Longitudinal follow-up study: effect of psychological co-morbidity on the prognosis of inflammatory bowel disease

BACKGROUND Psychological co-morbidity is more common in patients with inflammatory bowel disease (IBD), compared with the general population, but little is known about the cumulative effect of increasing psychological burden on disease behaviour. AIMS To examine the effect of psychological co-morbidity on inflammatory bowel disease in a longitudinal follow-up study. METHODS We collected complete demographic, symptom and psychological co-morbidity data (anxiety, depression and somatisation scores) at baseline from adults with IBD in biochemical remission (faecal calprotectin <250 µg/g). Objective markers of disease activity, including glucocorticosteroid prescription or flare of disease activity, escalation of therapy, hospitalisation or intestinal resection, were reviewed ≥2 years of follow-up. We performed multivariate Cox regression, controlling for patient characteristics and follow-up duration, to examine cumulative effect of psychological co-morbidities on subsequent IBD behaviour. RESULTS Among 218 participants, 48 (22.0%) had one, 13 (6.0%) two and nine (4.1%) three psychological co-morbidities at baseline. Following multivariate Cox regression analysis, glucocorticosteroid prescription or flare, and escalation of medical therapy were significantly higher among those with two (hazard ratio [HR] = 3.18; 95% confidence interval [CI] 1.44-7.02, and HR = 2.48; 95% CI 1.03-5.93, respectively) or three (HR = 3.53; 95% CI 1.26-9.92, and HR = 8.19; 95% CI 2.88-23.23, respectively) psychological co-morbidities. Occurrence of at least one endpoint of interest was significantly higher with increasing psychological co-morbidity (HR = 1.74; 95% CI 1.07-2.82 for one, HR = 2.47; 95% CI 1.12-5.46 for two and HR = 4.93; 95% CI 1.84-13.17 for three psychological co-morbidities). CONCLUSIONS Individuals with IBD in biochemical remission experienced a worse disease course with increasing psychological co-morbidity at baseline.