There is increasing evidence for an influence of the gut-brain axis on the natural history of inflammatory bowel disease (IBD). Psychological therapies could, therefore, have beneficial effects in individuals with IBD, but data are conflicting. We aimed to update our previous systematic review and meta-analysis to assess whether the inclusion of more randomised controlled trials (RCTs) showed any beneficial effects and whether these effects varied by treatment modality. In this systematic review and meta-analysis, we searched MEDLINE, Embase, Embase Classic, PsychINFO, and the Cochrane Central Register of Controlled Trials from Jan 1, 2016, to April 30, 2023, for RCTs published in any language recruiting individuals aged 16 years or older with IBD that compared psychological therapy with a control intervention or treatment as usual. We pooled dichotomous data to obtain relative risks (RR) with 95% CIs of inducing remission in people with active disease or of relapse in people with quiescent disease at final follow-up. We pooled continuous data to estimate standardised mean differences (SMD) with 95% CIs in disease activity indices, anxiety scores, depression scores, stress scores, and quality-of-life scores at completion of therapy and at final follow-up. We pooled all data using a random-effects model. Trials were analysed separately according to whether they recruited people with clinically active IBD or predominantly individuals whose disease was quiescent. We conducted subgroup analyses by mode of therapy and according to whether trials recruited selected groups of people with IBD. We used the Cochrane risk of bias tool to assess bias at the study level and assessed funnel plots using the Egger test. We assessed heterogeneity using the I2 statistic. The updated literature search identified a total of 469 new records, 11 of which met eligibility criteria. 14 studies were included from our previous meta-analysis published in 2017. In total, 25 RCTs were eligible for this meta-analysis, all of which were at high risk of bias. Only four RCTs recruited patients with active IBD; there were insufficient data for meta-analysis of remission, disease activity indices, depression scores, and stress scores. In patients with active IBD, psychological therapy had no benefit compared with control for anxiety scores at completion of therapy (two RCTs; 79 people; SMD -1·04, 95% CI -2·46 to 0·39), but did have significant benefit for quality-of-life scores at completion of therapy (four RCTs; 309 people; 0·68, 0·09 to 1·26), although heterogeneity between studies was high (I2=82%). In individuals with quiescent IBD, RR of relapse of disease activity was not reduced with psychological therapy (ten RCTs; 861 people; RR 0·83, 95% CI 0·62 to 1·12), with moderate heterogeneity (I2=60%), and the funnel plot suggested evidence of publication bias or other small study effects (Egger test p=0·046). For people with quiescent IBD at completion of therapy, there was no difference in disease activity indices between psychological therapy and control (13 RCTs; 1015 people; SMD -0·01, 95% CI -0·13 to 0·12; I2=0%). Anxiety scores (13 RCTs; 1088 people; -0·23, -0·36 to -0·09; 18%), depression scores (15 RCTs; 1189 people; -0·26, -0·38 to -0·15; 2%), and stress scores (11 RCTs; 813 people; -0·22, -0·42 to -0·03; 47%) were significantly lower, and quality-of-life scores (16 RCTs; 1080 people; 0·31, 0·16 to 0·46; 30%) were significantly higher, with psychological therapy versus control at treatment completion. Statistically significant benefits persisted up to final follow-up for depression scores (12 RCTs; 856 people; -0·16, -0·30 to -0·03; 0%). Effects were strongest in RCTs of third-wave therapies and in RCTs that recruited people with impaired psychological health, fatigue, or reduced quality of life at baseline. Psychological therapies have beneficial, short-term effects on anxiety, depression, stress, and quality-of-life scores, but not on disease activity. Further RCTs in selected groups are needed to establish the place for such therapies in IBD care. None.
Read full abstract