Demographic Data. Thirty-three (35.1%) of the coe­liac individuals were diagnosed between the ages of 31 and 45, with 26 (27.1%) diagnosed at age 46 years and over. Nineteen (20.2%) patients were diagnosed at an age younger than 15, while 16 (17.0%) were diagnosed between 16 and 30 years of age. The mean age of diag­nosis for all the patients was 34 years.

      The average age at diagnosis for the male coeliac pa­tients was 32 years, while for the female coeliac patients the average age at diagnosis was 34 years. An independent Student’s t-test, done between the age at diagnosis of males and females, showed no statistical significance (t = –0.65, p = 0.51), signifying that there is no difference in age at diagnosis between the two genders. Among the female population, the highest proportion (39%) were diagnosed between 31-45 years, with lower percentages [30% (46 or more), 17% (15 or less) and 14% (16-30)] for the other age groups. Among the male patients, the age distribution was approximately the same for all age groups. Twenty-six percent for age groups 15 or less, 16-30 years, 31-45 years, and 22% for the over 46 age group.

      The CTLA4 –318 (C→T) and +49 (A→G) Geno­types. The two polymorphisms were in Hardy-Weinberg equilibrium in both the neonatal control and coeliac popu­lation. The results are presented in Table 1. Possible asso­ciation between the studied CTLA4 gene polymorphisms and CD was tested by comparing the relative frequencies of the different genotypes between the neonatal controls and the coeliac samples. Using the Pearson chi-square test, no statistical difference was observed between the geno­types of the CTLA4 –318 (C→T) polymorphism among the neonatal controls and coeliac patients [χ2 = 1.20; degrees of freedom (df) = 1; p = 0.27]. The same test was used for the frequency of the CTLA4 +49 (A→G) geno­types among the neonatal controls and coeliac samples and showed no statistical difference (χ2 = 2.24; df = 1; p = 0.13). The Fisher’s exact test was employed to test for alleles of the neonatal controls with CD in both SNPs, but no statistically significant results were obtained for the –318 (C→T) (p = 0.54) and +49 (A→G) (p = 0.46) SNPs.

      Linkage Disequilibrium Between the –318 (C→T) and +49 (A→G) Polymorphisms. Evidence of linkage disequilibrium between alleles C and A was observed in the control group (χ2 = 10.49; p = 0.02). Haplotype fre­quencies were constructed by the estimated haplotypes program (ftp:/linkage.rockefeller.edu/software/utilities). The distribution of haplotype frequencies among the coe­liac patients and neonatal controls are shown in Table 2. No significant difference was observed between the two groups (χ2 = 0.04; p = 1.0).