Diet and Measurement Techniques Affect Small Intestinal Digesta Viscosity Among Dogs

10
Diet and measurement techniques affect small intestinal digesta viscosity among dogs Cheryl L. Dikeman, Kathleen A. Barry, Michael R. Murphy, George C. Fahey Jr 4 Division of Nutritional Sciences and Department of Animal Sciences, University of Illinois, Urbana, IL 61801, USA Received 8 February 2006; revised 20 December 2006; accepted 21 December 2006 Abstract Because increasing viscosity of gastrointestinal tract contents has been shown to alter physiologic responses in many species, and no single standardized method of intestinal viscosity measurement exists, effects of diet and measurement techniques were studied in a canine model. Three experiments were conducted to evaluate the use of mixer viscometry, effects of freezing, centrifugation, time of sampling, and dilution of digesta, and effects of diet on viscosity of canine ileal digesta and simulated small intestinal digesta viscosity. Digesta viscosity values measured at 218C were within 5% of those at 218C after a 24-hour freeze. Viscosity constants of whole digesta sampled between 8:00 am and 2:00 pm daily ranged from 6575 to 32 692 mPa d s and 7.47 to 9.03 mPa d s after centrifugation. Digesta sampled between 2:00 pm and 8:00 pm daily had viscosity constants of whole digesta ranging from 7137 to 17 345 mPa d s and 6.42 and 9.46 mPa d s after centrifugation. Digesta samples diluted with Millipore filtered water, to determine whether a dilution factor could be used to estimate viscosity, had drastically underestimated viscosity constants. Ileal digesta viscosity constants were similar for dogs fed test diets varying in carbohydrate source and ranged from 6901 to 12 590 mPa d s. During simulated small intestinal simulation, viscosity peaked between 6 and 9 hours. Data indicate that alteration of digesta by centrifugation or dilution underestimates viscosity data. Viscosity of whole digesta can be measured using mixer viscometry. Variations in diet ingredients appear to alter intestinal digesta viscosity, indicating further need for investigation of the role of diet in viscosity. D 2007 Elsevier Inc. All rights reserved. Keywords: Viscosity; Digesta; Mixer viscometry; Canine; Intestine 1. Introduction Increasing the viscosity of digesta in the gastrointestinal tract through diet manipulation with dietary fiber has been shown to significantly reduce postprandial glucose response in humans, pigs, canines, and rodents, indicating a potential mechanism for management of diabetes and abnormal carbohydrate metabolism [1-4]. Although it has been demonstrated that ingestion of viscous fibers added to the diets of animals (humans, canines, pigs, rodents) will lower postprandial blood glucose concentration and increase viscosity of intestinal contents, there is no single standard- ized method to measure viscosity of gastrointestinal tract contents. Interpretation of digesta viscosity data is difficult owing to lack of standardized methodology and presentation of data. In the case of fluid digesta from the stomach and small intestine of monogastric animals, an evaluation of viscosity is a complex and difficult task. The general equation to calculate viscosity is shear stress divided by shear rate. Shear rate is the velocity gradient established in a particular fluid due to an applied shear stress, such as a contraction in the gastrointestinal tract of an animal [5]. Gastrointestinal 0271-5317/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.nutres.2006.12.005 4 Corresponding author. Tel.: +1 217 333 2361; fax: +1 217 333 7861. E-mail address: [email protected] (G.C. Fahey). Nutrition Research 27 (2007) 56 – 65 www.elsevier.com/locate/nutres

description

Importance of viscosity in dogs food

Transcript of Diet and Measurement Techniques Affect Small Intestinal Digesta Viscosity Among Dogs

  • en

    a v

    Mimal S

    ecem

    ne

    Nutrition Research 27diets of animals (humans, canines, pigs, rodents) will lower1. Introduction

    Increasing the viscosity of digesta in the gastrointestinal

    tract through diet manipulation with dietary fiber has been

    shown to significantly reduce postprandial glucose response

    in humans, pigs, canines, and rodents, indicating a potential

    mechanism for management of diabetes and abnormal

    carbohydrate metabolism [1-4]. Although it has been

    demonstrated that ingestion of viscous fibers added to the

    postprandial blood glucose concentration and increase

    viscosity of intestinal contents, there is no single standard-

    ized method to measure viscosity of gastrointestinal tract

    contents. Interpretation of digesta viscosity data is difficult

    owing to lack of standardized methodology and presentation

    of data.

    In the case of fluid digesta from the stomach and small

    intestine of monogastric animals, an evaluation of viscosity

    is a complex and difficult task. The general equation toKeywords: Viscosity; Digesta; Mixer viscometry; Canine; Intestiinvestigation of the role of diet in viscosity.

    D 2007 Elsevier Inc. All rights reserved.0271-5317/$ see fro

    doi:10.1016/j.nutres.2

    4 CorrespondingE-mail address: gBecause increasing viscosity of gastrointestinal tract contents has been shown to alter physiologic

    responses in many species, and no single standardized method of intestinal viscosity measurement

    exists, effects of diet and measurement techniques were studied in a canine model. Three

    experiments were conducted to evaluate the use of mixer viscometry, effects of freezing,

    centrifugation, time of sampling, and dilution of digesta, and effects of diet on viscosity of canine

    ileal digesta and simulated small intestinal digesta viscosity. Digesta viscosity values measured at

    218C were within 5% of those at 218C after a 24-hour freeze. Viscosity constants of whole digestasampled between 8:00 am and 2:00 pm daily ranged from 6575 to 32692 mPa d s and 7.47 to 9.03mPa d s after centrifugation. Digesta sampled between 2:00 pm and 8:00 pm daily had viscosityconstants of whole digesta ranging from 7137 to 17345 mPa d s and 6.42 and 9.46 mPa d s aftercentrifugation. Digesta samples diluted with Millipore filtered water, to determine whether a dilution

    factor could be used to estimate viscosity, had drastically underestimated viscosity constants. Ileal

    digesta viscosity constants were similar for dogs fed test diets varying in carbohydrate source and

    ranged from 6901 to 12590 mPa d s. During simulated small intestinal simulation, viscosity peakedbetween 6 and 9 hours. Data indicate that alteration of digesta by centrifugation or dilution

    underestimates viscosity data. Viscosity of whole digesta can be measured using mixer viscometry.

    Variations in diet ingredients appear to alter intestinal digesta viscosity, indicating further need forAbstractDiet and measurem

    small intestinal digest

    Cheryl L. Dikeman, Kathleen A. Barry,Division of Nutritional Sciences and Department of Ani

    Received 8 February 2006; revised 20 Dnt matter D 2007 Elsevier Inc. All rights reserved.

    006.12.005

    author. Tel.: +1 217 333 2361; fax: +1 217 333 7861.

    [email protected] (G.C. Fahey).t techniques affect

    iscosity among dogs

    chael R. Murphy, George C. Fahey Jr4ciences, University of Illinois, Urbana, IL 61801, USA

    ber 2006; accepted 21 December 2006

    (2007) 5665

    www.elsevier.com/locate/nutrescalculate viscosity is shear stress divided by shear rate.

    Shear rate is the velocity gradient established in a particularfluid due to an applied shear stress, such as a contraction in

    the gastrointestinal tract of an animal [5]. Gastrointestinal

  • Approximately 30 mL of fresh ileal digesta was placed in

    ition Rtract shear rates in animals have not been established and

    may vary considerably with sampling location, individual

    animal, meal composition, and gut motility. It can be

    assumed that digesta is a slurry exhibiting non-Newtonian

    shear-thinning behavior. According to this flow behavior, as

    shear rate increases, viscosity decreases; therefore, multiple

    shear rates should be used to measure viscosity. Many

    researchers, however, have measured viscosity of digesta

    from experimental animals at only 1 shear rate, making

    comparisons among studies difficult [1,6-11]. Because a

    vast range of shear rates was used, comparing the resultant

    digesta viscosities is neither useful nor valid. Flow profiles

    should be obtained for liquid digesta across various shear

    rates rather than estimation of viscosity at only 1 shear

    rate [12,13].

    Another factor contributing to the difficulty in measuring

    viscosity of digesta is concentration and size of undigested

    particles in the sample. Particulate matter may interfere with

    viscosity instrumentation and, therefore, many researchers

    centrifuge and (or) dilute digesta samples with water before

    measuring viscosity to remove large particles or dilute them.

    Removal of particulate matter results in a more homogenous

    sample [7,14-17]; however, this may remove particles that

    contribute to viscosity. A solution to this problem may be

    the use of mixer viscometry to measure the viscosity of

    whole digesta. Mixer viscometry was developed to address

    problems including large particle size and (or) particles

    settling out of solution. Mixing involves intermingling of

    2 or more dissimilar materials to obtain a desired degree of

    uniformity. This uniformity is obtained by agitation to create

    motion in the material, usually in the form of a rotational

    viscometer with some form of mixing or vane spindles

    [5,18]. Similar to other rotational methods such as cone and

    plate geometries, mixer viscometry is limited by a limited

    shear rate range.

    To establish standard techniques to measure viscosity of

    digesta collected from the gastrointestinal tract of animals, it

    is necessary to understand the natural variation that exists

    among individual animals and the factors that affect this

    variation. An understanding of factors affecting viscosity

    will be valuable in controlling animal variation in future

    studies designed to measure viscosity of gastrointestinal con-

    tents as well as in aiding in accurate viscosity measurement.

    Although there is a great deal of literature available

    discussing the effects of viscous fibers on the viscosity of

    fluids such as gastric, small intestinal, and cecal contents

    from animals, few data are available on the effects of

    standard diets on digesta viscosity. The vast majority of

    intestinal content viscosity has been measured after the

    removal of solid particles through methods such as

    centrifugation and dilution. A repeatable measurement of

    viscosity on whole digesta would provide an excellent

    opportunity to explore the effects of viscosity changes in the

    gastrointestinal tract as affected by diet, ingredient(s), or diet

    C.L. Dikeman et al. / Nutrmatrix, and allow for further investigations of links between

    viscosity of gastric and intestinal contents and physiologic100-mL glass beakers with a diameter of 5 cm and allowed

    to equilibrate to room temperature (238C). Although thistemperature is significantly lower than physiologic temper-

    ature, it was necessary to avoid confusion between multiple

    viscometer configurations where temperature control couldresponses such as digestibility, gut morphology, transit time,

    blood glucose and cholesterol attenuation, and bowel health.

    The objectives of this study were to determine the effects of

    freezing, centrifugation, dilution, and various carbohydrate

    sources in diets on intestinal viscosity of dogs.

    2. Methods and materials

    2.1. Experiment 1

    2.1.1. Animals

    Purpose-bred female dogs (n = 3; Butler Farms USA,

    Clyde, NY) with hound bloodlines, an average initial body

    weight of 27.9 kg (range, 27-28.9 kg), and an average age

    of 3.7 years (range, 1.7-6.5 years) were used. Dogs had

    previously been surgically prepared with an ileal cannula

    according to Walker et al [19] with surgical and animal care

    procedures approved by the University of Illinois Animal

    Care and Use Committee. Dogs were housed individually in

    kennels in a temperature-controlled room (218C) at theanimal care facility in the Edward R Madigan Laboratory,

    University of Illinois. A 16-hour light/8-hour dark schedule

    was used.

    2.1.2. Procedures

    Dogs were fed 400 g of the same commercial diet (Iams

    Weight Control, The Iams Company, Dayton, OH) daily in

    2 equal feedings of 200 g at 8:00 am and 8:00 pm. The

    main ingredients of the diet were corn meal, chicken,

    ground whole grain sorghum, and chicken by-product meal.

    The dogs had been on this diet for several months before the

    initiation of this experiment. Water was available ad libitum.

    Ileal samples were collected for 1 hour between 10:00 am

    and 11:00 am on 3 consecutive days. Ileal samples were

    collected by attaching a sterile sampling bag (Fisher

    Scientific, Pittsburgh, Pa) to the cannula barrel and around

    the hose clamp with a rubber band. Before attachment of the

    bag, the interior of the cannula was scraped clean with a

    spatula and digesta discarded. During the collection of ileal

    effluent, the dogs were encouraged to move about freely. To

    deter the dogs from pulling the collection bag from the

    cannula, Bite-Not collars (Bite-Not Products, San Francisco,

    Calif) were used during collection times. After ileal effluent

    collection, the cannula plug was put in place and the cannula

    barrel and surrounding site were cleaned with a 10%

    Betadine solution (Purdue Frederick Co, Stamford, Conn).

    2.1.3. Viscosity

    esearch 27 (2007) 5665 57not be used. All viscosity measures in the study were

    determined at room temperature to alleviate any additional

  • ition Rvariation due to temperature. It would be assumed that

    viscosity measured at physiologic temperature would be

    lower than the viscosities presented in this study. In

    addition, there are data to suggest that gas bubbles due to

    continued fermentation at temperatures above 258C wouldlikely alter viscosity [20]. Ileal samples were mixed by hand

    with a metal spatula for 60 seconds to homogenize the

    sample. Viscosity was measured in duplicate with a Brook-

    field LV-DV-II+ viscometer. An SSV-vane standard spindle

    set (Brookfield Engineering, Middleboro, Mass) was used.

    Spindle multiplier constants used were 2.62, 11.1, and 53.5

    for spindles V-71, V-72, and V-73, respectively. These

    constants were programmed into Wingather software

    (Brookfield Engineering) for viscosity calculations. Viscos-

    ity was calculated as a function of revolutions per minute

    (rpm) over an rpm range of 0.3 to 10 (0.3, 0.6, 1, 2, 5, 10).

    After initial measurement, samples were frozen at 208C.After 24 hours in the freezer, samples were removed and

    allowed to equilibrate to 238C and viscosity measured aspreviously described.

    2.2. Experiment 2

    2.2.1. Animals

    Purpose-bred female dogs (n = 6; Butler Farms USA)

    with hound bloodlines, an average initial body weight of

    26.1 kg (range, 19.4-28.2 kg), and an average age of

    4.2 years (range, 3.0-7.8 years) were used. Dogs had

    previously been surgically prepared with an ileal cannula

    and were housed as previously described. Dogs were fed the

    same commercial diet as was used in Experiment 1. The diet

    met or exceeded the Association of American Feed Control

    Officials [21] recommendations for dogs at weight mainte-

    nance and had an ME concentration of 20628 kJ/kg. Dogs

    were fed a total of 400 g of each diet divided equally

    between 2 feedings occurring at 8:00 am and 8:00 pm. This

    study was conducted over a 14-day period. Initially, a 6-day

    diet adaptation phase preceded an 8-day collection of ileal

    digesta. Ileal digesta was collected once daily during the

    8-day collection period, and each sample collection lasted

    1.5 hours in length. Sample times rotated by 1.5 hours each

    day with the first-day collection beginning at 8:00 am.

    Collected ileal digesta was immediately frozen at 208C.Before the viscosity measurement, ileal digesta was

    composited for days 1 to 4 (8:00 am-2:00 pm) and days

    5 to 8 (2:00 pm-8:00 pm) and thawed in a 558C water bathuntil samples reached room temperature (238C). Water wasoffered ad libitum by providing 1500 mL at 8:00 am and

    8:00 pm daily into 2000-mL stainless steel bowls attached

    to the cages. Before offering new water at 8:00 am and

    8:00 pm, remaining water from the previous offering was

    measured with a graduated cylinder and amounts recorded.

    2.2.2. Viscosity

    C.L. Dikeman et al. / Nutr58Viscosity was measured in duplicate as described in

    Experiment 1. After the initial viscosity measurement, ilealdigesta samples were placed in 50-mL centrifuge tubes and

    centrifuged at 11000 g for 10 minutes. Viscosity of thesupernatant (2 mL) was measured with a Brookfield LV-DV-

    II+ viscometer with a Wells/Brookfield cone and plate using

    a CP-41 cone and plate.

    2.3. Experiment 3

    2.3.1. Animals

    Purpose-bred female dogs (n = 5; Butler Farms USA)

    with hound bloodlines, an average initial body weight of

    27.9 kg (range, 19.3-32.3 kg), and an average age of

    3.2 years (range, 1.8-6.5 years) were used. Dogs had

    previously been surgically prepared with an ileal cannula

    and were housed as described previously.

    2.3.2. Procedures

    Dogs were fed 1 of 5 dietary treatments that consisted

    of commercially available canine extruded diets. Data

    have shown that various carbohydrates impact the

    viscosity of intestinal fluid of many species including dogs

    [2,4,6,8-10,13]. Therefore, the diets chosen differed in

    carbohydrate ingredients and included either ground wheat

    (GW) (Kibbles n Bits Original, DLM Foods L.L.C, San

    Francisco, CA), brewers rice (BR) (Purina One, Nestle

    Purina, St Louis, Mo), corn meal (CM) (Eukanuba Adult

    Premium Performance Formula, The Iams Company), corn

    meal + ground sorghum + ground wheat (Mixed) (Science

    Diet Adult Original, Hills Pet Nutrition, Inc, Topeka, Ks),

    or oatmeal (OatM) (Cycle Custom Fitness, DLM Foods,

    LLC, San Francisco, Calif). Dogs were fed a total of 500 g

    of each diet daily divided equally between 2 feedings

    occurring at 8:00 am and 8:00 pm.

    The experimental design was a 5 5 Latin squareconsisting of 10-day periods. A 6-day adaptation phase

    preceded a 4-day collection of ileal digesta. Ileal digesta was

    collected 3 times a day, with an interval of 4 hours between

    collections. Ileal collections were 1 hour in length.

    Sampling times on the remaining 3 days rotated 1 hour

    from the previous days collection time. Ileal samples were

    collected as described above. Ileal digesta collected at each

    time was immediately frozen at 208C. At the end of eachperiod, ileal digesta was composited for each individual

    dog. Composited ileal samples were allowed to thaw in a

    558C water bath until they reached room temperature(238C). Samples were mixed by hand with a metal spatulafor 60 seconds. Approximately 30-mL duplicates of ileal

    digesta were placed in glass beakers (5 cm in diameter).

    Viscosity of ileal digesta was analyzed as described

    above. If samples were too viscous to measure directly, they

    were diluted 1:1 (w/w basis) with Millipore filtered water

    (viscosity, 1 mPa d s).

    2.3.3. In vitro digestion simulation

    esearch 27 (2007) 5665For the in vitro digestion simulation, diets from

    Experiment 3 were weighed (0.5 g) in duplicate and placed

  • 10.3 8.6 1.7 20628

    die

    ition Rin 50-mL plastic centrifuge tubes. To simulate gastric

    digestion, 0.2 N HCl (5 mL), 10% pepsin/HCl (w/v,

    0.5 mL), and 0.1 mol/L (12.5 mL) phosphate buffer (pH 6)

    were added to each tube. Solutions were adjusted to pH 2

    with HCl (0.2 N) or NaOH (0.6 N). Tubes were stoppered

    and incubated for 6 hours at 398C [22,23]. After the initial6-hour incubation, small intestinal simulation began with

    the addition of 0.6 N NaOH (2.5 mL), 0.2 mol/L phosphate

    buffer (pH 6.8, 5 mL), and 5% pancreatin solution (w/v,

    0.5 mL), with adjustment to pH 6.8 with HCl (0.2 N) or

    NaOH (0.6 N) [22,23]. Tubes were incubated at 398C for anadditional 18 hours. One set of substrates was removed from

    incubation and frozen at 208C at 0, 3, 6, 9, 12, 15, and18 hours from initiation of small intestinal digestion

    simulation. During the in vitro digestion simulation, vis-

    cosity was measured using a Brookfield digital viscometer

    (LV-DV-II+) with a Wells/Brookfield cone and plate

    extension. Solutions were assayed using a CP-41 cone and

    plate, and across rpm speeds of 0.3, 0.6, 1, 1.5, 2, and 3

    (shear rates, 0.6, 1.0, 2, 3, 4, and 6 per second, respectively).

    2.4. Chemical analyses

    Diets used in all 4 experiments were analyzed for dry

    matter (DM), organic matter (OM), and ash using the

    Association of Official Analytical Chemists [24] methods.

    Crude protein (CP) concentrations were calculated using

    Leco nitrogen (N) values (N 6.25) for all samples usingAOAC [24] methods. Total lipid content was determined by

    acid hydrolysis followed by ether extraction according to

    the American Association of Cereal Chemists [25] and

    Budde [26]. Total dietary fiber (TDF), soluble (SDF), and

    Table 1

    Chemical analyses of dietary treatmentsa

    Diet DM OM CP AHF

    %, Dry matter basis

    Diet A 91.0 93.4 23.4 11.7

    BR 93.3 92.8 28.2 19.4

    CM 93.3 91.7 33.5 22.6

    GW 91.1 91.6 24.9 11.9

    Mixed 92.5 95.0 26.8 15.4

    OatM 92.2 93.9 24.7 13.1

    a Diet A used in Experiments 1 and 2; BR, CM, GW, Mixed, and OatM

    C.L. Dikeman et al. / Nutrinsoluble dietary fiber (IDF) concentrations were deter-

    mined according to Prosky et al [27]. Gross energy content

    of the diets was determined by bomb calorimetry (Parr

    Instrument Co, Moline, Ill).

    2.5. Statistical analysis

    Viscosity data were analyzed using NLREG software

    (NLREG, Brentwood, Tenn) and SAS (SAS Institute,

    Cary, NC). NLREG was used to develop a working

    model of the viscosity flow curve data. Pseudoplastic

    fluids can be represented adequately by the power law

    equation ( y = a * xb) and termed power law fluids

    [5,28-31]. In the equation, shear stress ( y) is a function ofthe consistency index or constant (a), shear rate (x), and a

    dimensionless exponent (b) that indicates closeness to

    Newtonian flow. The exponent will equal 1 for New-

    tonian fluids and will be less than 1 for shear-thinning or

    pseudoplastic fluids. The constant is a parameter propor-

    tional to viscosity of power law fluids and is represented

    in units of millipascals per second. Model development

    allowed for the estimation of the constant and exponent

    parameters in the above equation.

    Viscosity data from Experiment 1 were analyzed for

    NLREG parameters using NLREG software.

    Nonlinear regression parameter data from Experiment 2

    were analyzed using the Mixed models procedure of SAS

    and included the fixed effect of treatment (whole or

    centrifuged digesta) and the random effect of dog. Treat-

    ment least squares means were compared using the

    Bonferroni method to control experimentwise error rate.

    A probability of P b .05 was accepted as statisticallysignificant. Data also were analyzed using the correlation

    procedure of SAS to determine the linear relationships

    between diet and water intake and viscosity parameters.

    Data obtained from Experiment 3 were analyzed to

    estimate nonlinear regression parameters using NLREG

    software and SAS. The experimental design was a 5 5 Latin square. The statistical model included the fixed

    effect of diet and the random effects of dog and period.

    In vitro data did not meet criteria of normality tested by

    the univariate procedure of SAS; therefore, data were log-

    transformed before the statistical analysis. Data were

    analyzed using the Mixed models procedure of SAS. The

    experimental design was a factorial randomized complete

    5.4 3.3 2.1 22190

    8.5 5.9 2.6 22948

    10.9 9.3 1.6 20105

    6.7 5.1 1.6 21436

    9.8 8.4 1.4 20704

    ts used in Experiments 3 and for in vitro digestion simulation.TDF IDF SDF Gross energy, kJ/kg

    esearch 27 (2007) 5665 59block design with diet serving as block. The statistical

    model included the fixed effect of diet and the random effect

    of replicate. Treatment least squares means were compared

    using the Bonferroni method to control experimentwise

    error rate. A probability of P b .05 was accepted asstatistically significant.

    3. Results

    3.1. Chemical analyses

    Chemical analyses of diets tested in all 3 experiments are

    presented in Table 1. Dry matter and OM concentrations

  • 24-hour freeze from 3 dogs consuming a standard dieta

    Frozen digesta

    R2 Viscosity constant, mPa d s Exponent R2

    0.99 13241 1.18 0.990.99 16084 1.31 0.990.99 15171 1.62 0.990.99 24878 1.41 0.990.99 17503 1.87 0.990.99 10846 1.50 0.990.99 12410 1.23 0.990.99 22270 1.00 0.990.99 16301 1.69 0.99

    = a * xb), where y is shear stress, a is the viscosity constant, x is shear rate, and

    exponent b1 for pseudoplastic fluids); R2 indicates the proportion of variation

    ition Research 27 (2007) 5665were similar among diets and averaged 92.2% and 93.1%,

    respectively. Crude protein and acid-hydrolyzed fat (AHF)

    concentrations ranged from 23.4% to 33.5% and 11.7% to

    22.6%, respectively. Total dietary fiber, SDF, and IDF

    concentrations averaged 8.6%, 1.8%, and 6.8%, respective-

    ly. The gross energy concentration of the 6 diets analyzed

    ranged from 20105 to 22948 kJ/kg.

    3.2. Experiment 1

    Nonlinear regression viscosity constants, exponents, and

    R2 values calculated for ileal digesta samples measured

    fresh and after a 24-hour freeze/thaw are presented in

    Table 2. Ileal digesta viscosity constants averaged 15052,

    17116, and 16614 mPa d s for dogs 1, 2, and 3, res-pectively, over the 3-day testing period. After a 24-hour

    freeze/thaw, viscosity values were within 5% of each other

    with the exception of dog 2 day 2 and dog 3 day 2. A high

    proportion of variation among digesta samples from dogs

    over the 3-day sampling period was accounted for with

    the nonlinear regression model based on high R2 values

    (Table 2). All digesta solutions sampled during the 3-day

    period exhibited non-Newtonian shear-thinning behavior

    (decreasing viscosity with increasing shear rate) indicated

    Table 2

    Nonlinear regression viscosity parameters for ileal digesta fresh and after a

    Dog Day Fresh digesta

    Viscosity constant, mPa d s Exponent

    1 1 13931 1.262 15563 1.303 15663 1.77

    2 1 23726 1.442 16206 1.913 11416 1.52

    3 1 12512 1.222 20813 0.833 16518 1.71

    a Nonlinear viscosity parameters are based on the power law equation ( y

    b is a dimensionless exponent indicating deviation from Newtonian flow (

    explained by the nonlinear regression model.

    C.L. Dikeman et al. / Nutr60by negative nonlinear regression exponent values ranging

    from 0.83 to 1.91.3.3. Experiment 2

    Dry matter intake (DMI), water intake, and nonlinear

    regression viscosity constant averages for whole and centri-

    fuged digesta samples are presented in Table 3. Nonlinear

    regression viscosity constants calculated for whole digesta

    ranged from a high of 22588 to a low of 7631 mPa d s.After centrifugation, viscosity constants were very low and

    ranged from a high of 9.23 to a low of 6.96 mPa d s. Nosignificant linear correlation was detected between DMI,

    water intake, and viscosity constants.

    Nonlinear regression viscosity parameters for digesta

    samples are presented in Table 4. When viscosity was

    measured on whole digesta samples, nonlinear regression

    constants ranged from 6575 to 32692 mPa d s duringmorning collections and 7137 to 17345 mPa d s duringafternoon collections. After centrifugation, nonlinear

    regression viscosity constants were lower (P b .05) thanwhole samples and ranged from 7.47 to 9.03 mPa d sduring morning collections and 6.42 to 9.46 mPa d sduring afternoon collections. In addition, the nonlinear

    regression model successfully explained 99% of the

    variation among samples measured whole, indicated by

    high R2 values; however, the model failed to successfully

    explain the variation in 5 of 6 samples after centrifuga-

    tion. Those samples that were not significantly modeled

    using nonlinear regression were labeled with exponent

    values of 1, indicating a lack of dependence of shear rate

    on viscosity.

    3.4. Experiment 3

    Nonlinear regression viscosity parameters, DMI, and

    ileal DM concentrations for dogs consuming diets varying

    in carbohydrate ingredients are presented in Table 5. Ileal

    DM concentration ranged from 11.7% to 14.6% and was

    highest (P b .05) for dogs fed CM compared with dogs fedOatM and GW. No additional differences were detected.

    During Experiment 3, 5 digesta samples were too viscousto measure using the Brookfield LV-DV-II+ viscometer

    Table 3

    Dry matter intake, water intake, and nonlinear regression viscosity

    constants for whole and centrifuged ileal digestaa,b

    Dog DMI, g/d Water intake, mL/d Viscosity constant, mPa d s

    Whole Centrifuged

    1 313 1066 15711 8.04

    2 296 1022 8808 7.70

    3 365 1095 22588 7.86

    4 232 568 7631 7.64

    5 361 902 12360 9.23

    6 198 628 10104 6.96

    a All nonlinear regression viscosity constants for whole digesta were

    higher ( P b .05) compared with centrifuged digesta (pooled SEM1397.25).

    b Correlations between DMI, water intake, and nonlinear viscosity

    constants measured on whole and centrifuged digesta were not significan

    ( P N .05).;

    ,

    t

  • Table 4

    Nonlinear regression viscosity parameters for digesta measured whole and after centrifugation in morning and afternoon hours from 6 dogs consuming a

    standard dieta,b,c,d

    Dog Morning hours Afternoon hours

    Whole digesta Centrifuged digesta Whole digesta Centrifuged digesta

    1

    Constant, mPa d s 14077 9.03 17345 7.05Exponent 1.58 1 1.20 0.61R2 0.99 NS 0.99 0.84

    2

    Constant, mPa d s 6575 7.47 11041 7.94

    C.L. Dikeman et al. / Nutrition Research 27 (2007) 5665 61Exponent 1.38 1R2 0.99 NS

    3

    Constant, mPa d s 32692 8.67Exponent 1.27 0.96R2 0.99 0.98

    4

    Constant, mPa d s 8124 8.23Exponent 1.29 1R2 0.99 NS

    5

    Constant, mPa d s 12236 8.99Exponent 1.46 1R2 0.99 NS

    6therefore, they were diluted with Millipore filtered water on

    a 1:1 (w/w) basis. After measurement and application of the

    appropriate dilution factor, nonlinear regression viscosity

    constants were much lower than expected, ranging from

    1585 to 5894 mPa d s (Table 6). These values should havebeen greater than the highest values calculated without

    dilution (Napproximately 22,000) based on the maximummeasurement capacity of the viscometer. Three of the

    diluted samples were obtained from the same dog consum-

    ing the CM, Mixed, and OatM diets. Samples obtained from

    one dog consuming the CM diet and one dog consuming

    the Mixed diet also were diluted. Because of the drastic

    Constant, mPa d s 10283 7.50Exponent 1.38 1R2 0.99 NS

    a Nonlinear viscosity parameters are based on the power law equation ( y = a

    b is a dimensionless exponent indicating deviation from Newtonian flow (expo

    regression was not significant (NS; P N .05); R2 indicates the proportion of varib Constant mean estimates were used in the case of nonsignificant nonlinearc Ileal digesta sampled in the morning hours between 8:00 am and 2:00 pm and All nonlinear regression viscosity constants for whole digesta were higher

    Table 5

    Nonlinear regression viscosity parameters, apparent total tract dry matter and o

    consuming diets containing various carbohydrate ingredients1,2

    Item BR CM GW

    Viscosity constant, mPa d s 12590 11291 9431Exponent 1.32 1.19 1.19Dry matter intake, g/d 358 353 303

    Ileal DM, % 13.0a,b 14.6a 12.4b

    a,b Least squares means in the same row that do not have common superscri1 Nonlinear viscosity parameters are based on the power law equation ( y = a

    b is a dimensionless exponent indicating deviation from Newtonian flow (expon2 Diluted samples were not included in the statistical analysis.1.46 10.99 NS

    12483 7.06

    1.30 10.99 NS

    7137 7.05

    1.36 10.99 NS

    12483 9.46

    1.70 10.99 NSreduction in viscosity, those 5 samples were not included in

    the statistical analysis.

    3.5. In vitro digestion simulation

    Nonlinear regression viscosity parameters for smallintestinal digestion simulations are presented in Table 7.

    During small intestinal digestion simulation, the BR

    treatment had a higher (P b .05) viscosity constant at6 hours compared with 9 and 15 hours.

    Viscosity constants for solutions containing CM were

    similar during the initial 6 hours of simulation and averaged

    221 mPa d s; however, there was a reduction (P b .05) in

    9925 6.42

    1.30 10.99 NS

    * xb), where y is shear stress, a is the viscosity constant, x is shear rate, and

    nent b1 for pseudoplastic fluids); exponents were given as 1 if nonlinearation explained by the nonlinear regression model.

    regression analysis.

    d ileal digesta sampled in the afternoon hours between 2:00 pm and 8:00 pm.

    ( P b .05) compared with centrifuged digesta (pooled SEM, 1397.25).

    rganic matter digestibilities, and ileal dry matter concentrations for dogs

    Mixed OatM SEM P value

    7934 10228 3134.7 NS

    1.21 1.17 0.09 NS372 379 38.17 NS

    12.7a,b 11.7b 0.59 b .05

    pt letters differ ( P b .05).* x2), where y is shear stress, a is the viscosity constant, x is shear rate, and

    ent b1 for pseudoplastic fluids).

  • viscosity beginning at 9 hours. Viscosity constants were

    lower (P b .05) at 15 and 18 hours compared with the6-hour value.

    Few differences were detected for the GW treatment. At

    the initiation of digestion simulation, the viscosity constant

    was lower (P b .05) than the viscosity constant after 6 hoursof digestion simulation. All other time point combinations

    were similar. As was the case for the BR and CM

    treatments, the GW treatment had the highest viscosity

    constant value at 6 hours.

    Viscosity constants for the Mixed treatment were lowest

    (P b .05) at 6 and 15 hours of simulation compared with the9-hour value. No other differences were detected.

    exhibit shear-thinning behavior [3,13,28-32]. Use of a

    working model such as nonlinear regression provides

    additional information about the viscosity/shear rate profile

    not available if a single shear rate measurement is obtained.

    This approach provides a more powerful tool for describing

    viscosity characteristics in the gastrointestinal tract, even

    though shear rates have not been established [13]. In the

    current studies, nonlinear regression analysis indicated that

    all solutions, with the exception of those that were

    centrifuged, were non-Newtonian and exhibited shear-

    thinning behavior as indicated by negative exponents.

    Greater negative exponents are associated with a greater

    dependence of viscosity on shear rate. It was expected that

    ileal digesta collected from dogs would exhibit this type

    of flow behavior because of the presence of particulate

    matter within the fluid matrix. In support, Reppas et al [13]

    reported that chyme collected from the duodenum and

    jejunum of dogs exhibited shear-thinning or pseudoplastic

    flow behavior.

    Multiple methods of viscosity measurement have been

    utilized in previous research to obtain a greater understand-

    ing of this property in the gastrointestinal tract of animals.

    Tube flow and rotational viscometers have been used to

    determine flow properties, and effects of particle size on

    Table 6

    Nonlinear regression viscosity constants calculated for digesta samples of

    dogs fed diets containing various carbohydrate ingredientsa

    Dog BR CM GW Mixed OatM

    1 21926 1585a 5362 9412 9094

    2 9216 9349 5505 6835 8431

    3 10331 3018a 12369 5894a 3026a

    4 4715 6712 7713 7557 2943

    5 16760 17813 16211 4811a 20446

    a Indicates samples diluted 1:1 with Millipore filtered water.

    h vari

    7

    7

    d

    5

    9

    C.L. Dikeman et al. / Nutrition Research 27 (2007) 5665624. Discussion

    In the current studies, fluid viscosities have been

    described using the power law equation to calculate a

    viscosity consistency index or constant that is proportional

    to viscosity. The power law equation describes solutions that

    Table 7

    Nonlinear regression viscosity parameters for solutions containing diets wit

    Sample

    0 3 6

    BR

    Constant, mPa d s 27d 47c,d 130c

    Exponent 1.04 1.53 0.8R2 0.95 0.99 0.9

    CM

    Constant, mPa d s 254c,d 148c,d 262c,

    Exponent 0.97 1.04 0.8R2 0.99 0.99 0.9

    GW

    Constant, mPa d s 14d 63c,d 77cExponent 1.50 0.92 0.78R2 0.99 0.99 0.99

    Mixed

    Constant, mPa d s 33c,d 14d,e 4e

    Exponent 1.30 1.62 2.18R2 0.98 0.96 0.97

    OatM

    Constant, mPa d s 32 29 17Exponent 1.05 0.95 0.69R2 0.99 0.99 0.95

    c, d, e Least squares means (5 diets, 7 time points; n = 35) in the same row that do1 Nonlinear viscosity parameters are based on the power law equation ( y = a

    b is a dimensionless exponent indicating deviation from Newtonian flow (expon

    explained by the nonlinear regression model.digesta collected from pigs, chickens, wallabies, and

    brushtail possums [20,29,30,33-36]. Because of the effect

    of particle size on viscosity, it seems relevant to use a

    method that accounts for particles in fluid. Research has

    been conducted on the application of mixer viscometry to

    measure the viscosity of biologic materials [5,18].

    ous carbohydrate ingredients during small intestinal digestion simulation1

    Time, h

    9 12 15 18

    22d 36c,d 23d 75c,d

    1.19 1.07 0.57 0.840.97 0.99 0.90 0.98

    84c,d 95c,d 34e 43e

    0.98 0.94 1.15 1.100.98 0.99 0.99 0.98

    29c,d 34c,d 31c,d 48c,d

    0.97 1.31 1.52 1.000.98 0.97 0.98 0.93

    89c 31c,d 4e 17d,e

    1.09 1.06 1.24 1.660.99 0.99 0.93 0.98

    18 24 25 26

    1.18 1.06 0.94 0.870.99 0.97 0.96 0.97

    not have common superscript letters differ ( P b .05) (pooled SEM, 21.20).* xb), where y is shear stress, a is the viscosity constant, x is shear rate, and

    2ent b1 for pseudoplastic fluids); R indicates the proportion of variation

  • ition RTo measure the viscosity of whole ileal digesta from dogs

    without altering or removing particles or diluting the digesta

    in any way, mixer viscometry was used. During the initial

    experiment, on average, calculated viscosity constants

    varied only by 12% among dogs. For each dog, there was

    a greater variation among days. Furthermore, with the

    exception of only 2 samples, viscosity values measured after

    a 24-hour freeze and thaw to room temperature were within

    5% of those measured on fresh digesta. Ability to freeze

    digesta before measuring viscosity allows additional con-

    venience when conducting experiments.

    In Experiment 2, there was a 66% variation among dogs.

    The larger variation in Experiment 2 may have been a result

    of the sampling times. During Experiment 1, digesta was

    sampled at the same time on each of the consecutive days.

    In Experiment 2, digesta was sampled at rotating times that

    changed every day during the collection phase.

    During Experiment 2, there was no interaction or

    correlation between DMI, water intake, and the calculated

    viscosity constant. Water intake and diet intake were

    measured but not controlled. Dogs were given a specified

    amount of food and water but were not forced to consume

    those amounts. Therefore, the variation in viscosity of ileal

    digesta collected from dogs is not surprising owing to the

    array of factors that might play a role in viscosity

    measurement such as fluid intake, diet intake, fluid

    secretions from the gastrointestinal tract, gut motility, meal

    composition, particle size, and shear rate differences in the

    gastrointestinal tract [5,30,31].

    When ileal digesta samples from Experiment 2 were

    centrifuged at 11000 g for 10 minutes, the viscosityconstants changed dramatically, resulting in very little

    correlation between the 2 methods. In addition, according

    to nonlinear regression analysis, 10 of 12 centrifuged

    samples no longer exhibited non-Newtonian behavior based

    on the exponent values of 1 and the nonsignificant model fit.

    The removal of particles from digesta has been shown to

    alter viscous characteristics of solutions, changing them

    from non-Newtonian to Newtonian [34-36].

    Researchers sometimes use centrifugation before mea-

    surement of viscosity to control and remove large

    particulate matter in samples. According to Tietyen et al

    [37], centrifuging solutions containing oat bran and

    hydrolyzed oat bran successfully separated the water-

    soluble fraction from the large, insoluble particulate

    fraction. This technique has been successfully used to

    detect differences in viscosity of gastrointestinal contents of

    rodents fed purified water-soluble fibers such as hydrox-

    ypropylmethylcellulose and guar gum at concentrations of

    40 to 50 g/kg of the diet [38-41]. When considering the

    contents of the gastrointestinal tract, insoluble and large

    particles contribute to viscosity; therefore, sampled digesta

    should remain intact for viscosity measurement. Although

    use of centrifugation might account for water-soluble

    C.L. Dikeman et al. / Nutrfibers that contribute to viscosity in solutions [37-41], the

    method removes the insoluble fibers and particulate matterthat also may contribute to viscosity characteristics in the

    gastrointestinal tract [20,33-36].

    Researchers have diluted extremely thick gastrointestinal

    tract samples from experimental animals to produce a

    digesta sample capable of being analyzed [6,42-44].

    Because water has a viscosity of 1 mPa d s and is aNewtonian fluid, it was assumed that a dilution factor could

    be applied to back calculate what the actual viscosity should

    be. After dilution and nonlinear regression analysis, the

    diluted samples had lower viscosity constants than their

    undiluted counterparts. It was expected that those 5 solutions

    would have much higher viscosity constants than the

    highest values calculated for undiluted samples Unfortu-

    nately, diluting the samples drastically underestimated the

    viscosity constants in the present study, particularly for the

    CM treatment. It might be expected that the CM treatment

    would promote high viscosity in the gastrointestinal tract

    owing to a relatively high concentration of SDF compared

    with the other 4 diets. In addition, this diet contained higher

    concentrations of CP and AHF compared with the other

    diets, and these components also may contribute to viscous

    characteristics. Lastly, ileal DM concentration was highest

    for dogs consuming the CM diet. The higher concentration

    of DM or solids in the ileal contents would be indicative of a

    higher viscosity due to increased particulate matter in the

    fluid and a lower water-to-solids ratio [5,30,34-36].

    Cameron-Smith et al [8] reported that dilution of a

    xanthan gumcontaining diet from an initial concentration

    in solution of 18 to 12 g/kg resulted in no change in

    viscosity. On the other hand, the same dilution of solutions

    containing guar gum and methylcellulose resulted in

    reductions in viscosity of 65% and 54%, respectively. These

    findings, along with the substantial reductions in viscosity

    found in the present study, call into question the validity of

    using a dilution technique before viscosity measurement. In

    addition, the drastic alteration in the viscosity constants after

    dilution may support the idea that gastrointestinal tract

    secretions play a very significant role in the viscosity of

    gastrointestinal tract contents [8].

    During small intestinal digestion simulation, dietary

    proteins and carbohydrates were enzymatically digested.

    The early phase of enzymatic digestion may contribute to

    the ability of diet matrix structure to interact with fluid,

    resulting in the increase in viscosities observed between

    6 and 9 hours of digestion. As the enzymatic processes

    continue, carbohydrate, protein, and fat structures in the diet

    matrix would be modified and macronutrients prepared for

    absorption and removal from the lumen of the small

    intestine. This modification of structure likely causes the

    reduction in viscosity observed with increased duration of

    simulated digestion.

    The drawback to in vitro investigations is the lack of

    accounting for absorption of macronutrients and water, or

    secretion of fluids into the gastrointestinal tract. It is still

    esearch 27 (2007) 5665 63unclear how these processes and other factors such as diet

    intake may impact gastrointestinal tract viscosity. Further

  • be avoided. These techniques of viscosity measurement will

    ition Rassist researchers in understanding the impact that diet has

    on characteristics in the gastrointestinal tracts of animals

    including relevance to humans. Much research is still

    needed to determine the effects of additional factors such

    as fluid and diet intake, fluid secretions, gut motility, meal

    composition, diurnal fluctuations, and shear rate differences

    on viscosity of intestinal contents. Understanding the

    viscous characteristics of animal intestinal contents will be

    useful in defining the mixing, diffusion, and flow of

    nutrients through the gastrointestinal tract.

    References

    [1] Gallaher DD, Schaubert DR. The effect of dietary fiber type on

    glycated hemoglobin and renal hypertrophy in the adult diabetic rat.

    Nutr Res 1990;10:1311-23.

    [2] Roberts FG, Smith HA, Low AG, Ellis PR, Morris ER, Sambrook IE.

    Influence of guar gum flour of different molecular weights on

    viscosity of jejunal digesta in the pig. Proc Nutr Soc 1989;49:53A.

    [3] Reppas C, Dressman JB. Viscosity modulates blood glucose response

    to nutrient solutions in dogs. Diabetes Res Clin Pract 1992;17:81 -8.

    [4] Wood PJ, Braaten JT, Scott FW, Riedel KD, Wolynetz MS, Collins

    MW. Effect of dose and modification of viscous properties of oat gum

    on plasma glucose and insulin following an oral glucose load. Br J

    Nutr 1994;72:731 -43.

    [5] Bourne M. Food texture and viscosity: concept and measurement.

    2nd ed. San Diego7 Academic Press; 2002.[6] Higham SE, Read NW. The effect of ingestion of guar gum on

    ileostomy effluent. Br J Nutr 1992;67:115-22.

    [7] Bedford MR, Classen HL. An in vitro assay for prediction of broiler

    intestinal viscosity and growth when fed rye-based diets in the

    presence of exogenous enzymes. Poult Sci 1993;72:137 -43.

    [8] Cameron-Smith D, Collier GR, ODea K. Effect of soluble dietary

    fibre on the viscosity of gastrointestinal contents and the acute

    glycaemic response in the rat. Br J Nutr 1994;71:563-71.

    [9] Danielson AD, Newman RK, Newman CW, Berardinelli JG. Lipidresearch is needed to determine the impact of dietary

    factors such as food and water intake on gastrointestinal

    tract viscosity.

    In summary, mixer viscometry can be used to measure

    the viscosity of power law fluids including ileal digesta

    sampled from dogs. This method allows for the measure-

    ment of viscosity without centrifugation or removal of large

    particulate matter that contributes to viscosity. In the present

    study, some samples were diluted because the viscometer

    did not provide enough torque to measure the extremely

    viscous samples. A viscometer with additional power could

    measure those samples without the need for dilution;

    therefore, the problem was due to the viscometer rather

    than the use of mixer viscometry.

    It was demonstrated in the present study that centrifuging

    and diluting digesta sampled from the ileum of the dog

    resulted in substantial reductions in viscous characteristics

    of the digesta. These data, along with data of previous

    experiments, provide information on the importance of

    particle size and inclusion on viscosity; therefore, methods

    of removing particles during viscous measurements should

    C.L. Dikeman et al. / Nutr64levels and digesta viscosity of rats fed a high-fiber barley milling

    fraction. Nutr Res 1997;17:515 -22.[10] Fuente JM, Perez de Ayala P, Flores A, Villamide MJ. Effect of

    storage time and dietary enzyme on the metabolizable energy and

    digesta viscosity of barley-based diets for poultry. Poult Sci 1998;

    77:90 -7.

    [11] McDonald DE, Pethick DW, Mullan BP, Hampson DJ. Increasing

    viscosity of the intestinal contents alters small intestinal structure and

    intestinal growth, and stimulates proliferation of enterotoxigenic

    Escherichia coli in newly-weaned pigs. Br J Nutr 2001;86:487 -98.

    [12] Schneeman BO. Physical and chemical properties, methods of

    analysis, and physiological effects. Food Tech 1986;40:104 -10.

    [13] Reppas C, Greenwood DE, Dressman JB. Longitudinal versus radial

    effects of hydroxypropylmethylcellulose on gastrointestinal glucose

    absorption in dogs. Eur J Nutr 1999;8:211-9.

    [14] Lee JT, Bailey CA, Cartwright AL. Guar meal germ and hull fractions

    differently affect growth performance and intestinal viscosity of

    broiler chickens. Poult Sci 2003;82:1589-95.

    [15] Hopwood DE, Pethick DW, Hampson DJ. Increasing the viscosity of

    the intestinal contents stimulates proliferation of enterotoxigenic

    Escherichia coli and Brachyspira pilosicoli in weaner pigs. Br J

    Nutr 2002;88:523 -32.

    [16] Dvir I, Chayoth R, Sod-Moriah U, Shany S, Nyska A, Stark AH, et al.

    Soluble polysaccharide and biomass of red microalga Porphyridium

    sp. alter intestinal morphology and reduce serum cholesterol in rats.

    Br J Nutr 2000;84:469-76.

    [17] Bedford MR, Classen HL. Reduction of intestinal viscosity through

    manipulation of dietary rye and pentosanase concentration is affected

    through changes in the carbohydrate composition of the intestinal

    aqueous phase and results in improved growth rate and food

    conversion efficiency of broiler chicks. J Nutr 1992;122:560-9.

    [18] Steffe JF. Rheological methods in food process engineering.

    East Lansing7 Freeman Press; 1996. p. 1 -91.[19] Walker JA, Harmon DL, Gross KL, Collings GF. Evaluation of

    nutrient utilization in the canine using the ileal cannulation technique.

    J Nutr 1994;124:2672S-6S.

    [20] Lentle RG, Hemar Y, Hall CE, Stafford KJ. Periodic fluid extrusion

    and models of digesta mixing in the intestine of a herbivore, the

    common brushtail possum (Trichosurus vulpecula). J Comp Physiol B

    2005;175:337 -47.

    [21] Association of American Feed Control Officials. Official publication.

    Oxford (Ind): Association of American Feed Control Officials; 2005.

    [22] Boisen S. In vitro digestion for pigs and poultry. New York7Commonwealth Agriculture Bureau International (CABI) Publishing;

    1991.

    [23] Boisen S, Eggum BO. Critical evaluation of in vitro methods for

    estimating digestibility in simple-stomach animals. Nutr Res Rev

    1991;4:141 -62.

    [24] Association of Official Analytical Chemists. Official methods of

    analysis. 17th ed. Washington (DC); 2002.

    [25] American Association of Cereal Chemists. Approved methods. 8th ed.

    St. Paul (Minn): American Association of Cereal Chemists; 1983.

    [26] Budde EF. The determination of fat in baked biscuit type of dog foods.

    J Assoc Off Agric Chem 1952;35:799 -805.

    [27] Prosky L, Asp N-G, Schweizer TF, De Vries JW, Furda I.

    Determination of insoluble and soluble dietary fiber in foods and

    food products: collaborative study. J AOAC Int 1992;75:360-7.

    [28] Fox RW, McDonald AT, Pritchard PJ. Introduction to fluid mechanics.

    Hoboken7 John Wiley & Sons, Inc; 2004. p. 2 -43.[29] Lentle RG, Stafford KJ, Kennedy MS, Haslett SJ. Rheological

    properties of digesta suggest little radial or axial mixing in the

    forestomach of the tammar (Macropus eugenii) and the parma

    (Macropus parma) wallaby. Physiol Biochem Zool 2002;75:572-82.

    [30] Takahashi T, Sakata T. Large particles increase viscosity and yield

    stress of pig cecal contents without changing basic viscoelastic

    properties. J Nutr 2002;132:1026-30.

    [31] Reppas C, Meyer JH, Sirois PJ, Dressman JB. Effect of hydrox-

    esearch 27 (2007) 5665ypropylmethylcellulose on gastrointestinal transit and luminal viscos-

    ity in dogs. Gastroenterology 1991;100:1217 -23.

  • [32] Carriere CJ, Inglett GE. Constitutive analysis of the nonlinear

    viscoelastic properties of cellulosic fiber gels produced from corn or

    oat hulls. Food Hydrocolloids 2003;17:605-14.

    [33] Takahashi T, Karita S, Ogawa N, Goto M. Crystalline cellulose reduces

    plasma glucose concentrations and stimulates water absorption by

    increasing the digesta viscosity in rats. J Nutr 2005;135:2405 -10.

    [34] Takahashi T, Goto M, Sakata T. Viscoelastic properties of the small

    intestinal and cecal contents of the chicken. Br J Nutr 2004;91:867 -72.

    [35] Takahashi T, Sakata S. Viscous properties of pig cecal contents and

    the contribution of solid particles to viscosity. Nutrition 2004;20(4):

    377-82.

    [36] Lam CD, Flores RA. Effect of particle size and moisture content on

    viscosity of fish feed. Cereal Chem 2003;80:20-4.

    [37] Tietyen JL, Nevins DJ, Shoemaker CF, Schneeman BO. Hypocho-

    lesterolemic potential of oat bran treated with an endo-b-d-glucanasefrom Bacillus subtilis. J Food Sci 1995;60:558 -60.

    [38] Carr TP, Wood KJ, Hassel CA, Bahl R, Gallaher DD. Raising intestinal

    contents viscosity leads to greater excretion of neutral steroids but not

    bile acids in hamsters and rats. Nutr Res 2003;23:93 -102.

    [39] Carr TP, Gallaher DD, Yang C-H, Hassel CA. Increased intestinal

    contents viscosity reduces cholesterol absorption efficiency in

    hamsters fed hydroxypropylmethylcellulose. J Nutr 1996;126:1463-9.

    [40] Gallaher DD, Hassel CG, Lee K-J, Gallaher CM. Viscosity and

    fermentability as attributes of dietary fiber responsible for the

    hypocholesterolemic effect in hamsters. J Nutr 1993;123:244 -52.

    [41] Gallaher DD, Hassel CA, Lee K-J. Relationships between viscosity

    of hydroxypropylmethylcellulose and plasma cholesterol in hamsters.

    J Nutr 1993;123:1732-8.

    [42] Wang L, Newman RK, Newman CW, Hofer PJ. Barley b-glucans alterintestinal viscosity and reduce plasma cholesterol concentrations in

    chicks. J Nutr 1992;122:2292 -7.

    [43] Ikegami S, Tsuchihashi F, Harada H, Tsuchihashi N, Nishide E,

    Innami S. Effect of viscous indigestible polysaccharides on pancre-

    atic-biliary secretion and digestive organs in rats. J Nutr 1990;120:

    353 -60.

    [44] Piel C, Montagne L, Se`ve B, Lalle`s J-P. Increasing digesta viscosity

    using carboxymethylcellulose in weaned piglets stimulates ileal goblet

    cell numbers and maturation. J Nutr 2005;135:86-91.

    C.L. Dikeman et al. / Nutrition Research 27 (2007) 5665 65

    Diet and measurement techniques affect small intestinal digesta viscosity among dogsIntroductionMethods and materialsExperiment 1AnimalsProceduresViscosity

    Experiment 2AnimalsViscosity

    Experiment 3AnimalsProceduresIn vitro digestion simulation

    Chemical analysesStatistical analysis

    ResultsChemical analysesExperiment 1Experiment 2Experiment 3In vitro digestion simulation

    DiscussionReferences