- Email: [email protected]
Epidemiologic and clinical characteristics of children with newly diagnosed inflammatory bowel disease in wisconsin: a statewide population-based study
EPIDEMIOLOGIC AND CLINICAL CHARACTERISTICS OF CHILDREN WITH NEWLY DIAGNOSED INFLAMMATORY BOWEL DISEASE IN WISCONSIN: A STATEWIDE POPULATION-BASED STUDY SUBRA KUGATHASAN, MD, ROBERT H. JUDD, MD, RAYMOND G. HOFFMANN, PHD, JANICE HEIKENEN, MD, GREGORZ TELEGA, MD, FARHAT KHAN, MD, SALLY WEISDORF-SCHINDELE, MD, WILLIAM SAN PABLO, JR, MD, JEAN PERRAULT, MD, ROGER PARK, MD, MICHAEL YAFFE, MD, CHRISTOPHER BROWN, MD, MARIA T. RIVERA-BENNETT, MD, ISSAM HALABI, MD, ALFONSO MARTINEZ, MD, ELLEN BLANK, MD, STEVEN L. WERLIN, MD, COLIN D. RUDOLPH, MD, PHD, AND DAVID G. BINION, MD, FOR THE WISCONSIN PEDIATRIC INFLAMMATORY BOWEL DISEASE ALLIANCE (APPENDIX)
Objective To define epidemiologic and clinical characteristics of newly diagnosed pediatric inflammatory bowel disease (IBD) in a large population-based model. Study design All pediatric gastroenterologists providing care for Wisconsin children voluntarily identified all new cases of IBD during a 2-year period. Demographic and clinical data were sent to a central registry prospectively for analysis. Results The incidence of IBD in Wisconsin children was 7.05 per 100,000, whereas the incidence for Crohn’s disease was 4.56, more than twice the rate of ulcerative colitis (2.14). An equal IBD incidence occurred among all ethnic groups, and children from sparsely and densely populated counties were equally affected. The majority (89%) of new IBD diagnoses were nonfamilial. Conclusions
This study provides novel, prospective, and comprehensive information on pediatric IBD incidence within the United States. The surprisingly high incidence of pediatric IBD, the predominance of Crohn’s disease over ulcerative colitis, the low frequency of patients with a family history, the equal distribution of IBD among all racial and ethnic groups, and the lack of a modulatory effect of urbanization on IBD incidence collectively suggest that the clinical spectrum of IBD is still evolving and point to environmental factors contributing to the pathogenesis. (J Pediatr 2003;143:525-31)
rohn’s disease (CD) and ulcerative colitis (UC), collectively known as inflammatory bowel disease (IBD), are idiopathic, lifelong, destructive chronic inflammatory conditions of the gastrointestinal tract. CD and UC were recognized as disease entities during the early 20th century, and pathogenesis has been linked to a combination of genetic and environmental factors.1,2 Population-based studies suggest that IBD is unevenly distributed throughout the world, with the highest disease rates occurring in first world, Westernized countries.3 Epidemiologic surveys have also suggested that IBD incidence rates have changed over the second half of the 20th century, with a steady, gradual increase for both UC and CD that appears to be independent of observer bias.4-6 Because increasingly high rates of IBD onset occur in childhood and adolescence, monitoring the incidence of pediatric IBD may reflect the changing trends in IBD demographics. Recent retrospective European studies and a single prospective populationbased survey from Great Britain have suggested that the incidence of IBD in children and adolescents has significantly increased over the last 35 years. These pediatric studies also suggest that there has been a change in the patterns of disease, because the incidence of CD has risen above that of UC.7,8 So far, there have been no systematic, population-based
C
CD CI IBD
Crohn’s disease Confidence interval Inflammatory bowel disease
IC UC
Indeterminate colitis Ulcerative colitis
From Division of Pediatric Gastroenterology and Nutrition, the Department of Epidemiology and Biostatistics, and the Division of Gastroenterology and Hepatology, Medical College of Wisconsin, Milwaukee; Children’s Hospital of Wisconsin, Milwaukee; the Division of Pediatric Gastroenterology, University of Wisconsin—Madison, School of Medicine, Madison; Dean Clinic, Madison; the Department of Pediatrics, Marshfield Clinic, Marshfield; Gunderson Clinic, La Crosse; and Gastroenterology Associates, Wausau, Wisconsin; and the Division of Pediatric Gastroenterology, University of Minnesota, Minneapolis; and the Division of Pediatric Gastroenterology, Mayo Clinic, Rochester, Minnesota. Supported by National Institutes of Health grant (RR016111, S. K.), the National Institutes of Health Clinical Scholars program (K30HL004111, S. K.), and the general clinical research center at the Medical College of Wisconsin and the Medical College of Wisconsin’s Digestive Disease Center (S. K., D. G. B.). Submitted for publication Feb 21, 2003; revision received June 18, 2003; accepted July 23, 2003. Reprint requests: Subra Kugathasan, MD, Department of Pediatrics, Medical College of Wisconsin, 8701 Watertown Plank Rd, Milwaukee, WI 53226. E-mail: [email protected]. Copyright ª 2003 Mosby, Inc. All rights reserved. 0022-3476/2003/$30.00 + 0 10.1067/S0022-3476(03)00444-X
525
studies in North America on the incidence of childhood IBD, because IBD is not a reportable condition, and there is no regional or national registry for tracking each IBD diagnosis within the United States. Our aim was to determine the pediatric IBD incidence and to define clinical characteristics from a statewide-defined population.
MATERIALS AND METHODS STUDY SETTING AND FORMATION OF THE WISCONSIN PEDIATRIC INFLAMMATORY BOWEL DISEASE ALLIANCE. Identification of new pediatric IBD diagnoses was based on voluntary reporting from pediatric gastroenterology subspecialists. All pediatric gastroenterologists within Wisconsin (n = 17) and those from neighboring states bordering the southern (Illinois, n = 2) and western (Minnesota, n = 2) areas of Wisconsin, who might provide care for resident children, participated in the study. These physicians were based in three academic institutions and five community-based practices within Wisconsin, two academic institutions from Minnesota, and one academic and one community-based practice in Illinois.
Recruitment of Patients and Data Collection Methods All children who were <18 years old at the time of diagnosis of IBD whose residential address was a Wisconsin zip code were eligible for inclusion in the study. Patient accrual began prospectively on January 1, 2000, and continued for a 2-year period ending December 31, 2001. A questionnaire for each new subject was completed by the pediatric gastroenterologist who made the IBD diagnosis. The study was conducted in Wisconsin, an upper midwestern state in the United States. The 2000 US Census Bureau data defining the total state population, pediatric population (age <18 years), county populations, and ethnic backgrounds were used in this analysis. Wisconsin has 72 counties, and the population of counties in the year 2000 was used to determine the distribution of population density throughout the state. Patients’ residential zip codes were used to determine county residence. Coded data were entered in a computerized database by using patient initials, date of birth, zip code, and the diagnosing physician’s name. Data recorded included clinical presenting symptoms, duration of each symptom, and family history of IBD. The date of diagnosis was set as the day of the definitive diagnostic procedure rather than symptom onset. All data from the time of the diagnostic work-up, including the laboratory findings, serologic tests, radiologic studies, endoscopic evaluation, and histologic findings, were obtained and entered into the central database. Any surgical intervention needed at the time of diagnosis was also recorded.
Criteria for Inflammatory Bowel Disease Diagnosis and Disease Subcategorization The diagnosis of IBD required endoscopic evaluation, including histologic assessment of mucosal pinch biopsies. All patients underwent colonoscopic examination with or without upper endoscopy with multiple mucosal biopsies for review by 526
Kugathasan et al
clinical pathologists. Stool studies were performed in all patients to exclude infectious causes of diarrheal illness. Patients whose IBD diagnosis was not confirmed by endoscopic or histologic examination were excluded from further analysis. Serologic testing of antineutrophil or anti-Saccharomyces cerevisiae antibodies, which correlate with an IBD diagnosis, was not considered a sufficient criterion for a diagnosis of IBD. New pediatric IBD cases were further subcategorized into three clinical groups—CD, UC, and indeterminate colitis (IC)—defined according to established clinical, biochemical, radiologic, endoscopic, and histologic criteria.9,10 Children with nonspecific chronic inflammation without firm evidence of IBD were excluded from the analysis. CD was defined as the presence of histologically confirmed discontinuous chronic inflammation of the gastrointestinal tract confirmed by endoscopy and supported by clinical, biochemical, and radiologic evidence of CD. UC was defined as the presence of continuous inflammation limited to the colon, extending proximally from the involved rectum, with histologically typical chronic inflammation limited to the mucosa, not extending beyond the muscularis mucosa. In all cases, the terminal ileum was evaluated by small bowel followthrough barium radiographs or colonoscopy and ileoscopy. IC was defined as inflammatory colitis in the setting of histopathological changes indicative of chronic IBD colitis, containing both endoscopic and microscopic findings that were consistent with both CD and UC. The presence of continuous endoscopic disease in the setting of either discontinuous microscopic disease or inflammation extending beyond the muscularis mucosa was included in the indeterminate category. In addition, patients with IC had no radiologic, endoscopic, or histologic evidence of small bowel disease. Anatomic regions of CD involvement at diagnosis, with the exception of jejunal and perianal CD, were confirmed by microscopic evidence of CD chronic inflammation. The presence of nonspecific inflammation was not a sufficient criterion for the diagnosis of upper gastrointestinal tract CD. In UC cases with inflammation limited to the left side of the colon and not extending proximally beyond the splenic flexure were categorized as left-sided colitis. Pancolitis was defined as gross and microscopic evidence of chronic inflammation extending proximally beyond the splenic flexure.
Analysis of Incidence Rates, Census Data, and Statistics Incidence rates were calculated per million pediatric populations per year, and the 95% confidence intervals (CIs) were estimated assuming a Poisson distribution of cases. Linear regression was used for multivariate analyses to calculate adjusted incidence rate ratios among the Wisconsin counties. Paired and unpaired t tests and v2 tests were used to determine significance. The Kolmogorov-Smirnov test was used to compare age of diagnosis, because the age distribution was skewed. This study was approved by the research and publication committee of Children’s Hospital of Wisconsin and the Medical College of Wisconsin. The Journal of Pediatrics October 2003
Fig 2. Age-related occurrence of all new-onset pediatric IBD, CD, and UC in Wisconsin between 2000 and 2001.
Table I. Demographics of newly diagnosed children with IBD children from January 1, 2000, to December 31, 2001
Fig 1. Distribution of new onset pediatric IBD cases in Wisconsin during the 24-month period from January 1, 2000, to December 31, 2001. Outline of the state of Wisconsin shows the localization of 129 CD (red circle), 60 UC (blue star), and 10 IC (yellow circle) newly diagnosed children per the patients’ home zip code. Clustering of cases in southeast Wisconsin corresponds with population density. The inlet at the right upper corner shows the exact geographic location of Wisconsin in North America.
RESULTS Population Characteristics and Inflammatory Bowel Disease Incidence In 2000, Wisconsin had 5.4 million persons, and children <18 years accounted for 26% of the total state population. During the 24-month study period, there were 2.8 million person-years of observation, with 199 incident cases of IBD (129 CD, 60 UC, 10 IC) in the pediatric population (Fig 1). The overall incidence of IBD was calculated as 7.05 (95% CI, 6.1-8.0) per 100,000. The incidence for CD was 4.56 (95% CI, 3.77-5.35), and the incidence for UC was 2.14 (95% CI, 1.6-2.68). Figure 2 shows the age-related incidence of IBD, CD, and UC. The overall incidence of IBD was <5 per 100,000 until the age of 8 years, rising to an incidence of 13 per 100,000 by the age of 10 years. The highest age-related occurrence for CD and UC was found in the 15-year-old age group. Nineteen reported cases did not fulfill the diagnostic criteria and were excluded from the analysis. The mean age of diagnosis for IBD was 12.5 years, whereas the mean age for CD was 13.5 years, and the mean age for UC was 11.8 years (Table I). The median age for CD and UC diagnosis was 15 years. The diagnosis was most commonly made after age 10
Total no. patients Male Female Male:Female ratio Age, y (mean + SE)
CD
UC
IC
129 80 49 1.6:1 (P < .02) 13.5 ± 4.5
60 33 27 1.2:1 (P = NS) 11.8 ± 4.4
10 6 4 1.5:1 10.5 ± 3.3
years for CD and UC; however, 39 children (20% of all cases) were diagnosed at <10 years of age.
Population Density and Pediatric Inflammatory Bowel Disease Diagnosis Previous epidemiologic reports have suggested that urban populations may be at higher risk for the development of IBD.11,12 The population density (population per square mile) in Wisconsin ranged from 8.6 to 3996 among the 72 counties (Fig 3), with the pediatric population distributed evenly throughout all counties (ranging 25.1%-26.4%). Linear regression analysis demonstrated similar IBD incidence rates between sparsely and densely populated regions of the state (Fig 4).
Ethnicity and Race and Pediatric Inflammatory Bowel Disease Diagnosis The ethnic and racial distribution of the newly diagnosed cases of IBD mirrored that of the Wisconsin population and is shown in Figure 5. The incidences of CD and UC were not different between groups.
Familial Incidence of Inflammatory Bowel Disease Only 21 (11%) of the newly diagnosed IBD cases had first-degree or second-degree relatives with a history of IBD. Of these, all were white, 15 had affected parents (10 mothers, five fathers), and four had histories of multiple family members with IBD. Of the 16 children with CD, all affected
Epidemiologic and Clinical Characteristics of Children With Newly Diagnosed Inflammatory Bowel Disease in Wisconsin: A Statewide Population-Based Study
527
Fig 3. Outline of Wisconsin showing population density in 72 counties. Population density is denoted by color shading, with densest populations appearing darkest. County population density ranged from 8.6 persons per square mile to 3996 persons per square mile.
Fig 4. Rate of new-onset IBD occurrences among children in 2000 and 2001 in each of the Wisconsin counties is shown in relation to the population density. When analyzed by using linear regression, there was no relationship between the IBD rate and the population density among the 72 Wisconsin counties. The average rate, independent of the population density, is shown as a horizontal line.
relatives had histories of CD. In the five children with UC, two relatives had CD, and three had UC.
Clinical Presentation of Crohn Disease and Ulcerative Colitis in Children The anatomic location of CD involvement at the time of diagnosis was categorized by using the Vienna classification 528
Kugathasan et al
Fig 5. The percentile of new-onset pediatric IBD diagnoses in Wisconsin among the various ethnic groups (white, African American, Hispanic, Asian American, and others [including Native American]) is shown in the right panel, and the ethnic composition of all Wisconsin children (1.4 million) is shown in the left panel.
for CD.13 Isolated ileal involvement occurred in 25%, colonic disease in 32%, ileocolonic disease in 29%, and significant upper gastrointestinal tract disease in 14%. Among the 60 children with UC, left-sided colitis (involvement limited to splenic flexure) was seen at presentation in 10%, whereas pancolitis (involvement of entire colon or beyond the splenic flexure) occurred in 90%. Presenting clinical features and laboratory findings of the newly diagnosed children with CD and UC are shown in Table II. Abdominal pain was significantly more common in CD, whereas rectal bleeding was more common in UC. The mean time between the onset of symptoms and the date of diagnosis was 4.5 months for CD and 3 months for UC, which was not significantly different. Among CD children, 10 had perianal fistulae (8% of all CD diagnoses), whereas 20 (15%) had perianal disease characterized by deep fissures and distortion of the anus with several large skin tags. Extraintestinal manifestations at the time of presentation were rare. Among CD, 3% had hepatobiliary abnormalities (one had sclerosing cholangitis and four had abnormal liver enzymes), 2% had arthritis, 1% had pancreatitis, 1% had erythema nodosum, and .5% had pyoderma gangrenosum. Only one child in the UC group had multiple axial joint involvement characterized by arthritis. Crohn disease may be subcategorized based on clinical phenotypes, which may include predominantly inflammatory, fistulizing, or stricturing disease.14 In the 129 newly diagnosed children with CD, the majority (115 children, 90%) demonstrated an inflammatory disease phenotype, whereas 8% (10 children) presented with fistulizing disease, and 2% (three children) had stricturing disease requiring surgery at diagnosis. Twenty percent of patients with CD and 35% of children with UC had a normal erythrocyte sedimentation rate The Journal of Pediatrics October 2003
Table II. Clinical manifestations among the 129 children with CD and 60 children with UC at presentation
Rectal bleeding Diarrhea Abdominal pain Weight loss Fatigue Aphthous ulcers
CD n (%)
UC n (%)
55 (43) 38 (30) 86 (67) 70 (55) 17 (13) 6 (5)
50 59 26 23 1 7
(83) (98) (43) (38) (2) (13)
Table III. Laboratory findings of 129 children with CD and 60 children with UC at presentation CD (n = 129) UC (n = 60) P Mean ± SE Mean ± SE value
P value .003 .0001 .009 .02 NS NS
at the time of diagnosis (Table III). Fifty-two percent of children with CD and 68% of children with UC had normal serum albumin. There was no difference in hemoglobin, erythrocyte sedimentation rate, and serum albumin between CD and UC, but the mean corpuscular volume was lower in CD than in UC.
DISCUSSION This study reports the incidence of pediatric CD and UC in a large, defined, ethnically diverse North American population. Important observations identified in this evaluation of pediatric IBD across Wisconsin include (1) the highest pediatric IBD incidence reported in the world to date, (2) a twofold predominance in pediatric CD incidence compared with UC, (3) a significantly higher rate of CD diagnosis among boys compared with girls, (4) a low frequency of patients with positive family histories, (5) no modulatory effect of urbanization on pediatric IBD incidence, (6) equal distribution of IBD incidence among all ethnic populations, and (7) the initial presentation of pancolitis among the majority of newly diagnosed children with UC. Previous pediatric epidemiologic studies in the United States have reported disease rates exclusively from tertiary referral centers, which are not representative of the patient population at large.14 To identify all new pediatric IBD diagnoses within an entire state, voluntary reporting by all pediatric gastroenterologists serving a defined population was the best available strategy. Participating physicians were distributed throughout the state and bordering regions, so access to subspecialty care was readily available. Some patients, particularly older adolescents, were likely diagnosed by adult gastroenterologists. There are approximately 200 adult gastroenterologists in Wisconsin, and it was not practical to have them all participate in the study. Our study identified the highest incidence of pediatric IBD reported to date. Retrospective and prospective incidence studies have shown stable incidence rates in younger age groups over time until the 1980s.15 In 1989, a report from Scotland described a threefold increase in the incidence of CD among pediatric patients, from .7 per 100,000 per year in 1968 to 2.3 in 1983.7 This increase in CD was present in both boys and girls; the incidence of UC remained unchanged during this period. In follow-up studies, the increase in CD
Hemoglobin, g/dL Mean corpuscular volume, fL Erythrocyte sedimentation rate, mm/h Serum albumin, gm/dL
11.91 ± .25 76.8 ± .9 34 ± 2.5
11.83 ± .3 81.3 ± 1 30.4 ± 4
NS .001 NS
3.57 ± .1
3.63 ± .1
NS
continued, with a reported incidence of 2.8 per 100,000 per year for 1990 to 1992.16 Similar findings were reported from Wales, where CD incidence rose from 1.3 per 100,000 per year during the period 1983 to 1988 to 3.1 per 100,000 per year for the period 1989 to 1993.17 The Welsh incidence of pediatric UC remained unchanged. Sawczenko et al8 from Great Britain used a prospective design to identify all newly diagnosed cases of IBD during a 13-month period starting in June 1998. They reported an overall pediatric IBD incidence of 5.2 per 100,000 per year, with an incidence of CD of 3.1 and an incidence of UC of 1.4 per 100,000 per year. In our study, the overall incidence of IBD was 7.05, with 4.56 per 100,000 for CD and 2.14 per 100,000 for UC. These high incidence rates of pediatric IBD are possibly an underestimate of the true incidence, because we suspect that not all of the older children were captured in our voluntary reporting survey. Our data demonstrated a decline in the age-specific incidence rate after 15 years of age, which likely represents a spurious event caused by children in late adolescence being diagnosed by adult gastroenterologists and therefore failing to be captured in our study. All the 199 IBD diagnoses in our study were definitive cases fulfilling our strict diagnostic criteria. Among the 129 new CD diagnoses, 126 either had small bowel involvement or severe perianal disease with or without colonic disease or presence of granulomas in the biopsy, whereas the remaining three cases of CD had pure colonic disease characterized by serpiginous ulcers and skip lesions. The ratio of CD to UC has changed over the past 50 years, with early reports in the mid-20th century describing a predominance of UC cases in the United States and Western Europe.3 A gradual rise in CD cases in the 1960s through the 1990s resulted in CD incidence surpassing UC incidence in most Westernized countries.3,6 Studies from Japan recapitulate the US experience, with a predominance of UC cases early and a rise in CD approximately 1 decade later.18 Our measured incidence rate for pediatric CD is twice that for UC; this finding is distinctly different from that for adult populations, which traditionally report an equal incidence and prevalence of CD and UC. The predominance of CD in newly diagnosed pediatric IBD is supported by a recent prospective survey from the United Kingdom8 and studies from Scotland,7,16 Wales,17 and Sweden.19 The increased pediatric incidence does not appear to result from an earlier diagnosis of children with IBD who previously were undiagnosed until adulthood, because
Epidemiologic and Clinical Characteristics of Children With Newly Diagnosed Inflammatory Bowel Disease in Wisconsin: A Statewide Population-Based Study
529
there are no signs of a corresponding decrease in the incidence in the adult population. Our study identified a significantly higher percentage of boys compared with girls diagnosed with CD. This finding contrasts with previous studies, which have reported a slight female preponderance of CD in adult populations,20,21 but agrees with recent British8 and Canadian studies.22 The inversion of the sex ratio for CD among children compared with adults could indicate that intrinsic age-related factors play a role in the cause of CD. The most important factor identified to date that confers risk for the development of IBD is a positive family history. Compared with the general population, there is a several-fold increased risk of developing either CD or UC in the setting of a positive family history with affected firstdegree relatives.23 Previous studies have shown that the pediatric onset of CD is associated with a 30% likelihood of having affected relatives.24 In our study, only 11% of children diagnosed with IBD had a family history of IBD. Because our study focused on incidence data in newly diagnosed children, there is the possibility that additional family members may develop IBD in the future. The relatively low rates of affected family members suggest that the incidence of IBD may be increasing in populations that were not previously at high risk. Past studies have shown a high incidence of IBD in Caucasian children, but we did not observe any differences across racial groups, including minority groups previously thought to be at lower risk for disease. However, none of the non-Caucasian children with IBD had family histories of IBD. These data, in conjunction with our equivalent incidence rates between sparsely and densely populated areas of the state, suggest that IBD is now a condition equally afflicting all populations within our region. The anatomic distribution of UC is limited to the left colon in approximately 36% of adult patients.25 At least two retrospective studies undertaken between the years of 1962 and 1994 have described the clinical presentation and the course of pediatric UC. In a US study, Hyams et al26 described a cohort of 171 children with UC diagnosed between the ages of 1.5 and 17.7 years. Only 43% of these children had pancolitis, and the remaining 57% had disease limited to the left side of the colon. In a European study of 80 children with UC, only 29% had pancolitis.27 In our study, 90% of children with UC had pancolitis at the time of diagnosis. The factors that account for the reversed incidence of pancolitis in our study are not clear. All the children included in our analysis underwent full colonoscopy. The higher rates of pancolitis diagnosed in our study may reflect previous underdiagnosis in children. Currently, pediatric gastroenterologists are more likely to use full colonoscopy instead of flexible sigmoidoscopy in the evaluation of children with chronic diarrhea and rectal bleeding, and this may provide a plausible explanation for the observed changes in the regional involvement. Among children with CD, the anatomical distribution of disease involvement and the presenting features are similar to those previously reported in adults13,28 and children.29 However, there were a few subtle differences in our study. We found that the most commonly 530
Kugathasan et al
affected anatomic segment in CD was the colon rather than the ileum. We also found that higher proportions of children with CD had upper tract involvement, which may be related to the routine use of upper endoscopy as the initial investigatory procedure in conjunction with colonoscopy in 92% of the children with CD in our series. Only a few children with IBD had extraintestinal manifestations at presentation, suggesting that the majority of the extraintestinal manifestations are seen in patients with long-standing IBD and may represent complications of chronic, uncontrolled intestinal inflammation. The pathogenesis of CD and UC is believed to involve interplay of environmental factors acting on genetically susceptible persons.30 Our study suggests that environmental factors are dominant in changing the pattern of disease, because the genetic make-up of human populations does not shift over short periods. A parallel phenomenon is the dramatic increase in asthma during the same period in the West.31,32 The concomitant emergence of chronic inflammation in the lung and gut also supports the concept that changing environmental factors play a pivotal role in the increased frequency of these disorders in children. In summary, our study provides new epidemiologic data regarding the incidence and characteristics of pediatric IBD within a well defined population. Overall, the results of this study suggest that the clinical spectrum of IBD continues to evolve, because pediatric IBD now afflicts all races and ethnicities as well as rural and urban populations with a similar frequency, pointing to changing environmental factors as increasingly important in the pathogenesis of IBD. We thank Promod Yannemsetty and Prasad Kothembaka for technical assistance and database development, Carolyn Raasch for data collection, and Dr. Theodore M. Kotchen for critical review of the manuscript. Finally, we thank Jan Lenz, executive director, Wisconsin Chapter, Crohn’s and Colitis Foundation of America, and Nadine Davis, Head, Wisconsin CCFA Youth Initiative, for their support of this project.
REFERENCES 1. Shanahan F. Inflammatory bowel disease: immunodiagnostics, immunotherapeutics, and ecotherapeutics. Gastroenterology 2001;120:622-35. 2. Podolsky DK. Inflammatory bowel disease. N Engl J Med 2002; 347:417-29. 3. Lashner BA. Epidemiology of inflammatory bowel disease. Gastroenterol Clin North Am 1995;24:467-74. 4. Calkins BM, Lilienfeld AM, Garland CF, Mendeloff AI. Trends in incidence rates of ulcerative colitis and Crohn’s disease. Dig Dis Sci 1984;29:913-20. 5. Miller DS, Kieghley AC, Langman MJ. Changing pattern in the epidemiology of Crohn’s disease. Lancet 1974;21:691-3. 6. Munkholm P, Langholz E, Nielsen OH, Kreiner S, Binder V. Incidence and prevalence of Crohn’s disease in the county of Copenhagen, 1962-87: a sixfold increase in incidence. Scand J Gastroenterol 1992;27:609-14. 7. Barton JR, Gillon S, Ferguson A. Incidence of inflammatory bowel disease in Scottish children between 1968 and 1983: marginal fall in ulcerative colitis, three-fold rise in Crohn’s disease. Gut 1989;30:618-22. 8. Sawczenko A, Sandhu BK, Logan RF, Jenkins H, Taylor CJ, Mian S, et al. Prospective survey of childhood inflammatory bowel disease in the British Isles. Lancet 2001;357:1093-4.
The Journal of Pediatrics October 2003
9. Hilderbrand H, Fredrikzon B, Holmquist L, Kristiansson B, Lindquist B. Chronic inflammatory bowel disease in children and adolescents in Sweden. J Pediatr Gastroenterol Nutr 1991;13:293-7. 10. Antonioli DA. Colitis in infants and children. In: Dahms BB, editor. Gastrointestinal diseases. Basel, Switzerland: Karger; 1997. p. 77-110. 11. Sandler RS. Epidemiology of inflammatory bowel disease. In: Targan SR, Shanahan F, editors. Inflammatory bowel disease: from bench to bedside. Baltimore: Williams & Wilkins; 1994. p. 5-30. 12. Sonnenberg A, McCarty DJ, Jacobsen SJ. Geographic variation of inflammatory bowel disease within the United States. Gastroenterology 1991;100:143-9. 13. Gasche C, Scholmerich J, Brynskov J, D’Haens G, Hanauer SB, Irvine EJ, et al. A simple classification of Crohn’s disease: report of the Working Party for the World Congresses of Gastroenterology, Vienna 1998. Inflamm Bowel Dis 2000;6:8-15. 14. Polito JM II, Childs B, Mellits ED, Tokayer AZ, Harris ML, Bayless TM Crohn’s disease: influence of age at diagnosis on site and clinical type of disease. Gastroenterology 1996;111:580-6. 15. Lindquist BL, Jarnerot G, Wickbom G. Clinical and epidemiological aspects of Crohn’s disease in children and adolescents. Scand J Gastroenterol 1984;19:502-6. 16. Armitage E, Drummond H, Ghosh S, Ferguson A. Incidence of juvenile-onset Crohn’s disease in Scotland. Lancet 1999;353:1496-7. 17. Cosgrove M, Al-Atia RF, Jenkins HR. The epidemiology of paediatric inflammatory bowel disease. Arch Dis Child 1996;74:460-1. 18. Yang SK, Loftus EV Jr, Sandborn WJ. Epidemiology of inflammatory bowel disease in Asia. Inflamm Bowel Dis 2001;7:260-70. 19. Lindberg E, Lindquist B, Holmquist L, Hildebrand H. Inflammatory bowel disease in children and adolescents in Sweden, 1984-1995. J Pediatr Gastroenterol Nutr 2000;30:259-64. 20. Calkins BM, Mendeloff AI. Epidemiology of inflammatory bowel disease. Epidemiol Rev 1986;8:60-91. 21. Russell MG, Stockbrugger RW. Epidemiology of inflammatory bowel disease: an update. Scand J Gastroenterol 1996;31:417-27. 22. Bernstein CN, Blanchard JF, Rawsthorne P, Wajda A. Epidemiology of Crohn’s disease and ulcerative colitis in a central Canadian province: a population based study. Am J Epidemiol 1999;149:916-24. 23. Peeters M, Nevens H, Baert F, Hiele M, de Meyer AM, Vlietinck R, et al. Familial aggregation in Crohn’s disease: increased age-adjusted risk
and concordance in clinical characteristics. Gastroenterology 1996;111: 597-603. 24. Feeney MA, Murphy F, Clegg AJ, Trebble TM, Sharer NM, Snook JA. A case-control study of childhood environmental risk factors for the development of inflammatory bowel disease. Eur J Gastroenterol Hepatol 2002;14:529-34. 25. Farmer RG, Easley K, Rankin GB. Clinical patterns, natural history and progression of ulcerative colitis: a long-term follow-up of 1116 patients. Dig Dis Sci 1993;38:1137-46. 26. Hyams JS, Davis P, Grancher K, Lerer T, Juntinich CJ, Markowitz J. Clinical outcome of ulcerative colitis in children. J Pediatr 1996;129:81-8. 27. Langholz E, Munkholm P, Davidsen M, Binder V. Course of ulcerative colitis: analysis of changes in disease activity over years. Gastroenterology 1994;107:3-11. 28. Ferguson A, Sedgwick DM. Juvenile-onset inflammatory bowel disease: predictors of morbidity and health status in early adult life. J R Coll Physicians Lond 1994;28:220-7. 29. Langholz E, Munkholm P, Krasilnikoff PA, Binder V. Inflammatory bowel diseases with onset in childhood: clinical features, morbidity, and mortality in a regional cohort. Scand J Gastroenterol 1997;32:139-47. 30. Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis. Gastroenterology 1998;115:182-205. 31. Myers TR. Pediatric asthma epidemiology: incidence, morbidity, and mortality. Respir Care Clin N Am 2000;6:1-14. 32. Smyth RL. Asthma: a major pediatric health issue. Respir Res 2002;3(suppl 1):S3-7.
APPENDIX Wisconsin Pediatric Inflammatory Bowel Disease Alliance: Ellen Blank, Christopher Brown, David Deusch, Istvan Danko, Glen Gourley, T. S. Gunasekaran, Issam Halabi, Janice Heikenen, Robert H. Judd, Farhat Khan, Subra Kugathasan, Alfonso Martinez, Roger Park, Jean Perrault, Maria T. Rivera-Bennett, Colin D. Rudolph, William San Pablo Jr., Gregorz Telega, Steven Werlin, Sally WeisdorfShendle, and Michael Yaffe.
CORRECTION The article, ‘‘Temporal variability in birth prevalence of congenital heart defects as recorded by a general birth defects registry,’’ by Bosi et al, which appeared in the June 2003 issue of The Journal (volume 142, pages 690-8) contains several errors. The following should be ‘‘per thousand’’and not ‘‘%’’: P 690 Results (in abstract): 5.1 (3.1-7.5) Opening paragraph: 3.5, 13.7, 4.5, 3.2, and 6 P 693 Second column, first paragraph in RESULTS: 7.5, 5.1, and 3.1 P 694 Second column, fourth paragraph: 8.3, 6.2, and 10.7 First paragraph in DISCUSSION: 3.5 to 13.7 P 695 First column, third paragraph: 5.1 and 3.1 to 7.5 P 697 First column: first paragraph, 6.2 to 10.7; second paragraph, 11; fifth paragraph, 8 to 10, 4 to 6 Epidemiologic and Clinical Characteristics of Children With Newly Diagnosed Inflammatory Bowel Disease in Wisconsin: A Statewide Population-Based Study
531
Objective To define epidemiologic and clinical characteristics of newly diagnosed pediatric inflammatory bowel disease (IBD) in a large population-based model. Study design All pediatric gastroenterologists providing care for Wisconsin children voluntarily identified all new cases of IBD during a 2-year period. Demographic and clinical data were sent to a central registry prospectively for analysis. Results The incidence of IBD in Wisconsin children was 7.05 per 100,000, whereas the incidence for Crohn’s disease was 4.56, more than twice the rate of ulcerative colitis (2.14). An equal IBD incidence occurred among all ethnic groups, and children from sparsely and densely populated counties were equally affected. The majority (89%) of new IBD diagnoses were nonfamilial. Conclusions
This study provides novel, prospective, and comprehensive information on pediatric IBD incidence within the United States. The surprisingly high incidence of pediatric IBD, the predominance of Crohn’s disease over ulcerative colitis, the low frequency of patients with a family history, the equal distribution of IBD among all racial and ethnic groups, and the lack of a modulatory effect of urbanization on IBD incidence collectively suggest that the clinical spectrum of IBD is still evolving and point to environmental factors contributing to the pathogenesis. (J Pediatr 2003;143:525-31)
rohn’s disease (CD) and ulcerative colitis (UC), collectively known as inflammatory bowel disease (IBD), are idiopathic, lifelong, destructive chronic inflammatory conditions of the gastrointestinal tract. CD and UC were recognized as disease entities during the early 20th century, and pathogenesis has been linked to a combination of genetic and environmental factors.1,2 Population-based studies suggest that IBD is unevenly distributed throughout the world, with the highest disease rates occurring in first world, Westernized countries.3 Epidemiologic surveys have also suggested that IBD incidence rates have changed over the second half of the 20th century, with a steady, gradual increase for both UC and CD that appears to be independent of observer bias.4-6 Because increasingly high rates of IBD onset occur in childhood and adolescence, monitoring the incidence of pediatric IBD may reflect the changing trends in IBD demographics. Recent retrospective European studies and a single prospective populationbased survey from Great Britain have suggested that the incidence of IBD in children and adolescents has significantly increased over the last 35 years. These pediatric studies also suggest that there has been a change in the patterns of disease, because the incidence of CD has risen above that of UC.7,8 So far, there have been no systematic, population-based
C
CD CI IBD
Crohn’s disease Confidence interval Inflammatory bowel disease
IC UC
Indeterminate colitis Ulcerative colitis
From Division of Pediatric Gastroenterology and Nutrition, the Department of Epidemiology and Biostatistics, and the Division of Gastroenterology and Hepatology, Medical College of Wisconsin, Milwaukee; Children’s Hospital of Wisconsin, Milwaukee; the Division of Pediatric Gastroenterology, University of Wisconsin—Madison, School of Medicine, Madison; Dean Clinic, Madison; the Department of Pediatrics, Marshfield Clinic, Marshfield; Gunderson Clinic, La Crosse; and Gastroenterology Associates, Wausau, Wisconsin; and the Division of Pediatric Gastroenterology, University of Minnesota, Minneapolis; and the Division of Pediatric Gastroenterology, Mayo Clinic, Rochester, Minnesota. Supported by National Institutes of Health grant (RR016111, S. K.), the National Institutes of Health Clinical Scholars program (K30HL004111, S. K.), and the general clinical research center at the Medical College of Wisconsin and the Medical College of Wisconsin’s Digestive Disease Center (S. K., D. G. B.). Submitted for publication Feb 21, 2003; revision received June 18, 2003; accepted July 23, 2003. Reprint requests: Subra Kugathasan, MD, Department of Pediatrics, Medical College of Wisconsin, 8701 Watertown Plank Rd, Milwaukee, WI 53226. E-mail: [email protected]. Copyright ª 2003 Mosby, Inc. All rights reserved. 0022-3476/2003/$30.00 + 0 10.1067/S0022-3476(03)00444-X
525
studies in North America on the incidence of childhood IBD, because IBD is not a reportable condition, and there is no regional or national registry for tracking each IBD diagnosis within the United States. Our aim was to determine the pediatric IBD incidence and to define clinical characteristics from a statewide-defined population.
MATERIALS AND METHODS STUDY SETTING AND FORMATION OF THE WISCONSIN PEDIATRIC INFLAMMATORY BOWEL DISEASE ALLIANCE. Identification of new pediatric IBD diagnoses was based on voluntary reporting from pediatric gastroenterology subspecialists. All pediatric gastroenterologists within Wisconsin (n = 17) and those from neighboring states bordering the southern (Illinois, n = 2) and western (Minnesota, n = 2) areas of Wisconsin, who might provide care for resident children, participated in the study. These physicians were based in three academic institutions and five community-based practices within Wisconsin, two academic institutions from Minnesota, and one academic and one community-based practice in Illinois.
Recruitment of Patients and Data Collection Methods All children who were <18 years old at the time of diagnosis of IBD whose residential address was a Wisconsin zip code were eligible for inclusion in the study. Patient accrual began prospectively on January 1, 2000, and continued for a 2-year period ending December 31, 2001. A questionnaire for each new subject was completed by the pediatric gastroenterologist who made the IBD diagnosis. The study was conducted in Wisconsin, an upper midwestern state in the United States. The 2000 US Census Bureau data defining the total state population, pediatric population (age <18 years), county populations, and ethnic backgrounds were used in this analysis. Wisconsin has 72 counties, and the population of counties in the year 2000 was used to determine the distribution of population density throughout the state. Patients’ residential zip codes were used to determine county residence. Coded data were entered in a computerized database by using patient initials, date of birth, zip code, and the diagnosing physician’s name. Data recorded included clinical presenting symptoms, duration of each symptom, and family history of IBD. The date of diagnosis was set as the day of the definitive diagnostic procedure rather than symptom onset. All data from the time of the diagnostic work-up, including the laboratory findings, serologic tests, radiologic studies, endoscopic evaluation, and histologic findings, were obtained and entered into the central database. Any surgical intervention needed at the time of diagnosis was also recorded.
Criteria for Inflammatory Bowel Disease Diagnosis and Disease Subcategorization The diagnosis of IBD required endoscopic evaluation, including histologic assessment of mucosal pinch biopsies. All patients underwent colonoscopic examination with or without upper endoscopy with multiple mucosal biopsies for review by 526
Kugathasan et al
clinical pathologists. Stool studies were performed in all patients to exclude infectious causes of diarrheal illness. Patients whose IBD diagnosis was not confirmed by endoscopic or histologic examination were excluded from further analysis. Serologic testing of antineutrophil or anti-Saccharomyces cerevisiae antibodies, which correlate with an IBD diagnosis, was not considered a sufficient criterion for a diagnosis of IBD. New pediatric IBD cases were further subcategorized into three clinical groups—CD, UC, and indeterminate colitis (IC)—defined according to established clinical, biochemical, radiologic, endoscopic, and histologic criteria.9,10 Children with nonspecific chronic inflammation without firm evidence of IBD were excluded from the analysis. CD was defined as the presence of histologically confirmed discontinuous chronic inflammation of the gastrointestinal tract confirmed by endoscopy and supported by clinical, biochemical, and radiologic evidence of CD. UC was defined as the presence of continuous inflammation limited to the colon, extending proximally from the involved rectum, with histologically typical chronic inflammation limited to the mucosa, not extending beyond the muscularis mucosa. In all cases, the terminal ileum was evaluated by small bowel followthrough barium radiographs or colonoscopy and ileoscopy. IC was defined as inflammatory colitis in the setting of histopathological changes indicative of chronic IBD colitis, containing both endoscopic and microscopic findings that were consistent with both CD and UC. The presence of continuous endoscopic disease in the setting of either discontinuous microscopic disease or inflammation extending beyond the muscularis mucosa was included in the indeterminate category. In addition, patients with IC had no radiologic, endoscopic, or histologic evidence of small bowel disease. Anatomic regions of CD involvement at diagnosis, with the exception of jejunal and perianal CD, were confirmed by microscopic evidence of CD chronic inflammation. The presence of nonspecific inflammation was not a sufficient criterion for the diagnosis of upper gastrointestinal tract CD. In UC cases with inflammation limited to the left side of the colon and not extending proximally beyond the splenic flexure were categorized as left-sided colitis. Pancolitis was defined as gross and microscopic evidence of chronic inflammation extending proximally beyond the splenic flexure.
Analysis of Incidence Rates, Census Data, and Statistics Incidence rates were calculated per million pediatric populations per year, and the 95% confidence intervals (CIs) were estimated assuming a Poisson distribution of cases. Linear regression was used for multivariate analyses to calculate adjusted incidence rate ratios among the Wisconsin counties. Paired and unpaired t tests and v2 tests were used to determine significance. The Kolmogorov-Smirnov test was used to compare age of diagnosis, because the age distribution was skewed. This study was approved by the research and publication committee of Children’s Hospital of Wisconsin and the Medical College of Wisconsin. The Journal of Pediatrics October 2003
Fig 2. Age-related occurrence of all new-onset pediatric IBD, CD, and UC in Wisconsin between 2000 and 2001.
Table I. Demographics of newly diagnosed children with IBD children from January 1, 2000, to December 31, 2001
Fig 1. Distribution of new onset pediatric IBD cases in Wisconsin during the 24-month period from January 1, 2000, to December 31, 2001. Outline of the state of Wisconsin shows the localization of 129 CD (red circle), 60 UC (blue star), and 10 IC (yellow circle) newly diagnosed children per the patients’ home zip code. Clustering of cases in southeast Wisconsin corresponds with population density. The inlet at the right upper corner shows the exact geographic location of Wisconsin in North America.
RESULTS Population Characteristics and Inflammatory Bowel Disease Incidence In 2000, Wisconsin had 5.4 million persons, and children <18 years accounted for 26% of the total state population. During the 24-month study period, there were 2.8 million person-years of observation, with 199 incident cases of IBD (129 CD, 60 UC, 10 IC) in the pediatric population (Fig 1). The overall incidence of IBD was calculated as 7.05 (95% CI, 6.1-8.0) per 100,000. The incidence for CD was 4.56 (95% CI, 3.77-5.35), and the incidence for UC was 2.14 (95% CI, 1.6-2.68). Figure 2 shows the age-related incidence of IBD, CD, and UC. The overall incidence of IBD was <5 per 100,000 until the age of 8 years, rising to an incidence of 13 per 100,000 by the age of 10 years. The highest age-related occurrence for CD and UC was found in the 15-year-old age group. Nineteen reported cases did not fulfill the diagnostic criteria and were excluded from the analysis. The mean age of diagnosis for IBD was 12.5 years, whereas the mean age for CD was 13.5 years, and the mean age for UC was 11.8 years (Table I). The median age for CD and UC diagnosis was 15 years. The diagnosis was most commonly made after age 10
Total no. patients Male Female Male:Female ratio Age, y (mean + SE)
CD
UC
IC
129 80 49 1.6:1 (P < .02) 13.5 ± 4.5
60 33 27 1.2:1 (P = NS) 11.8 ± 4.4
10 6 4 1.5:1 10.5 ± 3.3
years for CD and UC; however, 39 children (20% of all cases) were diagnosed at <10 years of age.
Population Density and Pediatric Inflammatory Bowel Disease Diagnosis Previous epidemiologic reports have suggested that urban populations may be at higher risk for the development of IBD.11,12 The population density (population per square mile) in Wisconsin ranged from 8.6 to 3996 among the 72 counties (Fig 3), with the pediatric population distributed evenly throughout all counties (ranging 25.1%-26.4%). Linear regression analysis demonstrated similar IBD incidence rates between sparsely and densely populated regions of the state (Fig 4).
Ethnicity and Race and Pediatric Inflammatory Bowel Disease Diagnosis The ethnic and racial distribution of the newly diagnosed cases of IBD mirrored that of the Wisconsin population and is shown in Figure 5. The incidences of CD and UC were not different between groups.
Familial Incidence of Inflammatory Bowel Disease Only 21 (11%) of the newly diagnosed IBD cases had first-degree or second-degree relatives with a history of IBD. Of these, all were white, 15 had affected parents (10 mothers, five fathers), and four had histories of multiple family members with IBD. Of the 16 children with CD, all affected
Epidemiologic and Clinical Characteristics of Children With Newly Diagnosed Inflammatory Bowel Disease in Wisconsin: A Statewide Population-Based Study
527
Fig 3. Outline of Wisconsin showing population density in 72 counties. Population density is denoted by color shading, with densest populations appearing darkest. County population density ranged from 8.6 persons per square mile to 3996 persons per square mile.
Fig 4. Rate of new-onset IBD occurrences among children in 2000 and 2001 in each of the Wisconsin counties is shown in relation to the population density. When analyzed by using linear regression, there was no relationship between the IBD rate and the population density among the 72 Wisconsin counties. The average rate, independent of the population density, is shown as a horizontal line.
relatives had histories of CD. In the five children with UC, two relatives had CD, and three had UC.
Clinical Presentation of Crohn Disease and Ulcerative Colitis in Children The anatomic location of CD involvement at the time of diagnosis was categorized by using the Vienna classification 528
Kugathasan et al
Fig 5. The percentile of new-onset pediatric IBD diagnoses in Wisconsin among the various ethnic groups (white, African American, Hispanic, Asian American, and others [including Native American]) is shown in the right panel, and the ethnic composition of all Wisconsin children (1.4 million) is shown in the left panel.
for CD.13 Isolated ileal involvement occurred in 25%, colonic disease in 32%, ileocolonic disease in 29%, and significant upper gastrointestinal tract disease in 14%. Among the 60 children with UC, left-sided colitis (involvement limited to splenic flexure) was seen at presentation in 10%, whereas pancolitis (involvement of entire colon or beyond the splenic flexure) occurred in 90%. Presenting clinical features and laboratory findings of the newly diagnosed children with CD and UC are shown in Table II. Abdominal pain was significantly more common in CD, whereas rectal bleeding was more common in UC. The mean time between the onset of symptoms and the date of diagnosis was 4.5 months for CD and 3 months for UC, which was not significantly different. Among CD children, 10 had perianal fistulae (8% of all CD diagnoses), whereas 20 (15%) had perianal disease characterized by deep fissures and distortion of the anus with several large skin tags. Extraintestinal manifestations at the time of presentation were rare. Among CD, 3% had hepatobiliary abnormalities (one had sclerosing cholangitis and four had abnormal liver enzymes), 2% had arthritis, 1% had pancreatitis, 1% had erythema nodosum, and .5% had pyoderma gangrenosum. Only one child in the UC group had multiple axial joint involvement characterized by arthritis. Crohn disease may be subcategorized based on clinical phenotypes, which may include predominantly inflammatory, fistulizing, or stricturing disease.14 In the 129 newly diagnosed children with CD, the majority (115 children, 90%) demonstrated an inflammatory disease phenotype, whereas 8% (10 children) presented with fistulizing disease, and 2% (three children) had stricturing disease requiring surgery at diagnosis. Twenty percent of patients with CD and 35% of children with UC had a normal erythrocyte sedimentation rate The Journal of Pediatrics October 2003
Table II. Clinical manifestations among the 129 children with CD and 60 children with UC at presentation
Rectal bleeding Diarrhea Abdominal pain Weight loss Fatigue Aphthous ulcers
CD n (%)
UC n (%)
55 (43) 38 (30) 86 (67) 70 (55) 17 (13) 6 (5)
50 59 26 23 1 7
(83) (98) (43) (38) (2) (13)
Table III. Laboratory findings of 129 children with CD and 60 children with UC at presentation CD (n = 129) UC (n = 60) P Mean ± SE Mean ± SE value
P value .003 .0001 .009 .02 NS NS
at the time of diagnosis (Table III). Fifty-two percent of children with CD and 68% of children with UC had normal serum albumin. There was no difference in hemoglobin, erythrocyte sedimentation rate, and serum albumin between CD and UC, but the mean corpuscular volume was lower in CD than in UC.
DISCUSSION This study reports the incidence of pediatric CD and UC in a large, defined, ethnically diverse North American population. Important observations identified in this evaluation of pediatric IBD across Wisconsin include (1) the highest pediatric IBD incidence reported in the world to date, (2) a twofold predominance in pediatric CD incidence compared with UC, (3) a significantly higher rate of CD diagnosis among boys compared with girls, (4) a low frequency of patients with positive family histories, (5) no modulatory effect of urbanization on pediatric IBD incidence, (6) equal distribution of IBD incidence among all ethnic populations, and (7) the initial presentation of pancolitis among the majority of newly diagnosed children with UC. Previous pediatric epidemiologic studies in the United States have reported disease rates exclusively from tertiary referral centers, which are not representative of the patient population at large.14 To identify all new pediatric IBD diagnoses within an entire state, voluntary reporting by all pediatric gastroenterologists serving a defined population was the best available strategy. Participating physicians were distributed throughout the state and bordering regions, so access to subspecialty care was readily available. Some patients, particularly older adolescents, were likely diagnosed by adult gastroenterologists. There are approximately 200 adult gastroenterologists in Wisconsin, and it was not practical to have them all participate in the study. Our study identified the highest incidence of pediatric IBD reported to date. Retrospective and prospective incidence studies have shown stable incidence rates in younger age groups over time until the 1980s.15 In 1989, a report from Scotland described a threefold increase in the incidence of CD among pediatric patients, from .7 per 100,000 per year in 1968 to 2.3 in 1983.7 This increase in CD was present in both boys and girls; the incidence of UC remained unchanged during this period. In follow-up studies, the increase in CD
Hemoglobin, g/dL Mean corpuscular volume, fL Erythrocyte sedimentation rate, mm/h Serum albumin, gm/dL
11.91 ± .25 76.8 ± .9 34 ± 2.5
11.83 ± .3 81.3 ± 1 30.4 ± 4
NS .001 NS
3.57 ± .1
3.63 ± .1
NS
continued, with a reported incidence of 2.8 per 100,000 per year for 1990 to 1992.16 Similar findings were reported from Wales, where CD incidence rose from 1.3 per 100,000 per year during the period 1983 to 1988 to 3.1 per 100,000 per year for the period 1989 to 1993.17 The Welsh incidence of pediatric UC remained unchanged. Sawczenko et al8 from Great Britain used a prospective design to identify all newly diagnosed cases of IBD during a 13-month period starting in June 1998. They reported an overall pediatric IBD incidence of 5.2 per 100,000 per year, with an incidence of CD of 3.1 and an incidence of UC of 1.4 per 100,000 per year. In our study, the overall incidence of IBD was 7.05, with 4.56 per 100,000 for CD and 2.14 per 100,000 for UC. These high incidence rates of pediatric IBD are possibly an underestimate of the true incidence, because we suspect that not all of the older children were captured in our voluntary reporting survey. Our data demonstrated a decline in the age-specific incidence rate after 15 years of age, which likely represents a spurious event caused by children in late adolescence being diagnosed by adult gastroenterologists and therefore failing to be captured in our study. All the 199 IBD diagnoses in our study were definitive cases fulfilling our strict diagnostic criteria. Among the 129 new CD diagnoses, 126 either had small bowel involvement or severe perianal disease with or without colonic disease or presence of granulomas in the biopsy, whereas the remaining three cases of CD had pure colonic disease characterized by serpiginous ulcers and skip lesions. The ratio of CD to UC has changed over the past 50 years, with early reports in the mid-20th century describing a predominance of UC cases in the United States and Western Europe.3 A gradual rise in CD cases in the 1960s through the 1990s resulted in CD incidence surpassing UC incidence in most Westernized countries.3,6 Studies from Japan recapitulate the US experience, with a predominance of UC cases early and a rise in CD approximately 1 decade later.18 Our measured incidence rate for pediatric CD is twice that for UC; this finding is distinctly different from that for adult populations, which traditionally report an equal incidence and prevalence of CD and UC. The predominance of CD in newly diagnosed pediatric IBD is supported by a recent prospective survey from the United Kingdom8 and studies from Scotland,7,16 Wales,17 and Sweden.19 The increased pediatric incidence does not appear to result from an earlier diagnosis of children with IBD who previously were undiagnosed until adulthood, because
Epidemiologic and Clinical Characteristics of Children With Newly Diagnosed Inflammatory Bowel Disease in Wisconsin: A Statewide Population-Based Study
529
there are no signs of a corresponding decrease in the incidence in the adult population. Our study identified a significantly higher percentage of boys compared with girls diagnosed with CD. This finding contrasts with previous studies, which have reported a slight female preponderance of CD in adult populations,20,21 but agrees with recent British8 and Canadian studies.22 The inversion of the sex ratio for CD among children compared with adults could indicate that intrinsic age-related factors play a role in the cause of CD. The most important factor identified to date that confers risk for the development of IBD is a positive family history. Compared with the general population, there is a several-fold increased risk of developing either CD or UC in the setting of a positive family history with affected firstdegree relatives.23 Previous studies have shown that the pediatric onset of CD is associated with a 30% likelihood of having affected relatives.24 In our study, only 11% of children diagnosed with IBD had a family history of IBD. Because our study focused on incidence data in newly diagnosed children, there is the possibility that additional family members may develop IBD in the future. The relatively low rates of affected family members suggest that the incidence of IBD may be increasing in populations that were not previously at high risk. Past studies have shown a high incidence of IBD in Caucasian children, but we did not observe any differences across racial groups, including minority groups previously thought to be at lower risk for disease. However, none of the non-Caucasian children with IBD had family histories of IBD. These data, in conjunction with our equivalent incidence rates between sparsely and densely populated areas of the state, suggest that IBD is now a condition equally afflicting all populations within our region. The anatomic distribution of UC is limited to the left colon in approximately 36% of adult patients.25 At least two retrospective studies undertaken between the years of 1962 and 1994 have described the clinical presentation and the course of pediatric UC. In a US study, Hyams et al26 described a cohort of 171 children with UC diagnosed between the ages of 1.5 and 17.7 years. Only 43% of these children had pancolitis, and the remaining 57% had disease limited to the left side of the colon. In a European study of 80 children with UC, only 29% had pancolitis.27 In our study, 90% of children with UC had pancolitis at the time of diagnosis. The factors that account for the reversed incidence of pancolitis in our study are not clear. All the children included in our analysis underwent full colonoscopy. The higher rates of pancolitis diagnosed in our study may reflect previous underdiagnosis in children. Currently, pediatric gastroenterologists are more likely to use full colonoscopy instead of flexible sigmoidoscopy in the evaluation of children with chronic diarrhea and rectal bleeding, and this may provide a plausible explanation for the observed changes in the regional involvement. Among children with CD, the anatomical distribution of disease involvement and the presenting features are similar to those previously reported in adults13,28 and children.29 However, there were a few subtle differences in our study. We found that the most commonly 530
Kugathasan et al
affected anatomic segment in CD was the colon rather than the ileum. We also found that higher proportions of children with CD had upper tract involvement, which may be related to the routine use of upper endoscopy as the initial investigatory procedure in conjunction with colonoscopy in 92% of the children with CD in our series. Only a few children with IBD had extraintestinal manifestations at presentation, suggesting that the majority of the extraintestinal manifestations are seen in patients with long-standing IBD and may represent complications of chronic, uncontrolled intestinal inflammation. The pathogenesis of CD and UC is believed to involve interplay of environmental factors acting on genetically susceptible persons.30 Our study suggests that environmental factors are dominant in changing the pattern of disease, because the genetic make-up of human populations does not shift over short periods. A parallel phenomenon is the dramatic increase in asthma during the same period in the West.31,32 The concomitant emergence of chronic inflammation in the lung and gut also supports the concept that changing environmental factors play a pivotal role in the increased frequency of these disorders in children. In summary, our study provides new epidemiologic data regarding the incidence and characteristics of pediatric IBD within a well defined population. Overall, the results of this study suggest that the clinical spectrum of IBD continues to evolve, because pediatric IBD now afflicts all races and ethnicities as well as rural and urban populations with a similar frequency, pointing to changing environmental factors as increasingly important in the pathogenesis of IBD. We thank Promod Yannemsetty and Prasad Kothembaka for technical assistance and database development, Carolyn Raasch for data collection, and Dr. Theodore M. Kotchen for critical review of the manuscript. Finally, we thank Jan Lenz, executive director, Wisconsin Chapter, Crohn’s and Colitis Foundation of America, and Nadine Davis, Head, Wisconsin CCFA Youth Initiative, for their support of this project.
REFERENCES 1. Shanahan F. Inflammatory bowel disease: immunodiagnostics, immunotherapeutics, and ecotherapeutics. Gastroenterology 2001;120:622-35. 2. Podolsky DK. Inflammatory bowel disease. N Engl J Med 2002; 347:417-29. 3. Lashner BA. Epidemiology of inflammatory bowel disease. Gastroenterol Clin North Am 1995;24:467-74. 4. Calkins BM, Lilienfeld AM, Garland CF, Mendeloff AI. Trends in incidence rates of ulcerative colitis and Crohn’s disease. Dig Dis Sci 1984;29:913-20. 5. Miller DS, Kieghley AC, Langman MJ. Changing pattern in the epidemiology of Crohn’s disease. Lancet 1974;21:691-3. 6. Munkholm P, Langholz E, Nielsen OH, Kreiner S, Binder V. Incidence and prevalence of Crohn’s disease in the county of Copenhagen, 1962-87: a sixfold increase in incidence. Scand J Gastroenterol 1992;27:609-14. 7. Barton JR, Gillon S, Ferguson A. Incidence of inflammatory bowel disease in Scottish children between 1968 and 1983: marginal fall in ulcerative colitis, three-fold rise in Crohn’s disease. Gut 1989;30:618-22. 8. Sawczenko A, Sandhu BK, Logan RF, Jenkins H, Taylor CJ, Mian S, et al. Prospective survey of childhood inflammatory bowel disease in the British Isles. Lancet 2001;357:1093-4.
The Journal of Pediatrics October 2003
9. Hilderbrand H, Fredrikzon B, Holmquist L, Kristiansson B, Lindquist B. Chronic inflammatory bowel disease in children and adolescents in Sweden. J Pediatr Gastroenterol Nutr 1991;13:293-7. 10. Antonioli DA. Colitis in infants and children. In: Dahms BB, editor. Gastrointestinal diseases. Basel, Switzerland: Karger; 1997. p. 77-110. 11. Sandler RS. Epidemiology of inflammatory bowel disease. In: Targan SR, Shanahan F, editors. Inflammatory bowel disease: from bench to bedside. Baltimore: Williams & Wilkins; 1994. p. 5-30. 12. Sonnenberg A, McCarty DJ, Jacobsen SJ. Geographic variation of inflammatory bowel disease within the United States. Gastroenterology 1991;100:143-9. 13. Gasche C, Scholmerich J, Brynskov J, D’Haens G, Hanauer SB, Irvine EJ, et al. A simple classification of Crohn’s disease: report of the Working Party for the World Congresses of Gastroenterology, Vienna 1998. Inflamm Bowel Dis 2000;6:8-15. 14. Polito JM II, Childs B, Mellits ED, Tokayer AZ, Harris ML, Bayless TM Crohn’s disease: influence of age at diagnosis on site and clinical type of disease. Gastroenterology 1996;111:580-6. 15. Lindquist BL, Jarnerot G, Wickbom G. Clinical and epidemiological aspects of Crohn’s disease in children and adolescents. Scand J Gastroenterol 1984;19:502-6. 16. Armitage E, Drummond H, Ghosh S, Ferguson A. Incidence of juvenile-onset Crohn’s disease in Scotland. Lancet 1999;353:1496-7. 17. Cosgrove M, Al-Atia RF, Jenkins HR. The epidemiology of paediatric inflammatory bowel disease. Arch Dis Child 1996;74:460-1. 18. Yang SK, Loftus EV Jr, Sandborn WJ. Epidemiology of inflammatory bowel disease in Asia. Inflamm Bowel Dis 2001;7:260-70. 19. Lindberg E, Lindquist B, Holmquist L, Hildebrand H. Inflammatory bowel disease in children and adolescents in Sweden, 1984-1995. J Pediatr Gastroenterol Nutr 2000;30:259-64. 20. Calkins BM, Mendeloff AI. Epidemiology of inflammatory bowel disease. Epidemiol Rev 1986;8:60-91. 21. Russell MG, Stockbrugger RW. Epidemiology of inflammatory bowel disease: an update. Scand J Gastroenterol 1996;31:417-27. 22. Bernstein CN, Blanchard JF, Rawsthorne P, Wajda A. Epidemiology of Crohn’s disease and ulcerative colitis in a central Canadian province: a population based study. Am J Epidemiol 1999;149:916-24. 23. Peeters M, Nevens H, Baert F, Hiele M, de Meyer AM, Vlietinck R, et al. Familial aggregation in Crohn’s disease: increased age-adjusted risk
and concordance in clinical characteristics. Gastroenterology 1996;111: 597-603. 24. Feeney MA, Murphy F, Clegg AJ, Trebble TM, Sharer NM, Snook JA. A case-control study of childhood environmental risk factors for the development of inflammatory bowel disease. Eur J Gastroenterol Hepatol 2002;14:529-34. 25. Farmer RG, Easley K, Rankin GB. Clinical patterns, natural history and progression of ulcerative colitis: a long-term follow-up of 1116 patients. Dig Dis Sci 1993;38:1137-46. 26. Hyams JS, Davis P, Grancher K, Lerer T, Juntinich CJ, Markowitz J. Clinical outcome of ulcerative colitis in children. J Pediatr 1996;129:81-8. 27. Langholz E, Munkholm P, Davidsen M, Binder V. Course of ulcerative colitis: analysis of changes in disease activity over years. Gastroenterology 1994;107:3-11. 28. Ferguson A, Sedgwick DM. Juvenile-onset inflammatory bowel disease: predictors of morbidity and health status in early adult life. J R Coll Physicians Lond 1994;28:220-7. 29. Langholz E, Munkholm P, Krasilnikoff PA, Binder V. Inflammatory bowel diseases with onset in childhood: clinical features, morbidity, and mortality in a regional cohort. Scand J Gastroenterol 1997;32:139-47. 30. Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis. Gastroenterology 1998;115:182-205. 31. Myers TR. Pediatric asthma epidemiology: incidence, morbidity, and mortality. Respir Care Clin N Am 2000;6:1-14. 32. Smyth RL. Asthma: a major pediatric health issue. Respir Res 2002;3(suppl 1):S3-7.
APPENDIX Wisconsin Pediatric Inflammatory Bowel Disease Alliance: Ellen Blank, Christopher Brown, David Deusch, Istvan Danko, Glen Gourley, T. S. Gunasekaran, Issam Halabi, Janice Heikenen, Robert H. Judd, Farhat Khan, Subra Kugathasan, Alfonso Martinez, Roger Park, Jean Perrault, Maria T. Rivera-Bennett, Colin D. Rudolph, William San Pablo Jr., Gregorz Telega, Steven Werlin, Sally WeisdorfShendle, and Michael Yaffe.
CORRECTION The article, ‘‘Temporal variability in birth prevalence of congenital heart defects as recorded by a general birth defects registry,’’ by Bosi et al, which appeared in the June 2003 issue of The Journal (volume 142, pages 690-8) contains several errors. The following should be ‘‘per thousand’’and not ‘‘%’’: P 690 Results (in abstract): 5.1 (3.1-7.5) Opening paragraph: 3.5, 13.7, 4.5, 3.2, and 6 P 693 Second column, first paragraph in RESULTS: 7.5, 5.1, and 3.1 P 694 Second column, fourth paragraph: 8.3, 6.2, and 10.7 First paragraph in DISCUSSION: 3.5 to 13.7 P 695 First column, third paragraph: 5.1 and 3.1 to 7.5 P 697 First column: first paragraph, 6.2 to 10.7; second paragraph, 11; fifth paragraph, 8 to 10, 4 to 6 Epidemiologic and Clinical Characteristics of Children With Newly Diagnosed Inflammatory Bowel Disease in Wisconsin: A Statewide Population-Based Study
531