Tests for Lactose

Comparison of a Real-Time Polymerase Chain Reaction Assay for Lactase Genetic Polymorphism With Standard Indirect Tests for Lactose Maldigestion

ANDREW SZILAGYI,*PAULA MALOLEPSZY,*ELISE HAMARD,?XIAOQING XUE,§NIR HILZENRAT,*MARY PONNIAH,* ELIZABETH MACNAMARA,?and GEORGE CHONG?

*Division of Gastroenterology,and§Department of Medicine,Sir Mortimer B.Davis Jewish General Hospital,and?Department of Clinical Biochemistry,McGill University School of Medicine,Montreal,Quebec,Canada

Background&Aims:There is a discrepancy in outcome between the lactose tolerance and breath hydrogen tests for

lactose maldigestion.The availability of a validated genetic

test for lactase polymorphism allows a reevaluation of these tests.Methods:Thirty healthy adults participated in a 50-g lactose challenge test at a university clinic.Blood was

drawn for genetic and timed blood glucose testing(2

hours),and breath hydrogen was measured(4.5hours).

Lactase genetic polymorphism was assessed by a real-time

polymerase chain reaction assay.Participants completed a

diet questionnaire,and symptoms were recorded during the

lactose challenge.Sensitivity and speci?city were calculated

for each indirect test.The2-way kappa coef?cient between

these tests was evaluated.Student t test and Wilcoxon rank sum test were used to compare variables.Results:The lactose tolerance test as a standard had an87.5%sensitivity and92.7%speci?city for genetic status.Only a moderate agreement between lactose tolerance test and breath hydro-gen test was observed(2-way kappa coef?cient,.53;95% con?dence interval,.22–.83).When genetic status was used as standard,symptoms had a moderate sensitivity and spec-i?https://www.360docs.net/doc/1812411435.html,ctose tolerance test had very good sensitivity,and the breath test had excellent speci?city.Conclusions: Both indirect tests independently have good to very good sensitivities and speci?cities for genetic lactase status.The noted disagreement likely re?ects variables that affect the tests independently of intestinal lactase status.The value of these tests in the light of the availability of genetic testing is discussed.

T he phenotypic dichotomy of lactose digestion and maldi-gestion is governed by the most common genetic trait worldwide.Whereas persistence of lactase phlorizin hydrolase(LP, lactase-persistent status)is considered to be a dominant genetic trait,1its loss in adulthood is recessive,2but the majority of the world’s population become lactase nonpersistent(LNP).3The con-sequences of this dichotomy have been linked to multiple effects,some bene?cial and some detrimental.4However,the most obvious perceived effect of LNP status is the development of abdominal symptoms when lactose-containing products are consumed irregularly.The uniqueness of lactose-causing symp-toms with dairy food intake is still somewhat unclear because apparent LP subjects can also experience symptoms on lactose challenge tests.5,6During the last several decades a number of tests for lactose maldigestion were developed in an effort to identify individuals who might bene?t from reduced lactose intake.A direct test such as the analysis of intestinal enzyme levels is accurate and to date is considered the gold standard for diagnosis,but it is not practical for large-scale screening.There-fore indirect tests have been developed.The2most common tests used today require measurements of blood glucose(lactose tolerance test[LTT])or exhaled breath hydrogen(hydrogen breath test[BHT])after a lactose challenge.7The latter test has surpassed the LTT as a?rst-line clinical tool.There is some discrepancy between outcomes of the2tests when they are compared,7–9and more recently there has been some questions as to whether the BHT re?ects true LNP status.10 During the last several years a Finnish group has developed a genetic test that accurately identi?es LP and LNP sub-jects.11,12The gene for lactase phlorizin hydrolase(LCT) resides on chromosome2(2q21),13but its DNA sequence does not readily distinguish LP/LNP dichotomy.11,14Instead, a cytosine to thymine mutation,approximately14-kilobase pairs upstream from the LCT locus,associates completely with promotion of the LP/LNP status.12,15This C/T-13910locus polymorphism controls LCT transcription.14The C/T-13910 mutation can be identi?ed by real-time polymerase chain reac-tion(PCR)analysis.15,16The C/C genotype is100%associated with a reduction to?10?g/g intestinal lactase.In a predom-inantly European population,the T/T genotype is also very close to100%associated with?10?g/g of intestinal lactase. The correlation of the C/T heterozygote to intestinal lactase is 97%.12Therefore,the genetic test has been validated against intestinal lactase;however,it also does not predict symptoms. The availability of this new test allowed us to reevaluate the reliability of the2existing,most commonly used indirect tests for lactose maldigestion with the C/T polymorphism.

Methods and Participants

This study was approved by the Research and Ethics Committee of the Sir Mortimer B.Davis Jewish General Hos-pital.All participants signed written consent.Healthy,nonpreg-nant subjects without obvious functional bowel symptoms were Abbreviations used in this paper:BHT,breath hydrogen test;LCT, lactase phlorizin hydrolase;LNP,lactase nonpersistence;LP,lactase persistence;LTT,lactose tolerance test;PCR,polymerase chain reac-tion;ppm,parts per million.

?2007by the AGA Institute

1542-3565/07/$32.00

doi:10.1016/j.cgh.2006.06.009

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY2007;5:192–196

recruited on the basis of results of previous lactose BHTs.Other participants were also sought if their ethnic background sug-gested high or low likelihood of LP or LNP status.These likelihoods were based on previously published population data.3,17,18For example,Asians or peri-Mediterranean peoples have a high likelihood of LNP status(?20%),whereas northern Europeans have a low likelihood of LNP status(?20%)and a high likelihood of LP status(?80%).Subjects were also required to have been off any pro-kinetic or anti-kinetic drugs(eg, domperidone or narcotics)or antibiotics during the month preceding recruitment.

We included30participants.All were asked to consume a low carbohydrate supper the night before,although water was permitted.They were asked to refrain from smoking and to perform only quiet activities during testing.Each person was asked to?ll out a previously validated questionnaire19targeting 27items of lactose-containing foods for the previous3days before the tests.The intake of mean daily lactose(g)was independently calculated from the results of other tests.Each person then underwent the3tests run simultaneously.Blood was drawn for genetic status and glucose.Breath hydrogen was measured.The latter2tests were continued after a50-g lactose challenge dissolved in200–250mL of water(25%–20%).

Lactase Genetic Test

An indwelling cannula was inserted into the antecubital vein,and blood(20mL)was drawn.DNA was prepared by using the DNA Isolation Kit from Gentra Systems(Minneapolis, MN).We used a real-time PCR assay based on?uorescence resonance energy transfer15,16and the LightCycler DNA Master Hybridisation Probes kit(Roche Diagnostics,Mannheim,Ger-many)for analysis of the C/T genetic polymorphism.The for-ward primer GCTTGGTAAGCATTTGAGT and the reverse primer GTTGAAGCGAAGATGGG were used for PCR ampli?-cation.The sensor probe ATGTAGCCCCTGGCCT is labeled with LC-640at the5=end,and the anchor probe CCTCT-GCGCTGGCAATACAGATAAGA is labeled with?uorescein at the3=end.The mutation is detected by melting curve analysis.

Lactose Tolerance Test

Blood is drawn via the indwelling cannula,which is heparinized.After a baseline measurement,blood is drawn at 30,60,and120minutes.Blood glucose is measured by a standard colorimetric method,which oxidizes glucose to glaci-ated form(Roche Diagnostics,Indianapolis,IN).A?at2-hour glucose curve signi?es lactose maldigestion and is de?ned by a failure to rise above the baseline by more than1.1mmol/L.20 Breath Hydrogen Test

Breath hydrogen in parts per million(ppm)is measured by using a validated hand-held H2chemical sensor(EC60 gastrolyzer;Bedfont Scienti?c Ltd,Rochester,Kent,United Kingdom).This instrument has a range of0–2000ppm.21,22At each time interval3exhaled breaths are measured,and the mean is used to de?ne the interval.The baseline(all?20ppm) is subtracted from values recorded at each subsequent30-minute interval up to4.5hours.A de?nite positive value is de?ned as20ppm above baseline,9,20and a probable positive is de?ned as10ppm above the baseline.23

Symptom Score

Symptoms are also recorded at the same time intervals as breath hydrogen.Each symptom associated with sugar mal-absorption—bloating,gas,and cramps—is assigned a score of 0–3.In this scheme,0means no symptoms,1is mild,2is moderate,and3is severe.Diarrhea is scored as0(none)or1 (present).The range was0to a maximum of90(9?3?3?9?1).However,no patient was expected to achieve such a score.

Statistical Analysis

Outcomes of tests were tabulated,and sensitivity and speci?city were calculated against genetic status.Two-way kappa statistics were used to determine agreement between the 2indirect tests.The kappa statistic measures the agreement beyond chance.Values of kappa between.4–.6are considered to represent moderate agreement and values between0.6–0.8to be representative of substantial agreement.Student t test was used to compare total symptom scores between the LNP and LP groups.Wilcoxon rank sum test was used to compare daily lactose intake between symptomatic and asymptomatic subjects and among different ethnic groups.All tests were2-tailed,and statis-tical signi?cance was accepted at P?.05.

Results

The18women and12men in the group had a mean age of26.7?9.4years(range,19–61years).None were of African origin.On the basis of the questionnaire there were15 with a northwest European,6with Mediterranean,and9with Asian background.The genetic test suggested that15subjects were C/C(LNP),12subjects were C/T(LP),and3subjects were T/T(LP).The LTT showed that14of15genetically analyzed LNP subjects and2of15genetically analyzed LP subjects failed to increase blood glucose above1.1mmol/L.If we accept the LTT to de?ne lactose digestion status,comparison with genetic analysis shows a sensitivity of87.5%(14/16)and speci?city of 92.7%(13/14).The2-way kappa coef?cient between LTT and BHT,however,was.53(95%con?dence interval,.22–.83),show-ing only a moderate level of agreement.Therefore we will consider the genetic test as the standard for further compari-sons.

The results of BHT,symptom scores,and LTT are displayed in Table1.In brief,there were no subjects with a BHT?20ppm incorrectly classi?ed as LP(Figure1A),and only1subject with a BHT?10ppm was incorrectly classi?ed as being LNP.Sub-jects with BHT between10–20ppm appeared to be a mix of LNP and LP(Figure1B).Symptoms were more severe and frequent in the LNP group than the LP group(P?.0003)but were less sensitive and speci?c for differentiating between LNP and LP on the basis of genetic analysis than either the BHT or LTT.Table2lists sensitivities and speci?cities for indirect tests and symptom scores.Overall,both BHT(when subjects with results of10–19ppm are included)and LTT have similar high sensitivity,but BHT has higher speci?city.Although excluding subjects with criteria of10–19ppm reduces sensitivity,speci-?city of the BHT remains100%.

In Table3the intake of daily lactose-containing foods is compared with genetic status,subjects with or without symptoms,or estimated ethnic distribution.A trend toward statistical signi?cance of increased lactose intake in non-symptomatic subjects is noted(P?.08).

February2007COMPARISON OF TESTS FOR LACTOSE MALDIGESTION193

Discussion

The results of our study describe the outcome of lactose maldigestion tests in a select multiethnic group.We used a newly described genetic test as a comparison with2existing indirect tests.With the LTT as a standard,there were very good sensitivity and speci?city for genetically de?ned LNP status. However,there was only a moderate agreement between LTT and BHT.With the genetic test as standard,both LTT and BHT independently related well to LNP status.The LTT was some-what more sensitive,and the BHT was more speci?c.The LNP group had signi?cantly more symptoms than the LP group,but symptoms were only moderately sensitive and speci?c for genetic status.There is a suggestion that symptomatic subjects,espe-cially Asians,consumed less daily lactose than asymptomatic ones.

The genetic test de?nes LNP/LP status by virtue of the level of intestinal lactase.12The residue of enzyme allows small

quantities of lactose to be digested.Indeed,a recent study in Japanese women estimated that a single dose ingestion of less than10g of lactose will not induce diarrhea.24However,larger quantities will not be split into glucose and galactose and will reach the lower intestine.Here bacteria(mainly Clostridium and Bacteroides species25,26in LNP subjects who consume dairy foods intermittently)metabolize it into short-chain fatty acids, carbon dioxide,methane,and hydrogen.27–29These divergent fates of lactose in LNP subjects allow the indirect tests of maldigestion.In the?rst instance,administering larger doses of lactose overwhelms residual lactase and leads to intestinal hurry,which further reduces lactose intestinal contact time.As such,there is a failure to increase blood glucose signi?cantly above the baseline.In the second instance,hydrogen that dif-fuses across the intestinal and ultimately the pulmonary mem-branes can be measured in the exhaled breath.In addition to qualitatively distinguishing LNP/LP status,the indirect tests have been used to quantify digested(as above24)or maldigested lactose.30In the case of the former outcome,insulin response has also been reported.24

The BHT for lactose maldigestion was popularized by Metz et al8and has been compared with other indirect tests,includ-ing the LTT in an older publication.7Newcomer et al7used quanti?ed intestinal lactase to sucrase ratio as a gold standard

Table1.Outcomes of BHT,Symptom Score,and LTT

Tabulated as Absolute Numbers

LNP LP BHT

?20ppm120 10–19ppm23a 0–9ppm112 Symptoms

Yes11b3 No412 LTT

?1.1mmol/L142?1.1mmol/L113 NOTE.Results of all30participants are https://www.360docs.net/doc/1812411435.html,ctose digestion phenotype is de?ned by genotype(C/C?LNP,C/T and T/T?LP).

a First time interval when test became positive was under3hr for CC subjects but well over3hr for T/T subjects.The1C/T subject met criteria for positivity for?10ppm within3hr and reached?20ppm on a single peak beyond3hr.

b Total symptom score was12.9?12.3(mean?standard deviation) in the LNP group vs1.5?3.5in the LP group(P?

.0003).

Figure1.(A)Result of a composite standard breath hydrogen curve

(ppm)plotted against minutes is shown for12subjects with genotype

C/C(homozygous mutants,LNP).All subjects met criteria for levels

?20ppm.(B)Composite results of5subjects(2C/C,2T/T,and1C/T)

who met criteria of positivity at?10ppm are shown.The1C/T subject

also achieved criteria at?20ppm late into the test.Results of the2

subjects with genotype T/T(phenotype LP)represent false-positive

outcomes,whereas the subject with genotype C/T was probably also

falsely elevated.

Table2.Sensitivities and Speci?cities for Indirect Tests

and Symptom Scores

Sensitivity Speci?city

BHT(?20ppm vs all)12/15(80)15/15(100)

BHT(?20ppm vs?10ppm)a12/13(92.3)12/12(100)

Symptoms(yes/no)11/15(73.3)12/15(80)

LTT(?1.1mmol/L)14/15(93.3)13/15(86.7)

NOTE.For each test,sensitivity and speci?city for lactose maldiges-

tion assessed by genetic analysis are presented as absolute counts

(%)of subjects because of small sample size.

a These?gures recognize that individuals with BHT between10–19ppm

have an equivocal test.This represented only5/30subjects in our

sample.

194SZILAGYI ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol.5,No.2

to classify LNP and LP subjects.They found that76%of LNP participants had the expected?at glucose curve in response to lactose challenge(24%false elevation),whereas4%of LP sub-jects had a false?at curve(96%appropriately elevated).The BHT provided complete separation between LP and LNP status by2hours.This study established the BHT as the most conve-nient noninvasive test for lactose maldigestion.

Normal cutoff values for breath hydrogen and the time of appearance have been debated.The standard positive test (hence LNP status)is an increase of20ppm above the base-line.9,20A result of10–19ppm above baseline is considered more sensitive by some,23but others believe that it is likely not related to lactose maldigestion.31In the current report we tested both de?nitions for qualitative purposes.However,use of the?10-ppm de?nition led to misclassi?cation of3subjects as LNP,when by genetic analysis they were LP.In at least1 publication even LP subjects maldigested about8%of con-sumed lactose,32and a response might occur when a high-dose lactose challenge(as used here)is given.Otherwise,2LNP subjects failed to reach the?20-ppm cutoff value but achieved ?10ppm within3hours.Therefore a cutoff value of10ppm is more sensitive and should be interpreted as suspicious for LNP status.Its accuracy might be enhanced by knowledge of ethnic background.However,for diagnosis,retaining the ?20-ppm criteria restricted to2–3hours remains the most speci?c test.

Quantitative intestinal enzyme measurement has been the primary comparator for genetic and indirect tests.Whereas genetic analysis is unlikely to vary,both the LTT and BHT can be affected by factors unrelated to lactase enzyme levels.In the case of glucose measurement,gastric emptying,intestinal tran-sit time,and metabolism of glucose(eg,diabetes)can alter results.7In the case of breath hydrogen,use of antibiotics, intestinal motility,drugs,or physical effects can alter measure-ments.4Colonic adaptation de?ned as the observation of im-proved symptoms of lactose maldigestion and decreased BHT results,as a result of expanded colonic bacterial populations (most likely of the lactic acid–producing variety)with in-creased metabolic capacity for lactose,3,33might lead to false-negative interpretations.34Alternatively,postulated bacterial overgrowth in patients with irritable bowel syndrome might result in false-positive results.10These and other different con-ditions likely explain the discrepancy observed between the2 indirect tests noted here and reported in the literature.

Nevertheless,independently both tests compare favorably to genetic status.The clinical consequence of these?ndings is the con?rmation of the appropriate current use of these indirect tests.Historical symptoms of lactose intolerance retain a mod-erate sensitivity and speci?city for LNP/LP status.The LTT increases these.However,the sensitivity of the BHT can be increased by choosing the lower cutoff value of10ppm,as predicted by Strocchi et al.23Furthermore,the BHT here has excellent speci?city.In addition,the physiologic variables af-fecting these indirect tests can be exploited,when genetic status is known.For example,glucose and insulin response to lactose in either LNP or LP subjects might be evaluated by the LTT. Similarly,BHT might be used to assess function and change in lower intestinal micro?ora in response to dairy food consump-tion in either of the2cases.33,35

Limitations of this study should be noted.First,it is a small study,and the validity and applicability of its outcome will have to be further evaluated in larger groups.Second,we enrolled a select group,and as such,the results might not apply to unselected patients.

In conclusion,both the LTT and BHT independently closely re?ect LNP/LP status in a select group of subjects without known functional gastrointestinal disorders.All3tests might be useful under a variety of clinical or research situations. Further work is required to place in proper perspective the biologic role of the most common known human genetic trait, lactase de?ciency.

References

1.Gilat https://www.360docs.net/doc/1812411435.html,ctase de?ciency:the world pattern today.Isr J Med Sci

1979;15:369–373.

2.Sahi T,Isokoski M,Jussila J,et al.Recessive inheritance of

adult-type lactose https://www.360docs.net/doc/1812411435.html,ncet1973;2:823–826.

3.Scrimshaw NS,Murray EB.The acceptability of milk and milk

products in populations with a high prevalence of lactose intoler-ance.Am J Clin Nutr1988;48(Suppl):1079–1159.

4.Szilagyi A.Review article:lactose—a potential prebiotic.Aliment

Pharmacol Ther2002;16:1591–1602.

5.Suarez FL,Savaiano D,Arbisi P,et al.Tolerance to the daily

ingestion of two cups of milk by individuals claiming lactose intolerance.Am J Clin Nutr1997;65:1502–1506.

6.Suarez FL,Savaiano DA,Levitt MD.A comparison of symptoms

after the consumption of milk or lactose-hydrolyzed milk by peo-ple with self-reported severe lactose intolerance.N Engl J Med 1995;333:1–4.

7.Newcomer AD,McGill DB,Thomas PJ,et al.Prospective compar-

ison of indirect methods for detecting lactase de?ciency.N Engl J Med1975;293:1232–1236.

8.Metz G,Jenkins DJ,Peters TJ,et al.Breath hydrogen as a

diagnostic method for https://www.360docs.net/doc/1812411435.html,ncet1975;1:1155–1157.

9.Romagnuolo J,Schiller D,Bailey https://www.360docs.net/doc/1812411435.html,ing breath tests wisely in

a gastroenterology practice:an evidence-based review of indica-

tions and pitfalls in interpretation.Am J Gastroenterol2002;97: 1113–1126.

10.Pimentel M,Kong Y,Park S.Testing to evaluate irritable bowel

Table3.Estimated Average Daily Lactose Intake(g/d) Mean?SD,Median(25th Quartile,75th Quartile)

Distributed by Genetic Classi?cation,Presence or

Absence of Symptoms,and Ethnic Groups

N Mean?SD

Median

(25th,75th Quartiles)

Genetic status

C/C a1521.8?34.112.4(5.5,20.0)

C/T1225.7?18.219.6(13.5,31.3)

T/T315.6?11.211.2(4.5,31.2)

Symptoms

Yes1414.1?12.612.6(6.0,17.8)

No b1627.3?32.432.4(11.9,28.4)

Ethnic c group

NWE1522.8?17.521.0(12.5,31.2)

M633.4?52.814.7(5.5,20.0)

Asian914.1?16.48.5(6.0,13.8)

SD,standard deviation.

a One CC subject consumed140.2g lactose/d;if excluded,the mean

intake is12.92?22.94.

b No vs Yes,P?.08(trend only).

c NWE ancestry from northern an

d western Europe;M ancestry from

Mediterranean region(eg,Italy,Greece).

February2007COMPARISON OF TESTS FOR LACTOSE MALDIGESTION195

syndrome correlates with lactulose testing and may not re?ect true lactose malabsorption.Am J Gastroenterol2003;98:2700–2704.

11.Enattah NS,Sahi T,Savilahti E,et al.Identi?cation of a variant

associated with adult-type hypolactasia.Nat Genet2002;30: 233–237.

12.Rasinpera H,Savilahti E,Enattah NS,et al.A genetic test which

can be used to diagnose adult-type hypolactasia in children.Gut 2004;53:1571–1576.

13.Kruse TA,Bolund L,Grzeschik KH,et al.The human lactase-

phlorizin hydrolase gene is located on chromosome2.FEBS Lett 1988;240:123–126.

14.Kuokkanen M,Enattah NS,Oksanen A,et al.Transcriptional

regulation of the lactase-phlorizin hydrolase gene by polymor-phisms associated with adult-type hypolactasia.Gut2003;52: 647–652.

15.Nilsson TK,Johansson CA.A novel method for diagnosis of adult

hypolactasia by genotyping of the?13910C/T polymorphism with Pyrosequencing technology.Scand J Gastroenterol2004;

39:287–290.

16.Chao CK,Sibley E.PCR-RFLP genotyping assay for a lactase

persistence polymorphism upstream of the lactase-phlorizin hy-drolase gene.Genet Test2004;8:190–193.

17.Kretchmer https://www.360docs.net/doc/1812411435.html,ctose and lactase.Sci Am1972;227:71–78.

18.Cramer DW,Xu https://www.360docs.net/doc/1812411435.html,ctase persistence,galactose metabolism,

and milk consumption as risk factors for ovarian cancer.In: Auricchio S.Semenza G,https://www.360docs.net/doc/1812411435.html,mon food intolerances2:milk in human nutrition and adult type hypolactasia.Dyn Nutr Res.

Basel,Switzerland,Karger1993:352–360.

19.Cooper GS,Busby MG,Fairchild AP.Measurement of lactose

consumption reliability and comparison of two methods.Ann Epidemiol1995;5:473–477.

20.Brummer RJ,Karibe M,Stockbrugger https://www.360docs.net/doc/1812411435.html,ctose malabsorp-

tion:optimalization of investigational methods.Scand J Gastro-enterol Suppl1993;200:65–69.

21.Metz G,Gassull MA,Leeds AR,et al.A simple method of mea-

suring breath hydrogen in carbohydrate malabsorption by end-expiratory sampling.Clin Sci Mol Med1976;50:237–240.

22.van der Klei-van Moorsel JM,Douwes AC,van Oeveren JP.New

principle for estimation of hydrogen in expired air.Eur J Pediatr 1984;141:221–224.

23.Strocchi A,Corazza G,Ellis CJ,et al.Detection of malabsorption

of low doses of carbohydrate:accuracy of various breath H2 criteria.Gastroenterology1993;105:1404–1410.

24.Oku T,Nakamura S,Ichinose M.Maximum permissive dosage of

lactose and lactitol for transitory diarrhea and utilizable capacity

for lactose in Japanese female adults.J Nutr Sci Vitaminol 2005;51:51–57.

25.Chinda D,Nakaji S,Fukuda S,et al.The fermentation of different

dietary?bers is associated with fecal clostridia levels in men.

J Nutr2004;134:1881–1886.

https://www.360docs.net/doc/1812411435.html,hijima T,Takahashi M,Tatewaki K,et al.A selective medium

for isolation and presumptive identi?cation of the Bacteriodes fragilis group.Microbiol Immunol1983;27:985–993.

27.Siigur U,Tamm A,Tammur R.The faecal SCFAs and lactose

tolerance in lactose malabsorbers.Eu J Gastroenterol Hepatol 1991;3:321–324.

28.Cummings JH,Macfarlane GT.The control and consequences of

bacterial fermentation in the human colon.J Appl Bacteriol1991;

70:443–459.

29.Calloway DH,Colasito DJ,Mathews RD.Gases produced by

human intestinal?ora.Nature1966;212:1238–1239.

30.Bond JH Jr,Levitt https://www.360docs.net/doc/1812411435.html,e of pulmonary hydrogen(H2)measure-

ments to quantitate carbohydrate absorption.J Clin Invest1972;

51:1219–1225.

31.Karcher RE,Truding RM,Stawick https://www.360docs.net/doc/1812411435.html,ing a cutoff of?10ppm

for breath hydrogen testing:a review of?ve years’experience.

Ann Clin Lab Sci1999;29:1–8.

32.Bond JH,Levitt MD.Quantitative measurement of lactose ab-

sorption.Gastroenterology1976;70:1058–1062.

33.Hertzler SR,Savaiano DA.Colonic adaptation to daily lactose

feeding in lactose maldigesters reduces lactose intolerance.

Am J Clin Nutr1996;64:232–236.

34.Szilagyi A,Cohen A,Vinokuroff C,et al.Deadaption and readap-

tation with lactose,but no cross-adaptation to lactulose:a case of occult colonic bacterial adaptation.Can J Gastroenterol2004;

18:677–680.

35.Szilagyi A,Malolepszy P,Yesovitch S,et al.Inverse dose effect of

pretest dietary lactose intake on breath hydrogen results and symptoms in lactase nonpersistent subjects.Dig Dis Sci2005;

2178–2182.

Address requests for reprints to:Andrew Szilagyi,MD,SMBD Jewish General Hospital,3755Cote St Catherine Road,Rm G327,Montreal, Quebec,Canada H3T1E2.e-mail:aszilagy@gas.jgh.mcgill.ca.;fax: (514)340-8282.

We would like to thank Ian Shrier,MD,PhD,for expert assistance with the preparation of the manuscript and Mr Albert Redman for devoted administrative and secretarial service.

196SZILAGYI ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol.5,No.2