Prediction and symptom clustering for post-infectious functional gastrointestinal disorders

Background and Aims: Acute gastroenteritis (GE) is an important risk factor for the development of irritable bowel syndrome (IBS). We utilized observational data from the Walkerton Health Study (WHS) to: i) develop and validate a risk score for post-infectious (PI) IBS; ii) evaluate the association between childhood exposure to acute GE and IBS in adulthood and identify risk factors for PI-IBS and iii) determine gastrointestinal symptom groupings in patients exposed to acute GE, assess their stability over time as well as assess how these symptoms associate with PI-IBS.;Results: For objective 1, nine variables were identified as important predictors of IBS. The area under the ROC was 0.70 and was similar in the validation set. Percentages of patients with PI-IBS in the low, intermediate, and high risk groups were 10%, 35% and 60% in the derivation cohort and 17%, 36% and 62% in the validation cohort, respectively.;For objective 2, the cumulative prevalence of IBS was significantly increased among exposed subjects vs. controls 10.5% vs. 2.5% OR (95%CI): 4.6(1.6, 13.3). In adjusted analysis, both female gender and older age remained independent predictors of PI-IBS after controlling for exposure propensity.;For objective 3, Two symptom groupings were identified which together explained 87.8% of the total variance. These included symptoms groupings related to abdominal pain with diarrhea, and, constipation. Cluster analysis identified four patients groupings based on these factors. These clusters could be qualitatively described as diarrhea predominant, constipation predominant, mixed and atypical. The results of the confirmatory factor analysis showed that the symptom groupings identified in year 1 were stable at 4 and 6 years, as demonstrated by borderline RMSEA of 0.08 for both years. At 8 years, RMSEA was slightly increased (0.09).;Methods: For objective 1, model derivation and validation were based on a split sample method from a cohort of patients with exposure to acute GE. Overall model performance was assessed using area under the receiver operating curve (ROC). Classification and Regression Tree (CART) modeling was used to determine cut-off values for high, intermediate, and low risk. For objective 2, the cumulative incidence of IBS was estimated as a proportion with 95% confidence interval (CI). Risk factors for IBS were identified by weighted logistic regression using propensity score as weights. For objective 3, factor analysis was used to identify the symptoms domains. Hierarchical cluster analysis using k-means method was used to create cluster groupings of patients based on these factors. Confirmatory factor analysis was performed at 4, 6 and 8 years after the outbreak to assess stability of symptom domains over time.;Conclusion: A simple risk tool that uses demographics and symptoms of acute GE can predict which patients with acute GE are at risk of developing PI-IBS. The risk of developing IBS is increased four-fold in children exposed to acute GE. Risk factors for PI-IBS in children are similar to those identified among adults. Our study findings confirms some of the findings from earlier studies conducted in community settings as well as in patients attending gastroenterology practice that functional gastrointestinal symptoms segregate into those associated with diarrhea, and, constipation.