• inflammatory bowel disease;
  • phenotype;
  • genetics;
  • population-based;
  • epidemiology


  1. Top of page
  2. Abstract
  5. CARD15
  7. Acknowledgements

Background and Aim: The aetiology of inflammatory bowel disease (IBD) is unknown, but it has become evident that genetic factors are involved in disease susceptibility. Studies have suggested a north–south gradient in the incidence of IBD, raising the question whether this difference is caused by genetic heterogeneity. We aimed to investigate the prevalence of polymorphisms in CARD15 and TLR4 and occurrence of anti-Saccharomyces cerevisiae (ASCA) and antineutrophil cytoplasmic antibodies (pANCA) in a European population-based IBD cohort.

Methods: Individuals from the incident cohort were genotyped for three mutations in CARD15 and the Asp299gly mutation in TLR4. Levels of ASCA and pANCA were assessed. Disease location and behaviour at time of diagnosis was obtained from patient files.

Results: Overall CARD15 mutation rate was 23.9% for CD and 9.6% for UC patients (P < 0.001). Mutations were less present in the Scandinavian countries (12.1%) versus the rest of Europe (32.8%) (P < 0.001). Overall population attributable risk was 11.2%. TLR4 mutation rate was 7.6% in CD, 6.7% in UC patients and 12.3% in healthy controls (HC), highest among South European CD patients and HC. ASCA was seen in 28.5% of CD patients with no north–south difference, and was associated with complicated disease. pANCA was most common in North European UC patients and not associated with disease phenotype.

Conclusion: The prevalence of mutations in CARD15 varied across Europe, and was not correlated to the incidence of CD. There was no association between mutations in TLR4 and IBD. The prevalence of ASCA was relatively low; however related to severe CD.

(Inflamm Bowel Dis 2007;13:24–32)

Crohn's disease (CD) and ulcerative colitis (UC) are two related inflammatory disorders of the intestinal tract, with unknown aetiology. However, during the past years it has become evident that genetic factors play a significant role in the development of inflammatory bowel disease (IBD). For several decades it has been known from epidemiological studies (including twin studies) that familial aggregation occurs, indicating some genetic predisposition.1–3 Epidemiological studies have also shown that the incidence and prevalence of IBD is dependent on geographic location and ethnic origin. In 1991–1993 the European Collaborative Study Group on Inflammatory Bowel Diseases (EC-IBD) investigated the incidence of IBD in well defined areas across Europe.4 The incidence of CD varied between 0.9/100.000 and 9.2/100.000, and for UC between 1.6/100.000 and 24.5/100.000, with a minor north–south gradient, but also considerable variations within similar geographical areas. The incidence of IBD seems to be rising in all Western countries, and seems to be more common in urban compared to rural areas, indicating that we are not dealing with a purely genetic disorder but with a complex multifactorial disease as the result of interplay between genetic and environmental factors. The genetic predisposition has been the subject of intensive research; beginning in the late nineties and leading to its first success in 2001 when it was discovered that mutations in the CARD15 gene contribute to CD susceptibility.5, 6 The prevalence of mutations in CARD15 in different CD populations has been reported to vary from 0% to 81.7%, also with a wide range among healthy controls.7–11 The identification of CARD15 has focused attention to the pathway of innate immunity and the molecules playing a role within this pathway. The family of the toll like receptors (TLR), of which TLR4 is associated with the gut, plays an important role in the innate immunity. TLR4 is up-regulated in epithelial cells in patients with IBD, and it has been debated whether there is an association between the TLR4 Asp299Gly mutation and susceptibility to IBD. Several studies have been conducted leading to divergent results.12–17 At present there is no unambiguous evidence that mutations in TLR4 are associated with CD or UC.

The aim of this study was to investigate the prevalence of CARD15, TLR4, ASCA and pANCA in a European population based cohort of IBD patients from different regions in Europe and Israel, and compare the results with the incidence and clinical characteristics of IBD at time of diagnosis.


  1. Top of page
  2. Abstract
  5. CARD15
  7. Acknowledgements

Study Population

During the period October 1991 to September 1993, 20 centres participating in the EC-IBD study group participated in forming a cohort of incident IBD patients. Patients came from 20 well defined geographical areas with a total background population of 7.479.216 inhabitants. All patients fulfilled the international diagnostic criteria, as defined by Lennard-Jones.4, 18 and were prospectively included. In 2000 it was decided to conduct a follow-up study in which 13 centres (from 9 countries) agreed to participate. Final participation required a minimum follow-up rate of 60%; 10 centres fulfilled this criterion and have provided data for this study. Patients were followed from inception between Oct 1st 1991 to Sept 30th 1993 until data inclusion between August 1st 2002 and January 31st 2004, or any date prior to the data inclusion period, indicated as lost to follow-up or death. From patient files each centre gathered information on vital status and disease course (disease location and behaviour, medical and surgical treatment) and filled in the information on electronic (web-based) “physician per patient” questionnaires.19 Patients were approached individually and asked to participate in an interview regarding family history of IBD, dietary habits and fertility and pregnancies. In connection to the interview, patients were asked to give a blood sample for genetic and serological analyses. Frozen whole blood and serum samples were shipped for analysis at the laboratory of gastroenterology, UZ Gasthuisberg (Leuven, Belgium). The local Ethics Committee of each participating centre approved the study and all subjects gave their informed consent.


CD and UC patients were classified according to disease location at diagnosis. Macroscopic disease location was assessed from endoscopic and radiological examinations (X-ray, MR-scan, CT-scan) or surgeries performed at time of diagnosis. CD patients were classified as having pure terminal ileal, pure colonic, ileo-colonic or upper gastrointestinal involvement according to the Vienna Classification.20 UC patients were classified according to disease extent as proctitis (rectum only), left-sided (sigmoideum involved), extensive disease (trans-versum involved) or pancolitis (all of colon involved).

CD patients were also classified according to disease behaviour as having stricturing (presence of stenosis in the small or large bowel), penetrating (presence of fistula or abscess) or inflammatory disease (non-stricturing, non-penetrating) at time of diagnosis.


DNA for genotyping was extracted from whole venous blood and genotyped for the three common single nucleotide polymorphisms (SNP) in the CARD15 gene known to be associated with CD: ARG702Trp (referred to as SNP8), Gly908Arg (SNP12) and Leu1007fsinsC (SNP13), and for the Asp299Gly SNP in TLR4. Genotyping analyses were performed using restriction fragment length polymorphism techniques as described before.21, 22

Serological Analyses

Anti-Saccharomyces cerevisiae antibodies (ASCA) were measured by a standardized ELISA (Medipan Diagnostica Germany) according to the manufacturer's instructions, as previously described.23 Results were interpreted by calculating the binding index (BI). ASCA IgG and IgA were considered positive at a BI > 1.

Antineutrophil cytoplasmic antibodies (pANCA) was determined by indirect immunofluorescence (IIF) using ethanol-fixated neutrophil slides (Inova Diagnostics, San Diego, Calif), as previously described.24

Statistical Analysis

All statistical analyses were performed using SPSS statistical software package (SPSS Inc., Chicago Ill). Comparison of different groups was done using the chi-square test or Fishers exact test where appropriate. Mann-Whitney test was used for comparison of medians. The threshold for statistical significance was set at P < 0.05. The Population Attributable Risk percent (PAR%) was defined as the excess rate of disease in individuals with a mutation compared with those without. This was estimated as described by Hennekens and Buring, and to calculate this, the frequency of all alleles in the control population was assumed to reflect that of the general population.25

Spearman rank correlation test for non-parametric values were used for correlation analysis.


  1. Top of page
  2. Abstract
  5. CARD15
  7. Acknowledgements

Data on 1164 CD and UC patients from the original cohort were available for this study. The patients were recruited from 10 centres in 8 countries in Europe and Israel: Oslo (Norway), Copenhagen (Denmark), South Limburg (The Netherlands), Cremona (North Italy (N-Italy 1)), Reggio Emilia (N-Italy 2), Vigo (Spain), Ioannina (North Greece (N-Greece)), Almada (Portugal), Heraklion (South Greece (S-Greece) and Beer Sheeva (Israel). One centre did not fulfil the 60% threshold for UC patients and therefore only their CD patients are included in the analysis. Three-hundred and eighty patients were diagnosed as having CD and 784 had UC. In all, 777 (67%) patients agreed to participate in the patient interview, 126 (10.7%) were not willing to participate; 111 patients (9.5%) had died, 138 (11.8%) were lost to follow-up and 12 (1.0%) turned out to be misdiagnosed. Of the 777 interviewed patients serum and/or DNA were available in 620 patients, equivalent to 53% of the initial 1164 patients.

Blood samples from 640 healthy controls (HC) were collected at the individual centres thus originated from the same geographical area as the patients.

The study includes North and South European centres. In the original paper centres located in France and the Netherlands were considered as being part of Northern Europe.4 This classification is maintained in this study. However, from a genetic point of view, we also aimed at analysing the Scandinavian countries separately given that these populations are genetically more closely related than the rest of Europe.26

Table 1 shows the clinical characteristics for the 213 CD patients. There was no significant difference between CD patients from North or South European centres regarding gender or age at diagnosis, disease location or behaviour at time of diagnosis (according to the Vienna classification). Looking at involvement of the terminal ileum regardless of any other location still showed no difference, the same was found for pure colonic involvement.

Table 1. Clinical Characteristics of 213 European Patients with Crohn's Disease. n (%)
 NorwayDenmarkNetherlandsNorthN-Italy 1N-Italy 2SpainN-GreecePortugalS-GreeceIsraelSouthTotal
  • *

    Irrespective of any other location.

  • a

    P = 0.07.

  • b

    P = 0.83.

  • c

    OR 0.62, 95% CI: 0.34–1.14, P = 0.12.

  • d

    OR 0.99, 95% CI: 0.51–1.90, P = 0.89.

CD cohort (n)62295314472420383469213
Male sex (%)32 (51.6)8 (27.6)24 (45.3)64 (44.4)a2 (28.6)12 (50.0)16 (80.0)2 (66.7)4 (50.0)2 (66.7)2 (50.0)40 (58.0)a104 (48.8)
Age at diagnosis (median, IQR)31 (24–43)31 (23–42)30 (24–39)30b (23–45)44 (28–56)35 (25–46)25 (21–30)43 (31–73)40 (27–52)25 (25–43)22 (18–43)31b (24–41)31 (34–43)
Location at diagnosis             
 Terminal ileum16 (25.8)7 (24.1)17 (32.1)40 (27.8)010 (41.7)4 (20.0)1 (33.3)2 (25)0017 (24.6)57 (26.8)
 Colon26 (41.9)12 (41.4)22 (41.5)60 (41.7)c5 (71.4)5 (20.8)7 (35)1 (33.3)3 (37.5)0021 (30.4)c81 (38.0)
 Ileocolon16 (25.8)6 (20.7)12 (22.6)34 (23.6)2 (28.6)8 (33.3)6 (30)03 (37.5)2 (66.7)1 (25.0)22 (31.9)56 (26.3)
 Upper GI4 (6.5)4 (13.8)2 (3.8)10 (6.9)01 (4.2)3 (15)1 (33.3)01 (33.3)3 (75)9 (13.0)19 (8.9)
Involvement of terminal ileum at diagnosis*36 (58.1)14 (48.3)30 (56.6)80 (55.6)d2 (28.6)19 (79.2)12 (63.2)2 (66.7)5 (62.5)2 (66.7)4 (100)46 (66.6)d125 (59)
Behaviour diagnosis             
 Inflammatory53 (85.5)17 (58.6)33 (62.3)103 (71.5)5 (71.4)20 (83.3)11 (55.0)3 (100)4 (50.0)3 (100)4 (100)50 (72.5)153 (71.8)
 Stricturing7 (11.3)8 (27.6)11 (20.8)26 (18.1)2 (28.6)3 (12.5)3 (15.0)03 (37.5)0011 (15.9)37 (17.4)
 Penetrating2 (3.2)4 (13.8)9 (17.0)15 (10.4)01 (4.2)6 (30.0)01 (12.5)008 (11.6)23 (10.8)

Table 2 shows the demographic data for UC patients. There were no north–south differences regarding gender, age or disease extent at diagnose.

Table 2. Clinical Characteristics of 407 European Patients with Ulcerative Colitis. n (%)
 NorwayDenmarkNetherlandsNorthN-Italy 1N-Italy 2SpainN-GreecePortugalS-GreeceIsraelSouthTotal
  • *

    Location at time of diagnosis is missing in 9 cases—3 from the north and 6 from the south.

  • a

    p > 0.05 when comparing North and South European centres.

UC cohort (n)1464376265143434301416142407
Male sex (%)69 (47.3)19 (44.2)50 (65.8)138 (52.1)a11 (78.6)19 (55.9)15 (44.1)14 (46.7)9 (64.3)11 (68.8)79 (55.6)217 (53.3)
Age at diagnosis38.037.037.737.8a45.632.931.440.443.031.337.937.8
 (median, range)16–7616–6416–7116–7622–6017–6916–7020–72 24–7218–7116–7216–76
Disease extent at diagnosis*             
–Pancolitis31 (21.2)5 (12.5)7 (9.2)43 (16.4)a2 (14.3)14 (43.8)5 (15.6)5 (16.7)6 (50.0)1 (6.2)33 (24.3)76 (18.7)
–Extensive10 (6.9)07 (9.2)17 (6.5)a02 (6.3)001 (8.3)03 (2.2)20 (4.9)
–Left-sided62 (42.5)21 (52.5)42 (55.3)111 (42.4)a7 (50.0)8 (2.5)15 (46.9)19 (63.3)5 (41.7)10 (62.5)64 (47.1)175 (43.0)
–Proctitis43 (29.5)14 (35.0)20 (26.3)91 (34.7)a5 (35.7)8 (2.5)12 (37.5)6 (20.0)05 (31.3)36 (26.5)127 (31.2)


  1. Top of page
  2. Abstract
  5. CARD15
  7. Acknowledgements

The allelic frequencies and the prevalence of mutations in the CARD15 gene is listed in Table 3.

Table 3. Allelic Frequencies of Mutations in CARD15 Mutations, n (%), in Patients with Crohn's Disease (CD) and Healthy Controls (HC) from 8 Different European Countries and Israel
CD (n = 62)HC (n = 136)CD (n = 29)HC (n = 122)CD (n = 53)HC (n = 114)CD (n = 144)HC (n = 372)CD (n = 31)HC (n = 95)CD (n = 20)HC (n = 75)CD (n = 8)HC (n = 13)CD (n = 6)HC (n = 53)CD (n = 4)HC (n = 32)CD (n = 69)HC (n = 268)
R702W9/124 (7.3)7/272 (2.6)2/58 (3.4)8/244 (3.3)12/106 (11.3)17/228 (7.5)23/288 (8.0)32/744 (4.3)8/62 (12.9)11/190 (5.8)6/40 (15.0)4/150 (2.7)1/16 (6.3)4/26 (15.4)02/106 (1.9)0015/138 (10.9)21/536 (3.9)
G908R1/124 (0.8)004/244 (1.6)4/106 (3.8)9/228 (3.9)5/288 (1.7)13/744 (1.7)1/62 (1.6)2/190 (1.1)000001/106 (0.9)2/8 (25.0)4/64 (6.3)3/138 (2.2)7/536 (1.3)
L1007fs3/124 (2.4)3/272 (1.1)06/244 (2.5)7/106 (6.6)6/228 (2.6)10/288 (3.5)15/744 (2.0)4/62 (6.5)7/190 (3.7)4/40 (10.0)2/150 (1.3)0002/106 (1.9)008/138 (5.8)11/536 (2.1)
1 risk allel5/62 (8.1)10/136 (7.4)2/29 (6.9)18/122 (14.7)15/53 (28.3)23/114 (20.2)22/144 (15.3)51/272 (13.7)7/31 (22.6)18/95 (18.9)6/20 (30.0)6/75 (8.0)1/8 (12.5)2/13 (15.4)03/53 (5.7)2/4 (50.0)4/32 (12.5)16/69 (4.3)33/268 (12.3)
2 risk allel1/62 (1.6)0004/53 (7.5)1/114 (0.9)5/144 (3.5)1/372 (0.27)2/31 (6.5)01/20 (5.0)00000003/69 (23.2)0
Homo-zygote3/62 (4.8)00004/114 (3.5)3/144 (2.1)4/372 (1.1)1/31 (3.2)1/95 (1.1)  1/20 (5.0)001/13 (7.7)01/53 (1.9)002/69 (2.9)3/268 (1.1)
At least one mutation9/62 (14.5)10/136 (7.4)2/29 (6.9)18/122 (14.7)19/53 (35.8)28/114 (24.5)30/144 (20.8)56/372 (15.1)10/31 (32.3)19/95 (20.0)  8/20 (40.0)6/75 (8.0)1/8 (12.5)3/13 (23.1)04/53 (7.6)2/4 (50.0)4/32 (12.5)21/69 (30.4)36/268 (13.4)

For the total CD cohort the prevalence of having at least one mutation in the CARD15 gene was 23.9%, significantly higher than the prevalence observed among UC patients (9.6%) (OR 2.89, 95% CI: 1.84–4.55, P < 0.0001) and healthy controls (14.4%) (OR 1.87, 95% CI: 1.28–2.75, P = 0.001). There was no significant difference between the prevalence of mutations in UC and controls. The prevalence for each centre varied from 0–50%, but with large variation also in the number of patients available for blood sampling. There was no difference in the prevalence of carrying at least one mutation in CARD15 between North or South European (including Israel) CD patients (OR 0.60, 95% CI: 0.31–1.15, P = 0.12). However, when looking at the Scandinavian CD patients alone, they had a significantly lower prevalence compared to the rest of Europe (OR 0.28, 95% CI: 0.14–0.59, P < 0.0001). In fact, in these countries the prevalence of carrying at least one mutation in CARD15 was not significantly higher in CD versus controls.

Regarding allelic frequencies in the CD cohort, the R702w (OR 2.27, 95% CI: 1.47–3.49, P = 0.0001) and 1007fs (OR 2.13, 95% CI: 1.15–3.92, P = 0.01) but not the G908R (OR 1.20, 95% CI: 0.53–2.76, P = 0.66) differed between CD patients and healthy controls (Table 3). There were no north–south differences when comparing the individual mutation for either CD patients or healthy controls. In the North the R702w (OR 1.93, 95% CI: 1.11–3.36, P = 0.02) but not the G908R (OR 0.99, 95% CI: 0.35–2.81, P = 0.99) or the 1007fs (OR 1.74, 95% CI: 0.78–3.94, P = 0.17) differed between CD and controls. In the South European centres both R702w (OR 2.99, 95% CI: 1.50–5.97, P = 0.001) and the 1007fs (OR 2.94, 96% CI: 1.16–7.45, P = 0.02) but not the G908R (OR 1.68, 95% CI: 0.43–6.58, P = 0.45) differed between CD and controls.

There was no correlation between the incidence of CD as it was described in 1991–19934 and the prevalence of at least one mutation in CARD15 (Spearman rank correlation coefficient 0.31, P = 0.46) (Fig. 1). Looking separately at the South European countries there was still no correlation between CARD15 and incidence (Spearman rank correlation coefficient 0.8, P = 0.1).

thumbnail image

Figure 1. Incidence of CD and prevalence of CARD15 mutations in CD patients in different European populations. Incidences are from the incidence study performed 1991–1193.3

Download figure to PowerPoint

The excess rate of CD in individuals with at least one mutation in CARD15 (PAR%) was 11.2% for the whole cohort, with large differences between the participating countries (range—13.7%–44.6%). Overall the PAR% was lower in the north compared to the south; 6.8% vs. 19.6%.


The prevalence of a mutation in TLR4 in the total cohort is listed in Table 4. There was no difference in the prevalence observed in CD patients versus UC (OR 1.14, 95% CI: 0.6–2.17, P = 0.68) or controls (OR 0.58, 95% CI: 0.33–1.03, P = 0.06), however the prevalence in UC patients was significantly lower than in controls (OR 0.51, 95% CI: 0.32–0.81, P = 0.003).

Table 4. The Prevalence (%), of Toll Like Receptor 4 (TLR4), anti-Saccharomyces cerevisiae Antibodies (ASCA) and Antineutrophil Cytoplasmic Antibodies (pANCA) in an Unselected European Cohort of Patients with Crohn's Disease (CD) and Ulcerative Colitis (UC) and Healthy Controls (HC)
- CD (n = 211)
- UC (n = 404)6.34.713.
- HC (n = 618)6.19.813.29.725.314.715.47.59.416.412.3
- CD (n = 207)25.827.623.125.222.652.637.533.35035.328.5
- UC (n = 394)
- HC (n = 640)13.20.811.48.6014.705.705.27.2
- CD (n = 207)
- UC (n = 394)37.430.215.829.816.718.215.418.816.925.3
- HC (n = 640)17.61.611.410.5016.005.705.68.4

Genotype frequencies for TLR4 did not differ between North European CD and UC (OR 0.63, 95% CI: 0.26–1.52, P = 0.30), CD and controls (OR 0.48, 95% CI: 0.21–1.10) or UC and controls (OR 0.77, 95% CI: 0.43–1.40, P = 0.37). In Southern Europe there was no difference between CD patients and controls (OR 0.72, 95% CI: 0.34–1.5, P = 0.40); however TLR4 mutations were less frequent in UC versus CD patients (OR 0.34, 95% CI: 0.12–0.97, P = 0.04) and controls (OR 0.26, 95% CI: 0.12–0.60, P < 0.001). The prevalence of mutations in TLR4 was significantly higher in Southern Europe among CD patients (OR 2.89, 95% CI: 1.0–8.1, P = 0.04) and healthy controls (OR 1.6 95% CI: 1.0–2.7, P = 0.03) but not in UC patients (OR 0.6, 95% CI: 0.26–1.50, P = 0.29) compared to northern populations.

When comparing the prevalence of mutation in TLR4 in Scandinavia to the rest of Europe there was no difference regarding CD (OR: 0.43, 95% CI: 0.13–1.38, P = 0.15) or UC patients (OR 0.87, 95% CI: 0.42–1.80, P = 0.71). However the prevalence of a mutation in TLR4 was lower in Scandinavian controls compared to controls from the rest of Europe (OR 0.48, 95% CI: 0.28–0.83, P < 0.007).

No interaction was found between Asp299Gly polymorphism and the CARD15 genotype.


Mutations in CARD15 were not associated with gender or age at diagnosis.

Fifty-one patients had at least one mutation in CARD15, of these 34 (66.7%) had involvement of the terminal ileum. Of the 162 patients without a mutation, 56.8% had CD in the terminal ileum, hence carriage of at least one mutation in CARD15 was not associated with involvement of the terminal ileum, neither in the total cohort (OR 1.52, 95% CI: 0.78–2.94, P = 0.21) nor in the Northern (OR 1.26, 95% CI: 0.55–2.85, P = 0.58) or Southern Europe separately (OR 1.98, 95% CI: 0.62–6.30, P = 0.24). When SNP8, 12 or 13 were considered separately, there was no association with involvement of the terminal ileum.

Mutations in TLR4 were not associated with sex or age at diagnosis in neither UC nor CD patients. No association was found between mutations in TLR4 and stricturing or penetrating disease behaviour in CD patients. For UC no association was found to extent of disease.


Table 4 shows the prevalence of ASCA and pANCA positive patients. 28.5% of all CD patients were ASCA positive compared to 7.1% of UC patients (OR 5.21, 95% CI: 3.19–8.49, P < 0.001) and 6.2% HC (OR 5.15, 95% CI: 3.36–7.88, P < 0.001). No difference in ASCA between UC patients and HC was observed; also were there no north–south differences.

For the total CD cohort ASCA positivity was associated with younger age at onset (median 30.4 vs. 35.8 years, P = 0.012). When analyzing North and South European centres separately this association however, was only present in the South European centres where the difference was more pronounced: median age at diagnosis in ASCA positive patients 28.1 years compared to 38.7 years in ASCA negative patients (P = 0.04).

CD patients with complicated disease at time of diagnosis (stricturing or penetrating according to Table 1) were more likely to be ASCA positive (OR 2.72, 95% CI: 1.46–5.08, P = 0.001). This was seen in the total cohort and when analyzing North and South of Europe separately.

We had information on both CARD15 and ASCA status on 207 CD patients. Thirty-nine percent of the ASCA positive CD patients were also CARD15 positive compared to 17.6% of the ASCA negative (OR 3.00, 95% CI: 1.53–5.88, P = 0.001).

In the total cohort pANCA positivity was more common among UC patients compared to CD (OR 3.60, 95% CI: 2.15–6.03, P < 0.001) and HC (OR 2.75, 95% CI: 1.93–3.93, P < 0.001) (Table 4). Significantly more UC patients from North compared to South European centres were pANCA positive (OR 2.51, 95% CI: 1.50–4.21, P = 0.003). pANCA was not correlated to gender, age or extent of disease at diagnosis in UC patients in Northern or Southern Europe.


  1. Top of page
  2. Abstract
  5. CARD15
  7. Acknowledgements

This is to our knowledge the first report of prevalences of mutations in CARD15 in an unselected cohort of IBD patients from well defined geographic areas in Europe and Israel.

Previous studies on these cohorts revealed varying incidences of CD across Europe. In this study we could not find a correlation between incidence of CD and prevalence of CARD15. On the contrary we found low prevalences of CARD15 mutations in some of the countries with the highest incidences of CD. The prevalence of CARD15 mutations in the Scandinavian countries was low compared to what has been reported elsewhere in Europe and North America,8, 10, 27, 28 but comparable to reports from Finland.9 In the present study, the prevalence of mutations in CARD15 was comparable in Scandinavian CD patients and controls indicating that in these populations other factors besides CARD15 are responsible for the development of CD. Environmental factors are likely to be involved since twin studies have described the presence of healthy siblings among CARD15 positive monozygotic twin pairs being discordant for CD.29 Furthermore several studies have shown a steep increase in the incidence of IBD since the 1980s, an increase that cannot be explained by genetic influence.30–35

There were no differences in disease location or age at diagnosis within this European cohort that could account for the difference seen in mutation rates, knowing that mutations in CARD15 have been associated with ileal disease and young age at diagnosis.

A recent study on Danish CD patients showed mutation rates of 22% in referral patients,36 which is higher than the rates described in this paper. The referral patients had a high frequency of fistulizing disease at time of diagnosis compared to the patients in this present study (31% vs. 13.8%). Other studies have reported mutations in CARD15 to be associated with a more aggressive clinical course;8 thus heterogeneity in disease course could also, to some extent, explain the observed differences in mutation rates between CD populations.

The Dutch and Scandinavian people (especially Danes) have a high genetic resemblance, and therefore one could expect them to have an almost identical prevalence in CARD15 mutations.26 However South Limburg, the region that took part in this study has had a large Mediterranean population throughout centuries. This could contribute to the higher mutation rate seen in the Dutch population since the Mediterranean population have little genetic resemblance to North European countries. This is also reflected in the population attributable risk, PAR%, which is almost similar in the Dutch and Italian population. When we counted The Netherlands as part of Southern Europe we found a north–south gradient, with mutations in CARD15 being more frequent in Southern Europe, presumably reflecting differences in the genetic origin.

When looking at the individual countries some have a high prevalence of mutations (e.g., Italy with 32.3%) but the prevalence in the healthy population is correspondingly high. This is reflected in the PAR% revealing that the prevalence of the mutation might be high, but its influence on the development of CD is relatively low. In some countries the PAR% is negative, thus the mutations does not seem to increase the susceptibility of CD in those populations. This however should be interpreted with caution, since the number of patients in some centres was very low.

Allele frequency of the Asp299Gly mutation presented in papers and abstracts range between 8–13% in CD, 0–10% in UC and 3–15% in HC.12, 16, 37–40 Mutations in TLR4 did not contribute to disease susceptibility in our IBD population nor was there any association with mutations in CARD15. The mutation was first described in a Belgian study, where the frequency of the allele mutation was higher in CD compared to UC patients.38 Replication studies have shown conflicting results. The association with CD was confirmed in Dutch and Greek populations but not in Irish CD patients.12, 16, 39 Török et al. found an association with UC which still remains to be confirmed.17 The diverging reports on the association between TLR4 and IBD most likely reflect the heterogeneity of populations and genetic complexity of IBD.

The prevalence of ASCA in this population was low compared to previous reports. The assay used to detect ASCA has in a comparative study been shown to be very specific, but with low sensitivity.23 Ours is a population-based study with both mild and severe cases of CD, and studies have indicated that ASCA is a marker of severe disease—confirmed in this study where ASCA was associated with stenosing/fistulizing disease at diagnosis. However; we have not performed a multivariate analysis on genotype-phenotype correlations and this implies that any conclusions should be drawn with great cautiousness. Nevertheless our findings fall in line with recent findings that CD patients with positive serological markers seem to experience a more severe disease course.41–43

This study is carried out on a population-based cohort. In a ten year follow-up it is unavoidable that some patients have died, are lost to follow-up or not willing to participate in a study that requires personal attendance. In order to overcome an eventual bias we defined a 60% follow-up limit for each of the centres prior to the data-collection. Extent of disease at diagnosis in patients participating in the present study is comparable with the distribution in the original cohort, indicating that the patients in this sub-cohort were not selected in this respect.19, 44 The weakness of this study is the low number of patients included especially by some of the South European centres. This increases the risk of type 2 errors and ideally our conclusions should be confirmed in larger cohorts.

In conclusion, in this population based European cohort of IBD patients, we found that the prevalence of mutations in CARD15 in patients with CD was comparable to prevalence found in other studies except in the Scandinavian countries where mutation rates were remarkably low. No correlation between incidence of CD and prevalence of mutations in CARD15 was found, and there was no European north–south gradient. However; the overall PAR% was low, indicating that the mutations do not play a major role for susceptibility in Crohn's disease. Furthermore we could not find an association between mutations in TLR4 and IBD. Presently, genetic variants cannot be used for screening in IBD, but further studies are warranted in order to evaluate whether certain genotypes or serotypes are associated with a specific disease course.


  1. Top of page
  2. Abstract
  5. CARD15
  7. Acknowledgements

The authors wish to thank the members of the EC-IBD Study Group for their contribution to this work: Marina Beltrami, Tom Bernklev, Paolo Bodini, Mercedes Butron, Mia Cilissen, Greet Claessens, Juan Clofent, Claudio Cortelezzi, Guiseppe d'Albasio, Michel Economou, Giovanni Fornaciari, Angie Harrington, Ole Høie, Sofie Joossens, Ioannis Koutroubakis, Charles Limonard, Idar Lygren, Andrea Messori, Estela Monteiro, Maria Grazia Mortilla, Marius Nap, Paula Borrhalho Nunes, Athanasios Pallis, Angelo Pera, Santos Pereira, Asghar Quasim, Tullio Ranzi, Marielle Romberg, Maurice G. Russel, Leo Schouten, Jildou Sijbrandij, Camilla Solberg, Epameinondas Tsianos, Maria Tzardi, Ed van Hees, Robert van Hees, Gilbert van Zeijl, Morten Vatn, Ioannis Vlachonikolis, Frederik Wessels.


  1. Top of page
  2. Abstract
  5. CARD15
  7. Acknowledgements
  • 1
    Orholm M, Munkholm P, Langholz E, et al. Familial occurrence of inflammatory bowel disease. N Engl J Med. 1991; 324: 8488.
  • 2
    Orholm M, Binder V, Sorensen TI, et al. Concordance of inflammatory bowel disease among Danish twins. Results of a nationwide study. Scand J Gastroenterol. 2000; 35: 10751081.
  • 3
    Kirsner JB, Spencer JA. Family occurrence of ulcerative colitis, regional enteritis, and ileocolitis. Ann Intern Med 1963; 59: 13344.
  • 4
    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. 1996; 39: 690697.
  • 5
    Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature. 2001; 411: 599603.
  • 6
    Ogura Y, Lala S, Xin W, et al. Expression of NOD2 in Paneth cells: a possible link to Crohn's ileitis. Gut. 2003; 52: 15911597.
  • 7
    Gazouli M, Zacharatos P, Mantzaris GJ, et al. Association of NOD2/CARD15 variants with Crohn's disease in a Greek population. Eur J Gastroenterol Hepatol. 2004; 16: 11771182.
  • 8
    Annese V, Lombardi G, Perri F, et al. Variants of CARD15 are associated with an aggressive clinical course of Crohn's disease—an IG-IBD study. Am J Gastroenterol. 2005; 100: 8492.
    Direct Link:
  • 9
    Helio T, Halme L, Lappalainen M, et al. CARD15/NOD2 gene variants are associated with familially occurring and complicated forms of Crohn's disease. Gut. 2003; 52: 558562.
  • 10
    Mendoza JL, Murillo LS, Fernandez L, et al. Prevalence of mutations of the NOD2/CARD15 gene and relation to phenotype in Spanish patients with Crohn disease. Scand J Gastroenterol. 2003; 38: 12351240.
  • 11
    Vermeire S, Wild G, Kocher K, et al. CARD15 genetic variation in a Quebec population: prevalence, genotype-phenotype relationship, and haplotype structure. Am J Hum Genet. 2002; 71: 7483.
  • 12
    Arnott ID, Nimmo ER, Drummond HE, et al. NOD2/CARD15, TLR4 and CD14 mutations in Scottish and Irish Crohn's disease patients: evidence for genetic heterogeneity within Europe? Genes Immun. 2004; 5: 417425.
  • 13
    Cario E, Podolsky DK. Differential alteration in intestinal epithelial cell expression of toll-like receptor 3 (TLR3) and TLR4 in inflammatory bowel disease. Infect Immun. 2000; 68: 70107017.
  • 14
    Lakatos PL, Lakatos L, Szalay F, et al. Toll-like receptor 4 and NOD2/CARD15 mutations in Hungarian patients with Crohn's disease: phenotype-genotype correlations. World J Gastroenterol. 2005; 11: 14891495.
  • 15
    Okayama N, Fujimura K, Suehiro Y, et al. Simple genotype analysis of the Asp299Gly polymorphism of the Toll-like receptor-4 gene that is associated with lipopolysaccharide hyporesponsiveness. J Clin Lab Anal. 2002; 16: 5658.
  • 16
    Oostenbrug LE, Drenth JP, de Jong DJ, et al. Association between Toll-like receptor 4 and inflammatory bowel disease. Inflamm Bowel Dis. 2005; 11: 567575.
  • 17
    Torok HP, Glas J, Tonenchi L, et al. Polymorphisms of the lipopolysaccharide-signaling complex in inflammatory bowel disease: association of a mutation in the Toll-like receptor 4 gene with ulcerative colitis. Clin Immunol. 2004; 112: 8591.
  • 18
    Lennard-Jones JE. Classification of inflammatory bowel disease. Scand J Gastroenterol Suppl. 1989; 170: 26.
  • 19
    Wolters FL, Zeijl Gv, Wessels F, et al. Internetbased data inclusion in a European epidemiological follow-up study of IBD patients: a description of methods used. World J Gastroenterol. 2005; 11: 71527158.
  • 20
    Gasche C, Scholmerich J, Brynskov J, 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: 815.
  • 21
    Pierik M, De HG, Vermeire S, et al. Epithelioid granulomas, pattern recognition receptors, and phenotypes of Crohn's disease. Gut. 2005; 54: 223227.
  • 22
    Schmitt C, Humeny A, Becker CM, et al. Polymorphisms of TLR4: rapid genotyping and reduced response to lipopolysaccharide of TLR4 mutant alleles. Clin Chem. 2002; 48: 16611667.
  • 23
    Vermeire S, Joossens S, Peeters M, et al. Comparative study of ASCA (Anti-Saccharomyces cerevisiae antibody) assays in inflammatory bowel disease. Gastroenterology. 2001; 120: 827833.
  • 24
    Peeters M, Joossens S, Vermeire S, et al. Diagnostic value of anti-Saccharomyces cerevisiae and antineutrophil cytoplasmic autoantibodies in inflammatory bowel disease. Am J Gastroenterol. 2001; 96: 730734.
    Direct Link:
  • 25
    Hennekens C, Buring JE. In: MayrentSL. Epidemiology in Medicine. Philadelphia: Lippincott Willians & Wilkins 1987: 9093.
  • 26
    Cavalli-Sforza LL, Menozzi P, Piazza A. The History and Geography of Human Genes. Princeton, NJ: Princeton University Press; 1996.
  • 27
    Hampe J, Grebe J, Nikolaus S, et al. Association of NOD2 (CARD 15) genotype with clinical course of Crohn's disease: a cohort study. Lancet. 2002; 359: 16611665.
  • 28
    Murillo L, Crusius JB, van Bodegraven AA, et al. CARD15 gene and the classification of Crohn's disease. Immunogenetics. 2002; 54: 5961.
  • 29
    Jess T, Riis L, Jespersgaard C, et al. Disease concordance, zygosity, and NOD2/CARD15 status: follow-up of a population-based cohort of Danish twins with inflammatory bowel disease. Am J Gastroenterol. 2005; 100: 24862492.
    Direct Link:
  • 30
    Ekbom A, Helmick C, Zack M, et al. The epidemiology of inflammatory bowel disease: a large, population-based study in Sweden. Gastroenterology. 1991; 100: 350358.
  • 31
    Langholz E, Munkholm P, Nielsen OH, et al. Incidence and prevalence of ulcerative colitis in Copenhagen county from 1962 to 1987. Scand J Gastroenterol. 1991; 26: 12471256.
  • 32
    Loftus EV Jr. Clinical epidemiology of inflammatory bowel disease: incidence, prevalence, and environmental influences. Gastroenterology. 2004; 126: 15041517.
  • 33
    Moum B, Vatn MH, Ekbom A, et al. Incidence of ulcerative colitis and indeterminate colitis in four counties of southeastern Norway, 1990–93. A prospective population-based study. The Inflammatory Bowel South-Eastern Norway (IBSEN) Study Group of Gastroenterologists. Scand J Gastroenterol. 1996; 31: 362366.
  • 34
    Munkholm P, Langholz E, Nielsen OH, et al. Incidence and prevalence of Crohn's disease in the county of Copenhagen, 1962–87: a sixfold increase in incidence. Scand J Gastroenterol. 1992; 27: 609614.
  • 35
    Vind I, Riis L, Jess T, et al. Increasing incidences of inflammatory bowel disease and decreasing surgery rates in Copenhagen City and County, 2003–2005: a population-based study from the Danish Crohn Colitis Database. Am J Gastroenterol. 2005; In press.
  • 36
    Vind I, Vieira A, Hougs L, et al. NOD2/CARD15 gene polymorphisms in Crohn's disease: a genotype-phenotype analysis in Danish and Portuguese patients and controls. Digestion. 2005; 72: 156163.
  • 37
    Bueno de Mesquita M, Henckaerts L, Joossens M, et al. Study on CARD15 and TLR4 polymorphisms in children with early diagnosed IBD and their family members. Gastroenterology. 2005; 128: A440.
  • 38
    Franchimont D, Vermeire S, El HH, et al. Deficient host-bacteria interactions in inflammatory bowel disease? The toll-like receptor (TLR)-4 Asp299gly polymorphism is associated with Crohn's disease and ulcerative colitis. Gut. 2004; 53: 987992.
  • 39
    Gazouli M, Mantzaris G, Kotsinas A, et al. Association between polymorphisms in the Toll-like receptor 4, CD14, and CARD15/NOD2 and inflammatory bowel disease in the Greek population. World J Gastroenterol. 2005; 11: 681685.
  • 40
    Hume GE, Fowler E, Lincoln D, et al. The role of TLR4 Asp200Gly and TLR5 Arg 392Stop in an Australian Inflammatory Bowel disease cohort. Gastroenterology. 2005; 128: A443.
  • 41
    Vasiliauskas EA, Kam LY, Karp LC, et al. Marker antibody expression stratifies Crohn's disease into immunologically homogeneous subgroups with distinct clinical characteristics. Gut. 2000; 47: 487496.
  • 42
    Walker LJ, Aldhous MC, Drummond HE, et al. Anti-Saccharomyces cerevisiae antibodies (ASCA) in Crohn's disease are associated with disease severity but not NOD2/CARD15 mutations. Clin Exp Immunol. 2004; 135: 490496.
  • 43
    Dubinsky MC, Lin YC, Dutridge D, et al. Serum immune responses predict rapid disease progression among children with Crohn's disease: immune responses predict disease progression. Am J Gastroenterol. 2006; 101: 360367.
    Direct Link:
  • 44
    Lennard-Jones JE, Shivananda S. Clinical uniformity of inflammatory bowel disease a presentation and during the first year of disease in the north and south of Europe. EC-IBD Study Group. Eur J Gastroenterol Hepatol. 1997; 9: 353359.