See “Rising Incidence of Inflammatory Bowel Disease in Young Children: What Does the Future Hold?” by Kunde et al on page 128.
Epidemiological studies report a high and rising incidence of inflammatory bowel disease (IBD) in western and northern Europe and the United States (1–3). An increase in IBD incidence has also been observed recently in eastern and southern Europe as well as in Asia (4). Most data on the epidemiology of paediatric IBD have been published in Europe and the United States (5–7). The majority of these studies report on increasing numbers of paediatric and adolescent patients with Crohn disease (CD) and ulcerative colitis (UC) (8–12), although some studies did not observe any change in incidence (13,14).
To explore the hypothesis that increasing numbers of paediatric IBD are not only a consequence of rising incidence but also of a general shift towards a younger age at disease onset, we analysed data of paediatric and adult patients recruited in the Swiss IBD Cohort Study (SIBDCS).
PATIENTS AND METHODS
The SIBDCS is a disease-oriented cohort of patients with IBD in Switzerland, which is collecting retrospective and prospective data on a large sample of paediatric and adult patients with IBD throughout Switzerland by means of physician and patient questionnaires, starting in November 2006. The study is supported by the Swiss National Science Foundation and approved by local ethical committees (15). Patients are recruited by paediatric and adult gastroenterologists during a routine consultation. Inclusion criteria consist of a diagnosis based on the Lennard-Jones diagnostic criteria, confirmed by radiology, endoscopy, or surgery. Diagnosis must be established at least 4 months before enrollment.
A total of 1910 paediatric and adult patients were recruited between November 2006 and May 2010 in the SIBDCS. Fifty-nine cases were excluded because of missing data on sex, age, or diagnosis. Forty patients, diagnosed as having an indeterminate colitis, were not considered in the analyses. We focused our analyses on the group diagnosed during the past 30 years (1980–2010), excluding therefore another 123 patients. The number of patients included in the present analysis was 1688.
Sociodemographic data were collected at the time of inclusion in the study, both by physician questionnaires (age, sex, diagnosis, calendar year of diagnosis, calendar year at first symptoms) and by patient questionnaires (education, country of birth). All of the data collected for the SIBDCS were entered, controlled, and validated at the cohort data centre.
Calendar year at first symptoms and calendar year at diagnosis were collected both by physician questionnaires and patient questionnaires; no significant differences were observed between physician and patient information. This observation was taken as a quality indicator, showing the reliability of data collection for those 2 variables. Henceforth, we decided to rely on variables from physician questionnaires to perform our analyses. Country of birth was dichotomised into 2 categories: patients born in Switzerland and patients born outside Switzerland. The latter group comprised mainly patients born in countries situated south of the Swiss latitude. Education was grouped into 3 levels: primary level (no education or compulsory school only), secondary level (elementary professional formation, apprenticeship, training school, professional graduate school, diploma school, college or household school), and tertiary level (superior professional formation, technical school, superior professional school or university). This variable was taken as a proxy for the socioeconomic situation of the patients. For paediatric patients, the level was based on the highest level of education of the father or mother.
Patients were stratified according to diagnosis and analyses were performed separately for each diagnosis. Descriptive analyses were made to characterise the patient population studied. Linear regressions were performed between age at first symptoms and age at diagnosis in relation to the calendar year at first symptoms and calendar year at diagnosis, respectively. Crude regression coefficients were calculated and compared with age-, sex- and education-adjusted regression coefficients. Reference categories used for regression modeling were female sex, born outside Switzerland and primary level of education. Regression coefficients are given with 95% confidence intervals (CI). A P value <0.05 was considered statistically significant.
A total of 1688 patients were included in the study: 986 patients with CD, 702 patients with UC. About two-thirds of the patients were recruited at tertiary centres and one-third at a district hospital or a doctor's office/private practice. The sociodemographic characteristics of the study population are shown in Table 1. Nearly half of the patients were males and half were females, with a slightly higher proportion of males with UC and females with CD. Seventy-eight percent of the patients were born in Switzerland and 22% outside Switzerland. More than half of the patients who responded to the education level questions had a secondary education level, 12% had a primary education level, and 30% had a tertiary education level. Patients with CD had a mean age at first symptoms and at diagnosis of 27 and 29 years, respectively, and patients with UC of 31 and 32 years, respectively. Identical mean ages and standard deviations were reported in the patient questionnaire.
Figures 1 and 2 and associated tables show results from unadjusted and adjusted linear regressions for both CD and UC. Considering unadjusted regressions, the tables show that there was no evidence for a linear association either between log age at diagnosis and year of diagnosis or between log age at first symptoms and year of first symptoms.
The table associated with Figure 1A shows that there is evidence for a linear association between log age at diagnosis and year at diagnosis for CD, and that male sex and being born in Switzerland were also significantly associated with log age at diagnosis. Year of diagnosis (coefficient 0.011; 95% CI [0.007–0.015]) as well as male sex (coefficient 0.078; 95% CI [0.015–0.140]) are positively correlated with log age at diagnosis, whereas being born in Switzerland is negatively correlated with log age at diagnosis (coefficient −0.130; 95% CI [−0.201 to −0.050]). Education is not associated with log age at diagnosis, considering the linear model. The same observations were made for UC (Fig. 1B), except for male sex, which was not significantly associated with log age at diagnosis (coefficient 0.058; 95% CI [−0.010 to 0.126]). The low R2 values show that only 6% to 7% of the observed variation can be explained by a linear regression.
Adjusted linear regressions show significant association with the same variables, considering age at first symptoms and year of first symptoms (Fig. 2A and B). Additionally, however, male sex was significantly associated with a linear increase in log age at first symptoms both for CD and UC. The low R2 values again show that only 5% to 8% of the observed variation can be explained by the linear regression.
Table 2 shows that the proportion of patients with CD with first symptoms at younger than 18 years has increased in the 2000s (24.8%) compared with the 1980s (14.5%); no difference in proportion was observed for patients with first symptoms at older than 40 years (17.6% vs 17.4%); however, for UC, the proportion of patients with first symptoms at older than 40 years has increased from 14.3% in the 1980s to 27.0% in the 2000s (Table 3). An increased proportion was also observed for young patients with UC. Considering year at diagnosis, we observed an increase in the proportion of patients, with more than 40 diagnosed as having CD (12.3%–22.5%; Table 2), and an increase of patients, with both <18 (9.0–12.9%) or >40 (16.4–27.5%) diagnosed as having UC between the 1980s and the 2000s (Table 3). All of the observed associations were statistically significant.
Our data have been collected in the course of the large SIBDCS. The results of our analysis do not support the hypothesis that altogether first symptoms and diagnosis of both patients with CD and patients with UC in Switzerland occur today at a younger age. On the contrary, the results of the SIBDCS show, taking account of sex, country of birth, and education, that there is a slow but significant trend for both first symptoms and diagnosis occurring at an older age today compared with the last 3 decades. A linear relation between log age and year of event cannot explain most of the observed phenomena, which seem to be additionally complex.
The analysis of patient's age at first symptoms and diagnosis by decade revealed that in CD—despite the fact that altogether there is significant association for first symptoms and diagnosis occurring at an older age—the number of younger patients (younger than 18 years) has increased significantly from the 1980s to the 2000s. The same observation was made in UC; however, we also observed increasing numbers of patients with UC with first symptoms at an older age (older than 40 years). The trend for both first symptoms and diagnosis occurring at an older age today compared with the last 3 decades in IBD may therefore be caused by both a rise in the number of younger and older patients, with older patients having a stronger effect.
The limitations of our study include a potential sampling bias, caused by the late acquisition of data for patients diagnosed earlier in the 20th century. For example, the data of some of these patients diagnosed in the 1980s at an older age may not have been collected because some of the patients may already have died at the time of the start of data collection for the SIBDCS in the year 2006. Nevertheless, the regression analysis of our data shows that there is a significantly linear increase in log age with year. A further limitation includes a potential selection bias, caused by the fact that a majority of the patients in the SIBDCS were recruited in a tertiary referral centre. Therefore, patients with mild disease may not have been included in this study; however, there is no indication that patients with mild disease may be diagnosed at a younger age compared with patients with moderate-to-severe disease.
1. Gismera CS, Aladrén BS. Inflammatory bowel diseases: a disease(s) of modern times? Is incidence still increasing? World J Gastroenterol
2. Lakatos PL. Recent trends in the epidemiology of inflammatory bowel diseases: up or down? World J Gastroenterol
3. Shivananda S, Lennard-Jones J, Logan R, et al. Incidence of inflammatory bowel disease across Europe: is there a difference between north and south? Results of the European Collaborative Study on Inflammatory Bowel Disease (EC-IBD). Gut
4. Ouyang Q, Tandon R, Goh KL, et al. The emergence of inflammatory bowel disease in the Asian Pacific region. Curr Opin Gastroenterol
5. Sawczenko A, Sandhu BK, Logan RF, et al. Prospective survey of childhood inflammatory bowel disease in the British Isles. Lancet
6. Kugathasan S, Judd RH, Hoffmann RG, et al. Epidemiologic and clinical characteristics of children with newly diagnosed inflammatory bowel disease in Wisconsin: a statewide population-based study. J Pediatr
7. Ravikumara M, Sandhu BK. Epidemiology of inflammatory bowel diseases in childhood. Indian J Pediatr
8. Armitage EL, Aldhous MC, Anderson N, et al. Incidence of juvenile onset Crohn's disease in Scotland: association with northern latitude and affluence. Gastroenterology
9. Hildebrand H, Finkel Y, Grahnquist L, et al. Changing pattern of paediatric inflammatory bowel disease in northern Stockholm 1990–2001. Gut
10. Auvin S, Molinie F, Gower-Rousseau C, et al. Incidence, clinical presentation and location at diagnosis of pediatric inflammatory bowel disease: a prospective population-based study in northern France (1988–1999). J Pediatr Gastroenterol Nutr
11. Pozler O, Maly J, Bonova O, et al. Incidence of Crohn disease in the Czech Republic in the years 1990 to 2001 and assessment of pediatric population with inflammatory bowel disease. J Pediatr Gastroenterol Nutr
12. Phavichitr N, Cameron DJ, Catto-Smith AG. Increasing incidence of Crohn's disease in Victorian children. J Gastroenterol Hepatol
13. Langholz E, Munkholm P, Krasilnikoff PA, et al. Inflammatory bowel diseases with onset in childhood. Clinical features, morbidity, and mortality in a regional cohort. Scand J Gastroenterol
14. Lindberg E, Lindquist B, Holmquist L, et al. Inflammatory bowel disease in children and adolescents in Sweden, 1984–1995. J Pediatr Gastroenterol Nutr
15. Pittet V, Juillerat P, Mottet C, et al. Cohort profile: the Swiss Inflammatory Bowel Disease Cohort Study (SIBDCS). Int J Epidemiol