a controlled trial of gluten-free diet in patients with irritable bowel syndrome-diarrhea- effects...

7/28/2019 A Controlled Trial of Gluten-Free Diet in Patients With Irritable Bowel Syndrome-Diarrhea- Effects on Bowel Frequen

1/12

CLINICALALIMENTARY TRACT

A Controlled Trial of Gluten-Free Diet in Patients With Irritable Bowel

Syndrome-Diarrhea: Effects on Bowel Frequency and Intestinal FunctionMARIA I. VAZQUEZROQUE,1,2 MICHAEL CAMILLERI,1 THOMAS SMYRK,3 JOSEPH A. MURRAY,1 ERIC MARIETTA,1

JESSICA ONEILL,1 PAULA CARLSON,1 JESSE LAMSAM,4 DENISE JANZOW,5 DEBORAH ECKERT,1 DUANE BURTON,1

and ALAN R. ZINSMEISTER6

1Clinical Enteric Neuroscience Translational and Epidemiological Research, 3Surgical Pathology, Department of Laboratory Medicine and Pathology,4Immunochemical Core Laboratory, 5Mayo Clinic Clinical Translational Science Awards Program, 6Division of Biomedical Statistics and Informatics, Department of

Health Sciences Research, College of Medicine, Mayo Clinic, Rochester, Minnesota; 2Division of Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, Florida

See Covering the Cover synopsis on page 859.

BACKGROUND & AIMS: Patients with diarrhea-pre-dominant irritable bowel syndrome (IBS-D) could benefitfrom a gluten-free diet (GFD). METHODS: We performeda randomized controlled 4-week trial of a gluten-containingdiet (GCD) or GFD in 45 patients with IBS-D; genotypeanalysis was performed for HLA-DQ2 and HLA-DQ8. Twen-ty-two patients were placed on the GCD (11 HLA-DQ2/8negative and 11 HLA-DQ2/8 positive) and 23 patients wereplaced on the GFD (12 HLA-DQ2/8 negative and 11 HLA-DQ2/8 positive). We measured bowel function daily, small-bowel (SB) and colonic transit, mucosal permeability (bylactulose and mannitol excretion), and cytokine production

by peripheral blood mononuclear cells after exposure togluten and rice. We collected rectosigmoid biopsy specimensfrom 28 patients, analyzed levels of messenger RNAs encod-ing tight junction proteins, and performed H&E stainingand immunohistochemical analyses. Analysis of covariancemodels was used to compare data from the GCD and GFDgroups. RESULTS: Subjects on the GCD had more bowelmovements per day (P .04); the GCD had a greater effecton bowel movements per day of HLA-DQ2/8positive thanHLA-DQ2/8negative patients (P .019). The GCD wasassociated with higher SB permeability (based on 02 hlevels of mannitol and the lactulose:mannitol ratio); SB per-

meability was greater in HLA-DQ2/8positive than HLA-DQ2/8negative patients (P .018). No significant differ-ences in colonic permeability were observed. Patients on theGCD had a small decrease in expression of zonula occludens 1in SB mucosa and significant decreases in expression ofzonula occludens 1, claudin-1, and occludin in rectosigmoid mu-cosa; the effects of the GCD on expression were significantlygreater in HLA-DQ2/8positive patients. The GCD vs theGFD had no significant effects on transit or histology. Pe-ripheral blood mononuclear cells produced higher levels ofinterleukin-10, granulocyte colony-stimulating factor, andtransforming growth factor- in response to gluten than rice(unrelated to HLA genotype). CONCLUSIONS: Gluten al-ters bowel barrier functions in patients with IBS-D, par-

ticularly in HLA-DQ2/8positive patients. These find-ings reveal a reversible mechanism for the disorder.Clinical trials.gov NCT01094041.

Keywords: Permeability; Transit; Immunity; Cytokines.

The relationship of gluten exposure and symptomgeneration in irritable bowel syndrome (IBS) is com-plex and not well understood although many patients areeither counseled or decide to follow a gluten-free diet(GFD). The prevalence of celiac disease in patients withfunctional chronic diarrhea or diarrhea-predominant IBS(IBS-D) is similar to that of healthy controls, approxi-mately 0.4%.1 Nonceliac IBS-D patients who are HLA-DQ2/8 positive can show symptom improvement on aGFD. In view of the observation of celiac diseaseassoci-

ated serum IgG (not IgA) in 37% of patients with IBS-D,an adaptive immune mechanism in response to glutenexposure has been proposed to explain the improvementin symptoms with gluten withdrawal from the diet.2

Other factors such as alterations in gut motility or secre-tion,3 gut permeability, or inflammation conceivably mayinteract with immune mechanisms to result in symptomsin IBS-D patients exposed to gluten.

Alterations in intestinal permeability and jejunal mu-cosal tight junction (TJ) signaling have been described inIBS-D,4,5 including postinfectious IBS-D.6 However, thedegree of increased small-bowel (SB) permeability in pa-

tients considered gluten sensitive was low compared withthat of celiac disease patients and healthy controls.7

Abbreviations used in this paper: CI, confidence interval; GAPDH,

glyceraldehyde 3-phosphate dehydrogenase; GCD, gluten-containing

diet; G-CSF, granulocyte colonystimulating factor; GFD, gluten-free

diet; GM-CSF, granulocyte-macrophage colonystimulating factor; IBS,

irritable bowel syndrome; IBS-D, diarrhea-predominant irritable bowel

syndrome; ITT, intention-to-treat; L:M ratio, lactulose:mannitol ratio;

mRNA, messenger RNA; PBMC, peripheral blood mononuclear cell;

PCR, polymerase chain reaction; SB, small bowel; TJ, tight junction;

ZO-1, zonula occludens 1.

2013 by the AGA Institute

0016-5085/$36.00

http://dx.doi.org/10.1053/j.gastro.2013.01.049

CLINICALAT

GASTROENTEROLOGY 2013;144:903911
http://www.clinicaltrials.gov/http://dx.doi.org/10.1053/j.gastro.2013.01.049http://dx.doi.org/10.1053/j.gastro.2013.01.049http://www.clinicaltrials.gov/


2/12

The aim of this 4-week, randomized, clinical trial inHLA-DQ2/8positive and HLA-DQ2/8negative patientswith IBS-D was to assess the effects of a gluten-containingdiet (GCD) compared with a GFD on bowel function, guttransit, SB, and colonic barrier functions measured func-tionally by the 2-sugar excretion permeability test andmessenger RNA (mRNA) expression of TJ proteins in the

mucosa of the SB and rectosigmoid. We also assessedmucosal morphology and immune activation in responseto the diets, and the proliferative and cytokine responsesof peripheral blood mononuclear cells (PBMCs) to glutenand rice antigens.

Materials and Methods

Study Design and Intervention

Between January 2010 and February 2012, we conducteda single-center, parallel-group, randomized, controlled, 4-weektrial of GCD and GFD in 45 gluten-ingesting patients withIBS-D. They were recruited from a database of more than 800

patients with IBS who had been evaluated clinically by theinvestigators or clinical staff at Mayo Clinic. These patientsreside within 150 miles of Mayo Clinic in Rochester, Minnesota,and the database is maintained in the laboratory of the principalinvestigator. A total of 307 participants with known IBS-D wereinvited to participate by letter or electronic communication. Thestudy was approved by the Mayo Clinic Institutional ReviewBoard, and all participants signed informed consent.

The study was registered at Clinicaltrials.gov, NCT01094041,and all authors had access to the study data and reviewed andapproved the final manuscript. There were no adverse effects ofthe interventions or treatments in the entire study.

All studies were conducted on an outpatient basis at Mayo

Clinic, and meals were ingested or prepared in the Mayo ClinicalResearch Unit. Participants also were provided snacks and ad-vised to eat only the foods provided by the study dietitiansduring the entire study period. The macronutrient distributionof the meals was 20% protein, 30% fat, and 50% carbohydrate.Three typical meals were prepared with the same macronutrientproportions in the gluten-free and gluten-containing diets. Thetypical food offerings in the diet are included in Appendix 1.Patients selected foods from these menus. We used the HarrisBenedict equation8 plus an additional correction of caloriesrequired for different levels of physical activity to ensure weightmaintenance. During the 4-week study, compliance to the dietwas assessed by direct questioning by the dietitians when par-ticipants picked up the meal and snack supplies.

Randomization and Masking

The randomization sequence was generated by com-puter and communicated by a statistician to the dietitians in theClinical Research Unit. Allocation to the different diets wasconcealed from all investigators except for the assigned leaddietitian (D.J.).

Participants and Screening for Celiac Disease

The IBS-D patients (43 women, 2 men) had symptomsconsistent with Rome II criteria9 and confirmed by a validatedquestionnaire.10 Psychological health was assessed at baselineusing the Hospital Anxiety and Depression Inventory.11 All pa-tients were ingesting gluten in their diet before starting the

study. Dietary gluten intake (food servings containing gluten)before enrollment in the trial was assessed by a questionnaire.

The 3 main exclusion criteria were as follows: dietary glutenexclusion before enrollment in the trial based on a dietaryquestionnaire; evidence on the history that the patient previ-ously had responded to gluten restriction or exclusion; andevidence of celiac disease by positive serum tissue transglutami-nase IgA or IgG, or a recorded SB biopsy suggestive of celiac

disease, or a positive serum anti-endomysial antibody test iftissue transglutaminase was equivocal. Tissue transglutaminaseIgA and IgG were measured by commercial enzyme-linked im-munoassay kits (Abnova, San Diego, CA).

Other procedurally related exclusion criteria were as follows:use of tobacco products within the past 6 months; use ofnonsteroidal anti-inflammatory drugs or aspirin within the pastweek (because they affect intestinal permeability); use of artifi-cial sweeteners within 2 days before the study (to avoid contam-ination of the 2-sugar permeability test); use of medications thataffect gastrointestinal motility within 2 days before the transitstudies; and use of medications that increase risk of bleedingfrom mucosal biopsies.

Twenty-eight of the 45 participants agreed to undergo upper-gastrointestinal endoscopy and flexible sigmoidoscopy. All theseparticipants underwent both procedures under conscious sedationperformed by the same 3 endoscopists (M.V.-R., J.A.M., M.C.).

Assessment of Stool Frequency, Consistency,and Ease of Passage

During the 14-day baseline and the 28-day study peri-ods, patients completed a daily bowel pattern diary (whichincluded the 7-point Bristol Stool Form Scale,12 ranging from1 hard lumpy stool, to 7 watery diarrhea) to record theirbowel habits. The bowel pattern diary was dispensed at thescreening visit, and the completed bowel diary was collected atthe end of the study. The diary also recorded the date and timing

of each bowel movement, the ease of passage (to assess evacua-tion) and completeness of evacuation, and any medicationsreceived. This instrument has been used extensively for morethan a decade since its incorporation in pharmacodynamic stud-ies in our laboratory.13

HLA Genotyping

DNA was extracted from peripheral blood for HLA typ-ing of DQ alleles. HLA-DQ2 and HLA-DQ8 were determinedusing 6 HLA-tagging single-nucleotide polymorphisms.14

Measurement of Gastric Emptying and Small-Bowel and Colonic Transit With Scintigraphy

Measurement of gastric emptying, small bowel, and co-lonic transit with scintigraphy was conducted as in prior studiesthat described the performance characteristics of the meth-ods1517 and involved ingestion of a radiolabeled solid 296-kcalmeal and radiolabeled activated charcoal delivered by means ofa pH-sensitive capsule to the distal small bowel to measurecolonic transit. Anterior and posterior gamma camera imageswere obtained immediately after radiolabeled meal ingestion,hourly through 8 hours, and then at 24 and 48 hours afterradiolabeled meal ingestion. Two standardized meals were in-gested at 4 hours (530-kcal chicken meal) and 8 hours (750-kcalroast beef sandwich, which included 2 slices of bread).

A variable region-of-interest program was used to measureisotope counts in each region and, thereby, derive a transitmeasurement (after correction for radioisotope decay and tissue

CLI NI CALAT

904 VAZQUEZROQUE ET AL GASTROENTEROLOGY Vol. 144, No. 5
http://www.clinicaltrials.gov/http://www.clinicaltrials.gov/http://www.clinicaltrials.gov/


3/12

attenuation) to estimate the proportion of technetium-99memptied from the stomach at each time point or filling the colonat 6 hours, and the proportion of indium-111 in each colonicregion at specified times. Gastric emptying was measured over 4hours after meal ingestion. Filling of the colon at 6 hours servedas a valid surrogate for SB transit.15

Overall colonic transit was summarized as the colonic geo-metric center (weighted average of counts in the colonic regions

[ascending, transverse, descending, rectosigmoid, and stool, re-spectively, numbers 15]) at 24 and 48 hours.

Measurement of Small-Bowel and ColonicPermeability

Measurement of SB and colonic permeability was con-ducted in the Clinical Research Unit, as in prior studies, withstandardized meals ingested during the first 8 hours and waterallowed ad libitum throughout the day.4,18 Lactulose 1000 mg(L7877; Sigma-Aldrich, St. Louis, MO) and mannitol 200 mg(M8429; Sigma-Aldrich) were used to determine the urine sugarexcretions at different times as markers of SB (0 2 h) andcolonic (824 h) mucosal permeability after oral ingestion of the

sugars in aqueous solution. Urine was collected every 30 minutesfor the first 2 hours and accumulated for the entire 2 hours(validated measure of SB permeability), every 2 hours for thenext 6 hours, and from 8 to 24 hours (validated measure ofcolonic permeability). The total volume of each collection wasmeasured and an aliquot was stored at 20C until it wasthawed for analysis.

Participants ingested standardized meals during the first 8hours, specifically, 500 mL water was given 30 minutes aftersugar administration to aid in the collection of urine, a breakfastof egg and toast was given after 2 hours, and a lunch of chickenand potato was offered after 6 hours. Water was allowed adlibitum throughout the day and during meals.

Urinary saccharide concentrations were measured by high-

performance liquid chromatographytandem mass spectrome-try. Details of the method were described previously.18,19 Cumu-lative excretion in each collection (02 h and 824 h) wascalculated as follows: (concentration of sugar [g/mL]) * totalurine volume (mL). The lactulose:mannitol ratio (L:M ratio) wascalculated as follows: L:M ratio 0.2 (cumulative excretionlactulose)/(cumulative excretion mannitol).

Quantitation of Tight Junction Proteins UsingReal-Time Polymerase Chain Reaction

We used real-time polymerase chain reaction (PCR) toquantitate mRNA expression of TJ proteins (zonula occludens 1[ZO-1], occludin, claudin-1, and glyceraldehyde 3-phosphate de-

hydrogenase [GAPDH, control]) in SB and rectosigmoid biopsysamples from 28 IBS-D patients who consented to undergoendoscopy solely for research.

Small-bowel and rectosigmoid colon biopsy samples weresubmerged in RNAlater Solution (Ambion, Austin, TX) andstored at 80C. RNA extraction was performed as per themanufacturers instructions (RNeasy Mini Kit; Qiagen, Valencia,CA). Complementary DNA synthesis was performed using 0.2g of total RNA with the High Capacity Reverse TranscriptionKit (Applied Biosystems, Foster City, CA). TaqMan gene expres-sion assays for ZO-1, occludin, claudin-1, and GAPDH wereperformed in triplicate for each gene on an ABI Prism 7900HTreal-time PCR system (Applied Biosystems) according to themanufacturers instructions using the comparative delta-deltaCT method for relative quantification.

The expression of each gene was normalized to the endoge-

nous control, GAPDH. The post-diet fold-change was calculated

individually with respect to the prediet biopsy. Mean values of

mRNA expressions were used for the statistical analysis.

Small-Bowel and Colonic Mucosal Morphology

The SB and colonic biopsy specimens (obtained in 28 of

45 IBS-D patients) were stained with H&E to quantitativelyassess the intraepithelial lymphocytes and the villous-to-crypt

ratio and CD3 and CD8 lymphocytes. Section orientation was

controlled by counting cells in areas with 3 adjacent villi in

small-bowel biopsy specimens and 3 adjacent crypts in colonic

biopsy specimen viewable for their entire length. The methods

used for immunochemistry and quantitative and semiquantita-

tive morphologic analyses are summarized in Appendix 2.

Proliferative Responses and In Vitro CytokineResponses of PBMCs to Gluten and Rice

In celiac disease, DQ2 and DQ8 restricted gluten-

specific inflammatory T cells mediate inflammation through the

production of interferon (IFN)-. There is also an inflammatoryinnate response to gluten that is mediated partly through the

production of TNF-.20 To help determine whether the immune

response to gluten in the IBS patients evaluated was an inflam-

matory adaptive T-cell response to gluten or an inflammatory

innate immune response to gluten, the PBMCs of patients were

tested for the ability to proliferate in response to in vitro gluten

stimulation, and the supernatants were evaluated for IFN-,

TNF-, and other cytokines.

The method used to assess the proliferative responses and in

vitro cytokine responses of PBMCs to gluten and rice is de-

scribed in Appendix 2.

Statistical AnalysisThe primary end points were colonic geometric center at

24 hours and urine mannitol excretion at 02 hours and 824

hours. Secondary end points were gastric emptying half-time

(min), colonic filling at 6 hours (%), geometric center at 48

hours, ascending colon half-time, urine lactulose at 824 hours,

urine lactulose:mannitol ratio at 02 hours and 824 hours,

mucosal inflammation, and bowel functions (frequency, consis-

tency, ease of passage). The level was set at .05 for the primary

end points, and we did not adjust the level for diet compari-

sons of the secondary end points.

The primary analyses used analysis of variance (ANOVA) or

analysis of covariance models (using prediet values and, where

relevant, body mass index as a covariate) to assess the effects of

diet. Thus, the effects of the diet on transit and bowel function

were direct comparisons of post-diet A vs post-diet B observa-

tions, adjusted for the prediet measurements in each individual

patient. Differential diet effects depending on HLA status were

assessed using analysis of covariance models by also including

an overall HLA term and an HLA by diet interaction term. An

intention to treat (ITT) analysis was used, including all subjects

randomized or those consenting to the biopsy portion.

The Student ttest was used to compare cytokine production in

vitro in response to rice and gluten. Other details of the statistical

analysis and statistical power are included in Appendix 2.

All authors had access to the study data and reviewed and

approved the final manuscript.

CLINICALAT

May 2013 DIETARY GLUTEN, HLA, BOWEL FUNCTIONS IN IBS 905


4/12

ResultsDemographics and Baseline Characteristics

Supplementary Figure 1 shows the trial flow andsummarizes the baseline demographic data. There wereno differences in age, sex distribution, Hospital Anxietyand Depression scores, body mass index, baseline 0 2hour excretion of mannitol and lactulose, and baselinecolonic transit in the groups assigned to each diet. Asexpected in this patient population, there was a femalepredominance.

The prestudy diet questionnaire showed that thepatients were not ingesting a gluten-free diet before thestudy. The range of the number of gluten-containingfood servings per day was 1 to 15 (mean, 3.10 0.46).Only 2 patients were ingesting 15 servings of gluten-containing food per day; 90% of participants ingestedbetween 1 and 4.4 servings of gluten-containing food

per day. Participant compliance with the diets wasuniformly excellent in the dietitians reports based ondirect questions when patients picked up assigned dietsand snacks.

Stool Frequency, Consistency, and Easeof Passage

There was a diet effect on stool frequency in theoverall groups, with a statistically significant decrease insubjects who were on a GFD compared with subjects on aGCD (P .04; Figure 1A and B). The 95% confidenceinterval (CI) of the difference in mean daily stool fre-

quency for the entire study period between GFD and GCDwas 0.652 to 0.015.

The diet effects on stool frequency were more pronouncedin subjects who were HLA-DQ2 or 8 positive (P .019). InHLA-DQ2 or HLA-DQ8negative patients, the 95% CI of

Figure 1. (A) Diet effect on stool frequency (*P .04), form, and ease of passage; the effect on stool frequency was greater in HLA-DQ2 or

HLA-DQ8positive patients (P .019).(B) Mean bowel movements per day during 14-day baseline and 28-day diet treatment periods in each patient

randomized to gluten-free and gluten-containing diets. HLA-DQ2/8negative patients are indicated by the open symbols.

Table 1. Effects of GCD and GFD on Barrier Functions, Tight Junction Proteins, and Colonic Transit

GFD GCD P value, effect of diet

P value, effect of

diet

HLA HLA

cumulative urine mannitol 02 h, mg 48.6 12.3 83.0 70.9 .028 .586 .01

cumulative urine lactulose 02 h, mg 3.7 4.0 3.5 2.5 .207 .150 .708

lactulose:mannitol ratio 02 h 0.008 0.004 0.005 0.004 .0012 .006 .043

cumulative urine mannitol 824 h, mg 21.6 13.4 35.2 12.7 .358 .999 .203

cumulative urine lactulose 824 h, mg 1.71 3.22 5.25 4.98 .858 .540 .396

lactulose: mannitol ratio 824 h 0.027 0.022 0.059 0.024 .531 .445 .919

ZO-1 fold-change, SB 1.57 0.24 1.11 0.24 .065 .119 .218

Occludin fold-change, SB 1.14 0.07 1.03 0.08 .28 .017 .490

Claudin fold-change, SB 1.64 0.31 1.13 0.12 .24 .32 .41

ZO-1 fold-change, colon 1.97 0.56 1.04 0.26 .025 .038 .161

Occludin fold-change, colon 1.47 0.16 0.96 0.13 .004 .006 .178

Claudin fold-change, colon 1.63 0.23 1.01 0.15 .036 .015 .203

Colonic transit, GC 24 h 3.2 0.19 2.6 0.2 .182 .364 .251

Colonic transit, GC 48 h 4.4 0.16 4.0 0.2 .304 .352 .548

NOTE. Data show mean standard error of the mean.

, difference post-diet/prediet or fold-change post-diet/prediet; GC, geometric center.

CLI NI CALAT



5/12

the difference in mean daily stool frequency for the entire

study period between GFD and GCD was0.5712 to 0.344;

in the HLA-DQ2 or HLA-DQ8positive patients, the 95%

CI was 1.005 to 0.092.There was no significant diet effect (GFD vs GCD) on

mean daily stool form for the entire study cohort (over-

all 95% CI, 0.475 to 0.251) with similar nonsignificant

differences in the HLA-DQ2 or HLA-DQ8positive orHLA-DQ2 or HLA-DQ8negative groups.

There was no significant diet effect (GFD vs GCD) on

mean ease of passage score for the entire study cohort

(overall 95% CI, 0.337 to 0.010; P .064) with similarnonsignificant differences in the HLA-DQ2 or HLA-

DQ8positive or HLA-DQ2 or HLA-DQ8 negative

groups.

Gastrointestinal and Colonic Transit

In the overall study cohort or HLA-DQ groups,

there were no diet effects on gastric emptying and colonicfilling at 6 hours; similarly, the GFD did not delay colonic

transit at 24 or 48 hours (Table 1).

Small-Bowel and Colonic Permeability

There were 2 participants who had more than 5times the ingested mass of mannitol excreted in urine atbaseline. This was deemed to be evidence of unrecognizedingestion of mannitol from a dietary contaminant. There-

fore, the permeability was analyzed using ITT (Table 1)and per-protocol principles.

There was no diet effect on SB permeability for cu-mulative lactulose excretion in the ITT analysis (P

.21); however, the per-protocol analysis showed a P

value of .097 for the diet effect in the overall cohort(Figure 2), and P .06 in the HLA-DQ2 or HLA-DQ8positive patients (detailed later).

In the overall study cohort, there was increased SB

permeability with GCD relative to GFD, as shown byboth the cumulative mannitol excretion (Figure 2) andthe L:M ratio in the ITT analysis (P .028 and P

.0012); these effects were also significant in the per-protocol analysis (P .05 and P .0035, respectively).

In the overall study cohort and the HLA-DQ sub-groups, there were no diet effects on colonic permeability

Figure 2. Small intestinal and

colonic permeability by the cu-

mulative (A) mannitol and (B) lac-

tulose excretion at 02 h and

824 h, respectively. There was

increased small intestinal per-

meability with GCD, as shown by

both cumulative mannitol excre-

tion and the lactulose-to-manni-

tol ratio (ITT analysis, #P .028

and P .0012). The effect on

lactulose excretion was border-

line significant (*P .097).

CLINICALAT



6/12

for the cumulative mannitol and lactulose excretion (Fig-ure 2) or the L:M ratio (Table 1).

Impact of HLA-DQ Status on Small-BowelMucosal Permeability

The effect of GCD on 02 hour cumulative man-

nitol excretion was stronger in HLA-DQ2 or HLA-DQ8positive patients in both ITT and per-protocol analysis(P .0097 and P .0086, respectively).

Similarly, the effects on L:M ratio at 02 hours weremore pronounced in HLA-DQ2 or HLA-DQ8positivepatients in the ITT and per-protocol analysis (P .006and P .014, respectively).

Tight Junction mRNA Expression inSmall-Bowel and Colonic Mucosa

There was aP value of .065 for the comparison ofthe diet effect on ZO-1 mRNA expression in the SB, butno significant effect on occludin and claudin-1 mRNA in

all patients in each diet group (Table 1). However, diet-associated changes in occludin expression in SB mucosawere significant (P .017) in the HLA-DQ2 or HLA-DQ8positive group (Table 1).

Expressions of ZO-1, occludin, and claudin-1 mRNA incolonic mucosa were significantly lower in GCD relativeto GFD in the overall groups, particularly in subjects withHLA-DQ2 or HLA-DQ8positive status.

Small-Bowel and Colonic Mucosal Morphology

There were no diet effects on SB intraepitheliallymphocytes, CD3, CD8 CD4, CD68, and CD79 cells, and

tryptase-positive (mast) cells. There were no diet effects onSB ZO-1 staining as determined by the percentage of cellsand intensity of the staining (Table 2). Normal villus-to-crypt ratios confirmed the absence of celiac disease in allparticipants at baseline; there were no diet effects onvillus-to-crypt ratios.

Relationships Between Stool Frequency,Transit, and Barrier Functions

After diet intervention, we found a significant corre-lation in the overall study cohort between stool frequencyand consistency and geometric center at 24 hours. r 0.42

(P

.03) and r

0.41 (P

.03), respectively. There was nosignificant correlation between SB and colonic mRNA ex-pression of TJ protein and bowel functions.

In Vitro Proliferative Responses and CytokineResponses of PBMCs to Gluten and Rice

The overall proliferation of PBMCs to gluten orrice in IBS patients was not different, and there was noincreased proliferative response to gluten in the DQ2 orDQ8positive patients, suggesting that gluten-inducedproliferation is not enhanced by DQ status and couldarise from nonT cells in both groups of patients.

Gluten induced greater IL-10 and granulocyte colonystimulating factor (G-CSF) by PBMCs compared with rice

stimulation (P .01 and P .05, respectively; Figure 3).IFN-was not changed consistently; however, TNF- pro-duction appeared to be increased in the PBMCs of IBSsubjects with some gluten specificity (P .076). In gen-eral, cytokine responses were not predicted by DQ status.

Discussion

In this 4-week, randomized, controlled, diet inter-vention study, the intervention was associated with severalsignificant effects. Subjects on a GCD had increased stoolfrequency compared with a GFD; this effect was greater in

Table 2. Small-Bowel and Colon Histology and

Immunohistochemistry

Data show mean SEM

GFD

(n 14)

GCD

(n 14)

P value,

effect of

diet

Duodenum

IEL/100 epithelial cells 28.5 5.0 23.5 6.1 NS

CD3 lymphocytes/100

IEL

30.5 5.4 23.9 5.5 NS

CD4 lymphocytes, n

1 1 4

2 11 10

3 1 0

CD8 lymphocytes/100

IEL

29.8 5.8 24.8 4.5 NS

CD68 lymphocytes, n

1 10 9

2 3 5

CD79 lymphocytes, n

1 1 3

2 11 10

3 1 1

Mast cells number/hpf 21.3

1.8 20.2

1.6 NSZO-1 IHC 152.3 22.3 165 21.5

Villus:crypt ratio, %

patients

3 38.5 35.7 NS

4 61.5 64.3 NS

Colon

IEL/100 epithelial cells 3.1 0.4 2.6 0.5 NS

CD4 lymphocytes, n

1 13 9

2 0 5

CD68 lymphocytes, n

1 8 10

2 5 4

CD79 lymphocytes, n

1 9 102 3 3

3 1 1

ZO-1 IHC 115.4 19.8 135.7 18.6

NOTE. Data were available in 28 patients. CD4, CD68, and CD79 were

scored semiquantitatively (grades 14) and ZO-1 immunohistochem-

istry was scored semiquantitatively as described in the Materials and

Methods section based on positive cells and intensity of staining. n

Number of patients with stated semiquantitative grades of CD4, CD68

or CD79 lymphocytes in duodenum or colon biopsies. No statistical

comparisons were conducted for the descriptive observations of CD4,

CD68, CD79 lymphocytes, and ZO-1 IHC, in contrast to the quantitative

data on IELs, CD3, CD8, and mast cells.

IEL, intraepithelial lymphocyte; IHC, immunohistochemistry; NS, non-

significant; SEM, standard error of the mean.

CLI NI CALAT



7/12

HLA-DQ2 or HLA-DQ8positive patients. Although theabsolute difference in stool frequency was small, it isimportant to appreciate that this increased frequency ison a background of the typical increase in stool frequency(average, 2.6 bowel movements/day at baseline) and con-sistency in patients with IBS-D. In addition, the patientsstudied were not selected on the basis of prior clinicaldemonstration of sensitivity to gluten; therefore, our ob-servations are generalizable to the broader spectrum ofIBS-D, rather than those with a prior response to GFD.Patients on a GCD also had increased SB permeability, asshown by increased mannitol excretion and an increasedlactulose-to-mannitol ratio and, overall, this effect wasgreater in the HLA-DQ2 or HLA-DQ8positive patients.A GCD resulted in reduced mRNA expression of mucosalTJ proteins (most clearly shown in colonic mucosa, andborderline significant for ZO-1 [P .065] in SB mucosa)

in the absence of mucosal inflammatory markers. Effectson mRNA expression of TJ proteins also were associatedwith HLA-DQ2 or HLA-DQ8positive status. Comparedwith GFD intervention, the GCD was not associated withsignificant effects on colonic transit, immunocyte activa-tion, or intraepithelial lymphocytes in SB or colonic mu-cosa, or SB villus:crypt ratios. A limitation in the quanti-fication of TJ protein expression was that it was basedsolely on PCR (mRNA expression), which may not neces-sarily match protein expression. Future studies will incor-porate Western blot and other methods to identify theseproteins and their distribution directly.

Patients on a GCD had increased SB permeability relativeto GFD. Although the clinical significance of these changesin permeability were not shown in the current study, theabundant experimental evidence from the published litera-ture that increased mucosal permeability enhances inflam-mation and leads to increased sensitivity is summarizedelsewhere.21,22 Increased intestinal permeability has been de-scribed in previous studies in postinfectious IBS patients,6,23

and we report the effect of gluten in intestinal permeabilityin IBS-D patients. Increased mucosal permeability conceiv-ably may result in greater fluid flux toward the lumen(thereby altering stool consistency) or activation of sensorypathways that result in hypersensitivity.24 The mechanismunderlying the host interaction with gluten that causes in-

creased SB permeability in IBS-D is not well understood.However, it appears that te HLA-DQ2 or HLA-DQ8 geno-type is a susceptibility factor because the overall effects ofgluten on barrier function (lactulose:mannitol ratio and onmRNA expression of TJ proteins) were more pronounced inHLA-DQ2 or HLA-DQ8positive IBS-D patients. A priorstudy in gluten-sensitive patients with IBS exposed to arandomized dietary intervention did not document the ef-fect of HLA genotype on permeability, but there was nosignificant effect of diet on permeability overall.25 DecreasedZO-1 expression in jejunal mucosa of IBS-D patients wasreported recently,5 and our study shows one dietary factor(gluten) that may alter mRNA expression of TJ proteins inIBS-D, particularly, occludin, with greater susceptibility inHLA-DQ2 or HLA-DQ8positive patients.

The association with HLA-DQ genotype suggests thatan adaptive immune response may explain the effects of

gluten on barrier function. It also is conceivable thatgluten may mediate an immune response through theinnate pathway because recent studies have shown in-creased levels of Toll-like receptors in mucosa of gluten-sensitive patients.7 One hypothesis, based on the latterobservation,7 is that the Toll-like receptors interact withgluten and activate immune responses that may lead tochanges in mucosal barrier function. Alternatively, thePBMC proliferative and cytokine responses to gluten andrice antigens suggest an increased response to gluten andrice, and, in support of this hypothesis, we have observedincreased gut mucosal permeability in patients with IBS-

D,4

which also was confirmed in these same patients atbaseline compared with controls.26 The PBMC cytokineresponses (particularly TNF- and granulocyte-macro-phage colonystimulating factor [GM-CSF]), which areincreased in response to gluten, suggest monocytic stim-ulation or innate immune responses that are not DQ2/8restricted. Increased PBMC cytokine responses to glutenhave been well documented in celiac disease,27 and herewe document PBMC cytokine responses in nonceliac pa-tients with IBS-D. The lack of proliferation and IFN-expression in response to gluten stimulation of the PB-MCs strongly suggests that the gluten-induced effectsthat are generated in the gluten-sensitive IBS patients arenot mediated by INF-producing gluten-specific CD4 T

Figure 3. In vitro cytokine pro-

duction by PBMCs in response

to stimulation with either rice

(control) or gluten. Note the in-

creased production of IL-10 and

GM-CSF, and the borderline in-

creased TNF- in response to

gluten, compared with rice.CLINICALAT



8/12

cells similar to those seen in celiac disease. The increasedproduction of the combination of cytokines IL-10, GM-CSF, and TNF- in response to gluten over rice (a controlgrain) would suggest that other cells (potentially mono-cytes, dendritic cells, eosinophils, or natural killer T cells)are being stimulated by gluten. Overall, the increasedproduction of TNF- in the absence of IFN-production

after in vitro treatment of PBMCs with gluten fragmentsis consistent with the stimulation of monocytes and,possibly, dendritic cells directly by gluten in an innatefashion.28 Further experiments are needed to determinewhich cell groups could be contributing to this particularcytokine pattern in the periphery and whether the samecytokine profile is observed in the gut.

Our data convincingly showed effects of gluten on theincreased mRNA expression of all the measured TJ pro-teins in colonic tissue relative to GFD. One limitation ofthe study was the inability to document alterations incolonic permeability using the 2-sugar excretion profilefrom 8 to 24 hours. We perceive that this may represent alack of sensitivity of the lactulose and mannitol excretiontest, for example, because of the metabolism of bothsugars by colonic bacteria.29 There are advantages to mea-suring both tissue and in vivo markers of barrier function.Another limitation was that the mechanism for improve-ment in stool frequency on a GFD in the absence ofchanges in colonic transit was not elucidated by ourstudies. Because it is unclear whether gluten or its meta-bolic products induce specific secretory mechanisms, thecurrent hypothesis is that change in stool frequency mayreflect change in fluid secretion from increased mucosalpermeability. Our current studies did not evaluate effects

of gluten on the microbiome, afferent functions, or cyto-kine expression in the mucosal biopsy specimens frompatients before and after the interventions. These wouldbe interesting parameters to include in future studies.Finally, our study did not specifically address the effects ofgluten protein per se, and it is possible that other proteins inwheat flour may be responsible for the changes observed.

Overall, our data provide mechanistic explanations forthe observation that gluten withdrawal may improve pa-tient symptoms in IBS. The data also explain, in part, theobservation of the relationship of HLA genotype to ben-eficial effects of gluten withdrawal2 in view of our results

showing that biological effects of gluten were associatedwith HLA-DQ2 or HLA-DQ8 genotype. The data suggestthat the relationship of dietary factors, innate and adaptiveimmune responses, and mucosal interactions in IBS-D de-serve further study, and they support the need for furtherclinical intervention studies to evaluate the clinical effects ofgluten withdrawal in patients with IBS-D.

Supplementary Materials

Note: To access the supplementary materialaccompanying this article, visit the online version ofGastroenterology at www.gastrojournal.org, and at http://dx.doi:10.1053/j.gastro.2013.01.049.

References

1. Cash BD, Rubenstein JH, Young PE, et al. The prevalence of celiac

disease among patients with nonconstipated irritable bowel syn-

drome is similar to controls. Gastroenterology 2011;141:1187

1193.

2. Wahnschaffe U, Schulzke JD, Zeitz M, et al. Predictors of clinical

response to gluten-free diet in patients diagnosed with diarrhea-

predominant irritable bowel syndrome. Clin Gastroenterol Hepatol

2007;5:844850.3. Verdu EF, Huang X, Natividad J, et al. Gliadin-dependent neuro-

muscular and epithelial secretory responses in gluten-sensitive

HLA-DQ8 transgenic mice. Am J Physiol 2008;294:G217G225.

4. Rao AS, Camilleri M, Eckert DJ, et al. Urine sugars for in vivo gut

permeability: validation and comparisons in irritable bowel syn-

drome-diarrhea and controls. Am J Physiol 2011;301:G919

G928.

5. Martnez C, Vicario M, Ramos L, et al. The jejunum of diarrhea-

predominant irritable bowel syndrome shows molecular altera-

tions in the tight junction signaling pathway that are associated

with mucosal pathobiology and clinical manifestations. Am J Gas-

troenterol 2012;107:736746.

6. Dunlop SP, Hebden J, Campbell E, et al. Abnormal intestinal

permeability in subgroups of diarrhea-predominant irritable bowel

syndromes. Am J Gastroenterol 2006;101:12881294.7. Sapone A, Lammers KM, Casolaro V, et al. Divergence of gut

permeability and mucosal immune gene expression in two gluten-

associated conditions: celiac disease and gluten sensitivity. BMC

Med 2011;9:2334.

8. Pellett PL. Food energy requirements in humans. Am J Clin Nutr

1990;51:711722.

9. Drossman DA, Corazziari E, Talley N, et al, eds. Functional bowel

disorders and functional abdominal pain. The functional gastroin-

testinal disorders. 2nd ed. McLean, VA: Degnon Associates,

2000.

10. Talley NJ, Phillips SF, Wiltgen CM, et al. Assessment of functional

gastrointestinal disease: the bowel disease questionnaire. Mayo

Clin Proc 1990;65:14561479.

11. Zigmond AS, Snaith RP. The hospital anxiety and depression

scale. Acta Psychiatr Scand 1983;67:361370.

12. Lewis SJ, Heaton KW. Stool form scale as a useful guide to

intestinal transit time. Scand J Gastroenterol 1997;32:920924.

13. Coulie B, Szarka LA, Camilleri M, et al. Recombinant human

neurotrophic factors accelerate colonic transit and relieve consti-

pation in humans. Gastroenterology 2000;119:4150.

14. Monsuur AJ, de Bakker PI, Zhernakova A, et al. Effective detection

of human leukocyte antigen risk alleles in celiac disease using tag

single nucleotide polymorphisms. PLoS One 2008;3:e2270.

15. Cremonini F, Mullan BP, Camilleri M, et al. Performance charac-

teristics of scintigraphic transit measurements for studies of ex-

perimental therapies. Aliment Pharmacol Ther 2002;16:1781

1790.

16. Burton DD, Camilleri M, Mullan BP, et al. Colonic transit scintig-

raphy labeled activated charcoal compared with ion exchangepellets. J Nucl Med 1997;38:18071810.

17. Deiteren A, Camilleri M, Bharucha AE, et al. Performance charac-

teristics of scintigraphic colon transit measurement in health and

irritable bowel syndrome and relationship to bowel functions.

Neurogastroenterol Motil 2010;22:415423.

18. Camilleri M, Nadeau A, Lamsam J, et al. Understanding measure-

ments of intestinal permeability in healthy humans with urine

lactulose and mannitol excretion. Neurogastroenterol Motil 2010;

22:e15e26.

19. Lostia AM, Lionetto L, Principessa L, et al. A liquid chromatogra-

phy/mass spectrometry method for the evaluation of intestinal

permeability. Clin Biochem 2008;41:887892.

20. Jelnkov L, Tuckov L, Cinov J, et al. Gliadin stimulates human

monocytes to production of IL-8 and TNF-alpha through a mecha-

nism involving NF-kappaB. FEBS Lett 2004;571:8185.

CLI NI CALAT

http://www.gastrojournal.org/http://www.gastrojournal.org/http://dx.http//dx.doi.org/10.1053/j.gastro.2013.01.049http://dx.http//dx.doi.org/10.1053/j.gastro.2013.01.049http://dx.http//dx.doi.org/10.1053/j.gastro.2013.01.049http://dx.http//dx.doi.org/10.1053/j.gastro.2013.01.049http://dx.http//dx.doi.org/10.1053/j.gastro.2013.01.049http://www.gastrojournal.org/


9/12

21. Camilleri M, Madsen K, Spiller R, et al. Intestinal barrier function

in health and gastrointestinal disease. Neurogastroenterol Motil

2012;24:503512.

22. Camilleri M, Lasch K, Zhou W. The confluence of increased per-

meability, inflammation, and pain in irritable bowel syndrome.

Am J Physiol 2012;303:G775G785.

23. Marshall JK, Thabane M, Garg AX, et al. Intestinal permeability in

patients with irritable bowel syndrome after a waterborne out-

break of acute gastroenteritis in Walkerton, Ontario. Aliment Phar-

macol Ther 2004;20:13171322.24. Bielefeldt K, Levinthal D. Pieces of a puzzle: permeability, proin-

flammatory pathways and pain? Pain 2009;146:78.

25. Biesiekierski JR, Newnham ED, Irving PM, et al. Gluten causes

gastrointestinal symptoms in subjects without celiac disease: a

double-blind randomized placebo-controlled trial. Am J Gastroenterol

2011;106:508514.

26. Vazquez-Roque MI, Camilleri M, Smyrk T, et al. Association of

HLA-DQ gene with bowel transit, barrier function and inflammation

in irritable bowel syndrome with diarrhea. Am J Physiol Gastroin-

test Liver Physiol 2012;303:G1262G1269.

27. Stepniak D, Koning F. Celiac diseasesandwiched between innate

and adaptive immunity. Hum Immunol 2006;67:460 468.

28. Cinova J, Palov-Jelnkov L, Smythies LE, et al. Gliadin peptides

activate blood monocytes from patients with celiac disease. J Clin

Immunol 2007;27:201209.

29. Meddings JB, Gibbons I. Discrimination of site-specific alterations

in gastrointestinal permeability in the rat. Gastroenterology 1998;

114:8392.

Received September 26, 2012. Accepted January 15, 2013.

Reprint requests

Address requests for reprints to: Michael Camilleri, MD, Mayo

Clinic, Charlton 8-110, 200 First Street SW, Rochester, Minnesota

55905. e-mail: [email protected].

Acknowledgments

The authors thank Michael Ryks and Debbie Rhoten for technical

support, and Cindy Stanislav for secretarial assistance.

Maria Vazquez-Roque was the fellow investigator, and was

responsible for participant recruitment and wrote the article;

Michael Camilleri was the principal investigator, wrote the

protocol, and wrote the article; Thomas Smyrk was responsible forsurgical pathology; Joseph Murray wrote the article; Jessica ONeill

was the study coordinator and was responsible for participant

recruitment; Eric Marietta performed proliferation and cytokine

expression studies; Paula Carlson was responsible for HLA typing,

tight junction protein expression, and preparation of

histopathology; Jesse Lamsam performed the intestinal

permeability measurements; Denise Janzow was responsible for

participants randomized controlled diets; Deborah Eckert was the

study coordinator and was responsible for participant recruitment;

Duane Burton performed analysis of gastrointestinal and colonic

transit and was responsible for database management; and Alan

Zinsmeister was the study statistician.

Conflicts of interestThis author discloses the following: Joseph Murray has received

grants from Alba Therapeutics for clinical trials with the drug

larazotide. The remaining authors disclose no conflicts.

Funding

This study was supported in part by National Institutes of Health

grants 1RC1-DK086182 and R01-DK092179 (M.C.) and by a

National Institutes of Health Clinical Translational Science Awards

Program grant (UL1 TR000135).

CLINICALAT

mailto:[email protected]:[email protected]:[email protected]


10/12

Immunochemistry

Immunohistochemical staining of 5-mthickparaffin sections was performed using a commercial kit(EnVision; Dako, Carpinteria, CA). Slides were depar-affinized and rehydrated to water. Heat-induced epitoperetrieval was performed in a citrate buffer target retrieval

solution (pH 6.0; Dako) in a waterbath for 20 minutes.Endogenous peroxidase activity was blocked for 10 min-utes at room temperature in 3% H2O2, and rinsed inrunning tap water. Nonspecific protein binding sites wereblocked by applying 5% normal goat serum diluted toslides for 10 minutes at room temperature. The serumwas blotted off and the sections subsequently were incu-bated with the primary antibodies for 30 minutes atroom temperature. After processing according to themanufacturers protocols, 3, 3=-diaminobenzidine wasused as the final chromogen, and nuclei were counter-stained with Mayers hematoxylin.

The primary antibodies used were CD3, CD4, CD8,CD68, CD79, tryptase (Dako), and ZO-1 (Invitrogen Cor-

poration, Camarillo, CA). Biopsy specimens were assessedby one experienced gastrointestinal pathologist (T.S.)who was blinded to diet allocation and each patientsHLA-DQ genotype.

Quantitative Morphologic Analyses

Intraepithelial lymphocytes were counted on theH&E slides and expressed as intraepithelial lymphocyte/100epithelial cells, and on the slides stained for CD3 and CD8the latter were expressed per 100 intraepithelial lymphocytecells. Mast cells were counted on the tryptase-stained slidesand expressed as mast cells per high-power field.

Semiquantitative Morphologic Analyses

CD4, CD68, and CD79a were scored on a semi-quantitative scale of 14: 1 small groups of cells inexpected location (superficial lamina propria in colon,superficial crypts in duodenum); 2 larger groups, still

in the expected location; 3

expansion of positive cellsbeyond the usual location; and 4 wall-to-wall positive.

Appendix Table 1. Menus of Food Choices Provided by the Metabolic Kitchen for the Randomized Controlled Trial

Breakfast

Egg or Egg Beaters (Con Agra Foods; Fresno, CA)

Shredded cheese or breakfast ham

Cereal: Rice Chex (General Mills, Minneapolis, MN), Honey

Nut Chex (General Mills, Minneapolis, MN) sugar if desired

milk (skim or whole)

Breada (1 or 2 slices) Patient must agree to eat white or wheat

Butter/margarine/Skippy Natural Peanut

Butter/jam/jelly/honey

Banana

Orange juice

Coffee/tea/cocoa

Brownberry (Bimbo Bakeries, Horsham, PA)

Wheat/white/toast (1 or 2 slices)

Lunch

Chicken noodle or tomato soupa (no saltines)

Turkey, ham, or roast beef sandwich on wheat or white

Cheddar or American cheese

(mayonnaise/Miracle Whip Free (Kraft Foods, Inc, Northfield,

IL)/butter/margarine/lettuce)

Peaches

Yogurt Stonyfield (Stonyfield Farm, Inc, Londonderry, NH)

onlyb

2% milk/skim milk

Dinner

Steak, chicken, pork tenderloin, or cod

Catsup/lemon

Baked potato butter, margarine, sour cream

Rice (plain or chicken) or macaronia butter, margarine

Broccoli

Romaine salad with choice of dressing

Ice cream with chocolate syrup or strawberries

Bread

Juices: apple, grape, grapefruit, pineapple, tomato, cranberry

Fruit: apricots, apple slices, applesauce, orange, mandarin oranges,

pears, raisins, strawberries, pineapple

Other Starch: waffle with syrup,a French toast (homemade)a with syrup,

muffin lemon streusel, blueberry, chocolate, bagel with cream cheese

Dairy: cottage cheese, frozen yogurt, sherbet shake, shake

chocolate/vanilla/strawberry, vanilla or chocolate pudding

Candy: Hershey Kisses, Hershey Milk Chocolate Bar, Snickers (The

Hershey Company, Hershey, PA)

Chips: corn chips, potato chips, popcorn

Frozen entrees (gluten rich only):c chicken, broccoli and cheese, meat

lasagna, Swedish meatball, chicken fettuccine, macaroni and cheese,

spaghetti and meatballs, chicken and vegetables, cheese cannelloni,

Thai style chicken

Sandwich: tuna or egg salad

Vegetables: green beans, peas, corn, asparagus, squash, carrots

Salad Dressings: Thousand Island, ranch, French, zesty Italian, blue

cheese, fat free Italian, fat free French, fat free Ranch, tabasco sauce

No food or drink with artificial sweeteners

NOTE. Typical beverage intake: water, coffee, tea, pop, milk, juice.aFoods available in both gluten-free and gluten-rich options.bThis brand was selected because it is certified by the manufacturer as gluten free whereas other brands include a warning that they may contain

gluten.cPrecautions were taken not to state brand names of frozen meals.

911.e1 VAZQUEZROQUE ET AL GASTROENTEROLOGY Vol. 144, No. 5


11/12

ZO-1 was scored as the product of the percentage ofpositive epithelial cells, along with a semiquantitativeassessment of stain intensity (on a scale of 03).

Proliferative Responses and In Vitro

Cytokine Responses of PBMCs toGluten and Rice

PBMCs were isolated after the administration ofthe diets using a Ficoll-Paque (GE Healthcare Life Sci-ences, Piscataway, NJ) gradient and then stimulated invitro with a peptic tryptic digest of gluten or rice. Prolif-eration was evaluated using incorporation of tritiatedthymidine and cytokine production, evaluated by multi-plex beads (Affymetrix-Panomics, Santa Clara, CA).

A buffy coat from each patient was obtained after theadministration of the diet and PBMCs were obtainedafter a Ficoll-Paque gradient. The cells then were washed

twice in 10% RPMI medium. Cells were plated at 2.5 106/mL and further stimulated with a peptic tryptic di-gest of food-grade rice or food-grade gluten at 100ug/mL for 24 hours, at which point supernatant wasextracted for cytokine production. Tritiated thymidinethen was added for an additional 24 hours, after whichincorporation into DNA was evaluated using a scintilla-tion counter. Cytokine evaluation was performed using abioplex (custom-made from Affymetrix-Panomics). Fordetermining gluten- or rice-specific production of thecytokines, background production of cytokines (additionof media only) was subtracted from the amount pro-

duced in response to gluten or rice. Rice flour used forthe peptic tryptic digest was obtained from Bobs RedMill (Milwaukie, OR), and the wheat flour used wasobtained from Manilda Milling Corporation (ShawneeMission, KS).

Additional Statistical Considerations

To mitigate the observed skewness in the excre-tion of mannitol and lactulose, the deltas (post-diet mi-nus prediet) first were computed, transformed to rankscale, and the diet effects were analyzed using an ANOVAmodel. In the subset of subjects with biopsy data, diet

effects on mucosal inflammation (SB and colon intraepi-thelial lymphocytes) and immunocyte and mast cellcounts per high-power field were assessed using the Wil-coxon rank-sum test. For tight junction gene expression,the expression of each gene, ZO-1, occludin, and clau-din-1, was normalized to the control, GAPDH; the post-diet fold-change was calculated individually with respect

to the prediet biopsy. These post-diet fold-changes thenwere transformed to rank scale, and ANOVA models(overall diet effects) and analysis of covariance models(potential differential effects depending on HLA status)were examined.

Subjects with missing end point values had these val-ues imputed using the corresponding mean overall sub-jects (with nonmissing data). A correction to the error

degrees of freedom in the analysis of covariance modelswas made (subtracting 1 for each value imputed) toadjust the residual error variance to account for theimputations.

The power to detect differences between treatmentgroups for the primary end points of interest was basedon the variation observed in the current study (AppendixTables 2AC). In general, the sample sizes in the 2 dietgroups would have provided 80% power to detect approx-imately 30%70% differences in the primary end points ofinterest.

Appendix Table 2A. Excretion of Mannitol and Lactulose in

Baseline, Prediet Study

Time period

Mannitol, mean

standard deviation

Lactulose, mean

standard deviation

Cumulative 02 h

excretion, mg

184 77 20.1 18.6

Cumulative 824 h, mg 60 66 24.7 21.9

Appendix Table 2B. Sample Size Assessment: Effect SizesDetectable With Approximately 80%

Power Between Two Diet Treatment

Groups

Response (data from

baseline study)

Difference

detectable with

80% powera Effect sizeb

N 12c N 24d N 12c N 24d

02 h mannitol excretion,

mg

92 64 40% 30%

02 h lactulose excretion,

mg

22 16 71% 57%

Cumulative 824 h

mannitol, mg

79 55 79% 63%

Cumulative 824 h

lactulose, mg

26 18 68% 53%

aBased on a 2-sample t test (2-sided .05).bEffect size is the difference as a percentage of the overall mean.cEffect of diet within HLA genotypes.dEffect size detectable for overall main effect of diet.

May 2013 DIETARY GLUTEN, HLA, BOWEL FUNCTIONS IN IBS 911.e2


12/12

Appendix Table 2C. Sample Size Assessment for Transit Data: Effect Sizes Detectable With Approximately 80% Power

between Two Diet Groups

Response (data from baseline study) Mean Standard deviation COV, %

Effect sizea detectable with 80% powerb

N 12/groupc N 24/groupd

Colon GC 24 h 3.12 1.21 39 61% 38%Ascending colon t, h 10.7 7.6 71 74% 51%

COV, coefficient of variation; GC, geometric center.aEffect size is the difference as a percentage of the overall mean.bBased on a 2-sample t test (2-sided .05).cEffect of diet within HLA genotypes.dEffect size detectable for overall main effect of diet.

Supplementary Figure 1. Trial flow according to the CONSORT guidelines and baseline demographics showing comparability of the 2 diet

treatment groups.

911.e3 VAZQUEZROQUE ET AL GASTROENTEROLOGY Vol. 144, No. 5

a controlled trial of gluten-free diet in patients with irritable bowel syndrome-diarrhea- effects...

Documents