Prevalence and determinants of fatigue in patients with IBD: a cross-sectional survey from the GETAID.

Conclusion: Determinants of fatigue are not restricted to IBD-related factors but also include social factors, sleep, and emotional disturbances, thus supporting a holistic approach to IBD patient care.


Aims:
To identify the factors associated with fatigue in a large population of patients with IBD.
Patients and methods: Fatigue and nine other IBD-related disability dimensions were assessed in a cohort of 1704 consecutive patients with IBD using the IBD-disk questionnaire in a cross-sectional survey of 42 French and Belgian centers.Fatigue and severe fatigue were defined as energy subscores >5 and >7, respectively.Determinants of fatigue were assessed using univariate and multivariate analyses (odds ratios (ORs) are provided with 95% confidence intervals).

Results:
The prevalence rates of fatigue and severe fatigue were 54.1% and 37.1%, respectively.Both fatigue and severe fatigue were significantly higher in patients with active disease than in patients with inactive disease (64.9% vs. 44.7%and 47.4% vs. 28.6%,respectively; p<0.001 for both comparisons).In the multivariate analysis stratified by age, sex, type of IBD and IBD activity, fatigue was associated with age >40 years (OR=0.71[0.54-0.93]),female sex (OR=1.48[1. 13

INTRODUCTION
Inflammatory bowel diseases (IBDs), including Crohn's disease (CD) and ulcerative colitis (UC), are chronic relapsing and remitting inflammatory diseases affecting the gastrointestinal tract 1,2 .IBD is responsible for significant disability [3][4][5] .Disability is an umbrella term encompassing impairments, activity limitations and participation restrictions, thus summarizing the functional dimension of the handicap 6,7 .Disability could be studied using two validated tools (the IBD-disability index and the IBD-disk) that assess ten dimensions of disability, including joint pain, abdominal pain, regulation of defecation, interpersonal interactions, education and work, sleep, energy, emotions, body image, and sexual functions [3][4][5] .
Fatigue is one of the most prevalent and burdensome symptoms reported by patients with IBD 8 .Recently, a systematic review with meta-analysis estimated the prevalence of fatigue in 12 025 patients with IBD to be 47%, with no difference between CD and UC 9 .Active IBD, anxiety and depression, sleep disturbances and anemia were the most commonly reported fatigue-associated factors.That study found high heterogeneity depending on the definition of fatigue and on the various datasets of patient characteristics.Indeed, fatigue is a multifactorial and multimodal symptom that could be linked not only to inflammation and micronutrient deficits but also to psychosocial and sleep disturbances as well as to exercise deconditioning 10,11 .
The aim of the present study was to identify the determinants of fatigue in a large population of patients with IBD.

Study population
Between 26 and 30 November 2018, all consecutive outpatients with IBD who were seen at the outpatient clinic or infusion unit were asked to participate in a cross-sectional multicenter survey 4,12 .Participating centers were all affiliated with the Groupe d'Etude Therapeutique des Affections Inflammatoires du tube Digestif (GETAID) in France or Belgium.This study was conducted in accordance with French ethical and legal principles through reference methodology MR-004 (Institut National des Donnees de Sante registration number 2210131).

Survey instrument
All patients completed the IBD-disk assessing ten dimensions of IBD-related disability: joint pain, abdominal pain, regulation of defecation, interpersonal interactions, education and work, sleep, energy, emotions, body image, and sexual functions (Table S1 and Figure S1).For each dimension of disability, patients had to score their level of agreement with a specific statement on a 0-10 visual analog scale (VAS; 0 = no burden; 10 = maximal burden).Fatigue was assessed according to the energy statement of the IBD-disk ('I have not felt rested and refreshed during the day and have felt tired and without energy').
According to the original instructions, any subscore > 5 represents a significant disability for each IBD-disk component 13 .Fatigue was defined as an IBD-disk energy subscore > 5, and severe fatigue was defined as > 7.
The recorded data included patient demographics, occupational status, weight and height, type of IBD, age at diagnosis of IBD, duration of IBD, history of surgical treatment of ACCEPTED MANUSCRIPT / CLEAN COPY IBD, current treatment, patient-global assessment of clinical remission, IBD-related sick leave rate, IBD daily life burden (0-10 visual analog scale), frequency of appointments with a gastroenterologist, perceived need for other health care professionals such as a dietician, a psychotherapist, a sexologist, a sports coach, an IBD nurse or a social worker.Current treatments included 5-aminosalicylates, conventional immunosuppressants (namely thiopurine or methotrexate), anti-TNF agents (namely adalimumab, infliximab, certolizumab pegol or golimumab), vedolizumab, ustekinumab, tofacitinib or any other treatment.

Study objectives
The aim of the present study was to assess the prevalence of fatigue and severe fatigue among a large cohort of outpatients with IBD and to identify factors associated with fatigue and severe fatigue.

Statistical analysis
The data are expressed as numbers (%) for qualitative data and as the mean ± standard deviation (SD) or median [interquartile range] for quantitative data.In the case of a partially incomplete IBD-disk questionnaire (≥ 6 out 10 completed items), missing data were generated using multiple imputations based on sociodemographic data and available IBD-disk subscores.Multiple imputations consisted of the generation of 50 different complete datasets finally combined in a pooled dataset.Patients with five or more IBD-disk missing items were excluded from the present study.Spearman's rho correlation coefficient® was determined to identify correlations between fatigue and other variables.To identify predictors of fatigue and severe fatigue, univariate analysis using the chi²-test was performed.Subsequent multivariate analysis using binary logistic regression models was ACCEPTED MANUSCRIPT / CLEAN COPY performed and adjusted for using the abovementioned variables.Variables with p<0.10 in the univariate analysis were considered potential adjustment variables for the multivariate analysis.Multivariate analysis was stratified by age, sex, type of IBD and IBD activity.All analyses were two-tailed, and p values less than 0.05 were considered significant.All statistics were calculated using SPSS statistical software (SPSS Inc., v23, Chicago, IL, USA).

Study population
We included 1704 patients in 42 tertiary care centers in the present study (802 (47.1%) males, median age of 39.0 [interquartile range (IQR) 29.0-51.0]years; 67.7% of Crohn's disease).Table 1 summarizes the main characteristics of the study population.The median duration of IBD was 10.5 [IQR 5.5-18.5]years.Current treatment included 5aminosalicylates in 8.6% of the cases, conventional immunosuppressants alone in 10.4%, anti-TNF agents in 54.8%, vedolizumab in 13.8% and ustekinumab in 6.5%, and for 6.3% of the cases, no treatment was currently prescribed.

Prevalence of fatigue and severe fatigue in the study population
The mean overall fatigue VAS was 5.7 ± 3.2 (Table 1).The prevalence of fatigue was 54.1%, and the prevalence of severe fatigue was 37.1%.The patient global assessment of clinical remission was missing in 205 out of 1704 patients.The prevalence of fatigue was significantly higher in patients with active disease according to the patient global assessment (64.9% vs. 44.7%,p<0.001).The prevalence of severe fatigue was significantly higher in patients with active disease according to the patient global assessment (47.4% vs. 28.6%, p < 0.001).

Correlations between fatigue and other variables
Significant correlations were observed between fatigue and all the other dimensions of IBD-related disability assessed by the IBD-disk (Table 2).The strongest correlations were observed between fatigue and sleep disturbances (R² = 0.624) and between fatigue and emotional disturbances (R² = 0.633).In patients with fatigue, sleep and emotional disturbances were significantly higher compared with those without fatigue (69.7% vs. 20.1% and 61.5% vs. 14.0%,respectively, P < 0.001 for both comparisons).

Predictors of fatigue and severe fatigue in the study population
In the multivariate analysis stratified by age, gender type of IBD and IBD activity,  5).

DISCUSSION
Fatigue is a burdensome, multidimensional, and multifactorial symptom implying a lack of energy or exhaustion that is out of proportion to physical exertion, that limits daily activities, and that is not relieved by rest 14 .In this large multicenter study, we showed that fatigue was influenced by age, sex, and many impaired dimensions assessed by the IBD-disk, mostly emotional and sleep disturbances.In parallel, fatigue was associated with IBDrelated sick leave and clinical symptoms, such as joint and abdominal pain, and difficulties in regulating defecation.
Sleep disturbances are commonly reported in the general population and are linked with worsening of health-related quality of life 15 .Sleep disturbances have also been widely reported in patients with IBD, mostly with questionnaire-based tools rather than with polysomnography.In a recent meta-analysis, patients with IBD had poorer sleep quality than controls 16 .Sleep disturbances and IBD activity are mutually impacted through the brain-gut axis.Indeed, in another meta-analysis, poor sleep quality was associated with an increased risk for disease activity without any clue of a potential causal relationship 17 .In the present study, we also observed a high prevalence of sleep disturbances (sleep subscore >5) and severe sleep disturbances (emotion subscore >7) in 47.0% and 30.9% of patients, respectively.Patients with active IBD had more sleep disturbances than those without active IBD (56.0% vs. 38.6%,p < 0.001).

ACCEPTED MANUSCRIPT / CLEAN COPY
Many patients with IBD may experience symptoms of common mental disorders such as anxiety or depression 18,19 .Such an association may be either fortuitous or induced by systemic inflammation, chronicity of symptoms and reduced social functioning.In a recent meta-analysis, the pooled prevalence of anxiety and depression in patients with IBD was 32.1% [28.3%-36.0%]and 25.2% [22.0%-285.5%],respectively, without any difference between CD and UC and with a higher prevalence in women.In the present study, emotional disturbances, including anxiety and depression, were observed in 39.7% of patients with IBD (emotion subscore >5).
The association of fatigue, sleep disorders and anxiety and depression has been previously reported in the general population as well as in patients with IBD 15,20,21 .Indeed, sleep disturbances have often been linked to most mental health conditions, including anxiety and depression 15 .In a large randomized controlled trial including 3755 patients, improving sleep using digital cognitive behavioral therapy was associated with the improvement of psychiatric symptoms 22 .However, this triple association is not easy to manage and is much more a multidirectional association between fatigue, sleep and emotional disturbances.Although not specific to IBD, those symptoms should be prospectively evaluated to lower the IBD-related burden irrespective of the specific management of IBD.
In the present study, the presence of IBD-related sick leave was associated with a higher rate of fatigue.Similarly, active workers and students had lower rates of fatigue than unemployed workers, remote workers and retired patients, but only in the univariate ACCEPTED MANUSCRIPT / CLEAN COPY analysis.Many studies have evaluated the impact of IBD on work productivity.Recently, a pan-European study estimated that the mean annual cost of absenteeism ranges from €1253 to €7915 23 .Patients in clinical remission had a lower annual indirect cost than those with active IBD.The authors suggest that the cost of absenteeism should be prospectively screened as an indirect marker of IBD-related disability.
Our study has several limitations.First, this cross-sectional study was not designed to specifically study fatigue and its determinants but to more broadly study IBD-related disability.Thus, we did not use a specific questionnaire and indices of fatigue but a 0-10 VAS included in the IBD-disk questionnaire.Second, being an anonymous study, our dataset did not include an objective assessment of systemic inflammation, such as C-reactive protein or fecal calprotectin, or nutritional status, such as hemoglobin or albumin.It is conceivable that malnutrition, anaemia and systemic inflammation may have an impact on perceived fatigue.
Third, such a cross-sectional study provides limited evidence concerning changes over time and the opportunity to lower fatigue through the management of IBD.As such, it is conceivable that we underestimate the impact of clinical symptoms on the emergence of IBD-related fatigue.Last, looking at outpatients, we did not ask specifically for current use of prednisone and/or prednisolone in the treatment list.However, very few patients reported taking steroids as "other treatment".The absence of a specific box for steroids in our questionnaire may have underestimate the proportion of patients with ongoing steroids.In conclusion, fatigue is a highly common symptom in patients with IBD that contributes to disability even in patients who report being in clinical remission.We

Conclusion:
Determinants of fatigue are not restricted to IBD-related factors but also include social factors, sleep, and emotional disturbances, thus supporting a holistic approach to IBD patient care.Keywords: Crohn's disease; ulcerative colitis; inflammatory bowel disease; disability; fatigue.

TABLE LEGENDS Table 1 :
Demographic, disease and medication characteristics of the study population

Table 2 :
Pearson's correlation coefficient R² assessing the convergent validity of Fatigue and other IBD-disk subscores evaluating other dimensions of IBD-related disability

Table 3 :
Proportion of patients with IBD-disk subscore >5 according to the presence of fatigue and severe fatigue

Table 4 :
Predictors of fatigue in our study population of 1704 patients

Table 5 :
Predictors of severe fatigue in our study population of 1704 patients ACCEPTED MANUSCRIPT / CLEAN COPY