Inflammatory bowel disease (IBD) is a chronic, disabling disease characterized by recurrent inflammation of the intestinal tract. As an incurable condition requiring lifelong treatment and having an unpredictable disease course, IBD often causes significant emotional distress. Common mental disorders (CMDs), including anxiety and depression, present major challenges for patients with IBD. These mental conditions can affect the patient’s ability to cope with the disease burden, leading to diminished treatment adherence and success.^1-3 Advances in understanding the brain-gut axis have highlighted the role of changes in brain signaling and structure, gut microbiome impairments, inflammatory cytokines, and genetic predispositions in contributing to CMDs in patients with IBD.^4-6

Recent studies have illuminated the complex bidirectional brain-gut interaction in IBD, highlighting how CMD may modify the intestinal environment, with potential implications for the disease course. Patients initially affected by CMDs subsequently face an increased risk of disease flares and the need for corticosteroids, hospitalization, emergency department visits, and IBD-related surgery.^7-12 Even in patients with quiescent IBD, the presence of abnormal anxiety at baseline has been linked to progression to active disease, necessitating glucocorticoid or therapeutic escalation.⁷ However, some reports have suggested that these psychological factors may not influence IBD outcomes as previously reported.^13,14 Therefore, a crucial gap remains in our understanding. For example, are patients with anxiety and depression less likely to achieve remission or to have a better quality of life (QoL)?

Patients with ulcerative colitis (UC), especially those with moderate-to-severe activity at diagnosis, are expected to have a more aggressive disease course, distressing symptoms (such as increased stool frequency, bleeding, abdominal pain, urgency, and tenesmus), and poor response to treatment, resulting in a greater burden on their mental health and QoL.^15,16 Our prior investigations have revealed that in patients newly diagnosed with moderate-to-severe UC, the prevalence of anxiety and depression was 33.7% and 41.8%, respectively, in the first 4 weeks after diagnosis.¹⁷ A strong association between significant anxiety or depression and poor health-related QoL (HRQoL) at diagnosis was observed in this study. However, only a few studies have examined long-term changes in HRQoL in patients with moderate-to-severe IBD and explored the predictors of changes in HRQoL. Therefore, a better understanding of the longitudinal impact of psychological distress on HRQoL and the course of IBD is required.

This prospective cohort study aimed to identify how CMDs might affect the long-term disease outcomes of patients with moderate-to-severe UC in terms of clinical improvement and HRQoL.

1. Study population and data collection

The Moderate-to-Severe Ulcerative Colitis in Korea (MOSAIK) inception cohort study is a well-established nationwide prospective study involving patients newly diagnosed with moderate-to-severe UC. Thirty academic referral hospitals in Korea enrolled UC patients who satisfied the inclusion criteria between August 2014 and March 2017.^17,18 The diagnosis was based on symptoms consistent with UC lasting for more than 4 weeks and supported by endoscopic, radiological, and histological findings. The disease severity was defined as a full Mayo score of 6 to 12 points (6 to 10 points for moderate; ≥11 for severe) and a Mayo endoscopic subscore of ≥2 points on colonoscopy or flexible sigmoidoscopy.

The patients included in the study were scheduled for annual follow-ups for 5 years after diagnosis. Comprehensive assessments of demographics, body mass index, perinatal status, vaccination, surgical history, familial history of IBD, extraintestinal manifestations, disease-related symptoms, and laboratory test results were performed at baseline. The following clinical information and patient-reported outcomes (PROs) were collected at each annual visit: UC-related symptoms, colonoscopic evaluation, Mayo scores, laboratory test results, medical records, UC-related hospitalizations, operative history, and UC treatment (Fig. 1). The study was approved by the Institutional Review Board of Kyung Hee University (IRB number: KHU 2014-01-402). Informed consent was obtained from all participants in the study.

Figure 1. Flowchart describing the MOSAIK cohort follow-up. UC, ulcerative colitis.

2. Assessment of psychological distress

Data concerning symptoms of anxiety or depression were collected at each annual visit using the Hospital Anxiety and Depression Scale (HADS). The HADS is a self-administered scale consisting of 14 items (seven items each for anxiety and depression)¹⁹ with a total score ranging from 0 to 21, with higher scores indicating more severe symptoms. The severity of HADS was graded as follows: 0–7 (normal), 8–10 (mild), 11–14 (moderate), and ≥15 (severe). We used a cutoff score of 8 to ascertain anxiety and depression, which has been thoroughly validated and identifies anxiety and depression at an early stage.²⁰

3. Ascertainment of clinical outcomes

Clinical disease activity was measured both at baseline and during follow-up using the Mayo score according to previously established protocols.¹⁸ Remission was defined as clinical improvement with a partial Mayo score (PMS) of ≤2 points, with no individual subscore >1 point. Clinical response referred to a decrease in PMS of ≥2 points and ≥30% from baseline PMS, plus either a decrease in the rectal bleeding subscore of ≥1 point or an absolute rectal bleeding subscore of 0 or 1. Relapse was defined as an increase of ≥3 points in the PMS from a state of clinical response or remission. If symptoms worsened, an unscheduled visit was recommended, and the investigator assessed for potential relapse. Depending on the investigator’s opinion, additional medication for UC treatment or an increased dose and UC-related surgery could be considered a relapse. Relapse patterns were further classified as infrequent (≤1 relapse/yr) or frequent (≥2 relapses/yr). Hospitalization owing to UC, proximal disease extension based on the Montreal classification, and colectomy were evaluated as objective clinical endpoints.

4. Ascertainment of HRQoL

HRQoL was evaluated at baseline and at follow-up using two validated Korean-translated versions of the Inflammatory Bowel Disease Questionnaire (IBDQ) and the 12-Item Short Form (SF-12) health survey. The IBDQ is widely used to measure the disease-specific HRQoL in patients with IBD. It consists of 32 items distributed across four dimensions: bowel symptoms, systemic symptoms, emotional function, and social function.^21,22 The total score ranges from 32 to 225, with higher scores indicating better HRQoL. An IBDQ score >170 was considered indicative of remission. The SF-12, a condensed version of the comprehensive SF-36 questionnaire, is a widely accepted tool for assessing generic HRQoL.²³ The SF-12 consists of 12 questions that assess various aspects of an individual’s well-being and generates two summary scores: the physical and mental component summaries (PCS and MCS, respectively). These scores offer a concise representation of a person’s physical and mental well-being based on their responses. The scale and summary scores range from 0 to 100, with higher scores indicating better HRQoL.

5. Statistical analysis

To compare the clinical characteristics and PROs at baseline and 3-year follow-up, we used a paired t-test or Wilcoxon signed-rank test for continuous variables and the McNemar test for categorical variables. We also compared clinical outcomes, including PROs after 3 years, to assess the statistical significance of differences between groups (anxiety vs non-anxiety or depression vs non-depression) based on the HADS at baseline. The Kaplan-Meier analysis was used to calculate the cumulative rates of medication use to evaluate the effects of psychological distress on medication use. To assess the influence of demographic and clinical variables on HRQoL outcomes, a linear mixed-effects model, which was employed to handle the longitudinal data using all available data, was used with subject-specific effects as random effects and an unstructured covariance structure for all models. The confounding covariates (age, sex, visit [0, 1, 2, and 3 years], and PMS) were included as fixed effects in all models. All statistical analyses were conducted using two-sided tests, and results with p<0.05 were considered statistically significant. All analyses were performed using SAS 9.4 (SAS Institute, Cary, NC, USA).

1. Changes in clinical characteristics of study population

A total of 354 consecutive patients newly diagnosed with moderate-to-severe UC between August 2014 and March 2017 were enrolled in the MOSAIK cohort. A total of 199 patients were followed up for 3 years and were included in the analysis. The mean age at diagnosis was 39.3±15.5 years and male sex was slightly predominant (60.3%). The demographic and clinical characteristics of the study population at diagnosis and 3-year follow-up are shown in Table 1. At the 3-year follow-up, 145 patients (73.6%) had achieved clinical remission. Additionally, 137 patients, accounting for 70%, had no major symptoms related to UC. All biochemical markers showed significant improvement at the 3-year follow-up compared to baseline. In our cohort of 354 patients, 197 had available baseline HADS scores and were followed for 3 years. The baseline distribution of CMDs based on HADS scores was as follows: for HADS-depression (HADS-D), 59.9% (118 patients) were classified as normal, 24.4% (48 patients) as mild, and 15.7% (31 patients) as moderate to severe. For HADS-anxiety (HADS-A), 69.5% (137 patients) were categorized as normal, 16.2% (32 patients) as mild, and 14.3% (28 patients) as moderate to severe. These baseline data illustrate the initial mental health status of the patients at the start of the follow-up period.

Table 1 . Demographic and Clinical Characteristics of the Study Population at Baseline and at the 3-Year Follow-up Visit.

Category	No. of patients (baseline/3 years)	Baseline	3 Years	p-value
Age, yr		39.3±15.51
Male sex		120 (60.3)
BMI, kg/m2	192/141	22.3±3.16	23.6±3.32	<0.001
Drinking history	194/109			0.007
Current		83 (42.8)	51 (46.8)
Never		53 (27.3)	17 (15.6)
Past		58 (29.9)	41 (37.6)
Smoking history	193/108			0.284
Current		19 (9.8)	13 (12.0)
Never		104 (53.9)	59 (54.6)
Past		70 (36.3)	36 (33.3)
CCI score	190/164	0.17±0.62	0.08±0.37	0.264
Disease extent	195/34			0.506
Proctitis (E1)		23 (11.8)	3 (8.8)
Left-sided colitis (E2)		90 (46.2)	14 (41.2)
Extensive colitis (E3)		82 (42.1)	17 (50.0)
Extraintestinal manifestation	193/180	14 (7.3)	18 (10.0)	1.000
Partial Mayo score	199/197	5.7±1.5	1.0±1.5	<0.001
Remission		0	145 (73.6)	<0.001
Mild		51 (25.6)	43 (21.8)
Moderate		121 (60.8)	9 (4.6)
Severe		27 (13.6)	0
UC-related main symptoms	197/190
None		0	137 (69.9)
Abdominal pain		19 (9.5)	11 (5.6)
Hematochezia		1 (0.5)	0
Diarrhea		92 (46.2)	22 (11.2)
Urgency		81 (40.7)	19 (9.7)
Nocturnal diarrhea		3 (1.5)	1 (0.5)
Others		1 (0.5)	0
Laboratory test
ESR, mm/hr	169/96	20.4±18.1	14.1±15.4	0.016
CRP, mg/dL	178/114	2.6±7.4	0.6±1.6	<0.001
WBC, 103/μL	196/118	8.4±3.3	6.3±1.8	<0.001
Hemoglobin, g/dL	196/118	12.8±2.0	14.0±1.6	<0.001
Albumin, g/dL	174/81	4.0±0.6	4.5±0.4	<0.001
Highest treatment before each visit*	199/199			<0.001
None		132 (66.3)	0
5-ASA		26 (13.1)	53 (26.6)
Corticosteroid		38 (19.1)	64 (32.2)
Immunomodulator		3 (1.5)	46 (23.1)
Biologics		0	36 (18.1)

Data are presented as mean±SD or number (%)..

BMI, body mass index; CCI, Charlson Comorbidity Index; UC, ulcerative colitis; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; WBC, white blood cells; 5-ASA, 5-aminosalicylic acid..

*Highest UC treatment received at baseline or in the year prior to the 3-year visit..

2. Changes in PROs

In this study, 108 patients completed the PROs at the 3-year follow-up. These subjects represent those with complete PRO data at both baseline and 3 years, allowing us to analyze changes over time in this subset. A significant improvement was observed in disease-specific HRQoL, with scores increasing from 140.6 to 188.9 (p<0.001). The proportion of patients in HRQoL remission (total IBDQ score >170) markedly improved from 22.7% to 82.6% during the 3-year follow-up period. Concurrently, both the PCS and MCS scores of the SF-12 showed incremental score increases; the mean final score exceeded 50, indicating an improvement in general HRQoL (Supplementary Table 1).

3. Subsequent clinical outcomes according to baseline psychological distress

Of the 197 patients without missing baseline psychological distress values, 95 (48.2%) had clinically significant psychological distress (encompassing anxiety and/or depression) at the time of UC diagnosis. There were 60 and 79 patients with anxiety or depression at baseline, respectively.

Table 2 provides a detailed comparison of clinical outcomes during follow-up between patients with anxiety and depression at baseline. No significant differences were observed in subsequent clinical outcomes, including the rates of clinical remission, continuous clinical response, relapse, hospitalization, and proximal disease extension. However, among the subset of patients who experienced relapse, those with both anxiety and depression were more likely to have multiple relapses (two or more) than those without these psychological conditions (p=0.026).

Table 2 . Comparison of Clinical Outcomes According to the Presence of Anxiety or Depression at Baseline.

Category	HADS-A at baseline				HADS-D at baseline
	<8 (n=137)	≥8 (n=60)	p-value		<8 (n=118)	≥8 (n=79)	p-value
Ever experienced remission	134 (97.8)	59 (98.3)	1.000		116 (98.3)	77 (97.5)	1.000
Continuous clinical response	60 (43.8)	30 (50.0)	0.421		51 (43.2)	39 (49.4)	0.396
Relapse	76 (55.5)	31 (51.7)	0.622		67 (56.8)	40 (50.6)	0.396
Aspect of relapse			0.023				0.026
Infrequent (≤1 relapse)	70 (92.1)	23 (74.2)			62 (92.5)	31 (77.5)
Frequent (>1 relapse)	6 (7.9)	8 (25.8)			5 (7.5)	9 (22.5)
Ever hospitalized	48 (35.0)	25 (41.7)	0.375		39 (33.1)	35 (44.3)	0.110
Proximal disease extension	8 (5.8)	3 (5.0)	0.813		7 (5.9)	4 (5.1)	1.000
Use of medication
Corticosteroid	87 (63.5)	35 (58.3)	0.492		67 (56.8)	56 (70.9)	0.045
Immunomodulator	55 (40.1)	22 (36.7)	0.645		46 (39.0)	31 (39.2)	0.971
Biologics	23 (16.8)	11 (18.3)	0.792		19 (16.1)	15 (19.0)	0.599

Data are presented as number (%)..

HADS-A, Hospital Anxiety and Depression Scale-Anxiety; HADS-D, Hospital Anxiety and Depression Scale-Depression..

A logistic regression analysis was conducted to pinpoint factors associated with major clinical outcomes, but neither HADS-A nor HADS-D scores at baseline were significantly associated with subsequent clinical results (such as relapse, hospitalization, and a PMS ≥2 at the 3-year follow-up) (data not shown).

Notably, when analyzing the impact of baseline psychological distress on medication use, patients with depression displayed a higher cumulative use of corticosteroids than those without depression. However, this difference was not significant (Supplementary Figs 1 and 2).

4. Changes in HRQoL according to baseline psychological distress

The 3-year follow-up results revealed discernible variations in HRQoL according to psychological distress at the onset of diagnosis. Specifically, patients diagnosed with anxiety displayed significantly lower scores not only in total but also across nearly all domains of IBD-specific QoL measured by the IBDQ at the end of the follow-up period compared to those without anxiety. Despite not reaching statistical significance, the decrease in IBDQ scores observed in patients with depression over the 3-year follow-up period indicated a clinically relevant trend (Fig. 2).

Figure 2. Inflammatory Bowel Disease Questionnaire (IBDQ) scores after a 3-year follow-up according to common mental disorders present at baseline. (A) Total IBDQ score. (B) Association of the IBDQ subscore with the Hospital Anxiety and Depression Scale-Anxiety (HADS-A) score. (C) Association of the IBDQ subscore with the HADS-Depression (HADS-D) score. *p<0.05.

Supplementary Table 1 shows the changes in the IBDQ scores of the patients according to their psychological distress at baseline. The discrepancy in IBDQ scores between the two groups was considerable at the time of UC diagnosis and at 1-year follow-up. However, this gap narrowed substantially 3 years after the diagnosis of UC.

Regarding general QoL, an analysis of the mean SF-12 subscores suggested that patients experiencing psychological distress at diagnosis generally reported poorer HRQoL than those without distress. Specifically, this difference was significant for the MCS scores. If anxiety or depression were present at baseline, the final MCS score did not exceed 50 during the follow-up period (Fig. 3).

Figure 3. The 12-Item Short Form (SF-12) summary score after a 3-year follow-up according to the common mental disorders present at baseline. (A) Physical component summary. (B) Mental component summary. HADS-A, Hospital Anxiety and Depression Scale-Anxiety; HADS-D, Hospital Anxiety and Depression Scale-Depression. *p<0.05.

5. Factors independently associated with changes in HRQoL

In the final stage of the analysis, parameters from both baseline and year 3 were incorporated into a longitudinal linear mixed model. This model accounts for the interdependencies between values at two timepoints for each individual. The initial parameter estimates were modified to account for the potential confounding factors. Disease activity and psychological distress were key determinants of HRQoL in the study cohort. The definitive model showed that elevated PMSs, along with higher HADS-A or HADS-D scores, were independently associated with diminished IBDQ scores while regular follow-up visits over time was associated with improved IBDQ scores (Table 3).

Table 3 . Factors Independently Associated with Improved HRQoL and Linear Mixed-Effect Models for the IBDQ (n=199).

Variable	Reference	Total IBDQ
		Individual model				Final model
		Beta	SE	p-value		Beta	SE	p-value
Age	-	-	-	-		0.10	0.078	0.199
Sex (female)	Male	-	-	-		–3.70	2.388	0.123
Visit (years)	-	-	-	-		3.77	1.106	<0.001
Partial Mayo score	-	-	-	-		–5.25	0.490	<0.001
BMI (kg/m2)		0.13	0.542	0.814		-	-	-
Past smoking	Never	–2.07	4.463	0.644		-	-	-
Current smoking	Never	–7.39	5.647	0.191		-	-	-
Past alcohol consumption	Never	–5.72	4.118	0.166		-	-	-
Current alcohol consumption	Never	–5.77	4.034	0.153		-	-	-
Extraintestinal manifestation	No	0.52	4.360	0.905		-	-	-
Disease extent (E2)	E1	2.20	6.403	0.731		-	-	-
Disease extent (E3)	E1	–9.44	6.681	0.159		-	-	-
Number of hospitalizations	-	–4.98	1.659	0.003		–1.36	1.305	0.299
Relapse	No	–3.20	3.446	0.353		-	-	-
Steroid use	No	–1.66	2.859	0.561		-	-	-
Immunomodulator use	No	–1.00	3.215	0.759		-	-	-
Biologics	No	2.63	4.624	0.570		-	-	-
HADS-A	-	–4.44	0.281	<0.001		–2.98	0.346	<0.001
HADS-D	-	–4.23	0.307	<0.001		–2.35	0.357	<0.001

HRQoL, health-related quality of life; IBDQ, Inflammatory Bowel Disease Questionnaire; SE, standard error; BMI, body mass index; HADS-A, Hospital Anxiety and Depression Scale-Anxiety; HADS-D, Hospital Anxiety and Depression Scale-Depression..

Similarly, in the MCS domain of SF-12, increased PMS, and higher HADS-A or HADS-D scores corresponded to reduced HRQoL, while regular follow-up visits over time was related with improved HRQoL. Within the PCS, higher HADS-D scores were significantly related to decreased HRQoL, although the association between PMS and HADS-A scores was not significant. Intriguingly, in contrast to the IBDQ, age exhibited an additional significant positive correlation with MCS and a negative correlation with PCS (Table 4).

Table 4 . Factors Independently Associated with Improved HRQoL and Linear Mixed-Effect Models for the SF-12 (n=199).

Variable	Reference	SF-12 (MCS)								SF-12 (PCS)
		Individual model				Final model				Individual model				Final model
		Beta	SE	p-value		Beta	SE	p-value		Beta	SE	p-value		Beta	SE	p-value
Age	-	-	-	-		0.05	0.022	0.026		-	-	-		–0.06	0.030	0.046
Sex (female)	Male	-	-	-		–0.37	0.678	0.584		-	-	-		–0.68	0.989	0.491
Visit (years)	-	-	-	-		0.73	0.330	0.028		-	-	-		0.80	0.581	0.170
Partial Mayo score	-	-	-	-		–0.70	0.146	<0.001		-	-	-		–0.49	0.220	0.026
BMI (kg/m2)		0.21	0.173	0.231		-	-	-		0.10	0.129	0.443		-	-	-
Past smoking	Never	–0.07	1.419	0.960						–0.89	1.032	0.391
Current smoking	Never	–2.42	1.812	0.183		-	-	-		0.05	1.332	0.972		-	-	-
Past alcohol consumption	Never	–0.99	1.353	0.466						0.60	0.996	0.545		–1.53	1.242	0.220
Current alcohol consumption	Never	–0.95	1.321	0.472		-	-	-		2.44	0.972	0.012		0.70	1.144	0.542
Extraintestinal manifestation	No	–2.00	1.472	0.175		-	-	-		–2.21	1.114	0.048		–0.97	1.554	0.532
Disease extent (E2)	E1	–0.29	2.058	0.890		-	-	-		0.27	1.531	0.859		0.51	1.405	0.716
Disease extent (E3)	E1	–2.86	2.143	0.183		-	-	-		–3.08	1.588	0.054		–1.51	1.468	0.304
Number of hospitalizations	-	–1.52	0.549	0.006		–0.04	0.379	0.909		–1.21	0.413	0.004		–1.26	0.730	0.085
Relapse	No	–1.00	1.202	0.404		-	-	-		–0.93	0.938	0.323		-	-	-
Steroid use	No	0.06	0.945	0.952		-	-	-		–0.49	0.710	0.492		-	-	-
Immunomodulator use	No	–0.43	1.076	0.690		-	-	-		0.52	0.819	0.528		-	-	-
Biologics	No	0.34	1.567	0.826		-	-	-		–1.18	1.196	0.323		-	-	-
HADS-A	-	–1.74	0.083	<0.001		–1.26	0.101	<0.001		–0.53	0.084	<0.001		–0.11	0.141	0.441
HADS-D	-	–1.61	0.095	<0.001		–0.80	0.104	<0.001		–0.73	0.083	<0.001		–0.64	0.151	<0.001

HRQoL, health-related quality of life; SF-12, 12-Item Short Form; MCS, mental component summary; PCS, physical component summary; SE, standard error; BMI, body mass index; HADS-A, Hospital Anxiety and Depression Scale-Anxiety; HADS-D, Hospital Anxiety and Depression Scale-Depression..

Using a well-defined inception cohort of patients with UC, we assessed how psychological distress at the time of diagnosis affected disease course and HRQoL. To the best of our knowledge, this is the first study to investigate the relationship between depression or anxiety and long-term disease outcomes in non-Western patients with UC, and in those with moderate or high disease activity. We found that, in moderate-to-severe UC, the presence of anxiety or depression at diagnosis had no effect on clinical outcomes. However, among patients who experienced relapse, those with anxiety and depression were more likely to experience multiple relapses than those without these conditions (p=0.026). The limited number of participants highlights the need for larger studies to validate these findings. The most important finding of this study is that underlying anxiety and depression affect the long-term outcomes of both IBD-specific and general HRQoL, independent of clinical activity. While this may sound obvious, it reiterates that managing mental health is critical for achieving therapeutic goals in long-term care strategies for patients with moderate-to-severe IBD.

Several studies have shown that the risk of psychiatric morbidity is the highest in the first year after IBD diagnosis.^24,25 For chronic illnesses, psychiatric symptoms present at the time of diagnosis can worsen over time. Herein, we examined time-series changes in QoL according to CMDs at diagnosis and found that differences in IBDQ scores between subgroups with and without anxiety and depression were maintained for up to 3 years. However, the difference appeared to be more pronounced earlier in the disease, suggesting that clearly defining this relationship in the early stages of the disease can be important for patients and physicians in maintaining tight disease control.

The most recent meta-analysis to date, which included 9,192 patients from 12 studies, noted that anxiety or depression at baseline was related to future risks of escalation of therapy, hospitalization, emergency department attendance, surgery, or a composite of any unfavorable outcomes.²⁶ While there is growing evidence regarding the association between psychological conditions and clinical courses, generalizing these findings to severe patients is limited due to the diverse characteristics of the participants, particularly with varying disease statuses. From this perspective, the MOSAIK cohort can be considered homogeneous, focusing on a well-defined, specific patient group. These results conflict with the findings of previous studies that demonstrated an association between antecedent psychological comorbidities and subsequent poor clinical outcomes.^7,27 The study conducted by Yoon et al.²⁸ using the same inception cohort to observe the early disease course of moderate-to-severe UC in Korea found that Korean patients may experience more favorable early outcomes than Western patients. At year 1, 95.3% of patients achieved clinical remission, and 58.1% showed endoscopic remission. When the observation period was extended to 3 years, we confirmed that this favorable prognosis was well maintained, with 73% of patients showing clinical remission and 70% of patients being free of UC-related symptoms. Early therapeutic intervention and the achievement of early clinical treatment goals are decisive factors that affect long-term clinical prognosis. In a previous study of the same cohort focusing on the initial stages of HRQoL, it was confirmed that consistent treatment over 1 year can improve HRQoL in patients with moderate-to-severe UC.²⁹ Indeed, this may have diluted the impact of psychological distress on clinical outcomes. It emphasizes the necessity of tailored therapeutic strategies that consider both the psychological and clinical dimensions of UC, particularly in severe cases. Future studies might benefit from exploring how different treatment strategies can modulate the interplay between mental health and clinical outcomes in such specific cohort.

Fairbrass et al.²⁶ followed 760 patients for 6.5 years, assessing the combined impact of psychological health and disease activity. Patients were divided into four groups based on their initial mental disorders and disease status. However, this study did not clearly define the influence of mental health on different IBD subtypes. There is potential heterogeneity in the effects of psychological symptoms on the disease course between patients with Crohn’s disease (CD) and UC. A previous study across multiple institutions found that mood or anxiety comorbidities were linked to a 28% higher surgery risk in CD patients, but this association was not observed in UC, indicating a more significant impact on CD.³⁰ Our study, which focused on moderate-to-severe UC, did not establish a clear link between CMD and long-term clinical outcomes. This disparity may stem from different primary mechanisms of CD and UC. This suggests that psychological stress may affect various components of the immune system differently, which warrants further investigation.

Mental disorders significantly affect the QoL and healthcare utilization in patients with IBD, independent of their clinical outcomes. Patients with IBD and active psychiatric comorbidities had notably more physician visits, hospital days, and drug use than those with inactive psychiatric comorbidities.³¹ While these patients do not necessarily have higher rates of treatment escalation or surgery, they require more clinical contact and examination.³² Our study aligns with these observations; there were no significant differences in clinical outcomes based on the initial HADS scores. However, following a 3-year longitudinal observation, we found that baseline CMDs are associated with lower HRQoL. This suggests that such conditions may also influence healthcare utilization and indirect healthcare costs, echoing findings from the literature. We also observed that an increase in the number of scheduled visits was independently associated with improvements in HRQoL. This suggests that rather than frequent visits being a predictive factor for HRQoL, regular follow-up visits over time are related to improvement in HRQoL.

The disparity between the associated factors, especially psychological factors, of PCS and MCS in the HRQoL can be attributed to the fundamentally different aspects of health they measure and the specific ways in which various factors influence these dimensions. Psychological factors such as anxiety and depression have a more pronounced effect on MCS because these conditions directly impact mental health and emotional functioning. Emotional distress can alter perceptions of vitality and social functioning, which are key components of MCS. In contrast, PCS may not show a strong association with anxiety because the physical symptoms might not directly correlate with anxiety levels. However, depression, which can have somatic symptoms, shows a significant relationship with PCS, affecting perceptions of physical health and pain.

Our study had several limitations. First, of the 354 patients initially enrolled, only 199 were successfully followed over a 3-year period. This smaller sample size might have reduced the power to fully evaluate outcomes. Nonetheless, our methodology aligns with that typical of real-world, non-interventional observational studies. Additionally, there was a concern about potential confusion due to varying numbers of analysis subjects. However, our study primarily focuses on longitudinally examining outcomes based on the baseline presence of CMDs among the 199 patients. By utilizing a linear mixed-effect model, we were able to incorporate all available data points from the 199 patients, which enhances the robustness and reliability of our findings. This approach ensures that we make full use of the available data, including valuable information from patients who did not complete the full 3-year follow-up. Second, recognizing CMDs can be particularly difficult for people with chronic conditions such as fatigue, pain, loss of appetite, sleep disturbances, depression/anxiety, and other physical illnesses. As the study population comprised patients with moderate-to-severe disease activity who attended a tertiary hospital, we cannot exclude the possibility that psychiatric illnesses were masked by severe physical distress. To overcome the underdiagnoses of mental disorders we used lower cutoff values (HADS ≥8) and tried to reduce false negatives. Third, the remission rate at the 3-year follow-up in this study is very high, and it has not been confirmed whether factors related to disease severity or status might impede the generalization of these results. Specifically, we did not investigate whether there were variations in HRQoL 3 years later among different subgroups, such as those based on treatment response after induction therapy or steroid-refractoriness. However, to evaluate how clinical factors impact HRQoL results, we utilized a linear mixed-effects model. This approach considered individual-specific impacts as random variables and adopted an unstructured covariance frame work for every model. To rule out the possibility that disease activity at 3-year follow-up may have influenced QoL in patients with UC at the time of the final HRQoL assessment, PMSs were included as fixed effects in each model. Fourth, due to a high rate of missing data from the Work Productivity and Activity Impairment questionnaire during the 3-year follow-up, we were unable to assess the impact of CMDs on work disability and social activity impairment. Finally, our study lacks detailed medication compliance data. We only collected information on the start and end dates of medication based on prescription records, which does not provide a complete picture of actual patient compliance. Anxiety and depression are known to influence medication adherence, and without detailed compliance data, we cannot fully assess their impact on clinical outcomes in our cohort. This limitation suggests that our conclusions regarding the relationship between emotional states and HRQoL should be interpreted with caution. Future research should include comprehensive compliance data to better understand this dynamic and strengthen the validity of the findings.

In conclusion, CMDs (especially anxiety at the time of diagnosis) were associated with long-term QoL in patients with moderate-to-severe UC. This association remained valid irrespective of subsequent clinical outcomes. Depression and anxiety are common consequences of UC and should be considered part of the disease course. Therefore, integrating physical and mental health when caring for patients with moderate-to-severe UC is critical for improving their QoL.

The Moderate-to-Severe Ulcerative Colitis in Korea (MOSAIK) cohort study has been funded and supported by Janssen Korea.

The authors thank the members of the MOSAIK study group for participating in this study. We are deeply grateful to Professor Hyo Jong Kim for his visionary guidance and invaluable mentorship during this study.

The Moderate-to-Severe Ulcerative Colitis in Korea (MOSAIK) cohort study has been funded and supported by Janssen Korea. No other potential conflict of interest relevant to this article was reported.

Study concept and design: S.J.O., C.K.L. Data acquisition: S.J.O., C.H.C., S.A.J., G.A.S., Y.K.K., J.S.K., S.J.S., G.S.S., K.M.L., B.I.K., E.S.J. Data analysis and interpretation: S.J.O., C.K.L., Y.K. Manuscript drafting: all authors. Critical revision of the manuscript for intellectual content and approval of the final manuscript: all authors.