Breath analysis for the detection o f gastrointestinal disorders has emerged as a useful diagnostic tool. Certain metabolic processes are more easily quantitated by breath analysis. This is because measurable gas concentrations may be produced by chemical reactions that take place in inaccessible regions o f the gastrointestinal tract. Exhaled breath is collected by simple means and does not require too much skill (Caspary 1978).
The association between carbohydrate meals and the presence o f gases in human flatus was established over one hundred years ago, but it was not until the late 1960's (Calloway et al 1969, Levitt 1969), that a relationship was demonstrated between ingested
carbohydrates and excreted breath hydrogen. Gaseous perfusion using a non-absorbable marker (sulphur hexafluoride), showed that human hydrogen production is essentially a result o f colonic bacterial degradation o f carbohydrates, the resorption o f hydrogen by peripheral tissues being negligible (Bond and Levitt 1972). The rate o f bacterial
degradation and hence hydrogen production was shown to be directly proportional to the availability o f substrate, similar amounts o f hydrogen being liberated when the
non-absorbable sugar lactulose was administered with different concentrations o f other test sugars such as glucose. Thus quantitation o f hydrogen production was possible. It was thought that this would provide a useful indicant o f carbohydrate malabsorption,
investigating the delivery o f carbohydrates to the colon. Maintaining the assumption that the clearance o f ingested carbohydrates from the stomach is almost immediate, the time course o f the passage o f administered non-adsorbable sugars to the subsequent rise in breath hydrogen levels was tested as an indicant o f mouth to caecum transit time o f subjects investigated (Bond and Levitt 1975). This hypothesis was supported by the finding that a significant correlation existed between the intestinal transit o f lactulose and polyethylene glycol. Passage o f the latter was determined by serial aspiration. Currently the breath hydrogen test is used for the assessment o f carbohydrate malabsorption using sugars such as lactose when brush-border enzyme deficiencies are suspected, and for the assessment o f gut transit time using lactulose a compound not metabolised by humans. The latter test has many advantages, namely; It is non-invasive, simple and rapid to perform.
The conventional use o f barium meal and follow through X-ray observation for the assessment o f transit time often shows a lack o f correlation with symptoms (Kim 1968). The barium meal is a dense mixture that may not mix well with gastric contents resulting in altered transit times.
Unlike the glucose tolerance test, breath hydrogen used as an indicant o f carbohydrate malabsorption, provides quantitative information that is distinct fi-om carbohydrate metabolism (Bond and Levitt 1972). The lactulose breath hydrogen test has also been advocated as a non-invasive test o f small intestinal bacterial colonisation. It has been postulated that a rapid and unexplained rise in breath hydrogen values is suggestive o f bacterial colonisation (Davidson and Robb 1985), and may serve as a first-line indicant o f this condition. However, recently the clinical value o f the lactulose breath hydrogen test has been questioned as it is known to be subject to many confounding variables that make performance and interpretation o f the test very difficult (Ladas et al 1989). Important assumptions are made when this test is performed and unless all are rigorously accounted for results may be erroneous. This may explain the variability seen between groups using this test. The main assumptions are; that all subjects have endogenous colonic bacteria that are capable o f degrading lactulose and that no degradation occurs until the
administered bolus enters the colon. Investigators have demonstrated that in some cases significant rises in breath hydrogen levels do not occur despite administration o f normal doses o f lactulose. This suggests a deficiency o f the lactulose dehydrogenase enzyme (Douwes et al 1985). Conversely in subjects with bacterial overgrowth o f the small intestine, falsely high and rapid increases in breath hydrogen levels may be seen. In a normal individual, bacteria range in number from 1 0^-1 0 "^ colony forming units (cfu)/ml in
the proximal small intestine to 10^-10^ cfii/ml in the distal ileum (Simon and Gorbach 1984). In cases o f overgrowth these numbers are 10^-10^ and 10" cfti/ml respectively. In such cases an early rise in breath hydrogen expiration is therefore not solely due to a rapid gut transit time. The key to interpretation o f such results appears to lie in the peak shape
Foods such as beans, rice and wheat taken upto twelve hours before a breath hydrogen test is performed can lead to high fasting expired hydrogen levels (Perman et al 1984, Corazza et al 1987). The reasons for this are unclear, delayed digestion and colonic bacterial breakdown o f complex carbohydrates may be an important factor.
Oropharyngeal hygiene should also be taken into account in unexplained cases o f high fasting breath hydrogen levels. Read and colleagues (Read et al 1980), showed that oral administration o f an antiseptic mouthwash containing chlorhexidine led to a decrease in fasting breath hydrogen levels in 14/20 o f their adult patients. It is beheved that anaerobic organisms are located in the buccal crevices and either produce an early hydrogen peak because they are not washed towards the intestine when food is ingested, or more
possibly, hydrogen from this region circulates to the lungs during ingestion and is expired relatively quickly. Clinical diagnoses o f patients could also have some bearing on initial expired hydrogen levels. Patients with coeliac disease (Perman et al 1984) and cystic fibrosis (Bali et al 1983) are often shown to have high fasting hydrogen levels. This may be a result o f gut and intestinal motility changes, leading to secondary bacterial
contamination, or the occurrence o f an alkaline pH, particularly in coeliac patients, leading to enhanced bacterial catabolism o f ingested sugars (Thompson et al 1986). Despite these shortcomings the breath hydrogen test does provide a simple and rapid assessment o f mouth to caecum transit time. The procedure is non-invasive, does not involve the use o f radioisotopes as is the case with many intestinal markers and is not dependent on
radiological examination as is the more conventional test o f intestinal transit, the Barium meal and follow through. Furthermore the breath hydrogen test has allowed the indirect assessment o f gut motility, relevant in the investigation o f the effect o f new drugs (Basilico et al 1985). The orally administered sugar dose mixes well with gastric contents hence giving a more realistic transit time and the procedure allows direct quantitation o f
carbohydrate malabsorption as distinct from carbohydrate metabolism (Bond and Levitt 1972).
6.2 Materials and Method.
6.2.1 Subjects.
Forty patients (18 female and 22 male) were investigated for the estimation o f gut transit time. The patient group were aged between 0.33-17.25 years o f age, median age 2.71yrs
(interquartile range 1.38-6.44yrs) and varied in weight from 3.84-52.10kg with a median weight o f 12.50kg interquartile range (9.5 8-20.00kg). All o f these children were being investigated for a variety o f suspected gastrointestinal conditions as characterised by the symptoms o f failure to thrive and/or persistently abnormal stools.
6.2.2 Protocol
All patients were fasted for a minimum o f 5hrs before testing and where possible were advised to eliminate any fibres or starch rich food from their diet for up to 24 hours. At the start o f the test a preload end expiratory breath sample was taken from each patient and analysed immediately to establish a baseline breath hydrogen level. Each patient was then given an oral dose o f the non-absorbable disaccharide, lactulose (0.5g/kg body to a maximum o f lOg). Further breath samples were then taken from the patient every half hour for a maximum o f 4hrs or until a significant hydrogen increase from the baseline reading was seen.