P-028 Predictors of Suicidal Severity Amongst Suicidal IBD Patients

Abstract

Background: Suicidal ideation is a serious symptom understudied in IBD patients. We aimed to assess demographic and clinical factors of suicidal IBD patients and evaluate significant predictors of suicidal severity in IBD patients who are screened for depression at routine outpatient visits. Methods: We conducted a prospective observational study of consecutive adults presenting to a tertiary care IBD clinic over an 18-month period. All included patients have an IBD diagnosis. Patients were screened for depression using patient health questionnaire (PHQ-9). Item-9 of PHQ9 measures self-reported suicidal ideation on a scale of 0 to 3. Patients scoring ≥1 on this item were selected for further analysis. IBD disease severity was measured by Harvey- Bradshaw Crohn's Disease Index (HBCDI) or Ulcerative Colitis Activity Index (UCAI). Quality of Life (QOL) was measured using Short IBD Questionnaire (SIBDQ). Demographic data including sex, age, race, social history and marital status were obtained from electronic medical records. Other data including labs, narcotic use, medications, pain, presence of current ostomy, prior intestinal surgery, and visits to the Visceral Inflammation and Pain Center were recorded. Logistic regression modeling techniques were used for statistical analysis. In an exploratory analysis, individual depression items of PHQ9 (minus suicide) and QOL items of SIBDQ were entered into a principal components analysis (varimax rotation). Eigen values and Scree plot revealed best fit with a 2-factor solution: Factor 1 (IBD-Related) consisting of somatic depressive symptoms and functional impairment, and Factor 2 (Depression-Related) consisting of emotional and cognitive symptoms. There was no significant correlation between the 2 factors (r = 0.009; P = 0.462). Results: Seventy-one (54.9% male, 81.7% white, 71.8% CD) out of 1352 consecutively screened outpatient IBD patients were suicidal. There was significant correlation between suicidal severity and depression severity, use of tricyclic antidepressant (TCA), IBD related QOL, low vitamin D levels, and current substance use, including narcotics. Univariate regression showed that depression severity (minus suicidal item), TCA use, QOL, and substance use each significantly predicted degree of suicidality. These variables were then added into a multivariate regression prediction model with depression severity (β = 0.51; P = 0.002) and TCA use (β = 0.29; P = 0.034) making significant unique contribution; and drug use at a trend level (β = 0.25; P = 0.061). Suicidal severity was not associated with IBD activity, IBD type, inflammatory biomarkers, use of IBD medications or other psychotropic agents (other antidepressant classes). In exploratory analysis of the 2 factor scores, only factor 2 significantly correlated with suicidal severity but did not correlate with IBD activity. Opposite was true for factor 1. Univariate regression analysis of factor 2 predicting suicidality showed that it explained 14% of variance in suicidality in this model (P > 0.003; β = 0.395; P = 0.003). Conclusions: Suicidal severity is associated with depression severity (especially mood symptoms: depressed, anxious, angry) and cognitive (low self-esteem) rather than disease activity, somatic depressive symptoms (fatigue, sleep, appetite) or inflammatory biomarkers. Substance abuse, especially narcotics and low dose TCA (as a marker for chronic pain) are also risk factors. Future work needs to identify other psychosocial risk factors for suicidality.