Association of lactose sensitivity with inflammatory bowel disease demonstrated by analysis of genetic polymorphism, breath gases and symptoms

Transcription

1 Alimentary Pharmacology and Therapeutics Association of lactose sensitivity with inflammatory bowel disease demonstrated by analysis of genetic polymorphism, breath gases and symptoms P. Eadala*, S. B. Matthews, J. P. Waud à, J. T. Green,1 & A. K. Campbell,1 *Department of Gastroenterology, Morriston Hospital, Swansea, UK. Welston Court Science Centre, Welston Court, Milton Pembrokeshire, UK. à Department of Medical Biochemistry and Immunology, University Hospital of Wales, Cardiff and Vale University Health Board, Cardiff, UK. Department of Gastroenterology, University Hospital Llandough, Cardiff & Vale University Local Health Board, Penarth Vale of Glamorgan, UK. Welsh School of Pharmacy, Cardiff University, Cardiff, Wales, UK. Correspondence to: Dr J. T. Green, Department of Gastroenterology, University Hospital Llandough, Cardiff & Vale University Local Health Board, Penlan Road, Penarth, Vale of Glamorgan,CF64 2XX, UK. 1 Equal contribution Publication data Submitted 9 May 2011 First decision 9 June 2011 Resubmitted 14 July 2011 Accepted 14 July 2011 EV Pub Online 4 August 2011 SUMMARY Background Sensitivity to lactose has been reported in Crohn s disease, but its true role in inflammatory bowel disease (IBD) is unclear. The genetic marker CC 13910,on chromosome2, with measurement of breath hydrogen and methane, and gut and systemic symptoms, are now the most comprehensive tests for evaluating sensitivity to lactose. Aim To investigate, for the first time, the prevalence of lactose sensitivity in IBD, using the most comprehensive tests for diagnosing this condition. Methods Prevalence of CC genotype was investigated using RT-PCR in 165 patients (Crohn s disease = 70, ulcerative colitis = 95), and 30 healthy volunteers. Genotype was correlated with breath hydrogen and methane up to 6 h after 50 g of oral lactose, all symptoms being recorded for up to 48 h. Critically, Crohn s disease and ulcerative colitis patients were selected with no record of lactose sensitivity, in remission at the time of the test. Results Lactose sensitivity occurred in a much higher proportion of patients, (approximately 70%), with IBD than previously thought. Seventeen per cent had raised methane, without raised breath hydrogen; those with ulcerative colitis exhibiting most symptoms. All CC patients were lactose sensitive. There was no correlation between genetic phenotype and IBD. As substantial numbers of IBD patients were CT or TT, and were lactose sensitive, this polymorphism cannot explain full down-regulation of the lactase gene. Conclusions Our results have implications for the clinical management of IBD. The high breath methane raised the possibility of a pathogenic role for methanogenic archaebacteria (Archaea) in IBD. This needs to be investigated. Aliment Pharmacol Ther 2011; 34: doi: /j x

2 P. Eadala et al. INTRODUCTION Sensitivity to lactose occurs as a result of deficient hydrolysis of lactose in the small intestine by the brush border enzyme lactase phlorizin hydrolase (LPH). 1 3 All mammals, apart from most white Northern Europeans, and few races, lose lactase after weaning. 1, 2 This results in deficient hydrolysis of lactose into its constituent monosaccharides, galactose and glucose. As a result, lactose is metabolised in the large intestine by colonic bacteria, where anaerobic metabolism produces hydrogen, and variety of metabolites. Archaebacteria (Archaea) convert some hydrogen into methane. The gases and bacterial metabolites cause abdominal symptoms pain, bloating, flatulence, diarrhoea or constipation and urgency 3, 4 and, systemic symptoms headaches, cognitive dysfunction, fatigue, muscle and joint pains, palpitations, eczema and mouth ulcers and increased micturition. 5 7 The mechanism causing the systemic symptoms, as well as either constipation or diarrhoea, involves the production of metabolites by bacteria in the large intestine, when they are exposed to carbohydrates not fully digested and absorbed in the small intestine. 8 The mechanism responsible for loss of lactase after weaning remains unknown. However, two polymorphisms within the introns of a helicase, MCM6, upstream of the LPH gene on chromosome2, have been identified 9, which correlate with hypolactasia. 10 The polymorphism involving a T to C change at position base pairs exhibits a close association with hypolactasia. 9 The three genotypes, CC, CT and TT, respectively, correlate with the levels of lactase activity in intestinal biopsy 9, 11 samples and their lactase: sucrase ratio. Lactase phlorizin hydrolase activity may also be reduced as a result of mechanisms independent of genetic pre-disposition. These include small intestinal bacterial overgrowth, coeliac disease, Crohn s disease (CD), drug-induced enteritis or infection The direct measurement of lactase activity in biopsies from the small intestinal mucosa has been used as a reference method to assess the level of lactase in the small intestine. 15 However, this method is invasive, expensive and not readily available. Furthermore, the lactase in a biopsy only represents a tiny fraction of the total lactase in the small intestine, and is usually taken during endoscopy, from the proximal duodenum, well away from where IBD might cause loss of lactase. Thus, hydrogen breath testing, after oral lactose, has been used in clinical practice, to assess the total active lactase. However, we have shown that measurement of breath hydrogen alone, without a record of symptoms, detects <50% of people who are lactose sensitive. 7, 16 Furthermore, other mechanisms responsible for incomplete absorption of sugars in the small intestine include insufficient contact time, due to accelerated intestinal passage, and the inhibition of the sodium dependent glucose transporter (SGLUT-1) by tri- and tetra-saccharides, found in many root vegetables and foods such as soya, generating flatus. Ulcerative colitis (UC) and CD are two types of idiopathic inflammatory bowel disease (IBD), caused by a combination of genetic factors and exposure to environmental agents, including dietary or bacterial antigens Increased prevalence of lactose sensitivity may 12, 21 occur in IBD, particularly small bowel CD, but, there have been no studies in patients with IBD, where the polymorphism C C 13910, associated with hypolactasia has been correlated properly with lactose sensitivity. 22 AIM The aim was, for the first time, to investigate the prevalence of lactose sensitivity in UC and CD, using the most comprehensive tests now available for diagnosing this condition. 7 Lactose sensitivity was defined on the basis of the patient s genotype (CC, CT and TT), together with a raised breath hydrogen (>20 ppm over the nadir) and methane (>5 ppm over the nadir), and or the occurrence of gut and systemic symptoms, after a lactose challenge. A further objective was to obtain clinical data to support our hypothesis that metabolic toxins produced by gut microbes play a key role in gastrointestinal disease. MATERIALS AND METHODS Subjects This was a prospective study enrolling newly referred patients from our gastroenterology clinic, and who had not been previously assessed for lactose sensitivity. We encountered just two patients who had been diagnosed with this and were on a restricted diet these were excluded from the present study to avoid bias and potential nocebo effects. 23 Patients with potentially life threatening co-morbidity that was apparent from the initial assessment were excluded. Similarly, healthy volunteers were selected without any known gut or systemic symptoms, to avoid confusion with undiagnosed gastrointestinal illness and who had not been assessed for lactose sensitivity. Antibiotic usages or bowel preparation for colonoscopy or radiological investigations during the preceding 4 weeks were also exclusion criteria, to prevent 736 Aliment Pharmacol Ther 2011; 34:

3 Lactose sensitivity and inflammatory bowel disease effects on breath gases. Thus, patients with known IBD who attended our GI clinics were compared with a control group of healthy volunteers. All were Caucasian over 18 years of age. UC and CD were diagnosed according to clinical, endoscopic, radiological and histological criteria. Critically, only patients were selected in remission, without active disease. Remission was defined by the Harvey Bradshaw Index 24 a score of 4 or less in those with CD and by the Simple Clinical Colitis Activity Index 25 a score of three or less with ulcerative colitis. Informed consent was obtained from each participant prior to recruitment. The study was approved by the South East Wales Research Ethics Committee. Patients with known lactose sensitivity, or who had concurrent life threatening illnesses, alcohol abuse or pregnancy, were excluded. The control group was healthy volunteers, recruited from hospital staff and medical students, on no regular treatment, nonsmokers, who had no previous abdominal surgery and who had no on-going illness of any kind. In particular, they had no record of symptoms associated with gut illness, including lactose sensitivity, inflammatory bowel disease, previous gastro-intestinal surgery, coeliac disease or irritable bowel syndrome, or gut symptoms, such as abdominal pain, distension or bloatedness and or change in stool frequency or form on most days in the preceding 12 months. Participants who could not attend for the tests or who had concurrent life threatening illness, alcohol abuse or pregnancy, were also excluded. Milk allergy was excluded in all patients and healthy volunteers by a negative blood test for IgE milk proteins using Phadia 250 (Phadia Ltd, Milton Keynes, UK). All patients and healthy volunteers were white Northern Europeans. No data were available as to how many were Ashkenazi Jews, who would be expected to be lactose sensitive. 1 Genotyping Lactase genotype was determined using RT-PCR for the C T polymorphism. Forward primer sequence coded JPWLPHCTL1-CTF: CTCTGCGCTGGCAATACAG; Reverse primer sequence coded JPWLPHCTL1-CTR: AAAT GCAACCTAAGGAGGAGAGTTC; Reporter 1 sequence coded JPWLPHCTL1-CTV1 VIC; ATAAGATAATGTAG CCCCTGGC; Reporter 2 sequence coded JPWLPHCTL 1-CTM1 FAM; ATAAGATAATGTAGTCCCTGGC. Validation of the RT-PCR assay was confirmed by comparison with restriction fragment length polymorphism analysis (RFLP) of PCR amplified DNA as previously described, with 100% correlation of C T genotypes. 5, 7 Lactose challenge test and lactose sensitivity It is important to distinguish between hypolactasia, assessed either by the polymorpisn CC or through small intestinal biopsy and lactose sensitivity. The key clinical criterion for diagnosing lactose sensitivity in this study was the effect of an oral lactose challenge on breath hydrogen and methane, and the appearance of gut and systemic symptoms. Thus, participants were given 50 g lactose as previously described. 7 They were asked not to smoke for 4 h prior to the breath test. They had also not received any antibiotic treatment or had bowel preparation for GI investigation in the 4 weeks before the study. Lactose sensitivity was defined on the basis of breath analysis for hydrogen (>20 ppm over the nadir) and methane (>5 ppm over the nadir) over 6 h as previously described, 7 together with a record of both gut and systemic symptoms for up to 48 h after the lactose challenge. Transit times through the small intestine vary considerably between individuals. This is why, on the basis of our previous study, we measured breath hydrogen and methane up to 6 h after the lactose challenge. The initial rise in breath gases can give an indication of the transit time. However, unlike one report in a Greek population, 26 we did not find this to be of value in assessing whether a patient was lactose sensitive or not. Gut and systemic symptoms Gastrointesinal symptoms (total = 7; abdominal pain, distension, flatulence, burping, loose stools, diarrhoea or constipation, nausea and vomiting), and systemic symptoms (total = 14; headache, memory impairment, loss on concentration, dizziness, fatigue, muscle and joint pain, skin itching, rhinitis, asthma, increased micturition, heart palpitations, hot and cold, mouth ulcers, bad taste) were recorded by the participants 0 = no symptoms, 10 = severe symptoms up to 48 h in a symptom diary. Patients and healthy volunteers were advised on how to assess the severity of any symptoms. This was based on conventional clinical methods and our previous study. 7 Our study shows that an objective, quantitative test for cognitive dysfunction is now required in the clinical management of both IBS and IBD. Statistical analysis Statistical significance was assessed using chi-squared statistics and ANOVA, and the statistical program SPSS version 12 (Chicago, IL, USA). Aliment Pharmacol Ther 2011; 34:

4 P. Eadala et al. RESULTS Patients The total number of people recruited was 165 IBD (95 with UC, 70 with CD, median age 47 years, duration of illness 6 years and 7 years respectively), in remission and on a range of treatments (Table 1), and 30 HV (median age 34 years), all being genotyped. However, several patients and healthy volunteers were not able to take part in the lactose challenge test. Thus, the total number of people who were analysed with the lactose challenge was 110 IBD (59 with UC and 51 with CD) and 28 HV (Tables 3 and 4). A full history of smoking or nonsmoking was taken, as well as a record of any previous surgery. Ninety-six per cent of UC patients were either ex-smokers or lifelong nonsmokers. However, 14% of CD patients were currently smokers. Only one of the UC patients had a previous colectomy, and therefore only underwent genotyping and not the breath test; 24 CD patients had previous abdominal surgery. There was no obvious correlation between the breath test results and symptoms after the lactose load in patients who had previous surgery. Genetics The polymorphism CC is within an intron of a helicase, MCM6, just upstream of the lactase gene on human chromosome2. 9 Measurement of levels of lactase in duodenal biopsies has shown that people who are CC have low lactase levels. Those who are CT have an intermediate level, and those who are TT have apparently high levels. Thus, the prediction is that those who are CC, with low lactase, should be most sensitive to lactose. In contrast, those who are TT, with apparently high lactase 1, 5, 7 levels, should be insensitive to lactose. Thus, the genetic polymorphism was analysed in 165 patients (95 with UC; 70 with CD), and compared with 30 HV (Table 2). As predicted from our previous study on patients with IBS, 7 and studies reported by others, all patients who were CC were lactose sensitive. In this study, only 7% of IBD patients were CC, 35% and 58% being CT or TT respectively. In contrast, as none of the Table 1 Patient details: the baseline characteristics of patients with CD, UC & HV who were included in the study Crohn s disease (CD) Ulcerative colitis (UC) Total Ileal (small bowel) Colonic (large bowel) Ileo-colonic (small and large bowel) Total Pan colitis Left sided Proctitis Healthy volunteers (HV) Number Age range (34) Sex M:F 32:38 10:8 13:15 9:15 51:44 20:15 26:22 5:9 15:15 Duration of NA illness (year) Family history of IBD Smoking Never ex Current Medications Yes No ASA Thiopurines Steroids Rectal therapy Surgery Yes IBD, inflammatory bowel disease; NA, not applicable; 5-ASA, 5-amino salicylic acid. For UC & CD, the results are broken down as per the extent of disease. 738 Aliment Pharmacol Ther 2011; 34:

5 Lactose sensitivity and inflammatory bowel disease healthy volunteers had reported any gut or systemic symptoms after ingesting milk, none of them turned out to be CC. There were no significant differences between IBD and HV (P = 0.14) or between genotype based on Table 2 Genetic analysis Healthy volunteers (30) Genotype result CC CT TT 0 (0%) 15 (50%) 15 (50%) Inflammatory bowel 12 (7.3%) 58 (35.2%) 95 (57.5%) disease (165) Ulcerative colitis Total (95) 7 (7.4%) 35 (36.8%) 53 (55.8%) Left sided (46) Proctitis (14) Pan colitis (35) Crohn s disease Total (70) 5 (7.1%) 23 (32%) 42 (60%) Large bowel (28) Small bowel (18) Both (24) A DNA analysis for polymorphism at position on the long arm of chromosome2 was carried out as described in the methods section. Previous data 7, 16 showed that all CC were sensitive to lactose, and are homozygous for lactase non persistence. 9 the affected segment of bowel in both UC (P = 0.6). Similar results were seen in CD, based on the segment of bowel affected (P = 0.4). When the CD or UC groups were individually compared with HV, there was also no significant difference. Effect of lactose challenge on breath hydrogen and methane Of the original 165 IBD patients in remission, only 110 were available for the lactose challenge test. Breath hydrogen and methane were analysed, up to 6 h, in 59 UC patients and 51 CD patients, and compared with the 28 HV, before and after a 50 g oral lactose challenge (Table 3; Figures 1 and 2). Raised basal hydrogen or methane, before lactose, was defined as >2 SDs over the mean in the control group of HV. Raised hydrogen or methane, after lactose, was defined as >20 ppm or >5 ppm over the nadir for hydrogen or methane respectively. None of the healthy volunteers showed either raised hydrogen or methane after lactose, the range being 0 14 and 0 5 ppm respectively. Those with high basal breath hydrogen or methane all exhibited a gradual fall to levels <20 ppm or <5 ppm, respectively, during the 6 h after ingestion of lactose. However, surprisingly, a substantial number of IBD patients had raised hydrogen and or methane after lactose (Table 3; Figures 1 and 2), being 27% of UC patients compared with 49% of CD Table 3 Effect of lactose load on breath hydrogen and methane Breath analysis Ulcerative colitis (n = 59) Crohn s disease (n = 51) Healthy volunteers (n = 28) Basal (before lactose load) Hydrogen range (ppm) Methane range (ppm) % of people with high hydrogen (>2 SDs over mean) % of patients with high methane (>2 SDs over mean) Maximum within 6 h after lactose load Hydrogen range (ppm) Methane range (ppm) % of people with high hydrogen only (>20 ppm) % of people with high methane only (>5 ppm)) % of people with both high hydrogen and methane (%) Total % of people positive, either or both hydrogen and methane Patients diagnosed with IBD (n = 110) or healthy volunteers (n = 28) ingested orally 50 g lactose, after an overnight fast, and breath hydrogen and methane were measured every 30 min for 6 h as described in the Methods. Results were expressed as ppm, and the percentage of each group calculated. Aliment Pharmacol Ther 2011; 34:

6 P. Eadala et al. patients. These differences were highly significant (P < 0.001). Importantly, a substantial number (17%) of patients in both IBD groups had raised breath methane Table 4 Relationship between symptoms and genotype Genotype Any symptom gut or systemic Gut symptoms only Systemic symptoms only No symptoms reported Ulcerative colitis CC (n = 3) 67% % (n =1) CT (n = 16) 50% 19% 6% 50% TT (n = 39) 77% 18% 18% 23% All (n = 58) 69% 17% 14% 29% Crohn s disease CC (n = 4) 100% CT (n = 17) 71% 29% 6% 29% TT (n = 30) 70% 17% 2% 30% All (n = 51) 73% 20% 6% 27% Patients with UC (n = 59) or CD (n = 51) were genotyped as described in the methods, and subjected to an oral dose of 50 g lactose. Breath hydrogen and methane were recorded every 30 min for 6 h, and gastrointesinal symptoms (total = 7; abdominal pain, distension, flatulence, burping, loose stools, diarrhoea or constipation, nausea and vomiting), and systemic symptoms (total = 14; headache, memory impairment, loss on concentration, dizziness, fatigue, muscle and joint pain, skin itching, rhinitis, asthma, increased micturition, heart palpitations, hot and cold, mouth ulcers, bad taste) recorded, together with their severity on a scale of 0 10, assessed on conventional clinical criteria. Results were expressed as a % of the number of patients within each genotype group. after lactose, with no detectable rise in breath hydrogen. There was a considerable variation in the absolute value in both the basal breath gases, and the maximum after lactose. The maximum breath hydrogen ranged from 0 to 227 ppm, the maximum methane ranged from 0 to 157 ppm and the lowest maximum breath hydrogen or methane values being 22 and 7 respectively (Figures 1 and 2). Interestingly, a significant number of IBD patients had raised basal hydrogen and or methane before the lactose challenge (Table 3; Figures 1 and 2), 12% of UC patients having raised basal methane, compared with only 2% in patients CD, there being no relationship with large or small bowel distribution and all IBD patients were in remission. Furthermore, UC patients exhibiting the most symptoms after the lactose challenge had raised breath methane, but no elevated breath hydrogen. In contrast, CD patients with problems in small bowel had raised breath hydrogen without raised breath methane. Raised breath hydrogen occurred in 73%, 56% & 15% in CD affecting the small bowel, small and large bowel and large bowel, respectively, these differences being highly significant (P = 0.002), with a statistically greater number of high breath hydrogen patients with CD affecting small bowel. In contrast, the prevalence of high breath hydrogen in patients with UC affecting the whole colon, left colon and rectum was 25%, 24% & 30% respectively. A comparison of breath hydrogen and the extent of the disease showed that there was no significant difference (P = 0.93). Breath hydrogen was raised in three of four smokers with UC and five of 10 smokers with CD. This was not due to any acute effects, as 250 Basal and maximum breath H 2 before and after lactose 200 Breath hydrogen ppm UC basal CD basal HV basal UC max CD max HV max Figure 1 Breath hydrogen before and after a lactose challenge. Patients or healthy volunteers were given a 50 g oral dose of lactose, and breath hydrogen measured every 30 min for up to 6 h, as described in the Methods. Results were plotted as ppm for basal values before the lactose challenge and maximum values after the lactose challenge. 740 Aliment Pharmacol Ther 2011; 34:

7 Lactose sensitivity and inflammatory bowel disease Breath methane ppm UC basal CD basal HV basal UC max CD max HV max Figure 2 Breath methane before and after a lactose challenge. Patients or healthy volunteers were given a 50 g oral dose of lactose and breath methane measured every 30 min for up to 6 h, as described in the Methods. Results were plotted as ppm for basal values before the lactose challenge and maximum values after the lactose challenge. patients were instructed not to smoke immediately prior to the breath sample being taken. 7, 16 As previously reported, all patients of genotype CC had a raised maximum breath hydrogen, none only having a raised breath methane, compared with 23% of those CT and 25% of those TT. The CT and TT genotypes were more likely to have positive breath hydrogen in CD than UC and HV (P < ). There was no difference in the positivity of the breath hydrogen in the genotypes CT or TT when analysed by the disease extent in UC. However, in CD, both CT and TT genotypes, Table 5 Diagnosis of lactose sensitivity based on breath gases and symptoms after an oral lactose load Breath test result Positive hydrogen (>20 ppm) Positive hydrogen (>20 ppm) & methane (5 ppm) Positive hydrogen (>20 ppm) & methane (5 ppm) & positive symptoms Ulcerative colitis (n = 59) Crohn s disease (n = 51) 10% 31% 27% 49% 68% 76% Patients diagnosed with IBD (n = 110) and healthy volunteers (28) were given an oral dose of 50 g lactose, as described in the Methods. Each individual filled in a questionnaire, recording gut and systemic symptoms, scoring the severity from 1 to10. Patients with positive symptoms were then assessed as to whether they had a positive breath test. None of the health volunteers exhibited any symptoms. Results were expressed as a % of the total number of patients within each clinical group. there was a significant difference in patients with a raised breath hydrogen when the CD affected small bowel alone, compared with that involving both the small and large bowel when compared with large bowel involvement alone (P < ). These results show that both breath hydrogen and methane must be measured if a correct assessment of lactose sensitivity is to be made in IBD patients. Gut and systemic symptoms following lactose challenge A substantial number of patients with IBD exhibited both gut and systemic symptoms after the lactose challenge, being 69% for UC and 73% in CD (Table 4), <20% in either group recording gut symptoms only. Furthermore, 14% and 6% with UC or CD, respectively, recorded systemic symptoms without any apparent gut symptoms. Recording both breath hydrogen and methane, with gut and systemic symptoms, increased the detection of lactose sensitivity in UC from 33% to 76% and in CD from 10% to 68%, compared with an assessment using raised breath hydrogen alone (Table 3). As previously reported, 7, 16 the association of gut and systemic symptoms in patients with genotype CC was very high, only one patient in this study reporting no increase in symptoms after lactose (Table 5). All the CC patients exhibited elevated breath hydrogen after the lactose challenge. A significant number of UC patients with the TT genotype (14%) exhibited only systemic symptoms after lactose. This was compared with only 2% of CD patients who were TT. There were no other major correlations of symptoms with genotype. The range of gut and systemic symptoms reported by the IBD patients Aliment Pharmacol Ther 2011; 34:

8 P. Eadala et al. after the lactose challenge ranged from 0 to 20. Interestingly, UC patients who exhibited most gut symptoms, i.e. three or more, had the highest elevated breath methane, in the range ppm, after a lactose load, compared with patients with Crohn s, who showed much lower or no elevation in breath methane after lactose (P < 0.03). Diarrhoea was a common symptom in 32% of patients with ulcerative colitis and 43% with CD. Interestingly, of these, in eight of UC patients and 12 of CD patients, the diarrhoea continued until the second day after the lactose load. This was not consistent with the effect of lactose being osmotic. The timing of both gut and systemic symptoms often correlated closely. For example, several patients suffered both diarrhoea and headache for at least 48 h. There was no evidence that any symptoms, particularly those lasting 48 h or more were due to a nocebo effect. 23 None of the healthy volunteers exhibited such an effect. It was concluded that it is essential to record both gut and systemic symptoms when assessing lactose sensitivity after an oral lactose challenge. DISCUSSION The results here show that sensitivity to lactose occurs in a high proportion (approximately 70%) of patients with IBD, either ulcerative colitis or CD (Table 5). This was in spite of the fact that patients were in remission. Our results also show that genetic polymorphism analysis alone is not sufficient to identify all patients who are lactose sensitive. The patients selected for this study were fresh referrals to the gastroenterology clinic, who had not previously been assessed for lactose sensitivity. The reason for this higher incidence of lactose sensitivity in IBD patients is because of the comprehensive tests used in this study, involved genetic testing, measurement of breath hydrogen and methane for 6 h and a full record of symptoms for up to 48 h. These results, together with those of our previous study, 7 argue strongly that this should be established as the gold standard for assessing lactose sensitivity in any patient. A further improvement would be the development of a quantitative, web based, psychological test for cognitive dysfunction. This would also provide an objective method for assessing the effectiveness of any therapy or dietary regime. We are at present addressing this. The lactose sensitivity in IBD cannot be attributed to any ongoing major inflammation, as the patients were in remission. Most importantly, the results show that it is essential to measure both breath hydrogen and methane for up 6 h after an oral lactose challenge, as well as recording both gut and systemic symptoms for up to 48 h. Elevated levels of breath hydrogen have been reported in IBD, 26 but this is the first time high breath methane levels have been reported in IBD, with or without an elevated breath hydrogen. Elevated breath hydrogen and or methane in a significant number of both UC and CD patients were consistent with an on-going carbohydrate malabsorption, even when in remission. The conclusions are consistent with our previous findings in 7, 16, 33 patients who were lactose sensitive, without IBD. Furthermore, the results here confirm our previous data and those of others that all patients with the CC genotype in chromosome2 were lactose sensitive. In the current study, only 7% of IBD patients were CC, 35% and 58% being CT or TT respectively. This was compared with 14.5% CC, 39% CT and 46.5% TT in our previous study on patients with IBS who were lactose sensitive. The higher percentage of CC in the previous study is to be expected, as these patients had been specifically referred to a food intolerance clinic because of possible lactose sensitivity. Thus, genotyping is recommended as a routine, those who are CC not needing the oral lactose challenge test, thereby avoiding major symptoms, which these patients frequently suffer for several days. In fact, IBD patients often admit to 7, 16, 33 avoiding dairy products on their own assessment of their sensitivity to milk. 21, Small bowel disease would be expected to lead to carbohydrate malabsorption, but our study shows that problems in the large bowel can also lead to malabsorption of sugars such as lactose. Although there was a high incidence of IBD in patients who were TT (Table 2), our results do not support any obvious increased risk of IBD in people who are TT, compared 37, 38 with CT, as has been reported. Although there have been several studies aimed at investigating whether there is a correlation between the CC CT TT genotype and gut illnesses, 27 29, 31, and an increased prevalence of lactose sensitivity has been reported in CD, 12, 21 our study is the first to correlate, in IBD, the CC CT TT genotype with both breath hydrogen and methane, together with both gut and systemic symptoms after a lactose challenge. The confusion over the value of the so-called lactose breath test is because other studies have failed to carry out this complete assessment of lactose sensitivity. Interestingly, a significant number of both UC and CD patients showed a rise in breath methane, without a rise in breath hydrogen (Table 3, Figures 1 and 2), with some IBD patients reporting symptoms without raised breath hydrogen or 742 Aliment Pharmacol Ther 2011; 34:

9 Lactose sensitivity and inflammatory bowel disease methane. The possibility that some of these patients were exhibiting a placebo nocebo 23 effect now needs to be ruled out by studying the effect of long-term removal of lactose from their diet, which previously showed that no symptoms in such patients were due to placebo nocebo. Several studies have now reported correlation 7, 16 between CC genotype and positive H 2 -breath test, >97% having raised breath hydrogen after a lactose challenge. 7, 16, 26 32, 42 However, as in our previous study, 7, 16 substantial numbers of the other two genotypes, CT and TT were also lactose sensitive, with raised breath hydrogen and or methane. As we previously reported that lactose sensitivity was found in several families who were either CT or TT, this confirms our previous conclusions, that the MCM6 genotype cannot wholly explain hypolactasia, assessed by biopsy from the duodenum. What is often ignored is that two molecular mechanisms can be responsible for a low level of lactase a down-regulation of the lactase gene, or, a reduction in the number of villi cells expressing lactase. The patchy occurrence of lactase in the mature small intestine of humans and animals suggests that the latter developmental mechanism is most likely to be responsible for hypolactasia in adults, being quite separate from the homeobox gene mechanisms responsible for the expression of lactase just before birth. Thus, simply measuring the level of lactase in one small area of the intestine is misleading. Our study is the first to do paired analysis, using both breath testing and genotyping in IBD patients. Given the fact that only 7% of our IBD patients were CC, our results confirm previous reports 37, 48 that there is no absolute association between the CC genotype and IBD. However, as over half of the Crohn s patients in our study had been subjected to surgery, a reduced level of lactase in the small intestine could be exacerbated by removal through the surgical procedure or by bacterial overgrowth. This might also affect the production of gases by microbes in the large intestine. Given the high occurrence of lactose sensitivity shown by our study, the question now arises whether this is a cause or consequence of IBD. Bacterial overgrowth would be expected to reduce the level of lactase in the small intestine, but this cannot explain all our results. Our results have important, potential implications for the pathogenesis of IBD. An intriguing result was the high levels of breath methane before and after lactose in the IBD patients, particularly those with ulcerative colitis (Table 3; Figures 1 and 2). Hydrogen is generated by bacteria, through the enzyme formate hydrogenase, from anaerobic metabolism of sugars, such as lactose. 1, 8 Methane, produced in the intestine, has been associated with IBD, but the importance of the cells that produce it has not been previously highlighted. Methane is only produced by archaebacteria (Archaea), through the reduction of CO 2 by H 2. Archaea represent a completely different group of organisms from bacteria, having glycerol-ether based phospholipids in their membranes, being able to grow in hostile environments as a result. They can also use hydrogen to reduce sulphates. However, it is now clear that Archaea also live in non-hostile environments, including the human body, and are therefore likely to play a role in the pathogenesis of disease. 54, 55 Archaea have been found in the gut, particularly the lower colon and faeces, mouth and vagina, and have been associated with periodontitis. Two archaebacteria species, originally discovered in the human gut, are Methanobrevibacter smithii and Methanosphaera stadtmanae, and halophilic Archaea (genus Halobacteriaceae) has been found in the intestinal mucosa. 57 Therefore, our results point to the possibility that Archaea may play a role in the pathogenesis of IBD, particularly in patients with ulcerative colitis and large bowel Crohn s. This now needs to be investigated. Our bacterial metabolic toxin hypothesis, based on previous work on IBS, 1, 7, 8, 58 and in vitro studies on E. coli, together with previous reports on the aetiology of IBD, suggest that three mechanisms by which Archaea could be involved in the pathogenesis of the intestine should now be investigated: (a) activation of immune responses, producing inflammation; (b) movement of microbial cells along nerve axons; and (c) generation of peptide and metabolic toxins. In addition to hydrogen, anaerobic metabolism of sugars by bacteria generates metabolites such as alcohols, 1, 16, diols, aldehydes, ketones and acids. This is because, to generate ATP under the anaerobic conditions of the gut, bacteria have to get rid of the hydrogen equivalents if glycolysis is to continue. It has been proposed that several of these act as metabolic toxins, controlling both the balance of microflora in the gut and affecting host cells, by acting on cell signalling mechanisms and ion channels in gut bacteria, and host neurons, muscle, heart, mast cells, cells of the immune system and others, thereby explaining the systemic symptoms. 1, 7, 8, 62 As in our previous study, 16 patients who suffered from diarrhoea after the lactose challenge often started having diarrhoea several hours, even a day, after the lactose would have disappeared from the intestine. In fact, 50% of the patients with CD and over Aliment Pharmacol Ther 2011; 34:

10 P. Eadala et al. 75% of those with UC, who had the diarrhoea after the lactose challenge, still had this 48 h later, also long after the lactose would have gone from the intestine. This supports our previous hypothesis that the mechanism causing diarrhoea involves cell signalling, analogous to 1, 8, 16, 61 the diarrhoea in severe gut infections, and is not simply an osmotic effect of the lactose, as has been previously thought. Important metabolic toxins are methylglyoxal, acetoin, diacetyl, butan 2,3 diol and related aldehydes and ketones. These generate Ca 2+ signals in bacteria, which affect the generation time, and thus enable these bacteria to compete with hundreds of others that exist in the human gut. The metabolic toxins can also chemically modify host signalling molecules, such as hormones and neurotransmitters, thereby affecting their biological activity. 1, 8 The question now is whether Archaea also produce these metabolic toxins. It is important to realise that, under anaerobic conditions, these metabolites will be generated by gut bacteria or Archaea from any sugar that is not absorbed in the small intestine, not just lactose. Thus, the bacterial metabolic toxin hypothesis provides a molecular mechanism to explain symptoms, particularly systemic, in IBS, coeliac disease, food intolerance and IBD, independent of inflammatory immune mechanisms. The mechanism causing IBD has been reported to be due to a combination of genetic and environmental factors. 20, 63 Thus, our study, for the first time, points towards a potential molecular mechanism that puts these factors, and the possible role of bacterial overgrowth in the small intestine, into perspective. 1, 8 It has been reported that restriction of a range of oligosaccharides, such as fructans and GOS, polyols, fructose and lactose, known as the FODMAP diet, has a major effect in reducing symptoms in most patients suffering from functional bowel disorders. 64, 65 Our bacterial toxin hypothesis explains this reported success in treating gut symptoms. 66 In conclusion, this study shows that there is a much higher prevalence in sensitivity to lactose in all types of IBD than previously thought, even in patients in remission with no active disease. The question now arises as to what active disease will do to this sensitivity to lactose. It would be expected to make it worse. This argues strongly for a full clinical trial to investigate the effect of removing lactose from the diet in patients with active IBD, or in remission. The work reported here shows that this should involve patients being genotyped. Those who are CC should be advised to remove all lactose from their diet, including lactose hidden in foods, which contain whey, other milk products and drugs. 58, 62 Those who are CT or TT should undergo a lactose challenge test, when both breath hydrogen and methane are measured, together with a record of gut and systemic symptoms. Those who are positive should go on a lactose free diet for 3 months and then be reassessed. Although it has been reported that 50 g oral lactose, equivalent to a litre of cow s milk, can result in some lactose reaching the large intestine, 67 our results confirm the benefit of using this relatively high dose of lactose for the lactose load, as the HV showed no raised breath gases or symptoms. Whether the sensitivity to lactose is a cause or consequence of IBD should now be investigated. Finally, the clinical data reported here are consistent with a role for metabolic toxins, produced either by gut bacteria or Archaea, in the pathogenesis of IBD. This should now be investigated using PCR, as it would provide a new pathway to understanding the true molecular basis of inflammatory bowel disease. ACKNOWLEDGEMENTS We thank all patients who were involved in this evaluation. We thank the Department of Infection, Immunity & Biochemistry, Cardiff University, and the Department of Medical Genetics Cardiff and Vale University Health Board. We also thank the staff of Clinical Research Facility of Cardiff and Vale University Local Health Board, Cardiff. Declaration of personal and funding interests: None. REFERENCES 1. Campbell AK, Waud JP, Matthews SB. The molecular basis of lactose intolerance. Sci Prog 2005; 88: Sahi T. Genetics and epidemiology of adult-type hypolactasia. Scand J Gastroenterol Suppl 1994; 202: Newcomer AD, McGill DB. Clinical importance of lactase deficiency. N Engl J Med 1984; 310: Arola H. Diagnosis of hypolactasia and lactose malabsorption. Scand J Gastroenterol Suppl 1994; 202: Matthews SB, Waud JP, Roberts AG, et al. Systemic lactose intolerance: a new perspective on an old problem. Postgrad Med J 2005; 81: Grimbacher B, Peters T, Peter HH. Lactose-intolerance may induce severe 744 Aliment Pharmacol Ther 2011; 34:

11 Lactose sensitivity and inflammatory bowel disease chronic eczema. Int Arch Allergy Immunol 1997; 113: Waud JP, Matthews SB, Campbell AK. Measurement of breath hydrogen and methane, together with lactase genotype, defines the current best practice for investigation of lactose sensitivity. Ann Clin Biochem 2008; 45: Campbell AK, Matthews SB, Vassel N, et al. Bacterial metabolic toxins : a new mechanism for lactose and food intolerance, and irritable bowel syndrome. Toxicology 2010; 278: Enattah NS, Sahi T, Savilahti E, et al. Identification of a variant associated with adult-type hypolactasia. Nat Genet 2002; 30: 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: Kuokkanen M, Enattah NS, Oksanen A, et al. Transcriptional regulation of the lactase-phlorizin hydrolase gene by polymorphisms associated with adulttype hypolactasia. Gut 2003; 52: von Tirpitz C, Kohn C, Steinkamp M, et al. Lactose intolerance in active Crohn s disease: clinical value of duodenal lactase analysis. J Clin Gastroenterol 2002; 34: Corazza GR, Ginaldi L, Furia N, et al. The impact of HIV infection on lactose absorptive capacity. J Infect 1997; 35: Flatz G, Lie GH. Effect of acetylsalicylic acid on symptoms and hydrogen excretion in the disaccharide tolerance test with lactose or lactulose. Am J Clin Nutr 1982; 35: Newcomer AD, McGill DB, Thomas PJ, et al. Prospective comparison of indirect methods for detecting lactase deficiency. N Engl J Med 1975; 293: Matthews SB, Waud JP, Matthews SB. The molecular basis of lactose intolerance. Sci Prog 2009; 92: Jones VA, Dickinson RJ, Workman E, et al. Crohn s disease: maintenance of remission by diet. Lancet 1985; 2: Jowett SL, Seal CJ, Pearce MS, et al. Influence of dietary factors on the clinical course of ulcerative colitis: a prospective cohort study. Gut 2004; 53: Satsangi J. Gene discovery in IBD: a decade of progress. J Pediatr Gastroenterol Nutr 2008; 46(Suppl. 1): E Van Limbergen J, Wilson DC, Satsangi J. The genetics of Crohn s disease. Annu Rev Genomics Hum Genet 2009; 10: Mishkin B, Yalovsky M, Mishkin S. Increased prevalence of lactose malabsorption in Crohn s disease patients at low risk for lactose malabsorption based on ethnic origin. Am J Gastroenterol 1997; 92: Buning C, Ockenga J, Kruger S, et al. The C C(-13910) and G G(-22018) genotypes for adult-type hypolactasia are not associated with inflammatory bowel disease. Scand J Gastroenterol 2003; 38: Vernia P, Di Camillo M, Foglietta T, et al. Diagnosis of lactose intolerance and the nocebo effect: the role of negative expectations. Dig Liver Dis 2010; 42: Harvey RF, Bradshaw JM. A simple index of Crohn s-disease activity. Lancet 1980; 1: Walmsley RS, Ayres RC, Pounder RE, et al. A simple clinical colitis activity index. Gut 1998; 43: Ladas S, Papanikos J, Arapakis G. Lactose-malabsorption in Greek adults correlation of small bowel transit-time with the severity of lactose-intolerance. Gut 1982; 23: Babu J, Kumar S, Babu P, et al. Frequency of lactose malabsorption among healthy southern and northern Indian populations by genetic analysis and lactose hydrogen breath and tolerance tests. Am J Clin Nutr 2010; 91: Hogenauer C, Hammer HF, Mellitzer K, et al. Evaluation of a new DNA test compared with the lactose hydrogen breath test for the diagnosis of lactase non-persistence. Eur J Gastroenterol Hepatol 2005; 17: Krawczyk M, Wolska M, Schwartz S, et al. Concordance of genetic and breath tests for lactose intolerance in a tertiary referral centre. J Gastrointestin Liver Dis 2008; 17: Piepoli A, Schirru E, Mastrorilli A, et al. Genotyping of the lactase-phlorizin hydrolase C T polymorphism by means of a new rapid denaturing highperformance liquid chromatographybased assay in healthy subjects and colorectal cancer patients. J Biomol Screen 2007; 12: Schirru E, Usai-Satta P, Jores RD, et al. Lactose malabsorption: genetic testing versus lactose hydrogen breath test. Dig Liver Dis 2009; 41: Usai Satta P, Congia M, Schirru E, et al. Genetic testing is ready to change the diagnostic scenario of lactose malabsorption. Gut 2008; 57: Matthews SB, Campbell AK. When sugar is not so sweet. Lancet 2000; 355: Mishkin S. Dairy sensitivity, lactose malabsorption, and elimination diets in inflammatory bowel disease. Am J Clin Nutr 1997; 65: Banos Madrid R, Salama Benerroch H, Moran Sanchez S, et al. Lactose malabsorption in patients with inflammatory bowel disease without activity: would it be necessary to exclude lactose products in the diet of all patients? An Med Interna 2004; 21: Bernstein CN, Ament M, Artinian L, et al. Milk tolerance in adults with ulcerative colitis. Am J Gastroenterol 1994; 89: Juste RA. Crohn s disease and ruminant farming. Got lactase? Med Hypotheses 2010; 75: Nolan DJ, Han DY, Lam WJ, et al. Genetic adult lactase persistence is associated with risk of Crohn s disease in a New Zealand population. BMC Res Notes 2010; 3: Nagy D, Bogacsi-Szabo E, Varkonyi A, et al. Prevalence of adult-type hypolactasia as diagnosed with genetic and lactose hydrogen breath tests in Hungarians. Eur J Clin Nutr 2009; 63: Nagy D, Tomory G, Csanyi B, et al. Comparison of lactase persistence polymorphism in ancient and present-day Hungarian populations. Am J Phys Anthropol 2011; 145: Schirru E, Corona V, Usai-Satta P, et al. Genetic testing improves the diagnosis of adult type hypolactasia in the Mediterranean population of Sardinia. Eur J Clin Nutr 2007; 61: Kerber M, Oberkanins C, Kriegshauser G, et al. Hydrogen breath testing versus LCT genotyping for the diagnosis of lactose intolerance: a matter of age? Clin Chim Acta 2007; 383: Maiuri L, Raia V, Potter J, et al. Mosaic pattern of lactase expression by villous enterocytes in human adult-type hypolactasia. Gastroenterology 1991; 100: Maiuri L, Rossi M, Raia V, et al. Patchy expression of lactase protein in adult rabbit and rat intestine. Gastroenterology 1992; 103: Maiuri L, Rossi M, Raia V, et al. Surface staining on the villus of lactase protein and lactase activity in adult-type hypolactasia. Gastroenterology 1993; 105: Rings EH, de Boer PA, Moorman AF, et al. Lactase gene expression during early development of rat small intestine. Gastroenterology 1992; 103: Stokkers P, van den Berg M, Buller H, et al. Patchy and mosaic protein expression in the small intestine. J Pediatr Gastroenterol Nutr 1994; 19: Nanji AA, Denardi FG. Primary adult lactose-intolerance protects against development of inflammatory bowel disease. Med Hypotheses 1986; 19: Hwang L, Low K, Khoshini R, et al. Evaluating breath methane as a diagnostic test for constipation-predominant IBS. Dig Dis Sci 2010; 55: Aliment Pharmacol Ther 2011; 34:

12 P. Eadala et al. 50. Kunkel D, Basseri RJ, Makhani MD, et al. Methane on breath testing is associated with constipation: a systematic review and meta-analysis. Dig Dis Sci 2011; 56: Pimentel M, Lezcano S. Irritable bowel syndrome: bacterial overgrowth what s known and what to do. Curr Treat Options Gastroenterol 2007; 10: Pimentel M, Mayer AG, Park S, et al. Methane production during lactulose breath test is associated with gastrointestinal disease presentation. Dig Dis Sci 2003; 48: Roccarina D, Lauritano EC, Gabrielli M, et al. The role of methane in intestinal diseases. Am J Gastroenterol 2010; 105: Conway de Macario E, Macario AJL. Methanogenic Archaea in health and disease: a novel paradigm of microbial pathogenesis. Int J Med Microbiol 2009; 299: Eckburg PB, Lepp PW, Relman DA. Archaea and their potential role in human disease. Infect Immun 2003; 71: Li CL, Liu DL, Jiang YT, et al. Prevalence and molecular diversity of Archaea in subgingival pockets of periodontitis patients. Oral Microbiol Immunol 2009; 24: Oxley APA, Lanfranconi MP, Wurdemann D, et al. Halophilic archaea in the human intestinal mucosa. Environ Microbiol 2010; 12: Campbell AK, Matthews SB. Tony s Lactose Free Cookbook: The Science of Lactose Intolerance and how to Live Without Lactose. Pembrokeshire: The Welston Press, Campbell AK, Naseem R, Wann K, et al. Fermentation product butane 2,3-diol induces Ca 2+ transients in E. coli through activation of lanthanum-sensitive Ca 2+ channels. Cell Calcium 2007; 41: Campbell AK, Naseern R, Holland IB, et al. Methylglyoxal and other carbohydrate metabolites induce lanthanum-sensitive Ca 2+ transients and inhibit growth in E-coli. Arch Biochem Biophys 2007; 468: Naseem R, Wann KT, Holland IB, et al. ATP regulates calcium efflux and growth in E. coli. J Mol Biol 2009; 391: Eadala P, Waud JP, Matthews SB, et al. Quantifying the hidden lactose in drugs used for the treatment of gastrointestinal conditions. Aliment Pharmacol Ther 2009; 29: Lees CW, Barrett JC, Parkes M, et al. New IBD genetics: common pathways with other diseases. Gut 2011 [Epub ahead of print]. 64. Gibson PR, Shepherd SJ. Personal view: food for thought western lifestyle and susceptibility to Crohn s disease. The FODMAP hypothesis. Aliment Pharmacol Ther 2005; 21: Halmos EP, Muir JG, Barrett JS, et al. Diarrhoea during enteral nutrition is predicted by the poorly absorbed shortchain carbohydrate (FODMAP) content of the formula. Aliment Pharmacol Ther 2010; 32: Gibson PR, Shepherd SJ. Evidence-based dietary management of functional gastrointestinal symptoms: the FODMAP approach. J Gastroenterol Hepatol 2010; 25: Bond JH, Levitt MD. Quantitative measurement of lactose absorption. Gastroenterology 1976; 70: Aliment Pharmacol Ther 2011; 34: