Design – Retrospective clinical study.
Setting – Community-based referral hospital.
Animals – Nondiabetic cats admitted to the hospital.
Interventions – None.
Measurements and Main Results – The medical records of nondiabetic cats admitted to the hospital over a
1-year period were reviewed. There were 182 cats that met the criteria for inclusion in the study. Information
obtained included signalment, length of hospitalization, initial and highest blood glucose measurement,
diagnosis, treatment, and final disposition. Sixty-three percent of cats (116/182) were hyperglycemic at the
time of presentation. Total incidence of hyperglycemia at any point during hospitalization was 64% (118/182).
No association was found between hyperglycemia either initially or at any point during the hospitalization
and mortality. However, a significant association was documented between the presence of hyperglycemia
and increased length of hospitalization (LOH) (P 5 0.04). The duration of LOH was also significantly
associated with the degree of hyperglycemia (P 5 0.01). A number of different disease processes were
represented in the study population. However, the number of cats in each disease category was small and no
association could be found between any of them and blood glucose affecting mortality and morbidity.
Conclusion – The prevalence of hyperglycemia in feline patients admitted to a primary referral hospital was
64%. Cats with hyperglycemia had a longer LOH when compared with normoglycemic cats; however,
presence of hyperglycemia did not impact mortality in this population of cats.
(J Vet Emerg Crit Care 2009; 19(4): 347–351) doi: 10.1111/j.1476-4431.2009.00435.x

Keywords: critical care monitoring, glucose metabolism, small animal critical care

Introduction the study had higher blood glucose (BG) concentrations

than did survivors.2 One investigation of hyper-
Human and veterinary studies have revealed a rela-
glycemia in cats presenting to an emergency depart-
tionship between hyperglycemia and morbidity in hos-
ment found that 40% of the presenting cats were
pitalized patients. Such studies show that in critically ill
hyperglycemic. This study also showed that cats with
humans and dogs, hyperglycemia increases both mor-
hyperglycemia were more likely to be euthanized than
bidity and mortality. A recent study of hyperglycemia
cats that did not have hyperglycemia. Hyperglycemic
in humans in a general hospital population found that
cats also had a higher rate of hospital admission.a
it was associated with a poor outcome.1 Furthermore, a
Stress hyperglycemia has been documented in hu-
study on dogs found that critically ill hyperglycemic
mans,3 dogs,4 and cats,5 with cats appearing to be par-
dogs had an increased length of hospitalization (LOH)
ticularly prone to this phenomenon. In this white coat
that was characterized as morbidity; nonsurvivors in
syndrome, BG increases when an animal is subjected to
physiological or psychological stress; it has been clas-
From the New England Animal Medical Center, West Bridgewater, Mas-
sachusetts, MA 01801.
sically attributed to the release of epinephrine. Illness,
hospital procedures (phlebotomy, catheterization, and
Address correspondence and reprint requests to
Dr. Casey Ray, Massachusetts Veterinary Referral Hospital, 20 Cabot Rd,
restraint), and hospitalization are common stressors.
Woburn, MA 01801, USA. Email: raycase04@yahoo.com Because cats are metabolically different from dogs and

& Veterinary Emergency and Critical Care Society 2009 347

C.C. Ray et al.
humans, results from studies of glucose metabolism in
humans and dogs cannot necessarily be applied to cats.
There is current interest in tight glucose control in
hospitalized humans.6,7 To our knowledge, no studies
have evaluated the implications of hyperglycemia for
cats admitted to a community-based referral veterinary
hospital. The purpose of this study is to report the
prevalence of hyperglycemia in feline patients admitted
to such a facility and to determine if hyperglycemic
patients had increased morbidity and mortality when
compared with normoglycemic controls. Additionally,
we attempted to determine if the reason for hospital-
ization and the presence of a fever had any influence
on hyperglycemia. Our hypothesis is that the preva-
lence of hyperglycemia in cats is high and that these
hyperglycemic cats would have an increased mor-
bidity and mortality as compared with normoglycemic
cats.
Materials and Methods
The medical records of all cats admitted into the New
England Animal Medical Center between January 2007
and December 2007 were examined to determine
eligibility. Criteria for inclusion in the study were
hospitalization for 24 hours or longer, and that a BG
measurement was taken on a point-of-care (POC)
chemistry analyzerb within 6 hours of admission. Cats
were excluded if they were admitted for routine pro-
cedures (ovariohysterectomy, onychectomy, and dental
cleaning), had a diagnosis of diabetes mellitus, or
received dextrose or glucocorticoids as part of their
treatment.
No intervention was done as a part of the study. The
cats were housed in a combined canine and feline gen-
eral hospital ward and critical care unit; healthy ani-
mals were housed in a separate ward. BG concentration
was measured at the discretion of the attending
clinician; however, in our hospital BG is typically eval-
uated at least once a day. Only BG measurements ob-
tained from the POC analyzer identified above were
included in the study. BG measurements were re-
corded, along with the time that the BG was taken,
signalment, diagnosis, treatment, days in hospital, and
the final disposition of the patient (died or discharged
from the hospital). All BGs taken were recorded as well
as the time after admission to the hospital (in hours)
each BG was taken. Morbidity was measured as LOH,
and patients that died naturally or were euthanized for
any reason were recorded as nonsurvivors. We also re-
corded the diagnosis and the presence of fever in the
cats. Cats were regarded as febrile if they had a tem-
perature of 39.71C (103.51F) or greater.
348 & Veterinary Emergency and Critical Care Society 2009, doi: 10.1111/j.1476-4431.2009.00435.x
Normoglycemia was defined as 0.44–7.22 mmol/L
(80–130 mg/dL), the reference interval established by
the POC analyzer manufacturer for BG in cats.c A
BG47.22 mmol/L (130 mg/dL) was considered to be
hyperglycemic. To determine if there was bias at a
BG47.22 mmol/L (130 mg/dL), we made additional
comparisons with BG: 7.23 mmol/L (131 mg/dL) to
8.33 mmol/L (150 mg/dL), 8.34 mmol/L (151 mg/dL)
to 9.44 mmol/L (170 mg/dL), 9.45 mmol/L (171 mg/
dL) to 11.10 mmol/L (200 mg/dL), and 411.1 mmol/L
(4200 mg/dL). Furthermore, a subset of cats whose BG
was taken within 2 hours of admission were also an-
alyzed separately. The cats were also grouped by dis-
ease process, including the presence of fever. We
compared all these groups to cats that were normo-
glycemic throughout their hospitalization.
The POC analyzer used in this study measures BG
amphometrically, and has been validated for use in
cats,c and has been shown to accurately measure BG in
cats.d The POC analyzer uses an electrochemical reac-
tion to obtain results from either fresh whole blood or
blood anticoagulated with lithium heparin.c Anticoag-
ulated samples were analyzed within 5 minutes of be-
ing drawn.
Logistic regression was used to determine the effects
of hyperglycemia, primary diagnosis, fever, age, and
breed on mortality. The significance of the Wald statistic
was specified at P  0.05.
Independent t-test procedures were used to deter-
mine the effects of binary independent measures of
hyperglycemia (eg, presence or absence of hyper-
glycemia during initial visit) on morbidity. A 2-tailed
significance level of P 0.05 was set. To assess the ef-
fects of continuous measures of hyperglycemia (eg,
value of highest BG [hBG] at any time point), linear
regression procedures were performed. To assess the
effects of independent variables with 3 or more cate-
gories, 1-way ANOVA procedures were utilized. A sig-
nificance level of P  0.05 was specified for linear
regression and 1-way ANOVA procedures. All statisti-
cal analysis was done by a commercial statistics con-
sulting firm.e
Results
During the study period, 926 cats were admitted to the
hospital. Of these 926 cats, 182 (19.7%) met the overall
criteria for inclusion in the study (the study group).
Background information on the 182 cats included age,
sex, breed, LOH, initial BG (iBG), and hBG recorded.
Ages in the group ranged from 0.2 to 20 years, with a
median age of 8.0 years. The study group included 8 of
182 (4.4%) intact females, 67 of 182 (36.8%) spayed fe-
males, 5 of 182 (2.7%) intact males, and 102 of 182 (56%)

were neutered males. The majority of the cats were do-
mestic shorthair (DSH) (124/182, [68.1%]), with other
prominent breeds represented consisting of domestic
medium hair (9/182, [4.9%]), domestic longhair (DLH)
(31/182, [17%]), and other breeds (18/182, [10%]).
LOH was not normally distributed. For the entire
group LOH ranged from 1 to 9 days, with a mean of
2.19 (SD  1.18) days. The iBG recorded ranged from
4.72 mmol/L (80 mg/dL) to 21.36 mmol/L (385 mg/
dL), with a mean of 8.93 (SD  3.23) mmol/L (161.0
[SD  58.3] mg/dL). The hBG recorded range was iden-
tical to that of the iBG but with a mean of 8.99
(SD  3.33) mmol/L (162.9 [SD  60.0] mg/dL). Sixty-
six of 182 (36.2%) cats were normoglycemic at presen-
tation; therefore, 116 of 182 (63.7%) were hyperglycemic
at presentation (the iBG group). Also, 118 of 182 (64.8%)
were hyperglycemic at some point during their hospi-
talization (the hBG group). Thirty-six of the 182 cats
died, consituting a mortality rate of 19.8%.
Of the 36 cats that died, 24 (66.7%) were hyper-
glycemic (see Table 1). No association was found be-
tween mortality and hyperglycemia either initially or at
any point during hospitalization. Also, no association
was found between different levels of hyperglycemia
and mortality. In the subset of cats whose BG was taken
within 2 hours of admission, no association was found
between hyperglycemia and mortality (P 5 0.73).
The mean LOH for the iBG group was 2.21 days. No
significant association was found between iBG and
LOH. The mean LOH of the hBG group was 2.32 days,
and a statistically significant association was found be-
tween hyperglycemia at any point during hospitaliza-
tion and LOH (P 5 0.04) in that cats with
hyperglycemia tended to have a longer LOH than
Table 1: Comparison of age, initial glucose, and length of hos-
pitalization between survivors and nonsurvivors
Overall Survivors
N 182 146
Age (y)
Range 0.2–20 0.2–19
Mean (SD) 8.03 (  4.95) 7.87 (  4.8)
iBG (mmol/L)
Range 4.4–21.4 4.1–21.4
Mean (SD) 8.9 (  3.2) 9.0 (  3.3)
iBG (mg/dL)
Range 80–385 80–385
Mean (SD) 161 (  58.3) 162 (  59.4)
LOH (d)
Range 1–9 1–9
Mean (SD) 2.19 (  1.18) 2.23 (  1.22)
N, number of cats; iBG, initial blood glucose; LOH, length of hospitaliza-
tion.
& Veterinary Emergency and Critical Care Society 2009, doi: 10.1111/j.1476-4431.2009.00435.x
Prevalence of hyperglycemia in cats
normoglycemic cats. Also, the higher the BG at any
point during hospitalization, the longer the LOH
(P 5 0.01). No statistical association was found between
LOH and either the different levels of hyperglycemia or
the hour the BG was taken, including within 2 hours of
admission.
The cats in the study had a number of different dis-
ease processes (see Table 2). No statistically significant
association could be found between hyperglycemia and
the disease process in regards to either mortality or
morbidity. Febrile cats were compared separately from
the primary disease process. In the study 16 of 182 cats
(8.8%) were febrile at some point during their hospi-
talization, including cats that were hospitalized for a
fever of unknown origin (included in the open disease
category). Of the 16 febrile cats, 15 (93.7%) were hyper-
glycemic. The febrile cats were compared with the iBG
and hBG groups and no statistically significant corre-
lation could be found.
We also compared multiple other parameters (age,
breed, and the change in BG [DBG]) with survival and
morbidity. The chances of a hospitalized cat dying in-
creased as the DBG became smaller (P 5 0.047). Also, in
our population, DSH cats were more likely to survive
than DLH cats (P 5 0.02). In regards to morbidity nei-
ther age, breed, nor DBG were statistically significant.
Discussion
In this retrospective study, 64.8% of hospitalized cats
were hyperglycemic during their hospitalization, which
contrasts to findings in a recent study of critically ill
dogs showing that only 16% of hospitalized dogs were
found to be hyperglycemic.2 Similarly to the aforemen-
tioned study, the current study showed an association
between hyperglycemia, at any point during hospital-
ization, and LOH. Cats with a higher BG tended to stay
in the hospital longer. However, there was no associ-
ation with an elevated BG at initial presentation and
Nonsurvivors LOH. Furthermore, this study found no association be-
tween the presence of hyperglycemia and mortality,
36
which contrasts the canine study. It is intriguing that
0.25–20 the lower the DBG the odds of death increase and that
9.0 (  5.8) DSH cats are less likely to die than DLH cats. However,
the clinical relevance of these findings is unclear.
4.5–15.9
Hyperglycemia is associated with stress in cats. Rand
8.9 (  2.9)
et al5 have demonstrated that a stressful event causes
81–286 an increase in BG in 100% of experimental cats, an in-
160.0 (  52.5) crease that was also correlated to an increase in blood
lactate and norepinephrine levels. Furthermore, a study
1–5
of hospitalized critically ill cats showed hyperglycemia
2.08 (  1.06)
along with increased cortisol, lactate, and norepineph-
rine levels, when compared with healthy normo-
glycemic controls.8 The increased BG, lactate, and
349
C.C. Ray et al.

Table 2: Disease categories represented in the study population

Total group Normoglycemic Hyperglycemic

N Percentage (%) N Percentage (%) N Percentage (%)

Neoplasia 18 9.9 6 9.4 12 10.2
Renal 22 12.1 11 17.2 11 9.3
Infectious 13 7.1 5 7.8 8 6.8
Anemia 2 1.1 0 0 2 1.7
Oral disease 3 1.6 1 1.6 2 1.7
GI 31 17.0 13 20.3 18 15.3
Open/FUO 12 6.6 2 3.0 10 8.5
Neurological 11 6.0 4 6.3 7 5.9
Cardiovascular 11 6.0 1 1.6 10 8.5
Trauma 23 12.6 4 6.3 19 16.1
Burn 1 0.5 1 1.6 0 0
Toxicity 6 3.3 6 9.4 0 0
Urinary 22 12.1 8 12.5 14 11.9
Liver disease 1 0.5 0 0 1 0.8
Orthopedic 1 0.5 0 0 1 0.8
Cardiac arrest 1 0.5 0 0 1 0.8
Respiratory 4 2.2 2 3.0 2 1.7
Total 182 100.0 64 100.00 118 100.00

N, number of cats; GI, gastrointestinal; FUO, fever of unknown origin.

norepinephrine levels in the stressed cats paralleled
those in the critically ill cats. Therefore, illness and pa-
tient procedures provoke a stress response and cause
hyperglycemia that could explain the development of
hyperglycemia in the majority of the hospitalized cats
in this study. Indicators of stress such as norepineph-
rine, cortisol, or lactate were not measured in this study.
The correlation with LOH indicates that hyper-
glycemia does affect cats in a similar manner as other
species, an outcome that might not be expected because
of the distinctive biochemistry of feline metabolism
when compared with dogs and humans. Cats are true
carnivores and as a result have minimal activity of he-
patic glucokinase, a liver enzyme that is responsible for
phosphorylation of glucose, and it is not inducible as it
is in other species. Furthermore, cats have minimal he-
patic glycogen synthetase. Thus, they do not efficiently
convert glucose to glycogen and may be more suscep-
tible to hyperglycemia. It has been shown that cats have
difficulty minimizing hyperglycemia.9
In the present study, the presence of fever was looked
at independently of the overall disease process. Al-
though no known association could be found between
the presence of hyperglycemia and fever, this relation-
ship was evaluated because of the high number of fe-
brile cats that were also hyperglycemic. Acute
hyperglycemia does increase the level of circulating
proinflammatory cytokines interleukin-6 (IL-6), tumor
necrosis factor-a, and IL-18 in humans.10 Two of these
cytokines, IL-6 and tumor necrosis factor-a, are princi-
ple mediators of fever.11 However, no association was
found between the presence of fever and hyper-
glycemia. This may have been due to the relative small
sample size. The authors urge additional study in this
intriguing subset of patients.
Seventeen different disease processes occurred in the
study population, yet none of them were correlated
with hyperglycemia. Because so many different disease
processes occurred in the population, the number of
cats in each disease category was relatively low. Cor-
relations may exist between hyperglycemia and partic-
ular disease processes, but in this study the number of
cats in some disease categories was not high enough to
be statistically significant. Further research needs to be
done on specific diseases. For example, hyperglycemia
has been shown to be a significant factor in coronary
disease. Hyperglycemia in humans with myocardial
infarction has long been correlated with an increase
in adverse outcomes12 and has proven to indepen-
dently predict mortality in myocardial infarction pa-
tients.13 A similar correlation has been shown in
hyperglycemic dogs.14 In this study, we found no cor-
relation between hyperglycemia in cardiac disease and
outcome, but the small sample size may have skewed
the results.
Cats were excluded from the study if they were re-
ceiving insulin, dextrose infusions, gluccorticoids, or if
they had been diagnosed with diabetes because these
medications and diabetes have been known to affect
glucose metabolism15 and could artificially influence
the study results. Furthermore, these are situations in
which the BG might influence a clinician’s decisions.

350 & Veterinary Emergency and Critical Care Society 2009, doi: 10.1111/j.1476-4431.2009.00435.x

This study shows that hyperglycemia had no clinical
effect on mortality in a population of cats hospitalized
because of illness, without regard to age, sex, or disease
process. However, hyperglycemia does affect morbidity
in that the higher the BG, the longer the LOH. These
results suggest that the presence of hyperglycemia can-
not be used to predict mortality in sick hospitalized
cats, but the degree of hyperglycemia might be an in-
dicator of the severity of illness in cats. The strengths of
this study lie in its relatively large sample size and
standardized BG sampling, while its weaknesses in-
clude its retrospective nature, lack of cats with diseases
known to affect or to be affected by hyperglycemia, and
the inability to determine if euthanasia was the result of
medical or financial reasons. Further prospective stud-
ies into the effect of hyperglycemia in cats with heart
disease and febrile cats are warranted.
Acknowledgements
The authors would like to thank Karen Martin and
Helen Hammond for help with data collection, Chris-
tine Koiralla for help with statistics and data organiza-
tion, and Bernard Kaetz for editorial support.
Footnotes
a
Chan DL, Freeman LM, Rozanski EA, Rush JE. Prevalence of
hyperglycemia in cats presented to the emergency service (abstr). J Vet
Emerg Crit Care 2002;12(3):199.
b
I-STAT Heska Corp, Loveland, CO.
c
Corporate communication, Heska Corp.
d
Cohn LA, McCaw DL, Tate, D. Comparison of glucose measurement
devices in cats (abstr). J Vet Intern Med 2001;15(3):309.
e 2006; 67(4):583–587.
Minimax Consulting LLC., Providence, RI.
& Veterinary Emergency and Critical Care Society 2009, doi: 10.1111/j.1476-4431.2009.00435.x
Prevalence of hyperglycemia in cats
References
1. Umpierrez G, Isaacs S, Bazargan N, et al. Hyperglycemia: an in-
dependent marker of in-hospital mortality in patients with undi-
agnosed diabetes. J Clin Endocrinol Metab 2002; 87(3):978–982.
2. Torre D, deLaforcade A, Chan D. Incidence and clinical relevance
of hyperglycemia in critically ill dogs. J Vet Intern Med 2007;
21:971–975.
3. McCowen K, Malhotra A, Bistrian B. Stress-induced hyper-
glycemia. Crit Care Clin 2001; 17(1):107–124.
4. McGuinness O, Fugiwara T, Murrell S, et al. Impact of chronic
stress hormone infusion on hepatic carbohydrate metabolism in
the conscious dog. Am J Physiol 1993; 265:E314–E322.
5. Rand J, Kinnaird E, Baglioni A, et al. Acute stress hyperglycemia
in cats is associated with struggling and increased concentrations
of lactate and norepinephrine. J Vet Intern Med 2002; 16:123–132.
6. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin
therapy in critically ill patients. N Engl J Med 2001; 345(19):
1359–1367.
7. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin
therapy in the medical ICU. N Engl J Med 2006; 354(5):449–461.
8. Chan D, Freeman L, Rozanski E, et al. Alterations in carbohydrate
metabolism in critically ill cats. J Vet Emerg Crit Care 2006;
16(2):S7–S13.
9. Zoran D. The carnivore connection to nutrition in cats. J Am Vet
Med Assoc 2002; 221(11):1559–1567.
10. Esposito K, Nappo F, Marfella R, et al. Inflammatory cytokine
concentrations are acutely increased by hyperglycemia in humans:
role of oxidative stress. Circulation 2002; 106:2067–2072.
11. Marik P. Fever in the ICU. Chest 2000; 117:855–869.
12. Capes S, Hunt D, Malmberg K, et al. Stress hyperglycemia and
increased risk of death after myocardial infarction in patients with
and without diabetes: a systemic overview. Lancet 2000; 355:
733–778.
13. Bolk J, van der Ploeg T, Arnold A, et al. Impaired glucose metab-
olism predicts mortality after a myocardial infarction. Int J Cardiol
2001; 79:207–214.
14. Brady C, Hughes D, Drobatz K. Association of hyponatremia and
hyperglycemia with outcome in dogs with congestive heart fail-
ure. J Vet Emerg Crit Care 2004; 14(3):177–182.
15. Ployngam T, Tobias A, Smith S, et al. Hemodynamic effects of
methylprednisolone acetate administration in cats. Am J Vet Res
351