Dysregulation of glucose metabolism in preclinical type 1 diabetes

Transcription

1 Pediatric Diabetes 2016: 17(Suppl. 22): doi: /pedi All rights reserved Pediatric Diabetes Review Article Dysregulation of glucose metabolism in preclinical type 1 diabetes Veijola R, Koskinen M, Helminen O, Hekkala A. Dysregulation of glucose metabolism in preclinical type 1 diabetes. Pediatric Diabetes 2016: 17 (Suppl. 22): Long-term prospective studies have provided valuable information about preclinical type 1 diabetes (T1D). Children who have seroconverted to positive for islet autoantibodies have also, in follow-up, had metabolic tests to understand the timing and development of abnormal glucose tolerance and declining insulin secretion before the clinical diagnosis of T1D. First phase insulin response (FPIR) in the intravenous glucose tolerance test (IVGTT) is lower in the progressors positive for multiple islet autoantibodies in all age groups and as early as 4 6 years before the diagnosis when compared with the non-progressors positive for only islet cell antibodies (ICA). An accelerated decline in FPIR is seen in the progressors during the last 1.5 years before the diagnosis. These results indicate that the progressors may have an early intrinsic defect in beta cell development or function. In the oral glucose tolerance test (OGTT) the peak C-peptide response is delayed in the progressors at least 2 years before diagnosis. Glucose levels and HbA1c are increasing about 2 years before clinical diagnosis. An increase in HbA1c and detection of abnormal glucose tolerance in OGTT are useful in the prediction of the timing of clinical onset of T1D. Continuous glucose monitoring (CGM) may be useful in the prediction of T1D as an early indicator of increased glycemic variability but more data from larger series are needed for confirmation. Children followed in the prospective studies are diagnosed earlier and have a decreased frequency of ketoacidosis at the diagnosis of T1D when compared with age-matched cases from the population. Riitta Veijola a,b, Maarit Koskinen c,d, Olli Helminen a,b and Anne Hekkala a,b a Department of Pediatrics, Research Unit for Pediatrics, Dermatology, Clinical Genetics, Gynecology and Obstetrics (PEDEGO), Medical Research Center (MRC) Oulu, University of Oulu, Oulu, Finland; b Department of Children and Adolescents, Oulu University Hospital, Oulu, Finland; c Department of Pediatrics, University of Turku, Turku, Finland; and d Department of Pediatrics, Turku University Hospital, Turku, Finland Key words: glucose metabolism HbA1c insulin secretion prediction T1D Corresponding author: Riitta Veijola, MD, PhD, Faculty of Medicine, Department of Pediatrics, University of Oulu, P.O. Box 5000, Oulu 90014, Finland. Tel: (358) ; fax: (358) ; riitta.veijola@oulu.fi Submitted 28 February Accepted for publication 23 March 2016 The prevalence of type 1 diabetes (T1D) approaches 1% among adolescents aged 15 years in the highincidence countries in northern Europe. In Finland, the incidence of T1D is the highest in the world reaching 64.9/ /yr in 2006 in children <15 years, which is almost five times higher than the incidence in the 1950s in this country (1). It is now more frequently diagnosed in young children aged 0 4 years (2). This trend may increase the life-time risk of vascular complications, which are known to be associated with long disease duration and insufficient metabolic control. The reasons for these unfavorable epidemiologic trends for pediatric T1D are unknown. Considerable research efforts have taken place over the latest three decades to understand the pathogenesis of T1D, improve prediction, and find preventive treat The Authors. Pediatric Diabetes published by John Wiley & Sons Ltd. 25 This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

2 Veijola et al. ment. Prospective studies focusing on individuals who are at increased risk for T1D, either first degree relatives (FDRs) of patients with the disease or children from the general population who carry the risk-conferring class II human leukocyte antigen (HLA) genotype, have provided essential new information about the development of T1D (3). Initiation and persistence of autoimmunity against the pancreatic islets is characteristic for the preclinical stage of T1D. One of the most important findings of prospective studies with regular follow-up from birth has been the observation that seroconversion for islet autoantibodies peaks very early, during the first 2 years of life. (4, 5). Thereafter, the incidence of new seroconversions decreases and remains stable at least until the start of puberty. Subjects with persistent multiple islet autoantibodies have a very high risk for progression to clinical T1D. However, the progression rate seems to be relatively steady after the appearance of multiple biochemical islet autoantibodies. In a large international series combining data from long-term prospective follow-up studies in Finland, USA, and Germany 43.5% of children had developed clinical T1D within 5 years, 69.7% within 10 years and 84.2% within 15 years from seroconversion for multiple biochemical islet autoantibodies (6). Young age (<3 years) at seroconversion, high-risk HLA genotype and female gender were associated with faster progression in that analysis. Recently, the prediction of the timing of clinical diagnosis has become more accurate by analyzing the development of various markers of glucose metabolism during the preclinical stage of T1D (7, 8). This review summarizes the data on beta cell function and the development of abnormal glucose tolerance (dysglycemia) before and until the clinical diagnosis of T1D. We are reviewing here data from several prospective studies including The Diabetes Prevention Trial Type 1 (DPT-1), The TrialNet Natural History Study (TNNHS), The Diabetes Autoimmunity Study in the Young (DAISY) Study, The Diabetes Prediction in Skåne (DiPiS) Study, and The Environmental Determinants of Diabetes in the Young (TEDDY) Study. In particular, we discuss the recently published glucose metabolism data from The Finnish Type 1 Diabetes Prediction and Prevention (DIPP) Study. The DIPP Study was launched in November 1994 in Finland and has screened more than newborns for HLA-conferred genetic susceptibility for T1D. Over children have started clinical follow-up with regular study visits and systematic collection of longitudinal biological samples. More than 800 children have seroconverted to positivity for multiple ( 2) biochemical islet autoantibodies (IAA, GADA, IA-2A and ZnT8A), and more than 400 children have progressed to clinical T1D with DIPP follow-up from birth until diagnosis. Currently there are about 7200 children between 3 months and 15 years of age participating in the DIPP Study follow-up. In addition, many subjects older than 15 years and testing positive for multiple autoantibodies remain in the follow-up for their glucose metabolism and may participate in secondary prevention studies. According to the DIPP protocol HbA1c and random plasma glucose have been measured every 3 months after the child seroconverted positive for islet autoantibodies. In addition, IVGTT and OGTT have been performed once a year for subjects with multiple islet autoantibodies to follow endogenous insulin secretion and the glycemic status. In a subcohort of the DIPP children IVGTTs have also been performed in subjects who were positive for classical ICA only, and actually have a low risk for developing type 1 diabetes, which was not yet known when the protocol was planned. Preclinical stage of type 1 diabetes in children Positivity for islet autoantibodies (i.e. preclinical T1D) has been a prerequisite for the performance of the metabolic tests to analyze beta cell function and glucose tolerance in the protocols of prospective follow-up studies. It is noteworthy that there are some practical limitations when studying young children, for example the volume of blood that can be drawn. Various metabolic tests have been used in different studies to characterize the changes in glucose metabolism before the clinical diagnosis. A schematic summary of the beta cell insulin secretory capacity and the development of dysglycemia during the preclinical stage of T1D in children is presented in Fig. 1. Intravenous glucose tolerance test (IVGTT) The insulin secretory capacity of the beta cells in young children has been evaluated in several prospective studies by performing an IVGTT and calculating the first phase insulin response (FPIR) as the sum of serum insulin concentrations at 1 and 3 min after the end of glucose infusion. In the DIPP Study reduced FPIR was shown to be predictive of progression to T1D in children with multiple autoantibodies (9). Earlier studies focusing on first-degree relatives of patients with T1D had reported similar results. In 35 FDRs positive for islet autoantibodies FPIR less than the first percentile as compared with 225 normal control subjects was associated with a 50% risk of developing T1D within 1 year (10). In the DPT-1 Study, serial IVGTTs were performed in children positive for islet autoantibodies, and during the progression to T1D an accelerated decline in FPIR was observed, especially during the last 1.5 years before diagnosis as illustrated in Fig. 1 (11). Recent data from the DIPP Study confirmed that there is a rapid decrease in FPIR during 26 Pediatric Diabetes 2016: 17 (Suppl. 22): 25 30

3 Glucose metabolism in preclinical type 1 diabetes Fig. 1. Development of dysglycemia during preclinical type 1 diabetes. Schematic illustration of first phase insulin response (FPIR) measured in intravenous glucose tolerance test (IVGTT), insulin resistance index (HOMA-IR), glycosylated hemoglobin A1c (HbA1c), randomly measured plasma glucose, fasting plasma glucose and 2-h plasma glucose values in oral glucose tolerance test (OGTT). Black lines represents progressors in all graphs. Long dashed lines in the two top graphs represent non-progressors with classical islet cell antibodies (ICA) only. Dotted lines in the middle and bottom graphs represent non-progressors who are positive for multiple islet autoantibodies. The dotted lines and the dashed line in HOMA-IR are on the same level as solid lines, but drawn separately here to keep the figures clear. FPIR is constantly higher in non-progressors than in progressors. Adapted from the original data from refs Koskinen et al. (12), Helminen et al. (7, 8), Sosenko et al. (11). the last 2 years preceding the diagnosis (12). The DIPP data also showed that FPIR was reduced as early as 4 6 years before the diagnosis in the progressors when compared with the non-progressors who were healthy children with positivity for classical islet cell antibodies (ICA) only (Fig. 1). In age-dependent longitudinal analysis in the same series FPIR was constantly lower in the progressors than in non-progressors even when FPIR values from the last 2 years before the diagnosis were excluded from the analysis. The interpretation of these observations is that children who progress to T1D may have an early intrinsic defect in their beta cell function and with advancing age and body size they fail to increase the insulin secretory capacity. (12). It is interesting to note that in an early analysis from the DIPP series, reduced FPIR (defined as lower than the 5th percentile in age-matched children without autoantibodies) was observed in as many as 42% of young children who had recently seroconverted positive for islet autoantibodies (13). Hemoglobin A1c (HbA1c) Increasing HbA1c levels have been shown to be useful in the prediction of the timing of T1D diagnosis. In the DAISY Study which has prospectively followed general population newborns carrying HLA genotypes conferring risk for T1D and young siblings or offspring of subjects with T1D, increasing HbA1c levels within the normal limits were observed in 15 autoantibody positive participants who progressed to clinical disease, whereas HbA1c levels remained stable in autoantibody positive non-progressors (14). The TNNHS dataset included 1143 relatives of people with T1D who were positive for at least two islet autoantibodies and were monitored every 6 months by HbA1c and OGTT. A Pediatric Diabetes 2016: 17 (Suppl. 22):

4 Veijola et al. 10% increase in HbA1c was associated with an 84% 3-year risk for T1D and a 20% increase in HbA1c with nearly 100% risk within the next 3 5 years (15). In the DIPP Study 466 children with multiple islet autoantibodies were monitored by HbA1c every 3 12 months, and a 10% increase in HbA1c predicted development of T1D with a hazard ratio (HR) of 5.7, and the median time to diagnosis was 1.1 years from the observed rise. If HbA1c was 41 mmol/mol in two consecutive samples the HR was 11.9 and median time to diagnosis only 0.9 years (7). HbA1c levels started to be consistently higher in the progressors 2 years before the diagnosis when compared with autoantibody positive non-progressors. These data show that HbA1c is a simple and cost effective predictive marker for T1D and is particularly useful for the estimation of the timing of the diagnosis in subjects with multiple islet autoantibodies. Oral glucose tolerance test (OGTT) The development of dysglycemia as well as the dynamics of the C-peptide response in OGTT have been analyzed in the DPT-1 study and the TNNHS series. In these studies dysglycemic OGTT has been defined as fasting glucose values mg/dl ( mmol/l) and/or 30-, 60-, or 90-min glucose values 200 mg/dl ( 11.1 mmol/l) and/or 120-min glucose values mg/dl ( mmol/l). In the DPT-1 study 515 of ICA-positive relatives were followed at every 6 months with OGTTs. Dysglycemia occurred at least once in 60% of the subjects over a period of 7 years. Dysglycemia detected at the 6-month visit from the start of follow-up was highly predictive of T1D, particularly among relatives <13 years of age (16). Interestingly, the dynamics of the C-peptide response in OGTT started to differ in the progressors at least 2 years before clinical diagnosis when compared with the non-progressors. The early 0 30 min C-peptide response was decreased, whereas the late C-peptide response (the sum of C-peptide values after 30-min in OGTT) was increased in the progressors 2 years before the diagnosis (17, 18). The delayed peak C-peptide value at 2 years before the diagnosis occurred even later as the diagnosis approached. The area under the curve (AUC) of C-peptide and the peak level of C-peptide started to decrease only 6 months before the diagnosis, and fasting C-peptide increased during that interval. After the diagnosis the C-peptide levels declined in an accelerated rate. In the DPT-1 series the deteriorating glucose tolerance was associated with lower C-peptide at 30 min and a delay in peak C-peptide in the OGTT (19). In a series of 403 DIPP children with multiple islet autoantibodies impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) in an OGTT as defined by WHO were predictive for T1D with a HR of 3.2 and 8.3, respectively. The median time to diagnosis after IGT was 0.7 years showing that the simple 2-point OGTT may also provide important information when estimating the time to clinical diagnosis of T1D (8). Random plasma glucose measurements The DAISY study followed 92 children with multiple autoantibodies for a mean period of 3.4 years with point-of-care testing for random plasma glucose, and found that random plasma glucose levels only marginally predicted development of T1D with a HR of 1.4 (14). In the DIPP Study random plasma glucose levels were followed at every 3 months in a larger series of 505 autoantibody positive children, 204 of whom progressed to T1D. In subjects with a random plasma glucose value of 7.8 mmol/l ( 140 mg/dl), the HR for progression to clinical diabetes was 6.0 and the median time to diagnosis 1.0 year indicating that this simple and low-cost measure is also useful in the prediction of the timing of diagnosis (8). Continuous glucose monitoring (CGM) In a group of 14 children positive for islet autoantibodies a 4-day CGM showed that these children have increased glycemic variability and spend more time above 140 mg/dl (7.8 mmol/l) when compared with autoantibody negative controls (20). Another study including 22 autoantibody positive FDRs of patients with T1D also found that glycemic variability during a 5-day CGM was increased in these subjects as compared with 20 healthy autoantibody negative controls without a relative with T1D (21). These preliminary observations will be validated in multiple ongoing CGM studies. Hyperglycemic clamp A Belgian group has used the hyperglycemic clamp technique to investigate the variables contributing the deteriorating glucose metabolism during the preclinical stage of type 1 diabetes. A total of 81 persistently autoantibody positive FDRs, aged 5 39 years (mostly adults) were studied by OGTT and hyperglycemic clamp, and the area under the curve (AUC 5 10 min ) of clamp-derived first phase C-peptide release was calculated. For the AUC 5 10 min serum C-peptide was measured at 5, 7.5, and 10 min after the desired plateau 10 mmol/l for plasma glucose was achieved in the clamp. A low C-peptide AUC 5 10 min below the 10th percentile of controls was associated with 50 70% risk of clinical T1D within 3 years, and reported to be a better predictor for T1D than the peak C-peptide in the OGTT (22). The clamp procedure, however, is not 28 Pediatric Diabetes 2016: 17 (Suppl. 22): 25 30

5 feasible in large scale follow-up studies involving young children. Diagnosis of type 1 diabetes in children participating in prospective studies Children who have participated in prospective followup studies and progress to clinical T1D have been diagnosed at an earlier stage of the disease when compared with other children diagnosed with T1D at the same age in the same region (23, 24). The earlier diagnosis results most probably from the awareness of the families and the study personnel about the fact that that positivity for multiple islet autoantibodies is associated with a high risk for progression to clinical diabetes. The families are educated about the classical symptoms of hyperglycemia and when these appear the condition is recognized by the parents without delay. The families have been advised to make an immediate contact with the health care professionals to confirm the diagnosis and start insulin treatment when the symptoms appear. Some families have a glucometer at home and can easily check the plasma glucose value if any suspicion of hyperglycemia arise. Early diagnosis is clearly an advantage for these children and their families. Insulin treatment can often be started with fewer daily injections and lower doses and the metabolic control is easier to achieve than after ketoacidosis. The families have time to concentrate on the education and to adapt to the requirements of daily routines such as insulin injections, glucose monitoring, and carbohydrate counting. In the participants of the prospective studies type 1 diabetes may be diagnosed very early, even before any clinical symptoms appear (24, 25). Many study protocols include the measurement of random plasma glucose at every study visit and the performance of OGTTs at regular intervals for subjects who are persistently positive for islet autoantibodies. For example, in the large international TEDDY Study following more than 8000 risk children from birth the frequency of asymptomatic diabetes was 36% among the first 100 cases diagnosed with type 1 diabetes at the mean age of 2.3 years. The majority of these very young children were diagnosed on the basis of random plasma glucose (24). When the diagnosis is based on a diabetic OGTT in an asymptomatic subject, it is recommended that another confirmatory OGTT will be performed after a minimum interval of 1 week. Children followed regularly before the diagnosis of T1D have marked reduction in the frequency of diabetic ketoacidosis (DKA) at the time of diagnosis (24 28). For example, only 2% of children followed in the Swedish DiPiS Study since birth and diagnosed with T1D at the age of 2 12 years had diabetic ketoacidosis compared with 18% of children diagnosed from the Glucose metabolism in preclinical type 1 diabetes same birth cohort but without prospective follow-up (28). Similarly, in the DIPP study 2% of children who participated in the nasal insulin trial and progressed to clinical diabetes presented with diabetic ketoacidosis (26) whereas the overall ketoacidosis frequency was 19% in children with newly diagnosed T1D in Finland during that time period (29). Several other studies have also recorded very low DKA frequencies of 3 8% in subjects participating in regular preclinical follow-up and progressing to clinical diabetes. Conclusions and future prospects The prospective longitudinal studies with information from the genetic, autoimmune, and metabolic parameters have improved our understanding about the sequence of the events leading to type 1 diabetes. On the basis of this knowledge a classification of the stages of type 1 diabetes starting from seroconversion for multiple ( 2) islet autoantibodies has recently been suggested (30), and refinement of this staging is expected. More detailed information about the preclinical metabolic and immunologic process is needed, e.g. the significance of glycemic variability as a predictive marker, the regulation of early beta cell function and the interplay between the beta cell and the immune system. The current clinical infrastructure of the DIPP Study is exceptional because it includes a large follow-up cohort of children, adolescents, and young adults, and also maintains continuous recruitment of newborn infants who have an increased risk for T1D. This infrastructure offers excellent possibilities to investigate the whole disease process from increased genetic risk until the clinical diagnosis of T1D and even beyond. In addition, subjects at various stages of the disease process can be identified and recruited for clinical trials to prevent the disease. Prevention is the key in our campaign to reduce the prevalence of T1D, and new innovative prevention trials are badly needed. Acknowledgments The DIPP study has been supported by the Juvenile Diabetes Research Foundation International and the Special Research Funds for University Hospitals in Finland. We also thank all children and families participating in the DIPP study and the personnel at all DIPP centers. References 1. Harjutsalo V, Sund R, Knip M, Groop PH. Incidence of type 1 diabetes in Finland. JAMA 2013: 310: Harjutsalo V, Sjöberg L, Tuomilehto J. Time trends in the incidence of type 1 diabetes in Finnish children: a cohort study. Lancet 2008: 371: Atkinson MA, Eisenbarth GS, Michels AW. Type 1 diabetes. Lancet 2014: 383: Pediatric Diabetes 2016: 17 (Suppl. 22):

6 Veijola et al. 4. Parikka V, Näntö-Salonen K, Saarinen M et al. Early seroconversion and rapidly increasing autoantibody concentrations predict prepubertal manifestation of type 1 diabetes in children with genetic risk. Diabetologia 2012: 55: Krischer JP, Lynch KF, Schatz DA et al. The 6 year incidence of diabetes-associated autoantibodies in genetically at-risk children: the TEDDY study. Diabetologia 2015: 58: Ziegler AG, Rewers M, Simell O et al. Seroconversion to multiple islet autoantibodies and risk of progression to diabetes in children. JAMA 2013: 309: Helminen O, Aspholm S, Pokka T et al. HbA1c predicts time to diagnosis of type 1 diabetes in children at risk. Diabetes 2015: 64: Helminen O, Aspholm S, Pokka T et al. OGTT and random plasma glucose in the prediction of type 1 diabetes and time to diagnosis. Diabetologia 2015: 58: Siljander HT, Hermann R, Hekkala A et al. Insulin secretion and sensitivity in the prediction of type 1 diabetes in children with advanced beta cell autoimmunity. Eur J Endocrinol 2013: 169: Vardi P, Crisa L, Jackson RA. Predictive value of intravenous glucose tolerance test insulin secretion less than or greater than the first percentile in islet cell antibody positive relatives of type 1 (insulin-dependent) diabetic patients. Diabetologia 1991: 34: Sosenko JM, Skyler JS, Beam CA et al. Acceleration of the loss of the first-phase insulin response during the progression to type 1 diabetes in diabetes prevention trial-type 1 participants. Diabetes 2013: 62: Koskinen MK, Helminen O, Matomäki J et al. Reduced β-cell function in early preclinical type 1 diabetes. Eur J Endocrinol 2016: 174: Keskinen P, Korhonen S, Kupila A et al. First-phase insulin response in young healthy children at genetic and immunological risk for type I diabetes. Diabetologia 2002: 45: Stene LC, Barriga K, Hoffman M et al. Normal but increasing hemoglobin A1c levels predict progression from islet autoimmunity to overt type 1 diabetes: Diabetes Autoimmunity Study in the Young (DAISY). Pediatr Diabetes 2006: 7: Krischer JP, Type 1 Diabetes Trial Net Study Group. The use of intermediate endpoints in the design of type 1 diabetes prevention trials. Diabetologia 2013: 56: Sosenko JM, Palmer JP, Rafkin-Mervis L et al. Incident dysglycemia and progression to type 1 diabetes among participants in the diabetes prevention trial-type 1. Diabetes Care 2009: 32: Sosenko JM, Palmer JP, Rafkin LE et al. Trends of earlier and later responses of C-peptide to oral glucose challenges with progression to type 1 diabetes in diabetes prevention trial-type 1 participants. Diabetes Care 2010: 33: SosenkoJM, Skyler JS, Herold KC, Palmer JP, Type 1 Diabetes TrialNet and Diabetes Prevention Trial-Type 1 Study Groups. The metabolic progression to type 1 diabetes as indicated by serial oral glucose tolerance testing in the Diabetes Prevention Trial-type 1. Diabetes 2012: 61: Greenbaum CJ, Buckingham B, Chase HP, Krischer J, Diabetes Prevention Trial, Type 1 Diabetes (DPT- 1) Study Group. Metabolic tests to determine risk for type 1 diabetes in clinical trials. Diabetes Metab Res Rev 2011: 27: Steck AK, Dong F, Taki I, Hoffman M, Klingensmith GJ, Rewers MJ. Early hyperglycemia detected by continuous glucose monitoring in children at risk for type 1 diabetes. Diabetes Care 2014: 37: Van Dalem A, Demeester S, Balti EV et al. Relationship between glycaemic variability and hyperglycaemic clamp-derived functional variables in (impending) type 1 diabetes. Diabetologia 2015: 58: Balti EV, Vandemeulebroucke E, Weets I et al. Hyperglycemic clamp and oral glucose tolerance test for 3-year prediction of clinical onset in persistently autoantibody-positive offspring and siblings of type 1 diabetic patients. J Clin Endocrinol Metab 2015: 100: Elding Larsson H, Vehik K, Bell R et al. Reduced prevalence of diabetic ketoacidosis at diagnosis of type 1 diabetes in young children participating in longitudinal follow-up. Diabetes Care 2011: 34: Elding Larsson H, Vehik K, Gesualdo P et al. Children followed in the TEDDY study are diagnosed with type 1 diabetes at an early stage of disease. Pediatr Diabetes 2014: 15: Triolo TM, Chase HP, Barker JM, DPT-1 Study Group. Diabetic subjects diagnosed through the Diabetes Prevention Trial-Type 1 (DPT-1) are often asymptomatic with normal A1C at diabetes onset. Diabetes Care 2009: 32: Näntö-Salonen K, Kupila A, Simell S et al. Nasal insulin to prevent type 1 diabetes in children with HLA genotypes and autoantibodies conferring increased risk of disease: a double-blind, randomised controlled trial. Lancet 2008: 372: Winkler C, Schober E, Ziegler AG, Holl RW. Markedly reduced rate of diabetic ketoacidosis at onset of type 1 diabetes in relatives screened for islet autoantibodies. Pediatr Diabetes 2012: 13: Lundgren M, Sahlin Å, Svensson C et al. Reduced morbidity at diagnosis and improved glycemic control in children previously enrolled in DiPiS follow-up. Pediatr Diabetes 2014: 15: Hekkala A, Reunanen A, Koski M, Knip M, Veijola R, Finnish Pediatric Diabetes Register. Age-related differences in the frequency of ketoacidosis at diagnosis of type 1 diabetes in children and adolescents. Diabetes Care 2010: 33: Insel RA, Dunne JL, Atkinson MA et al. Staging presymptomatic type 1 diabetes: a scientific statement of JDRF, the Endocrine Society, and the American Diabetes Association. Diabetes Care 2015: 38: Pediatric Diabetes 2016: 17 (Suppl. 22): 25 30