HIGH STRESS REACTIVITY AND SYMPTOM FLARES IN ULCERATIVE COLITIS PATIENTS

Abstract Psychological stress has long been implicated as a cause of symptom relapse in IBD. Whether psychological stress is associated with alterations in the brain-gut-microbiome (BGM) axis and intestinal inflammation requires further investigation. Aims: We hypothesize that stress-induced changes in BGM interactions play a role in the development of symptomatic flares in patients with ulcerative colitis (UC). We aimed to test this hypothesis in a prospective study of 100 UC patients in clinical remission to identify gut microbiome, brain network connectivity, and stress-associated biomarkers of UC patients at baseline that can predict the risk of future symptomatic flares. Methods: Ninety-four subjects in clinical remission (Mayo score ≤ 1 AND Simple Colitis Clinical Activity Index (SCCAI) score < 5) underwent comprehensive phenotyping, including clinical and behavioral assessments, brain imaging, and autonomic testing for heart rate variability (HRV) and electrodermal activity (EDA) before, during, and after a 3 minute arithmetic stress test at baseline, with longitudinal follow-up every 3 months and at time of flare. Clinical flares (SCCAI >5) and subjective flares (“Do you think you are in a flare?”) were assessed. Fecal calprotectin has been collected at baseline and every 3 months and at time of flare. Results: To date, there have been 18 clinical flares (19%) and 8 symptomatic flares (8.5%) in 94 patients over a mean follow-up time of 8.1 months. Baseline measurements have identified two clusters of patients (low stress reactivity in cluster 1, high stress reactivity in cluster 2), based on trait anxiety (International Personality Item Pool [IPIP]-Neuroticism), and perceived stress (Cohen’s Perceived Stress Scale) measured at baseline. These groups differed in the frequency of clinical flares (9.3% vs. 27.4%, p=0.03) and subjective flares (3.6% vs. 11.8%, p=0.28), respectively. There were no cluster differences in heart rate variability, but there was overall greater EDA noted in the high stress reactivity group (EDA: 3.99 (SE: 0.27) vs. 5.02 (SE: 0.25); p=0.007). Conclusions: Based on our current results, the frequency of symptomatic flares in UC patients in clinical remission is predicted by measures of perceived stress and trait anxiety and by heightened sympathetic arousal as measured by electrodermal activity. We expect that gut microbiome profiles and brain imaging networks will further delineate the stress-responsive subset of UC patients with heightened flare risk and that fecal calprotectin levels collected will confirm whether the stress-responsive subset of UC patients experiences biochemical flare with increase in symptoms.