Evaluation of fermentable oligosaccharides in diets fed to dogs

in comparison to fiber standards

I. S. Middelbos, N. D. Fastinger, and G. C. Fahey Jr.1

Department of Animal Sciences, University of Illinois, Urbana 61801

ABSTRACT: Blends of fermentable oligosaccharides bifidobacteria concentrations compared with the diets
in combination with nonfermentable fiber, cellulose, without supplemental fermentable fiber. Lactobacilli
were evaluated for their ability to serve as dietary fibers concentrations tended to be greater (P < 0.08) in treat-
in dog foods. Using a 6 × 6 Latin square design, 6 diets ments containing fermentable fiber compared with the
were evaluated that contained either no supplemental cellulose treatment. Bifidobacteria and lactobacilli con-
fiber, beet pulp, cellulose, or blends of cellulose, fructoo- centrations were similar for the beet pulp treatment
ligosaccharides, and yeast cell wall added at 2.5% of compared with the fermentable oligosaccharide blends.
the diet. Six ileal-cannulated dogs were fed 175 g of Total fecal short-chain fatty acid concentration was
greater for the beet pulp treatment (P < 0.05) compared
their assigned diet twice daily. Chromic oxide served
with the control and cellulose treatments. The treat-
as a digestibility marker. Nutrient digestibility, fecal
ments containing fermentable fiber had greater (P <
microbial populations, fermentative end products, and
0.05) fecal butyrate concentrations compared with cel-
immunological indices were measured. Total tract DM lulose and control treatments. Immune indices were
and OM digestibilities were lowest (P < 0.05) for the not affected by treatment. Our results suggest that dog
cellulose treatment. Crude protein digestibility was foods containing blends of fermentable and nonfer-
lower (P < 0.05) for the treatments containing carbohy- mentable carbohydrates produce similar physiological
drate blends. The cellulose treatment had the lowest results as dog food containing beet pulp as a fiber
(P < 0.05) concentration of bacteria, and all diets con- source. Therefore, blends of these carbohydrates could
taining fermentable fiber had greater (P < 0.05) fecal be useful substitutes for beet pulp in dog foods.

Key words: beet pulp, dietary fiber, dog, intestinal microbiota, oligosaccharide

©2007 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2007. 85:3033–3044
doi:10.2527/jas.2007-0080

INTRODUCTION tains about 60 to 80% total dietary fiber (TDF; Fahey

et al., 1990b; Sunvold et al., 1995a,b), of which about
Commercial pet foods contain significant amounts of 80% is insoluble (Sunvold et al., 1995c); however, beet
carbohydrates. These carbohydrates are either digest- pulp tends to vary in quality.
ible (mostly starch) or nondigestible (generally classi- Fermentable carbohydrates such as fructooligosac-
fied as dietary fiber). Dietary fiber is physiologically charides (FOS) and mannanoligosaccharides (MOS)
important, because it affects gastric emptying (Russell from yeast cell wall (YCW) have been evaluated for use
and Bass, 1985), digesta transit time (Burrows et al., in dog foods. Both are fermented by intestinal microbi-
1982; Fahey et al., 1990a), fecal bulk (Fahey et al., ota of dogs but MOS at a more moderate rate than FOS
1992; Sunvold et al., 1995b), and short-chain fatty acid (Vickers et al., 2001). Additionally, FOS and MOS may
(SCFA) production in the intestine (Muir et al., 1996; increase lactobacilli and bifidobacteria concentrations
Silvio et al., 2000), depending on fiber type and source. (Swanson et al., 2002c) and decrease production of pu-
Beet pulp is often used in pet foods because of its trefactive compounds (phenols and indoles; Swanson et
fermentation characteristics (Sunvold et al., 1995a) and al., 2002b). The purity and consistency of FOS and MOS
its desirable effect on stool consistency (Fahey et al., products may make them useful as fiber sources in
1992; Sunvold et al., 1995b). Beet pulp typically con- pet foods.
In this study, nondigestible, but fermentable, oligo-
saccharides were evaluated as potential replacements
1
Corresponding author: gcfahey@uiuc.edu for more traditional dietary fiber sources. The blends
Received February 5, 2007. tested contained both fermentable and nonfermentable
Accepted July 25, 2007. fibers to create a balance of insoluble and soluble fiber

3033

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 3034 Middelbos et al.

Table 1. Composition of diets containing select dietary fiber sources and fed to adult
dogs (as-fed basis)
Treatment

Ingredient, % Control Cellulose Beet pulp CF1 CFY12 CFY23

Brewers rice 45.22 42.72 42.72 42.72 42.72 42.72

Poultry by-product meal 37.00 37.00 37.00 37.00 37.00 37.00
Poultry fat 14.00 14.00 14.00 14.00 14.00 14.00
Dried egg 2.40 2.40 2.40 2.40 2.40 2.40
Salt 0.45 0.45 0.45 0.45 0.45 0.45
Potassium chloride 0.56 0.56 0.56 0.56 0.56 0.56
Choline chloride4 0.13 0.13 0.13 0.13 0.13 0.13
Vitamin mix5 0.12 0.12 0.12 0.12 0.12 0.12
Mineral mix6 0.12 0.12 0.12 0.12 0.12 0.12
Cellulose — 2.50 — 1.00 1.00 1.00
Beet pulp — — 2.50 — — —
Fructooligosaccharides7 — — — 1.50 1.20 0.90
Yeast cell wall8 — — — — 0.30 0.60
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.
4
Provided the following per kilogram of diet: choline, 2,284.2 mg.
5
Provided the following per kilogram of diet: vitamin A, 11,000 IU; vitamin D3, 900 IU; vitamin E, 57.5
IU; vitamin K, 0.6 mg; thiamin, 7.6 mg; riboflavin, 11.9 mg; pantothenic acid, 18.5 mg; niacin, 93.2 mg;
pyridoxine, 6.6 mg; biotin, 12.4 mg; folic acid, 1,142.1 ␮g; and vitamin B12, 164.9 ␮g.
6
Provided the following per kilogram of diet: manganese (MnSO4), 17.4 mg; iron (FeSO4), 284.3 mg; copper
(CuSO4), 17.2 mg; cobalt (CoSO4), 2.2 mg; zinc (ZnSO4), 166.3 mg; iodine (KI), 7.5 mg; and selenium (Na2SeO3),
0.2 mg.
7
Nutraflora P-95, GTC Nutrition, Golden, CO.
8
Safmannan, LeSaffre Yeast Corp., Milwaukee, WI.

components. Additionally, the effects of these fer- Bioprocessing and Industrial Value-Added Program fa-
mentable carbohydrates on intestinal health, intestinal cility (Manhattan) under the direction of Pet Food and
microbiota, and immune status of the animal were in- Ingredient Technology Inc. (Topeka, KS). All fiber treat-
vestigated. ments were incorporated into the diets before extrusion.
A total of 6 diets were prepared with the following fiber
MATERIALS AND METHODS sources incorporated:
1) control diet – no supplemental fermentable carbo-
Animals and Diets
hydrate (formulated to analyze as approximately
All surgical and animal care procedures were ap- 1.5% TDF);
proved by the University of Illinois Institutional Animal 2) as (1) + 2.5% cellulose (a highly refractory, poorly
Care and Use Committee before initiation of the ex- fermentable carbohydrate);
periment. 3) as (1) + 2.5% beet pulp (a moderately fermentable
Six purpose-bred adult female dogs (Marshall Biore- fiber source);
sources, North Rose, NY) with hound bloodlines, an 4) as (1) + 1.0% cellulose + 1.5% short-chain FOS
average initial BW of approximately 23 kg, and an aver- (Nutraflora P-95, GTC Nutrition, Golden, CO; CF);
5) as (1) + 1.0% cellulose + 1.2% short-chain FOS
age age of 4.5 yr were surgically prepared with an ileal
+ 0.3% YCW (Safmannan, LeSaffre Yeast Corp.,
T-shaped cannula according to Walker et al. (1994).
Milwaukee, WI; CFY1); and
After the surgery, dogs were closely monitored daily for
6) as (1) + 1.0% cellulose + 0.9% short-chain FOS +
clinical abnormalities and given a 2-wk recovery period
0.6% YCW (CFY2).
before the beginning of the experiment. Dogs were
housed individually in kennels (2.4 × 1.2 m) in a temper- Dogs were offered 175 g of their assigned diet twice
ature-controlled room with a 16-h light:8-h dark cycle daily (0800 and 2000). Chromic oxide was used as a
at the animal care facility of the Edward R. Madigan digestion marker. On d 6 through 14 of each period,
Laboratory on the University of Illinois campus. dogs were dosed with 0.5 g of Cr2O3 at each feeding via
Oligosaccharide-free ingredients were used in diet a gelatin capsule, for a total of 1.0 g of marker/d. Fresh
formulation, with brewers rice, poultry by-product water was available at all times.
meal, and poultry fat constituting the main ingredients
of the dry, extruded kibble diets (Table 1). The diet Sample Collection
formulation was milled at Lortscher Agri Service Inc. A 6 × 6 Latin square design with 14-d periods was
(Bern, KS) and extruded at Kansas State University’s used. A 10-d adaptation phase preceded a 4-d collection

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 Dietary fiber sources in dog foods 3035
of feces and ileal effluent. Ileal effluent was collected 3 Cincinnati, OH) for subsequent bacterial enumeration
times/d, at 4-h intervals. Each collection was 1 h in (total anaerobes, total aerobes, bifidobacteria, lactoba-
length. Sampling times were rotated 1 h from the previ- cilli, Clostridium perfringens, and Escherichia coli). Ali-
ous day’s collection time. For example, sampling times quots of fresh feces were transferred to sterile cryogenic
on the first collection day were 0800, 1200, and 1600; vials (Nalgene, Rochester, NY) and frozen at −80°C un-
on the second day, samples were collected at 0900, 1300, til DNA extraction for microbial analysis (bifidobact-
and 1700, etc. Ileal samples were collected by attaching eria, lactobacilli, C. perfringens, and E. coli). Aliquots
a sterile sampling bag (Whirlpack, Fisher Scientific, for analysis of phenol, indole, and biogenic amine con-
Pittsburgh, PA) to the cannula barrel with a rubber centrations were frozen at −20°C immediately after col-
band. Before attachment of the bag, the cannula plug lection. One aliquot was collected and put in 10 mL of
was removed, the interior of the cannula scraped clean, 2 N hydrochloric acid for SCFA, branched-chain fatty
and old digesta discarded. During collection of ileal ef- acids (BCFA), and ammonia analyses. Additional ali-
fluent, the dogs were encouraged to move around freely. quots were used for pH measurement and fresh fecal
To prevent the dogs from pulling the collection bag from DM determination.
the cannula, Bite-Not collars (Bite-Not Products, San
Francisco, CA) were used during collections as needed. Chemical Analyses
After ileal effluent collection, the cannula plug was put
in place, and the cannula site was cleaned with a dilute Diets, feces, and ileal samples were analyzed for DM,
betadine solution. OM, and ash using AOAC (2000) methods. Crude pro-
Although nutrient digestibility was calculated based tein was calculated from Leco total N values (AOAC,
on digestion marker recovery, total feces excreted dur- 2000). Amino acid concentrations in the diets were ana-
ing the collection phase of each period was removed lyzed at the University of Missouri Experiment Station
from the floor of the pen, weighed, and composited to Chemical Laboratories using a Beckman 6300 AA ana-
obtain the most representative sample. Fecal samples lyzer (Beckman-Coulter Inc., Fullerton, CA) according
were frozen at −20°C until analysis. On d 14 of each to AOAC (2000) methods. Total lipid content (acid hy-
period, a fresh fecal sample was collected within 15 min drolyzed fat, AHF) was determined by acid hydrolysis
of defecation for the measurement of pH and bacterial followed by ether extraction according to AACC (1983)
enumeration. All fecal samples during the 4-d collection and Budde (1952). Dietary fiber concentrations [TDF,
phase were scored for consistency according to the fol- soluble dietary fiber (SDF), and insoluble dietary fiber
lowing system: 1 = hard, dry pellets; small, hard mass; (IDF)] were determined according to Prosky et al.
2 = hard-formed, dry stool; remains firm and soft; 3 = (1984, 1992). Gross energy was measured using an oxy-
soft, formed, and moist stool, retains shape; 4 = soft, gen bomb calorimeter (model 1261, Parr Instruments,
unformed stool; assumes shape of container; and 5 = Moline, IL). Chromium concentrations in digesta and
watery; liquid that can be poured. fecal samples were analyzed according to Williams et
On d 14 of each period, a blood sample (5 mL) was al. (1962) using atomic absorption spectrophotometry
collected via jugular puncture into nonheparinized (model 2380, Perkin-Elmer, Norwalk, CT). Short-chain
evacuated tubes to obtain serum immunoglobulin A, G, fatty acid and BCFA concentrations were determined
and M concentrations. Another 5-mL blood sample was by gas chromatography according to Erwin et al. (1961)
collected in evacuated tubes with EDTA for a complete using a Hewlett-Packard 5890A series II gas chromato-
blood count (total white blood cells, neutrophil, eosino- graph (Palo Alto, CA) and a glass column (180 cm × 4
phil, lymphocyte, and monocyte). mm i.d.) packed with 10% SP-1200/1% H3PO4 on 80/
100+ mesh Chromosorb WAW (Supelco Inc., Bellefonte,
Sample Handling PA). Nitrogen was the carrier with a flow rate of 75
mL/min. Oven, detector, and injector temperatures
Ileal samples were frozen at −20°C in their individual were 125, 175, and 180°C, respectively. Phenol and in-
bags. At the end of the experiment, all ileal effluent dole concentrations were determined using gas chroma-
samples were composited for each dog for each period tography according to the methods of Flickinger et al.
and then refrozen at −20°C. Before analysis, ileal efflu- (2003). Biogenic amine concentrations were measured
ent was lyophilized in a Dura-Dry MP microprocessor- by HPLC according to the methods of Flickinger et
controlled freeze-drier (FTS Systems, Stone Ridge, NY). al. (2003).
Composited fecal samples and diets were dried at 55°C
in a forced-air oven. After drying, diets, fecal samples, Microbial Analyses
and ileal samples were ground through a 2-mm screen
in a Wiley mill (model 4, Thomas Scientific, Swedesb- Microbial populations were determined by serial dilu-
oro, NJ). tion and plating and by DNA extraction from fecal sam-
On d 14 of each period, fresh fecal samples were col- ples, followed by quantitative PCR (qPCR) and dena-
lected within 15 min of defecation, and an aliquot was turing gradient gel electrophoresis techniques as de-
immediately transferred to a preweighed Cary-Blair scribed in detail elsewhere (Middelbos et al., 2007).
transport media container (Meridian Diagnostics Inc., The purpose of using 2 distinct methods for microbial

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 3036 Middelbos et al.

analysis was to compare the 2 techniques under the consistent among diets, but the CP and AHF concentra-
same conditions (the same fecal samples). tions varied slightly. Total dietary fiber, IDF, and SDF
varied consistently with the added fiber components.
Immunological Analyses Nutrient intakes and apparent ileal and total tract
nutrient digestibilities are presented in Table 3. Feed
Ileal immunoglobulin A concentrations were mea-
refusals were minimal, and nutrient intakes were simi-
sured according to the methods of Nara et al. (1983).
lar among treatments. Crude protein intake varied
Freshly collected ileal fluid was frozen at −20°C in ster-
slightly with the CP concentrations in the diet. The
ile collection bags. The frozen samples were lyophilized
intakes for TDF, IDF, and SDF also varied according to
and crushed using a mortar and pestle. A 2-g aliquot
their respective concentrations in the treatment diets.
of each lyophilized and crushed sample was suspended
Total tract DM digestibility was lower (P < 0.05) for
in 20 mL of PBS solution (pH 7.2) and mixed for 30 min
the cellulose treatment compared with the control treat-
at room temperature. Samples then were centrifuged at
ment (83.1 vs. 86.2%). Trends were noted for DM digest-
20,000 × g for 30 min at 4°C. The supernatant was
ibility between the CF and control treatment (P = 0.06;
collected and ileal immunoglobulin A concentrations
84.1 vs. 86.2%) and between the CFY2 and cellulose
determined using a radial immunodiffusion kit (MP
treatments (P = 0.10; 83.2 vs. 85.1%). Organic matter
Biomedicals, Aurora, OH).
digestibility was decreased (P < 0.05) for the cellulose,
After blood was collected in nonheparinized evacu-
CF, and CFY1 treatments compared with control (88.7,
ated tubes, samples were centrifuged at 2,000 × g for
89.8, and 90.0 vs. 91.7%, respectively). Additionally,
20 min at 4°C, and the serum was collected. Serum
the cellulose treatment had decreased (P < 0.05) OM
immunoglobulin A, immunoglobulin G, and immuno-
digestibility compared with the beet pulp and CFY2
globulin M concentrations were determined using ra-
treatments and tended (P = 0.09) to have decreased OM
dial immunodiffusion kits. The blood collected in evacu-
digestibility compared with the CFY1 treatment (88.7
ated tubes containing EDTA was used for complete
vs. 90.0%). Crude protein digestibility was decreased
blood count determination, which was performed on a
(P < 0.05) for the treatments supplemented with fer-
Cell-Dyn 3500 hematology analyzer (Abbott Labora-
mentable oligosaccharides (∼84.7%) compared with the
tories, Abbott Park, IL).
control, beet pulp, and cellulose treatments (∼86.8%).
Calculations Fat digestibility, in general, was high (96 to 97%) but
was decreased (P < 0.05) for the CFY1 treatment com-
Dry matter (g/d) recovered as ileal effluent was calcu- pared with the control and cellulose treatments. Gross
lated by dividing the Cr intake (mg/d) by ileal Cr concen- energy digestibility was decreased (P < 0.05) for the
trations (mg of Cr/g of ileal effluent). Ileal nutrient flows cellulose, CF, and CFY1 treatments compared with con-
were calculated by multiplying DM flow by the concen- trol (89.8, 90.6, and 90.5 vs. 91.9%, respectively).
tration of the nutrient in the ileal DM. Ileal nutrient Total dietary fiber digestibility for the beet pulp treat-
digestibilities were calculated as nutrient intake (g/d) ment (39.1%) was greater (P < 0.05) than for the cellu-
minus ileal nutrient flow (output, g/d), and this value lose (11.5%), CF (15.3%), and CFY1 (14.2%) treatments.
was then divided by nutrient intake (g/d). Similar calcu- Total dietary fiber digestibility values for the control
lations were performed on fecal samples to determine (27.3%) and CFY2 (25.2%) treatments were not differ-
total tract nutrient digestibilities. ent from any of the other treatments.
Complete blood counts and serum and ileal immuno-
Statistical Analysis globulin concentrations are presented in Table 4. No
differences were detected among treatments in white
Data for continuous variables were analyzed by the
blood cell counts or immunoglobulin concentrations.
MIXED procedure, and data for discontinuous variables
Fecal microbial concentrations are presented in Table
were analyzed by the GLIMMIX procedure (SAS Inst.,
5. According to serial dilution and plating methods,
Cary, NC). The experimental design was a 6 × 6 Latin
bifidobacteria concentrations were greatest for the CF
square. The statistical model included the random ef-
fects of animal and period and the fixed effect of treat- treatment, and the only difference detected was be-
ment. All treatment least squares means were com- tween the CF and cellulose treatments (P < 0.05).
pared with each other, and the Tukey adjustment was Trends were noted for decreased bifidobacteria concen-
used to control for experimentwise error. Differences trations for the control treatment compared with CF
among least squares means with a probability of P < (P = 0.06) and CFY2 compared with the cellulose treat-
0.05 were accepted as statistically significant, although ment (P = 0.09). Clostridium perfringens, E. coli, and
mean differences with P-values ranging from 0.06 to lactobacilli concentrations were not different among
0.10 were accepted as trends. treatments, although trends (P = 0.07) were detected
for decreased lactobacilli concentrations in the control
RESULTS and cellulose treatments compared with the CF treat-
ment. Total aerobic bacteria concentrations were
The chemical composition of the diets is presented in greater (P < 0.05) for the CF treatment compared with
Table 2. Dry matter, OM, and GE concentrations were the cellulose treatment and tended (P = 0.06) to be

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 Dietary fiber sources in dog foods 3037
Table 2. Chemical composition of diets containing select dietary fiber sources and fed to
adult dogs
Treatment

Item Control Cellulose Beet pulp CF1 CFY12 CFY23

DM, % 91.6 91.4 90.9 91.1 91.2 90.6

%, DM basis
OM 91.9 92.1 91.7 91.6 92.0 92.4
CP 33.4 33.2 33.1 32.1 30.3 29.4
Acid hydrolyzed fat 21.3 21.7 20.4 22.3 19.9 20.5
GE, kcal/g 5.4 5.4 5.2 5.4 5.2 5.3
Total dietary fiber 2.53 5.07 4.03 3.21 3.50 3.46
Insoluble fiber 1.68 4.20 2.66 2.31 2.27 2.37
Soluble fiber 0.85 0.87 1.37 0.90 1.23 1.09

Essential AA
Arg 1.98 1.97 1.87 1.94 1.81 1.70
His 0.58 0.57 0.54 0.55 0.51 0.49
Ile 1.08 1.07 1.00 1.02 0.90 0.89
Leu 1.98 1.97 1.85 1.89 1.76 1.68
Lys 1.60 1.59 1.48 1.53 1.40 1.32
Met 0.57 0.59 0.56 0.58 0.54 0.51
Phe 1.14 1.12 1.06 1.08 1.01 0.97
Thr 1.02 1.02 0.97 0.98 0.93 0.88
Trp 0.22 0.22 0.20 0.19 0.19 0.19
Val 1.36 1.34 1.27 1.32 1.16 1.13
Nonessential AA
Ala 1.80 1.82 1.74 1.79 1.66 1.57
Asp 2.28 2.31 2.18 2.24 2.09 1.96
Cys 0.34 0.34 0.31 0.32 0.30 0.29
Glu 3.55 3.53 3.35 3.44 3.25 3.11
Gly 2.62 2.69 2.58 2.66 2.47 2.32
Pro 1.86 1.90 1.82 1.80 1.74 1.64
Ser 1.00 1.00 0.95 0.95 0.93 0.89
Tyr 0.81 0.78 0.75 0.75 0.76 0.70
Total essential AA 11.53 11.46 10.80 11.08 10.21 9.76
Total nonessential AA 14.26 14.37 13.68 13.95 13.20 12.48
Total AA 25.79 25.83 24.48 25.03 23.41 22.24
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.

greater for the CF treatment compared with the control represent comparisons between 2 treatments and, thus,
treatment (10.0 vs. 9.0 log10 cfu/g). Total anaerobic bac- cannot be separated in a classical statistical manner.
terial concentrations were greater (P < 0.05) for the The pooled SEM indicates overall variation present in
CF and CFY2 treatments compared with the cellulose the matrix of comparisons.
treatment. Additionally, there was a trend (P = 0.09) There is a clear pattern in the similarity of fecal DNA
for greater total anaerobic bacterial concentrations for among treatments. The control treatment and cellulose
the CF treatment compared with the control. treatment had a high degree of similarity (87%), as
Fecal bacterial enumeration by qPCR did not indicate did the treatments containing fermentable fiber sources
significant differences in C. perfringens or E. coli popu- (beet pulp, CF, CFY1, and CFY2; 85 to 88%). The control
lations among treatments. Bifidobacteria concentra-
treatment and the treatments containing fermentable
tions, however, were greater (P < 0.05) in the treat-
fiber had slightly lower similarity (84 to 85%), whereas
ments supplemented with fermentable substrate (beet
the cellulose treatment resulted in the least similar
pulp, CF, CFY1, and CFY2). The CF and CFY2 treat-
ments resulted in greater (P < 0.05) lactobacilli concen- fecal DNA compared with the treatments containing
trations compared with the cellulose treatment. Addi- fermentable fibers (80 to 83%).
tionally, trends for increased lactobacilli concentrations Fecal pH and score, daily wet fecal output, and con-
were noted with the beet pulp (P = 0.07) and CFY1 (P = centrations of fecal ammonia, SCFA, BCFA, phenol,
0.08) treatments compared with the cellulose and indole are presented in Table 7. Fecal pH was unaf-
treatment. fected by the dietary fiber treatments, as was fecal
Analysis of total fecal bacterial DNA is presented score. Wet fecal output was not significantly different
in Table 6. The mean Dice’s scores given in the table among treatments, but the beet pulp treatment tended

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 3038 Middelbos et al.

Table 3. Food intake and apparent ileal and total tract digestibility by dogs fed diets
containing select dietary fiber sources
Treatment

Item Control Cellulose Beet pulp CF1 CFY12 CFY23 SEM

Intake, g/d
DM 296 278 302 282 263 285 18.0
OM 272 256 277 259 242 263 16.5
CP 99 92 100 90 80 84 5.7
Acid hydrolyzed fat 63 60 62 63 52 58 3.8
GE, kcal/d 1,584 1,499 1,568 1,511 1,374 1,497 95.3
Total dietary fiber 7.5 14.1 12.2 9.0 9.2 9.9 0.65
Insoluble fiber 5.0 11.7 7.8 6.5 5.9 6.8 0.49
Soluble fiber 2.5 2.4 4.2 2.6 3.2 3.1 0.19
Ileal digestibility, %
DM 78.0 57.8 67.5 73.1 77.2 72.6 6.0
OM 83.9 67.4 75.7 79.7 82.8 79.5 4.6
CP 74.0 55.4 65.0 70.7 74.2 68.4 6.5
Acid hydrolyzed fat 95.9 93.1 93.6 96.1 96.0 95.5 1.0
GE 86.1 71.6 77.9 82.5 85.0 81.8 4.1
Total dietary fiber −26.2 −36.3 −11.5 −12.2 −2.6 −40.7 26.2
Total tract digestibility, %
DM 86.2a 83.2b 85.4a 84.1ab 84.2ab 85.1ab 0.9
OM 91.7a 88.7c 91.0ab 89.8bc 90.0bc 90.5ab 0.6
CP 86.9a 86.7a 87.0a 84.9b 84.7b 84.8b 1.0
Acid hydrolyzed fat 97.2a 97.1a 96.6ab 96.7ab 95.9b 96.5ab 0.3
GE 91.9a 89.8c 91.3ab 90.6bc 90.5bc 90.9abc 0.6
Total dietary fiber 27.3ab 11.5b 39.1a 15.3b 14.2b 25.2ab 6.2

Means in the same row not sharing common superscript letters are different (P < 0.05).
a–c
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.

(P = 0.09) to result in greater wet fecal output compared Fecal propionate concentrations were greatest for the
with the CFY1 treatment. beet pulp treatment and greater (P < 0.05) than for the
Fecal ammonia concentration was not affected sig- control, cellulose, and CF treatments. Both CFY1 and
nificantly by dietary treatment. Per gram of fecal DM, CFY2 treatments had greater (P < 0.05) propionate con-
the beet pulp treatment resulted in the greatest fecal centrations than did the control and cellulose treat-
concentration of acetate, which was different (P < 0.05) ments. Additionally, the propionate concentration for
from values for the cellulose and control treatments. the CF treatment tended (P = 0.08) to be greater than

Table 4. Complete blood count and serum and ileal immunoglobulin concentrations in
adult dogs fed diets containing select dietary fiber sources
Treatment

Item Control Cellulose Beet pulp CF1 CFY12 CFY23 SEM

Blood cell count, thousands/␮L

Total white cells 11.3 11.7 11.8 9.9 10.1 10.9 1.0
Neutrophils 7.3 7.7 8.4 6.9 6.4 7.1 0.8
Lymphocytes 2.5 2.0 1.9 1.9 2.3 2.4 0.3
Monocytes 1.0 1.1 0.9 0.6 0.7 0.7 0.2
Eosinophils 0.5 0.8 0.6 0.6 0.7 0.6 0.1
Serum immunoglobulins, mg/dL
Immunoglobulin A 38.6 39.6 43.8 40.4 35.3 39.0 7.3
Immunoglobulin G 1,209 1,306 1,218 1,264 1,332 1,244 145.0
Immunoglobulin M 165.8 143.6 148.7 148.8 146.8 151.1 15.3
Ileal immunoglobulins, mg/g of DM
Immunoglobulin A 4.3 4.5 3.4 4.1 4.7 4.4 1.3
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 Dietary fiber sources in dog foods 3039
Table 5. Fecal microbial populations for dogs fed diets containing select dietary fiber
sources
Treatment

Item Control Cellulose Beet pulp CF1 CFY12 CFY23 SEM

Population (plating) cfu log10/g of fecal DM

Bifidobacteria 10.1ab 10.0b 10.3ab 10.6a 10.2ab 10.5ab 0.13
Clostridium perfringens 9.8 9.9 9.8 9.8 9.9 10.0 0.19
Escherichia coli 8.0 8.1 8.4 8.0 8.7 7.9 0.25
Lactobacilli 9.3 9.4 9.6 10.4 9.8 10.1 0.34
Total aerobes 9.0ab 8.9b 9.6ab 10.0a 9.8ab 9.8ab 0.28
Total anaerobes 10.7ab 10.5b 10.8ab 11.0a 10.9ab 11.0a 0.11
Population (quantitative PCR)
Bifidobacteria 7.7b 7.8b 8.9a 8.7a 9.1a 8.7a 0.30
C. perfringens 10.7 10.2 11.3 11.2 11.2 11.0 0.38
Lactobacilli 11.3ab 11.2b 12.0ab 12.2a 12.1ab 12.1a 0.23
E. coli 10.3 10.1 10.6 10.7 10.7 10.1 0.32
Means in the same row not sharing common superscript letters are different (P < 0.05).
a,b
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.

for the cellulose treatment. Butyrate concentrations not affected by dietary treatment. The CFY2 treatment
were decreased (P < 0.05) for the control and cellulose resulted in the greatest concentration of biogenic
treatments compared with treatments that were sup- amines (6.53 ␮mol/g of DM), whereas the cellulose
plemented with some form of fermentable fiber where treatment resulted in the lowest concentration (2.55
values were similar to each other. Isobutyrate concen- ␮mol/g of DM). Of the individual amines, putrescine
trations were not different among treatments. Isovaler- was greater (P < 0.05) for the CFY2 treatment compared
ate concentrations were greater (P < 0.05) for the CF with the cellulose treatment.
and CFY1 treatments than for the cellulose treatment.
Valerate concentrations were greater (P < 0.05) for the DISCUSSION
CF and CFY1 treatments compared with the cellulose
treatment. Trends were noted for decreased valerate In this experiment, nondigestible, but fermentable,
concentrations for the cellulose compared with the beet oligosaccharides were evaluated as potential replace-
pulp (P = 0.07) and CFY2 (P = 0.06) treatments. ments for more traditional dietary fiber sources. The
Phenol, p-cresol, and indole concentrations were not blends of carbohydrates tested contained both fer-
affected by treatment. Phenol concentrations for the mentable and nonfermentable fibers to create a balance
CFY2 treatment were below the detection limit for all of insoluble and soluble fiber components.
samples. No 4-ethyl-phenol, 7-methyl-indole, 3-methyl- The variation in CP intake was related directly to
indole, 2-methyl-indole, or 2,3-dimethyl-indole was de- the CP concentrations in the diets. This is evident, be-
tected in any of the samples. cause the DM and OM intakes were very similar. How-
Fecal biogenic amine concentrations are presented ever, dietary CP concentrations in all diets were well
in Table 8. Total biogenic amine concentrations were above minimum requirements, and protein intake was

Table 6. Matrix of Dice’s similarity coefficients for total fecal DNA in dogs fed diets
containing select dietary fiber sources as determined by denaturing gradient gel electro-
phoresis
Treatment Control Cellulose Beet pulp CF1 CFY12 CFY23 SEM
4
Control X 86.98 84.95 84.82 84.19 83.53
Cellulose X 82.93 80.78 80.09 80.22
Beet pulp X 87.78 86.23 84.58
CF X 87.69 86.57
CFY1 X 87.99
CFY2 X
SEM 2.32
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.
4
Means in rows and columns cannot be separated, because they represent comparisons between 2 treat-
ments.

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 3040 Middelbos et al.

Table 7. Fecal pH, score, and concentrations of ammonia, short-chain fatty acids, branched-
chain fatty acids, phenols, and indoles for dogs fed diets containing select dietary fiber
sources
Treatment

Item Control Cellulose Beet pulp CF1 CFY12 CFY23 SEM

pH 6.7 6.5 6.3 6.5 6.4 6.3 0.14

Score4 2.9 2.7 2.7 2.9 2.7 2.6 0.18
Fecal output (as-is), g/d 73.2 87.6 89.6 80.1 69.0 76.6 7.7
Ammonia, ␮mol/g of DM 164 132 191 177 169 175 15.1
Short-chain fatty acids, ␮mol/g of DM
Acetate 172 b 127 b 276 a 187 ab 196 ab 187 ab 24.3
Propionate 63 bc 49 c 93 a 70 bc 84 ab 85 ab 6.2
Butyrate 28 b 21 b 42 a 40 a 41 a 42 a 4.6
Total 262 b 197 b 411 a 297 ab 321 ab 314 ab 33.2
Branched-chain fatty acids, ␮mol/g of DM
Isobutyrate 5.6 4.8 6.6 6.5 6.9 6.0 0.7
Isovalerate 10.4ab 8.3b 12.2ab 13.4a 13.5a 11.8ab 1.3
Valerate 12.7ab 9.6b 16.1ab 18.7a 16.9a 16.3ab 2.0
Phenols and indoles,5 ␮g/g of DM
Phenol 20.1 9.1 4.7 7.4 6.9 Trace 8.0
p-Cresol 79.9 50.5 30.6 40.6 73.7 19.7 17.0
Indole 224.1 193.3 254.5 238.9 173.7 201.2 28.4
a–c
Means in the same row not sharing common superscript letters are different (P < 0.05).
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.
4
Score based on the following scale: 1 = hard, dry pellets; small, hard mass; 2 = hard-formed, dry stool;
remains firm and soft; 3 = soft, formed, and moist stool; retains shape; 4 = soft, unformed stool; assumes
shape of container; and 5 = watery; liquid that can be poured.
5
4-Ethyl-phenol, 7-methyl-indole, 3-methyl-indole, 2-methyl-indole, and 2,3-dimethyl-indole were not de-
tected in any sample.

at least double the recommended CP intake on a meta- cellulose treatment had the greatest intake, which is
bolic BW basis (NRC, 2006). caused by the refractory nature of cellulose (>90% TDF;
Intake of TDF, IDF, and SDF also varied among Sunvold et al., 1995a,b,c). The beet pulp treatment re-
treatment. This observation is not surprising, because sulted in a greater fiber intake compared with most
treatment diets were formulated to contain different other treatments due to the relatively high fiber concen-
concentrations of dietary fiber. The control treatment tration in beet pulp (∼60 to 80%; Fahey et al., 1990b;
had the lowest fiber intake as expected, whereas the Sunvold et al., 1995c). The fermentable oligosaccharide

Table 8. Fecal biogenic amine concentrations in dogs fed diets containing select dietary
fiber sources
Treatment

Item Control Cellulose Beet pulp CF1 CFY12 CFY23 SEM

␮mol/g of DM
Total biogenic amines4,5 4.01 2.55 5.11 4.03 5.14 6.53 1.55
Agmatine 0.38 0.25 0.29 0.07 0.59 0.47 0.29
Cadaverine 0.42 0.23 0.55 0.26 0.61 0.67 0.18
Histamine 0.12 0.10 0.07 0.00 0.14 0.11 0.09
Putrescine 1.02ab 0.35b 1.98ab 2.21ab 1.27ab 3.04a 0.73
Spermidine 1.10 0.90 1.16 0.95 1.17 1.12 0.31
Spermine 0.30 0.26 0.19 0.03 0.30 0.22 0.14
Tryptamine 0.32 0.26 0.42 0.35 0.46 0.40 0.13
Tyramine 0.36 0.21 0.45 0.16 0.52 0.48 0.18
Means in the same row not sharing common superscript letters are different (P < 0.05).
a,b
1
CF = 1% cellulose + 1.5% fructooligosaccharides.
2
CFY1 = 1% cellulose + 1.2% fructooligosaccharides + 0.3% yeast cell wall.
3
CFY2 = 1% cellulose + 0.9% fructooligosaccharides + 0.6% yeast cell wall.
4
Phenylethylamine was analyzed but present only in trace amounts (<0.01 ␮mol/g of DM).
5
Total biogenic amines equal the sum of the individual amines listed in Table 8 plus phenylethylamine.

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 Dietary fiber sources in dog foods 3041
treatments resulted in lower fiber intakes, not because Immunological indices were not affected by dietary
fermentable carbohydrate was present in lower concen- treatment in this study, although previous research
trations, but because FOS in particular do not analyze with fermentable oligosaccharides in dogs (O’Carra,
as TDF due to their low degree of polymerization and 1997; Swanson et al., 2002b,c; Grieshop et al., 2004;
the resulting inability to precipitate in 80% ethanol. Middelbos et al., 2007) suggested that white blood cell
The high total tract nutrient digestibility for all diets counts and immunoglobulin concentrations may be al-
is consistent with the dietary formulation that con- tered. All values reported herein, however, are within
tained high-quality ingredients. Total tract DM digest- physiological ranges for adult dogs. Perhaps greater
ibility was greater for the control and beet pulp treat- concentrations of fibrous constituents are necessary to
ment compared with cellulose treatment. Cellulose in- affect immunological outcomes.
creases DM output and can thereby decreases DM Fecal microbial populations were affected by dietary
digestibility. Muir et al. (1996) reported that increasing treatment. Bifidobacteria concentrations were signifi-
concentrations (2.5 to 7.5%) of Solka floc depressed total cantly greater for the CF treatment compared with the
tract DM, OM, and CP digestibility while not affecting cellulose treatment. In combination with the tendencies
ileal nutrient digestibility. Total tract OM digestibility of greater bifidobacteria counts for the CF compared
was greater for the control treatment than for the cellu- with the control treatment and the CFY2 treatment
lose, CF, and CFY1 treatments. This could possibly be compared with the cellulose treatment, these observa-
attributed to TDF composition, because the cellulose, tions suggest preference of bifidobacteria for treatments
CF, and CFY1 treatments numerically had the lowest containing FOS and MOS. The effect of fermentable
TDF digestibilities. This would also account for the dif- fiber is more striking when qPCR is used for microbial
ference in OM digestibility between the beet pulp and analysis. Both the control and cellulose treatments had
CFY2 treatments compared with cellulose. Digestible significantly decreased bifidobacteria counts compared
energy responded similarly to OM digestibility, with with the beet pulp, CF, CFY1, and CFY2 treatments.
the exception of the difference between the CFY2 and These results suggest that blends of fermentable oligo-
cellulose treatments, and may be related to TDF digest- saccharides, but also beet pulp, are able to increase
ibility. Total tract CP digestibility was greater for the bifidobacteria in the canine intestine. Clostridium per-
control, cellulose, and BP treatments compared with fringens and E. coli concentrations were not affected
the CF, CFY1, and CFY2 treatments. The high CP di- by dietary treatment according to microbiota analysis
gestibility of the control and cellulose treatments (com- by both serial dilution and plating and qPCR.
pared with the treatments with fermentable oligosac- Using serial dilution and plating, lactobacilli concen-
charides) is probably due to the lack of fermentable trations tended to be greater for the CF treatment than
substrate and a resulting limited formation of bacterial for the control and cellulose treatments. As with bifido-
biomass. Increased synthesis of bacterial biomass and bacteria, these results are not surprising, because the
the subsequently decreased CP digestibility is the re- available substrate for fermentation favors lactobacilli
sult of GE derived from oligosaccharide fermentation. proliferation. Analysis using qPCR provides stronger
This effect will, in turn, slightly affect DM and OM support for the CF treatment effect. The CF treatment
digestibility, as was noted herein. The high CP digest- had significantly greater lactobacilli concentrations
ibility of the beet pulp treatment is slightly surprising, than the cellulose treatment, and trends existed for all
because bacterial counts were high for this treatment treatments containing fermentable components to have
(discussed later). Although the control and cellulose greater lactobacilli counts than the cellulose treatment.
treatments had statistically greater AHF digestibilities The observed differences between the treatments with-
compared with the CFY1 treatment, the small numeri- out and with added fermentable components were 0.7
cal difference (<1.3%) is not likely to affect the animal. to 1.0 log counts (or a 5- to 10-fold increase in number
The greater TDF digestibility by dogs fed the beet pulp of cells), which could be considered a biologically sig-
treatment compared with the cellulose, CF, and CFY1 nificant difference.
treatments is not surprising, because this treatment Total anaerobic and aerobic bacterial counts were
contained the greatest concentration of soluble fiber. significantly lower for the cellulose treatment compared
The refractory nature of cellulose resulted in lower TDF with the CF treatment, and total anaerobic bacteria
digestibility by dogs fed the cellulose treatment and is were also lower for the cellulose treatment compared
probably responsible for the decreased TDF digestibil- with the CFY2 treatment. Additionally, the CF treat-
ity of the CF and CFY1 treatments. It is unlikely that ment also tended to have greater total aerobic bacterial
the presence of FOS and MOS from YCW contribute to counts compared with the control treatment. The inert
differences in TDF digestibility, because these oligosac- fiber, cellulose, is likely the reason why this dietary
charides are too soluble to analyze as TDF. The digest- treatment resulted in lower bacterial counts compared
ibility of TDF for the beet pulp treatment is comparable with some of the treatments containing fermentable
to previous work from our laboratory (Fahey et al., substrates.
1990b), and although beet pulp at greater concentra- This is the first report that beet pulp inclusion in dog
tions (>7.5%) may affect nutrient digestibility, the 2.5% diets produced bifidobacteria and lactobacilli popula-
concentration used here does not have this drawback. tions in similar concentrations as diets containing fer-

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 3042 Middelbos et al.

mentable oligosaccharides. The latter treatments cate that beet pulp stimulates similar bacterial species
would have a theoretical advantage in their potential as FOS, but a mixture of FOS and YCW does so to a
to increase lactic acid-producing bacteria such as bifid- lesser extent. Although no quantitative inferences can
obacteria and lactobacilli because of the selective fer- be made based on these data, it is evident that low-
mentation of FOS and MOS by these species. However, level fiber supplementation is able to alter total fecal
these data suggest that beet pulp as a fiber source is bacterial DNA composition. This is likely the result of
just as potent in increasing beneficial gut bacteria as substrate-specific alteration of bacterial populations in
are prebiotic oligosaccharides. This finding may have the intestine.
implications for the use of beet pulp as a fiber source Fecal pH and fecal score were not affected by dietary
in control diets used in studies evaluating possible pre- treatment. Increased fermentation and the resulting
biotic compounds. Indeed, the 2 microbial enumeration SCFA production are thought to lower pH and possibly
methods used here tended to agree on the order of the increase fecal water content. However, the pH effect
treatments in terms of fecal microbial concentrations. may be limited, because the dog colon absorbs SCFA
However, with the qPCR method, the effect of treat- rapidly, and absorption increases as SCFA concentra-
ment on bifidobacteria and lactobacilli was much more tions increase (Herschel et al., 1981). The beet pulp
distinct compared with the plating method, despite the treatment tended to have a greater wet fecal output
greater variation for bifidobacteria. Given the composi- compared with the CFY1 treatment. Compared with
tion of the fiber substrates, the patterns in bifidobact- other fiber sources, beet pulp has been reported to in-
eria and lactobacilli concentrations measured using crease fecal output (Fahey et al., 1992; Sunvold et al.,
qPCR were as expected. 1995b), but the inclusion level of beet pulp used herein
Differences in C. perfringens concentrations among is up to 80% lower than in previously published work.
treatments were larger (although not significant) using Additionally, fecal output increases linearly with beet
qPCR enumeration compared with plating methods. A pulp inclusion level (Fahey et al., 1990b), and, there-
possible reason for this is the difficulty of growing C. fore, the low inclusion level of fiber used here may not
perfringens from a fecal inoculum, because the cultiva- have been sufficient to increase fecal bulk significantly
tion is susceptible to fungal contamination and fast in excess of that for a very low fiber diet (the control
overgrowth of the colonies on the plate. This makes treatment).
distinguishing separate colonies difficult and may im- Fecal ammonia concentrations were not affected by
pair the detection of differences in colony-forming units. dietary treatment, but fecal SCFA and BCFA were af-
It is interesting to note that, even though the trends fected. The beet pulp treatment had significantly
between methods are similar, the absolute numbers of greater acetate, propionate, and butyrate concentra-
colony-forming units enumerated are markedly differ- tions per gram of DM compared with the control and
ent. In the case of bifidobacteria, the plating colony- cellulose treatments. The high concentration of SCFA
forming unit counts are greater than the qPCR colony- in the beet pulp treatment is likely caused by the fact
forming unit counts. Normally, qPCR tends to report that beet pulp contains a blend of several fermentable
greater colony-forming units because it also accounts substrates, all with different fermentation rates. By
for dead (but intact) bacterial cells, whereas plating is the time of defecation, it is possible that fermentation
able only to analyze viable cells in the fecal sample. activity is greater for beet pulp than for the treatments
Nevertheless, the observation that different analytical containing no fermentable substrates (control and cel-
methods yield similar differences in bacterial concen- lulose). Oligosaccharides such as MOS and, especially,
trations among treatments (i.e., the absence of a FOS are highly fermentable (compared with natural
method × treatment interaction) is much more im- fibers) and are rapidly used up once they enter the
portant than the actual numbers they generate. Based large intestine. Combined with the fast and efficient
on the results noted here, the 2 methods yield similar intestinal absorption of SCFA, this may affect SCFA
results, although qPCR appears to have more discrimi- presence in feces (at least numerically) in comparison
natory power. with beet pulp. It has been estimated that only 5% of
The similarity in total fecal DNA is most striking SCFA produced in the intestinal tract are excreted in
between the cellulose treatment and the treatments feces (McNeil et al., 1978). A very interesting effect
containing fermentable oligosaccharides, with values noted in this experiment is the significantly greater
of 80 to 81%. This was expected, because the cellulose butyrate concentrations for all treatments containing
treatment does not allow for extensive bacterial growth fermentable components compared with the control and
as was outlined above. Among the 3 treatments con- cellulose treatments. Butyrate is the main fuel source
taining fermentable oligosaccharides (CF, CFY1, and of colonocytes, and high butyrate concentrations are
CFY2), similarity of fecal DNA was high (87 to 88%). thought to be associated with gut health and colonocyte
The beet pulp treatment was most similar to the CF proliferation (Roediger, 1995). However, this butyrate
treatment (88%), but similarity with the CFY1 and production may not be solely due to altered bifidobact-
CFY2 treatments was lower (85 to 86%). This latter eria and lactobacilli concentrations, because these bac-
observation could be due to fermentation characteris- teria produce mainly lactate. Lactate, however, can be
tics associated with each treatment, which could indi- used by other species such as Eubacterium spp. to form

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 Dietary fiber sources in dog foods 3043
butyrate. Unfortunately, neither lactate nor Eubacte- In conclusion, fermentable carbohydrates used in
rium or other lactate-consuming bacterial concentra- blends with nonfermentable fiber results in similar re-
tions were measured in this experiment. Nevertheless, sponses in digestive physiology as beet pulp, which is
if the increased butyrate concentrations are indigenous regarded as one of the superior natural fiber sources
to the type of fermentable carbohydrate blends used in pet foods. Especially compared with a diet supple-
here, it would be of interest to investigate how lactate mented with cellulose only, it is evident that fer-
produced by bifidobacteria and lactobacilli is further mentable fiber plays an important role in the canine
metabolized to butyrate by other bacterial species. Re- intestine. Although nutrient digestibility may be
gardless of the mechanism of butyrate formation, our slightly decreased when using fermentable fiber blends,
data suggest that fermentable fiber sources are able to this effect is offset by beneficial properties. For example,
produce large amounts of butyrate compared with diets fermentable fiber stimulates bacterial growth and fer-
containing no supplemental fiber or cellulose. mentation, thereby increasing production of butyrate.
Branched-chain fatty acids are generated when en- It appears, however, that fermentable fiber also may
ergy is limiting in the large intestine. In this experi- increase BCFA concentrations. Nevertheless, if a fiber
ment, the CF treatment had significantly greater fecal blend containing fermentable oligosaccharides can be
BCFA concentrations compared with the cellulose produced economically and at a constant quality, it may
treatment. The CF treatment also had significantly be an alternative to the use of beet pulp, which is known
greater total bacterial counts. The rapid fermentation to vary in consistency and quality.
of FOS in the proximal colon may have resulted in a
limited energy environment in the distal region, leading LITERATURE CITED
to an increased catabolism of AA with the end result
AACC. 1983. Approved Methods. 8th ed. Am. Assoc. Cereal Chem.,
being greater BCFA concentrations. Previous research
St. Paul, MN.
with fermentable carbohydrates (inulin, oligofructose, AOAC. 2000. Official Methods of Analysis. 17th ed. Assoc. Off. Anal.
and short-chain FOS) in our laboratory reported mixed Chem., Arlington, VA.
effects on fecal BCFA concentrations ranging from no Budde, E. F. 1952. The determination of fat in baked biscuit type
effect to a decrease in BCFA, depending on type and dog foods. J. Assoc. Off. Agric. Chem. 35:799–805.
Burrows, C. F., D. F. Kronfeld, C. A. Banta, and A. M. Merritt. 1982.
concentration of the oligosaccharide (Swanson et al.,
Effects of fiber on digestibility and transit time in dogs. J. Nutr.
2002a,b; Flickinger et al., 2003; Propst et al., 2003). 112:1726–1732.
Fecal biogenic amine concentrations reported here Erwin, E. S., G. J. Marco, and E. M. Emery. 1961. Volatile fatty acid
are comparable to results previously reported by our analyses of blood and rumen fluid by gas chromatography. J.
laboratory in adult dogs (Swanson et al., 2002a,b; Flick- Dairy Sci. 44:1768–1771.
Fahey, G. C., Jr., N. R. Merchen, J. E. Corbin, A. K. Hamilton, L. L.
inger et al., 2003; Propst et al., 2003). Additionally, the Bauer, E. C. Titgemeyer, and D. A. Hirakawa. 1992. Dietary
pattern of biogenic amine concentrations is similar to fiber for dogs. III. Effects of beet pulp and oat fiber additions to
previously reported data with putrescine and spermi- dog diets on nutrient intake, digestibility, metabolizable energy,
dine and accounts for at least 31% (but often >50%) of and digesta mean retention time. J. Anim. Sci. 70:1169–1174.
total biogenic amines, suggesting significant decarbox- Fahey, G. C., Jr., N. R. Merchen, J. E. Corbin, A. K. Hamilton, K.
A. Serbe, and D. A. Hirakawa. 1990a. Dietary fiber for dogs. II.
ylation of Met, Arg, and ornithine, the parent AA of Iso-total dietary fiber (TDF) additions of divergent fiber sources
putrescine and spermidine (and spermine). Although to dog diets and their effects on nutrient intake, digestibility,
total amine concentration is not statistically different metabolizable energy and digesta mean retention time. J. Anim.
among treatments, the numerically decreased concen- Sci. 68:4229–4235.
Fahey, G. C., Jr., N. R. Merchen, J. E. Corbin, A. K. Hamilton, K.
tration for the cellulose treatment compared with the
A. Serbe, S. M. Lewis, and D. A. Hirakawa. 1990b. Dietary fiber
other treatments is in agreement with other indices of for dogs. I. Effects of graded levels of dietary beet pulp on nutrient
bacterial activity (e.g., SCFA concentrations and fecal intake, digestibility, metabolizable energy and digesta mean re-
bacterial counts). tention time. J. Anim. Sci. 68:4221–4228.
Although polyamines (cadaverine, putrescine, sper- Flickinger, E. A., E. M. W. C. Schreijen, A. R. Patil, H. S. Hussein, C.
M. Grieshop, N. R. Merchen, and G. C. Fahey Jr. 2003. Nutrient
midine, and spermine) have been linked to the incidence
digestibilities, microbial populations, and protein catabolites as
of colorectal cancers (Milovic and Turchanowa, 2003), affected by fructan supplementation of dog diets. J. Anim. Sci.
they are required for normal development and repair 81:2008–2018.
of intestinal mucosal cells throughout the intestinal Grieshop, C. M., E. A. Flickinger, K. J. Bruce, A. R. Patil, G. L.
tract (Wang and Johnson, 1990; Wang et al., 1991; Czarnecki-Maulden, and G. C. Fahey Jr. 2004. Gastrointestinal
and immunological responses of senior dogs to chicory and man-
Löser et al., 1999). It is, therefore, undesirable to have
nan-oligosaccharides. Arch. Anim. Nutr. 58:483–493.
significantly decreased polyamine concentrations in the Herschel, D. A., R. A. Argenzio, M. Southworth, and C. E. Stevens.
intestinal lumen, because depletion of (intracellular) 1981. Absorption of volatile fatty acid, Na, and H2O by the colon
polyamines affects epithelial cell apoptosis. The exact of the dog. Am. J. Vet. Res. 42:1118–1124.
mechanisms and effects of polyamine depletion on Löser, C., A. Eisel, D. Harms, and U. R. Fölsch. 1999. Dietary poly-
amines are essential luminal growth factors for small intestinal
apoptosis, however, are not yet completely understood, and colonic mucosal growth and development. Gut 44:12–16.
and research in this area has yielded contradictory re- McNeil, N., J. Cummings, and W. James. 1978. Short chain fatty
sults (Seiler and Raul, 2005). acid absorption by the human large intestine. Gut 19:819–822.

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018
 3044 Middelbos et al.
Middelbos, I. S., M. R. Godoy, N. D. Fastinger, and G. C. Fahey Jr. dog and cat fecal inoculum: Influence of diet composition on
2007. A dose-response evaluation of spray-dried yeast cell wall substrate organic matter disappearance and short-chain fatty
supplementation of diets fed to adult dogs: Effects on nutrient acid production. J. Anim. Sci. 73:1110–1122.
digestibility, immune indices, and fecal microbial populations. Sunvold, G. D., G. C. Fahey Jr., N. R. Merchen, E. C. Titgemeyer,
J. Anim. Sci. 85:3022–3032. L. D. Bourquin, L. L. L. Bauer, and G. A. Reinhart. 1995b.
Milovic, V., and L. Turchanowa. 2003. Polyamines and colon cancer. Dietary fiber for dogs. IV. In vitro fermentation of selected fiber
Biochem. Soc. Trans. 31:381–383. sources by dog fecal inoculum and in vivo digestion and metabo-
Muir, H. E., S. M. Murray, G. C. Fahey Jr., N. R. Merchen, and G. lism of fiber-supplemented diets. J. Anim. Sci. 73:1099–1109.
A. Reinhart. 1996. Nutrient digestion by ileal cannulated dogs Sunvold, G. D., H. S. Hussein, G. C. Fahey Jr., N. R. Merchen, and
as affected by dietary fibers with various fermentation character- G. A. Reinhart. 1995c. In vitro fermentation of cellulose, beet
istics. J. Anim. Sci. 74:1641–1648.
pulp, citrus pulp, and citrus pectin using fecal inoculum from
Nara, P. L., K. Winter, J. B. Rice, R. G. Olsen, and S. Krakowda.
cats, dogs, horses, humans, and pigs and ruminal fluid from
1983. Systemic and local intestinal antibody response in dogs
cattle. J. Anim. Sci. 73:3639–3648.
given both infective and inactivated canine parvovirus. Am. J.
Swanson, K. S., C. M. Grieshop, E. A. Flickinger, L. L. Bauer, J.
Vet. Res. 44:1989–1995.
Chow, B. W. Wolf, K. A. Garleb, and G. C. Fahey Jr. 2002a.
NRC. 2006. Nutrient requirements of dogs and cats. Natl. Acad.
Press, Washington, DC. Fructooligosaccharides and Lactobacillus acidophilus modify
O’Carra, R. 1997. An assessment of the potential of mannanoligosac- gut microbial populations, total tract nutrient digestibilities and
charides as immunostimulants. MS Thesis. Natl. Univ. Ireland, fecal protein catabolite concentrations in healthy adult dogs. J.
Galway, Ireland. Nutr. 132:3721–3731.
Propst, E. L., E. A. Flickinger, L. L. Bauer, N. R. Merchen, and G. Swanson, K. S., C. M. Grieshop, E. A. Flickinger, L. L. Bauer, H. P.
C. Fahey Jr. 2003. A dose-response experiment evaluating the Healy, K. A. Dawson, N. R. Merchen, and G. C. Fahey Jr. 2002b.
effects of oligofructose and inulin on nutrient digestibility, stool Supplemental fructooligosaccharides and mannanoligosacchar-
quality, and fecal protein catabolites in healthy adult dogs. J. ides influence immune function, ileal and total tract nutrient
Anim. Sci. 81:3057–3066. digestibilities, microbial populations and concentrations of pro-
Prosky, L., N. G. Asp, I. Furda, J. W. DeVries, T. F. Schweizer, and tein catabolites in the large bowel of dogs. J. Nutr. 132:980–989.
B. F. Harland. 1984. Determination of total dietary fiber in foods Swanson, K. S., C. M. Grieshop, E. A. Flickinger, H. P. Healy, K. A.
and products: Collaborative study. J. Assoc. Off. Anal. Chem. Dawson, N. R. Merchen, and G. C. Fahey Jr. 2002c. Effects of
67:1044–1052. supplemental fructooligosaccharides plus mannanoligosacchar-
Prosky, L., N. G. Asp, T. F. Schweizer, J. W. de Vries, and I. Furda. ides on immune function and ileal and fecal microbial popula-
1992. Determination of insoluble and soluble dietary fiber in tions in adult dogs. Arch. Anim. Nutr. 56:309–318.
foods and food products: Collaborative study. J. Assoc. Off. Anal. Vickers, R. J., G. D. Sunvold, R. L. Kelley, and G. A. Reinhart. 2001.
Chem. 75:360–366. Comparison of fermentation of selected fructooligosaccharides
Roediger, W. E. W. 1995. The place of short-chain fatty acids in and other fiber substrates by canine colonic microflora. Am. J.
colonocyte metabolism in health and ulcerative colitis: The im- Vet. Res. 62:609–615.
paired colonocyte barrier. Pages 337–351 in Physiological and Walker, J. A., D. L. Harmon, K. L. Gross, and G. F. Collings. 1994.
Clinical Aspects of Short-Chain Fatty Acids. J. H. Cummings,
Evaluation of nutrient utilization in the canine using the ileal
J. L. Rombeau, and T. Sakata, ed. Cambridge Univ. Press, Cam-
cannulation technique. J. Nutr. 124:2672S–2676S.
bridge, UK.
Wang, J. Y., and L. R. Johnson. 1990. Luminal polyamines stimulate
Russell, J., and P. Bass. 1985. Canine gastric emptying of fiber meals:
repair of gastric mucosal stress ulcers. Am. J. Physiol. Gastroint-
Influence of meal viscosity and antroduodenal motility. Am. J.
Physiol. Gastrointest. Liver Physiol. 249:G662–G667. est. Liver Physiol. 259:G584–G592.
Seiler, N., and F. Raul. 2005. Polyamines and apoptosis. J. Cell. Mol. Wang, J. Y., S. A. McCormack, M. J. Viar, and L. R. Johnson. 1991.
Med. 9:623–642. Stimulation of proximal small intestinal mucosal growth by lu-
Silvio, J., D. L. Harmon, K. L. Gross, and K. R. McLeod. 2000. Influ- minal polyamines. Am. J. Physiol. Gastrointest. Liver Physiol.
ence of fiber fermentability on nutrient digestion in the dog. 261:G504–G511.
Nutrition 16:289–295. Williams, C. H., D. J. David, and O. Iismaa. 1962. The determination
Sunvold, G. D., G. C. Fahey Jr., N. R. Merchen, and G. A. Reinhart. of chromic oxide in feces samples by atomic absorption spectro-
1995a. In vitro fermentation of selected fibrous substrates by photometry. J. Agric. Sci. 59:381–385.

Downloaded from https://academic.oup.com/jas/article-abstract/85/11/3033/4779189

by guest
on 01 September 2018