Abnormal permeability precedes the development of
a gluten sensitive enteropathy in Irish setter dogs
E J Hall, RM Batt
Intestinal permeability to 15Cr-EDTA was
examined during the development of gluten
affected Irish setters and reared on a normal
wheat containing diet. Comparisons were
made with litter mates reared on a gluten free
diet and with a control group of age matched,
clinically healthy Irish setters reared on the
normal diet. Studies at4, 6, 8, and 12 months of
age were correlated with morphometric and
biochemical examinations of peroral jejunal
biopsy specimens. Permeability was increased
at all ages in the group fed gluten free diet
compared with control dogs, although there
were no differences in villus height, intra-
epithelial lymphocyte density, and alkaline
phosphatase activity. At four months, perme-
ability in the normal diet group was greater
than in controls, although comparable with
that in the gluten free diet group. Permeability
in the normal diet group increased further in
conjunction with the development of partial
villus atrophy and reduced alkaline phos-
phatase activity, andby 12monthspermeability
was significantly greater than in their gluten
free diet litter mates and the control dogs. The
findings suggest that an underlying perme-
ability abnormality may be involved in the
pathogenesis ofgluten sensitive enteropathy in
Irish setter dogs.
in dogs bred from
Pathology, University of
E J Hall
Dr E J Hall, Department of
University ofLiverpool, PO
Box 147, Liverpool L69 3BX.
Accepted for publication
3 August 1990
patients with untreated coeliac disease by many
workers using 11Cr-EDTA'4 or non-digestible
sugars" aspermeability probes, maybeimpor-
tant as amarkerofintestinaldamage. In addition,
it has been suggested that an underlying defect in
permeability could play a primary role in the
pathogenesis of coeliac disease.8 According to
this hypothesis, a primary increase in mucosal
permeability allows ingested gluten, or peptides
derived from gluten digestion,
damage either by direct toxicity or by immune
Support for this hypothesis has been provided
by evidence of increased intestinal permeability
to 5'Cr-EDTA not only in patients with untreated
coeliac disease but also in patients successfully in
remission on a gluten free diet.'214 However,
considerable overlap or no significant difference
between the permeability
treated coeliac patients and control subjects has
been found by others.35 15
ative permeability probes have shown increased
permeability in some patients investigated for
intestinal disease but with apparently normal
jejunal histology," 12 although careful morpho-
intestinal permeability, shown
to cross the
to 5'Cr-EDTA in
6 Studies using altern-
metric analysis has detected subtle histological
changes in these cases.9'7 No primary abnor-
mality in coeliac disease has been shown using
treated with a gluten free diet.'8 19 In one study,
permeability decreased significantly after the
introduction of a gluten free diet, although it
remained greater than controls in five of 13
patients; however, this was correlated with the
persistence of morphological abnormalities and
Attempts to identify a primary permeability
abnormality in treated coeliac disease patients
may thus be frustrated by the difficulty of
excluding inadvertent gluten ingestion and any
potential long term effects of severe mucosal
damage. An alternative approach to the inherent
problems ofinvestigating permeability in treated
patients has been the examination of unaffected
first degree relatives of patients with coeliac
disease, but no increase in permeability to
1'Cr-EDTA was shown despite increased intra-
Examination ofa suitable animal model might
elucidate potential relations between enhanced
intestinal permeability and mucosal damage by
gluten. Recent studies have characterised a
naturally occurring gluten sensitive enteropathy
in Irish setter dogs.22 23 Affected Irish setters fed a
diet containing wheat show inappetance, poor
weight gain or loss of weight, and chronic
improve after introduction ofa cereal free diet.23
Jejunal changes include partial villus atrophy,
accompanied by an increase in the relative
number of intraepithelial lymphocytes,22 24 and a
specific but secondary loss of brush border
alkaline phosphatase on a normal diet containing
wheat.25 These histological and biochemical
abnormalities improve on feeding a wheat free
challenge.23 Further studies have clearly shown
that these abnormalities are related to the
presence of gluten in the diet.26 Investigation of
this naturally occurring animal model should
therefore provide insight into the mechanism of
damage to the intestinal mucosa by gluten.
The 1'Cr-EDTA absorption test has been
validated for the determination of intestinal
permeability in the dog.27 In the present study,
permeability to 11Cr-EDTA was examined in
dogs bred from affected Irish setters and reared
on a normal wheat containing diet, in litter mates
reared on a cereal free diet, and in age matched
normal Irish setters fed the wheat containing
provide an opportunity to determine whether
any changes in intestinal permeability are merely
the result of mucosal damage by gluten or
represent primary abnormalities which precede
the development ofthe disease.
GROUPS OF ANIMALS
Thirteen Irish setter dogs, bred from Irish
enteropathy, were studied. Details ofthese dogs
have been reported previously.24 The progeny
were allotted to two groups at 5-6 weeks of age,
housed separately, and weaned. One group ofsix
dogs was reared exclusively on a gluten free diet
of rabbit and turkey meat (Mr Dog, Pedigree
Petfoods). The other group of seven dogs was
reared on a normal diet consisting of rabbit and
turkey meat (Mr Dog) with cereals containing
wheat (Pedigree Chum Mixer, Pedigree Pet-
clinically healthy Irish setters, bred from un-
affected parents, and fed the same diet of meat
and cereal as the normal diet group. Investiga-
tions were performed at approximately 4, 6, 8,
and 12 months ofage, as detailed previously.24
perorally from a site approximately 10 cm distal
to the duodenal-jejunal flexure using a Quinton
multiple suction biopsy capsule,23 after food had
been withheld for approximately 18 hours. The
tissues were fixed in 10% (w/v) formalin, and
after conventional processing, 5 [im paraffin
embedded sections were cut and stained with
haematoxylin and eosin. The height of five
sequential, well-orientated villi in coded sections
were measured by use of a calibrated eyepiece
graticule. Intraepithelial lymphocytes (numbers
per 100 enterocytes) were counted in the mid-
zone regions ofthese villi.
Portions of jejunal biopsy specimens were
homogenised in sucrose medium (0 3 mol/l
sucrose, 1 mmol/lNa2EDTA, 22 mmol/l ethanol,
marker enzyme activities, including alkaline
phosphatase, were assayed in the fractions. Sub-
cellular fractionation was performed and the
organelle marker enzymes have been presented
elsewhere.25 Proteinwas determined accordingto
Schacterle and Pollack2' using bovine serum
for alkaline phosphatase and other
PERMEABILITY TO 5'CR-EDTA
Permeability to 51Cr-EDTA was assessed in each
dog one to twoweeks before jejunal biopsy, using
the procedure reported previously.27 The test
solution contained 50
11Cr-EDTA (specific activity 1 to 2 mCi/mgCr,
Amersham International) in 50 ml of distilled
water; a 5 ml aliquot was retained as a counting
,tCi (135 MBq) of
standard. After food had been withheld from the
dogs for 15 hours, 45 ml of the 5'Cr-EDTA
solution was given intragastrically by a stomach
tube, whichwas then flushed with 55 ml ofwater.
Eachdogwasthenplacedin ametabolismcage for
24 hours; waterwas available ad libitum and food
was given after six hours. At the end of24 hours,
the bladder was catheterised and the urine was
pooled with any urine collected in the cage, and
the total volume recorded. No dog passed faeces
duringthe24hourperiod in themetabolism cage.
administered dose of 5'Cr-EDTA was calculated
by counting the radioactivity in five 2.5 ml
aliquots of urine samples and 1:100 dilutions of
the standards in a LKB-Wallac 1270Rackgamma
counter. Each sample was counted for five
minutes; the coefficient ofvariation was between
1.2 and 6-0% depending on the level of radio-
activity. Sensitivity was 0.03% of the adminis-
tered dose per litre of urine.
After testing for homogeneity of variance by
Levene's test, comparisons of results from the
three groups at each age were performed by
analysis of variance and Newman-Keuls test.
Differenceswith apvalue <0 05 were considered
to be significant.
Offspring of affected dogs fed the normal diet
containing wheat showed clinical signs of poor
appetite and chronic intermittent diarrhoea
typical ofgluten sensitive enteropathy,22 whereas
littermates fed a gluten free diet and controls
were clinically healthy.
Results of the morphometric analysis of the
sections ofbiopsyspecimens have been presented
Percentage 24 hour urinary excretion ofingested "Cr-EDTA
from offspring ofaffected Irish setters andfromcontrol setters
(C; closed square) after intragastricadministration of50[tCi
of3"Cr-EDTA. Offspring ofaffected dogswerefedeither a
nonnal wheat containing diet (ND;closed circle)and
eventually showed intestinal damage, or werefedaglutenfree
diet (GFD; open circle) and showed no intestinal damage.
Data are expressed as mean (SEM). Significance ofstatistical
comparisons: **=p<0.01 v C; t=p<0-05; tt=p<001
Abnormalpermeability precedes the developmentofagluten sensitive enteropathy in Irish setterdogs
Examinationofjejunalbiopsy specimensfrom affected dogs on a normal diet(ND)or a gluten
free diet (GFD), andfrom control dogs (C). Values are mean (SEM)
(Per 100 enterocytes)
Significance of statistical comparisons: *=p<0.01 v C; t=p<0-05; t=p<001 v GFD.
elsewhere24 and are summarised in the Table.
Villus height in the normal diet group was
unaffected at 4 months but was significantly
boththecontrolsandthe progenyofaffected dogs
intraepithelial lymphocytes was
feature in the normal diet group at all ages,
whereas there were no differences comparing
progeny of affected dogs reared on a gluten free
diet with controls.
Full results of the biochemical analyses of the
biopsy specimens have been presented else-
where25; results of alkaline phosphatase activity
are shown in the Table. In the affected progeny
fed a normal diet, specific activities of alkaline
phosphatase were unaffected at 4 months, but
were significantly lower than controls by 6
months, and had fallen noticeably by 12 months
to values significantly different not only from
controls but also from litter mates fed a gluten
PERMEABILITY TO 51CR-EDTA
The percentage 24 hour urinary excretion of
51Cr-EDTA in both normal and gluten free diet
groups of affected progeny was significantly
greater than in controls at all times (Figure). At 4
months excretion in the normal diet group was
comparable with the gluten free diet group, but
increased by 8 months to become significantly
greater compared not only with the controls but
also with their litter mates on a gluten free diet.
It has been shown that intestinal permeability to
5'Cr-EDTA is increased in approximately 80% of
coeliac disease patients successfully treated with
a gluten free diet.
intestinal permeability in treated patients lends
support to the hypothesis that deranged perme-
ability could be a primary defect involved in the
aetiopathogenesis of the condition.8"'Persistent
ultrastructural lesions have been described in
biopsy specimens from children with coeliac
disease in remission on a gluten free diet, despite
normal dissecting and light microscopy find-
ings,29 and these might be relevant to a perme-
Despite the finding of apparently normal
2 Such a persistent increase in
jejunal mucosa in treated coeliac patients with
increased permeability to 5'Cr-EDTA, however,
suspicion may remain as to the complete absence
of gluten from the diet,' and the patients'
compliance to a strictly gluten free diet with
abstinence from substances known to increase
permeability to `Cr-EDTA, such as alcohol and
non-steroidal anti-inflammatory drugs.30 Other
studies of treated coeliac disease patients, using
5'Cr-EDTA35 1516 or other probes'8"20 to measure
intestinal permeability, have not provided any
abnormality. These conflicting results may be
explained by the different sensitivities of the
tests and the failure to include appropriate
controls,30 variations in control values between
osmolarities ofthe test solutions used.32
abnormality in patients with coeliac disease may
be impossible because of the difficulties of inter-
pretation arising from the severe disruption of
mucosal integrity that occurs, causing non-
specific secondary increase in permeability.
Examination oftreated coeliac disease patients is
complicated by the difficulties of complete
gluten exclusion and the potential long term
effects onmucosa after apparent recovery. Thus,
measurement of intestinal permeability in a
suitable animal model ofgluten sensitivity would
provide an opportunity to determine whether a
primary permeability defect may be relevant to
the mechanism of gluten toxicity. The study of
Irish setter dogs with naturally occurring gluten
sensitive enteropathy permits examination of
whether a primary permeability abnormality
could be involved in the mechanism of mucosal
damage by gluten.
Sequential examination of biopsy specimens
in the present study showed the characteristic
changes - partial villus atrophy, intraepithelial
lymphocyte infiltration, and reduced alkaline
enteropathy22-25 in affected dogs reared on a
normal wheat containing diet. The reduction in
villus height and alkaline phosphatase activity in
affected dogs in response to dietary wheat was
evident by 6 months of age and was progressive,
with noticeable reductions by 12 months of age.
As expected, affected progeny that were fed a
diet containing wheat had morphological and
biochemical evidence of intestinal damage, and
indeed had the greatest permeability to 5'Cr-
EDTA. Permeability was also increased at 4
months of age, however, compared with con-
In the absence of dietary wheat, villus height,
intraepithelial lymphocyte density, and alkaline
phosphatase activity in progeny of affected dogs
were no different from the control dogs. Yet
significantly, these progeny on a gluten free diet
also exhibited increased permeability when com-
pared with controls, despite the absence of
morphological or biochemical evidence of intes-
tinal damage; these dogs did excrete less 5'Cr-
EDTA than the normal diet group, but this
would be expected in the absence of gross
The pathway across the mucosa taken by 5'Cr-
EDTA is believed to be paracellular, and per-
meation to 5Cr-EDTA in experimental models
has been shown to correlate with permeation to
oligopeptides33 and macromolecules,34 and the
formation of circulating immune complexes in
vivo.35 The mucosa in coeliac disease has been
shown to have increased permeability to dietary
antigens,36 and the size of "Cr-EDTA (mol wt
340) is within an order of magnitude of the
smallest gliadin peptides (mol wt 500-1000),
produced by enzymatic digestion, that have been
shown to be toxic in coeliac disease.3738 Access of
gluten or gluten derived peptides across the
intestinal epithelium could initiate a pathological
process, either by direct toxicity or by activating
immune cascade mechanisms.
The increase in intraepithelial lymphocyte
density seen in the normal diet group was an
early response to the presence of gluten in the
diet, and could represent an immune response to
entry ofgluten or gluten derived peptides across
the epithelial barrier. It remains to be deter-
mined whether such an intraepithelial lympho-
cyte infiltration is merely a secondary response to
an abnormally permeable mucosa, or represents
a primary immunological abnormality. The
further increase in permeability, presumably
secondary to the intestinal damage present in the
normal diet group fed wheat, would perhaps
permit greater access of potential antigens into
the lamina propria thereby exacerbating the
In mature control Irish setters the 24 hour
urinary excretion of "Cr-EDTA after peroral
administration is approximately 10%, and much
greater than that reported for man30; the signifi-
cance of this is discussed elsewhere.27 Despite
this high natural intestinal permeability in
canines compared with other species, increased
permeability apparently exists in affected Irish
setters compared with control Irish setters,
although it may be speculated that the apparent
increase in permeability in the gluten free diet
group, despite the lack of evidence of intestinal
damage, was caused by extraneous factors.
Mucosal integrity can be damaged by EDTA39
and its presence in the test solution, to stabilise
5Cr-EDTA, could theoretically damage the
mucosal barrier of affected progeny more than
controls. The concentration of EDTA reported
to produce changes in permeability, however, is
one thousand times greater than that of the test
solution. Another possiblefactor is the diet ofthe
gluten free diet group, which is an all meat
product and could increase intestinal transit
time.'I However, the total transit time in all these
dogs was greater than 24 hours, and so an effect
on the permeation of 5Cr-EDTA would only be
seen if the permeability of small intestinal and
colonic mucosa were different. Furthermore,
dogs in the normal diet group were fed the same
developed intestinal damage they had increased
permeability compared with the controls.
The diet fed to the gluten free diet dogs was
strictly gluten free, but could contain some other
sensitive. Carrageenan, for example, is present
in some proprietary dog foods and has been
shown to cause an increased intestinal perme-
as the controls, and yet before they
ability in rats without producing any histological
abnormalities.4' Thus, the permeability abnor-
mality may not be a primary defect but secon-
thereby permits access of gluten to the mucosa.
Elemental diets have been shown to reduce
intestinal permeability in Crohn's
where permeability to macromolecules may be
involved in the pathogenesis ofthe condition. 13 It
would be interesting to measure the intestinal
permeability of these dogs while they are main-
tained on an elemental diet in order to eliminate
the effects ofany potential dietary sensitivity.
In conclusion, increased permeability to 5Cr-
EDTA was found in progeny of affected Irish
setters fed a gluten free diet, although there was
no morphological or biochemical evidence of
intestinal damage. Increased permeability was
also found preceding the development of gluten
sensitive enteropathy in affected dogs reared on a
diet containing wheat. These findings suggest
that an underlying permeability abnormality
may be important in the pathogenesis of gluten
sensitive enteropathy in Irish setters. Further
increases in permeability reflected the develop-
ment ofovert intestinal damage secondary to the
ingestion of gluten. Any correlation between
increased 5'Cr-EDTA permeation and perme-
ability to gluten or gluten derived peptides in
affected Irish setters remains to be determined.
that increases permeability and
Supported by the Wellcome Trust. Presented in abstract form at
the British Society ofGastroenterology Meeting, Warwick, March
1 Bjarnason I, Peters TJ, Veall N. A persistent defect in
intestinal permeability in coeliac disease demonstrated by a
"Cr-labelled EDTA absorption test. Lancet 1983; i: 323-5.
2 Bjarnason I, MarshMN, Price A, et al. Intestinal permeability
in patients with coeliac disease and dermatitis herpetiformis.
Gut 1985; 26: 1214-9.
3 Behrens RH, Szaz KF, Northrop C, Elia M, Neale G.
Radionucleide tests for the assessment of intestinal perme-
ability. EurJ Clin Invest 1987; 17: 100-5.
4 Martines D, Morris AI, Gilmore IT, et al. Comparison
between the cellobiose/mannitol and "Cr-labelled ethylene-
diaminetetra-acetate absorption tests in the detection of
coeliac disease. Clin Sci 1988; 75: 375-8.
5 Fotherby KJ, Wraight EP, Neale G. 'Cr-EDTA/'4C-Mannitol
intestinal permeability test. Clinical use in screening for
coeliac disease. ScandJ Gastroenterol 1988; 23: 171-7.
6 Turck D, Ythier H, Maquet E, et al. Intestinal permeabilityto
['Cr]EDTA in children with Crohn's disease and celiac
disease. JPediatr GastroenterolNutr 1987; 6: 535-7.
7 Menzies IS. Absorption ofintact oligosaccharidein health and
disease. Biochem Soc Trans 1974; 2: 1042-7.
8 Menzies IS, Laker MF, Pounder R, et al. Abnormal intestinal
permeability to sugars in villous atrophy. Lancet 1979; ii:
9 Strobel S, Brydon WG, Ferguson A. Cellobiose/mannitol
sugar permeability test complements biopsy histopathology
in clinical investigation of the jejunum. Gut 1984; 25:
10 Stenhammar L, Stromberg S. Intestinal permeability
lactulose/L-rhamnose in children with celiac disease and
other gastrointestinaldisorders. J Pediatr Gastroenterol Nutr
1988; 7: 304-5.
11 Juby LD, Rothwell J, Axon ATR. Cellobiose/mannitol sugar
test: a sensitive tubeless test for coeliac disease: results on
1010 unselected patients. Gut 1989; 30: 476-80.
12 Juby LD,Rothwell J, Axon ATR. Lactulose/mannitol test: an
ideal screen for celiac disease. Gastroenterology 1989; 96:
13 Bjarnason I, Peters TJ. Helping the mucosa make sense of
macromolecules. Gut 1987; 28: 1057-61.
14 Dawson DJ, Lobley RW, Burrows PC, Notman JA, Mahon
M, Holmes R. Changes in jejunal permeability and passive
permeation of sugars in intestinal biopsies in coeliac disease
and Crohn's disease. Clin Sci 1988; 74: 427-31.
15 O'Mahony CP, Stevens FM, Bourke M, McCarthy CF, Weir
DG, Feighery CF. 5'Cr-EDTA test for coeliac disease.
Lancet 1984; i: 1354-5.
16 Peled Y, Watz C, Gilat T. Measurements of intestinal perme-
ability using 5'Cr-EDTA. Am J Gastroenterol 1985; 80:
17 Juby LD, Dixon MF, Axon ATR. Abnormal intestinal
Abnormalpermeabilityprecedes thedevelopmentofagluten sensitive enteropathy in Irish setterdogs
permeability and jejunal morphometry. J Clin Pathol 1987;
18 Hamilton I, Cobden I, Rothwell J, Axon ATR. Intestinal
permeability in coeliac disease: the response to gluten
withdrawal and single-dose gluten challenge. Gut 1982; 23:
19 Stenhammar L, Falth-Magnusson K, Jansson G, Magnusson
K-E, Sundqvist T. Intestinal permeability to inert sugars
and different polyethyleneglycols in children with celiac
disease.jPediatr Gastroenterol Nutr 1989; 8: 281-9.
20 Ukabam SO, Cooper BT. Small intestinal permeability as an
indicator of jejunal mucosal recovery in patients with celiac
sprue on a gluten-free diet. J Clin Gastroenterol 1985; 7:
21 Marsh MN, Bjarnason I, Shaw J, Ellis A, Baker R, Peters TJ.
Studies of intestinal lymphoid tissue. XIV-HLA status,
mucosal morphology, permeability and epithelial lympho-
cyte populations in first degree relatives of patients with
coeliac disease. Gut 1990; 31: 32-6.
22 Batt RM, Carter MW, McLean L. Morphological and bio-
chemical studies of a naturally occurring enteropathy in the
Irish setter dog: a comparison with coeliac disease in man.
Res VetSci 1984; 37: 339-46.
23 Batt RM, McLean L, Carter MW. Sequential morphologic
and biochemical studies of naturally occurring wheat-
sensitive enteropathy in Irish setter dogs. Dig Dis Sci 1987;
24 Hall EJ, Batt RM. Development of wheat-sensitive entero-
pathy in Irish setters: morphologic changes. Am J Vet Res
1990; 51: 978-82.
25 Hall EJ, Batt RM. Development of wheat-sensitive entero-
pathy in Irish setters: biochemical changes. Am J Vet Res
1990; 51: 983-9.
26 Hall EJ, Batt RM. Challenge studies demonstrate gluten
sensitivity ofa naturally occurring enteropathy in Irish setter
dogs [Abstract]. Gastroenterology 1988; 94: A167.
27 Hall EJ, Batt RM, Brown A. Assessment of canine intestinal
permeability, using "chromium-labeled ethylenediamine-
tetra-acetate. AmJ VetRes 1989; 50: 2069-74.
28 Schacterle GR, Pollack RL. A simplified method for the
quantitative assay of small amounts of protein in biologic
material. AnalBiochem 1973; 51: 654-5.
29 Stenling R, Fredrikzon B, Engberg S, Falkmer S. Surface
ultrastructure ofthe small intestinal mucosa in children with
coeliac disease. I. Untreated disease and effects oflong-term
gluten elimination and challenge. UltrastructPathol 1984; 6:
30 Bjarnason I, Peters TJ, Veall N. 5Cr-EDTA test for intestinal
permeability. Lancet 1984; ii: 523.
31 Bourke S, Murphy B, Stafford F, Maher K, O'Morain C.
chromium EDTA. IrJ MedSci 1988; 157: 287-9.
32 Smethurst P, Menzies IS, Levi AJ, Bjarnason I. Intestinal
permeability: osmolarity revisited [Abstract]. Gut 1988; 29:
33 Ferry DM, Butt TJ, Broom MF, Hunter J, Chadwick VS.
Bacterial chemotactic oligopeptides and
mucosal barrier. Gastroenterology 1989; 97: 61-7.
34 Ramage JK, Stanisz A, Scicchitano R, Hunt RH, Perdue MH.
Effect ofimmunologic reactions on rat intestinal epithelium.
chromium 51-labeled ethylenediaminetetraacetic acid and
ovalbumin during acute inflammation and anaphylaxis.
Gastroenterology 1988; 94: 1368-75.
35 Davin J-C, Forget P, Mahieu PR. Increased intestinal per-
meability to (51Cr) EDTA is correlated with IgA immune
levels in children with IgA-associated
nephropathies. Acta PaediatrScand 1988; 77: 118-24.
36 Husby S, Foged N, H0st A, Svehag S-E. Passage of dietary
antigens into the blood of children with coeliac disease.
Quantification and size distribution of absorbed antigens.
Gut 1987; 28: 1062-72.
37 Bronstein HD, Haeffner LJ, Kowlessar OD. Enzymatic
digestion of gliadin; the effect of the resultant peptides in
adult celiac disease. Clin Chim Acta 1966; 14: 141-55.
38 Offord RE, Anand BS, Piris J, Truelove SC. Further subfrac-
tionation of digests of gluten. In: McNicholl B, McCarthy
CF, Fottrell PF, eds. Perspectives in coeliac disease. Lan-
caster: MTP Press, 1978: 25-9.
39 Ahrens FA, Aronson AL. A comparative study of the toxic
effects of calcium and chromium chelates of ethylene-
diaminetetraacetate in the dog. Toxicol Appl Pharmacol
1970; 18: 10-25.
40 Cherbut Ch, Meirieu 0, Ruckebush Y. Effect of diet on
intestinal xylose absorption in dogs. Dig Dis Sci 1986; 31:
41 Delahunty T, Recher L, Hollander D. Intestinal permeabilitv
changes in rodents: a possible mechanism for degraded
carrageenan-induced colitis. Food Chem Toxicol 1987; 86A:
42 Sanderson IR, Boulton P, Menzies I, Walker-Smith JA.
Improvement ofabnormal lactulose/rhamnose permeability
in active Crohn's disease of the small bowel by an elemental
diet. Gut 1987; 28: 1073-6.