Association of Childhood-Onset Inflammatory Bowel Disease With Risk of Psychiatric Disorders and Suicide Attempt

Key Points

Questions  Are patients with childhood-onset inflammatory bowel disease at increased risk of psychiatric disorders and suicide attempt, and is familial confounding important in this association?

Findings  In this population-based cohort study, childhood-onset inflammatory bowel disease was associated with increased risk of any psychiatric disorder and suicide attempt, as well as several specific psychiatric disorders. The association was observed when patients with inflammatory bowel disease were compared with the general population and with siblings without inflammatory bowel disease.

Meaning  Long-term psychological support should be considered for patients with childhood-onset inflammatory bowel disease.

Importance  Inflammatory bowel disease (IBD) has been associated with psychiatric morbidity in adults, although previous studies have not accounted for familial confounding. In children, IBD has an even more severe course, but the association between childhood-onset IBD and psychiatric morbidity remains unclear.

Objective  To examine the risk of psychiatric morbidity in individuals with childhood-onset IBD, controlling for potential confounding shared between siblings.

Design, Setting, and Participants  A population-based cohort study was conducted using data from the Swedish national health care and population registers of all children younger than 18 years born from 1973 to 2013. The study included 6464 individuals with a diagnosis of childhood-onset IBD (3228 with ulcerative colitis, 2536 with Crohn disease, and 700 with IBD unclassified) who were compared with 323 200 matched reference individuals from the general population and 6999 siblings of patients with IBD. Cox proportional hazards regression was used to estimate hazard ratios (HRs) with 95% CIs. Statistical analysis was performed from January 1, 1973, to December 1, 2013.

Main Outcomes and Measures  The primary outcome was any psychiatric disorder and suicide attempt. Secondary outcomes were the following specific psychiatric disorders: psychotic, mood, anxiety, eating, personality, and behavioral disorders; substance misuse; attention-deficit/hyperactivity disorder; autism spectrum disorders; and intellectual disability.

Results  The study included 6464 individuals with a diagnosis of childhood-onset IBD (2831 girls and 3633 boys; mean [SD] age at diagnosis of IBD, 13 [4] years). During a median follow-up time of 9 years, 1117 individuals with IBD (17.3%) received a diagnosis of any psychiatric disorder (incidence rate, 17.1 per 1000 person-years), compared with 38 044 of 323 200 individuals (11.8%) in the general population (incidence rate, 11.2 per 1000 person-years), corresponding to an HR of 1.6 (95% CI, 1.5-1.7), equaling 1 extra case of any psychiatric disorder per 170 person-years. Inflammatory bowel disease was significantly associated with suicide attempt (HR, 1.4; 95% CI, 1.2-1.7) as well as mood disorders (HR, 1.6; 95% CI, 1.4-1.7), anxiety disorders (HR, 1.9; 95% CI, 1.7-2.0) eating disorders (HR, 1.6; 95% CI, 1.3-2.0), personality disorders (HR, 1.4; 95% CI, 1.1-1.8), attention-deficit/hyperactivity disorder (HR, 1.2; 95% CI, 1.1-1.4), and autism spectrum disorders (HR, 1.4; 95% CI, 1.1-1.7) Results were similar for boys and girls. Hazard ratios for any psychiatric disorder were highest in the first year of follow-up but remained statistically significant after more than 5 years. Psychiatric disorders were particularly common for patients with very early-onset IBD (<6 years) and for patients with a parental psychiatric history. Results were largely confirmed by sibling comparison, with similar estimates noted for any psychiatric disorder (HR, 1.6; 95% CI, 1.5-1.8) and suicide attempt (HR, 1.7; 95% CI, 1.2-2.3).

Conclusions and Relevance  Overall, childhood-onset IBD was associated with psychiatric morbidity, confirmed by between-sibling results. Particularly concerning is the increased risk of suicide attempt, suggesting that long-term psychological support be considered for patients with childhood-onset IBD.

Inflammatory bowel disease (IBD) is a group of chronic inflammatory conditions of the gut; Crohn disease (CD) and ulcerative colitis (UC) are the major subtypes. Patients with IBD often experience symptoms such as diarrhea, abdominal pain accompanied by weight loss, and fatigue.

The challenges of living with arduous symptoms, together with the relapsing-remitting character of the disease and the need for demanding treatment,¹ have generated considerable interest in the mental health of patients with IBD. Among the potential mechanisms for the observed association between IBD and psychiatric disorders are psychological stress, abdominal pain, surgical procedure, disability,² use of corticosteroids,³ malabsorption, and vitamin and mineral deficiency,⁴ followed by chronic inflammation and its association with the function of the central nervous system.⁵ For adults, IBD has been associated with psychiatric morbidity, such as suicide,^6-10 mood disorders,^11-13 and psychotic disorders¹⁴ (eTable 1 in the Supplement). Childhood-onset IBD tends to have an even more severe disease course and longer duration than IBD diagnosed in adulthood.^15-18 However, to our knowledge, there are only 2 studies on the risk of psychiatric disorders in individuals with childhood-onset IBD. Both studies evaluated only the outcome before 18 years of age. The first study, by Loftus et al,¹⁹ explored the US MarketScan medical claims database and found that patients with childhood-onset CD were at increased risk of both anxiety disorders (hazard ratio [HR], 2.28; 95% CI, 1.65-3.17) and depression (HR, 1.74; 95% CI, 1.35-2.25). However, it is unclear whether patients in a medical claims database are representative of the average patient with IBD, thereby threatening the external validity of the study. The second study, based on 248 patients with IBD and 992 controls, examined only the risk of depression (using antidepressant medication as a proxy for diagnosis) and had a median follow-up of only 3.1 years, during which only 8 patients with IBD received antidepressants (3.2% of patients with IBD vs 1.2% of controls).²⁰ Previous studies have so far failed to take into account confounding from familial factors. This is an important limitation because both IBD and psychiatric disorders are associated with familial factors.^21,22 Thus, any causal claims behind the association may be questioned.

We undertook a nationwide cohort study of individuals with childhood-onset IBD to examine the risk of any psychiatric disorder, suicide attempt, and specific categories of psychiatric disorders until, and beyond, 18 years of age. We hypothesized that IBD diagnosed in childhood would increase the risk of developing psychiatric disorders. A sibling comparison design was used to control for shared familial confounders.

This population-based cohort study encompasses all children born in Sweden between 1973 and 2013. We used data from the Swedish registers to calculate the risks of developing psychiatric morbidity among individuals with IBD who received a diagnosis when they were younger than 18 years compared with matched reference individuals from the general population. In a separate analysis, we compared patients with IBD with their siblings without IBD. This study was approved by the Regional Ethical Review Board in Stockholm (2013/862-31/5). We obtained anonymized data from Statistics Sweden. Because this was a register-based study, no patient or individual from the general population was contacted.²³

Swedish pediatric health care²⁴ is publicly funded, with universal access to both primary and nonprimary health care. All residents in Sweden receive a unique personal identity number²⁵ that allows for extensive linkages between administrative and health registers. Through the Total Population Register,²⁶ all migrations, births, and deaths are tracked, enabling researchers to follow up with all residents in Sweden. The Patient Register started in 1964 and became nationwide in 1987.²⁷ Psychiatric diagnoses have been entered into the register since 1973. Although initially limited to inpatient care, the Patient Register contains data on specialized outpatient care since 2001.

From the Patient Register, we identified individuals with at least 2 International Classification of Diseases (ICD) codes representing IBD (eTable 2 in the Supplement). Previous studies^17,28 have used 2 or more diagnostic IBD codes because this approach had a high positive predictive value (93%) in a recent validation study.²⁹ For the present study, patients were defined as having CD when the first 2 IBD diagnoses were consistent with CD, and they were defined as having UC when the first 2 diagnoses consisted of UC codes. Patients with a code of K53.2 (indeterminate colitis in the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision [ICD-10]) and with a combination of UC and CD codes were classified as having IBD-unclassified. Childhood-onset IBD was defined as having a first IBD diagnosis before 18 years of age.

Each patient with IBD was matched with 50 reference individuals from the Swedish Population Register.²⁶ Reference individuals were required to be alive and to have lived in Sweden at the time of the second IBD diagnosis for patients with IBD. Matching variables consisted of sex, birth year, and county of birth. Individuals with IBD and their reference individuals with a diagnosis of any psychiatric disorder and those who emigrated from Sweden prior to cohort entry (date of the first IBD diagnosis for patients with IBD and the corresponding date for matched individuals) were excluded from the analyses (Figure 1). The risk of psychiatric outcomes was estimated for IBD as 1 group and among individuals with specific IBD diagnoses (CD, UC, and IBD-unclassified) to identify potential differences in psychiatric comorbidity previously described in adult patients.^7,14

Siblings of IBD probands were identified through the Multi-generation Register. It contains all individuals born in 1932 or later and who have been registered as Swedish residents since 1961.³⁰ Siblings of patients with IBD were required to be born in Sweden in 1973 or later and have no diagnosis of IBD.

Psychiatric disorders were identified from the National Patient Register²⁶ and the clinical databases for Child and Adolescent Mental Health Services³¹ and habilitation³² in Stockholm county and ascertained using relevant ICD codes; 1 code was required for a positive event (eTable 3 in the Supplement). Our primary outcome measures were any psychiatric disorder and suicide attempt. We explicitly examined the following disorders as secondary outcomes: psychotic, mood, anxiety, and eating disorders; substance misuse; personality and behavioral disorders; attention-deficit/hyperactivity disorder (ADHD); autism spectrum disorders (ASD); and intellectual disability. Dispensed prescriptions with ADHD medications (Anatomical Therapeutic Chemical Classification System codes N06BA01, N06BA02, N06BA04, and N06BA09) from the Drug Register were also used to identify individuals with ADHD.³³

Statistical analysis was performed from January 1, 1973, to December 31, 2013. The follow-up period began on the first visit of an individual with IBD (and corresponding date in the reference individuals) and ended with the first psychiatric diagnosis of interest (separate censoring date for each outcome), emigration, death, or December 31, 2013. Hazard ratios with 95% CIs were estimated from Cox proportional hazard regression models stratified on matched sets to account for the matching by sex, birth year, and county of birth. A design with sibling analyses was used to adjust for potential unmeasured familial confounding. The risk of psychiatric disorders among patients with IBD was estimated in relation to their siblings without IBD, with adjustment for sex and year of birth. A robust sandwich variance estimator was used to accommodate for the dependence in the data. In specific analyses, we stratified for sex, age at IBD diagnosis, version of ICD classification at the time of diagnosis (ICD-8, ICD-9, or ICD-10), year of diagnosis, presence of complications, surgery, and parental psychiatric history (eTables 4 and 5 in the Supplement). Although the Patient Register contains data on specialized outpatient care since 2001, the year 2001 contains a mix of truly incident IBD cases and prevalent IBD cases detected only with the introduction of outpatient care records. Hence, we chose to analyze patients diagnosed in 2002 and later separately. Analyses were performed with SAS software, version 9.3 (SAS Institute Inc).

We identified 6464 patients with a diagnosis of childhood-onset IBD (3228 with UC [49.9%], 2536 with CD [39.2%], and 700 with IBD-unclassified [10.8%]) (Table 1). The median age at IBD diagnosis was 14 years (interquartile range [IQR], 11-16 years). Patients were followed up for a median of 9 years (IQR, 4-15 years); the median age at the end of follow-up was 23 years (IQR, 18-29 years).

Individuals with childhood-onset IBD were followed up for a total of 65 503 person-years, during which 1117 individuals received a diagnosis of any psychiatric disorder (incidence rate [IR], 17.1 per 1000 person-years; HR, 1.6; 95% CI, 1.5-1.7) (Table 2). Meanwhile, 117 individuals with IBD attempted suicide within 71 147 person-years of observation (IR, 1.6 per 1000 person-years; HR, 1.4; 95% CI, 1.2-1.7). A total of 38 044 reference individuals received a diagnosis of a first psychiatric disorder during 3 396 346 person-years (IR, 11.2 per 1000 person-years), and 4335 reference individuals attempted suicide during 3 570 412 person-years of follow-up (IR, 1.2 per 1000 person-years). Hence, there was 1 extra case of any psychiatric disorder per 170 person-years. Inflammatory bowel disease was associated with mood disorders (HR, 1.6; 95% CI, 1.4-1.7), anxiety disorders (HR, 1.9; 95% CI, 1.7-2.0), eating disorders (HR, 1.6; 95% CI, 1.3-2.0), personality disorders (HR, 1.4; 95% CI, 1.1-1.8), ADHD (HR, 1.2; 95% CI, 1.1-1.4), and ASD (HR, 1.4; 95% CI, 1.1-1.7) (Figure 2).

The sibling comparison confirmed the observed associations, with similar estimates noted for any psychiatric disorder (HR, 1.6; 95% CI, 1.5-1.8) and suicide attempt (HR, 1.7; 95% CI, 1.2-2.3). Patients with IBD were also at statistically significantly higher risk for mood, anxiety, eating, and personality disorders and substance misuse compared with siblings without IBD (Table 2 and Figure 2).

The risks of any psychiatric disorder and suicide attempt were increased in all IBD subgroups, although the estimate was not statistically significant for suicide attempt in the UC subgroup (Table 3). When we restricted follow-up to 18 years of age, childhood IBD was associated with a 2.1-fold (95% CI, 2.0-2.3) increased risk of any psychiatric disorder (eTable 6 in the Supplement). The observed excess risk was lower but still significant (HR, 1.3; 95% CI, 1.2-1.4) during adulthood. An increased risk of suicide attempt was not observed until adulthood. Increased risks within the first 18 years of age were seen for mood disorders, anxiety disorders, eating disorders, ASD, and intellectual disability (eTable 6 in the Supplement), with similar patterns observed for specific IBD subtypes (eTable 7 in the Supplement).

The risk of any psychiatric disorder was tripled in the first year after IBD diagnosis (HR, 3.5; 95% CI, 3.0-4.0) and remained statistically significant after 5 or more years of follow-up (HR, 1.3; 95% CI, 1.2-1.5). A minor increase in risk for suicide attempt was observed throughout the whole follow-up period, but the estimates were statistically significant only 5 or more years after IBD diagnosis (HR, 1.5; 95% CI, 1.2-1.9) (eTable 8 in the Supplement).

Stratified analyses revealed a particularly increased risk of any psychiatric disorder among patients with very early onset IBD (<6 years of age; HR, 2.4; 95% CI, 1.9-3.1), as well as associations with extraintestinal manifestations (HR, 2.0; 95% CI, 1.7-2.4), bowel surgery (HR, 1.9; 95% CI, 1.6-2.2), perianal surgery (HR, 2.0; 95% CI, 1.7-2.4), and parental psychiatric history (HR, 3.4; 95% CI, 1.7-6.7) in comparison with reference individuals (eTable 9 in the Supplement). There was no essential difference in risk estimates according to sex, ICD classification, or source of data. The risk of suicide attempt was especially elevated among patients with surgical intervention. A similar pattern of results was obtained in sensitivity analyses in which the follow-up period began on the second visit with IBD diagnosis.

To our knowledge, this is the first large population-based study of psychiatric disorders in individuals with childhood-onset IBD with follow-up through adulthood and the use of a sibling comparison. We found that individuals with childhood-onset IBD were at increased risk of any psychiatric disorder and suicide attempt, including several categories of psychiatric disorders. Most of this association remained significant when comparing individuals with IBD with their siblings without IBD, thus accounting for environmental and genetic factors shared within siblings.

Our findings confirm and further support the results of 2 previous original studies on childhood-onset IBD (eTable 1 in the Supplement). The increased risk of mood and anxiety disorders seen in our study is consistent with findings from both America¹⁹ and Finland,²⁰ as well as a recent meta-analysis by Stapersma et al.³⁴ However, none of those studies included IBD-unclassified, had a follow-up longer than 5 years, or had a follow-up in adulthood. Those studies were unable to consider unmeasured familial confounding owing to a lack of genetically informative data.

In adult IBD, high rates of depressive and anxiety symptoms have previously been described.¹² An increased risk of psychiatric diagnoses associated with IBD has been consistently reported for mood disorders,^11,13,14 with ambiguous findings for anxiety disorders.^11,13 A study by Bhandari et al³⁵ recently found that patients with IBD with depressive symptoms reported suicidal ideation more often than did other patients with depressive symptoms (27% vs 12%), but this increase was not statically significant (P = .08). To our knowledge, the risk of suicide attempt has not been studied among patients with IBD, although some studies on mortality among patients with IBD have reported data on completed suicide,^7,9,10,36,37 with 1 study demonstrating an association between IBD and suicide.⁷ However, meta-analyses have failed to confirm these findings,^8,38 mostly because of insufficient statistical power. The Danish study by Gradus et al⁶ reported an association between completed suicide and both UC and CD in adults. We chose not to examine the risks for completed suicide owing to the low power for that measure (there were 14 suicide deaths in our cohort). However, the observed increased risk of suicide attempt in our study underlines the severity of psychiatric disorders in patients with IBD.

Studies on an association between eating disorders and IBD are scarce³⁹ and mostly limited to case reports or general research on autoimmune disease in patients with eating disorders.^40,41 Patients with IBD may experience abdominal pain and defecation with food intake, making them more likely to exclude certain food items from their diet. Therefore, various aspects of food intake may be associated with the symptoms of eating disorders in patients with IBD.

A Taiwanese study⁴² reported that adults with ADHD have a 2-fold increased risk for UC, but a previous study by Virta and Kolho²⁰ on 248 children with IBD failed to find a statistically significant association between ADHD and childhood-onset IBD. Our study, using a much larger sample, found a 1.2-fold increased risk for ADHD with IBD. It has been speculated that some patients with ASD have alterations of the intestinal barrier⁴³ and that this alteration might serve as a link to IBD. We found a 1.4-fold increased risk for ASD among individuals with IBD. This result is supported by a previous report by Kohane et al⁴⁴ on a possible association between IBD and ASD, even though no supportive statistical measures were reported. In contrast with the previous report, we presented data from subgroups of patients with IBD and calculated IRs, along with the relative risks in comparison with reference individuals from the general population.

Consistent with a recent report on adult IBD by Bernstein et al,¹³ the results of our study indicate no increased risk of psychotic disorders associated with IBD. Although psychotic disorders are rare among children, follow-up through adulthood enabled us to evaluate this outcome among individuals with childhood-onset IBD. In our study, the median age at the end of follow-up was 23 years (IQR, 18-29 years), which was older than the median age of 20.1 years (IQR, 18.3-22.3 years) at the first diagnosis of schizophrenia observed in the Swedish-born population.⁴⁵

The main strengths of this study are its population-based approach, the virtually complete follow-up of more than 9 years, the large numbers of participants, and the sibling comparisons. Sweden has a publicly funded health care system with equal access to health care regardless of social status.²⁴ Swedish research reported an even distribution of mental health problems among youths irrespective of parental educational level and occupation.⁴⁶ This finding is important because socioeconomic status is a factor associated with psychiatric disorders.⁴⁷ We used the Total Population Register²⁶ to select random reference individuals, matched by sex, county, and year of birth. This register also tracks all emigrations, allowing for a virtually complete follow-up and ensuring a close estimation of time at risk. The Patient Register, which was used to identify both IBD and psychiatric disorders, has been extensively validated,²⁷ with a positive predictive value of 85% to 95% for most diseases, including a positive predictive value of 93% for IBD.²⁹ For psychiatric disorders, the positive predictive value differs between diagnoses: 86% to 95% for schizophrenia^48,49 and 96% for ASD.³² Our large sample of patients with IBD and their follow-up until adulthood allowed us to calculate precise risk estimates but also enabled us to examine subgroups of psychiatric disorders that are otherwise uncommon in childhood.

To minimize intrafamilial confounding, we carried out a sibling comparison. Potential confounding is an issue in all observational research, and many factors signaling vulnerability are not included in traditional large-scale registers and are difficult to measure even if using small-scale, well-designed questionnaires and interviews. Siblings share early environmental and important genetic factors. Sibling analyses, therefore, allow researchers to better separate an association between IBD and psychiatric disorders than do studies exclusively using the general population as reference individuals.

We also acknowledge several limitations. We did not have any direct measurements of IBD severity. Instead, we used IBD surgery and the presence of extraintestinal manifestations as proxies for severity. We also did not have data on endoscopy or laboratory measures. We cannot rule out that IBD severity is associated with psychiatric disorders. A reverse association cannot be ruled out for early-onset neurodevelopmental disorders. Even if they are diagnosed after onset of IBD, neurodevelopmental disorders may be present before the onset of IBD.

We found an overall increased risk of psychiatric disorders and suicide attempt among individuals with childhood-onset IBD. These associations remained when genetic and environmental factors shared by siblings were taken into account, suggesting an association between IBD and psychiatric disorders. The psychological aspects of IBD may be associated with the increased risk of psychiatric disorders. Increased mood and anxiety disorders¹² may come from perceived stress, abdominal pain, and fear of fecal incontinence but also from socioeconomic stressors such as an increased burden on education and working life.² Several psychological factors have been hypothesized to be associated with the development of eating disorders in individuals with IBD, such as dietary preoccupation, abdominal discomfort related to food intake, body concerns, and physical limitations due to use of a colostomy bag.³⁹ Similar psychological factors may trigger personality disorders. Biological mechanisms, including malnutrition, vitamin deficiencies,^4,50 use of corticosteroids,³ and immunoinflammatory deregulation,⁵¹ may be other factors in the cause of psychiatric comorbidity. The high-risk estimates of psychiatric disorders in the first year after diagnosis of IBD, when disease activity is highest, indicate that psychological stress and biological factors may play a role in the cause of psychiatric comorbidity, but we also acknowledge that patients with IBD are intensively surveilled at this period of life and that bias may also be associated with the high risk of psychiatric disorders. This high risk of psychiatric disorders was observed among individuals with very early onset IBD and those whose parents had a history of psychiatric disorders, suggesting that these groups may be particularly vulnerable.

To our knowledge, this is the first study on the risk of suicide attempt and psychiatric disorders among individuals with childhood-onset IBD that uses a sibling comparison approach. An increased risk of psychiatric morbidity highlights the importance of mental health surveillance among patients with IBD. By reducing familial confounding, we have been able to show that familial factors do not seem to explain the increased risk of psychiatric outcomes, suggesting an association between IBD and those psychiatric disorders.

The highest risk of anxiety and mood disorders during the first year after a diagnosis of IBD suggests the need for psychological support for these patients. Particularly concerning is the increased risk of suicide attempt. Long-term psychological support should therefore be considered for patients with childhood-onset IBD.

Accepted for Publication: May 20, 2019.

Corresponding Author: Agnieszka Butwicka, MD, PhD, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, 171 77 Stockholm, Sweden (agnieszka.butwicka@ki.se).

Published Online: August 19, 2019. doi:10.1001/jamapediatrics.2019.2662

Author Contributions: Dr Butwicka had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Butwicka, Olén, Lichtenstein, Frisén, Ludvigsson.

Acquisition, analysis, or interpretation of data: Butwicka, Olén, Larsson, Halfvarson, Almqvist, Lichtenstein, Serlachius, Ludvigsson.

Drafting of the manuscript: Butwicka, Ludvigsson.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Butwicka, Ludvigsson.

Obtained funding: Butwicka, Larsson, Almqvist, Lichtenstein, Frisén.

Administrative, technical, or material support: Larsson, Lichtenstein, Serlachius.

Supervision: Halfvarson, Ludvigsson.

Conflict of Interest Disclosures: Dr Olén reported being principal investigator for projects at Karolinska Institutet partly financed by investigator-initiated grants from Janssen and Pfizer and receiving fees for lectures and participation on advisory boards from Janssen, Ferring, and Takeda. Dr Larsson reported serving as a speaker for Eli Lilly and Shire and receiving research grants from Shire and personal fees from Evolan outside the submitted work. Dr Halfvarson reported receiving consulting and lecture fees from AbbVie, Celgene, Celltrion, Ferring, Hospira, Janssen, Medivir, MSD, Pfizer, Prometheus, RenapharmaVifor, Sandoz, Shire, Takeda, Tillotts Pharma, and Vifor Pharma and research grants from Janssen, MSD, and Takeda. Dr Almqvist reported receiving grants from the Swedish Research Council during the conduct of the study. Dr Ludvigsson reported coordinating a study on behalf of the Swedish IBD quality register that received funding from Janssen Corporation. No other disclosures were reported.

Funding/Support: Dr Butwicka was supported by grant 2017-00788 from the Swedish Research Council, grant 2018-0718 from the Stockholm County Council (clinical research appointment), Karolinska Institutet, Strategic Research Programme in Neuroscience (StratNeuro), and grant 2016-00254 from Fredrik O. Ingrid Thurings Stiftelse while working on this project. Dr Olén was supported by grants from the Swedish Medical Society, Mag-tarmfonden, the Jane and Dan Olsson Foundation, the Mjölkdroppen Foundation, the Bengt Ihre research fellowship in gastroenterology, and Karolinska Institutet Foundations. Financial support was also provided through the Regional Agreement on Medical Training and Clinical Research between Stockholm County Council and Karolinska Institutet (ALF), the Swedish Cancer Society, the Swedish Research Council through the Swedish Initiative for Research on Microdata in the Social and Medical Sciences (SIMSAM), framework grant 340-2013-5867, and the Swedish Foundation for Strategic Research.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Information: No additional data available owing to Swedish regulation.