(PS6-C58) Characterizing Empirically-derived Psychological Profiles of Patients with Chronic Constipation

Introduction: Chronic constipation (CC) is currently defined by symptom criteria reflecting heterogenous physiology.  Significant psychological comorbidities are present in many patients with CC, including depression, anxiety, avoidant/restrictive food intake disorder (ARFID), and eating disorder (ED) symptoms.  These symptoms are relevant in selecting treatments, as patients with specific psychological profiles may differentially benefit from certain treatments (e.g., cognitive behavioral therapy, medications).  However, no study has empirically derived unique psychological symptom subgroups (i.e., profiles) that could inform an alternative biopsychosocial classification model and optimize treatment selection for CC.  We sought to: 1) empirically-derive psychological symptom profiles of patients with CC using latent profile analysis (LPA) and 2) externally validate these profiles by comparing them on symptom severity, GI-specific anxiety, body mass index (BMI), and anorectal manometry parameters.

Methods: Participants included adults presenting to a tertiary care center for anorectal manometry for CC (N=468, 82% female, M_age=47, M_BMI=25 kg/m²). Included patients completed a series of questionnaires. The Hospital Anxiety and Depression Scale (HADS) and Eating Attitudes Test (EAT-26) subscales were used as indicators (i.e., variables used to derive profiles) representing unique, unexplored psychological constructs in CC. The Visceral Sensitivity Index (VSI), BMI, and anorectal manometry results were used as validator variables (i.e., variables used to examine the clinical utility of the resulting profiles). 

Results: A 5-profile solution provided the best statistical fit, and comprised the following profiles: Profile 1 comprising individuals with “high dieting, low bulimia;” Profile 2 comprising individuals with “high eating symptoms;” Profile 3 comprising individuals with “moderate eating symptoms;” Profile 4 comprising individuals with “high anxiety and depression, low eating symptoms;” and Profile 5 comprising individuals with “low psychological symptoms.”  Profiles significantly differed on each indicator variable as well as on GI-specific anxiety, CC symptoms, but not on anorectal manometry parameters or BMI.  Specifically, Profile 5 had significantly lower GI-specific anxiety, abdominal CC symptoms, and overall CC symptoms than all other profiles. 

Conclusions: Unique psychological symptom profiles characterize patients with CC, and these profiles differ on symptom severity and GI-specific anxiety.  Interestingly, profiles characterized by higher psychological symptoms of any type tended to have more severe abdominal and overall CC symptoms.  Consistent with precision medicine efforts, future research should test whether these profiles predict differential treatment outcomes.