Coeliac disease in children with type 1 diabetes
Kurppa K, Laitinen A & Agardh D.
Lancet Child Adolesc Health. Published online December 13, 2017
This review paper summarises approaches to diagnosis and treatment of children with type 1 diabetes (T1DM) and coeliac disease (CD) with a particular emphasis on those with a double diagnosis. Over the last few decades, the incidences of both CD and T1DM have rapidly increased as demonstrated by data from Finland which saw a five-fold increase in T1DM between 1980 and 2011 with a parallel change observed for CD.
Lancet Child Adolesc Health. Published online December 13, 2017
This review paper summarises approaches to diagnosis and treatment of children with type 1 diabetes (T1DM) and coeliac disease (CD) with a particular emphasis on those with a double diagnosis. Over the last few decades, the incidences of both CD and T1DM have rapidly increased as demonstrated by data from Finland which saw a five-fold increase in T1DM between 1980 and 2011 with a parallel change observed for CD.
Both conditions are also over-represented in relatives of affected individuals and also among those with other autoimmune disease. In addition, the prevalence of CD in patients with T1DM is higher than for the general population and conversely patients with CD have an increased prevalence of diabetes. Children affected by either or both of these diseases are also at risk of other autoimmune conditions.
Rapid changes in disease incidence cannot be explained by genetics and point to environmental changes which also play a role including presence of viruses, timing of gluten introduction to the diet along with the amount of gluten in the diet. Other possible contributing factors include imbalanced intestinal microbiota, caesarean section and early infant feeding practices such as breastfeeding.
Challenges in histological diagnosis and the high specificity of current serological markers led the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) to publish new diagnostic criteria for children in 2012. It recommended a diagnosis of CD without biopsy can be made in those who were symptomatic, with high concentrations of tTG (10 x the upper limit of normal) confirmed by positive endomysial autoantibodies (EMA) and coeliac-type HLA genotypes. It is important to note a biopsy is still required in some cases. The reliability of the ESPGHAN criteria has been reinforced but are not yet accepted in all countries.
Many patients with T1DM have low or fluctuating levels of tTG autoantibodies but do not always have apparent mucosal lesions required for a histological diagnosis. Currently, there is insufficient evidence to make recommendations on diagnosis and treatment for this group of patients. Choice of treatment of CD in children with T1DM should be made on an individual basis and after careful discussion with the child and parents.
A GFD can be burdensome to children with T1DM who have to carefully monitor their carbohydrate intake with some studies reporting dietary problems or increased insulin requirements following the diagnosis of CD. However, a GFD is usually well adopted by patients and strict adherence might improve glycaemic control and HbA1c concentration whereas poor adherence predisposes patients to hypoglycaemia and reduces quality of life. Both conditions require regular monitoring during childhood and adolescence and a well-planned transition from paediatric to adult care is also important.
Screening for CD in at-risk groups has led to an increasing number of seropositive children presenting with milder intestinal lesions that do not fulfil the current diagnostic criteria for CD. Increasing evidence shows that these patients may benefit from dietary treatment. However, an unsolved issue is that some children with T1DM have low or fluctuating levels of tTG and normal histology (potential CD) may later become seronegative or remain having normal mucosal villi even after several years on a gluten-containing diet. Ongoing studies will hopefully provide further answers for these scenarios as well as other unanswered research questions including whether the development of CD and T1DM could be prevented or slowed by modifying the infant diet.
Source: https://doi.org/10.1016/S2352-4642(17)30172-4
Pathogenesis
Both CD and T1DM involve an autoimmune process in genetically predisposed individuals. In CD, gliadin peptides penetrate the intestinal epithelium and activate innate and adaptive immunity. This may be preceded by initial damage triggered, for example, by an infection resulting in increased permeability of the epithelial barrier. There is also growing evidence to suggest that viruses (e.g. enteroviruses) and other environmental factors, such as dietary factors, contribute to disease development and activation of the innate immune system in T1DM.Shared genetics and environmental factors
Evidence of gene involvement in the aetiology of T1DM and CD is the high concordance in monozygotic twins of developing T1DM (approximately 42%) and CD (approximately 75-90%). The increased prevalence of CD in patients with T1DM can predominately be explained by the shared genes that have been shown to be associated with both diseases. The strongest genetic risk is attributed to HLA genes on chromosome 6 but a large genome-wide association study found 26 non-HLA loci were also associated with both conditions.Rapid changes in disease incidence cannot be explained by genetics and point to environmental changes which also play a role including presence of viruses, timing of gluten introduction to the diet along with the amount of gluten in the diet. Other possible contributing factors include imbalanced intestinal microbiota, caesarean section and early infant feeding practices such as breastfeeding.
Relevance of autoantibodies
The main difference between the autoantibodies in CD & T1DM is their specificity. Whilst tTG autoantibodies have a near 100% positive predictive value (PPV) for untreated CD, the five-year PPV of a single positive autoantibody for T1DM is only 10-25%. In those positive for multiple autoantibodies (i.e. ≥3), the PPV can be upto 75% if a family member also has the condition. The pattern of islet and coeliac autoantibody appearance depends on the combination of HLA risk genes. The prospective TEDDY cohort study has shown that islet cell antibodies appear before tTG antibodies in approximately 2/3 of children. Both conditions have a long subclinical phase in which autoantibodies are present prior to clinical manifestations. In T1DM, upto 70-90% of β cells can be destroyed before the occurrence of symptoms. Similarly, CD patients can have decades of subclinical features before intestinal damage occurs, however, contrary to T1DM, the association between degree of small bowel damage and antibody levels is more obvious. In addition, symptoms, or even complications, can appear in seropositive individuals even if only mild histological injury is present.Presentation and Diagnosis
Upto 1/3 of patients with T1DM have unrecognised CD symptoms for more than 5 years before a diagnosis is made. Some clinical recommendations in relation to screening for CD in children with T1DM exist which has led to earlier identification of CD and a milder presentation. However, no difference has been found in mucosal lesions between patients with T1DM diagnosed via screening and those diagnosed following onset of symptoms. Even asymptomatic patients can have advanced mucosal damage.Challenges in histological diagnosis and the high specificity of current serological markers led the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) to publish new diagnostic criteria for children in 2012. It recommended a diagnosis of CD without biopsy can be made in those who were symptomatic, with high concentrations of tTG (10 x the upper limit of normal) confirmed by positive endomysial autoantibodies (EMA) and coeliac-type HLA genotypes. It is important to note a biopsy is still required in some cases. The reliability of the ESPGHAN criteria has been reinforced but are not yet accepted in all countries.
Many patients with T1DM have low or fluctuating levels of tTG autoantibodies but do not always have apparent mucosal lesions required for a histological diagnosis. Currently, there is insufficient evidence to make recommendations on diagnosis and treatment for this group of patients. Choice of treatment of CD in children with T1DM should be made on an individual basis and after careful discussion with the child and parents.
Screening for CD in patients with T1DM
Diagnostic guidelines differ in their approach to screening for CD in this patient group. The high coexistence and frequently missed cases seems to support the case for screening, however, more data is warranted, particularly on the long-term prognosis of children with coexisting T1DM and asymptomatic CD. Ongoing trials are aiming to address this issue. A further important question is whether clinical T1DM could be prevented if CD is diagnosed earlier, prior to symptoms of T1DM, by screening at-risk groups. Simplifying screening by limiting to the at-risk HLA genotypes could make this approach more cost-effective but is challenged by some.Long-term complications and follow up
In both conditions, inadequate treatment and poor compliance are associated with poorer long-term outcomes. In addition, coexistence of the two conditions further increases the risk of complications with long-term CD (≥15 years post diagnosis) having been associated with a 2.8 times increased mortality risk in patients with T1DM. Furthermore, CD might be an independent risk factor for earlier development of certain diabetic complications such as retinopathy and nephropathy and, adults with a double diagnosis, .are more prone to cardiovascular complications. It remains unclear whether a gluten-free diet (GFD) could reduce the risk of these complications but poorly treated CD has been shown to predispose patients with coexisting T1DM to low bone mineral density.A GFD can be burdensome to children with T1DM who have to carefully monitor their carbohydrate intake with some studies reporting dietary problems or increased insulin requirements following the diagnosis of CD. However, a GFD is usually well adopted by patients and strict adherence might improve glycaemic control and HbA1c concentration whereas poor adherence predisposes patients to hypoglycaemia and reduces quality of life. Both conditions require regular monitoring during childhood and adolescence and a well-planned transition from paediatric to adult care is also important.
New challenges of the GFD & Future Perspectives
Whilst an increase in availability of gluten-free products is beneficial to patients in need of gluten-free options it is important for clinicians to be aware of a patient’s intake of gluten prior to commencing screening for the condition. Furthermore, prospective studies of optimal nutrition in children with a double diagnosis are warranted.Screening for CD in at-risk groups has led to an increasing number of seropositive children presenting with milder intestinal lesions that do not fulfil the current diagnostic criteria for CD. Increasing evidence shows that these patients may benefit from dietary treatment. However, an unsolved issue is that some children with T1DM have low or fluctuating levels of tTG and normal histology (potential CD) may later become seronegative or remain having normal mucosal villi even after several years on a gluten-containing diet. Ongoing studies will hopefully provide further answers for these scenarios as well as other unanswered research questions including whether the development of CD and T1DM could be prevented or slowed by modifying the infant diet.
Source: https://doi.org/10.1016/S2352-4642(17)30172-4
www.drschaer-institute.com