Literature DB >> 30700920

Prevalence of markers of celiac disease in Colombian children with diabetes mellitus type 1.

Carlos Alberto Velasco-Benítez1,2, Ángeles Ruíz-Extremera3, Audrey Mary Matallana-Rhoades1,4, Sandra Carolina Giraldo-Lora1, Claudia Jimena Ortíz-Rivera1.   

Abstract

INTRODUCTION: Although the association between diabetes mellitus type 1 (T1DM) and celiac disease (CD) is well established; there are only a few studies that focus on South American children, haplotypes and their possible associations.
OBJECTIVE: To determine the prevalence of CD markers in a group of children with T1DM and to analyze the associated clinical, immunological and genetic manifestations.
METHODS: A prevalence study focusing on children with T1DM who were assessed based on variables including sociodemographics, anthropometric information, disease characteristics, laboratory results and family medical history. In partitipants a positive tTG2 (Ig A anti-transglutaminase), a duodenal biopsy and genotype were performed. The proportion of children with T1DM and CD was estimated (CI 95%). Determinations of central tendency, univariate and bivariate analysis, were also performed; p <0.05 was considered significant.
RESULTS: Thirteen (8.4%) of the 155 children (53.6% girls, 11.0 ±3.6 years, 2-18 years) with T1DM were tTG2 positive, four had CD (2.6%), seven had potential CD (4.5%) and nine were HLA DQ2/DQ8 positive (5.8%). Children with T1DM and CD had their last ketoacidotic episode (21.5 ±30.4 months versus 69.5 ±38.8 months, p= 0.0260) earlier than children with T1DM and potential CD. There were no differences with anthropometry or with the laboratory results regarding glycemic control.
CONCLUSIONS: The prevalence of CD in these children with T1DM is higher than that reported in other South American countries. The prevalence of CD was found to be associated with the time of presentation of T1DM and its main allele, the DQ2/DQ8. These findings are different from what has been described in other places around the world.

Entities:  

Keywords:  HLA-DQ antigens; Prevalence; Transglutaminases; celiac disease; children; genotype; potential celiac disease; prolamins, glutens; type 1 diabetes mellitus

Mesh:

Substances:

Year:  2018        PMID: 30700920      PMCID: PMC6342086          DOI: 10.25100/cm.v49i3.3650

Source DB:  PubMed          Journal:  Colomb Med (Cali)        ISSN: 0120-8322


Introduction

Celiac disease (CD) is a systemic disorder mediated by the immune system and caused by gluten and related prolamins, it is found in genetically susceptible individuals and is characterized by the presence of a variable combination of gluten-dependent clinical manifestations, CD-specific antibodies, HLA-DQ2 or HLA-DQ8 haplotypes and enteropathy . The association between type 1 diabetes mellitus (T1DM) and CD is well established; however, in recent years, several studies have shown that the prevalence of CD in diabetic patients is even higher than previously thought . The guidelines for the diagnosis of CD from the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) suggest testing asymptomatic children for CD when they have an increased genetic risk for developing it, as is seen in patients with type 1 diabetes mellitus (T1DM) . The highest prevalence reported worldwide of CD is in the Sahara Desert, Africa (5.6%), followed by Oceania (1.2%), Europe (1.0%), the United States (0.8%), Asia (0.3% to 0.7%), and Brazil and Argentina (0.1% to 0.6%) . This prevalence is higher in patients with T1DM (2.4% to 16.4%) . In a systematic and meta-regression review conducted in Colombia, it is concluded that CD seems to be a rare condition among Colombians . It is well known worldwide that there is a higher prevalence of CD in children with T1DM compared to the general population , and that most of these children are asymptomatic at the time of diagnosis , . However, it has yet to be established in Colombia whether this risk is greater in children under 5 years old (as has been demonstrated by the Europeans) , whether there is a predominance of the female gender to develop CD (as there is in the general adult population) , and whether there is a relationship among CD, T1DM and glycemic control (as is still controversial) - . Additionally, there have been no previous Colombian studies about the prevalence rates, clinical characteristics and laboratory results in children with coexisting CD and T1DM. Therefore, the databases of the Pediatric Endocrinology Service of the Hospital Universitario del Valle "Evaristo García" in Cali, Colombia were analyzed. The data of children diagnosed with T1DM were analyzed in order to determine the prevalence of CD markers in a group of children with T1DM. Clinical, immunological and genetic manifestations were included in the analysis.

Materials and Methods

T1DM was diagnosed in participants when there was evidence of beta cell destruction in children older than 6 months of age, regardless of whether they presented with ketoacidosis, and whether they had other autoimmune diseases. T1DM was diagnosed in participants older than 10 years of age if they were obese, in the same way, until there is evidence of absence of autoimmunity . CD was diagnosed when anti-transglutaminase IgA (tTG2) was positive, the HLA DQ2 and HLA DQ8 haplotypes were compatible, and when duodenum histology showed intestinal villus abnormalities with Marsh grade II or higher in the setting of gluten-dependent clinical manifestations. Potential CD was diagnosed in the absence of histological abnormalities on the duodenal biopsy when the tTG2 was positive and the HLA haplotypes were compatible, with or without signs and symptoms . An observational prevalence study was performed in children diagnosed with T1DM who presented between 2 August 2013 and 23 February 2017 at the Pediatric Endocrinology Service of the Hospital Universitario del Valle "Evaristo García" in Cali, Colombia. This hospital is a third-level care institution located in the southwest of the country. The inclusion criteria were children diagnosed with T1DM, both male and female; age older than 6 months of age; and previous consumption of gluten in their complementary diet. The exclusion criteria were children diagnosed with congenital diabetes mellitus, being in a ketoacidotic coma, the presence of associated chromosomal abnormalities such as Down syndrome, the presence of other associated autoimmune diseases such as hypothyroidism, a previous diagnosis of CD, and the presence of inflammatory bowel disease. Sociodemographic variables that were considered included age, sex, race and origin. Anthropometric variables included weight and height. The study also considered the length of time since being diagnosed with T1DM, the number of ketoacidotic comas and the date of the last ketoacidotic coma. Laboratory results, such as glycosylated hemoglobin, hemoglobin and glycemia were recorded, and the metabolic relationships between hypertension, diabetes, being overweight and obesity were considered. The digestive clinical symptoms that were studied were constipation, vomiting, abdominal distension, steatorrhea, diarrhea, abdominal pain, flatulence and weight loss. Screening was carried out using an anti-transglutaminase IgA (tTG2) Biocard ™ Celiac Test (Ani Biotech, Vantaa, Finland, 97.4% sensitivity, 96.9% specificity) , . In the participants with a positive tTG2, a minimum of 4 biopsies of the bulb and second duodenal portion were taken. The evaluation of the biopsy material was made by the same pathologist after hematoxylin-eosin staining. Immunohistochemistry of the common leukocyte antigen was used for the evaluation of the intraepithelial leukocyte count, and the presence of more than 30 intraepithelial lymphocytes versus 100 epithelial cells detected by common leukocyte antigen was considered as intraepithelial lymphocytosis , . According to Marsh-Oberhuber histopathology, the sample was classified as Marsh I when the results showed an "infiltrative lesion", as Marsh II when there was an "infiltrative-hyperplastic lesion", and as Marsh III when there was "hairy atrophy" (partial IIIa, subtotal IIIb and total IIIc) . These children underwent genotyping of HLA DQ2 and HLA DQ8, by polymerase chain reaction . The parents or guardians of the participants signed informed consent forms, as did the children who were over 8 years of age. This study was approved by the Ethics Committee of the Universidad del Valle and the Hospital Universitario del Valle "Evaristo García" in Cali, Colombia. The sample size included all children diagnosed with T1DM who met the inclusion criteria of the study, agreed to participate in the study and presented to the Pediatric Endocrinology Service of the Hospital Universitario del Valle "Evaristo García" in Cali, Colombia. Percentages, percentiles, averages, medians and other descriptive measures were estimated to a CI 95% with their corresponding standard deviations and ranges. To evaluate the possible associations, univariate analysis was performed for each of the variables. In addition, we explored the possible association between the variables of exposure of greatest interest and other covariates, and between the outcome variable of interest (CD) and the other covariates, in order to evaluate the possible existence of confusion. To do this, graphs and 2x2 tables were constructed and the ORs with their respective confidence intervals (95%) were estimated. To assess the statistical significance, the Fisher`s exact test was used and a p <0.05, two-tailed value, was considered statistically significant.

Results

A total of 155 children were included (83 male), with an average age of 11 years (2 to 18) and a diagnosis of T1DM. They had an average duration of the disease of 46.1 months (0 to 180 months), with an average of one episode of ketoacidotic coma (0 to 7 episodes). The last average ketoacidotic episode was 20.5 months prior (0 to 163 months). The majority (73.6%, 114/155) of the participants presented with some family history of metabolic disorders. The laboratory results were: glycosylated hemoglobin of 9.1% (5.2 to 19.0%), glycemia of 193.1 g/dL (36.1 to 600.0 g/dL) and hemoglobin of 12.9 g/dL (9.1 to 15.0 g/dL). Most of the participants came from urban cities (n=109, 70.3%). The tTG2 was positive in 8.4% (13/155) of the cases. The anthropometric parameters are shown in Table 1.
Table 1

Characteristics of children with type 1 diabetes mellitus and anti-transglutaminase IgA (n = 155)

VariablestTG2
Positive (n= 13)Negative (n= 142)
Sociodemographic 8.4%. (CI 95%: 6.2 to 10.6) 91.6% (CI 95%, 89.3 to 93.8)
Age (years) X (range)10 (5-16)11 (2-18)
Sex (female:male)7:676:66
Origin (urban:rural)10:399:43
Race (white:other)6:773:69
Antecedents    
Family members with metabolic diseases (n, %)10 (76.9)104 (73.2)
Evolution of the disease    
Duration (months) X (range)38.1 (0-140)46.8 (0-180)
No. ketoacidotic comas X (rank)1 (0-3)1 (0-7)
Last ketoacidotic coma (months) X (rank)36.4 (0-97)19.4 (0-163)
Nutritional status according to WHO
Normal:Malnutrition8:5102:40
Normal:Altered height13:0126:16
Laboratory results    
Hemoglobin (g/dL) X (range)13.2 (11.0-15.8)12.9 (9.1-15.8)
Glycosylated hemoglobin (%) X (range)9.3 (5.6-12.0)9.1 (5.2-9.0)
Glycemia (g/dL) X (range)179.2 (57-560)194.4 (36.1-600.0)

X=mean; n=number

X=mean; n=number Of the 155 children included, 13 had positive tTG2 (8.4%), and eleven of them underwent endoscopy and biopsy of the upper digestive tract (EUDT). Their HLA DQ2 and HLA DQ8 haplotypes were also determined (Fig. 1). The 4 children with CD as determined by positive immunohistochemistry had a number and location consistent with CD in the CD3, CD8 and CD45 antibodies compared with appropriate controls. In Table 2, the general characteristics of the 11 children with T1DM, potential CD (n= 7) and CD (n= 4) are described.
Figure 1

Flow chart for the study of children with T1DM and CD

Table 2

General characteristics in children with type 1 diabetes mellitus and celiac disease (n = 11)

Age (years) SexRaceOriginDigestive Symptoms Nutritional Status t T1DM (months) HbA1cImmuno histochemistryMarshHLADx
16FMixedUrbanDiarrheaNormal418.2PositiveIIDQB1* 02:01CD
FlatulenceDQB1* 03:02*(08)
Distension
Vomiting
11MAfroRuralNoNormal3511.0PositiveIIDQB1* 02:02CD
DQB1* 03:02*(08)
9FWhiteUrbanNoOverweight6112.0Negative0DQB1* 02:01pCD
DQB1* 03:02*(08)
9MWhiteUrbanVomitingOverweight609.7Negative0DQB1* 02:01pCD
DQB1* 02:02
5FWhiteRuralAbdominal painNormal188.3Negative0DQB1* 02:01pCD
10MMixedUrbanAbdominal painNormal<19.2PositiveIIDQB1* 02:01CD
VomitingDQB1* 03:02*(08)
Constipation
14FWhiteUrbanDiarrheaObese14011.0Negative0DQB1* 02:01pCD
VomitingDQB1* 03:02*(08)
Constipation
7FWhiteUrbanNoNormal445.6Negative0DQB1* 02:02pCD
DQB1* 03:02*(08)
9MAfroRuralNoNormal179.2Negative0DQB1* 02:01pCD
DQB1* 03:02*(08)
6FWhiteUrbanDiarrheaNormal268.1Negative0DQB1* 02:01pCD
FlatulenceDQB1* 03:02*(08)
Abdominal pain
16FAfroUrbanSteatorrheaNormal449.0PositiveIIDQB1* 02:02CD
DistensionDQB1* 03:02*(08)
Abdominal pain
Weightloss
Constipation

Sx=symptoms, Dx=diagnosis, HLA=histocompatibility antigen, F=female, M=male

CD=celiac disease (positive immunochemistry, Marsh II/III and/or present HLA DQ2/DQ8), pCD=potential CD (negative immunohistochemistry, Marsh 0/I and/or present HLA DQ2/DQ8), t=duration

Sx=symptoms, Dx=diagnosis, HLA=histocompatibility antigen, F=female, M=male CD=celiac disease (positive immunochemistry, Marsh II/III and/or present HLA DQ2/DQ8), pCD=potential CD (negative immunohistochemistry, Marsh 0/I and/or present HLA DQ2/DQ8), t=duration The comparison between children with T1DM diagnosed with CD and those with T1DM diagnosed as having potential CD is shown in Table 3. The time of presentation of the last ketoacidotic episode after diagnosis of T1DM and CD was significantly higher in children with potential CD than in children with CD (p <0.0260). There was a higher risk of presenting with CD (n= 4) in children with T1DM who were between 13 and 18 years of age, of the male sex and with digestive symptoms (constipation, vomiting, distension, steatorrhea, diarrhea, abdominal pain and flatulence), but these findings were not statistically significant (p >0.05). The risk of presenting with potential CD (n= 7) was highest for children with T1DM between the ages of 2 and 5 years old who were malnourished, had altered glycosylated hemoglobin and glycemia levels, and had digestive symptoms such as diarrhea, vomiting, flatulence and abdominal pain, but these findings were not statistically significant (p >0.05).
Table 3

Characteristics of children with T1DM, celiac disease and potential celiac disease. (n=11)

VariablestTG2
CD (n=4)Potential CD (n=7)
Sociodemographic
Age (years) X (range)13 (10-16)8 (5-14)
Sex (female:male)2:25:2
Origin (urban:rural)3:15:2
Race (white:other)0:46:1
Antecedents    
Family members with metabolic diseases (n, %)4 (100.0)5 (71.4)
Evolution of the disease    
Duration (months) X (range)30.0 (0-44)53.6 (17-140)
No. ketoacidotic comas X (range)1 (0-2)0,6 (0-3)
Last ketoacidotic coma (months) X (range)21.5 (0-43)69.5 (42-97)
Nutritional status according to WHO
Normal:Malnutrition4:04:3
Normal:Altered height4:07:0
Paraclinics    
Hemoglobin (gr/dL) X (range)14.3 (12.9-15.8)13.3 (12.1-14.5)
Glycosylated hemoglobin (%) X (range)9.4 (8.2-11.0)9.1 (5.6-12.0)
Glycemia (gr/dL) X (range)145.5 (57-246)228.8 (64.0-560.0)

X=mean; n=number

X=mean; n=number

Discussion

Prevalence and seroprevalence

Based on the histopathological findings that 7.1% of children with T1Dm have CD or potential CD, the prevalence in Colombia is higher than that in other Latin American countries such as Brazil (2.6% -4.8%) - and Venezuela (1.7%) . It is also higher than that in Asia , Europe and North America ; but very similar to that of Oceania , and lower than that of Africa (3.0-11.0%) , . The seroprevalence for CD was determined in this study by testing for tTG2, which has a reported sensitivity/specificity of 97.0% , . This finding allowed the identification of CD in 8.4% of participants; this rate is lower than that reported in Asian, European and other South American countries (11.3%) , , . The differing results of these prevalences and seroprevalence rates may be due, among others factors, to genetic and regional characteristics; but primarily to the different antibodies used for CD screening; Therefore, in order to unify and standardize the study of CD, it is suggested that researchers rely on the current algorithms proposed by the ESPGHAN guidelines for the diagnosis of asymptomatic children with a high risk of CD. These guidelines recommend the use of anti-transglutaminase antibodies (tTG2) and/or anti-endomysium (EMA) .

Possible associations

Among children with T1DM, there are various factors that increase their risk of presenting with CD. Those risk factors include sex , age - , and the presence of thyroid disease , , ; and of greater symptoms , decreased weight and height - , anemia , , hypoalbuminemia , rickets , and hypophosphatemia . Similar to other authors , , we found significant differences related to the time of T1DM, specifically in the time (months) of the presentation of the last ketoacidotic coma between children with potential CD and without CD (69.5 ±38.9 months versus 19.1 ±30.7 months, p= 0.0260). This variety of risk factors may depend on sample size, genetics, geographic area, and the specific antibodies used for CD screening, and they should be the focus of future studies, looking for other possible risk factors such as the environmental factors related to breastfeeding, the amount of gluten ingested and the age of introduction in complementary feeding which recently have begun to show controversial results , .

Genotype

The most frequent allele of the 11 children with T1DM and CD in our study was DQ2/DQ8. This result is different from those reported in Africa , Europe , and Oceania , where DQ2 predominates. It is also different from Asia and the United States , where DQ8 is the predominant one. The consulted Latin American studies do not report the alleles - . Given that 40.0% of the general population carries HLA-DQ2 or DQ8 and that the risk of presenting in the following 10 years with CD or autoimmunity for CD is increased in these patients , periodic monitoring is necessary in children with T1DM. Monitoring should be done with serological markers, corroboration with endoscopy and by HLA measurement. This concept makes sense, especially given that HLA DQ2 and DQ8 haplotypes are found in almost all children with CD and are essential for the recognition of gliadin epitopes by antigen-presenting cells. Additionally, if a child is negative for both types of HLA DQ, it is very unlikely that he will have CD, since the negative predictive value is more than 99% . Children with a diagnosis of CD, including potential CD, were referred to a pediatric nutritionist, who initiated the nutritional recommendations of a gluten-free diet. Additionally, their parents and siblings (the first-degree relatives) were screened for CD; and together with the children with negative screening results, they will be monitored every six months/annually for CD. The strengths of the study include that all participants belong to the same cohort of children and were seen by the same healthcare professionals (endocrinologist, gastroenterologist and pathologist) for several years of follow-up (in-hospital cohort). Among the limitations of the study, it is noted that the sample size was limited. Although the population of a tertiary care hospital is described, where a large number of children from southwestern Colombia attend, the results cannot be generalized to all of Colombia, or even to the whole city. In the same way, other possible risk factors such as quality of life, psychological, social, nutritional, and environmental, among others, that could explain the multifactorial model of this entity were not evaluated. Finally, our data were obtained in an intrahospital environment, which allows for some degree of bias.

Conclusion

The prevalence of CD in these children with T1DM is higher than that reported in other South American countries. The prevalence of CD was found to be associated with the time of presentation of T1DM and its main allele, the DQ2/DQ8. These findings are different from what has been described in other places around the world.

Introducción

La enfermedad celíaca (EC) es un trastorno sistémico mediado por el sistema inmune, provocado por el gluten y las prolaminas relacionadas, en individuos genéticamente susceptibles y se caracteriza por la presencia de una combinación variable de manifestaciones clínicas dependientes del gluten, anticuerpos específicos para EC, haplotipos HLA-DQ2 o HLA-DQ8 y enteropatía . La asociación entre la diabetes mellitus tipo 1 (DMT1) y la EC está bien establecida; sin embargo, dentro de los últimos años, varios estudios han demostrado que la prevalencia de EC en pacientes diabéticos es aún mayor de lo considerado anteriormente . Las Guías para el diagnóstico de EC de la Sociedad Europea de Gastroenterología, Hepatología y Nutrición Pediátrica ESPGHAN sugieren realizar pruebas a los niños asintomáticos con riesgo genético para desarrollar EC, como Diabetes mellitus tipo 1 (DMT1) . La prevalencia más alta reportada a nivel mundial de EC es en El Sahara, África (5.6%); seguida en Oceanía (1.2%); Europa (1.0%); Estados Unidos (0.8%); Asia (0.3%-0.7%), y Brasil y Argentina (0.14%-0.60%) ; siendo esta prevalencia mayor en pacientes con DMT1 (2.4%-16.4%) . En una revisión sistemática y de meta-regresión realizada en Colombia, se concluye que la EC parece ser una condición rara en los colombianos . A pesar que es bien conocido a nivel mundial, que hay mayor prevalencia de EC en niños con DMT1 en comparación con la población general , y que la mayoría de estos niños son asintomáticos al momento del diagnóstico ,; aún nos falta en Colombia, entre otras, resolver varias preguntas de investigación, tales como si este riesgo es mayor en los menores de 5 años como lo han demostrado los europeos , o con predominio del género femenino como en la población general adulta o sobre los datos controversiales de la coexistencia de EC y DMT1 y el control glucémico - e incluso reportar nuestra prevalencia, características clínicas y paraclínicas de niños con EC y DMT1 coexistentes; por lo que se analizaron las bases de datos del Servicio de Endocrinología Pediátrica del Hospital Universitario del Valle “Evaristo García” de Cali, Colombia de los niños con diagnóstico de DMT1 con el objetivo de determinar la prevalencia de marcadores de EC en un grupo de niños con DMT1, analizando las manifestaciones clínicas, inmunológicas y genéticas.

Materiales y Métodos

Se diagnosticó DMT1 por destrucción de las células beta en todos los niños mayores de 6 meses de edad sin importar si debuta o no en cetoacidosis, haya o no presencia de otras enfermedades autoinmunes; y en mayores de 10 años si están obesos, de igual manera, hasta que se demuestre ausencia de autoinmunidad . Se diagnosticó EC cuando la IgA anti-transglutaminasa (tTG2) fue positiva; el haplotipo HLA DQ2 y HLA DQ8 compatible; la histotología del duodeno demostró anormalidad de la vellosidad intestinal con un Marsh grado II o mas y se presentó una combinación variable de manifestaciones clínicas dependientes del gluten; y EC potencial cuando la tTG2 fue positiva, el HLA compatible y sin anomalías histológicas a la biopsia duodenal, con o sin signos y síntomas . Se realizó un estudio observacional descriptivo de prevalencia en niños con diagnóstico de DMT1 que consultaron entre el 2 de agosto de 2013 y el 23 de febrero de 2017 a la Consulta de Endocrinología Pediátrica del Hospital Universitario del Valle “Evaristo García” de Cali, Colombia, institución de tercer nivel de atención del suroccidente del país. Los criterios de inclusión fueron niños con diagnóstico de DMT1, de ambos sexos, mayores de 6 meses de edad que en su alimentación complementaria ya habían consumido gluten; y los criterios de exclusión fueron niños con diagnóstico de diabetes mellitus congénita, en coma cetoacidótico, con cromosomopatías asociadas como Síndrome de Down, con otras enfermedades autoinmunes asociadas como hipotiroidismo, con diagnóstico previo de EC, y con enfermedad inflamatoria intestinal. Se tuvieron en cuenta variables sociodemográficas como edad, sexo, raza y orígen; antrompométricas como peso y talla; de la enfermedad como duración de la DMT1, número de comas cetoacidóticos y fecha de último coma cetoacidótico; paraclínicas como hemoglobina glicosilada, hemoglobina y glicemia, y familiares metabólicas como hipertensión, diabetes, sobrepeso y obesidad. Las variables clínicas digestivas que se tuvieron en cuenta fueron el estreñimiento, el vómito, la distensión abdominal, la esteatorrea, la diarrea, el dolor abdominal, las flatulencias y la pérdida de peso. Se realizó tamizaje mediante una Ig A anti-transglutaminasa (tTG2) BiocardTM Celiac Test Ani Biotech, Vantaa, Finland (97.4% sensibilidad, 96.9% especificidad) ,. A los niños tTG2 positivos, se les realizó toma de mínimo 4 biopsias del bulbo y segunda porción duodenal. La evaluación del material de biopsia fue realizado por el mismo patólogo luego de la tinción de hematoxilina-eosina. La inmunohistoquímica del antígeno leucocito comun fue usada para la evaluación del conteo de leucocitos intraepiteliales, y la presencia de mas de 30 linfocitos intraepiteliales versus 100 células epiteliales detectados por antígeno leucocito comun fue considerado como linfocitosis intraepitelial ,. Según Marsh-Oberhuber la histopatología fue clasificada como Marsh I cuando se demostró “lesión infiltrativa”, “lesión infiltrativa-hiperplásica” como Marsh II, y “atrofia vellosa” como Marsh III (parcial IIIa, subtotal IIIb y total IIIc) . A estos niños se les realizó genotipificación del HLA DQ2 y HLA DQ8, por reacción en cadena de polimerasa . La aceptación de participar en el estudio se logra mediante la firma del consentimiento informado de los padres o tutores y el asentimiento informado de los niños mayores de 7 años, siendo este trabajo aprobado por el Comité de Ética de la Universidad del Valle y del Hospital Universitario del Valle “Evaristo García” de Cali, Colombia. El tamaño de muestra fue por conveniencia, participando todos los niños con diagnóstico de DMT1 que cumplieron con los criterios de inclusión del estudio, aceptaron participar en el estudio y asistieron a la Consulta de Endocrinología Pediátrica del Hospital Universitario del Valle “Evaristo García” de Cali, Colombia. Se estimó de la proporción de niños con DMT1 y EC y su correspondiente intervalo de confianza al 95%; en el total de niños participantes se estimaron porcentajes, percentiles, promedios, medianas y las demás medidas descriptivas con sus correspondientes desviaciones estándar y rangos, y para evaluar las posibles asociaciones, se realizó inicialmente análisis univariado entre cada una de las variables de exposición de interés y la variable efecto. Además, se exploró la posible ocurrencia asociación entre las variables de exposición de mayor interés y otras covariables, y entre la variable resultado de interés (EC) y las demás covariables, con el objetivo de evaluar la posible existencia de confusión. Para ello, se construyeron gráficos y tablas de 2x2 y se estimaron los ORs con sus respectivos intervalos de confianza (95%). Para valorar la significancia estadística se utilizó la prueba exacta de Fisher y un valor de p <0.05, a dos colas, se consideró como estadísticamente significativo.

Resultados

Fueron incluidos 155 niños, 83 del sexo masculino, con edad promedio de 11 años (2-18), con diagnóstico de DMT1, quienes tuvieron una duración promedio de la enfermedad de 46.1 meses (0-180 meses), con promedio de un episodio de coma cetoacidótico (0-7 episodios) y con presencia del último episodio cetoacidótico promedio de 20.5 meses (0-163 meses). La mayoría (73.6%; 114/155) presentaron algún antecedente familiar con problemas metabólicos. Los paraclínicos fueron: hemoglobina glicosilada de 9.1% (5.2-19.0%), glicemia de 193.1 g/dL (36.1-600.0 g/dL) y hemoglobina de 12.9 g/dL (9.1-15.0 g/dL). La mayoría de ellos provenían de ciudad urbana (n=109, 70.3%). La tTG2 fue positiva en el 8.4% (13/155). Los parámetros antropométricos se muestran en la Tabla 1.
Tabla 1

Características de los niños con Diabetes mellitus tipo 1 e Ig A anti-transglutaminasa(n= 155)

VariablestTG2
Positiva (n= 13 )Negativa (n= 142)
Sociodemográficas 8.4% (IC 95% 6.2-10.6) 91.6% (IC 95% 89.3-93.8)
Edad (años) X (rango)10 (5-16)11 (2-18)
Sexo femenino:masculino7:676:66
Origen urbano:rural10:399:43
Raza blanca:otra6:773:69
Antecedentes  
Familiares metabólicos presentes n, %10 (76.9)104 (73.2)
Evolución de la enfermedad  
Duración (meses) X (rango)38.1 (0-140)46.8 (0-180)
No. comas cetoacidóticos X (rango)1 (0-3)1 (0-7)
Último coma cetoacidótico (meses) X (rango)36.4 (0-97)19.4 (0-163)
Estado nutricional según OMS
Normal:Malnutrición8:5102:40
Normal:Talla alterada13:0126:16
Paraclínicos  
Hemoglobina (gr/dl) X (rango)13.2 (11.0-15.8)12.9 (9.1-15.8)
Hemoglobina glicosilada (%) X (rango)9.3 (5.6-12.0)9.1 (5.2-9.0)
Glicemia (gr/dl) X (rango)179.2 (57-560)194.4 (36.1-600.0)

X=promedio; n=número

X=promedio; n=número De los 155 niños incluidos, 13 tuvieron tTG2 positiva (8.4%), y a once de ellos, se les realizó endoscopia de vías digestivas altas con toma de biopsias intestinales y haplotipos HLA DQ2 y HLA DQ8 (Figura 1). Los 4 niños con EC por inmunohistoquímica positiva, presentaron en los anticuerpos CD3, CD8 y CD45 comparados con testigos adecuados, un número y localización concordante con EC. En la Tabla 2, se describen las características generales de los 11 niños con DMT1, EC potencial (n= 7) y EC (n= 4).
Figura 1

Flujograma para el estudio de niños con DMT1 con EC

Tabla 2

Características generales en niños con Diabetes mellitus tipo 1 y Enfermedad celíaca (n=11)

Edad (años) SexoRazaOrigenSíntomas Digestivos Diagnóstico Nutricional t DMT1 (meses) HbA1cInmuno Histoquímica MarshHLADx
16FMestizaUrbanoDiarreaEutrófico418.2PositivaIIDQB1* 02:01EC
FlatosDQB1* 03:02*(08)
Distensión
Vómito
11MAfroRuralNoEutrófico3511.0PositivaIIDQB1* 02:02EC
DQB1* 03:02*(08)
9FBlancaUrbanoNoSobrepeso6112.0Negativa0DQB1* 02:01Ecp
DQB1* 03:02*(08)
9MBlancaUrbanoVómitoSobrepeso609.7Negativa0DQB1* 02:01ECp
DQB1* 02:02
5FBlancaRuralDolor abdominalEutrófico188.3Negativa0DQB1* 02:01ECp
10MMestizaUrbanoDolor abdominalEutrófico<19.2PositivaIIDQB1* 02:01EC
VómitoDQB1* 03:02*(08)
Estreñimiento
14FBlancaUrbanoDiarreaObeso14011.0Negativa0DQB1* 02:01ECp
VómitoDQB1* 03:02*(08)
Estreñimiento
7FBlancaUrbanoNoEutrófica445.6Negativa0DQB1* 02:02ECp
DQB1* 03:02*(08)
9MAfroRuralNoEutrófico179.2Negativa0DQB1* 02:01ECp
DQB1* 03:02*(08)
6FBlancaUrbanoDiarreaEutrófico268.1Negativa0DQB1* 02:01ECp
FlatosDQB1* 03:02*(08)
Dolor abdominal
16FAfroUrbanoEsteatorreaEutrófico449.0PositivaIIDQB1* 02:02EC
DistensiónDQB1* 03:02*(08)
Dolor abdominal
Pérdida de peso
Estreñimiento

Sx=síntomas, Dx=diagnóstico, HLA=antígeno de histocompatibilidad, F=femenino, M=masculino

EC=enfermedad celíaca (inmunohistoquímica positiva, Marsh II/III y/o HLA DQ2/DQ8 presente), Ecp=EC potencial (inmunohistoquímica negativa, Marsh 0/I y/o HLA DQ2/DQ8 presente), t=duración

Sx=síntomas, Dx=diagnóstico, HLA=antígeno de histocompatibilidad, F=femenino, M=masculino EC=enfermedad celíaca (inmunohistoquímica positiva, Marsh II/III y/o HLA DQ2/DQ8 presente), Ecp=EC potencial (inmunohistoquímica negativa, Marsh 0/I y/o HLA DQ2/DQ8 presente), t=duración La comparación entre los niños con DMT1 diagnosticados como EC y con DMT1 diagnosticados como EC potencial se muestran en la Tabla 3. El tiempo de presentación del último episodio cetoacidótico luego del diagnóstico de DMT1 y EC, fue significativamente mayor en los niños con EC potencial que en los niños con EC (p < 0.0260). Tuvieron mayor riesgo de presentar EC (n= 4) los niños con DMT1 entre los 13 y 18 años edad, del género masculino y con síntomas digestivos (estreñimiento, vomito, distensión, esteatorrea, diarrea, dolor abdominal y flatulencias), pero sin significancia estadística (p >0.05); y de presentar EC potencial (n= 7) los niños con DMT1 entre los 2 y 5 años, malnutridos, con hemoglobina glicosilada y glicemia alteradas, y con síntomas digestivos como diarrea, vomito, flatulencias y dolor abdominal, pero sin diferencias estadísticas (p >0.05).
Tabla 3

Características de los niños con DMT1, Enfermedad celíaca y Enfermedad celíaca potencial. (n= 11)

VariablestTG2
EC (n= 4) EC potencial (n= 7)
sociodemográficas
Edad (años) X (rango)13 (10-16)8 (5-14)
Sexo femenino:masculino2:25:2
Origen urbano:rural3:15:2
Raza blanca:otra0:46:1
Antecedentes  
Familiares metabólicos presentes n, %4 (100.0)5 (71.4)
Evolución de la enfermedad  
Duración (meses) X (rango)30.0 (0.0-44.0)53.6 (17.0-140.)
No. comas cetoacidóticos X (rango)1.0 (0.0-2.0)0.6 (0.0-3.0)
Último coma cetoacidótico (meses) X (rango)21.5 (0.0-43.0)69.5 (42.0-97.0)
Estado nutricional según OMS
Normal:Malnutrición4:04:3
Normal:Talla alterada4:07:0
Paraclínicos  
Hemoglobina (g/dL) X (rango)14.3 (12.9-15.8)13.3 (12.1-14.5)
Hemoglobina glicosilada (%) X (rango)9.4 (8.2-11.0)9.1 (5.6-12.0)
Glicemia (g/dL) X (rango)145.5 (57-246)228.8 (64.0-560.0)

X=promedio; n=número

X=promedio; n=número

Discusión

Prevalencia y seroprevalencia

Nuestra prevalencia basada en los hallazgos histopatológicos del 7,1% de EC, incluida la EC potencial, en niños con DMT1, es superior a la de otros países latinoamericanos como Brasil (2.6%-4.8%) - y Venezuela (1.7%) ; así como a la de Asia , Europa y Norteamérica ; pero muy similar a la de Oceanía (279, e inferior a la de África (3.0-11.0%) ,. Con relación a la seroprevalencia para EC en este estudio, el tamizaje se llevó a cabo con la tTG2, cuya sensibilidad/especificidad reportada del 97.0% ,; permitió identificarla en un 8.4%; inferior a lo reportado en otros países asiáticos, europeos y suramericanos, que es del 11.3% ,,. Los resultados variables y diversos de estas prevalencias y seroprevalencias, se pueden deber, entre otras, a características genéticas y regionales; y principalmente a los diferentes anticuerpos utilizados para el tamizaje de EC; por lo que para unificar y estandarizar conceptos, se sugiere basarse en los actuales algoritmos propuestos por las Guías de la ESPGHAN para el diagnóstico de niños asintomáticos con riesgo elevado de EC, que recomiendan el uso de anticuerpos anti-transglutaminasa (tTG2) y/o anti-endomisio (EMA) .

Posibles asociaciones

Entre los diversos factores de riesgo en niños con DMT1 para presentar EC están el sexo , la edad -, la presencia de enfermedad tiroidea ,,; y de mayores síntomas , (bajo peso y talla -, anemia ,, hipoalbuminemia ; raquitismo , hipofosfatemia . Al igual que otros autores ,, nosotros encontramos diferencias significativas relacionadas con el tiempo de la DMT1, específicamente en el tiempo (meses) de la presentación del último coma cetoacidótico entre los niños con EC potencial y sin EC (69.5 ±38.9 meses versus 19.1 ±30.7 meses, p= 0.0260). Esta diversidad de factores de riesgo, dependientes del tamaño de muestra, de la genética, del área geográfica, de los anticuerpos utilizados para el tamizaje de EC; deberán ir enfocados a futuro, a la búsqueda de otros posibles factores de riesgo como los ambientales relacionados con lactancia materna, cantidad de gluten ingerido y edad de introducción en la alimentación complementaria; los cuales, al día de hoy comienzan a mostrar resultados controversiales ,.

Genotipo

El alelo más frecuente de los 11 niños con DMT1 y EC de nuestro estudio, fue el DQ2/DQ8, datos diferentes a los reportados en África , Europa , y Oceanía , donde predomina el DQ2 y a los descritos en Asia y Estados Unidos , donde el DQ8 es el predominante. Los estudios Latinoamericanos consultados no informan los alelos -. Sabiendo que el 40.0% de la población general porta HLA-DQ2 o DQ8, y el riesgo a presentar en los siguientes 10 años EC o autoinmunidad para EC , es necesario el control periódico en los niños con DMT1 con marcadores serológicos, corroboración endoscópica y medición de HLA. Este concepto es mas que válido, teniendo en cuenta que los haplotipos HLA DQ2 y DQ8 se encuentran en casi todos los niños con EC, que son esenciales para el reconocimiento de epítopes de gliadina por las células presentadoras de antígenos; y que si un niño es negativo para ambos tipos de HLA DQ, es muy poco probable que presente EC, ya que el valor predictivo negativo es más del 99% . Los niños con diagnóstico de EC, incluida la EC potencial, fueron remitidos a la nutricionista infantil, quien da inicio a las recomendaciones nutricionales sobre dieta libre en glúten, y a sus padres y hermanos (primera línea de consanguinidad) se les inicia el tamizaje para EC; y junto con los niños con tamizaje negativo, se les continua monitorización semestral/anual para EC. Las fortalezas del estudio incluyen que todos los participantes pertenecen a una misma cohorte de niños que son vistos por los mismos profesionales de la salud (endocrinólogo, gastroenterólogo y patólogo) durante varios años de seguimiento (cohorte intrahospitalaria). Entre las limitaciones del estudio; a pesar que se describe la población de un hospital de tercer nivel de atención donde asiste una gran cantidad de niños del suroccidente colombiano, el tamaño de la muestra puede considerarse limitado, y así, los resultados no puedan ser generalizados a toda Colombia, ni siquiera a toda la ciudad. De la misma manera, no se preguntaron otros posibles factores de riesgo como calidad de vida, sicológicos, sociales, nutricionales, y ambientales, entre otros, que puedan explicar el modelo multifactorial de esta entidad. Finalmente, nuestros datos fueron obtenidos en un ambiente intrahospitalario, lo que permite algún grado de sesgo.

Conclusión

La prevalencia de EC en estos niños con DMT1 es superior a la de otros países suramericanos; estando asociada al tiempo de presentación de la DMT1 y su principal alelo el DQ2/DQ8, hallazgos diferentes a lo descrito a nivel mundial.
  42 in total

Review 1.  Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity ('celiac sprue').

Authors:  M N Marsh
Journal:  Gastroenterology       Date:  1992-01       Impact factor: 22.682

2.  High prevalence of coeliac disease in Danish children with type I diabetes mellitus.

Authors:  D Hansen; F N Bennedbaek; L K Hansen; M Høier-Madsen; L S Hegedü; B B Jacobsen; S Husby
Journal:  Acta Paediatr       Date:  2001-11       Impact factor: 2.299

3.  The old and new tests for celiac disease: which is the best test combination to diagnose celiac disease in pediatric patients?

Authors:  Ignazio Brusca; Antonio Carroccio; Elio Tonutti; Danilo Villalta; Renato Tozzoli; Maria Barrale; Filippo M Sarullo; Pasquale Mansueto; Stella Maria La Chiusa; Giuseppe Iacono; Nicola Bizzaro
Journal:  Clin Chem Lab Med       Date:  2011-09-26       Impact factor: 3.694

4.  Prevalence of celiac disease in Brazilian children and adolescents with type 1 diabetes mellitus.

Authors:  Márcia Luiza Baptista; Yu Kar Ling Koda; Renato Mitsunori; Sérgio Ossamu Ioshii
Journal:  J Pediatr Gastroenterol Nutr       Date:  2005-11       Impact factor: 2.839

5.  Clinical benefit of a gluten-free diet in type 1 diabetic children with screening-detected celiac disease: a population-based screening study with 2 years' follow-up.

Authors:  Dorte Hansen; Bendt Brock-Jacobsen; Elisabeth Lund; Christina Bjørn; Lars P Hansen; Christian Nielsen; Claus Fenger; Søren T Lillevang; Steffen Husby
Journal:  Diabetes Care       Date:  2006-11       Impact factor: 19.112

6.  HLA-DQA1*05-DQB1*0201 positivity predisposes to coeliac disease in Czech diabetic children.

Authors:  Z Sumník; S Kolousková; O Cinek; R Kotalová; J Vavrinec; M Snajderová
Journal:  Acta Paediatr       Date:  2000-12       Impact factor: 2.299

7.  Prevalence of celiac disease in Brazilian children with type 1 diabetes mellitus.

Authors:  Mariella Guarino Tanure; Ivani Novato Silva; Magda Bahia; Francisco José Penna
Journal:  J Pediatr Gastroenterol Nutr       Date:  2006-02       Impact factor: 2.839

8.  Screening of coeliac disease in north Italian children with type 1 diabetes: limited usefulness of HLA-DQ typing.

Authors:  G Contreas; E Valletta; D Ulmi; S Cantoni; L Pinelli
Journal:  Acta Paediatr       Date:  2004-05       Impact factor: 2.299

9.  Younger age at onset and sex predict celiac disease in children and adolescents with type 1 diabetes: an Italian multicenter study.

Authors:  Franco Cerutti; Graziella Bruno; Francesco Chiarelli; Renata Lorini; Franco Meschi; Carla Sacchetti
Journal:  Diabetes Care       Date:  2004-06       Impact factor: 19.112

10.  Celiac disease and autoimmune thyroid disease in children with type 1 diabetes mellitus: clinical and HLA-genotyping results.

Authors:  Ayça Törel Ergür; Gönül Oçal; Merih Berberoğlu; Pelin Adıyaman; Zeynep Sıklar; Zehra Aycan; Olcay Evliyaoğlu; Aydan Kansu; Nurten Girgin; Arzu Ensari
Journal:  J Clin Res Pediatr Endocrinol       Date:  2010-11-03
View more

北京卡尤迪生物科技股份有限公司 © 2022-2023.