Type D Personality Is Associated With Depressive Symptoms and Clinical Activity in Inflammatory Bowel Disease

Sebastian Bruno Ulrich Jordi; Federica Botte; Brian Matthew Lang; Thomas Greuter; Niklas Krupka; Bianca Auschra; Philipp Schreiner; Michael Christian Sulz; Luc Biedermann; Roland von Känel; Gerhard Rogler; Stefan Begré; Benjamin Misselwitz

Disclosures

Aliment Pharmacol Ther. 2021;54(1):53-67. 

In This Article

Results

Study Participants

For our analysis, data of 2275 Swiss IBD cohort study patients (672 with Type D; 29.5%) were available with regular follow-ups for up to 4250 days. Baseline characteristics of our cohort patients at enrolment are presented in Table 1, confirming a mixed IBD population (1005 with UC/IBDu, 1270 with CD) with mild, moderate and severe disease. At enrolment, Type D was significantly associated with higher disease activity (standardised CDAI/MTWAI, P < 0.001, Figure 1A,B), extraintestinal manifestations and depressive symptoms.

Figure 1.

Distribution of Type D and depression (HADS-D ≥11). (A) Distribution of disease activity levels, stratified for the presence of Type D. (B) Distribution of disease activity levels, stratified for the presence of depression. (C) Distribution of disease activity levels and HADS-D scores, stratified for the presence of Type D. The dashed blue and red lines represent the respective groups' medians. P values were calculated using Mann-Whitney U or Kruskal-Wallis tests, respectively. CDAI, Crohn's Disease Activity Index; HADS, Hospital Anxiety and Depression Scale; HADS-D, HADS subscale for depressive symptoms; IBD, inflammatory bowel disease; MTWAI, Modified Truelove and Witts Severity Index

Effects of Type D on IBD Course

Patients with Type D had a significantly higher hazard (HR) to experience active disease (CDAI ≥150/MTWAI ≥10; HR: 1.62, q = 0.004), a finding which remained significant after correction for confounders (CDAI ≥150/MTWAI ≥10; adjusted HR (aHR): 1.60, q = 0.007, Table 2 and Figure 2). Associations between Type D and any new extraintestinal manifestation during follow-up lost statistical significance after correction for multiple testing in univariable (HR: 1.20, q = 0.164, Table 2) as well as in multivariable models (aHR: 1.21, q = 0.151, Table 2 and Figure 2).

Figure 2.

Kaplan Meier curves showing how Type D increases the hazard for active disease (CDAI ≥150/MTWAI ≥10) and new extraintestinal manifestations over time. (A) Percentage of IBD patients without experiencing occurrence of active disease, stratified according to the presence of Type D. (B) Percentage of IBD patients without experiencing new occurrence of any extraintestinal manifestation, stratified according to the presence of Type D. Extraintestinal manifestations included peripheral arthritis/arthralgia, uveitis/iritis, pyoderma gangrenosum, erythema nodosum, aphthous oral ulcers/stomatitis, ankylosing spondylitis/sacroiliitis and primary sclerosing cholangitis. Analyses: multivariable Cox proportional hazards models; q values were obtained by applying Bonferroni correction for four tested endpoints (q = P × 4). aHR, adjusted hazard ratio; CI, 95% confidence interval; EIM, extraintestinal manifestation

We also observed a significant negative impact of Type D on the clinical exploratory endpoints physician-reported flares (aHR: 1.19, P = 0.017) and primary non-response to therapy (aHR: 1.42, P = 0.037) after correction for confounders (Supplemental digital content 1B,F, univariable results: Supplemental digital content 2B,F). Patients with Type D also tended to experience most other clinical endpoints earlier except for new stenosis and new or changed TNF inhibitor therapy (Supplemental digital content 1). Furthermore, Type D also increased the risk for new complications (aHR: 1.36, P = 0.013) and peripheral arthritis/arthralgia (aHR: 1.23, P = 0.040) with a similar trend for other individual extraintestinal manifestations (Supplemental digital content 3A,C, univariable results: Supplemental digital content 4).

Moreover, patients with Type D were also more likely to experience IBD recurrence defined by composite endpoints FNCE[6] and AFFSST[5] (Figure 3). For FNCE, Type D was a significant predictor for recurrence in univariable (HR: 1.25, 95% confidence interval [CI]: 1.10–1.42, q = 0.002) and in multivariable Cox proportional hazards models (aHR: 1.23, CI: 1.08–1.40, q = 0.006; Table 2). For the association of Type D with AFFSST, we observed nominal significance but Type D's predictive value for lost statistical significance after multiple testing correction in both univariable (HR: 1.16, CI: 1.03–1.31, q = 0.069) and multivariable models (aHR: 1.15, CI: 1.02–1.30, q = 0.103; Table 2).

Figure 3.

Kaplan Meier curves showing how Type D increases the hazard for clinical recurrence (composite endpoints) over time. Percentage of IBD patients without clinical recurrence according to the FNCE (A) and AFFSST (B) definition, stratified according to the presence of Type D. Analyses: multivariable Cox proportional hazards models; q values were obtained by applying Bonferroni correction for four tested endpoints (q = P × 4). AFFSST, active disease, physician-reported flare, new fistula, new stenosis, surgery, or new systemic therapy; CI, 95% confidence interval; FNCE, physician-reported flare, non-response to therapy, new complication or extraintestinal manifestation; HR, hazard ratio; IBD, inflammatory bowel disease

Subgroup analyses did not reveal any differences between CD and UC/IBDu subgroups regarding the impact of Type D on active disease or the occurrence of any new extraintestinal manifestation and composite endpoints (FNCE, AFFSST); however, with these smaller subgroups, significance was lost for most endpoints (Supplemental digital content 5).

A patient's diagnosis (CD vs UC/IBDu) and sex[41] are potentially relevant prognostic factors in IBD. However, both seem to be independent of Type D status as (a) Type D prevalence was similar in patients with CD (29.9%) and UC/IBDu (29.1%, P = 0.687) and patients of both sexes with a prevalence of 30.8% amongst women compared with 28.2% in men (P = 0.188) and (b) diagnosis and sex showed partially significant prognostic value in multivariable analyses independent of Type D (Table 2 and Table 3).

Type D and Depressive Symptoms

Type D was more prevalent in patients with high HADS-D scores (Figure 1C). Furthermore, logistic regression models showed significant (P < 0.001) associations between Type D and depressive symptoms at enrolment with odds ratios of 6.7 and 6.8, respectively (Supplemental digital content 6).

Type D also had a predictive value for the occurrence of depressive symptoms after enrolment. In Cox proportional hazards models, Type D was significantly (P < 0.001) associated with depressive symptoms over time showing HRs of 3.3–3.35 (Supplemental digital contents 7 and 8). These findings remained robust when patients that showed depressive symptoms at enrolment were excluded (Supplemental digital contents 9 and 10).

Interaction Between Type D and Depressive Symptoms

The predictive value of Type D for the occurrence of active disease (CDAI ≥150/MTWAI ≥10), any new extraintestinal manifestation and disease recurrence (FNCE, AFFSST) decreased in magnitude and significance when corrected for depressive symptoms (Supplemental digital contents 11 and 12). Depression was coded as a time-varying variable over a patient's follow-up time. In such depression-corrected Cox proportional hazards models, Type D remained a significant predictor for FNCE (aHR: 1.19, CI: 1.04–1.36, q = 0.043). However, no significant association was detected regarding the occurrence of active disease (CDAI ≥150/MTWAI ≥10, aHR: 1.32, CI: 0.97–1.79, q = 0.292), any new extraintestinal manifestation (aHR: 1.15, CI: 0.95–1.38, P = 0.564) and AFFSST events (aHR: 1.10, CI: 0.97–1.25, q = 0.575).

To test for synergistic interactions of Type D and depressive symptoms, we compared patients with both, Type D and depression and either condition alone with patients having neither Type D nor depression. Patients with Type D and depressive symptoms in combination were at greatest risk regarding the occurrence of active disease (CDAI ≥150/MTWAI ≥10, aHR: 3.98, CI: 2.56–6.20, q < 0.001), any new extraintestinal manifestation (aHR: 1.60, CI: 1.13–2.28, q = 0.105), FNCE (aHR: 1.55, CI: 1.21–1.98, q = 0.006) and AFFSST events (HR: 1.74, CI: 1.37–2.21, q < 0.001, Table 3 and Figures 4 and 5).

Figure 4.

Kaplan Meier curves showing how Type D and depression (HADS-D ≥11) in combination increase the hazard for active disease (CDAI ≥150/MTWAI ≥10) and new extraintestinal manifestations over time. (A) Percentage of IBD patients without experiencing occurrence of active disease, stratified for different combinations of Type D and/or depression. (B) Percentage of IBD patients without experiencing new occurrence of any extraintestinal manifestation, stratified for different combinations of Type D and/or depression. Extraintestinal manifestations included peripheral arthritis/arthralgia, uveitis/iritis, pyoderma gangrenosum, erythema nodosum, aphthous oral ulcers/stomatitis, ankylosing spondylitis/sacroiliitis and primary sclerosing cholangitis. Analyses: multivariable Cox proportional hazards models; q- values were obtained by applying Bonferroni correction for 12 tests (q = P × 12). aHR, adjusted hazard ratio; CI, 95% confidence interval; EIM, extraintestinal manifestation

Figure 5.

Kaplan Meier curves showing how Type D and depression (HADS-D ≥11) in combination increase the hazard for clinical recurrence (composite endpoints) over time. Percentage of IBD patients without clinical recurrence according to the FNCE (A) and AFFSST (B) definition, stratified for different combinations of Type D and/or depression. Analyses: multivariable Cox proportional hazards models; q values were obtained by applying Bonferroni correction for 12 tests (q = P × 12). AFFSST, active disease, physician-reported flare, new fistula, new stenosis, surgery, or new systemic therapy; CI, 95% confidence interval; FNCE, physician-reported flare, non-response to therapy, new complication or extraintestinal manifestation; HR, hazard ratio; IBD, inflammatory bowel disease

Type D and depression in combination showed also the highest hazards for almost all clinical exploratory endpoints and individual extraintestinal manifestations; however, significance (without correction for multiple testing) was only observed for physician-reported flare (aHR: 1.39, P = 0.024), new surgery (aHR: 2.35, P < 0.001), primary non-response to therapy (aHR: 2.38, P = 0.004), new peripheral arthritis/arthralgia (aHR: 1.76, P = 0.002) and new primary sclerosing cholangitis (aHR: 4.13, P = 0.030, Supplemental digital contents 13 and 14).

Subgroup analyses did not reveal any differences between CD and UC/IBDu subgroups. Despite smaller sample size, was maintained in both subgroups for several endpoints (Supplemental digital content 15).

Type D and Therapy Adherence

For a subsample of 690 patients (211 with Type D, 479 without), data on self-reported adherence to a total of 1288 prescribed pharmacological therapies during the first follow-up period were available. In this subsample, Type D was significantly associated with lower levels of therapy adherence during the first follow-up period (P < 0.001, Supplemental digital content 16A). Among patients with Type D, 36.8% did not respect either the dosage or frequency of their prescribed medication or did not take it at all, compared to 23.5% of patients without Type D (OR: 1.89, CI: 1.47–2.43, P < 0.001, Supplemental digital content 16). Moreover, significantly (χ 2 test, P = 0.008) more pharmacological therapies were prescribed to patients with Type D (2.08 per patient) compared to patients without Type D (1.77 per patient). In contrast, types of medications prescribed did not differ significantly between patients with and without Type D (P = 0.713, Supplemental digital content 16B).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....