J Vet Intern Med 2000;14:20–26

Metabolic and Hormonal Alterations in Cats with Hepatic Lipidosis

Bonnie Brown, Glenna E. Mauldin, Joel Armstrong, Scott D. Moroff, and G. Neal Mauldin

Hepatic lipidosis in cats is a commonly diagnosed hepatobiliary disease of unknown cause. The purpose of this prospective study
was to characterize the blood hormone and lipid status of cats with hepatic lipidosis, and to compare this status to that of cats
with other types of liver disease and to control cats. Twenty-three cats with hepatic disease were assigned to 1 of 2 groups on the
basis of cytopathologic or histopathologic examination of the liver: group 1, hepatic lipidosis (n 5 18); or group 2, cholangiohepati-
tis (n 5 5). Ten healthy young adult cats were used as controls. Food was withheld from control animals for 24 hours before blood
collection. Concentrations of plasma glucagon and serum insulin, cortisol, thyroxine, triglycerides, cholesterol, phospholipids, and
nonesterified fatty acids (NEFAs) were determined in all cats, in addition to routine hematologic and serum biochemical testing.
Cats with hepatic lipidosis had higher serum NEFA concentrations than cats with cholangiohepatitis or control cats (P , .05). Cats
with cholangiohepatitis had higher serum cholesterol and phospholipid concentrations than those of cats with lipidosis or control
cats (P , .05); their plasma glucagon concentrations were higher than those of control cats (P , .05), but were not different from
those of cats with hepatic lipidosis. Serum insulin concentrations were significantly higher in control cats than in diseased cats (P
, .05), but neither serum insulin nor the insulin to glucagon ratio was significantly different among the cats with hepatic disease.
The high concentration of NEFAs in cats with hepatic lipidosis suggests that at least 1 factor in the pathogenesis of this syndrome
may involve the regulation of hormone-sensitive lipase.
Key words: Anorexia; Cholangiohepatitis; Glucagon; Insulin; Nonesterified fatty acid; Triglyceride.

H epatic lipidosis is a commonly diagnosed hepatobili-

ary disease of cats.1,2 Clinical signs in cats with he-
patic lipidosis include anorexia, weight loss, jaundice, he-
patomegaly, vomiting, depression, and hypersalivation. The
disease may occur in association with underlying disease
such as pancreatitis,3 neoplasia, or small intestinal disease,
or it may be idiopathic.2 Obesity or stress are considered
contributory factors,1,2,4–7 but the underlying etiology of the
disease is unknown. The diagnosis of hepatic lipidosis is
based on cytologic or histopathologic evaluation of the liv-
er.2,8 The prognosis for affected cats was poor (survival rate
of 5–20%)5,9 until the institution of aggressive nutritional
support, which improved the survival rate to 50–60%.2,6,7,10–13
The pathophysiology of feline hepatic lipidosis is incom-
pletely understood. Obese but otherwise healthy cats un-
dergoing food restriction sufficiently severe to result in a
30–40% loss of body weight develop a syndrome similar
to naturally occurring hepatic lipidosis.6,13–17 However, the
variability in reported historical, physical, and clinicopath-
ologic findings in cats with naturally occurring hepatic lip-
idosis suggests that this is a syndrome with many causative
factors. Proposed etiologies include increased peripheral li-
From the Departments of Medicine (Brown, Mauldin, Mauldin) and
Pathology (Moroff), The Elmer and Mamdouha Bobst Hospital of The
Animal Medical Center, New York, NY; and the Radionuclide and
Hormone Radioimmunoassay Laboratory of the University of Wiscon-
sin School of Veterinary Medicine, Madison, WI (Armstrong). Dr
Brown is presently affiliated with The Alex Lewyt Veterinary Medical
Center, Port Washington, NY. Drs Mauldin and Mauldin are presently
affiliated with the Department of Veterinary Clinical Sciences, School
of Veterinary Medicine, Louisiana State University, Baton Rouge, LA.
Dr Moroff is presently affiliated with Antech Diagnostics, Farming-
dale, NY.
Reprint requests: Glenna E. Mauldin, DVM, MS, Department of
Veterinary Clinical Sciences, Louisiana State University School of Vet-
erinary Medicine, Baton Rouge, LA, 70803; e-mail: gmauldin@
vetmed.lsu.edu.
Submitted July 15, 1998; Revised March 23, 1999, and June 21,
1999; Accepted August 4, 1999.
Copyright q 2000 by the American College of Veterinary Internal
Medicine
0891-6640/00/1401-0003/$3.00/0
polysis secondary to an absolute or relative insulin defi-
ciency; fatty liver induced by protein–calorie malnutrition
similar to kwashiorkor or marasmus in humans; essential
amino acid deficiencies that result in the inability to syn-
thesize sufficient apolipoproteins to mobilize hepatic fat;
deficiency of lipotrophic compounds; an inborn or acquired
error of fatty acid oxidation; and hepatic peroxisomal dam-
age due to oxidative stress.1,2,4–7,13–17 Toxic causes of hepatic
lipidosis also have been suggested, including overgrowth of
intestinal anaerobes, as well as intoxication by drugs, chem-
icals, and plant toxins.1,4–6,9 The supposition that nutritional
deficiencies are involved in the etiopathogenesis of this syn-
drome has led to the addition of arginine, citrulline, thre-
onine, carnitine, choline, vitamin B12, and inosital to the
enteral feeding solutions used to treat cats with hepatic lip-
idosis.18–21
Perturbations in both nutritional and hormonal status of
cats with hepatic lipidosis are likely to be important factors
in the development of the fatty liver syndrome. We hy-
pothesized that, regardless of underlying etiology, cats af-
fected with hepatic lipidosis would have alterations in cir-
culating lipid and hormone concentrations that were distinct
from those observed in normal fasted cats, or in cats with
other liver diseases. This paper reports the results of a pro-
spective study that compared serum lipid variables and
plasma or serum hormone concentration in 3 groups of cats:
cats with hepatic lipidosis, cats with other liver disorders,
and a group of control cats. We measured serum triglyc-
erides, cholesterol, phospholipids, and nonesterified fatty
acids (NEFAs) in an attempt to identify a serum lipid profile
specific to hepatic lipidosis. We also examined the concen-
tration in plasma of factors that regulate hormone-sensitive
lipase activity and peripheral lipolysis through measure-
ment of plasma glucagon and serum insulin, cortisol, and
thyroxine concentrations in our 3 patient groups. Historical,
physical, and routine clinicopathologic findings were com-
pared among the 3 groups as well.
Materials and Methods
Plasma glucagon and serum insulin, cortisol, thyroxine, triglyceride,
cholesterol, phospholipid, and NEFA concentrations, as well as sig-
 Metabolic Alterations in Lipidosis
nalment, historical data, results of physical examination, CBC, and
routine serum biochemical variables, were compared in 3 groups of
cats. Cats in group 1 had severe hepatic lipidosis. Cats in group 2 had
cholangiohepatitis. Cats in group 3 were clinically normal cats.
Cats with Hepatic Lipidosis and Cats with Other
Liver Disease
Thirty-four sequentially identified cats with primary hepatic disease
that were admitted to the Animal Medical Center between February
1991 and February 1992 were evaluated for inclusion in this study.
Cats were suspected of having liver disease based on historical data
and results of physical and clinicopathologic examination. Plasma for
measurement of plasma glucagon and serum insulin, cortisol, thyrox-
ine, triglyceride, cholesterol, phospholipid, and NEFA concentrations
was obtained from all cats before initiation of enteral or parenteral
feeding. Pretreatment complete blood counts,a serum biochemical pro-
files,b feline leukemia,c and feline immunodeficiency virus testingd
were also completed on all animals.
Diagnosis of liver disease was confirmed by either cytologic ex-
amination of a fine-needle aspirate or by histopathologic examination
of liver tissue. Liver tissue for histopathology was obtained by ultra-
sound-guided percutaneous collection of liver biopsy, during explor-
atory celiotomy, or at the time of postmortem examination. Specimens
were fixed in 10% buffered formalin, embedded in paraffin, and
stained with hematoxylin and eosin. Cats with liver disease were as-
signed to 1 of 2 groups based on the results of cytologic or histopath-
ologic examination of liver tissue. Cats with lipidosis were placed in
group 1; cats with all other forms of liver disease were placed in group
2. All specimens of liver tissue were reviewed by 1 of the authors
(SDM).
Control Cats
Ten mixed-source, adult cats without evidence of liver disease
served as controls. Complete blood counts,a serum biochemical pro-
files,b feline leukemia virus,c and feline immunodeficiency virusd tests
were performed on each animal. Plasma glucagon and serum insulin,
cortisol, thyroxine, triglyceride, cholesterol, phospholipid, and NEFA
concentrations were measured after a 24-hour fast, as outlined below.
Specimen Collection and Analysis
Pretreatment blood samples for determination of plasma glucagon
and serum insulin, cortisol, thyroxine, triglycerides, cholesterol, phos-
pholipids, and NEFA concentrations were collected from all 44 cats
in the study (control, group 1, and group 2).
Glucagon. Collection tubes for glucagon assay were prepared by
adding 100 mL of aseptically prepared aprotinin solution in 0.9% so-
dium chloride and 0.9% benzyl alcohol containing 5–10 trypsin in-
hibitor units per milliliter of solution into 3-mL draw potassium ethy-
lenediaminetetraacetic acid (EDTA) tubes already containing 0.068
mL of 7.5% potassium EDTA. Two milliliters of patient whole blood
were added directly into these tubes. Blood samples were centrifuged
immediately at 620 3 g for 5 minutes. Plasma was harvested and
stored at 2208C. Glucagon concentrations were determined using a
commercial radioimmunoassay kit.e The primary antibody in this kit
is produced from human glucagon and is polyclonal. The detection
limit of the assay, defined as the apparent concentration 2 standard
deviations above the response at zero dose, is 13 pg/mL. The intraas-
say coefficient of variation for the kit as determined for each of 3
samples from the results of 20 pairs of tubes at 35 pg/mL is 15.7%,
at 151 pg/mL is 4.4%, and at 564 pg/mL is 4.1%. The interassay
coefficient of variation as determined for each of 3 samples from the
results of 20 different assays at 37 pg/mL is 15.7%, at 159 pg/mL is
6.5%, and at 534 pg/mL is 5.7%.22 The assay was validated for use
with feline plasma by taking equal amounts of plasma from 18 of the
44 samples submitted for analysis for this study, and combining them
21
to form a pooled plasma sample. Each of the assay’s standards was
diluted 1 : 1 with the pooled plasma and then assayed along with the
study samples. The pooled plasma itself was assayed as 5 different
samples, to accurately determine its glucagon content and to indicate
within-assay variation. All samples were run in duplicate. The 5
pooled plasma results were 104, 108, 98, 104, and 111 pg/mL. The
average of these 5 results was used for calculating the expected results
from mixing it in equal part with each of the standards. The observed
results as a percent of expected results were 20 pg/mL standard 5
128%; 50 pg/mL standard 5 120%; 100 pg/mL standard 5 102%;
200 pg/mL standard 5 88%; 500 pg/mL standard 5 81%; and 1,000
pg/mL standard 5 73%.
Insulin, Cortisol, and Thyroxine. Serum for determination of in-
sulin, cortisol, and thyroxine concentrations was frozen and stored at
2208C until assayed. Insulin,f cortisol,g and thyroxineh concentrations
were determined using commercial radioimmunoassay kits previously
validated for use on feline plasma.23–25
Triglycerides, Cholesterol, Phospholipids, and NEFA. Serum for
determination of triglyceride, cholesterol, phospholipid, and NEFA
concentration was frozen and stored at 2208C until assayed. Com-
mercial enzymatic assays for triglyceride,i cholesterol,i phospholipid,j
and NEFAsk were performed by an automated analyzer.l
Statistical Analysis
Statistical analyses were performed using a computer software pack-
age.m Ordinal data including sex and the proportion of cats with an-
orexia, lethargy, or hyperglycemia were compared between the 2
groups of diseased cats using a Fisher exact test. All continuous data
were tested for normality using a Kolmogorov–Smirnov test. Those
parameters passing the normality test were compared between all 3
groups of cats using a 1-way analysis of variance (ANOVA), followed
by a Tukey test for assessment of all pair-wise multiple comparisons.
When data failed the normality test, comparisons between the 3 groups
were made using a Kruskal–Wallis 1-way ANOVA on ranks. Post hoc
multiple comparisons were made in the latter case according to Dunn’s
method for nonparametric data. Data for 2 continuous variables, age
and length of anorexia, were compared between cats with hepatic lip-
idosis and cats with cholangiohepatitis only and a Mann–Whitney rank
sum test was used for this purpose. The potential relationship between
the presence or absence of hepatic lipidosis in the 2 groups of diseased
cats and serum or plasma insulin and glucagon concentrations was
analyzed using multiple logistic regression. A value of P , .05 was
considered significant in all cases.
Results
Group 1 initially consisted of 21 cats with idiopathic he-
patic lipidosis. Eighteen cats had severe lipidosis; in order
to provide as uniform a group as possible for comparison,
3 cats with moderate lipidosis were excluded from further
study. Severe hepatic lipidosis in the cat has been previ-
ously defined as diffuse involvement of greater than 50%
of hepatocytes with cytoplasmic vacuoles consistent in ap-
pearance with lipid.2 Fourteen of the 18 cats with severe
hepatic lipidosis were diagnosed by examination of liver
tissue biopsy, whereas 4 were diagnosed by cytologic ex-
amination of specimens of liver. Group 2 initially consisted
of 13 cats with liver disease other than hepatic lipidosis.
Five cats had chronic lymphocytic-plasmacytic
cholangiohepatitis with a suppurative component. Eight
cats had 5 other histologic diagnoses: Ito cell hyperplasia
(3), lymphosarcoma (2), hepatocellular degeneration (1),
hepatic abscessation (1), and granulomatous inflammation
(1). The latter 8 cats were excluded from further study,
again in order to provide as uniform a group as possible
22 Brown et al

Table 1. Clinicopathologic and clinical parameters in cats with hepatic lipidosis and cholangiohepatitis and normal control
cats.
Hepatic lipidosis (n 5 18) Cholangiohepatitis (n 5 5) Controls (n 5 10)
Parameter Median Range Median Range Median Range
Age (years) 5 (3–16) 14 (11–17) NA NA
Days of anorexia 14 (4–75) 10 (4–56) NA NA
Albumin (g/dL) 3.25 (2.4–4.2) 3.5 (2.8–3.8) 3.65 (3.2–4.0)
BUN (mg/dL) 18.10 (8.8–39.2) 22.65 (18.9–33.9) 28.45 (22.0–35.8)
Tbili (mg/dL) 3.85 (0.3–20.0) 11.85 (11.1–18.8) 0.55 (0.1–0.6)
ALP (IU/L) 518.5 (26–1191) 294.5 (278–366) 38.5 (24–63)
ALT (IU/L) 316 (90–812) 755 (443–787) 56.5 (43–81)
Insulin (pmol/L) 25.5 (2.9–378) 17 (10–180) 65 (23–147)
Glucagon (pg/mL) 82 (26–1252) 156 (54–227) 45.5 (20–65)
Cortisol (nmol/L) 67.5 (3–499) 69 (50–177) 73 (30–224)
T4 (nmol/L) 5.5 (1–19) 9 (5–180) 15 (6–19)
Chol (mg/dL) 124.5 (55–201) 259 (88–334) 117.5 (76–253)
TG (mg/dL) 115.5 (27–754) 99 (82–149) 32 (17–46)
Plipid (g/L) 2.70 (0.98–4.47) 4.66 (2.45–5.72) 2.64 (2.27–4.94)
NEFA (mmol/L) 1.69 (0.80–3.03) 0.99 (0.72–1.62) 0.50 (0.26–0.73)

NA, not available or not applicable; BUN, blood urea nitrogen; Tbili, total bilirubin; ALP, serum alkaline phosphatase; IU, International Unit;
ALT, alanine aminotransferase; T4, thyroxine; Chol, cholesterol; TG, triglyceride; Plipid, phospholipid; NEFA, nonesterified fatty acid.

for comparison. All 5 cats with cholangiohepatitis were
identified by histopathologic examination of biopsy speci-
mens.
The control cats included 4 spayed females and 6 cas-
trated males. These cats ranged in age from 2 to 4 years.
Their CBCs and serum biochemical profiles were within
reference ranges. All were feline leukemia and feline im-
munodeficiency virus test negative.
The median age of cats with hepatic lipidosis was sig-
nificantly younger than that of cats with cholangiohepatitis
(P 5 .008; Table 1). No difference in the proportion of
male and female cats was detected between the 2 liver dis-
ease groups (P 5 1.0), although a majority of the cats in
both groups were female: 12 of 18 cats (67%) with severe
hepatic lipidosis were female, and 4 of 5 cats (80%) with
cholangiohepatitis were female. No significant difference
was found between the hepatic lipidosis and cholangiohep-
Fig 1. Duration of anorexia for cats with hepatic lipidosis and cho-
langiohepatitis. The large box in the box plot extends from the 25th
to the 75th percentile. The small open box is the median. The whiskers
represent the 10th through 90th percentiles. The P value refers to
Mann–Whitney rank sum test results.
atitis groups in the proportion of cats with weight loss or
anorexia at the time of diagnosis. Fifteen of 18 cats (83%)
with hepatic lipidosis were anorexic, whereas 3 of 5 cats
(60%) with cholangiohepatitis were anorexic (P 5 .291).
Ten of 18 cats (60%) with hepatic lipidosis had documented
weight loss, and 3 of 5 cats (60%) with cholangiohepatitis
had documented weight loss (P 5 1.0). The median length
of anorexia before hospitalization was not significantly dif-
ferent between cats with hepatic lipidosis and cats with
cholangiohepatitis (P 5 .722; Table 1, Fig 1).
Significant differences in all 4 serum lipid variables were
observed between the 3 groups of cats in this study (see
Table 1). Cats with hepatic lipidosis had higher median se-
rum NEFA concentrations than did control cats or cats with
cholangiohepatitis (P , .05 for both comparisons), but no
significant difference was found in serum NEFA concen-
trations between cats with cholangiohepatitis and controls
(see Fig 2). Median serum triglyceride concentration was
significantly lower in control cats than it was in cats with
hepatic lipidosis or cholangiohepatitis (P , .05 for both
comparisons), but no significant difference could be found
between the 2 groups of diseased cats. Finally, cats with
cholangiohepatitis had higher serum concentrations of both
cholesterol and phospholipid when compared to hepatic lip-
idosis cats and controls (P , .05 for all comparisons), but
these 2 variables were not significantly different when cats
with hepatic lipidosis and control cats were compared to
each other.
Significant differences were found among the 3 groups
of cats with respect to all plasma or serum hormones as-
sayed in this study except cortisol (see Table 1). The me-
dian serum insulin concentration in the control cats was
significantly higher than in cats with liver disease, whether
they had hepatic lipidosis (P , .05) or cholangiohepatitis
(P , .05). However, no significant difference in serum in-
sulin concentration was found between cats with hepatic
lipidosis and cats with cholangiohepatitis (see Fig 3). Me-
 Metabolic Alterations in Lipidosis
Fig 2. Serum nonesterified fatty acid concentrations for cats with
hepatic lipidosis and cholangiohepatitis and control cats. The large box
in the box plot extends from the 25th to the 75th percentile. The small
open box is the median. The whiskers represent the 10th through 90th
percentiles. The open circles represent outliers. The P value refers to
1-way analysis of variance test results.
dian plasma glucagon concentrations were significantly
higher in cats with cholangiohepatitis than in control cats
(P , .05), but plasma glucagon concentrations in cats with
hepatic lipidosis were not significantly different from those
of cats in either of the other 2 groups (see Fig 4). The
insulin to glucagon ratio was significantly higher in control
animals than in cats with hepatic lipidosis (P , .05) or
cholangiohepatitis (P , .05), but no significant difference
in this ratio could be found between cats with hepatic lip-
idosis and cats with cholangiohepatitis. Furthermore, mul-
tiple logistic regression did not detect a significant relation-
ship between hepatic lipidosis, serum insulin, and plasma
glucagon concentrations. Lastly, median serum thyroxine
concentration was significantly higher in control cats than
in cats with hepatic lipidosis (P , .05), but no significant
difference was found when cats with cholangiohepatitis
were compared to cats with hepatic lipidosis or controls. A
single hyperthyroid cat was present in the cholangiohepa-
titis group.
CBCs in cats with hepatic lipidosis and cholangiohepa-
titis were variable, but no significant difference was found
in any hematologic variable among the 3 groups. Signifi-
cant differences were found between control cats and cats
with liver disease with respect to serum albumin, serum
urea nitrogen (SUN), total bilirubin concentrations, and se-
rum alkaline phosphatase (SAP), alanine transaminase
(ALT), and aspartate transaminase (AST) activities (see Ta-
ble 1). The median serum albumin concentration in control
cats was significantly higher than in cats with hepatic lip-
idosis (P , .05), but no statistical difference was found in
median serum albumin concentration between cats with he-
patic lipidosis and cats with cholangiohepatitis. Median
SUN was higher in control cats than in cats with hepatic
lipidosis (P , .05), but was not different when control cats
were compared to cats with cholangiohepatitis or when the
2 groups of cats with liver disease were compared to each
other. The median total serum bilirubin concentration and
SAP, ALT, and AST activities were significantly lower in
control cats than in either group of diseased cats (P , .05
23
Fig 3. Serum insulin concentrations for cats with hepatic lipidosis
and cholangiohepatitis and control cats. The large box in the box plot
extends from the 25th to the 75th percentile. The small open box is
the median. The whiskers represent the 10th through 90th percentiles.
The P value refers to Kruskal–Wallis 1-way analysis of variance on
ranks test results.
for each of the 8 pair-wise comparisons), but no significant
differences were found in these 4 variables between cats
with hepatic lipidosis and cats with cholangiohepatitis.
Serum glucose concentration was determined at least
twice in all cats with liver disease, from their time of hos-
pitalization up to the initiation of enteral feeding. Multiple
glucose determinations were used in order to differentiate
between hyperglycemia associated with glucose intolerance
and stress hyperglycemia. Nine of 18 cats with hepatic lip-
idosis had serum glucose concentrations within the labo-
ratory reference range (70–150 mg/dL), and 9 cats had per-
sistent mild hyperglycemia (from 150 to 240 mg/dL). Two
of 4 cats with cholangiohepatitis had normal serum glucose
concentrations, and 2 cats had persistent mild fasting hy-
perglycemia. Serial glucose determinations in 1 cat with
cholangiohepatitis were not available. The proportion of
cats with hyperglycemia was not different between the 2
groups of diseased cats (50% for both; P 5 1.0).
Fig 4. Plasma glucagon concentrations for cats with hepatic lipidosis
and cholangiohepatitis and control cats. The large box in the box plot
extends from the 25th to the 75th percentile. The small open box is
the median. The whiskers represent the 10th through 90th percentiles.
The P value refers to Kruskal–Wallis 1-way analysis of variance on
ranks test results.
24 Brown et al
Discussion
Significant differences in serum lipid variables and plas-
ma or serum hormone concentrations were clearly demon-
strated among the 3 groups of cats in this study. The most
striking difference was in the serum NEFA concentrations.
The median concentration of serum NEFAs was signifi-
cantly higher in cats with hepatic lipidosis than in cats with
cholangiohepatitis, even though the prevalence of weight
loss and the duration of anorexia were not significantly dif-
ferent between these 2 groups. Additionally, serum triglyc-
eride concentrations were increased in cats with hepatic lip-
idosis and cholangiohepatitis compared to controls. In con-
trast, serum concentrations of phospholipid and cholesterol
were higher in cats with cholangiohepatitis than either of
the other 2 groups. These observations closely parallel
those reported in human patients with fatty liver syndrome,
where NEFAs and triglycerides are the major serum lipid
fractions that are increased. Phospholipids and cholesterol
are usually only minimally elevated.26
NEFAs that enter the liver in anorexic animals are de-
rived from lipolysis of peripheral fat stores. Once inside the
liver these fatty acids may be oxidized for energy, con-
verted to phospholipids or cholesterol, or reesterified into
triglycerides. The rate of hepatic triglyceride synthesis is
usually balanced by the rate of hepatic secretion of triglyc-
erides, and hepatic triglyceride formation is proportional to
the concentration of fatty acids that enter the liver. The liver
responds to an increased influx of NEFAs with limited in-
creases in lipoprotein synthesis and secretion of triglycer-
ides in very-low-density lipoprotein (VLDL).26 The capac-
ity for increased rates of mitochondrial oxidation and ke-
tone body synthesis is also small, although the serum con-
centration of beta-hydroxybutyrate has been measured in
cats with hepatic lipidosis and is increased compared to
healthy cats.10,15,16 However, the rate of fatty acid esterifi-
cation to triglycerides has no limits. Therefore, a marked,
persistently increased mobilization of NEFAs from periph-
eral adipose tissue to the liver may lead to lipid accumu-
lation and storage within hepatocytes.26 Recent work has
confirmed that the predominant lipid that accumulates with-
in the hepatocytes of cats with naturally occurring hepatic
lipidosis is triglyceride, and the source of hepatic triglyc-
eride is mobilization of peripheral adipose stores.27 A block
in 1 of the lipoprotein synthetic or secretion pathways with-
in the liver of cats with hepatic lipidosis is likely, and the
results of the present study indirectly support this hypoth-
esis. Normal metabolism of VLDL in the cat results in pro-
duction of high-density lipoproteins (HDLs), which are rich
in both cholesterol and phospholipid.28 The increased serum
concentrations of both of these lipid fractions in cats with
cholangiohepatitis compared to cats with hepatic lipidosis
may be the result of more effective VLDL synthesis in
cholangiohepatitis cats, leading to relatively increased syn-
thesis of HDL. However, a marked increase also could oc-
cur in peripheral lipolysis with increased delivery of NE-
FAs to the liver secondary to starvation in cats with hepatic
lipidosis. Further work will be necessary to determine the
precise role played in the pathogenesis of feline hepatic
lipidosis by each of these pathways.
Having established that our lipidotic cats had signifi-
cantly elevated serum concentrations of NEFAs, we also
examined some of the hormonal influences on lipolysis.
The release of fatty acids from adipocytes is under the in-
fluence of hormone-sensitive lipase, which catalyzes the hy-
drolysis of triglycerides into NEFAs and glycerol. Hor-
mone-sensitive lipase is tightly regulated by multiple hor-
monal and neural inputs, via the cyclic adenosine mono-
phosphate (cAMP) 2nd messenger system. Insulin
deficiency favors lipolysis by placing hormone-sensitive li-
pase under control of the counterregulatory hormones. Cat-
echolamines, glucocorticoids, thyroxine, growth hormone,
and glucagon all promote lipolysis. Thyroxine sensitizes the
adipocyte to the lipolytic actions of the catecholamines, and
glucocorticoids inhibit carbohydrate metabolism, which po-
tentiates the lipolytic actions of the catecholamines and glu-
cagon. Stimulation of the sympathetic nervous system re-
leases epinephrine, which stimulates lipolysis and suppress-
es insulin release.26
Serum insulin concentration was significantly lower in
cats with liver disease than in control cats, but was not
different between cats with hepatic lipidosis and cats with
cholangiohepatitis. Plasma glucagon concentration was not
different between cats with hepatic lipidosis and the control
cats; however, cats with cholangiohepatitis had significantly
higher plasma glucagon concentrations than controls. The
development of insulin resistance2 or absolute insulin de-
ficiency2,5 have been proposed as possible causes of the
mild fasting hyperglycemia seen in some cats with hepatic
lipidosis, as well as the disorder itself. Abnormal glucose
tolerance tests have been documented in some animals.5
Glucose intolerance has been implicated in the pathogenesis
of human fatty liver syndrome, but is commonly associated
with primary liver diseases other than hepatic lipidosis.26
Based on the current study, insulin deficiency or resistance
and glucose intolerance also seem unlikely to be primary
etiologic factors in feline hepatic lipidosis. Serum insulin
concentrations did not vary with the type of liver disease
present. Furthermore, hyperglycemia was not universally
present in cats with hepatic lipidosis, and was common
among cats with cholangiohepatitis. No association could
be found between the presence of hepatic lipidosis and mild
fasting hyperglycemia, or concentrations of insulin and glu-
cagon in this study. The decreased insulin concentrations
in all diseased cats and the increased glucagon concentra-
tions in cats with cholangiohepatitis might result in in-
creased delivery of NEFAs to the liver in these animals,
but obviously cannot account for the significant increase in
serum NEFA concentrations in cats with hepatic lipidosis.
Clinical, hematologic, and biochemical abnormalities
documented here were comparable to those observed by
other investigators.1,2,4–7,9–13 The apparent female predispos-
tion has been reported.1,2,4 Cats with hepatic lipidosis in the
present study were found to be significantly younger than
cats with cholangiohepatitis; cats with idiopathic hepatic
lipidosis were significantly younger than cats with second-
ary hepatic lipidosis in 1 previous study as well.2 As ex-
pected, serum total bilirubin concentration and SAP, ALT,
and AST activities were significantly higher in both groups
of cats with liver disease compared to control cats.29 Serum
albumin concentrations in cats with lipidosis in this study
were also significantly lower than in control cats, and could
 Metabolic Alterations in Lipidosis
have resulted from severe protein–calorie malnutrition of
any cause. Rates of both albumin and lipoprotein synthesis
are diminished in humans with fatty liver syndrome.26 The
decreased serum albumin concentration observed in the cats
in the current study may also be etiologically important if
it reflects concurrent impairment of hepatic triglyceride re-
lease, due to decreased availability of amino acids for li-
poprotein synthesis. Human fatty liver patients fed protein
show a marked rise in serum VLDL released from the
liver,26 and in 1 study where hepatic lipidosis was experi-
mentally induced in cats by fasting, protein supplementa-
tion reduced the severity of resulting lipidosis.16 Correction
of specific deficiencies in essential amino acids such as me-
thionine or arginine has been suggested to explain this ob-
servation.15,16 Still, it is difficult to explain why anorexia
and weight loss of the same prevalence and duration would
result in nutritional deficiency and hepatic lipidosis among
cats in group 1 in this study, but not in the cats with cho-
langiohepatitis.
Significantly increased hepatic NEFA concentrations
have previously been reported in normal cats fed a com-
pletely taurine-free diet,30 and were attributed to increased
peripheral lipolysis similar to that potentially contributing
to the increased serum NEFA concentrations among the cats
with hepatic lipidosis in the present study. The increased
NEFA concentrations documented in both these studies
suggest that taurine deficiency may also be involved in the
pathogenesis of feline hepatic lipidosis. Taurine is involved
in membrane function and stability, and may affect hepa-
tocellular and hepatocellular organelle membrane func-
tion.30 Malfunction of the endoplasmic reticulum, Golgi
complex, or peroxisome membranes could contribute to the
development of clinical hepatic lipidosis.31 Taurine defi-
ciency is known to occur during starvation in the cat. Sig-
nificant decreases in serum taurine concentration have been
observed in obese but otherwise normal cats undergoing
extended fasts,15 as well as in sick cats with clinical hepatic
lipidosis.10 Urinary losses of conjugated bile acids and tau-
rine are very large in cats with naturally occurring hepatic
lipidosis, although the capacity to conjugate bile acids
seemingly is preserved.32 However, cats with hepatic lipi-
dosis may have insufficient taurine available to maintain
other functions. Although hepatic lipid accumulation in cats
with experimentally induced hepatic lipidosis can be re-
duced by taurine-free protein supplementation,16 taurine
could still be involved in the pathogenesis of hepatic lipi-
dosis. Feeding a mixture of arginine, threonine, methionine,
and cystine has been shown to significantly improve taurine
status in previously deprived cats through increased taurine
synthesis.33
The specific underlying pathogenesis of hepatic lipidosis
in cats remains elusive. Based on the analysis of the data
reported here, we hypothesize that hepatic lipidosis in cats
may be the result of a combination of factors: excessive
peripheral lipid mobilization, possibly due to adrenergic re-
lease during illness or stress, and the subsequent develop-
ment of nutrient deficiencies that compromise the formation
of lipoproteins and the mobilization of hepatic triglycerides
during complete or nearly complete food deprivation. Test-
ing of this hypothesis will require clarification of the role
of individual nutrients, including taurine. The potential in-
25
volvement of hormone-sensitive lipase must also be ex-
plored further. Although no critical alterations could be
found here in the hormone status of cats with hepatic lipi-
dosis compared to cats with cholangiohepatitis, a valid crit-
icism of this study is the small number of cats included,
especially in the cholangiohepatitis group. Further work
with larger numbers of animals is necessary to confirm our
result and investigate other hormones, particularly the cat-
echolamines. Plasma lipids in cats with hepatic lipidosis
should also be assessed in more detail, using ultracentri-
fugation and immunoelectrophoretic techniques. Finally, if
sympathetic input proves to be a major contributor to pe-
ripheral lipolysis in these cats, adrenergic-blocking drugs
could be evaluated for their ability to reduce peripheral fat
mobilization by direct interference with catecholamine re-
lease.26
Footnotes
a
Serono-Baker 9000 Hematology Analyzer, Serono-Baker Diagnos-
tics, Inc, Allentown, PA.
b
Hitachi 717 Serum Biochemical Analyzer, Boehringer Mannheim Di-
agnostics Division of Boehringer Mannheim Corp, Indianapolis, IN.
c
ViraCHEK/FeLV Feline Leukemia Virus Antigen Test Kit, Synbiotics
Corp, San Diego, CA.
d
PetChek FIV Ab Feline Immunodeficiency Virus Antibody Test Kit,
IDEXX Laboratories, Inc, Westbrook, ME.
e
Glucagon Double Antibody, Diagnostic Products Corporation, Los
Angeles, CA.
f
Coat-A-Count Insulin Kit, Diagnostic Products Corporation, Los An-
geles, CA.
g
Coat-A-Count Cortisol Kit, Diagnostic Products Corporation, Los
Angeles, CA.
h
Coat-A-Count Total T4 Kit, Diagnostic Products Corporation, Los
Angeles, CA.
i
Boehringer Mannheim Diagnostics Division of Boehringer Mann-
heim Corp, Indianapolis, IN.
j
Wako BioProducts, Richmond, VA.
k
Sigma Diagnostics, St Louis, MO.
l
Cobas Fera2 Automated Analyzer, F. Hoffmann-La Roche Ltd, Di-
agnostics Division, Basel, Switzerland.
m
SigmaStat Version 2.0 for Windows 95, Copyright 1997, SPSS Inc,
Chicago, IL.
Acknowledgments
We gratefully acknowledge Samantha Mooney, MA, for
her assistance in preparing the manuscript and Dr Dan Le-
vine of the Rogosin Institute, New York, NY, for technical
assistance.
References
1. Barsanti JA, Jones BD, Spano JS, et al. Prolonged anorexia as-
sociated with hepatic lipidosis in three cats. Feline Pract 1977;7:52–
57.
2. Center SA, Crawford MA, Guida L, et al. A retrospective study
of 77 cats with severe hepatic lipidosis: 1975–1990. J Vet Intern Med
1993;7:349–359.
3. Akol KG, Washabau RJ, Saunders HM, et al. Acute pancreatitis
in cats with hepatic lipidosis. J Vet Intern Med 1993;7:205–209.
4. Burrows CF, Chiapella AM, Jezyk P. Idiopathic feline hepatic
26 Brown et al
lipidosis: The syndrome and speculations on its pathogenesis. Fla Vet
J 1981;Winter:18–20.
5. Zawie D, Garvey M. Feline hepatic disease. Vet Clin North Am
1984;14:1201–1230.
6. Armstrong PJ. Feline hepatic lipidosis. 7th Annual ACVIM Fo-
rum, San Diego, CA, May 1989.
7. Dimski DS. Feline hepatic lipidosis. Semin Vet Med Surg 1997;
12:28–33.
8. Kristensen AT, Weiss DJ, Klausner J, et al. Liver cytology in
cases of canine and feline hepatic disease. Compend Cont Educ Pract
Vet 1990;12:797–808.
9. Hitt ME, Jones BD, Constantinescu G. The feline liver: Identi-
fying and treating its diseases. Vet Med 1987/b:139–147.
10. Center SA. Hepatic ultrastructural and metabolic derangements
in cats with severe hepatic lipidosis. 9th Annual ACVIM Forum, New
Orleans, LA, May 1991.
11. Jacobs G, Cornelius L, Allen S, et al. Treatment of idiopathic
hepatic lipidosis in cats: 11 cases (1986–1987). J Am Vet Med Assoc
1989;195:635–638.
12. Biourge VC. Nutrition and liver disease. Semin Vet Med Surg
1997;12:34–44.
13. Biourge VC, Pion P, Lewis J, et al. Spontaneous occurrence of
hepatic lipidosis in a group of laboratory cats. J Vet Intern Med 1993;
7:194–197.
14. Biourge VC, Groff JM, Munn RJ, et al. Experimental induction
of hepatic lipidosis in cats. Am J Vet Res 1994;55:1291–1302.
15. Biourge VC, Groff JM, Fisher C, et al. Nitrogen balance, plas-
ma free amino acid concentrations and urinary orotic acid excretion
during long-term fasting in cats. J Nutr 1994;124:1094–1103.
16. Biourge VC, Massat B, Groff JM, et al. Effects of protein, lipid,
or carbohydrate supplementation on hepatic lipid accumulation during
rapid weight loss in obese cats. Am J Vet Res 1994;55:1406–1415.
17. Biourge VC, Nelson RW, Feldman EC, et al. Effect of weight
gain and subsequent weight loss on glucose tolerance and insulin re-
sponse in healthy cats. J Vet Intern Med 1997;11:86–91.
18. Hayes KC. Inosital as part of therapy for feline hepatic lipi-
dosis. Viewpoints Vet Med 1992;2(2):1–4.
19. Rogers KS, Cornelius LM. Feline icterus. Compend Cont Educ
Pract Vet 1985;7:391–399.
20. Biourge VC, Pion P, Lewis J, et al. Dietary management of
idiopathic feline hepatic lipidosis with a liquid diet supplemented with
citrulline and choline. J Nutr 1991;21:S155-S156.
21. Frank A, Feinstein RE. Hepatic lipidosis associated wtih severe
vitamin B12 deficiency recognized by liver cobalt status in three cats.
Feline Pract 1991;19(5):16–20.
22. Diagnostic Products Corporation. Product Insert. Glucagon
Double Antibody Radioimmunoassay Kit. Los Angeles, CA: Diagnos-
tic Products Corporation; 1998:5.
23. Nelson RW, Himsel CA, Feldman EC, et al. Glucose tolerance
and insulin response in normal-weight and obese cats. Am J Vet Res
1990;51:1357–1362.
24. Kemppainen RJ, Mansfield PD, Sartin JL. Endocrine responses
of normal cats to TSH and synthetic ACTH administration. J Am
Anim Hosp Assoc 1984;20:737–740.
25. Peterson ME, Broussard JD, Gamble DA. Use of the thyrotro-
pin releasing hormone stimulation test to diagnose mild hyperthyroid-
ism in cats. J Vet Intern Med 1994;8:279–286.
26. Alpers DH, Sabesin SM, White HM. Fatty liver: Biochemical
and clinical aspects. In: Schiff L, Schiff ER, eds. Diseases of the Liver.
Philadelphia, PA: Lippincott; 1993:825–855.
27. Hall JA, Barstad LA, Connor WE. Lipid composition of hepatic
and adipose tissues from normal cats and from cats with idiopathic
hepatic lipidosis. J Vet Intern Med 1997;11:238–242.
28. Demacker PNM, van Heijst PJ, Hak-Lemmers HLM, et al. A
study of the lipid transport system in the cat, Felix domesticus. Ath-
erosclerosis 1987;66:113–123.
29. Center SA, Baldwin BH, Dillingham S, et al. Diagnostic value
of serum gamma-glutamyl transferase and alkaline phosphatase activ-
ities in hepatobiliary disease in the cat. J Am Vet Med Assoc 1986;
188:507–510.
30. Cantafora A, Blotta I, Rossi SS, et al. Dietary taurine content
changes liver lipids in cats. J Nutr 1991;121:1522–1528.
31. Center SA, Guida L, Zanelli MJ, et al. Ultrastructural hepato-
cellular features associated with severe hepatic lipidosis in cats. Am J
Vet Res 1993;54:724–731.
32. Center SA, Thompson M, Guida L. Three alpha-hydroxylated
bile acid profiles in clinically normal cats, cats with severe hepatic
lipidosis, and cats with complete extrahepatic bile duct occlusion. Am
J Vet Res 1993;54:681–688.
33. Trautwein EA, Hayes KC. Gender and dietary amino acid sup-
plementation influence the plasma and whole blood taurine status of
taurine-depleted cats. J Nutr 1991;121:S173-S174.