Prevalence of Celiac Disease in Egyptian children and adolescents with Diabetes Mellitus: a Clinical, Biochemical & Histopathologic Study

Prevalence of Celiac Disease in Egyptian children and adolescents with Diabetes Mellitus: a Clinical, Biochemical & Histopathologic Study 

Sherif El Saadany¹, Farrag W², Saleh M.A³, E smail S.A,¹ Menessy A,⁴ and Hala Hamouda⁵

Tropical Medicine1, Internal Medicine², Pediatric³Pathology⁴ and Medical Biochemistry⁵ Departments, Faculty of Medicine, Tanta University, Egypt

ABSTRACT:

Background /Aim: Diagnosis of atypical and silent celiac disease (CD) is important because of its serious complications. The association of type 1 diabetes and CD has been reported worldwide. The aim of this study was to determine the prevalence and clinical, biochemical and histopathologic characteristics of CD among Egyptian children and adolescents with type 1diabetes.

Subjects and methods: A total of 116 children and adolescents with type 1 diabetes (62 males and 54 females, age rage 2-21 yrs) and 25 age and sex matched healthy children were screened for CD using anti-gliadin (AGA), anti-tissue transglutaminase (t-TG), anti-reticulin (AGA) and anti-endomysial (EmA) antibodies. Clinical data, hemoglobin Alc, insulin requirements, hemoglobin concentration, mean red cell volume and serum ferritin levels were evaluated.

Results: Twenty-six (22.4%) patients were positive for AGA and t-TG antibodies, 14 of them were ARA positive. Ten of these patients were EmA positive and four were EmA negative. From the EmA negative patients three sera with IgA deficiency had high IgG class in AGA, anti-t-TG and ARA antibodies. All these 14 patients (EmA positive and negative) underwent intestinal biopsy. Thirteen had histological evidence of CD including the EmA negative patients with IgA deficiency, giving a prevalence of CD in diabetic children of 11.2% (13/116). Compared with the other diabetic patients, those with CD had a significantly higher height SD scores with no statistical significant changes in any other parameters

Conclusion: the prevalence of CD in Egyptian type 1 diabetes children and adolescents was found to the high. Serologic markers for CD are useful for identifying asymptomatic type 1 diabetes children who should undergo a small intestinal biopsy. 
 

                    INTRODUCTION 

Gluten -sensitive enteropathy or celiac disease (CD) is a common cause of chronic malabsorption in children and is characterized by mucosal damage of the small intestine induced by gluten-containing food, including wheat, rye, barley and possibly oats. Clinically, it ranges from silent asymptomatic form to active malabsorption syndrome⁽¹⁾.The reported incidence/ prevalence of CD has varied widely between different geographical locations and over the past several decades. This in part related to the association of disease with genetic factors that are unequally distributed among various racial and ethnic groups⁽²⁾.

Several recent screening studies have found a prevalence of CD of one in 100 to one in 300 ^(3'4). A very high prevalence (5.6%) of CD has been reported in Saharawi children⁽⁵⁾.

Type 1 diabetes is an organ-specific autoimmune disease characterized by cell destruction resulting in glucose intolerance and finally insulin dependence⁽⁶⁾.  Genetic finally insulin dependence⁽⁶⁾. Genetic susceptibility and environmental triggering factor yet to be identified seem to be responsible for type 1 diabetes ^(7,8).Type 1 diabetes has also been reported to be characterized by increased prevalence of various autoantibodies, and of manifest autoimmune disease such as chronic thyroiditis, atrophic gastritis, and celiac disease⁽⁹⁾. An association between type 1 diabetes and CD has been previously reported⁽¹⁰⁾. Both disease are associated with the genetic loci on the short arm of chromosome 6. The genetic susceptibility to CD is strongly associated with the HLA class II antigen DQW2 , aDqab heterodimer encoded by DQA1*0.501 and DQB2* 0201.The most commonly seen HLA types in type 1 diabetes  are HLA DR3, HLA DW3, HLADR4 ad HLAD4 ^(7,11).

Selective IgA deficiency is a primary immunodeficiency disorder characterized by a reduction in the production of IgA antibodies. It may predispose to respiratory infections, gastrointestinal infection, and atuoimmune disease. The increased prevalence of selective IgA deficiency among patients with CD has been reported previously ⁽¹²⁾. The frequently of selective IgA deficiency has also been examined in patients with type I diabetes ⁽¹³⁾.

The prevalence of CD in children with type 1 diabetes ranges between 1.3 and 12% through the world and may contain a high proportion of clinically asymptomatic and atypical cases^(10,14,15). CD is associated with an increased risk of malignancy and may be complicated by other conditions such as osteoporosis, infertility, and neurological disorders, problems that may be averted by gluten   restriction ^(16,17).

Several serological tests are used as a screening method for CD and included antigliadin, antireticulin anti-tissue transglutaminase and anti-endomysial antibodies⁽¹⁸⁾.The main role of these tests in clinical practice is in screening patients with diseases associated with CD and who may have atypical or silent forms of CD, (e.g type 1 diabetes) , in order to avoid an unnecessary intestinal biopsy⁽¹⁹⁾.

Till now there are no available data on the incidence of CD in Egyptian diabetic children. The aim of this study, therefore, was to determine the prevalence and the clinical, biochemical and histopathologic characteristics of CD among Egyptian children and adolescents suffering from type 1 diabetes. 

SUBJECTS AND METHODS

Subjects:

One hundred-sixteen children and adolescents (62 males and 54 females) with type 1 diabetes  attending the outpatient clinics at the Departments of Pediatrics and Internal Medicine (diabetes and endocrinology unit), Tanta University Hospital were enrolled in the study during the period between April 2002 and January 2004. Their age ranged between 2-21 years (mean ±SD, 13.4 ±2.4 yr). Diagnosis of type 1 diabetes was based on World Health Organization criteria⁽²⁰⁾. All patients were receiving human insulin and were asymptomatic i.e none of them had gastrointestinal symptoms suggestive of celiac disease such as diarrhea, abdominal pain, abdominal distention, vomiting or flatulence.

Diabetic patients were subjected to the following:

-Full medical history including (i) age at onset of illness (defined as first injection of insulin), (ii) duration of disease, (iii) daily insulin requirements.

-Thorough clinical examination: including anthropometric assessment. Height, weight and body mass index (BMI) were assessed and expressed as standard deviation score (SDS) using the NCHS standards⁽²¹⁾. 

Methods :

I-Laboratory investigations :

-Complete blood count (CBC):

-Estimation of serum ferritein levels by nephelometry (Array protein system, Beckman, USA).

-Estimation of glycosylated hemoglobin A1C (HbA1c) biochemically by calorimetric methods.

-Urine and stool examination.

-Measurement of total serum immunoglobulin A (IgA) levels by nephelometry.

-Determination of celiac disease-associated antibodies: In all participants a venous blood sample was taken. Serum samples for the determination of IgA- and IgG-antigliadin antibodies (AGA), IgA- and IgG-anti-tissue transglutaminase (tTG), IgA- and IgG- antireticulin antibodies (ARA), and IgA- anti-endomysial antibodies (EmA) were collected from each patient and stored in separate aliquots. These antibody assays were also performed in 25 age-and sex- matched apparently normal children and adolescents. None of them had been suffering from gastrointestinal problems or has a family history of diabetes.

II- Intestinal biopsy:

CD was diagnosed by intestinal biopsy, which showed characteristic morphologic features according to the European Society for Pediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) criteria⁽²²⁾.

-Informed consent was obtained from the parents before the study.

Measurement of gliadin, tissue trans-glatiminase, and reticulin antibodies.

IgA- and IgG -AGA, IgA- and IgG-tTG and IgA - and IgG- ARA antibodies were   measured   by   ELISA  technique ^(23,24,25) Briefly, microtitier plates were coated with 5 ug/ml of purified wheat α gliadin (Sigmas, St. Louis, Mo, USA) for AGA, 5ug/ml of tissue transflutaminase (Sigma, USA) for anti-tTG antibodies, or 5 ug/ml of human recombinant glutathione-5- transferase (GST) reticulin for ARA antibodies⁽²⁵⁾. The following day, sera was applied in triplicate at dilutions of 1: 20 and 1: 100 for IgA and 1: 100 and 1: 500 for IgG. After 2h of incubation, peroxidase conjugated affinity purified swine anti-human IgA or IgG (Binding site, Brimingham, UK) was incubated for 1h, substrate solution was applied , and the colorimetric reaction was recorded at 492 nm by Titretek Multiscan photometer. The results were expressed in percentage of the optical density of positive reference sera. Values for AGA- IgG more than 32 arbitrary units (AU), for AGA -IgA more than 40 AU, for tTG- IgG more than 48 AU, and for ARA -IgG more than 60 AU were considered positive.

Measurement of endomysial antibodies :

This was performed by indirect immunofluorescence on cryostat sections of human umbilical cord ⁽²⁶⁾. Briefly, 6-_m sections of the cord were cut on a cryostat (Diadorum Minn, VSA) at -20oC, the sections were fixed with cold acetone (-4°C) for 10 min and then with chlorofrom for 30 min. at room temperature. After washing three times in Phosphate Buffer Saline (PBS), the section were pretreated with 1% bovine serum albumin (BSA), (Sigma, USA) in PBS for 30 min. The sections were then washed again three times in PBS and incubated with patient's sera diluted 1 : 20 for 30 min. in a humidity chamber. After washing there times in PBS, the sections were incubaed with fluorescein isothiocyanate (FITC)- conjugated antibody against human IgA (Binding site, Brimingham, UK) diluted 1:100, in a humidity chamber in the dark. The slides were washed again, mounted in glycerol and examined under a fluorescent microscope. Both positive and negative reference sera were used in all tests. Positive sera were titrated , the results being expressed as the highest dilution factor yielding a positive fluorescence pattern. All sera manifesting fluorescence (titer > 5) were considered to be positive.

Intestinal biopsy and its evaluation

Fourteen type 1 diabetes patients with positive results in AGA, tTG and ARA antibodies (10 EmA positive and 4 EmA negative) as shown in Figure 1 underwent intestinal mucosal biopsy. Small bowel biopsy was performed under sedation by upper gastrointestinal endoscopy. Precautions were taken to prevent hypoglycemia in fasting diabetic children. Multiple mucosal specimens were obtained from the distal part of the duodenum or the proximal part of the jejunum. The specimens were put in fromalehyde solution and hematoxylin and eosin-stained sections were examined histologically. The intestinal mucosa was assessed according to the scoring system of Marshal⁽²⁷⁾. Grade 0 = normal ; grade 1 = infiltration lesion ; grade ll= partial villous atrophy; and grade III = sub total villus atropy. 

STATISTICAL ANALYSIS

The optical density of reference serum was taken as 100% (100 arbitrary units, AU). The levels of antibodies presented in Table 2 were calculated as a percentage of the reference serum sample ⁽²⁸⁾. The results of estimation of antibody level in the sera of 25 normal children were used to calculate cut-off values defined as the mean value plus 2 SD. Statistical comparisons were carried out with Student's t-test. Correlation between variables were determined using the Spearman's rank test. Significance was assumed if p values were < 0.05. 

RESULTS

The clinical and laboratory characteristics of study patients were shown is Table (1). Figure (1) shows the diabetic patients with positive and negative results of AGA, tTG, ARA and EmA antibodies, and the number of biopies done and the biopsy results.

The first step of investigations was the determination of AGA and tTG antibodies in all diabetic patients. We found that 5 patients (4.3%) were positive for IgG -AGA, 7 patients (6%) for IgA and IgG -AGA, 4 patients (3.4%) were positive for IgG -AGA and IgG -tTG antibodies, and 10 (8.6%) for IgA- and IgG- AGA and IgA and IgG- tTG antibodies.

The second step of investigation was carried out in 26 patients (22.4%) with positive results of AGA and tTG antibodies to perform a serum IgA levels and IgA- and IgG- antireticulin (ARA) antibodies. From the 26 patients, 3 patients (11.5%) were IgA- deficient with serum level < 0.53 g/L and 23patients (88.5%) were within normal IgA levels. The 10 patients with a positive IgA and IgG -ARA antibodies were from the same patients with positive results on IgA-and IgG-AGA and anti tTG antibodies. Four patients with positive results on IgG ARA antibody also showed positive results on IgG AGA antibody and IgG anti-tTG antibody; three of them were found to have a selective IgA deficiency.

The third step of investigation was to determine the presence of IgA EmA in patients ( n =14) with positive IgA-and/ or IgG -ARA antibodies. Ten patients (38.5%) were EmA positive and 4 (15.4%) were EmA negative. The IgA -EmA negative patients were positive for IgG class of AGA, and tTG, and anti ARA antibodies. These patients with positive and   negative  EmA   (n   =14)   results underwent small bowel biopsy.

All the patients with EmA positive (n=10) had typical histological features of CD, demonstrating subtotal or total villous atrophy in the absence of parasites. From the four patients with positive IgG class antibody, with low IgA levels, and negative in EmA, 3 patients showed partial or subtotal villous atrophy and one patient had normal mucosal appearance. A celiac disease was thus confirmed in the 13 patients (6 boys and 7 girls) and this study revealed that the prevalence of CD was 11.2%. Further analysis (Fig. 2) revealed that biopsy specimens from two patients had only increased numbers of intraepthelial lymphocytes (lELs) and received a score of I, indicating possible CD. Four biopsy specimens received a score of II, and 7 specimens received a score of III. Thus 11 children (79%) had definite histologic evidence of CD. In this study, the predictive value of the combined tested antibodies for histologic evidence of CD was between 79% (11 of 14 with Marsh score II or III) and 93 (13 of 14 with Marsh score I,II or III).

Data on clinical, serologic markers and biopsy results and associated manifestations in children with type 1 diabetes and suspected CD are shown in Table (2). Typical geastrointestinal complaints of CD were not detected , but atypical isolated signs or symptoms of CD were evident in particular poor linear growth in 7 patients (53.8%), iron deficiency anemia in 3 (33.1%) and delayed puberty in one patient (7.7%). Short stature (i.e height < 3 rd percentile ) was found in 7 (53.8%) patients whereas weight reduction in 2 (15.4%) patients. Mild hypertransa-minasemia (serum ALT/AST level 3-4 time higher than normal values) without hepatomegaly was observed in 5 (38.5%) patients. Results of other standard liver function tests were in the normal range, and serologic markers for hepatis A, B and C were not detected. Ultrasound revealed mild fatty liver (steatosis) in 3 children. Liver biopsy showed histologic evidence of micro/macro vesicular steatosis in 3 and minimal inflammatory changes in the other 2 children.

Comparison of clinical and laboratory data between diabetic patients with and without CD are shows in Table (3). There were no significant differences between the two groups with regard to age, duration and diabetes, body weight, BMI, insulin requirement and glycated Hb. Height SDS measurements were significantly different between the two groups of patients (P < 0.01). A difference in hematological indices (Hb, MCV and serum ferritin) was observed between the two groups, however it was not statistically significant.

A significant correlation was identified between Marsh score and both tTG and EmA antibody titers (Table 4). In addition, all four subjects with EmA strongly positive results had biopsy specimens with a Marsh score of III. Thus high titer of EmA antibodies had a 100% positive predictive value for small bowel biopsy evidence of CD. Height SDS inversely correlated with Marsh score (Fig. 3), whereas weight SDS and BMI SDS did not. 
 
 
 
 

Table (1): Clinical and laboratory characteristics of study of type 1 diabetes patients

Data expressed as mean ±SD

Fig. (1): flow diagram of celiac-disease associated antibody screening schedule and small bowel biopsy in 116 type I diabetes children. 

Fig. 2  : Distribution of small bowel biopsy histology score according to system of Marsh in 14 children with type 1 diabetes screened for celiac disease. 
 

Table ( 3 ) : Comparison of clinical and laboratory data between type 1 diabetes patients with and without celiac disease.

Data expressed as mean ±SD

NS, non significant. 
 

Table (4): Relationship of serologic markers and auxological data with small bowel histology in type 1 diabetes with celiac disease (n=13)

 

Figure 3. Relationship between small bowel histology score according to system of Marsh and SD scores for height.

                          DISCUSSION  

Type 1 diabetes  and celiac disease are disorders that show distinct similarities as an underlying genetic susceptibility associated with HLA class II antigens and autoimmune phenomena but striking geographic differences in the prevalence rates. One likely explanation for the frequent simultaneous occurrence of these two diseases is their partly similar genetic background : Both are associated with HLA DQA1* 501 DQB1 * 201⁽²⁹⁾. Recently , a linkage of CD and a region on chromosome 15 q 26 was described -a region where the type 1 diabetes -3 susceptibility locus for type I diabetes has been mapped ⁽³⁰⁾. This locus may in addition confer to the concurrent existence of CD and type 1 diabetes.

In the present study we found that the prevalence of CD in type 1 diabetes children and adolescents was 11.2%. The diagnosis of CD was confirmed in 13 cases from 116 patients by serological analysis and intestinal biopsy histologically consistent with CD. Similar CD prevalence rates in type 1 diabetes children have been described for the European countries^(14,3,32). There was a scarcity of data assessing the prevalence of CD in diabetic patients among the Arab population. There is only one study available to date from Western Alegria where a higher rate (16.4%) of CD was reported in diabetic children⁽³³⁾. However, the investigators combined previously diagnosed cases, with CD cases found by serologic screening, and the incidence was therefore relatively high. In addition, a recent abstract by Shabib et al., ⁽³⁴⁾ showed a prevalence of 8.1% in symptom -free type 1 diabetic children in Saudia Arabia.

Variably lower prevalence rates in diabetic patients at approximately 1.4% to 5.8% have been reported for Germany ⁽³⁵⁾, United States ⁽³⁶⁾, and Scotland051. The different rates reported between these studies may be due to a number of factors including methodological reasons (differences in type of antibody used for screening), the size and age of the cohort, and possible genuine underlying geographic differences that may be explained by genetic or environmental factors.

The diagnosis of CD in our study was confirmed in 10 patients with positive EmA results and in 3 patients with EmA negative, who consented to intestinal biopsy. The EmA test is widely recognized to be the most specific and sensitive serologic test for the diagnosis and screening of CD, having a high correlation with mucosal findings⁽³²⁾ as was remarkably observed in our study. Three of the four patients in this study with biopsy-proven CD, but EmA negative results were subsequently found to have selective IgA deficiency. All three were identified because they had isolated elevation of their IgG class in AGA, tTG and ARA levels prompting us to determine their serum IgA levels. Consistently, Collin et al.,⁽³⁷⁾ and Rittmeyer et al.,⁽³⁸⁾ reported that selective IgA deficiency is considered a risk factor for celiac disease and occurring with a high frequency of approximately 3% in both diabetes and CD.

The clinical features of CD can vary between the typical form with malabsorption, and silent disease with flat intestinal mucosa without symptoms of illness and a potential form with positive autoantibodies and /or increased intraepithelial lymphocytes only ⁽¹⁾. All of histologically-confirmed cases of CD in our study were asymptomatic at their presentation. As in other studies ⁽³²⁾, we found CD is nearly equal in diabetic males and females. In most patients with type 1 diabetes the diagnosis (silent) CD was made using screening studies. The reported rates are probably underestimates, because many patients   with   diabetes   with   positive antibody results did not agree to small bowel biopsy⁽³⁵⁾.

The evaluation of demographics, including age, sex of patients, duration of diabetes, and glycemic control in our study did not help identifying diabetic patients with CD. Investigators in several studies have reported that CD develops later during the follow -up of patients with diabetes ⁽³⁹⁾ and not at the time of diagnosis of diabetes. This is in line with observations in our cross-sectional study, that the duration of diabetes in patients with CD was comparable to those without CD.

The clinical significance of celiac disease in diabetic children with no symptoms is worthwhile. There may be considerable benefits derived from screening and early detection of CD. Without treatment , long-term risks may include malignant disease of the small intestine particularly lymphoma⁽¹⁶⁾, osteoporosis⁽¹⁷⁾, and other autoimmune diseases ⁽⁴⁰⁾. Moreover, several studies reported that the introduction of gluten-free diet (GFD) may lead to an improvement in diabetic control and to a reduction in hypoglycemic episodes, in addition to reduction of CD-associated risks.

We have identified a possible clinical effect of CD in our type 1 diabetes. Growth parameter assessment showed a significant decrease in height SDS values in children with the combination of CD and diabetes . In further analysis , short stature was observed in 7 (53.8%) of the 13 celiac type 1 diabetes. This occurs in spite of glycemic control as reflected by equivalent results of HbA_1c values and insulin requirements in these patients as compared to the other diabetic subjects. Thus, this findings have suggested the underlying diagnosis in our asymptomatic patients and reflects the significant adverse effects of the sufficient long duration of gluten exposure. Other studies described retarded linear growth in type 1 diabetic children with CD with a frequency ranging from 30% to 96% ^(34,36). Our data confirm that all diabetic children should be screened for CD, particularly if they have stunted growth. Growth impairment may be the result of nutritional deficiencies, mediators of chronic inflammation such as y interferon and tumor necrosis factor-α^(1,41), or other factors.

Despite the existence of abnormal mucosal appearances on biopsy, evidence suggesting disruption to the absorptive processes of the small intestine, in the present study , was otherwise lacking. It is noticeable that sporadic cases presenting with clinical features of hypochromic/ microcytic anemia (3 cases) and underweight (2 cases) while the remaining subjects were all entirely asymptomatic. Mean weight SDS and body mass index SDS values as well as hematological indices (Hb concentration, MCV and serum ferritin level) in subjects with mucosal changes were found to be not significantly different from the other diabetic patients. Similar observations have been reported in other studies in type 1 diabetes^(36,39), although other have reported a high frequency of hematological abnormalities in association with CD⁽⁴²⁾.

One of the manifestations of the expanding atypical spectrum of celiac disease is the important finding that five of the 13 (38.5%) diabetic subjects with CD in the current study were exhibiting liver dysfunction as evidenced by mild hypertransaminasemia. This observation suggests that elevated serum aminotransferase activity, unrelated to the commonest causes of liver disease in children, in our asymptomatic patients is an early manifestation of celiac disease in children with type 1 diabetes . These results confirm preliminary observations that elevated serum aminotransferase activity may be an early sign of  gluten-sensitive enteropathy, suggesting that CD should be considered when elevated ALT and /or AST levels of unknown origin is present in children⁽⁴³⁾. Furthermore, liver biopsy of the five studied patients revealed mild hepatic macrovesicular steatosis in 3 of them and non-specific inflammation with mild fibrosis in the other two cases. These histologic features are in agreement with most series dealing with celiac patients^(43,44).

Increased serum levels of hepatic transaminase enzymes have recently been reported to be present in patients with CD at the time of diagnosis , with a prevalence of about 40% in adults ⁽⁴⁴⁾ and 54% in children⁽⁴⁵⁾. The pathogenesis of liver dysfunction in patients with CD is unclear. CD and liver disease may share common risk factors, and the consequences of CD may themselves predispose to liver dysfunction. It was found that autoimmune liver disease and CD occur in patients who possess certain HLA class II molecules and haplotypes⁽¹¹⁾. The trigger for the mucosal inflammatory process in CD may potentially predisposes to autoimmune reaction in the liver. Maggiore et al.⁽⁴⁵⁾, suggested that the antigen involved in liver disease could either be derived from the damaged mucosa or from the dietary peptide gliadin. Biochemical normalization and histologic remission of liver abnormalities have been achieved in celiac children following gluten exclusion.⁽⁴⁵⁾

The observation that a positive combination of antibody assay had a positive predictive value (PPV) of 79% to 93% for histologic evidence of CD in the current study is comparable to other reports ⁽³²⁾. However, this value is lower than that reported in symptomatic children (98%)⁽¹⁸⁾.

The correlation between either tTG and EmA antibody titers and severity of intestinal  injury,  demonstrated  in  this study, suggests that antibodies may have a role in immunologic injury, even though such injury is cell –mediated⁽⁴¹⁾.

Type 1 diabetes is considered to be an autoimmune disease caused by the destruction of pancreatic (3-cells by T cells (M6). The most important genetic determinants are located within HLA complex on chromosome 6. Diabetic patients carrying the HLA types B 8 and DR 3 frequently show other autoimmune endocrinopathies in contrast to patients with HLA type B 15 andDR4⁽⁴⁷⁾.

In addition to the disease-associated risk genes, environmental factors are important because even identical twins have only a 30 to 60% concordance for DM.⁽⁴⁸⁾. The enteroviral infection and cow's milk proteins were suggested as the two main environmental factors influencing the development of type 1 diabetes, probably by initiating or modifying an autoimmune process^(6,8,49). Several epidemiological and ecological investigations have suggested that early exposure to cow's milk proteins during infancy as well as a short duration of breast feeding are risk factors for induction of the autoimmune process leading to DM⁽⁴⁹⁾.

Some cow's milk proteins, such as BSA, (3-lactoglobulin, and p-casein have antigenic similarities to an islet-cell antigen and are thus immunogenic determinants of type 1 diabetes⁽⁵⁰⁾.The food antigen gluten also has been suggested to play a role in type 1diabetes. The increased association between I type 1 diabetes and CD⁽¹⁰⁾, the reduced prevalence of type 1 diabetes  autoimmunity after gluten deprivation in patients with CD ⁽⁵¹⁾, and the reduced incidence of autoimmune diabetes in the nonobese diabetic (NOD) mouse receiving a gluten-free diet⁽⁵²⁾ are some of the evidence supporting this claim. The gut immune system can play a role in the triggering of diabetes similarly as known in celiac disease⁽⁵³⁾. 

CONCLUSIONS AND RECOMMENDATIONS

The present study revealed an estimated prevalence of CD of 11.2% among children and adolescents with type 1 diabetes by screening the serologic markers anti-gliadin, reticulin, tissue transglutaminase and endomysial antibodies followed by intestinal biopsy in test positive patients, which is distinctly higher than that in several other countries. This relatively high incidence of CD in the asymptomatic children and adolescents is not fortuitous and suggests that serologic analysis using highly sensitive and specific tests for CD should be included in the health assessment of patients with type 1 diabetes. This will help the patients to cope early with this additional burden. Long-term and follow-up studies of a larger population of type 1 diabetes   will be required to determine the need for gluten-free diet. Finally, in future, further efforts to improve knowledge about the genetic background of Egyptian children with coexisting type 1 diabetes and CD are warranted. 

ABBREVIATIONS

DM, diabetes mellitus; ; CD, celiac disease IgA and   IgG, Immunoglobulin A and G; AGA, antigliadin antibodies, tTG, anti-tissue transglutaminase; ARA. antireticulin antibodies; EmA; anti-endomysial antibodies. GFD, gluten-free diet; HbA,c glycosylated hemoglobin, MCV, mean red cell volume; ALT, and AST, alanine and aspartate aminotransferases; Ht., height; Wt., weight; BMI, body mass index; SDS, standard deviation scores. 

REFERENCES

1. Catassi C and Fabiami E. The spectrum of celiac disease in children. Baillieres Clin Gastroenterol. 11: 487, 1997.
2. Ascher H, Holm K and Maki M. Different features of celiac disease in two neighbouring countries. Arch Dis Child 69: 375, 1993.
3. Dieterich   W,   Laag    E   and   Schopper  H. Autoantibodies  to  tissue  transglutaminase  as predictors of celiac disease. Gastroenterology. 115 : 1317, 1998.
4. Troncone R,  Maurano F,  and Rossi M.  IgA antibodies to tissues transglutaminase: an effective diagnostic test for celiac disease. J Pediatr 134 : 166, 1999.
5. Catassi C, Ratsch IM and Gandolfi L Why is coeliac disease endemic In the people of the Sahara? Lancet 354:647,1999.
6. Atkinson M and Eisenbarth GS. Type 1 diabeties : new perspectives on disease pathogenesis and treatment. Lancet 358:221, 2001.
7. Davies JL, and Bennett ST. A genome wide search for human type 1 diabetes susceptibility genes. Nature 371: 130, 1994.
8. Leslie RD and Elliott RB. Early environmental events  as  a  cause  of IDDM:  evidence  and implications. Diabetes. 43:843, 1994.
9. Elsenbarthe GS. Type 1 diabetes mellitus. N Engl JMed 314: 1360, 1986.
10. Gronin CC, and Shanahon F. Insulin dependent diabetes mellitus and celiac disease. Diab Med, 18: 169, 2001.
11. Sollid LM and Thorsby E. HLA susceptibility genes in coeliac disease : genetic mapping and role in pathogenesis. Gastroenterology 105 : 910, 1993.
12. Catal do F, Marino V and Bottaro G. Celiac disease and selective IgA deficiency . J Pediatr. 131:306, 1997.
13. Hoddinott  S,   Dornon   J   and   Farid   NR. Immunoglobulin levels in typei diabetes mellitus diabetologia 32: 326, 1991.
14. De  Vitis I and Ghirlanda G. Prevalence of coeliac disease in type 1 diabetes. A multicenter study. Acta Paediatr; 412: (Suppl) 56, 1996.
15. Gillett HR , Noques K and Kelnar C. Screening for celiac disease in the pediatric insulin-dependent diabetic population of South East Scotland. Horm Res. 48: 145, 1997
16. Ferguson A and Kingstone K. Coeliac disease and malignancies. Acta Paediatr. 85 (Suppl) 412: 78, 1996.
17. Kemppainen  T,   Kroger H and Kosma   V. Osteoporosis in adult patients with celiac disease. Bone, 24 :244,1999.
18. EMRC/ESPGAN Group Meeting. Serological Screening for Celiac Disease, Aikammer-Kirrwiler, Germany, November 1994.
19. Grodzinsky E, Hed J and Skogh T. IgA anti- endomysium antibodies have a  high positive predictive    value    for   coeliac   disease    in asymptomatic patients. Allergy 49:596, 1994.
20. World Health Organization Group. Report of diabetes mellitus technical report, series no. 727. World Health Organization, Geneva, 1985.
21. Hamil PV, Drizd TA, Reed RB and Roche AF. Physical growth:  National  Center for Health Statistics percentiles. Am J Clin Nutr 32: 607, 1979.
22. Walker. Simth JA, Schitz J and Visakorpi JK. Revised criteria for the diagnosis of celiace disease.  Report from a working group . Arch Dis Child. 65 : 909, 1990.
23. Ascher H and Lanner A. Anew laboratory kit for anti-gliadin IgA at diagnosis and follow-up of childhood celiac disease. J Pediatr Gastroenterol Nutr 10 : 443, 1990.
24. Dieterich W, Ehnis T and Bauer M. Identification of the tissue transglutaminase as the autoantigen of coeliac disea'e. Nat Med. 3: 797,199
25. Sanchez D, Tuckova L and Sebo P. Occurrence of IgA autoantibodies to calreticulin in coeliac disease and various autoimmune disease. J Autoimmunity 15:441, 2000.
26. Grodzinsky E, Jansson G and Skog T. Anti-endomysium and anti-gliadin antibodies as serological markers of coeliac disease : a clinical study to develop a practical routine. Ada Paediatr 84:294,1995.
27. Marsh MN and Crow Pt. Morphology of the mucosal lesion in gluten sensitivity. Baillieres Clin Gastroenterol 9:273, 1995.
28. Ascher M, Hahn Soric M and Hanson LA. Value of the serological markers for clinical diagnosis and population studies of coeliac diseases. Scand J Gstroenterol 31: 61, 1996.
29. Sollid LM, Gerde H and Vartdal F. Evidence for primary association of coeliac diseae to a particular HLA.DQalb heterodimer. J Exp Med.69:345, 1989.
30. Houlston RS,  Tomlinson IP and Ford D. Linkage analysis of candidate regions      celiac disease genes. Hum Mol Genet 9: 1353, 2000.
31. Cronin  CC,   Ferriss  B and Liddy C.   High prevalence of celiac disease among patinets with insluin dependent diabetes mellitus. Gastroenterol. 92:210,1997.
32. Schober E, Bittmann B, Huppe A and Jager A. Screening by anti-endomysium antibody for celiac disease in diabetic children and adolescents in Austria. J Ped. Gastroenterol Nutr, 30:391, 2000.
33. Boudraa G, Hachelaf W, Benbousbdellah M and Belkadi M. Prevalence of celiac disease in diabetic children and their first-degreerelatives in West Algeria : screening with serological markers. Ada Paediatr 442 (Suppl: 58, 1996.
34. Shabib   SM,   Attia   N   and   AL-Ashwal  A. Prevalence and characteristics of celiac disease in insulin  dependent diabetes  mellitus in  Saudi Arabia. Saudi J Gastroenterol 9: 110, 2003.
35. Kordonouri O, Dietrich W, Becker Becker M and Muller C. Antibodies to tissue transglutaminase predict latent celiac disease in  children  with diabetes mellitus. Diab Res Clin Pract 44 : 6, 1999.
36. Bird G Ahmed ML and Rossk M G. Coeliac disease in children and adolescents with IDDM: clinical characteristics and response to gluten free diet. Diab Med 15: 38, 1998.
37. Collin P, Maki M, and Hallston O. Selective IgA deficiency   and   celiac   disease.   Scand  J. Gastroenterol 27:367,1992.
38. Rittmeyer C and Rhoads JM. IgA deficinency cause false negative endomysial antibody results in celiac disease . Pediatr Gastroenterol Nutr. 23: 504, 1996.
39. Soukkonen T, Salivaht. E and Reijonen H. Childhood diabetes diabetes in Finland Study Group Coeliac Disease : Frequent occurrence after clinical onset of IDDM. Diab Med. 13:464, 1996.
40. Ventura A, Magazzu G and Greco L. Duration of exposure to gluten and risk for autoimmune disorders in patients with celiac disease. Gastroenterology 117:297, 1999.
41. Pena A, Garrote J and Crusius J. Advances in the immunogenetics of celiac disease. Clues for understanding the pathogenesis and disease heterogeneity . Scand J Gastroenterol 38 : 56, 2002.
42. Lorini R, Vitali L, Avanzini M and Severi F. Clinical aspects of coeliac disease in children with insulin-dependent diabetes mellitus, J Pediatr Endocrinol Metab 9: 101, 1996.
43. Lorini R.Maggiore G and DE Giacomo C. Elevated serum aminotrans ferase activity : an early manifestation of gluten- sensitive enteropathy in children with insulin dependent diabetes mellitus. J. Pediatr 124 :995, 1994.
44. Novacek G and Wrba F. Prevalence and clinical importance of hypertransaminasemia in coeliac disease Eur J Gastroenteral Hepatol 11:283, 1999.
45. Moggiore   G,   Massei  F  and   Ventural A. Autoimmune liver disorders and coeliac disease in children :  The SIGEP-CALD Study.  J Pediatr Gastroenterol Nutr 32:387,2001.
46. Tisch R and Me Devitt H. Insulin dependent diabetes mellitus. Cell 85:291, 1991.
47. Throsby E.  HLA-associated diseases Hum Immunol 53: 1, 1997.
48. Olmos P, Hern R Heaton DA. The significance of the concordance rate for type 1 diabetes in identical twins .Diabetologia 31: 747, 1998.
49. Gerstein HC. Cow's milk exposure and type 1 diabetes mellitus. Diabetes Care 17: 13, 1994.
50. Vaarala O, KlemettP and Savilahti E. Cellular immune response to cow's milk (1-Lactoglobulins in recent onset-insulin dependent diabetes. Diabetes 45: 178, 1996.
51. Ventura A, Neri E and Vghi C. Gluten-dependent diabetes-related and thyroid- relaed autoantibodies in patients with celiac disease. J Pediatr 137:263, 2000.
52. Fuda DP, Kaas A end Trods B. Gluten- free diet prevents diabetes in NOD mice. Diabetes Metab Res Rev 15:323, 1999.
53. Vaarala O. Gut and the induction of immune tolerance in type 1 diabetes . Diabetes Metalo Res Rev. 15:353, 1999.