Incidence, Outcomes, and Health Services Burden of Very Early Onset Inflammatory Bowel Disease

  • Eric I. Benchimol
    Correspondence
    Reprint requests Address requests for reprints to: Eric Benchimol, MD, PhD, FRCPC, Division of Gastroenterology, Hepatology and Nutrition, Children’s Hospital of Eastern Ontario, 401 Smyth Road, Ottawa, Ontario, K1H 8L1 Canada. fax: (613) 738-4854.
    Affiliations
    Children’s Hospital of Eastern Ontario Inflammatory Bowel Disease Centre, Division of Gastroenterology, Hepatology and Nutrition, Children’s Hospital of Eastern Ontario, Ottawa, Canada

    Department of Pediatrics, University of Ottawa, Ottawa, Canada

    Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Canada

    Institute for Clinical Evaluative Sciences, Toronto, Canada
    Search for articles by this author
  • David R. Mack
    Affiliations
    Children’s Hospital of Eastern Ontario Inflammatory Bowel Disease Centre, Division of Gastroenterology, Hepatology and Nutrition, Children’s Hospital of Eastern Ontario, Ottawa, Canada

    Department of Pediatrics, University of Ottawa, Ottawa, Canada
    Search for articles by this author
  • Geoffrey C. Nguyen
    Affiliations
    Institute for Clinical Evaluative Sciences, Toronto, Canada

    Department of Medicine, University of Toronto, Toronto, Canada

    Mount Sinai Centre for Inflammatory Bowel Disease, Toronto, Canada
    Search for articles by this author
  • Scott B. Snapper
    Affiliations
    Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Children’s Hospital Boston, Boston, Massachusetts

    Department of Medicine, Harvard University, Boston, Massachusetts
    Search for articles by this author
  • Wenbin Li
    Affiliations
    Institute for Clinical Evaluative Sciences, Toronto, Canada
    Search for articles by this author
  • Nassim Mojaverian
    Affiliations
    Institute for Clinical Evaluative Sciences, Toronto, Canada
    Search for articles by this author
  • Pauline Quach
    Affiliations
    Children’s Hospital of Eastern Ontario Inflammatory Bowel Disease Centre, Division of Gastroenterology, Hepatology and Nutrition, Children’s Hospital of Eastern Ontario, Ottawa, Canada

    Institute for Clinical Evaluative Sciences, Toronto, Canada
    Search for articles by this author
  • Aleixo M. Muise
    Affiliations
    Department of Paediatrics, University of Toronto, Toronto, Canada

    SickKids Inflammatory Bowel Disease Center, Division of Gastroenterology Hepatology and Nutrition, Cell Biology Program, Research Institute, Hospital for Sick Children, Toronto, Ontario, Canada
    Search for articles by this author

      Background & Aims

      The Paris pediatric modification of the Montreal classification defines very early onset inflammatory bowel disease (VEO-IBD) as a form of IBD distinct from that of older children. We compared the incidence and outcomes of VEO-IBD with those of IBD in older children.

      Methods

      We performed a population-based retrospective cohort study of all children diagnosed with IBD in Ontario, Canada, from 1994 through 2009. Trends in standardized incidence were calculated using Poisson regression. We compared outpatient and emergency department visits, hospitalizations, and surgeries among children diagnosed with IBD when they were younger than age 6, ages 6–9.9, and older than age 10 years. Multivariable models were adjusted for income and stratified by sex.

      Results

      The incidence of IBD increased from 9.4 per 100,000 children (95% confidence interval [CI], 8.2–10.8/100,000 children) in 1994 to 13.2 per 100,000 children (95% CI, 11.9–14.6/100,000 children) in 2009 (P < .0001). The incidence increased by 7.4% per year among children younger than 6 years old and 6–9.9 years old, and by 2.2% per year among children ≥10 years old. IBD-related outpatient visits were less frequent among children <6 years old than ≥10 years old (odds ratio for female patients, 0.67; 95% CI, 0.58–0.78; odds ratio for male patients, 0.86; 95% CI, 0.75–0.98). Hazard ratios [HRs] for hospitalization were lower for children <6 years old (female HR, 0.70; 95% CI, 0.56–0.87; male HR, 1.12; 95% CI, 0.94–1.33) than for older children. HRs for surgery among children <6 years old with Crohn’s disease were 0.35 for female patients (95% CI, 0.16–0.78) and 0.59 for male patients (95% CI, 0.34–0.99). HRs for children <6 years old with ulcerative colitis were 0.88 for female patients (95% CI, 0.47–1.63) and 0.42 for male patients (95% CI, 0.21–0.85). There was no difference in hospitalization or surgery rates among children 6–9.9 years old vs those ≥10 years old.

      Conclusions

      Based on a retrospective cohort study, the incidence of VEO-IBD increased from 1994 through 2009. Children diagnosed with IBD before they were 6 years old used fewer health services and had lower rates of surgery than children diagnosed when they were 10 years or older.

      Keywords

      Abbreviations used in this paper:

      CD (Crohn's disease), CI (confidence interval), ED (emergency department), HR (hazard ratio), IBD (inflammatory bowel disease), OHIP (Ontario Health Insurance Plan), VEO (very early onset), UC (ulcerative colitis)
      The incidence of childhood-onset inflammatory bowel disease (IBD) is increasing internationally,
      • Benchimol E.I.
      • Fortinsky K.J.
      • Gozdyra P.
      • et al.
      Epidemiology of pediatric inflammatory bowel disease: a systematic review of international trends.
      with the most striking increase in Ontario, Canada, described in children younger than 10 years of age.
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      A small number of unique genetic mutations
      • Kotlarz D.
      • Beier R.
      • Murugan D.
      • et al.
      Loss of interleukin-10 signaling and infantile inflammatory bowel disease: implications for diagnosis and therapy.
      • Muise A.M.
      • Xu W.
      • Guo C.H.
      • et al.
      NADPH oxidase complex and IBD candidate gene studies: identification of a rare variant in NCF2 that results in reduced binding to RAC2.
      • Moran C.J.
      • Walters T.D.
      • Guo C.H.
      • et al.
      IL-10R polymorphisms are associated with very-early-onset ulcerative colitis.
      • Glocker E.O.
      • Kotlarz D.
      • Boztug K.
      • et al.
      Inflammatory bowel disease and mutations affecting the interleukin-10 receptor.
      have been identified in children with a diagnosis of IBD at a very young age, but genome-wide association studies have not detected large differences between adult-onset and early onset disease.
      • Imielinski M.
      • Baldassano R.N.
      • Griffiths A.
      • et al.
      Common variants at five new loci associated with early-onset inflammatory bowel disease.
      • Kugathasan S.
      • Baldassano R.N.
      • Bradfield J.P.
      • et al.
      Loci on 20q13 and 21q22 are associated with pediatric-onset inflammatory bowel disease.
      However, the phenotype of children with onset of Crohn’s disease (CD) occurring younger than the age of 10 is predominantly colonic, with a lower risk of ileal disease.
      • Heyman M.B.
      • Kirschner B.S.
      • Gold B.D.
      • et al.
      Children with early-onset inflammatory bowel disease (IBD): analysis of a pediatric IBD consortium registry.
      • Markowitz J.
      • Kugathasan S.
      • Dubinsky M.
      • et al.
      Age of diagnosis influences serologic responses in children with Crohn's disease: a possible clue to etiology?.
      • Meinzer U.
      • Idestrom M.
      • Alberti C.
      • et al.
      Ileal involvement is age dependent in pediatric Crohn's disease.
      • Shaoul R.
      • Karban A.
      • Reif S.
      • Weiss B.
      • Shamir R.
      • Tamir A.
      • Davidovich O.
      • Halevi J.
      • Silver E.L.
      • Levine A.
      Disease behavior in children with Crohn's disease: the effect of disease duration, ethnicity, genotype, and phenotype.
      In children with earlier onset of ulcerative colitis (UC), the requirement for second-line therapy (biologics or colectomy) was reported to be lower than for those diagnosed in the second decade of life.
      • Levine A.
      • Griffiths A.
      • Markowitz J.
      • et al.
      Pediatric modification of the Montreal classification for inflammatory bowel disease: the Paris classification.
      These findings led to a change in the classification of pediatric patients with IBD. Although the Montreal classification of IBD previously denoted all children diagnosed at younger than 17 years as having “early-onset IBD,”
      • Silverberg M.S.
      • Satsangi J.
      • Ahmad T.
      • et al.
      Toward an integrated clinical, molecular and serological classification of inflammatory bowel disease: Report of a Working Party of the 2005 Montreal World Congress of Gastroenterology.
      the Paris modification of the Montreal classification suggested that patients with age at onset of younger than 10 years (A1a) be classified differently from those diagnosed at or older than 10 years of age (A1b).
      • Levine A.
      • Griffiths A.
      • Markowitz J.
      • et al.
      Pediatric modification of the Montreal classification for inflammatory bowel disease: the Paris classification.
      Patients diagnosed at younger than this 10-year age cut-off value have been deemed to have very early onset (VEO)-IBD, although some investigators also have suggested 6 years as the cut-off value.
      • Muise A.M.
      • Snapper S.B.
      • Kugathasan S.
      The age of gene discovery in very early onset inflammatory bowel disease.
      This new classification and the apparent growing burden of VEO-IBD led to the creation of the interNational Early Onset Paediatric IBD Cohort Study (NEOPICS), a large cohort study aimed at discovering the cause and cure of VEO-IBD.

      interNational Early Onset Paediatric IBD Cohort Study (NEOPICS). NEOPICS, 2013. Available at: http://neopics.org. Accessed May 20, 2014.

      In addition, a group of VEO-IBD experts met in Amsterdam in January 2013 to develop a research agenda for the field.

      Snapper SB, Klein C, Muise AM. Very early onset IBD: mapping the research agenda. Amsterdam. NEOPICS, 2013. Available at: http://neopics.org/uploads/VEOIBD_MEETING_SUMMARY.pdf. Accessed May 20, 2014.

      They identified the importance of assessing the burden and prognosis of VEO-IBD. Considering that most very young patients with IBD typically are treated in specialized IBD care centers by pediatric gastroenterologists, whereas older children and adolescents with IBD may be treated by a variety of specialist pediatric or adult physicians,
      • Benchimol E.I.
      • Guttmann A.
      • To T.
      • et al.
      Changes to surgical and hospitalization rates of pediatric inflammatory bowel disease in Ontario, Canada (1994-2007).
      any comparative study of prognosis and burden is best conducted with population-level cohorts.
      We aimed to describe the incidence and prevalence of pediatric IBD in Ontario, Canada, updating our previous report of the increasing incidence in children younger than age 10.
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      In addition, we used a population-based cohort to determine the health services burden and outcomes of children with IBD, comparing patients with VEO-IBD with patients with onset after the age of 10 years.

      Methods

       Study Design

      This study was approved by the research ethics boards of the Children’s Hospital of Eastern Ontario and The Ottawa Hospital, and the privacy officer of the Institute for Clinical Evaluative Sciences. We conducted a population-based, retrospective cohort study of all children <18 years diagnosed with IBD in Ontario, Canada. We included children diagnosed with IBD between fiscal years (April 1 to March 31) 1994 and 2009 identified from the Ontario Crohn’s and Colitis Cohort, a population-based registry of all IBD patients living in Ontario, Canada (population, 13.5 million), derived from health administrative data.
      To create our cohort, we drew from the health administrative data of all residents of Ontario, Canada, who were <18 years during the study period of fiscal years 1994–2009 who qualified for universal government health care (>99% of the population). Although the Montreal and Paris classifications use <17 years as the age cut-off age for pediatric-onset disease, we elected to use <18 years to reflect the age at which adolescents must be treated by adult providers in the Ontario health system. Ontario’s administrative databases are maintained by the Institute for Clinical Evaluative Sciences through a comprehensive data-sharing agreement with the Ontario Ministry of Health and Long Term Care. The Ontario Crohn’s and Colitis Cohort is a data-derived cohort created using validated algorithms of health care contacts to classify patients as having or not having IBD,
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      which is linked to other Ontario administrative databases by an encrypted patient identifier. This algorithm identified children with IBD within Ontario’s health administrative databases with the following diagnostic accuracies: sensitivity of 89.6%–91.1%, specificity of 99.5%–100%, and positive and negative predictive values of 59.2%–76.0% and 99.9%–100%, respectively.
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      In addition, a validated algorithm based on the diagnoses assigned to the last 7 outpatient visits distinguished CD from UC patients with a sensitivity of 95.1%, a specificity of 86.0%, a positive predictive value of 92.0%, and a negative predictive value of 91.2%.
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      The date of diagnosis was defined as the date of first contact with a physician or first hospitalization with an associated IBD-specific diagnostic code. Patients were excluded when their first IBD-specific health contact occurred at younger than 6 months of age (because of the possible misclassification of allergic enterocolitis), if date of birth or sex were unavailable, or they did not have continuous enrollment from diagnosis to the end of the follow-up period (death or March 31, 2012). We compared patients aged 6 months to <6 years at diagnosis with those aged 6.0 to <10 years (6–9.9 y) and ≥10 years at diagnosis.
      Annual sex- and age-adjusted annual incidence rates per 100,000 population were determined for 1994–2009, with corresponding 95% confidence intervals (CIs) based on gamma distribution.
      • Fay M.P.
      • Feuer E.J.
      Confidence intervals for directly standardized rates: a method based on the gamma distribution.
      We used the Canadian censuses from 1991, 1996, 2001, and 2006 to calculate the annual intercensal population estimates of children aged <18 years.

      Statistics Canada. Population and dwelling counts, for Canada, provinces, territories and census subdivisions (municipalities), 2006 census, 100% data (table). Population and dwelling counts. Statistics Canada, 2007. Available at: http://www12.statcan.ca/english/census06/data/popdwell/Table.cfm?T=101. Accessed May 20, 2014.

      Statistics Canada. Population and dwelling counts, for Canada, provinces, territories and census subdivisions (municipalities), 2001 and 1996 censuses, 100% data (table). Population and dwelling counts. Statistics Canada, 2002. Available at: http://www12.statcan.ca/english/census01/products/standard/popdwell/tables.cfm. Accessed May 20, 2014.

      We assessed age group differences in baseline demographic characteristics, including IBD type (CD, UC, or unclassifiable based on administrative data). We also determined the likelihood of a changed diagnosis from one IBD type to another by age group. We described the 10 most frequent reasons for outpatient physician visits by age group in the 2 years before the date of diagnosis with IBD. In anticipation of possible age group differential misclassification of diagnosis, or misdiagnosis with IBD, we calculated the rate of misclassification by assessing the rate of those without an IBD health system contact in the most recent 3 years. In addition, we determined the rates of concomitant diagnosis with immunodeficiency syndromes (International Classification of Diseases codes 271.0, 279.X, 288.1, 288.2, and E74.0), comparing the proportion between age groups.
      We determined health services utilization rates (outpatient physician visits, hospitalizations, and emergency department [ED] visits), as well as likelihood of hospitalization and ED visits at 1, 3, 5, and 10 years after diagnosis, comparing utilization in children within the 2 age groups of VEO-IBD with those who were diagnosed at ≥10 years. For outpatient physician visits and ED visits, we conducted subgroup analyses to assess overall utilization, and IBD-specific and IBD-related visits (code list is shown in Supplementary Table 1). We also determined surgical rates by age group, determining the likelihood of intra-abdominal surgery (intestinal resection or colectomy) in CD patients, and colectomy in UC patients (codes are listed in Supplementary Table 2).

       Data Sources

      To report on hospitalizations, we used discharge abstract data mandatorily collected from all hospitals and reported to the Canadian Institute for Health Information, with diagnostic codes known to be 92%–99% accurate.
      Surgical events also were derived from the Canadian Institute for Health Information discharge abstract database. Surgical codes were validated previously in this database for children with UC.
      • Ma C.
      • Crespin M.
      • Proulx M.C.
      • et al.
      Postoperative complications following colectomy for ulcerative colitis: a validation study.
      • Soon I.S.
      • Wrobel I.
      • deBruyn J.C.
      • et al.
      Postoperative complications following colectomy for ulcerative colitis in children.
      Data on outpatient physician visits and ED utilization were obtained from physician billing information obtained from the Ontario Health Insurance Plan (OHIP) database. The Registered Persons Database was used to obtain date of birth, sex, and postal code at diagnosis. To control for confounding by socioeconomic status, we linked postal codes to 1996, 2001, and 2006 Canadian census data using the Postal Code Conversion File Plus
      • Wilkins R.
      PCCF+ version 4E user's guide: automated geographic coding based on the Statistics Canada postal code conversion files, including postal codes to July 2004. Catalogue no. 93-387-XIE.
      to obtain the mean neighborhood income quintile at diagnosis of individual children with IBD at the enumeration area (1996 census) or dissemination area (2001 and 2006 censuses) levels.

       Statistical Analysis

      Comparisons of the likelihood of diagnosis change, misclassification rate, likelihood of immunodeficiency syndromes, and reasons for prediagnosis outpatient visits were performed between age groups using chi-square analysis or the Fisher exact test where appropriate. Trends in incidence over time were reported as percentage change per year from 1994 to 2009 by exponentiating the β value of the sex-adjusted Poisson regression and subtracting it by 1, as follows: exp(β) - 1. This Poisson model was also used to assess statistical significance in trends in incidence over time, stratified by age category with age-appropriate denominator populations. To assess whether temporal trends were different between age groups, we determined whether the interaction between the year of diagnosis and age category was significantly different.
      The likelihood of changed diagnosis by age group was analyzed using a 2-sample test for equality of proportions with continuity correction. The association between age group and the number of physician visits, hospitalizations, and ED visits was analyzed using Poisson regression models. The association between age group and the likelihood of hospitalization, surgery, or ED visit was analyzed using Cox proportional hazard multivariable survival models. The association between age group and the likelihood of hospitalization or surgery within 1, 3, 5, and 10 years after diagnosis was assessed using logistic regression analysis. All multivariable analyses included age group (<6 y at diagnosis vs 6–9.9 y at diagnosis vs ≥10 y at diagnosis) as the main predictor. The mean neighborhood income quintile and sex were included a priori because of the risk of confounding. There were significant interactions between age group and sex for almost all models. Therefore, multivariable regression analyses were stratified by sex. For goodness-of-fit testing, we used the -2 log likelihood ratio for Cox proportional hazard models, the Pearson chi-square method for Poisson regression models, and the c-statistic for logistic regression models. Where there was overdispersion, zero-inflation negative binomial Poisson regression models were used. All statistical analyses were performed using SAS version 9.3 (SAS Institute, Inc, Cary, NC).

      Results

       Descriptive Characteristics

      The characteristics of the cohort by age group are presented in Table 1. The diagnosis of CD was more likely in the ≥10-year group, and became less likely with younger age at diagnosis (P < .0001). Patients in the younger age groups were more likely to have a changed diagnosis from the initial IBD subtype diagnosis. Diagnosis changed in 62 of 383 (16.2%) of the patients in the <6 year group, 119 of 657 (18.1%) of the patients in the 6–9.9 y group, and 845 of 6103 (13.8%) of the patients in the ≥10 year group (P = .007). Among the 3 age groups (youngest to oldest), the likelihood of changing from UC to CD was 11.9% vs 13.1% vs 12.6%, respectively, of those initially diagnosed with UC (P = .93). The likelihood of changing from CD to UC was 1.6% vs 3.2% vs 3.3%, respectively, of those initially diagnosed with CD (P = .74). The likelihood of changing from IBD-U to UC was 16.7% vs 24.0% vs 26.8%, respectively (P = .33), among those initially diagnosed with IBD-U. Finally, the likelihood of changing from IBD-U to CD was 14.3% vs 26.7% vs 28.1%, respectively (P = .15).
      Table 1Demographic Characteristics of the Cohort
      <6 y at diagnosis (n = 383)6–9.9 y at diagnosis (n = 657)≥10 y at diagnosis (n = 6103)
      Mean age at diagnosis ± SD, y2.7 ± 1.97.90 ± 1.0714.4 ± 2.15
      Female, N (%)156 (40.7)287 (43.7)2719 (44.6)
      Length of follow-up period, mean ± SD, y9.2 ± 6.29.2 ± 5.89.9 ± 6.1
      First diagnostic classification
       Crohn's disease106 (27.7%)314 (47.8%)3518 (57.6%)
       Ulcerative colitis235 (61.4%)268 (40.8%)2126 (34.8%)
       Unclassifiable42 (11.0%)75 (11.4%)459 (7.5%)
      Most recent diagnostic classification
       Crohn's disease127 (33.2%)346 (52.7%)3720 (61.0%)
       Ulcerative colitis213 (55.6%)244 (37.1%)1987 (32.6%)
       Unclassifiable43 (11.2%)67 (10.2%)396 (6.5%)
      Mean neighborhood income quintile, N (%)
       1 (lowest)75 (19.6)79 (12.1)859 (14.2)
       276 (19.8)110 (16.8)1012 (16.7)
       378 (20.4)142 (21.7)1176 (19.4)
       474 (19.3)152 (23.2)1363 (22.5)
       5 (highest)80 (20.9)171 (26.1)1651 (27.2)
      Urban residence at diagnosis, N (%)345 (90.1)584 (88.9)5302 (87.0)
      The proportion of patients deemed to have been misclassified or misdiagnosed was low. In total, 36 patients (0.5% of the cohort) were misclassified, and had no IBD-specific outpatient physician visits within 3 years of their latest contact with the health system. There was no significant difference in the rate of misclassification in patients <6, 6–9.9, and ≥10 years (1.0% vs 0.9% vs 0.4%, respectively; P = .054). The proportion of patients with concomitant codes for immunodeficiency syndromes was low (8 of 7143 [0.1%] in all ages) patients of the full cohort and there was no difference by age group (0.3% vs 0.3% vs 0.1%, respectively; P = .13).

       Incidence of Pediatric IBD

      Temporal trends in incidence of pediatric IBD in children <18 years at diagnosis are shown in Figure 1. For all age groups, the standardized incidence increased from 9.4 (95% CI, 8.2–10.8) per 100,000 children in 1994, to 13.2 (95% CI, 11.9–14.6) per 100,000 children in 2009 (P < .0001 by Poisson regression). The incidence of CD increased from 5.2 (95% CI, 4.3–6.2) per 100,000 children in 1994 to 7.9 (95% CI, 6.9–9.0) per 100,000 children in 2009 (P < .0001) (Supplemental Figure 1). The incidence of UC increased from 3.9 (95% CI, 3.2–4.8) per 100,000 children in 1994 to a peak of 5.1 (95% CI, 4.3–6.0) per 100,000 children in 2007. It decreased again to 4.1 (95% CI, 3.4–4.9) per 100,000 children by 2009 Supplemental Figure 2. Overall, the change in UC incidence was statistically significant by Poisson regression (P < .0001). Trends in incidence by age groups are shown in Table 2. The rate of incidence increase per year for children <6 years was 7.4% for IBD (P = .002), 6.2% for CD (P = .51), and 5.0% for UC (P = .44). For children aged 6–9.9 years at diagnosis, incidence increased by 7.4% per year for overall IBD (P < .0001), 7.4% per year for CD (P < .0001), and 7.2% per year for UC (P < .0001). For children diagnosed at ≥10 years, the rate of incidence increase per year was 2.2% for IBD (P < .0001), 1.9% for CD (P < .0001), and 2.2% for UC (P = .0001). Compared with the reference ≥10-year group, the rate of IBD incidence increase was faster in the <6-year group (P = .003) and the 6–9.9-year group (P < .0001). For CD, the rate of increase was faster in the 6–9.9-year (P = .001), but not the <6-year group (P = .18), compared with the ≥10-year reference group. In UC patients, the rate of increase was faster in the 6–9.9-year group (P = .01) but not the <6-year group (P = .22).
      Figure thumbnail gr1
      Figure 1Standardized incidence of IBD from 1994 to 2009 by age group.
      Table 2Trends in Age- and Sex-Standardized Incidence of IBD From 1994 to 2009 by Age Group and IBD Subtype
      1994 incidence per 100,000 (95% CI)2002 incidence per 100,000 (95% CI)2009 incidence per 100,000 (95% CI)Average yearly change
      Annual trends were calculated using Poisson regression, adjusting for sex. The average yearly change was calculated using the following equation: exp(β) - 1.
      (P value)
      All patients <18 y
       IBD9.4 (8.2–10.8)8.1 (7.0–9.2)13.2 (11.9–14.6)+3.5% (<.0001)
       CD5.2 (4.3–6.2)5.2 (4.4–6.1)7.9 (6.9–9.0)+3.2% (<.0001)
       UC3.9 (3.2–4.8)2.4 (1.8–3.0)4.1 (3.4–4.9)+3.5% (<.0001)
      0–5 y
       IBD1.3 (0.7–2.4)0.6 (0.2–1.5)2.1 (1.2–3.4)+7.4% (.002)
       CD0.5 (0.1–1.2)No cases0.4 (0.08–1.1)+6.2% (.51)
       UC0.8 (0.3–1.7)0.6 (0.2–1.5)0.9 (0.3–1.8)+5.0% (.44)
      6–9 y
       IBD3.8 (2.4–5.7)3.8 (2.4–5.6)7.7 (5.6–10.3)+7.4% (<.0001)
       CD1.2 (0.5–2.5)2.2 (1.2–3.7)5.1 (3.5–7.3)+7.4% (.0001)
       UC2.6 (1.4–4.3)1.2 (0.5–2.5)2.2 (1.2–3.8)+7.2% (<.0001)
      10–17 y
       IBD17.1 (14.8–19.8)14.8 (12.8–17.1)23.4 (20.8–26.2)+2.2% (<.0001)
       CD10.1 (8.3–12.2)10.0 (8.3–11.8)14.3 (12.3–16.5)+1.9% (<.0001)
       UC6.4 (5.0–8.1)4.0 (3.0–5.3)7.2 (5.5–8.4)+2.2% (.0001)
      a Annual trends were calculated using Poisson regression, adjusting for sex. The average yearly change was calculated using the following equation: exp(β) - 1.

       Prediagnosis Health Care Contacts

      The 10 most common reasons for outpatient physician visits in the 2 years before diagnosis are shown in Supplementary Table 3. These were significantly different between patients in the different age groups (P < .0001). Overall, most children were noted by their physicians to have gastrointestinal-related symptoms leading up to IBD diagnosis. Children <6 years were significantly more likely to be diagnosed with diarrhea, gastroenteritis, or viral gastroenteritis (OHIP code 009), and rectal bleeding or other intestinal disorder (OHIP code 569). Children ≥10 years were more likely to be diagnosed with myalgias or arthralgias (OHIP code 781), and anemia or iron deficiency (OHIP code 280). These diagnoses did not appear in the 10 most frequent reasons for visits in children <10 years.

       Outpatient Physician Visits

      The number of IBD-specific and IBD-related outpatient visits per year was lower in patients diagnosed at <6 year, but not in those diagnosed at 6–9.9 years compared with those diagnosed ≥10 years (Table 3). When stratified by IBD subtype, this association was present in females <6 years with CD and UC, but not males. For females <6 years with CD, there was a lower rate of outpatient visits for IBD-specific (OR, 0.70; 95% CI, 0.50–0.98) and IBD-related reasons (OR, 0.74; 95% CI, 0.57–0.97). This was not present for males <6 years with CD for either IBD-specific (OR, 1.00; 95% CI, 0.81–1.24) or IBD-related reasons (OR, 1.00; 95% CI, 0.83–1.21). Similarly, for females <6 years with UC there were lower visit rates for IBD-specific (OR, 0.75; 95% CI, 0.63–0.94) and IBD-related reasons (OR, 0.75; 95% CI, 0.63–0.89). This association was not present for males <6 years with UC (IBD-specific: OR, 0.86; 95% CI, 0.71–1.05; IBD-related: OR, 0.89; 95% CI, 0.75–1.05). For both males and females 6–9.9 years, IBD-related visits were higher in CD patients, but not in UC patients. In CD patients aged 6–9.9 years, there were higher IBD-specific (females: OR, 1.29; 95% CI, 1.09–1.52; males: OR, 1.28; 95% CI, 1.16–1.41) and IBD-related reasons (females: OR, 1.21; 95% CI, 1.05–1.40; males: OR, 1.25; 95% CI, 1.14–1.37). There was no difference in UC patients aged 6–9.9 years compared with those aged ≥10 years at diagnosis (Supplementary Table 4).
      Table 3IBD-Specific and IBD-Related Outpatient Physician and Emergency Department Visit Rates Comparing Age Groups by Multivariable Poisson Regression
      Outpatient physician visitsEmergency department visits
      <6 y (n = 383)6–9.9 y (n = 657)≥10 y (n = 6103)<6 y (n = 383)6–9.9 y (n = 657)≥10 y (n = 6103)
      IBD-specific
       Number of visits per year, mean ± SD2.9 ± 2.93.8 ± 4.13.4 ± 3.50.06 ± 0.220.07 ± 0.220.12 ± 0.36
       Females: adjusted OR (95% CI)0.66 (0.56–0.79)

      P < .0001
      1.06 (0.94–1.20)

      P = .35
      Ref0.24 (0.16–0.37)

      P < .0001
      0.66 (0.46–0.93)

      P = .02
      Ref
       Males: adjusted OR (95% CI)0.84 (0.72–0.97)

      P = .02
      1.14 (1.04–1.25)

      P = .004
      Ref0.50 (0.31–0.79)

      P = .003
      0.55 (0.43–0.71)

      P < .0001
      Ref
      IBD-related
       Number of visits per year, mean ± SD3.5 ± 3.104.3 ± 4.544.2 ± 3.800.22 ± 0.560.19 ± 0.450.34 ± 0.78
       Females: adjusted OR (95% CI)0.67 (0.58–0.78)

      P < .0001
      1.02 (0.91–1.13)

      P = .78
      Ref0.33 (0.25–0.44)

      P < .0001
      0.60 (0.47–0.77)

      P < .0001
      Ref
       Males: adjusted OR (95% CI)0.86 (0.75–0.98)

      P = .02
      1.11 (1.02–1.20)

      P = .01
      Ref0.61 (0.45–0.84)

      P = .002
      0.53 (0.43–0.65)

      P < .0001
      Ref

       Hospitalization Rates

      The likelihood of hospitalization within 1, 3, 5, and 10 years after diagnosis for patients diagnosed ≥10 years was 27.4%, 43.2%, 53.5%, and 67.3%, respectively. In patients <6 years, the likelihood was 34.0%, 44.7%, 50.2%, and 54.4%, respectively. In patients 6–9.9 years, the likelihood was 23.6%, 40.1%, 49.1%, and 63.4%, respectively. Comparisons of the likelihood of hospitalization within 1, 3, 5, and 10 years by IBD subtype and sex are shown in Supplementary Table 5. The hazard of hospitalization was lower in younger females with IBD (<6 y: hazard ratio [HR], 0.70; 95% CI, 0.56–0.87; P = .001; 6–9.9 y: HR, 0.81; 95% CI, 0.69–0.95; P = .008) compared with those ≥10 years. There was no difference in males with VEO-IBD (<6 y: HR, 1.12; 95% CI, 0.94–1.33; P = .21; 6–9.9 y: HR, 0.93; 95% CI, 0.81–1.08; P = .34) compared with those with onset ≥10 years. In patients with CD, the hazard of hospitalization was lower for females diagnosed at <6 years (HR, 0.56; 95% CI, 0.37–0.85; P = .006), but not at 6–9.9 years (HR, 0.85; 95% CI, 0.69–1.06; P = .15). There was no difference in males diagnosed at <6 years (HR, 1.07; 95% CI, 0.81–1.41; P = .65) or at 6–9.9 years (HR, 1.02; 95% CI, 0.85–1.22; P = .86), compared with males diagnosed at ≥10 years. In patients with UC, the hazard of hospitalization was not different in patients diagnosed at <6 years (females: HR, 0.89; 95% CI, 0.67–1.18; P = .42; males: HR, 1.24; 95% CI, 0.97–1.58; P = .08) or at 6–9.9 years (females: HR, 0.78; 95% CI, 0.60–1.01; P = .055; males: HR, 0.88; 95% CI, 0.68–1.15; P = .35).

       Emergency Department Utilization

      The likelihood of visiting an ED within 1, 3, 5, and 10 years after diagnosis for patients diagnosed at ≥10 years was 20.3%, 35.5%, 46.5%, and 62.8%, respectively. The likelihood in the <6-year group was 15.6%, 25.1%, 28.3%, and 41.9% at 1, 3, 5, and 10 years, respectively. The likelihood in the 6–9.9-year group was 13.5%, 23.9%, 29.7%, and 46.6% at 1, 3, 5, and 10 years, respectively. The numbers of IBD-specific and IBD-related ED visits were significantly lower in both the <6-year and 6–9.9-year age groups, compared with those ≥10 years at diagnosis (Table 3). The overall adjusted hazard of IBD-related ED visits was lower in females with CD diagnosed at <6 years compared with those diagnosed at ≥10 years (females: HR, 0.53; 95% CI, 0.34–0.82; P = .004; males: HR, 1.03; 95% CI, 0.77–1.37; P = .85). Similarly, hazard of IBD-related ED visits was lower in female CD children diagnosed at 6–9.9 years compared with those diagnosed at ≥10 years (females: HR, 0.61; 95% CI, 0.47–0.77; P < .0001; males: HR, 0.85; 95% CI, 0.70–1.04; P = .11). By contrast, in UC patients the hazard of IBD-related ED visits was lower in both young females and young males. This was true for children diagnosed at <6 years (females: HR, 0.68; 95% CI, 0.50–0.92; P = .01; males: HR, 0.57; 95% CI, 0.42–0.78; P = .0004) and at 6–9.9 years (females: HR, 0.58; 95% CI, 0.44–0.77; P = .0002; males: HR, 0.63; 95% CI, 0.46–0.85; P = .003), compared with those diagnosed at ≥10 years.

       Surgical Rates

      For patients with CD, the likelihood of intestinal resection and intra-abdominal surgery is shown in Table 4, including sex-stratified odds ratios. Overall, the hazard of undergoing surgery for CD was lower in patients diagnosed at <6 years compared with those diagnosed at ≥10 years (females: HR, 0.35; 95% CI, 0.16–0.78; males: HR, 0.58; 95% CI, 0.34–0.99). For patients age 6–9.9 years at diagnosis, the hazard of surgery was not significantly lower (females: HR, 0.73; 95% CI, 0.51–1.06; males: HR, 0.89; 95% CI, 0.67–1.17). The Kaplan–Meier survival curves for these comparisons are shown in Figure 2. Very few CD patients underwent repeated (≥2) intestinal resections in our cohort, but the likelihood of requiring a repeat surgery was lower in children diagnosed at age <10 years compared with those diagnosed at age ≥10 years (5 of 473 [1.1%] vs 98 of 3720 [2.6%]; P = .039). The hazard of CD patients requiring perianal surgery was not different in the <10-year age group compared with the ≥10-year age group (HR, 0.78; 95% CI, 0.46–1.31).
      Table 4Likelihood and Odds of Surgery Comparing Age Groups by Multivariable Logistic Regression
      Crohn’s disease: risk of intestinal resectionUlcerative colitis: risk of colectomy
      <6 y (n = 127)6–9.9 y (n = 346)≥10 y (n = 3720)<6 y (n = 213)6–9.9 y (n = 244)≥10 y (n = 1987)
      Within 1 y of diagnosis7/119 (5.9%)4/332 (1.2%)235/3568 (6.6%)7/205 (3.4%)12/238 (5.0%)95/1933 (4.9%)
       Females: adjusted OR (95% CI)1.08 (0.38–3.08)

      P = .89
      0.26 (0.08–0.83)

      P = .02
      Ref1.20 (0.46–3.12)

      P = .71
      1.37 (0.63–3.00)

      P = .42
      Ref
       Males: adjusted OR (95% CI)0.76 (0.23–2.44)

      P = .64
      0.09 (0.01–0.63)

      P = .02
      Ref0.32 (0.08–1.35)

      P = .12
      0.66 (0.23–1.86)

      P = .43
      Ref
      Within 3 y of diagnosis10/104 (9.6%)14/272 (5.1%)466/3167 (14.7%)13/179 (7.3%)16/217 (7.4%)172/1695 (10.1%)
       Females: adjusted OR

      (95% CI)
      0.71 (0.28–1.85)

      P = .49
      0.37 (0.17–0.81)

      P = .01
      Ref1.21 (0.58–2.51)

      P = .62
      0.99 (0.51–1.93)

      P = .98
      Ref
       Males: adjusted OR

      (95% CI)
      0.54 (0.21–1.35)

      P = .19
      0.28 (0.13–0.60)

      P = .001
      Ref0.36 (0.13–1.01)

      P = .05
      0.42 (0.17–1.06)

      P = .07
      Ref
      Within 5 y of diagnosis10/86 (11.6%)36/240 (15.0%)611/2744 (22.3%)11/148 (7.4%)20/186 (10.8%)191/1429 (13.4%)
       Females: adjusted OR

      (95% CI)
      0.44 (0.15–1.27)

      P = .13
      0.55 (0.31–0.98)

      P = .04
      Ref1.00 (0.46–2.16)

      P = .99
      1.18 (0.64–2.18)

      P = .59
      Ref
       Males: adjusted OR

      (95% CI)
      0.47 (0.20–1.11)

      P = .08
      0.67 (0.42–1.08)

      P = .10
      Ref0.24 (0.07–0.76)

      P = .02
      0.44 (0.19–1.03)

      P = .06
      Ref
      Within 10 y of diagnosis8/52 (15.4%)43/136 (31.6%)658/1803 (36.5%)8/98 (8.2%)19/110 (17.3%)162/915 (17.7%)
       Females: adjusted OR

      (95% CI)
      0.09 (0.01–0.67)

      P = .02
      0.76 (0.41–1.42)

      P = .40
      Ref0.90 (0.36–2.22)

      P = .81
      1.31 (0.67–2.54)

      P = .43
      Ref
       Males: adjusted OR (95% CI)0.45 (0.19–1.06)

      P = .07
      0.83 (0.52–1.33)

      P = .43
      Ref0.16 (0.04–0.66)

      P = .02
      0.59 (0.24–1.44)

      P = .25
      Ref
      Figure thumbnail gr2
      Figure 2Kaplan–Meier survival curve for the estimate of abdominal surgery over time in patients with Crohn’s disease.
      For patients with UC, the likelihood of colectomy is shown in Table 4, including sex-stratified odds ratios. The hazard of undergoing colectomy was similar in females age <6 years (HR, 0.88; 95% CI, 0.47–1.63) and females age 6–9.9 years (HR, 0.97; 95% CI, 0.59–1.61) compared with females age ≥10 years. The hazard of colectomy was lower in males age <6 years (HR, 0.42; 95% CI, 0.21–0.85), but not males age 6–9.9 years (HR, 0.56; 95% CI, 0.30–1.03). The Kaplan–Meier survival curves for colectomy in UC patients are shown in Figure 3.
      Figure thumbnail gr3
      Figure 3Kaplan–Meier survival curve for the estimate of colectomy in children with ulcerative colitis.

      Discussion

      By using a population-based cohort, we found that the incidence of IBD continues to increase in children, and is increasing most rapidly in children age <10 years. However, we determined that children age <6 years at diagnosis with IBD generally had lower rates of outpatient physician visits compared with their ≥10-year counterparts. Similarly, ED visits and hospitalizations were lower in the younger age groups. In addition, the likelihood of undergoing intestinal resection was lower in children with CD diagnosed at <6 years compared with those with older-age onset. These findings would imply more mild disease in patients with VEO-IBD of CD subtype. Alternatively, unique health care choices may have occurred in patients with VEO-CD, resulting in less urgent use of the health care system, and lower surgical rates. For example, concern about extension of CD involvement from the colon to the small bowel,
      • Van Limbergen J.
      • Russell R.K.
      • Drummond H.E.
      • et al.
      Definition of phenotypic characteristics of childhood-onset inflammatory bowel disease.
      general reluctance about performing surgery on young children, or less clinical need for surgery in colonic disease may have reduced the likelihood of surgery in the VEO-CD group. In addition, younger children may not be able to verbalize symptoms adequately, resulting in a lower likelihood of intervention and health services utilization.
      Our previous report from the Ontario Crohn’s and Colitis Cohort
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      included data through 2005, and reported no significant increase in children diagnosed at age 10–18 years, or an increase in UC incidence. By contrast, the addition of recent data in this study has shown a statistically significant increased incidence in the older and younger age groups, for both CD and UC. This is consistent with a recent French study that found a doubling of CD incidence in children age 10–19 years.
      • Chouraki V.
      • Savoye G.
      • Dauchet L.
      • et al.
      The changing pattern of Crohn's disease incidence in northern France: a continuing increase in the 10- to 19-year-old age bracket (1988-2007).
      However, the rate of increase is still fastest in the younger age groups. These results imply earlier disease onset or more prompt diagnosis of childhood-onset disease. Whatever the reason, the burden of IBD in children in Ontario is increasing, with an overall increase in standardized rates by more than 50% between 1994 and 2009.
      The age categories of the Paris modification of the Montreal classification system are based on differences in phenotype and serologic response in children with IBD onset before age 10 years.
      • Levine A.
      • Griffiths A.
      • Markowitz J.
      • et al.
      Pediatric modification of the Montreal classification for inflammatory bowel disease: the Paris classification.
      Children with VEO-CD are more likely to present with isolated colitis, and ileal disease is more likely in children who present at an older age.
      • Meinzer U.
      • Idestrom M.
      • Alberti C.
      • et al.
      Ileal involvement is age dependent in pediatric Crohn's disease.
      • Mamula P.
      • Telega G.W.
      • Markowitz J.E.
      • et al.
      Inflammatory bowel disease in children 5 years of age and younger.
      • Aloi M.
      • Lionetti P.
      • Barabino A.
      • et al.
      Phenotype and disease course of early-onset pediatric inflammatory bowel disease.
      However, this relationship was not present for children with NOD2/CARD15 mutations,
      • Levine A.
      • Kugathasan S.
      • Annese V.
      • et al.
      Pediatric onset Crohn's colitis is characterized by genotype-dependent age-related susceptibility.
      suggestion that susceptibility genes, age of onset, and disease location are linked inextricably. Unfortunately, we did not have access to genotypic or phenotypic information within the health administrative data, and therefore could not adjust for the likely effect modification. However, we confirmed the results of another study that showed that UC was more common than CD in those with VEO-IBD.
      • Mamula P.
      • Telega G.W.
      • Markowitz J.E.
      • et al.
      Inflammatory bowel disease in children 5 years of age and younger.
      Although children with VEO-IBD onset at age <6 years were more likely to have their initial diagnosis changed sometime throughout their course of illness, the majority had a final diagnosis of UC after a mean follow-up time of almost 10 years. By comparison, the majority of patients with onset after age 6 years had a final diagnosis of CD. Some studies have suggested that children age <2 years at diagnosis had a greater likelihood of a CD-like phenotype, more aggressive disease, and a greater genetic burden.
      • Cannioto Z.
      • Berti I.
      • Martelossi S.
      • et al.
      IBD and IBD mimicking enterocolitis in children younger than 2 years of age.
      • Mamula P.
      • Markowitz J.E.
      • Piccoli D.A.
      • et al.
      Immune response to influenza vaccine in pediatric patients with inflammatory bowel disease.
      • Ruemmele F.M.
      • El Khoury M.G.
      • Talbotec C.
      • et al.
      Characteristics of inflammatory bowel disease with onset during the first year of life.
      In fact, these very young patients may have a concomitant genetic predisposition to immunodeficiency syndromes,
      • Uhlig H.H.
      Monogenic diseases associated with intestinal inflammation: implications for the understanding of inflammatory bowel disease.
      including interleukin-10 and interleukin-10–receptor mutations,
      • Kotlarz D.
      • Beier R.
      • Murugan D.
      • et al.
      Loss of interleukin-10 signaling and infantile inflammatory bowel disease: implications for diagnosis and therapy.
      • Moran C.J.
      • Walters T.D.
      • Guo C.H.
      • et al.
      IL-10R polymorphisms are associated with very-early-onset ulcerative colitis.
      or reduced nicotinamide adenine dinucleotide phosphate oxidase genes NCF2/RAC211.
      • Muise A.M.
      • Xu W.
      • Guo C.H.
      • et al.
      NADPH oxidase complex and IBD candidate gene studies: identification of a rare variant in NCF2 that results in reduced binding to RAC2.
      The scarcity of patients diagnosed at age <2 years emphasizes the importance of international cohort studies of this important group.
      Our findings of decreased likelihood of surgery in children with VEO-CD confirm an American cohort study,
      • Schaefer M.E.
      • Machan J.T.
      • Kawatu D.
      • et al.
      Factors that determine risk for surgery in pediatric patients with Crohn's disease.
      and may be reflective of differences in phenotype, severity, or treatment choices. Mamula et al
      • Mamula P.
      • Telega G.W.
      • Markowitz J.E.
      • et al.
      Inflammatory bowel disease in children 5 years of age and younger.
      reported that 59% of young children had extensive small- and large-bowel disease, and 89% of those with VEO-CD had large-bowel involvement. With large-bowel involvement being very common in young children, these patients may have a decreased likelihood of undergoing resection owing to the requirement for permanent ostomy after colectomy. Younger patients also may be more likely to escalate medical therapy because of the extensiveness of their IBD, and therefore are less likely to undergo surgery. For example, young children may be more likely to be treated by pediatric gastroenterologists who may be more inclined to treat IBD medically rather than surgically.
      • Benchimol E.I.
      • Guttmann A.
      • To T.
      • et al.
      Changes to surgical and hospitalization rates of pediatric inflammatory bowel disease in Ontario, Canada (1994-2007).
      However, escalation of medical therapy typically would result in more frequent physician visits and hospitalizations, which was not observed in our cohort. Therefore, we suspect that patients with VEO-CD had more easily treatable inflammatory disease and therefore required fewer surgical interventions and hospitalizations. This may be secondary to more rapid diagnosis, and less time with undiagnosed intestinal inflammation, therefore resulting in less intestinal fibrosis. A recent Italian cohort study did not show significantly lower rates of surgery in children with VEO-IBD.
      • Aloi M.
      • Lionetti P.
      • Barabino A.
      • et al.
      Phenotype and disease course of early-onset pediatric inflammatory bowel disease.
      However, that cohort had a median follow-up time of 4 years, and our survival curves show that lower surgical rates in VEO-CD patients are apparent only after 5 years from disease onset.
      This study assessed differences in health system utilization between patients with VEO-IBD and older children. We found that children with VEO-CD had lower utilization of the health system in the outpatient and emergency settings. Although this implies more mild disease, the nature of health administrative data prevents this conclusion with certainty. This finding may have been secondary to health system access issues or personal preference. The families of young patients may choose to see physicians less frequently, although this would run counter to previous reports from the Ontario health system, which found increased utilization in younger children.
      • Guttmann A.
      • Lam K.
      • Schultz S.E.
      • et al.
      Primary care in children.
      The IBD-related outpatient code list has not been validated, however, the consistency with hospitalization and surgical findings (which use validated coding) is reassuring of code accuracy. Nevertheless, differential misclassification may have played a role in age-related differences. For example, the IBD-related diagnoses overlap between IBD and irritable bowel syndrome. Older children and adults with IBD may be more likely to experience irritable bowel syndrome symptoms than younger children, therefore seeing their physician more frequently. Although the IBD-specific codes were validated in various age groups to identify children with IBD,
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      they were not validated to distinguish functional from inflammatory symptoms in children with IBD.
      We previously described increased health services utilization and worse outcomes in females with pediatric IBD,
      • Benchimol E.I.
      • To T.
      • Griffiths A.M.
      • et al.
      Outcomes of pediatric inflammatory bowel disease: socioeconomic status disparity in a universal-access healthcare system.
      as have other cohorts.
      • Gupta N.
      • Bostrom A.G.
      • Kirschner B.S.
      • et al.
      Gender differences in presentation and course of disease in pediatric patients with Crohn disease.
      The statistical interaction between sex and age group in this study suggests a differential effect of early age of onset in females compared with males, perhaps owing to greater differential in disease severity between females with VEO-IBD and those with onset in adolescence. We therefore stratified our analyses by sex to allow for transparency in reporting of gender effect.
      The use of health administrative data to examine outcomes had limitations. The databases contained no clinical or biological information, such as genotype, phenotype, location, or severity. Although we could describe the health services events occurring in Ontario patients, we could not make conclusions as to whether the health care contacts were clinically appropriate. In addition, we did not have access to prescription medication records. Differential use of biologics or immunosuppressive agents may have contributed to differences in age groups, however, reliable medication data are not available for Ontario children. The majority of our patients were diagnosed before Health Canada approval for infliximab for pediatric CD (in 2006) and UC (in 2011). In addition, we would suspect lower utilization of immunosuppressive agents, thereby resulting in more uncontrolled disease in young children owing to concerns about adverse events. Misclassification bias is always a concern when using health administrative data. To identify children with IBD, we used algorithms validated in Ontario pediatric patients of all ages. The algorithms were shown to accurately classify patients as having IBD and distinguish CD from UC in children age <14 years and <18 years at diagnosis.
      • Benchimol E.I.
      • Guttmann A.
      • Griffiths A.M.
      • et al.
      Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
      In this study, the rate of misclassified or misdiagnosed patients was low in all 3 groups (0.5% of the overall cohort). Therefore, misclassification of non-IBD patients as having IBD was not likely contributing to the age differential observed. The advantages of using health administrative data to investigate outcomes include the large number of available patients, its population-based nature, and the availability of longitudinal data.
      We chose not to correct P values for multiple testing and instead emphasize odds/hazard ratios with 95% confidence intervals to show the clinical relevance of our findings. The primary objectives of this study make 2 comparisons per analysis (<6 vs ≥10 y, and 6–9.9 vs ≥10 y). Therefore, the Bonferroni method would consider P < .025 (0.05/2) to be significant, although this may be an overly conservative value.
      • Olejnik S.
      • Li J.
      • Supattathum S.
      • et al.
      Multiple testing and statistical power with modified Bonferroni procedures.
      In addition, more than 100 comparisons were conducted if one considers the odds of events at each time point, and the Bonferroni would consider P < .0005 to be significant (0.05/100). We chose to present the unadjusted P value as recommended by some methodologists,
      • Rothman K.J.
      No adjustments are needed for multiple comparisons.
      • Proschan M.A.
      • Waclawiw M.A.
      Practical guidelines for multiplicity adjustment in clinical trials.
      but would advise interpreting results of P > .01 with caution.
      In conclusion, we found an increased incidence of childhood-onset IBD in Ontario and lower health services utilization and surgical rates in children with VEO-CD with onset at age <6 years. This was not the case for children with VEO-UC or for children with onset at age 6–9.9 years, who had similar rates of health services utilization compared with those diagnosed at an older age. Future research should examine subgroups within the VEO-IBD population to assess differences in phenotype, genotype, intestinal microbiome, and response to medications.

      Acknowledgments

      The authors wish to thank Mr Steven Hawken (Institute for Clinical Evaluative Sciences uOttawa) for his review of the analytic code and manuscript.
      The following authors were part of the interNational Early Onset Pediatrics IBD Cohort Study (www.NEOPICS.org): Eric I. Benchimol, David R. Mack, Scott B. Snapper, Aleixo M. Muise. This work was made possible with the support of ICES which receives funding from the Ontario Ministry of Health and Long-Term Care.

      Supplementary Material

      Supplementary Table 1Classification of IBD-Related Health Services Use
      ConditionICD-9OHIP diagnostic codeICD-10
      IBD specific
       Crohn's disease555.x555K50.x
       UC556.x556K51.x
      IBD signs/symptoms
       Anorexia783.0787R63.0
       Abnormal weight gain783.1R63.5
       Abnormal weight loss783.2R63.4
       Underweight783.22R62.8
       Failure to thrive, child783.41R62.8

      R62.9
       Failure to thrive, adult783.7R62.8

      R62.9
      Symptoms involving digestive system, including

       (787.0) Nausea and vomiting

       (787.01) Nausea with vomiting

       (787.02) Nausea, alone

       (787.03) Vomiting, alone

       (787.1) Heartburn

       (787.2) Dysphagia

       (787.3) Gas/bloating

       (787.6) Encopresis, fecal incontinence

       (787.9) Other symptoms involving digestive system

       (787.91) Diarrhea, not otherwise specified
      787.x787R11.x

      R12.x

      R13.x

      R14.x

      R15.x

      R19.x
      Abdominal pain789.0787R10.x
      Dyspepsia536.8536K30.x
      Cachexia799.4R64.x
      Esophagitis530.1530K20.x

      K22.1
      GERD530.81K21.x
      Gastric ulcer531531K25.x
      Duodenal ulcer532532K26.x
      Peptic ulcer533K27.x
      Gastrojejunal ulcer534534K28.x
      Gastritis/duodenitis535.x535K29.x
      Intestinal obstruction560.8

      560.9
      560K31.5

      K56.6
      Rectal/anal hemorrhage569.3569K62.5
      Other disorder of rectum/anus, including

       (569.41) Ulcer

       (569.42) Pain

       (569.43) Sphincter tear (healed)

       (569.44) Dysplasia

       (569.45) Other specified, including proctitis, inflammation
      569.4569K62.6

      K62.8
      Abscess of the intestine569.5569K63.0
      Other disorders of intestine, including

       (569.81) Fistula (excluding rectum)

       (569.82) Ulcer of intestine

       (569.83) Perforation

       (569.84) Angiodysplasia, no hemorrhage

       (569.85) Angiodysplasia, with hemorrhage

       (569.86) Dieulafoy

       (569.89) Other, including

      Enteroptosis

      Granuloma of intestine

      Prolapse of intestine

      Pericolitis

      Perisigmoiditis

      Visceroptosis
      569.8569K63.2

      K63.3

      K63.1

      K55.2

      K63.8
      Malabsorption262.x

      263.0

      263.1

      263.2

      263.8

      263.9

      579.8

      579.9
      579E43.x

      E44.x

      E45.x

      E46.x

      K90.8

      K90.9
      Extraintestinal manifestations
      Anal fistula565.1565K60.3
      Anal abscess566
      Ureteral fistula593.82N28.8
      Urethral fistula599.1N36.0
      Fistula of stomach and duodenum537.4K31.6
      Vesical fistula596.2N32.2
      Fistula involving female genitourinary system619.xN82.x
      Hemorrhoids, including

      (455.9) anal skin tags
      455.x455I84.x
      Rheumatoid arthritis714.x714M06.99

      M05.3

      M08.x

      M09.x
      Arthropathy associated with erythema nodosum713.3M14.8
      Inflammatory spondylopathies, including

       (720.0) Ankylosing spondylitis

       (720.1) Spinal enthesopathy

       (720.2) Sacroiliitis

       (720.8) Other inflammatory

       (720.9) Other unspecified inflammatory
      720.x720M45.x

      M46.x

      M47.x

      M48.x
      Scleritis and episcleritis379.0379H15.x
      Unspecified iridocyclitis (uveitis not otherwise specified)364.3364H20.9
      Chorioretinitis, unspecified (uveitis, posterior not otherwise specified)363.2363H30.9
      Acute and subacute iridocyclitis364.0364H20.0
      Erythema nodosum695.2695L52.x
      Acrodermatitis enteropathica686.8E83.2
      Pyoderma686.0686L08.0
      Pyogenic granuloma of the skin and soft tissue686.1686L98.0
      Oral aphthae528.2528K12.0
      Short stature783.43E34.3
      Osteoporosis733.0733M80.x

      M81.x

      M82.x

      M83.x
      Osteomyelitis730.0

      730.1

      730.2
      730M86.x
      Acute glomerulonephritis580.x580N00.x

      N01.x

      N05.x
      Nephrolithiasis592.x592N20.x
      Primary sclerosing cholangitis576.1576K83.0
      Venous embolism/thrombosis453.xI82.x
      NOTE. The description of IBD-related diagnoses in the text refer to all diagnostic codes classified as IBD-specific, IBD signs/symptoms, and extra-intestinal manifestations.
      GERD, gastroesophageal reflux disease; ICD, International Classification of Diseases.
      Supplementary Table 2Classification of Intestinal Resection and Colectomy Using Canadian Institute for Health Information Procedure Codes
      Resection/colectomy for CDColectomy for UC
      5741: Multiple segmental resection of small intestine
      5742: Other partial resection of small intestine
      5743: Total removal of small intestine
      575: Partial excision of large intestine575: Partial excision of large intestine
      5751: Multiple segmental resection of large intestine5751: Multiple segmental resection of large intestine
      5753: Right hemicolectomy5753: Right hemicolectomy
      5755: Left hemicolectomy5755: Left hemicolectomy
      576: Total colectomy576: Total colectomy
      5752: Cecectomy5752: Cecectomy
      5754: Resection of transverse colon5754: Resection of transverse colon
      5756: Sigmoidectomy5756: Sigmoidectomy
      5759: Other partial excision of large intestine5759: Other partial excision of large intestine
      Supplementary Table 3Most Common Diagnoses Associated With Outpatient Physician Visits in the Two Years Before the Date of Diagnosis With IBD
      Code<6 y (n = 383)6–9.9 y (n = 657)≥10 y (n = 6103)
      Diseases of the gastrointestinal system not yet diagnosed787189/383

      (49.3%)
      399/657

      (60.7%)
      3324/6103

      (54.5%)
      Upper respiratory tract infection460208/383

      (54.3%)
      344/657

      (52.4%)
      2220/6103

      (36.4%)
      Diarrhea, gastroenteritis, viral gastroenteritis009217/383

      (56.7%)
      349/657

      (53.1%)
      2218/6103

      (36.3%)
      Other undefined conditions, general symptoms799133/383

      (34.7%)
      190/657

      (28.9%)
      1126/6103

      (18.4%)
      Rectal bleeding or other rectal problems (polyp, prolapse, stricture)569122/383

      (31.9%)
      165/657

      (25.1%)
      1014/6103

      (16.6%)
      Irritable bowel syndrome, constipation564101/383

      (26.4%)
      148/657

      (22.5%)
      1006/6103

      (16.5%)
      Rash69194/383

      (24.5%)
      113/657

      (17.2%)
      793/6103

      (13.0%)
      Iron deficiency28033/383

      (8.6%)
      54/657

      (8.2%)
      688/6103

      (11.3%)
      Leg cramps, leg pain, muscle pain, joint pain, arthralgia, joint swelling, masses78120/383

      (5.2%)
      55/657

      (8.4%)
      661/6103

      (10.8%)
      Acute bronchitis46669/383

      (18.0%)
      89/657

      (13.5%)
      658/6103

      (10.8%)
      Asthma, allergic bronchitis49355/383

      (14.4%)
      98/657

      (14.9%)
      643/6103

      (10.5%)
      Other viral diseases07974/383

      (19.3%)
      89/657

      (13.5%)
      469/6103

      (7.7%)
      Serous otitis media, eustachian tube disorders38190/383

      (23.5%)
      101/657

      (15.4%)
      375/6103

      (6.1%)
      Suppurative otitis media38275/383

      (19.6%)
      74/657

      (11.3%)
      207/6103

      (3.4%)
      Well-baby care916144/383

      (37.6%)
      6/657

      (0.9%)
      N/A
      NOTE. Diagnoses are ranked in order from most to least common in the ≥10-year age group.
      Supplementary Table 4IBD-Specific and IBD-Related Outpatient Physician Visit Rates Comparing Age Groups by Multivariable Poisson Regression
      <6 y at diagnosis6–9.9 y diagnosis≥10 y at diagnosis
      Crohn’s disease, n1273463720
       IBD-specific
      Number of visits per year, mean ± SD3.5 ± 3.34.8 ± 4.33.7 ± 3.1
      Females: adjusted OR (95% CI)0.70 (0.50–0.98)

      P = .04
      1.29 (1.09–1.52)

      P = .003
      Ref
      Males: adjusted OR (95% CI)1.00 (0.81–1.24)

      P = .99
      1.28 (1.16–1.41)

      P < .0001
      Ref
       IBD-related
      Number of visits per year, mean ± SD4.3 ± 3.65.6 ± 4.94.4 ± 3.5
      Females: adjusted OR (95% CI)0.74 (0.57–0.97)

      P = .03
      1.21 (1.05–1.40)

      P = .01
      Ref
      Males: adjusted OR (95% CI)1.00 (0.83–1.21)

      P = .98
      1.25 (1.14–1.37)

      P < .0001
      Ref
      Ulcerative colitis, n2132441987
       IBD-specific
      Number of visits per year, mean ± SD2.5 ± 2.72.7 ± 3.82.8 ± 3.0
      Females: adjusted OR (95% CI)0.77 (0.63–0.94)

      P = .01
      0.93 (0.80–1.08)

      P = .34
      Ref
      Males: adjusted OR (95% CI)0.86 (0.71–1.05)

      P = .14
      0.99 (0.86–1.15)

      P = .93
      Ref
       IBD-related
      Number of visits per year, mean ± SD3.0 ± 2.83.3 ± 4.03.4 ± 3.3
      Females: adjusted OR (95% CI)0.75 (0.63–0.89)

      P = .0009
      0.91 (0.80–1.05)

      P = .20
      Ref
      Males: adjusted OR (95% CI)0.89 (0.75–1.05)

      P = .17
      0.98 (0.85–1.13)

      P = .78
      Ref
      Supplementary Table 5Likelihood and Odds of Hospitalization in IBD, CD, and UC Patients at 1, 3, 5, and 10 Years After Diagnosis Comparing Age Groups by Multivariable Logistic Regression
      <6 y6–9.9 y≥10 y
      Overall IBD
       Hospitalization within 1 y124/365 (34.0%)149/631 (23.6%)1608/5872 (27.4%)
      Female, OR (95% CI)1.31 (0.93–1.86)

      P = .13
      0.96 (0.73–1.26)

      P = .76
      Ref
      Male, OR (95% CI)1.39 (1.03–1.86)

      P = .03
      0.71 (0.54–0.93)

      P = .01
      Ref
       Hospitalization within 3 y139/311 (44.7%)216/539 (40.1%)2233/5165 (43.2%)
      Female, OR (95% CI)0.89 (0.62–1.27)

      P = .52
      0.91 (0.69–1.19)

      P = .49
      Ref
      Male, OR (95% CI)1.22 (0.90–1.65)

      P = .20
      0.87 (0.68–1.11)

      P = .27
      Ref
       Hospitalization within 5 y126/251 (50.2%)228/464 (49.14%)2059/4424 (53.5%)
      Female, OR (95% CI)0.73 (0.50–1.08)

      P = .12
      0.73 (0.55–0.98)

      P = .03
      Ref
      Male, OR (95% CI)1.01 (0.72–1.42)

      P = .94
      0.96 (0.74–1.24)

      P = .77
      Ref
       Hospitalization within 10 y87/160 (54.4%)170/268 (63.4%)1922/2857 (67.3%)
      Female, OR (95% CI)0.48 (0.29–0.79)

      P = .004
      0.79 (0.53–1.18)

      P = .25
      Ref
      Male, OR (95% CI)0.66 (0.43–1.01)

      P = .05
      0.89 (0.63–1.26)

      P = .50
      Ref
      CD
       Hospitalization within 1 y42/119 (35.3%)84/332 (25.3%)1020/3568 (28.6%)
      Female, OR (95% CI)1.45 (0.78–2.69)

      P = .24
      0.90 (0.60–1.33)

      P = .59
      Ref
      Male, OR (95% CI)1.32 (0.81–2.15)

      P = .27
      0.82 (0.58–1.15)

      P = .25
      Ref
       Hospitalization within 3 y47/104 (45.2%)116/272 (42.6%)1450/3167 (45.8%)
      Female, OR (95% CI)0.90 (0.48–1.69)

      P = .73
      0.91 (0.61–1.35)

      P = .62
      Ref
      Male, OR (95% CI)1.07 (0.65–1.76)

      P = .80
      0.89 (0.65–1.23)

      P = .49
      Ref
       Hospitalization within 5 y46/86 (53.5%)133/240 (55.4%)1571/2744 (57.2%)
      Female, OR (95% CI)0.69 (0.34–1.40)

      P = .31
      0.74 (0.48–1.13)

      P = .16
      Ref
      Male, OR (95% CI)1.02 (0.59–1.77)

      P = .93
      1.12 (0.80–1.58)

      P = .51
      Ref
       Hospitalization within 10 y29/52 (55.8%)95/136 (69.8%)1316/1803 (73.0%)
      Female, OR (95% CI)0.42 (0.17–1.06)

      P = .07
      0.73 (0.39–1.36)

      P = .32
      Ref
      Male, OR (95% CI)0.48 (0.24–0.99)

      P = .045
      0.96 (0.60–1.56)

      P = .88
      Ref
      UC
       Hospitalization within 1 y71/205 (34.6%)52/238 (21.8%)494/1933 (25.6%)
      Female, OR (95% CI)1.56 (0.99–2.47)

      P = .06
      1.03 (0.67–1.60)

      P = .89
      Ref
      Male, OR (95% CI)1.52 (1.00–2.29)

      P = .049
      0.60 (0.36–1.00)

      P = .049
      Ref
       Hospitalization within 3 y79/179 (44.1%)79/217 (36.4%)674/1695 (39.8%)
      Female, OR (95% CI)1.07 (0.66–1.71)

      P = .79
      0.89 (0.59–1.34)

      P = .58
      Ref
      Male, OR (95% CI)1.32 (0.87–2.01)

      P = .19
      0.86 (0.56–1.32)

      P = .48
      Ref
       Hospitalization within 5 y72/148 (48.6%)76/186 (40.9%)678/1429 (47.4%)
      Female, OR (95% CI)1.02 (0.61–1.70)

      P = .94
      0.78 (0.50–1.20)

      P = .25
      Ref
      Male, OR (95% CI)1.07 (0.68–1.69)

      P = .77
      0.77 (0.49–1.21)

      P = .26
      Ref
       Hospitalization within 10 y51/98 (52.0%)61/110 (55.5%)530/915 (57.9%)
      Female, OR (95% CI)0.64 (0.34–1.22)

      P = .18
      0.93 (0.54–1.60)

      P = .78
      Ref
      Male, OR (95% CI)0.94 (0.54–1.65)

      P = .83
      0.92 (0.50–1.68)

      P = .78
      Ref
      Figure thumbnail fx1
      Supplementary Figure 1Standardized incidence of CD from 1994 to 2009 by age group.
      Figure thumbnail fx2
      Supplementary Figure 2Standardized incidence of UC from 1994 to 2009 by age group.

      References

        • Benchimol E.I.
        • Fortinsky K.J.
        • Gozdyra P.
        • et al.
        Epidemiology of pediatric inflammatory bowel disease: a systematic review of international trends.
        Inflamm Bowel Dis. 2011; 17: 423-439
        • Benchimol E.I.
        • Guttmann A.
        • Griffiths A.M.
        • et al.
        Increasing incidence of paediatric inflammatory bowel disease in Ontario, Canada: evidence from health administrative data.
        Gut. 2009; 58: 1490-1497
        • Kotlarz D.
        • Beier R.
        • Murugan D.
        • et al.
        Loss of interleukin-10 signaling and infantile inflammatory bowel disease: implications for diagnosis and therapy.
        Gastroenterology. 2012; 143: 347-355
        • Muise A.M.
        • Xu W.
        • Guo C.H.
        • et al.
        NADPH oxidase complex and IBD candidate gene studies: identification of a rare variant in NCF2 that results in reduced binding to RAC2.
        Gut. 2012; 61: 1028-1035
        • Moran C.J.
        • Walters T.D.
        • Guo C.H.
        • et al.
        IL-10R polymorphisms are associated with very-early-onset ulcerative colitis.
        Inflamm Bowel Dis. 2013; 19: 115-123
        • Glocker E.O.
        • Kotlarz D.
        • Boztug K.
        • et al.
        Inflammatory bowel disease and mutations affecting the interleukin-10 receptor.
        N Engl J Med. 2009; 361: 2033-2045
        • Imielinski M.
        • Baldassano R.N.
        • Griffiths A.
        • et al.
        Common variants at five new loci associated with early-onset inflammatory bowel disease.
        Nat Genet. 2009; 41: 1335-1340
        • Kugathasan S.
        • Baldassano R.N.
        • Bradfield J.P.
        • et al.
        Loci on 20q13 and 21q22 are associated with pediatric-onset inflammatory bowel disease.
        Nat Genet. 2008; 40: 1211-1215
        • Heyman M.B.
        • Kirschner B.S.
        • Gold B.D.
        • et al.
        Children with early-onset inflammatory bowel disease (IBD): analysis of a pediatric IBD consortium registry.
        J Pediatr. 2005; 146: 35-40
        • Markowitz J.
        • Kugathasan S.
        • Dubinsky M.
        • et al.
        Age of diagnosis influences serologic responses in children with Crohn's disease: a possible clue to etiology?.
        Inflamm Bowel Dis. 2009; 15: 714-719
        • Meinzer U.
        • Idestrom M.
        • Alberti C.
        • et al.
        Ileal involvement is age dependent in pediatric Crohn's disease.
        Inflamm Bowel Dis. 2005; 11: 639-644
        • Shaoul R.
        • Karban A.
        • Reif S.
        • Weiss B.
        • Shamir R.
        • Tamir A.
        • Davidovich O.
        • Halevi J.
        • Silver E.L.
        • Levine A.
        Disease behavior in children with Crohn's disease: the effect of disease duration, ethnicity, genotype, and phenotype.
        Dig Dis Sci. 2009; 54: 142-150
        • Levine A.
        • Griffiths A.
        • Markowitz J.
        • et al.
        Pediatric modification of the Montreal classification for inflammatory bowel disease: the Paris classification.
        Inflamm Bowel Dis. 2011; 17: 1314-1321
        • Silverberg M.S.
        • Satsangi J.
        • Ahmad T.
        • et al.
        Toward an integrated clinical, molecular and serological classification of inflammatory bowel disease: Report of a Working Party of the 2005 Montreal World Congress of Gastroenterology.
        Can J Gastroenterol. 2005; 19: 5-36
        • Muise A.M.
        • Snapper S.B.
        • Kugathasan S.
        The age of gene discovery in very early onset inflammatory bowel disease.
        Gastroenterology. 2012; 143: 285-288
      1. interNational Early Onset Paediatric IBD Cohort Study (NEOPICS). NEOPICS, 2013. Available at: http://neopics.org. Accessed May 20, 2014.

      2. Snapper SB, Klein C, Muise AM. Very early onset IBD: mapping the research agenda. Amsterdam. NEOPICS, 2013. Available at: http://neopics.org/uploads/VEOIBD_MEETING_SUMMARY.pdf. Accessed May 20, 2014.

        • Benchimol E.I.
        • Guttmann A.
        • To T.
        • et al.
        Changes to surgical and hospitalization rates of pediatric inflammatory bowel disease in Ontario, Canada (1994-2007).
        Inflamm Bowel Dis. 2011; 17: 2153-2161
        • Fay M.P.
        • Feuer E.J.
        Confidence intervals for directly standardized rates: a method based on the gamma distribution.
        Stat Med. 1997; 16: 791-801
      3. Statistics Canada. Population and dwelling counts, for Canada, provinces, territories and census subdivisions (municipalities), 2006 census, 100% data (table). Population and dwelling counts. Statistics Canada, 2007. Available at: http://www12.statcan.ca/english/census06/data/popdwell/Table.cfm?T=101. Accessed May 20, 2014.

      4. Statistics Canada. Population and dwelling counts, for Canada, provinces, territories and census subdivisions (municipalities), 2001 and 1996 censuses, 100% data (table). Population and dwelling counts. Statistics Canada, 2002. Available at: http://www12.statcan.ca/english/census01/products/standard/popdwell/tables.cfm. Accessed May 20, 2014.

      5. CIHI data quality study of the 2009-2010 discharge abstract database. Canadian Institute for Health Information, Standards and Data Submission, Ottawa2012
        • Ma C.
        • Crespin M.
        • Proulx M.C.
        • et al.
        Postoperative complications following colectomy for ulcerative colitis: a validation study.
        BMC Gastroenterol. 2012; 12: 39
        • Soon I.S.
        • Wrobel I.
        • deBruyn J.C.
        • et al.
        Postoperative complications following colectomy for ulcerative colitis in children.
        J Pediatr Gastroenterol Nutr. 2012; 54: 763-768
        • Wilkins R.
        PCCF+ version 4E user's guide: automated geographic coding based on the Statistics Canada postal code conversion files, including postal codes to July 2004. Catalogue no. 93-387-XIE.
        Health Analysis and Measurement Group, Statistics Canada, Ottawa, Canada2005
        • Van Limbergen J.
        • Russell R.K.
        • Drummond H.E.
        • et al.
        Definition of phenotypic characteristics of childhood-onset inflammatory bowel disease.
        Gastroenterology. 2008; 135: 1114-1122
        • Chouraki V.
        • Savoye G.
        • Dauchet L.
        • et al.
        The changing pattern of Crohn's disease incidence in northern France: a continuing increase in the 10- to 19-year-old age bracket (1988-2007).
        Aliment Pharmacol Ther. 2011; 33: 1133-1142
        • Mamula P.
        • Telega G.W.
        • Markowitz J.E.
        • et al.
        Inflammatory bowel disease in children 5 years of age and younger.
        Am J Gastroenterol. 2002; 97: 2005-2010
        • Aloi M.
        • Lionetti P.
        • Barabino A.
        • et al.
        Phenotype and disease course of early-onset pediatric inflammatory bowel disease.
        Inflamm Bowel Dis. 2014; 20: 597-605
        • Levine A.
        • Kugathasan S.
        • Annese V.
        • et al.
        Pediatric onset Crohn's colitis is characterized by genotype-dependent age-related susceptibility.
        Inflamm Bowel Dis. 2007; 13: 1509-1515
        • Cannioto Z.
        • Berti I.
        • Martelossi S.
        • et al.
        IBD and IBD mimicking enterocolitis in children younger than 2 years of age.
        Eur J Pediatr. 2009; 168: 149-155
        • Mamula P.
        • Markowitz J.E.
        • Piccoli D.A.
        • et al.
        Immune response to influenza vaccine in pediatric patients with inflammatory bowel disease.
        Clin Gastroenterol Hepatol. 2007; 5: 851-856
        • Ruemmele F.M.
        • El Khoury M.G.
        • Talbotec C.
        • et al.
        Characteristics of inflammatory bowel disease with onset during the first year of life.
        J Pediatr Gastroenterol Nutr. 2006; 43: 603-609
        • Uhlig H.H.
        Monogenic diseases associated with intestinal inflammation: implications for the understanding of inflammatory bowel disease.
        Gut. 2013; 62: 1795-1805
        • Schaefer M.E.
        • Machan J.T.
        • Kawatu D.
        • et al.
        Factors that determine risk for surgery in pediatric patients with Crohn's disease.
        Clin Gastroenterol Hepatol. 2010; 8: 789-794
        • Guttmann A.
        • Lam K.
        • Schultz S.E.
        • et al.
        Primary care in children.
        in: Jaakkimainen L. Upshur R.E. Klein-Geltink J.E. Primary care in Ontario. Institute for Clinical Evaluative Sciences, Toronto2006: 35-52
        • Benchimol E.I.
        • To T.
        • Griffiths A.M.
        • et al.
        Outcomes of pediatric inflammatory bowel disease: socioeconomic status disparity in a universal-access healthcare system.
        J Pediatr. 2011; 158 (e1–4): 960-967
        • Gupta N.
        • Bostrom A.G.
        • Kirschner B.S.
        • et al.
        Gender differences in presentation and course of disease in pediatric patients with Crohn disease.
        Pediatrics. 2007; 120: e1418-e1425
        • Olejnik S.
        • Li J.
        • Supattathum S.
        • et al.
        Multiple testing and statistical power with modified Bonferroni procedures.
        J Educ Behav Stat. 1997; 22: 389-406
        • Rothman K.J.
        No adjustments are needed for multiple comparisons.
        Epidemiology. 1990; 1: 43-46
        • Proschan M.A.
        • Waclawiw M.A.
        Practical guidelines for multiplicity adjustment in clinical trials.
        Control Clin Trials. 2000; 21: 527-539

      Linked Article