Beneficial effects of chronic administration of dietary

c0-3 polyunsaturated fatty acids in dogs with renal

insufficiency

SCOTT A. BROWN, CATHY A. BROWN, WAYNE A. CROWELL, JEANNE A. BARSANTI,TIMOTHY

ALLEN, CHRISTOPHER COWELL, and DELMAR R. FINCO
ATHENS,GEORGIA, and TOPEKA, KANSAS

Dietary supplementation with polyunsaturated fatty acids (PUFA) alters the course
of experimental renal disease in rats. However, chronic renal disease in other lab-
oratory animals and in human beings frequently responds differently to experi-
mental manipulations. We investigated the effects of variations in dietary PUFA
composition on the chronic course of induced renal disease in dogs. Two months
after 15/16 nephrectomy, dogs were randomly divided into three groups of seven
animals each. For the next 20 months, each group of dogs was fed a low-fat basal
diet supplemented with one of three sources of lipid to achieve a final concentra-
tion of 15% a d d e d fat. Fat sources provided o)-3 PUFA (menhaden fish oil, group
FO), 0)-6 PUFA (safflower oil, group SO), or saturated fatty acids (beef tallow, group
BT). Throughout the dietary trial, the magnitude of proteinuria and the plasma con-
centrations of creatinine, cholesterol, and triglyceride were lower in group FO. The
mean overall glomerular filtration rate was 0.89 +_0.18 ml/min per kilogram of body
weight in group SO, a value that was significantly less (p < 0.05) than the corre-
sponding values for groups BT and FO (1.21 _+ 0.18 and 1.43 _+ 0.20 ml/min/kg,
respectively). Renal interstitial fibrosis also was significantly elevated in group SO.
The extents of mesangial matrix expansion, glomerulosclerosis, and renal intersti-
tial cellular infiltrate were similar in groups BT and SO, but lower (p < 0.05) in group
FO. We conclude that supplementation with o)-6 PUFA enhanced renal injury; sup-
plementation with ~-3 PUFA was renoprotective. (J Clin Lab Med 1998;131:447-55)

Abbreviations: BT= Beef tallow; BUN = blood urea nitrogen; CCr = urinary clearance of exoge-
nously administered creatinine; FO = menhaden fish oiL; GFR = glomerular filtration rate; HDL-
cholesterol = cholesterol contained in high-density lipoprotein particles; PAS= periodic acid-
Schiff; PUFA: polyunsaturated fatty acids; SCr = serum creatinine concentration; SO = saf-
flower oil

~ t has been suggested that dietary fatty acid compo-

sition modifies the course of induced renal disease
in rats. 1-3 However, results in rodent models are
renoprotection; in others, 2,4 fish oil supplementation
was associated with worsening glomerulosclerosis or
decrements in GFR, or both. Results of studies in
contradictory. In some laboratory studies,i, 3 dietary human beings are similarly contradictory, with both
supplementation with a source of 0)-3 PUFA led to benefits 5 and no effect6, 7 being reported.
In dogs, marked reduction of renal mass is followed
From the Department of Physiology and Pharmacology and the by proteinuria, glomerulosclerosis, tubulointerstitial
Department of Pathology,College of Veterinary Medicine, Universi- damage, and progressive loss of renal function. 8-11
ty of Georgia,Athens, and Mark Morris Associates, Topeka.
Although the precise causes of this structural and func-
Supported by the Morris Animal Foundation.
tional deterioration of the kidney remain poorly eluci-
Submitted for publication March 13, 1997; revision submitted
November 3, 1997; accepted December 5, 1997. dated, this model has been used to study progressive
Reprint requests: Dr. Scott A. Brown, Department of Physiology,Col- renal injury 8-13 and the effects of nutrients on that
lege of Veterinary Medicine, University of Georgia, Athens, GA process.S,9,11,14,15 In particular, the long-term response
30602. of dogs with reduced renal mass to dietary manipula-
Copyright © 1998 by Mosby, Inc. tions 9-11 differs considerably from that of inbred strains
0022-2143/98 $5.00 + 0 5/1/88744 of rodents.

447
 J Lab Clin Med
448 Brown et al. May 1998

The purpose of the current study was to compare the erides, urine protein-to-creatinineratio, and CCr 16were deter-
effects of dietary supplementation with saturated fatty mined. The dogs~were randomly assigned to three groups of
acids, 0)-6 PUFA, and 0)-3 PUFA on the chronic course seven animals each on the basis of gender and the measured
of induced renal disease in dogs. values for CCr and were fed the experimental diet supple-
mented with lipids as described earlier. SCr, BUN, choles-
terol, HDL-cholesterol, triglycerides, total protein, albumin,
METHODS
and hematocrit were determined after a 12- to 16-hour fast
Animals. Experiments were performed on 21 adult mixed- once a month. Every 4 months during the dietary trial, CCr
breed dogs of either sex weighing 11.0 + 0.7 kg. The dogs and the urine protein-to-creatinine ratio were determined as
were individually maintained in runs in an environmentally previously described. 16 After the determinations performed
controlled facility. All dogs had renal mass reduced by right at 20 months, the dogs were killed. During the long-term
nephrectomy and infarction of 7/8 of the left kidney, as pre- dietary trial, any dog that developed terminal renal failure--
viously described, 9 with both procedures performed simulta- defined as CCr lower than 0.2 ml/min/kg in the absence of
neously. A portion of the right kidney was preserved in 10% dehydration or other intercurrent disease--was killed.
neutral buffered formalin solution at the time of partial SCr, BUN, total protein, and albumin were measured by a
nephrectomy. All research was conducted in accordance with semiautomated method (Impact 400, Gilford Instruments,
the National Institutes of Health Guide for Care and Use of Oberlin, OH). HDL-cholesterol was determined with the aid
Laboratory Animals after approval by the Institutional Ani- of an HDL precipitating reagent (Sigma Chemical Company,
mal Care Committee. St. Louis, Mo.). Serum concentrations of cholesterol and
Diets. Throughout all studies, each dog was fed a preweighed triglyceride were determined by enzymatic methods (Stanbio
amount of food each morning. The amount of food remaining Laboratories, San Antonio, Tex.). Whole kidney CCr was cal-
the next morning was determined. Initially, each dog was fed culated by the standard clearance formula. Urine protein con-
132 kcal/kg0.75. Subsequently, quantities of food intake were centration was determined by the Coomassie brilliant blue
adjusted on the basis of monthly weights, with a goal of main- technique. 17
taining stable body weight. For 2 months after renal mass Histology. After each dog was killed, the remnant renal
reduction, all dogs were fed a canine maintenance chow (Hill's tissue was removed, excised, stripped of surrounding tissue
Science Diet, Hill's Petfoods, Topeka, KS) to allow for estab- and capsule, and blotted dry. The scar from the area previ-
lishment of stable, chronic renal failure before dietary trials. ously infarcted was removed from each remnant kidney by
The experimental diet (Mark Morris Associates, Topeka, KS) sharp dissection under stereoscopic magnification, and the
contained 20.8% protein, 1.8% fat, 0.3% sodium, 0.8% potas- viable portion of the remnant was weighed. A single 2- to 3-
sium, 0.9% calcium, and 0.4% phosphorus on a dry matter mm thick midsagittal section was placed in 10% buffered for-
basis (analysis by Woodson-Tenent Laboratories, Nashville, malin solution for subsequent processing and examination by
Tenn.). Immediately before each feeding, a quantity of lipid light microscopy. Formalin-fixed tissue was processed by rou-
equal to 15% of daily intake on an as-fed weight basis was tine histologic methods, and sections were stained with hema-
added to each dog's ration. The final offered combination of toxylin and PAS dyes. For glomerular morphologic analyses,
diet plus lipid, on a dry matter basis, contained 16.8% fat, only glomeruli from the outer third of the cortex were evalu-
17.7% protein, 0.3% sodium, 0.7% potassium, 0.8% calcium, ated. A minimum of 25 outer cortical glomeruli were exam-
0.4% phosphorus. Vitamin E, 5 IU per kilogram of body ined by the pathologist (C. A. Brown) without knowledge of
weight, was added to the diet of each dog on a dally basis. the study group of origin and scored for the presence of
The lipids were stored at -20 ° C until 12 hours before feed- mesangial matrix expansion, using a numerical scoring sys-
ing and were stored at 4 ° C overnight before feeding; the two tem previously described 18 (0 = normal; 1 = minimal expan-
oils were stored in preweighed vials under nitrogen. The fat sion; 2 = moderate expansion; 3 = severe, obllterative expan-
sources were as follows: 0)-3 PUFA (group FO), menhaden fish
sion). The prevalence of glomerulosclerosis was quantified
oil (Zapata-Haynie Menhaden Oil Refinery, Reedville, Va.); 0)- as the number of glomeruli exhibiting segmental or global
6 PUFA (group SO), safflower oil (Oil Seeds International, San glomerulosclerosis divided by the number of glomeruli exam-
Francisco, Ca.), and saturated fatty acids (group BT), edible ined, expressed as a percentage. Tubular lesions, interstitial
beef tallow (Addison Foods, Dallas, Tx.). The fatty acid pro- fibrosis, interstitial inflammatory cell infiltrate, and glomeru-
file of the added lipids for group FO averaged 13.6% eicos- lar cellularity were scored by a similar qualitative scoring sys-
apentaenoic acid, 12.5% docosahexaenoic acid, 1.2% linoleic tem (0 = no change; 1 = mild change; 2 = moderate change;
acid, 33.4% monounsaturated fatty acids, and 30.6% saturated and 3 = severe change from normal).
fatty acids. For group SO it was 77.7% linoleic acid, 13.0% Planar area of 20 randomly selected outer cortical glomeru-
monounsaturated fatty acids, and 9.9% saturated fatty acids, lar capillary tufts was measured in PAS-stained, formalin-fixed
and for group BT it was 4.3% linoleic acid, 47.1% monoun- sections of renal tissue obtained at the time of nephrectomy
saturated fatty acids, and 48.1% saturated fatty acids (analysis (initial) and at the time of the renal hemodynamic studies
by Woodson-Tenent Laboratories). The combination of beef
(final) with the aid of a planar morphometry image analysis
tallow plus basal diet fed to group BT contained 1.05% linole- system, as previously described. 12 Glomerular volumes were
ic acid and 0.1% linolenic acid on a dry weight basis. calculated by the following formula:
Study protocol. Two months after partial nephrectomy in
21 dogs, the SCr, BUN, cholesterol, HDL-cholesterol, triglyc- V= ~/K (area 3/e)
J Lab CIin Med
Volume 131, Number 5 Brown et al, 449

where ~ = 1.38 (the shape coefficient for spheres), ~ = 1.1 tion of 2.84 _ 0.10 gm/dl were significantly different
(the size distribution coefficient for glomerular profiles), and from the corresponding values for the other two dietary
area = the mean glomemlar profile planar area.12 groups. The mean overall value for hematocrit for
Statistics. Values are reported as means + SEM. Data were group SO (see Table II) was less than the values in the
compared among groups by analysis of variance, with repeat- other two groups (p < 0.05). In group SO, the final
ed measures analysis of variance models used where appro- value for hematocrit in those dogs that survived to the
priate. Where a significant effect was identified, group means
end of the study (n = 3; 40.7% -+ 1.2%) was less (p <
were compared by the Fisher protected least signifigant dif-
0.01) than in those that reached end-stage renal disease
ference comparison test. A probability value lower than 0.05
was considered indicative of a statistically significant differ- during the dietary trial (n = 4; 25.9% -+ 12.5%).
ence. Renal function. At the time of division into dietary
groups, there were no differences among groups for
RESULTS CCr, SCr, BUN, or urine protein-to-creatinine ratio (see
Division into dietary groups. A t the time of random- Table I). The values for CCr were approximately 70%
ized division into dietary groups, there were no differ- less (p < 0.05) than the reference value for CCr in nor-
ences among groups for CCr, SCr, BUN, hematocrit, mal dogs in our laboratory of 4.08 _ 0.50 ml/min/kg. 19
urine protein-to-creatinine ratio, or serum concentra- The baseline values for SCr and BUN at 2 months after
tions of triglyceride, cholesterol, albumin, or total pro- nephrectomy (see Table I) were similar among groups
tein. There were four male and three female dogs in and exceeded (p < 0.05) the mean prenephrectomy val-
each group. ues of 0.81 +- 0.03 mg/dl for SCr and 14.9 _ 1.2 mg/dl
Dietary intake. During the dietary trial the mean for BUN. All three groups had a similar significant (p
dietary intakes were not significantly different among < 0.05) increase in the urine protein-to-creatinine ratio
groups. There was no difference in mean daily intake at the beginning of the dietary trial compared with the
of protein, which was 2.89 _ 0.16 gm/kg body weight mean prenephrectomy value of 0.11 _ 0.01. Although
per day for group FO, 2.94 _ 0.21 for group BT, and the magnitude of proteinuria was similar in all three
2.76 _ 0.18 for group SO. Mean daily intakes of sodi- groups at the time of initiation of the dietary trial, it
um, phosphorus, calcium, and potassium during the increased progressively in groups SO and BT, and the
dietary trial were not different among groups. mean overall value for these groups exceeded the cor-
The mean daily intakes of fat were 2.9 -+ 0.1 gm/kg responding value for group FO (p < 0.05; see Table II).
body weight per day for group FO, 2.9 _+0.2 for group Throughout the dietary trial, mean CCr was consistent-
BT, and 2.8 - 0.2 for group SO. The mean daily intakes ly higher in group FO. Renal function declined progres-
for saturated fatty acids were statistically unique (p < sively in group SO, and at 12 months the mean CCr for
0.05) among the three groups and were 0.49 +- 0.03 group SO was significantly (p < 0.05) lower than the CCr
gm/kg body weight per day for group FO, 1.20 + 0.08 for either of the other groups. The mean final value for
for group BT, and 0.23 -+ 0.01 for group SO. The mean CCr (Fig. 1) for group SO (0.59 _ 0.12 ml/min/kg) rep-
daily intakes for vitamin E were 4.6 _ 0.4 IU/kg body resented a mean change o f - 4 3 % _+ 16.9% from the ini-
weight/day for group FO, 4.9 _+0.4 for group BT, and tial value. The final value for group FO (1.29 _+ 0.18
4.8 +_0.3 for group SO. ml/min/kg) was significantly higher than for groups SO
General. Body weights were not significantly differ- and BT and represented an increase of 11.4 +_7.5% dur-
ent among groups at any time during the dietary trial. ing the 20-month dietary trial. The final value for CCr in
The initial mean body weight, at the time of division group BT (0.85 _ 0.09 ml/min/kg was intermediate and
into dietary groups, was 11.4 _+ 1.1 kg for group SO, represented a mean change during the trial of-14.9% +
9.9 + 0.9 for group BT, and 12.3 +- 1.4 for group FO. 15.9%. Overall CCr declined in six dogs (86%) in group
There was a small increase (p < 0.05) in mean body SO, five dogs (71%) in group BT, and two dogs (29%) in
weight during the 20-month dietary trial that averaged group FO. As a result of these differences in the pattern
4.0%. The final mean body weight was 11.7 +_ 1.5 kg of change in renal function over time, progression to end-
for group SO, 10.2 _+0.8 for group BT, and 13.1 _+ 1.3 stage renal disease was significantly hastened in group
kg for group FO. SO compared with group FO (p < 0.05; Fig. 2). All deaths
Plasma concentrations of total protein and albumin were caused by euthanasia preceded by progressive
and hematocrit were not different among groups at the decrements in CCr and the development of end-stage
beginning of the dietary trial (Table I). Although there renal failure. The mean overall values for CCr and BUN
was a trend for plasma albumin concentration to (see Table I) reflected the differences in CCr between
decrease mildly during the dietary trial in group SO groups.
(Table II), neither the mean overall value for this pa- Plasma lipids. Compared with the prenephrectomy
rameter nor the mean final plasma albumin concentra- vg.ue of 123.1 _ 6.0 mg/dl, plasma cholesterol concen-
 J Lab Clin M e d
450 Brown e t al. M a y 1998

T a b l e I. S y s t e m i c a n d r e n a l p a r a m e t e r s a t i n i t i a t i o n o f t h e d i e t a r y trial, 2 m o n t h s a f t e r p a r t i a l n e p h r e c t o m y ,
by dietary group (mean + SEM)
Parameter Group SO Group BT Group FO

GFR (ml/min/kg) 1.24 4- 0.17 1.21 4- 0.21 1.22 4- 0.20

SCr (mg/dl) 3.0 4- 0.3 3.2 4- 0.3 3.0 4- 0.4
BUN (mg/dl) 50.0 4- 9.6 61.0 4- 9.5 59.5 4- 10.9
Urine protein-to-creatinine ratio 0.31 4- 0.07 0.24 4- 0.03 0.24 4- 0.02
Triglycerides (mg/dl) 52.8 + 5.6 45,4 4- 5.8 45.3 4- 5.7
Total cholesterol (mg/dl) 211 4- 18 222 4- 23 215 4- 16
HDL-cholesterol (mg/dl) 156 4- 9 168 4- 11 155 4- 7
Ratio of HDL-cholesterol to total cholesterol 0.75 4- 0.03 0.77 _+0.02 0.73 4- 0.04
Albumin (gm/dl) 3,01 4- 0.09 3.03 4- 0.06 3.13 4- 0.11
Total protein (gm/dl) 6,96 _+ 0.33 6.94 4- 0,36 6.96 4- 0.23
Hematocrit (%) 44.6 4- 1.8 45.4 4- 1.2 45.3 4- 1.5

T a b l e II. O v e r a l l v a l u e s for s y s t e m i c a n d r e n a l p a r a m e t e r s during the 20-month d i e t a r y trial, b y d i e t a r y g r o u p

(mean + SEM)
Parameter Group SO Group BT Group FO

G F R (rnl/min/kg) 0.89 ± 0.18 1.21 _+ 0.18" 1.43 __-0.20*

SCr (mg/dl) 3.48 _+ 0.74 3.01 4- 0.37* 2.29 4- 0.27"t
BUN (mg/dl) 68.9 4- 11.1 41.54-5.1" 37.4 4- 4.5*
Urine protein-to-creatinine ratio 2.55 4- 0.79 1.76 4- 0.68 0.60 4- 0.24"t
Triglyceride (mg/dl) 40.8 _+5.0 44.7 4- 5.6 32,5 4- 3.7"t
Total cholesterol (mg/dl) 297 _+32 298 4- 20 227 4- 31"t
HDL-cholesterol (mg/dl) 215 _+ 16 224 ± 12 153 4- 10*t
Ratio of HDL-cholesterol to total cholesterol 0.75 4- 0.04 0.76 _+0.03 0.72 _+0.06
Total protein (gm/dl) 6.56 4- 0.22 6.83 4- 0.22 6.85 4- 0.13
Albumin (gm/dl) 2.93 _+0107 3.02 4- 0.05 3.08 4- 0.07
Hematocrit (%) 37.4 4- 2.0 42.7 4- 2.1" 44.4 4- 0.9*

All measurements o b t a i n e d in e a c h d o g were a v e r a g e d to determine an overall value for e a c h dog.

*p < 0,05 versus group SO.
t p < 0.05 versus group BT.

tration was significantly (p < 0.05) elevated throughout throughout the dietary trial.
the dietary trial (see Table II). Two months after insti- Compared with the prenephrectomy value of 43.9 +
tution of the dietary change, the plasma cholesterol con- 2.5 mg/dl, there was no significant change in plasma
centration was 236.1 +- 18.7 mg/dl in group SO, 268.2 triglyceride concentration 2 months after partial renal
_+ 18.5 in group BT, and 202.3 _+ 23.2 in group FO. ablation (see Table I). However, there was a significant
Although this represented a trend for plasma choles- decrement in plasma triglyceride concentration in
terol values (group BT > group SO > group FO), these group FO during the dietary trial.
differences were significant (p < 0.05) only for group Renal m o r p h o l o g y . The mean final kidney weights
FO versus group BT. Within group SO, there was an were not significantly different among groups, averag-
inverse correlation (R = 0.52, p < 0.05) between CCr ing 1.40 _ 0.10 gm/kg body weight for group SO, 1.55
and plasma cholesterol concentration, which tended to ___0.13 for group BT, and 1.57 + 0.16 for group FO.
rise as CCr progressively declined. The mean overall Histologic studies of tissue obtained at the time of
plasma cholesterol concentration during the dietary trial nephrectomy and at the end of the dietary trial were
was significantly lower in group FO (see Table II) than evaluated to determine the effect of treatment group and
in the other two groups. Plasma HDL-cholesterol was the combined effect of time and partial nephrectomy
also significantly lower in group FO. However, there on the extent of glomerular and tubulointerstitial
was no change in any group in the ratio of HDL-cho- lesions. There was a significant (p < 0.001) difference
lesterol to total cholesterol, which remained similar to between the histologic scores for prenephrectomy tis-
the overall mean prenephrectomy value of 0.72 _+0.02 sue (initial) and those for renal tissue obtained at the
J Lab Clin Med
Volume 131, Number 5 Brown et al. 451

1.6-

90
1.4

~0-
1.2

70'

1.0" r./l 60"

-- Group FO I
• Group BT
0.8 50 " A Group SO

40 ~
*t o lO 2o
0.6
10 20 Time (months)
Time (months)
Fig. 1. Serial values for CCr in partially n e p h r e c t o m i z e d dogs fed Fig. 2. Percentsurvival in partially nephrectomized dogs fed a basal
basal diet s u p p l e m e n t e d with 15% fish oil (group FO, n = 7), b e e f diet supplemented with 15% fish oil (group FO, n = 7), b e e f tallow
tallow (group BT, n = 7), or safflower oil (group SO, n = 7). Values (group BT, n = 7), or safflower oil (group SO, n = 7). *p < 0.05 ver-
are means _+ SEM. *p < 0.05 versus group FO for same time point. sus group FO.
tp < 0.05 versus group BT for same time point.

DISCUSSION

end of the study (final). These respective scores were After reduction of renal mass, a pattern of declining
glomerular mesangial expansion, 0.13 -+ 0.04 versus renal function developed in six of seven dogs main-
1.58 +__0.17; interstitial fibrosis, 0.02 _+0.02 versus 0.52 tained on a diet enriched with c0-6 PUFA. This progres-
_+ 0.09; cellular infiltrate, 0.02 _+ 0.02 versus 0.62 _+ sive deterioration of renal function was associated with
0.10; and tubular lesions, 0.00 _+ 0.00 versus 0.45 -+ proteinuria, hypercholesterolemia, morphologic evi-
0.09. There was a significant effect of treatment group dence of glomerular and tubulointerstitial injury, and
on histologic severity of glomerular injury, with the an increased prevalence of end-stage renal failure in
glomerular score significantly higher (p < 0.05; Fig. 3) this group. Dietary supplementation with BT, a source
in groups SO and BT than in group FO. The prevalence of saturated fatty acids, was also associated with evi-
of glomerulosclerosis was significantly lower (p < 0.05; dence of progressive decrements of renal function and
see Fig. 3) in group FO than in groups SO and BT. structure. However, the rate of decay of renal function
There was a mild increase in glomerular cellularity in was slower in this group, and fewer dogs achieved end-
all three groups, with a nonsignificant (p = 0.12) trend stage renal failure. Progression to end-stage renal dis-
for reduced cellularity in group FO (mean score 0.7 _+ ease in this group was also associated with proteinuria,
0.2, versus 1.0 -+ 0.0 for groups SO and BT). Renal hypercholesterolemia, and morphologic evidence of
tubulointerstitial fibrosis was more severe (p < 0.05; glomerular and tubulointerstitial injury.
Fig. 4) in group SO than in group FO, and renal inter- In contrast, dogs fed a diet supplemented with fish
stitial cellular infiltrate, composed primarily lympho- oil, a rich source of o)-3 PUFA, did not exhibit progres-
cytes and plasma cells with occasional macrophages, sive decrements of renal function, and the final values
was more severe (p < 0.05; see Fig. 4) in groups BT for CCr and SCr actually reflected an increase in renal
and SO than in group FO. Although there was a trend function during the 20-month trial. These results indi-
for preservation of normal morphology of renal tubules cate that a o-3 PUFA-enriched diet may be renoprotec-
in group FO, this was not statistically significant (p = tive. In fact, the findings of stable or increasing GFR,
0.17). minimal proteinuria, and absence of end-stage renal
There was a significant (p < 0.05) increase in failure in this group of markedly nephrectomized dogs
glomerular volume between the mean initial and mean contrast with our previous results in dogs fed a low-
final values: group FO, 3.64 _+0.49 versus 6.57 + 0.91; protein, low-phosphorus diet for 24 months after 15/16
group BT, 3.73 + 0.40 versus 6.76 + 0.70; and group nephrectomy. 9 Dogs in that study were fed a diet that
SO, 3.49 +_0.34 versus 6.53 +_0.26 x 106gm 3. Howev- contained 16.7% protein, 12% fat, and 0.44% phospho-
er, there were no significant differences among the rus on a dry matter basis, similar to the diet used in the
groups in mean final values for glomerular volume. present study. In the previous study, proteinuria (mean
 J Lab Clin Med
452 Brown et al. May 1998

1.2
80
I •[::] Grouu
Group FO]
BT |

60
2 0.8

il 20
0,4

0 0
Mesangial expansion GlomeruIosclerosis 0.0
fibrosis interstitial tubular
cellular infiltrate les~ons
Fig. 3. Glomerular lesions in partially nephrectomized dogs fed a
basal diet supplemented with 15% fish oil (group FO), beef tallow Fig. 4. Renal structural lesions in partially nephrectomized dogs fed
(group BT), or safflower oil (group SO). Results of histologic scores a basal diet supplemented with 15% fish oil (group FO), beef tallow
for mesangial matrix expansion and prevalence (%) of glomeru- (group BT), or safflower oil (group SO). Histologic scores for renal
losclerosis at the end of the dietary trial. *p < 0.05 versus group FO tissue at the end of the dietary trial. *p < 0.05 versus group FO for
for same parameter. same parameter.

protein-to-creatinine ratio, 3.12 _+ 0.44), progression laboratory rodents have suggested that fish oil supple-
(mean rate of change of renal function, -2.6% ___1.1%), mentation is of no benefit 32 or actually enhances renal
and end-stage renal failure (25% prevalence) were evi- injury.2, 4 The reasons for these disparate results are
dent. However, dogs fed the m-3 PUFA-enriched diet unclear, although it has been suggested 3 that the incor-
in the present study had minimal proteinuria (mean pro- poration of antioxidants may be critical when PUFA-
tein-to-creatinine ratio, 0.60 _+0.24) and a mean over- rich diets are used in long-term studies. Increased lipid
all increase in renal function (+11.4% __.7.5%), and no peroxidation has been observed in rats exhibiting
dogs of this group developed end-stage renal failure. In adverse effects while being fed fish oil enriched diets. 4
addition to preservation of renal function, the magni- When dietary antioxidants were used, 3 0)-3 PUFA sup-
tudes of glomerular and tubulointerstitial lesions were plementation was beneficial to rats with renal disease.
dramatically reduced, further demonstrating the reno- In the present study, both PUFA-rich oils were stored
protective effect of dietary 0}-3 PUFA supplementation under nitrogen at - 2 0 ° C (or at 4 ° C for the last 12
in this study. hours) before use to minimize lipid peroxidation in the
These results are consistent with those of studies diet. Dietary supplementation with approximately 5 IU
demonstrating preservation of renal function and/or vitamin E per kilogram of body weight per day, or
structure in rats after reduction of renal mass. 1,3 Bene- approximately 2.5 mg a-tocopherol per gram of PUFA
ficial effects of dietary fish oil supplementation have in group SO, was used to limit lipid peroxidation of
also been observed in dogs 20 and rats 21 with experi- PUFA once these fatty acids became incorporated into
mentally induced acute renal injury, in various other plasma membranes. This intake of vitamin E exceeds
models of chronic renal disease in rats, 22-30 and in the National Research Council minimum requirements
some 5 but not all 6,7 studies of renal disease in humans. of 0.5 IU/kg/day and more than 0.5 mg of a-tocopherol
In partially nephrectomized rats, two dietary maneu- per gram of PUFA 33 and may have lessened the amount
vers, fish oil supplementation and protein restriction, of lipid peroxidation of PUFA in vivo.
have been shown to be beneficial. 31 In our study, fish Compared with the other groups, dogs in group SO
oil supplementation lessened the severity of both exhibited a significantly enhanced rate of progression
glomerular and tubulointerstitial lesions. Some studies and a trend for more proteinuria, reduced survival, and
in rodents have also reported less glomerular22, 31 or more severe renal structural injury. Although the ane-
tubulointerstitia129, 30 injury associated with dietary fish mia observed in this group of dogs may have been
oil supplementation. caused by a lack of erythropoietin, oxidant damage to
Results of some studies of models of renal disease in circulating erythrocytes, or other contributing factors,
J Lab Clin Med
Volume 131, Number 5 Brown et al. 453

we did not determine the cause of this abnormality. Other mechanisms for renoprotection by supplemen-
Taken in concert, these results suggest that dietary saf- tation with dietary fish oil, such as an antioxidant
flower oil supplementation may be nephrotoxic. If these actionSO,56 or an effect to limit intrarenal calcifica-
findings can be extended to spontaneous renal disease, tion, 29 have been suggested. A mechanism for the
dietary supplementation with similar levels of 0)-6 effects observed in the present study of partially
PUFA in an attempt to increase PUFA intake may prove nephrectomized dogs was not determined.
undesirable. In summary, progressive deterioration of renal func-
Several factors may have contributed to observed tion associated with proteinuria, hypercholesterolemia,
results in the present study. One hypothesis for reno- morphologic evidence of glomerular and tubulointer-
protection by 03-3 PUFA-supplemented diets is their stitial injury, and an increased prevalence of end-stage
tendency to reduce plasma concentrations of choles- renal failure was observed in partially nephrectomized
terol and triglycerides, 34 which could be renoprotec- dogs fed an 60-6 PUFA-enriched diet. In contrast,
tive. 35-38 Dogs in group FO did exhibit a reduction in dietary supplementation with 0)-3 PUFA was renopro-
plasma cholesterol concentration but no change in the tective, preventing deterioration of renal function and
ratio of HDL-cholesterol to total cholesterol. It seems preserving renal structure.
likely that this was primarily a response to the diet and
not a secondary response to changes in renal function, We thank Hill's Science and TechnologyCenter and Hill's Pet-
as observed in the dogs of group SO. Although hyper- foods for donation of the diets used in this study.
cholesterolemia could have contributed to progressive
renal injury in groups SO and BT, variations in total REFERENCES
cholesterol or HDL-cholesterol cannot explain the 1. Barcelli U, MiyataJ, Ito Y, et al. Beneficial effects of polyun-
observed differences between these groups. saturated fatty acids in partially nephrectomized rats.
The present investigation did not address other pos- Prostaglandins 1986;32:211-9.
sible mechanisms of the observed response to the 2. ScharschmidtLA, Gibbons NB, McGarry L, et al. Effects of
dietary manipulations studied. One study demonstrat- dietary fish oil on renal insufficiency in rats with subtotal
nephrectomy. Kidney Int 1987;32:700-9.
ed effects of dietary lipid composition on glomerular 3. Clark WF, Parbtani A, Philbrick DJ, Holub BJ, Huff MW.
hemodynamics that could explain some of the benefi- Chronic effects of omega-3 fatty acids (fish oil) in a rat 5/6
cial effects of fish oil supplementation. 39 We previ- renal ablation model. J Am Soc Nephrol 1991;1:1343-53.
ously reported that dogs with renal insufficiency 4. Logan JL, Michael UF, Benson B. Dietary fish oil interferes
exhibit glomerular hypertension and hyperfiltration, 12 with renal arachidonic acid metabolism in rats: correlation
with renal physiology.Metabolism 1992;41:382-9.
adaptive changes which may be detrimental to the 5. Donadio JV, Bergstralh EJ, Offord KP, Spencer DC, Holley
long-term preservation of renal function. 40,41 If KE. A controlled trial of fish oil in IgA nephropathy. N Engl
dietary supplementation with fish oil modifies these J Med 1994;331:1194-9.
glomerular adaptations in dogs, the course of renal 6. Homan JJ, Biol HJ, Donker AJ, Wilmink JM, Sluiter WJ,
disease could be altered; these mechanisms were not TegzessAM. The effects of dietary supplementation with fish
oil on renal function and the course of early postoperative
studied in the present investigation. It has been argued rejection episodes in cyclosporine-treated renal transplant
that glomerular hypertrophy is maladaptive, leading recipients. Transplantation 1992;54:257-63.
to progressive renal injury. 42 Although all three groups 7. Clark WF, Parabtani A, Naylor CD, et al. Fish oil in lupus
of dogs exhibited an increase in glomerular volume nephritis: clinical findings and methodologicalimplications.
during the study, there were no differences in the Kidney Int 1993;44:75-86.
8. PolzinDL, LeiningerJR, Osborne CA, Jeraj K. Development
degree of glomerular enlargement among the groups. of renal lesions in dogs after 11/12 reduction of renal mass.
However, we evaluated glomerular volume only ini- Lab Invest 1988;58:172-83.
tially and terminally; differences in extent of glomeru- 9. Brown SA, Crowell WA, Barsanti JA, White JV, Finco DR.
lar hypertrophy may have been present at intermedi- Beneficial effects of dietary mineral restriction in dogs with
ate times of the study. Others have argued that dietary marked reduction of functional renal mass. J Am Soc
Nephrol 1991;1:1169-79.
fish oil supplementation provides protection because 10. Brown SA, Finco DR, Crowell WA, Navar LG. Dietary pro-
of its ability to suppress inflammation 43-46 or coagu- tein intake and the glomerular adaptations to partial nephrec-
lation 47 by interfering with the production of proin- tomy in dogs. J Nutr 1991;121:S125-7.
flammatory, procoagulation prostanoids, thrombox- 11. Finco DR, Brown SA, Crowell WA, Duncan RJ, Barsanti JA,
anes, and/or leukotrienes. 48-5° Renoprotection by 0)-3 Bennett SE. Effects of dietary phosphorus and protein in
dogs with chronic renal failure. Am J Vet Res 1992;53:2264-
PUFA could be caused by the tendency of human 71.
beings 51,52 or laboratory rodents24,53,54 ingesting such 12. Brown SA, Finco DR, Crowell WA, Choat DC, Navar LG.
diets to exhibit lower systemic arterial pressure, a Single-nephron adaptations to partial renal ablation in the
change that could reduce progressive renal injury. 55 dog. Am J Physiol 1990;258:F495-F503.
 J Lab Clin Med
454 Brownet al. May 1998

13. Brown SA, Walton CA, Crawford P, Bakris GL. Long-term mentation in the chronic remnant nephron model. Clin
effects of antihypertensive regimens on renal hemodynamics Nephrol 1993;39:295-304.
and proteinuria. Kidney Int 1993;43:1210-8. 32. Aukema HM, Ogbom MR, Tomobe K, Takashi H, Hibino T,
14. Bovee KC, Kronfeld DS, Ramberg C, Goldschmidt M. Long- Holub BJ. Effects of dietary protein restriction and oil type
term measurement of renal function in partially nephrec- on the early progression of murine polycystic kidney disease.
tomized dogs fed 56, 27, or 19% protein. Investig Urol (Berl) Kidney Int 1992;42:837-42.
1979;16:378-84. 33. Council NR. Nutrient requirements of domestic animals:
15. Robertson JL, Goldschmidt M, Kronfeld DS, Tomaszewski nutrient requirements of dogs. Washington, DC: National
JE, Hill S, Bovee KC. Long-term renal responses to high Academy Press, 1985.
dietary protein in dogs with 75% nephrectomy. Kidney Int 34. Clark WF, Parbtani A, Huff MW, Redi B, Holub B J,
1986;29:511-9. Falardeau P. Omega-3 fatty acid dietary supplementation in
16. Finco DR, Brown SA, Crowell WA, Barsanti JA. Exogenous systemic lupus erythematosus. Kidney Int 1989;36:653-60.
creatinine clearance as a measure of glomerular filtration rate 35. Kaisiske B, O'Donnell M, Cleary M, Keane WE Treatment
in dogs with reduced renal mass. Am J Vet Res of hyperlipidemia reduces glomerular injury in obese Zuck-
1991;52:1029-32. er rats. Kidney Int 1988;33:667-2.
17. Bradford MM. A rapid and sensitive method for the quanti- 36. Guijarno C, Keane WF. Lipid-induced glomerular injury.
tation of microgram quantities of protein utilizing the prin- Nephron 1994;67:1-6.
ciple of protein-dye binding. Anal Biochem 1976;72:248-54. 37. Keane WF, Kasiske BL, O'Donnell ME Hyperlipidemia and
18. Brown SA, Brown CA. Single-nephron adaptations to par- the progression of renal disease. Am J Clin Nutr
tial renal ablation in cats. Am J Physiol 1995;269:R1002-8. 1988;47:157-60.
19. Finco DR, Coulter DB, Barsanti JA. Procedure for a simple 38. Wanner C, Kramer-Guth A, Nauck M, Quaschning T, Pavan-
method of measuring glomerular filtration rate in the dog. J stadt H, Schollmeyer P. Cholesterol metabolism in glomem-
Am Anita Hosp Assoc 1982;18:804-6. lar cells: effect of lipoproteins from nephrotic patients. Miner
20. Neumaryer HH, Heinrish M, Schmissas M, Hailer H, Wag- Electrolyte Metab 1996;22:39-46.
ner K, Luft FC. Amelioration of ischemic acute renal failure 39. Schmitz PG, O'Donnell MP, Kasiske BL, Keane WE
by dietary fish oil administration in conscious dogs. J Am Glomerular hemodynamic effects of dietary polyunsaturated
Soc Nephrol 1992;3:1312-20. fatty acid supplementation. J Lab Clin Med 1991; 118:129-
21. Abdel-Gayoum AA, Bashir AA, el-Fakhri MM. Effects of 35.
fish oil and sunflower oil supplementation on gentamicin- 40. Brenner BM, Meyer TW, Hostetter TH. The role of hemo-
induced nephrotoxicity in rat. Hum Exp Toxicol dynamically mediated glomerular injury in the pathogenesis
1995;14:884-8. of progressive glomerular sclerosis in aging, renal ablation,
22. Kasiske BL, O'Donnell MR Lee H, Kim Y, Keane WE and intrinsic renal disease. N Engl J Med 1982;307:652-9.
Impact of dietary fatty acid supplementation on renal injury 41. Hostetter TH, Olson JL, Rennke HG, Venkatachalam MA,
in obese Zucker rats. Kidney Int 1991;39:1125-34. Brenner BM. Hyperfiltration in remnant nephrons: a poten-
23. Rahman MA, Sauter DC, Young MR. Effects of fish oil on tially adverse response to renal ablation. Am J Physiol
the induction of experimental membranous nephropathy in 1981;241:F85-F92.
the rat. Lab Invest 1991;64:371-6. 42. Dworkin LD, Benstein JA, Parker M, Tolbert E, Feiner HD.
24. Shimamura T, Wilson AC. Influence of dietary fish oil on the Calcium antagonists and converting enzyme inhibitors
aortic, myocardial, and renal lesions of SHR. J Nutr Sci Vit- reduce renal injury by different mechanisms. Kidney Int
aminol (Tokyo) 1991 ;37:581-90. 1993;43:808-14.
25. Wheeler DC, Nair DR, Persaud JW, et al. Effects of dietary 43. Prickett JD, Robinson DW, Steinberg AD. Effects of dietary
fatty acids in an animal model of focal glomerulosclerosis. enrichment with eicosapentaenoic acid upon autoimmune
Kidney Int 1991;39:930-7. nephritis in female NZB X NZW/P1 mice. Arthritis Rheum
26. Robinson DR, Xu LL, Tateno S, Guo M, Colvin RB. Sup- 1983;26:133-9.
pression of autoimmune disease by dietary n-3 fatty acids. J 44. Donadio JV. Omega-3 polyunsaturated fatty acids: a poten-
Lipid Res 1993;34:1435-44. tial new treatment of immune renal disease. Mayo Clin Proc
27. Weise WJ, Natori Y, Levine JS, et al. Fish oil has protective 1991;66:1018-28.
and therapeutic effects on proteinuria in passive Heymann 45. Boutard V, Fouqueray B, Philippe C, Perez J, Baud L. Fish
nephritis. Kidney Int 1993;43:359-68. oil supplementation and essential fatty acid deficiency reduce
28. Chandrasekar B, Fernandes G. Decreased pro-inflammatory nitric oxide synthesis by rat macrophages. Kidney Int
cytokines and increased antioxidant enzyme gene expression 1994;46:1280-6.
by omega-3 lipids in murine lupus nephritis. Biochem Bio- 46. Collie-Duguid ES, Wahale KW. Inhibitory effect of fish oil
phys Res Commun 1994;200:893-8. n-3 polyunsaturated fatty acids on the exPression of endothe-
29. Burgess NA, Reynolds TM, Williams N, Pathy A, Smith S. lial cell adhesion molecules. Biochem Biophys Res Com-
Evaluation of four animal models of intrarenal calcium depo- mun 1996;220:969-74.
sition and assessment of the influence of dietary supplemen- 47. Barcelli U, Glass-Greenwalt P, Pollak VE. Enhancing effects
tation with essential fatty acids on calcifications. Urol Res of dietary supplementation with omega-3 fatty acids on plas-
1995;23:239-42. ma fibrinolysis. Thromb Res 1985;39:307-12.
30. Goldstein DJ, Wheeler DC, Sandstrom DJ, Kawachi H, 48. Clark WF, Parbtani A, Philbrick D, et al. Comparative effi-
Salant DJ. Fish oil ameliorates renal injury and hyperlipi- cacy of dietary treatments on renal function in rats with sub-
demia in the Milan normotensive rat model of focal glomeru- total nephrectomy: renal polyunsaturated fatty acid incorpo-
losclerosis. J Am Soc Nephrol 1995;6:1468-75. ration and prostaglandin excretion. Clin Nephrol 1990;33:25-
31. Clark WF, Parbtani A, Philbrick DJ, Spanner E, Huff MW, 34.
Holub BJ. Dietary protein restriction versus fish oil supple- 49. Spurney R.F, Ruiz P, Albrightson CR, Pisetsky DS, Coffman
J Lab Clin Med
Volume 131, Number 5 Brown et al, 455

TM. Fish oil feeding modulates leukotriene production in dietary fatty acids in the borderline hypertensive rat. Trans-
murine lupus nephritis. Prostaglandins 1994;48:331-48. plantation 1992;53:649-54.
50. Chandrasekar B, Fernandes G. Decreased pro-inflammatory 54. Kimura S, Minami M, Saito H, Kobayashi T, Okuyama H.
cytokines and increased antioxidant enzyme gene expression Dietary docosahexaenoic acid (22:6n-3) prevents the devel-
by omega-3 lipids in routine nephritis. Biochem Biophys Res opment of hypertension in SHRSP. Clin Exp Pharmacol
Commun 1994;200:893-8. Physiol 1995;22:$308-9.
51. Radack K, Deck C, Huster G. The effects of low doses of n- 55. Bidani A, Mitchell K, Schwartz MM, Navar LG, Lewis EJ.
3 fatty acid supplementation on blood pressure in hyperten- Absence of glomemlar injury or nephron loss in a normoten-
sive subjects. Arch Intern Med 1991;151:1173-80. sive rat remnant kidney model. Kidney Int 1990;38:28-38.
52. Morris MC, Sacks F, Rosner B. Does fish oil lower blood 56. Panzetta O, Cominacini L, Garbin U, et al. Increased suscep-
pressure? Circulation 1993;88:523-533. tibility of LDL to in vitro oxidation in patients on mainte-
53. Mills DE, Ward RP, McCutcheon D, Dixon H, Ly H, Scho- nance hemodialysis: effects of fish oil and vitamin E admin-
ley J. Attenuation of cyclosporine-induced hypertension by istration. Clin Nephrol 1995;44:303-9.