PSIII-4 - Evaluating the variability of the dual sugar absorption test as a measure of intestinal permeability in pigs

Abstract: The intestinal barrier of pigs can be compromised by many factors including health challenges, off-feed events, and environmental stressors. If dysregulated barrier function is caught early, proper treatment can be initiated to potentially prevent more severe complications to overall health. However, there are few procedures to measure intestinal function in pigs that are both validated, sensitive, and minimally invasive. In this research, the dual sugar test using lactulose and mannitol was conducted on a total of 534 pigs of varying age, sex, body weight, and health status. Each pig received the same oral dose of 500 mg/kg body weight (BW) lactulose and 50 mg/kg BW mannitol, and blood was collected either 60 or 240 minutes following the dose. Plasma concentrations of lactulose and mannitol were quantified by high performance liquid chromatography using validated methods. The ratio of lactulose to mannitol absorption (L:M) served as the indicator of barrier function such that a larger L:M ratio was indicative of greater intestinal permeability. In these pigs, L:M ratios ranged from 0 to 6.96. The median L:M ratio was 0.41 while the average was 0.70, indicating that the distribution of the data was right skewed, influenced by 21.0% of pigs exhibiting a L:M of 0. Collecting blood sooner after oral gavage yielded better detectability as only 2.71% of L:M in pigs from the 60 min group was 0 compared with 43.5% from the 240 min group. Ostensibly, a L:M ratio of 0 is biologically relevant, but interpretation of such values is challenging both from statistical and biological perspectives. Intestinal barrier permeability, absorption kinetics between pigs, and delivery efficiency of the L:M solution may all affect lactulose absorption yielding a ratio of 0. In general, the data was highly variable, with a standard deviation of 0.95, so a regression analysis was performed to determine which factors may be influencing this variation. Neither a forward nor a backward elimination model determined any of the five explanatory variables (i.e., BW, timing of blood collection, health status, age, or sex) to be significant. Additional correlational analyses further confirmed that there were no significant interactions. Overall, it appears that these factors do not help to explain the variability in the L:M ratio. Further investigation is warranted to validate this process and understand the source and impact of L:M ratios equaling 0.