Limitations of indirect methods of estimating small bowel transit in man.

Experiments were carried out in healthy volunteers to explore the utility of a new [14C]lactulose breath test for measuring small intestinal transit time in man and to use this procedure to test whether two antidiarrheal agents, codeine and clonidine, alter small intestinal transit time during digestion of a liquid meal. In an initial validation study performed in 12 subjects (three studies in each subject), a liquid test meal containing 10 g [14C]lactulose was administered and the colonic entry time estimated from the time course of 14CO2 excretion in breath compared with that of H2 excretion. There was a fair correlation (r = 0.77; P less than 0.001) between results obtained by the two methods; both methods gave similar results, but 14CO2 output was delayed when compared to H2 output and was incomplete. The meal also contained xylose and [13C]glycine, permitting the duodenal entry time of the meal to be estimated by the appearance of xylose in blood and 13CO2 in breath, respectively. The same liquid meal was then used to examine the effect on small intestinal transit time (colonic entry time minus duodenal entry time) of codeine or clonidine. 99Tc-sulphur colloid was also added to the meal to permit a comparison of small intestinal transit estimated by imaging with that estimated by the 14CO2-lactulose breath test. 99Tc radioactivity appeared in the cecum (as assessed using gamma scintigraphy) about 2 hr before 14CO2 radioactivity appeared in breath; the correlation between transit time estimated by the two methods was moderate (r = 0.61; P less than 0.05). Based on the [14C]lactulose data, small intestinal transit time ranged from less than 1 to 3 hr for a liquid meal containing 10 g lactulose; within-subject variation (coefficient of variation 17%) was considerably less than between-subject variation (coefficient of variation 56%). Codeine increased the small intestinal transit time significantly (from 2.7 +/- 0.3 hr to 5.0 +/- 0.9 hr; mean +/- SE), whereas clonidine did not alter small intestinal transit time, as estimated by the colonic entry time minus duodenal entry time. Neither drug influenced duodenal entry time. These results suggest that the [14C]lactulose breath test, which has only moderate accuracy, may have occasional utility as a convenient, noninvasive method for estimating small intestinal transit time in man. However, this study also suggests that indirect methods of estimating small bowel transit in man have limitations, variability, and possibly may lack the desired sensitivity.