Controversies regarding the role of dairy products in inflammatory bowel disease

Dairy products may affect inflammatory bowel disease (IBD) patients who are either lactose intolerant or who are allergic to the proteins in these foods. The actual incidence of these conditions in IBD patients is not entirely clear. Whether either of these conditions results in benign symptomatic discomfort or can actually contribute to the relapse and/or clinical activity of IBD is also unclear. Physicians differ widely in the advice they give their patients; some dogmatically advise avoidance of dairy products when the diagnosis is made while others discount their possible role in the management of IBD. On the basis of the author’s and his group’s experience and review of the literature, a balanced and exploratory approach by patients, physicians and dieticians is advised.

Controverses sur le role des produits laitiers clans la maladie inflammatoire de l'intestin RESUME : Les prcxiuits laiticrs pourraienr agb,raver l'etat Jes patients atteints d'unc maladie inflammatoire de l'intcstin et qui ont demonrre une intolerance au lactose ou unc allergic aux proteincs laitieres.On ne connait pas l'incidcnce recite de ces troubles chcz cc type de patients ct on ne sait pas s'ils provoquent seulemcnr une sensation Jc malaise benin ou s'ils contribuent vraiment aux rechutc~ et a l'activice clinique de la maladie mflammacoire de l'incestin.Les conseils que lcs medecins dispensent a leurs patients a ce sujet sont tres contradictoircs.Certains medecins preconisent !'abandon total des produ1ts laitiers quand le diagnostic est pose tandis que d'autres nc pensent pas que ces produits puissent entraver le traitcment de la maladic inflammatmre de l'intestin.L'auteur, sc basant sur son experience ct sur ccllc de ses collaboratcurs, ainsi quc sur um: revue de la litterature, recommande aux patients, aux mcdecins e t aux dieteticiens d'adopter une approche exploratoire et mesurec.matory bowel disease (IBD) often face the di lemma of consum ing expensive med ications or considering newer experimental therapies.Medical therapy of this condition b frustrating because it is noncurative and may be associated wi th unpleasant side effects.Va rious forms of alternative medicine (ie, na turopachy, homeopathy, acupuncture) as well as many popu lar d iet a nd self-help medical books often make fantastic claims of positive results in the treatment of IBD.Who is the patient to bel ieve?W ha t is the patient to do?In view of these considerations, we fe lt that the time was right for a n object ive review of nu tri tional approaches to IBD.The newly formed and still evolving IBO Nutri tion Review Forum hopes to examine some of th e claims made in an attempt to sort out fact from fantasy, what is worth trying although not yet proven scientifically, and what is down right misleading and based on erroneous reason ing.This is the firsr of what we hope will become a series of reviews and position papers on re levant nutrition-re lated topics.We invite submissions and inquiries from interested physicians, d ieticians, scie n tists and patients.A lthough th is manuscript has been presented and circulated to 180 Nutrition Review Forum members, the conclusions arc t hose of the author.ULCERATIVE COLITIS Andresen ( l ) in 1925 was lhe fi rst to suggest that ulcera live colit is (UC) may be due lo food allergy.He later reported (2) chal in 66% of his patients, one or mo re specific dietary items appeared to be involved in the e tio logy of the disease.Whi le cow's milk was the mosl impo n a nt (55%), olher o ffenders included wheat (12%), tomalOcs ( 10%), o ra nges (8%), potatoes (8%) and eggs (6%).S imilar conclus ions were reached by Mackie in 1938 (3) and by Rowe in 1942 (4).Truelove (5), in 196 1, presented five patiems with UC who had experienced definite improvement when milk was removed from thei r diets.Reimrod uctio n of dairy products was associated with a reactiva• tion of the disease within two to 4 2 days.Biopsy eviden ce documeming reactivation of disease aclivi ty was o btained in a patient previously in c linical and h1stological remission who began consuming a 'h e lping o f uncooked ch eese' each day for approximate ly 10 days.Wrighl and Truelove (6) estimated chat a mi lk-free diet was beneficial in approximately 20% of pa• tients with active ' uncomplicated' UC who were treated with a combinatio n o f o ral and rec ta lly adminislered steroids fo r lwo mo nths.Twenty-four pa• tients consumed a no rmal 'dummy' diel and 26 a milk-free diet fo r o n e year.Patients o n the milk-free diet had fewer relapses th an chose o n the dummy diet.The milk-free diet appeared to be more effective in first attacks th an in chronic disease.In a ll o f the studies c ited above, there was no apparent corre lation between c irculating antibodies to cow's milk and the clinical o bservations.
In a search for alternative h ypoth eses to expla in the apparent into lerance and/or a llergy to cow's milk in UC, the poss ible role of lactose intolerance sec• ondary co a deficiency of the d,saccharidase, lactase, was encenai ned by Binder e t a l (7) who identified lactose into lerance in 49% of 39 patients usin g a 100 g lactose ch alle nge and measurement o f blood glucose levels (lactose tolerance cesl).A lactose-free d iet benefitced 47% of UC patients including fou r who h ad a previo us history of milk intolerance.The flaw in the study 206 rela tes lo the fact chat because beta• galactosidase ( ie, lactaid) was not yet available commercially to break down the lactose in milk, a 'lactose-free diet' was in fact a milk -free <liet.
The tenns 'milk a llergy' o r 'h ypersensitivity' are reserved fo r chose reactio ns sh own to be medi ated by chc immune system.ln tolernnce to dairy produces sh o uld be used co <lescribe no nimmuno log,cal adverse reacu ons, ie, laclose and/or far inlo lerance (8).Allergic reactions co milk proce,m (or rarely lo contaminant antibiotics) may be immediate (within 2 h after ingestion) inte rmedi ate (between 2 a nd 24 h ) o r delaye<l (longer than 24 h) (9).lmmunoglobulin (lg) E-mediace<l milk a llergies usually manifcsl within 30 mins bul may occur within a few <lays.The presen ce of lgE antibodies against milk protein can be <locume nred by prick skin tests as well as in vitro tests such as the rad ioallergosorbent test (RAST) and enzyme-linkc<l immunoadsorbenc assay (ELISA).C linical features supporting the diagnosis of milk a lle rgy inclu<le respiratory (rhinorrhea, wheeze, strido r an d cough) cutan eous (urticaria, angioedema, eczem a) and gastrointestinal symptoms ( vomiting, c ramping, distension and diarrhea).Most patients with cow's milk a llergy are skin-test n egative to milk ex tract and <lo no t sh ow serological evidence of lgE hypersen silivity to cow's milk.C hildren with this condilion can <levelop three fo rms of IBD-like syn• dro m es: milk-induced colitis, milkinduced eosinophilic enteropathy an<l milk-in<luce<l benign procutis ( 10).A child is decla red to be n on-cow's milk a llerg ic if 300 mL of cow's milk <la ily is tolerated for four weeks.
Further evidence chal lactose intole rance <lid n ol accoun t for ' milk a llergy' came from the observations of Gudmand-1 loyer and Jamum ( 11 ) who documented a 24% ben eficial response to a milk-free <liel in 2 1 patients with UC, n one of whom h ad lactose intolerance (accord ing to the I 00 g lactose toleran ce test ing used by Binder).The bonom line of 45 years of clinical observatio n (1925-70) was lhat at leasl 20% o f patiems with UC' benefit from a milk-free diel irrespective of whether they arc 'milk a llergic' o r lactose intolerance.TI1e resuln, of these studies arc summarized in Table I.Newcomer and McGill ( 12) who summarized the 'state of lhe art' in 1967 nmed that fou r of their 24 palienls ( 17%) with UC were milk imolcranr while only one of the fou r were lactose mwleranr.TI1ey con• eluded that "fmm a practical s tandpoint, withdrawal of milk from the <l1cb of patients who had acuve ulcerauve colilis see ms juslifie<l regardless of the history.Later, when symptoms have subs,<le<l, tolerance to lactose can be determine<l o r milk can be added lo the diet with careful observation for any change in gastroimescinal sympwms."The aut hors a lso documented that there was no significam <lccrease 111 the mucosa( concentranom of lactase, su• erase a nd malrnse activilies m UC compared with healthy con trols.The consensus with respect co laccase activity in UC is thal a tempornry re<luct ion may occur du ring an acwal attack but no pcnnanem deficiency develops (1 3,1 4 ).By studyi ng a population with an inherently low incidence of lactose intolerance, Busk et a l ( 15) docu-mente<l thal there was no evidence that active UC was associate<l with a higher inci<lence o f lactose intolerance.Base<l o n diminished lactase enzyme activily in jejuna!biopsies and a flat blood glucose curve after a I 00 g lactose challenge, laclose intoleran ce was <locumenre<l in only 9 .2% of 120 pa• licn ts with UC'.This prevalence was 111 keeping wilh that previously recor<led in t he normal Danish population.
In spite of multiple i.tudies, there 1s no eviden ce, to dace, of ,111 increased inc idence of lacwse inlolcrance in pa• tiencs with UC when ethnic h eritage 1s taken imo account.Di Palma and Narvaez (16) analyzed lhc ir hy<lrogen breach resting <lam after a lactose challenge of SO g 111 terms of the ethnic h e ritage of their subjects.On the hasis of elhnic herilage, subjects were as-signe<l to three pre<licteJ prevalence categories: high prevalen ce (greater than 90%) -ie, Orientals, Nal ivc Americans; moderate prevalence (60 to 70%) -ie, A merican Blacks, Arabs, Jews, I lispanics and South Europeans ( Ita lia ns an<l Greeks); and low preva-   (18) reported on 20 patients in whom remission had been induced with total parenteral nutrition (n=l3 ) or an elemental diet (n=7).Patients were then randomized to either an unrefined carbohydrate fibre-rich diet or a diet that excluded specific foods to which a patient was intolerant.Seven of 10 patients on the exclusion diet remained in remission for six months compared with none of 10 on the other diet.
In an uncontrolled study (19), an exclusion diet allowed 51 of 77 patients to remain well on diet alone for up to 51 months, with an average annual relapse rate of less than 10%.These authors introduced foods one a day in the order that allows a nutritionally adequate diet to be built up most rapidly.Three portions of a food were eaten on its test day, and if no symptoms were no ted, it was subsequently eaten ad libitum.If it appeared to provoke symptoms, it was avoided and only 'safe' foods were eaten until the patient was symptom-free again, when further testing was resumed.Patients were instructed to take only elemental diet and spring water for three to four days before returning cautiously to their personal diet if they developed any symptoms.
Sixty-four of the 77 patients completed the process of 'food testing' to find a diet on which they remained well fo r at least three months with no other treatment.The foods most commonly associated with intolerance were wheat (28 patients, 44%) and dairy products (24 patients, 37.5%).The estimated incidence of lactose intolerance in this population is approximately 15% (personal communication).
Subjective improvement was accompanied by normalization of the erythrocyte sedimentation rate and serum orosomucoid levels and by radiological improvement in those willing to undergo a follow-up x-ray.Jones, in a follow-up report (19), concluded that uncontrolled clinical experience with 77 patients showed that 'personalized food exclusion d iets' were associated with an average annual relapse rate of only 11 % for the first five years of therapy with diet alone.She suggests that the rate of relapse observed by the European Cooperative Crohn's Disease Study (ECCDS)  (24) can be used as a control.In the European study, 90% of patients, allocated to placebo once in remission, had relapsed at the end of follow-up at two years, while approximately 60% of patients treated with steroids or steroids plus salazopyrine had suffered a relapse.Jones concluded that the avoidance of specific foods appears co be superior to the medical management of CD and equals the results following successful surgery, the average annual relapse rate for both being approximately 10%.
Having compared the placebo management offered by both the Cambridge group ( 18) and the ECCDS, l question the validiry of the comparisons made.In the ECCDS, follow-up occurred at three-month intervals.At each visit, results of a brief medical history, physical examination and laboratory assays of blood, serum and urine were obtained.The CD activity index was also determined.In the Cambridge study, patients were seen by a physician every month and by a dietician as often as was thought necessary to give them adequate 'guidance and encouragement' in keeping to their diets.l have often wondered about the role of frequent 'guidance and encouragement' on the clinical course of !BD.Until this matter is properly studied, l deem the placebo group chosen by Jones as inappropriate.
Riordan and Hunter (20), regarding patients with CD who had achieved remission on an elemental diet, claim that 21 of 40 patients (52.5%) on diet alone for 12 months were in remission compared with 10 of 38 patients (263%) who were treated with decreasing doses of prednisone during a 12-wcek period.It is noted that 43 of L36 patients (31 %) initially entering the trial abandoned the initia l treatment with the elemental diet.Of the remaining 93 patients, 78 (84%) subsequently achieved remission within 14 days and were then randomized to either corticosteroids or diet (Figure l).The authors conclude that an "elemental diet followed by identification of food intolerance presents an effective strategy for long term management of acute CD".I would Uke to point out that in the September 1992 issue of The Journal (Canadian Foundation far lletis and Colitis), one page was devoted to this rather complex study.No comments or figures were appended to discuss or interpret the results.(The work has recently been published again [25]).
No other centre has yet succeeded in reproducing the findings claimed by the Cambridge group.
In 1991, Giaffer et al (26) reported somewhat different results in 27 patients with CD who attained clinical remission after four weeks of enteral feeding.Five of the 14 patients who completed testing for specific food intolerances could not identify any trigger foods; the remaining nine were maintained on exclusion diets, three of whom relapsed early.In this study, over 30% of patients tested for food intolerance did not identify specific trigger foods; this contrasts the claim by the Cambridge group that over 90% of CD patients could identify specific foods to which they were intolerant (18).Four of these nine patients agreed to undertake double-blind re-challenge tests, all of which were negative.The authors commented that the negative results of the blind challenges may indicate either that food aversion is responsible or that food-related reactions are delayed for over 24 h.In spite of this, six of nine patients (66%) on exclusion diets and two of 11 ( 18%) on a normal diet remained in remission.The authors concluded that while not statistically significant, there was a trend in favour of exclusion diet.The authors also observed that disease location was the single most important determinant of the subsequent course after treatment with elemental diet-Of the patients with large bowel involvement, 80% relapsed early.In Levi's review (27) of diet in CD management, he expresses his scepticism regarding the work of Hunter and co-workers."Aided by the media ... these studies have stimulated enormous public interest, so that now gastroenterologists are constantly asked by their patients whether they should go onto an elimination diet and whether a spe- Dairy products in IBD cific food intolerance is the cause of their disease ... " Levi notes, and concludes that "ultimately double blind challenges as well as controlled trials are required.At present, the place of exclusion diets in the management of CD is not known." Ginsberg and Albert (21) report on a patient whose longstanding steroiddependent CD went into remission after 10 weeks of consuming only Ensure Plus (Ross Laboratories) and tap water.During this time the patient became completely asymptomatic and his prednisone was tapered over an eight-week period and then discontinued for the first time in three years.Following this, his diet was gradually liberalized to include 'safe' foods (no adverse effect consuming this food along with Ensure Plus three times a day for three days; if symptoms developed the patient was instructed to eat only Ensure Plus and the previously determined 'safe' foods until symptoms subsided).Lactaidtreated milk gave a violent reaction after six glasses on the first day.A relapse of Crohn's-related symptoms lasting one week ensued and gradually disappeared one week later without the use of prednisone.Following remission for one year, a repeat small bowel series showed marked improvement.A 25 g lactose challenge {equivalent to the lactose contained in two glasses of milk) produced no symptoms while a double-blind milk challenge using 5 mL of whole milk resulted in a recurrence of severe cramping and diarrhea.This attack was aborted by taking 20 mg of prednisone for three days."Eighteen months after total clinical and laboratory remission had been induced by Ensure Plus, the patient continued on a strict milk-free diet and remained well without requiring any medication, and with all laboratory studies in the normal range.At that point, the patient voluntarily underwent a rechallenge with whole milk, lactose and lactoglobulin, all of which now failed to elicit any symptoms.He has since returned to his strict milk-free diet and remains in total remission."Clearly in this case of CD, one can conclude that a hypersensitivity to dairy products, but not to lactose, was operative in the ac-MIS11KIN tivation of the disease process.The authors make the interesting comment that "although we have been able to induce remission in 40% of our steroiddependent CD patients with Ensure Plus (28), in no other patient have we yet been able to isolate a specific food that consistently reproduced symptoms".I was disappointed, when I reviewed the abstract cited, to learn that these conclusions were based on experience with only seven CD patients.The question remains whether the incidence of sensitivity to dairy products really exists in 37.5% of patients with CD -as suggested by the work of Jones and Hunter -or is a rare occurrence as described above ma single patient The discussion of exclusion diets in this paper is primarily related to the withdrawal of dairy products and is not intended to deal with the issue of exclusion diets in general.Russell (29), in a recent review, comes to the following conclusion: "there is a suspicion that some specific improvement in activity of CD can be achieved by dietary and nutritional manipulation.However, there is a cogent need for well-planned prospective studies in much larger numbers of patients for longer periods of time, in which good scientific methods of assessing improvement in disease activity are used in conjunction with good nutritional data .... "

LACTOSE INTOLERANCE
le has been said that the incidence of lactose intolerance or malabsorpcion seems to be increased in adult patients with CD, especially in patients who have undergone intestinal resection (30).ln CD patients without resection, 33%compared with 16% of normal controls -met the criteria of lactose intolerance during hydrogen breath testing after a lactose challenge of 12.5 g (equivalent to the lactose content in one cup of mi lk).1n patients who underwent intestinal resection, the incidence of lactose intolerance was 58%.ln the entire group of CD patients studied (all Caucasian, non-Jewish, originating from northern and central Italy), 48% were lactose malabsorbers after a lactose challenge of 12.5 g, while only 8% experienced immcdi-ate symptoms of intolerance after the ingestion of one cup of milk (250 mL containing 12.4 g of lactose).Lactose loads used ranged between 12.5 and 100 g.Clearly 100 g, which corresponds to the lactose contents of 2 L of milk, is unrealistic in terms of normal intake, but 12.5 g, which corresponds to one glass of milk, is less than taken in normally and picks up too few patients with lactose intolerance.Less comprehensive analyses were carried out after lactose challenges of 25 and 50 g.The effect of dose on the apparent incidence of lactose intolerance was studied by Pironi ct al (30).In 67 healthy Italian subjects after the ingestion of 12.5, 25 and 50 g lactose, the cumulative percentages of malabsorbers rose from 16 to 31 to 65%.It 1s the bias of the author that a 25 g challenge representing the Lactose content of two cups of milk (500 mL) is realistic.In addition, we will advise patients to use lactaid only if symptoms were experienced during or immediately after the test.
ln some centres lactose intolerance testing is carried out by measuring the change in plasma glucose concentration following an oral challenge of lactose.To achieve a consistent separation of absorbers and nonabsorbers via changes in plasma glucose (at least 20 rng/mL), relatively large doses (ie, 50 g) must be used.Breath hydrogen testing, a noninvasive method that can use more physiological doses, has displaced blood testing for glucose in most labs.Hydrogen, which is not normally manufactured by humans, is liberated during the colonic fermentation of unabsorbed carbohydrate.False-negative results may occur in a minoriry of subjects who have colonic flora that does not produce appreciable amounts of hydrogen during fermentation (31 ).For the same reason, hydrogen breath testing should not be undertaken within two weeks of taking antibiotics.
False-positive results can be prevented by avoiding the intake of complex carbohydrates in various beans and vegetables 24 h before testing.It is claimed that smoking, sleeping or eating shortly before or during the test can give a false-positive test (32).Bacterial overgrowth, in which case colonic flora is found in the small bowel, will elevate fasting breath hydrogen.
In a study of children (given 12 g lactose) and adolescents (given 25 g lactose) by Kirschner et al (33) a higher than expected incidence of lactose intolerance was noted in 'Caucasian gentiles' with CD but not in a group with UC matched for ethnic heritage.This difference in lactose intolerance was not found when Caucasian Jews with CD were compared with a group matched for ethnic heritage with UC, with the exception of patients with diffuse small bowel disease.The location of intestinal involvement with CD and the severity of clinical symptoms did not affect the incidence of lactose malabsorption.ln a recent pr~-pcctive survey of 222 adult patients with IBD (118 with UC and 104 with CD) (34), a history of sensitivity to dairy products (a subjective assessment, described by the patient, of symptoms associated with disease activity either during remission or relapse after consuming dairy products with or without lactaid fbeta-galactosidase]) was obtained in 14.4 and 11.5% of all patients with UC and CD, respectively.The highest incidence of dairy sensitivity, 16.9%, was recorded in the group of patients with UC who had a moderate risk (60 to 70%) according to their ethnic heritage of being lactose intolerant.The incidence of dairy sensitivity in UC patients with a low risk of lactose intolerance (10 to 15%) was 12.9%.The corresponding incidences of dairy sensitivity in CD patien~ at moderate and low risk for lactose intolerance were 12.9 and 9.1 %, respectively.Thus, there appears co be an excess of dairy sensitivity in lactose intolerant patients.This may be due to the difficulty in differentiating between the symptoms resulting from dairy sensitivity or lactose intolerance.The incidence of lactose intolerance did not appear to be greater in dairy sensitivity patients compared with the entire lBD population studied.The corresponding figures for lactose intolerance in patients with UC and CD were 16.0 and 10.6%, respectively, after a lactose challenge of 25 g (equivalent to the lactose content of two cups of milk).In this study, a history of drug allergy was not h elpful in identifying patiems who were dairy sensitive.
The amount of lactose administered fo r hydrogen breath testi ng is arbitrary, ranging from 12 to 50 g (equivalent to the lactose content in one to four cups of milk).The apparent incidence of lactose intolerance, as defined by a rise in hreath hydrogen of at least 200 ppm, will increase as the dose of lactose is increased.In addition, lactose intolerance patients will ex hibit a wide range of sensitivity to a given amount of lactose as indicated hy the recorded rise in breath hydrogen, as well as the severity of symptoms expe rienced.l recommend that a 25 g lactose chall enge be used as the standard test dose.I have been impressed by the symptomatic improvement in lactose intolerance patients and high degree of compliance with the appropriate diet and enzyme replacement (beta-galactosidase). Lactose intolerance patients who are restricting their intake of dairy products shou ld receive appropriate calcium supplementation.In a study of 65 patients with lactose anJ other into lerances who completed a questionnaire (35), 75% categorized their lactose intolerance as a 'major problem'.ln excess of 60% found that their symptoms improved by more than 50% with appropriate measures, which were adhered co in mo re than 90% of cases.N ine ty-chree per cent of those surveyed felt that breath testing and dietary gu idance had been worthwhile.
The figures for dairy sensitivity in CD patients in this study are similar to the 8% observed in lcalian patients with CD by Pironi et al (30), but are significantly lower than the 37.5% incidence quoted by Jones et al (18).The incidences of dairy sensitivity in 14.4 to 16.0% of patients with UC arc reminiscen t of the estimates of approximately 20% made by Wright and Truelove (6).The incidence of lactose intolerance in patients expected to have a moderate incidence (60 to 70%) based on ethnic he ritage was not different for patients with CD or UC (66.6 and 61.0%, respectively).In contrast , in CD patients with a low predicted incidence, ic, ' low risk ethnic heritage' ( LO to 15%), the docu-

CONCLUSIONS
The incidence of dairy sensitivity in IBD palients is probably in the range of 10 to 20%.The incidence of lactose into lerance is no greater than expected by ethnic heritage except for a possible inc reased incidence in a subgroup of ' low risk' o r 'Caucasian gentile' patients with CD, most of whom are females who have undergone surgery for their disease.Additional studies arc needed to validate the above-mentioned conclusions.Reactions to dairy produces may occur as a result of lactose intolerance or a reaction to the proteins and other components (dairy sensitivity).It is also important to realize chat these reactions, especially those unrelated to lactose intolerance, may occur after many weeks of repeated consumption rather than immediately afler intake.To dace, skin testing or measurement of circulating an tibodies to cow's milk have not been helpful in the prediction of dairy sensitivity.With respect to UC in the active phase, it appears thal at least 20% of patients will benefit from a milk-free diet.There are no published data on the response of UC in remission .CD, on the ocher hand, appears to respond in hoth the active and quiescent phases.Elemental diets that are also milk-free will uniformly benefit these patients in rhe acute phase.O nce in remission, apparently 30% or more of CD patients manage to stay in remission with avoidance of da iry produces and ocher 'offending' foodstuffs.The scientific basis for these observations (Table 3) is lacking and except for one study (26), b linded food challenges have not heen carried out.We (the author and the IBD Nutrition Review Forum) recommend chat this informatio n be taken into account in the d ietary management of patients with IBD, especially when they arc not responding to conventional therapy.We urge physicians and dieticians to offer their patients adequate information and guidelines to enable them co determine whether dairy products are responsible for some of their gastrointestinal symptoms and whether further investigation is warranted.It is my bias that lactose breath testing should be offered to most patients with IBD.We feel tha t the information provided by this noninvasive test is important in the nutritional management of IBD patients.Elimination o r exclusion diets involving milk withdrawal where indicated are best worked out by specially trained dietic ians working with the treating physic ians.Dairy products constitute an important source of calcium and other nutrients and their intake should not be discontinued arbicrarily.lf there is a valid reason to curtail the intake of chiry products, provisions should be CAN J GASTROENTEROL VOL 8 No 3 MAY/JUNE 1994 made to obtain calcium and other nutrients from alternative sources.
We advise that lactose breath testing to confirm lactose intolerance be carried out before beta-galactosidase preparations are used to reduce the lactose content of dairy products.We also recommend that organizations that publish and circulate the results of nutritional studies to their members should ensure that this material has been carefully screened, properly 'digested' and evaluated for their readers who are eagerly searching for new breakthroughs in the management of their IBO.

Pfs Criteria for 'food allergy' %OS %LI Response Time frame Specific Rx for UC 50
'Dx Diagnostic criteria for ulcerative col/tis (UC): 1 By cl/ntcal criteria; 2 By cltnlcal and biopsy criteria:1No correlation with skin testing;' No correlation with circulating antibodies to cow's milk; 5 26 on milk-free diet and 24 on 'normal' diet.DS Dairy sensitivity -a subjective assessment, described by the patient, of symptoms associated with disease activity either during remission or relapse after consuming dairy products with or without lacta/d (beta ga/actos/dase): LI Lactose Intolerance: Pfs Number of patients studied

TABLE 2 Response of Crohn's disease to mllk-tree diets
(22,23)e rate of 11%/yeor: 'Not confirmed by blind challenge.DS Dairy sens/Nv/ty-o subjective assessment, described by the patient, of symptoms associated with disease activity either during remission or relapse ofter consuming dairy products with or without factoid (beta ga/actosldase): LI Lactose Intolerance -on objective assessment by a variety of methods (le, H2 breath testing or blood sugar measurements after oral Intake of12.5to 100 g of lactose).U Is caused by the Incomplete breakdown of lactose Into glucose and galactose as a result of Inadequate or absent beta-galactosldase In the mucosa/ lining of the small bowelr-2 '<I'" ~ .... § ~ ~ Figure 1) Schema.ticdiagramdepicting the various patient groups and the results of the pharmaceutical and nutritional managements offered to 136consecutive patients with Crohn's disease of varying severity and duration.The raw data were obtained from reference 20 and from The Journal (Canadian Foundation for lletis and Colitis) 1992;Sept:7 lence ( 10 to 15%) -ie, northern and western Europeans and Americans of similar extraction.This scheme is only applicable to the hereditary form of lac-CD patients, none of whom was lactose intolerant, benefitted from a milk-free diet.Except for a very interesting milksensitive nonlactose intolerant patient whose steroid refractory CD went into remission with the avoidance of dairy products (21), there is no documenta-208 tion of milk allergy or sensitivity as a factor in the activity of CD.The efficacy of elemental diets, which are in fact milk-free, in the management of acute CD(22,23)is a separate issue and will not be dealt with in this review.The earliest study from the Cambridge group