Age-period-cohort modelling of type 1 diabetes incidence rates among children included in the EURODIAB 25-year follow-up study.

AIMS: Specific patterns in incidence may reveal environmental explanations for type 1 diabetes incidence. We aimed to study type 1 diabetes incidence in European childhood populations to assess whether an increase could be attributed to either period or cohort effects. METHODS: Nineteen EURODIAB centres provided single year incidence data for ages 0-14 in the 25-year period 1989-2013. Case counts and person years were classified by age, period and cohort (APC) in 1-year classes. APC Poisson regression models of rates were fitted using restricted cubic splines for age, period and cohort per centre and sex. Joint models were fitted for all centres and sexes, to find a parsimonious model. RESULTS: A total of 57,487 cases were included. In ten and seven of the 19 centres the APC models showed evidence of nonlinear cohort effects or period effects, respectively, in one or both sexes and indications of sex-specific age effects. Models showed a positive linear increase ranging from approximately 0.6 to 6.6%/year. Centres with low incidence rates showed the highest overall increase. A final joint model showed incidence peak at age 11.6 and 12.6 for girls and boys, respectively, and the rate-ratio was according to sex below 1 in ages 5-12. CONCLUSION: There was reasonable evidence for similar age-specific type 1 diabetes incidence rates across the EURODIAB population and peaks at a younger age for girls than boys. Cohort effects showed nonlinearity but varied between centres and the model did not contribute convincingly to identification of environmental causes of the increase.

Substantial proportion of MODY among multiplex families participating in a Type 1 diabetes prediction programme.

AIMS: Patients with maturity-onset diabetes of the young (MODY) might be over-represented in families with histories of Type 1 diabetes. Our aim was to re-evaluate families participating in the Czech T1D Prediction Programme (PREDIA.CZ) with at least two members affected with diabetes to assess the proportion of MODY among these families and determine its most significant clinical predictors. METHODS: Of the 557 families followed up by the PREDIA.CZ, 53 (9.5%) had two or more family members with diabetes. One proband with diabetes from these families was chosen for direct sequencing of the GCK, HNF1A, HNF4A and INS genes. Non-parametric tests and a linear logistic regression model were used to evaluate differences between MODY and non-MODY families. RESULTS: MODY was genetically diagnosed in 24 of the 53 families with multiple occurrences of diabetes (45%). Mutations were detected most frequently in GCK (58%), followed by HNF1A (38%) and INS (4%). MODY families were more likely to have a parent with diabetes and had a higher proportion of females with diabetes than non-MODY families. Higher age (P < 0.001), a lower level of HbA1c (P < 0.001) at clinical onset and at least two generations affected by diabetes were the variables most predictive for probands of MODY families already presenting with diabetes. CONCLUSIONS: A prediction programme for Type 1 diabetes would provide a useful new source of patients with MODY most likely to benefit from an accurate diagnosis. This identification has implications for patient treatment and disease prognosis.

Seasonality at the clinical onset of type 1 diabetes-Lessons from the SWEET database.

BACKGROUND: Seasonality at the clinical onset of type 1 diabetes (T1D) has been suggested by different studies, however, the results are conflicting. This study aimed to evaluate the presence of seasonality at clinical onset of T1D based on the SWEET database comprising data from 32 different countries. METHODS: The study cohort included 23 603 patients (52% males) recorded in the international multicenter SWEET database (48 centers), with T1D onset ≤20 years, year of onset between 1980 and 2015, gender, year and month of birth and T1D-diagnosis documented. Data were stratified according to four age groups (<5, 5-<10, 10-<15, 15-20 years) at T1D onset, the latitude of European center (Northern ≥50°N and Southern Europe <50°N) and the year of onset ≤ or >2009. RESULTS: Analysis by month revealed significant seasonality with January being the month with the highest and June with the lowest percentage of incident cases (P < .001). Winter, early spring and late autumn months had higher percentage of incident cases compared with late spring and summer months. Stratification by age showed similar seasonality patterns in all four age groups (P ≤ .003 each), but not in children <24 months of age. There was no gender or latitude effect on seasonality pattern, however, the pattern differed by the year of onset (P < .001). Seasonality of diagnosis conformed to a sinusoidal model for all cases, females and males, age groups, northern and southern European countries. CONCLUSIONS: Seasonality at T1D clinical onset is documented by the large SWEET database with no gender or latitude (Europe only) effect except from the year of manifestation.

Seasonal variation in month of diagnosis in children with type 1 diabetes registered in 23 European centers during 1989-2008: little short-term influence of sunshine hours or average temperature.

BACKGROUND: The month of diagnosis in childhood type 1 diabetes shows seasonal variation. OBJECTIVE: We describe the pattern and investigate if year-to-year irregularities are associated with meteorological factors using data from 50 000 children diagnosed under the age of 15 yr in 23 population-based European registries during 1989-2008. METHODS: Tests for seasonal variation in monthly counts aggregated over the 20 yr period were performed. Time series regression was used to investigate if sunshine hour and average temperature data were predictive of the 240 monthly diagnosis counts after taking account of seasonality and long term trends. RESULTS: Significant sinusoidal pattern was evident in all but two small centers with peaks in November to February and relative amplitudes ranging from ± 11 to ± 38% (median ± 17%). However, most centers showed significant departures from a sinusoidal pattern. Pooling results over centers, there was significant seasonal variation in each age-group at diagnosis, with least seasonal variation in those under 5 yr. Boys showed greater seasonal variation than girls, particularly those aged 10-14 yr. There were no differences in seasonal pattern between four 5-yr sub-periods. Departures from the sinusoidal trend in monthly diagnoses in the period were significantly associated with deviations from the norm in average temperature (0.8% reduction in diagnoses per 1 °C excess) but not with sunshine hours. CONCLUSIONS: Seasonality was consistently apparent throughout the period in all age-groups and both sexes, but girls and the under 5 s showed less marked variation. Neither sunshine hour nor average temperature data contributed in any substantial way to explaining departures from the sinusoidal pattern.

Trends in childhood type 1 diabetes incidence in Europe during 1989-2008: evidence of non-uniformity over time in rates of increase.

AIMS/HYPOTHESIS: The aim of the study was to describe 20-year incidence trends for childhood type 1 diabetes in 23 EURODIAB centres and compare rates of increase in the first (1989-1998) and second (1999-2008) halves of the period. METHODS: All registers operate in geographically defined regions and are based on a clinical diagnosis. Completeness of registration is assessed by capture-recapture methodology. Twenty-three centres in 19 countries registered 49,969 new cases of type 1 diabetes in individuals diagnosed before their 15th birthday during the period studied. RESULTS: Ascertainment exceeded 90% in most registers. During the 20-year period, all but one register showed statistically significant changes in incidence, with rates universally increasing. When estimated separately for the first and second halves of the period, the median rates of increase were similar: 3.4% per annum and 3.3% per annum, respectively. However, rates of increase differed significantly between the first half and the second half for nine of the 21 registers with adequate coverage of both periods; five registers showed significantly higher rates of increase in the first half, and four significantly higher rates in the second half. CONCLUSIONS/INTERPRETATION: The incidence rate of childhood type 1 diabetes continues to rise across Europe by an average of approximately 3-4% per annum, but the increase is not necessarily uniform, showing periods of less rapid and more rapid increase in incidence in some registers. This pattern of change suggests that important risk exposures differ over time in different European countries. Further time trend analysis and comparison of the patterns in defined regions is warranted.

HbA1c-based diabetes diagnosis among patients with glucokinase mutation (GCK-MODY) is affected by a genetic variant of glucose-6-phosphatase (G6PC2).

AIMS: Genetic variation at the rs560887 locus of the glucose-6-phosphatase, catalytic 2 gene (G6PC2) is known to affect regulation of fasting glycaemia. We determined the rs560887 genotype of patients with monogenic diabetes and glucokinase gene mutations (GCK-MODY) and correlated the genotypes with HbA(1c) levels. METHODS: Patients from families with GCK-MODY were recruited from two large cohorts from Poland (n = 128) and the Czech Republic (n = 154). Genotypes at the rs560887 polymorphic site in G6PC2 were examined using real-time quantitative polymerase chain reaction. The effect of rs560887 genotype on age at diagnosis of GCK-MODY and initial HbA(1c) levels were evaluated separately within both cohorts. Following that, a meta-analysis of rs560887 genotype-HbA(1c) associations of both Polish and Czech cohorts was performed to confirm homogeneity of findings and validate cohort-specific results. RESULTS: GG homozygosity at rs560887 was associated with marginally elevated HbA(1c) levels (P = 0.07 in both cohorts). The effects observed in both groups were very homogeneous (Q = 0.18; P = 0.68). Meta-analysis showed that GG homozygosity at rs560887 was associated with mean HbA(1c) levels higher by 2.4 mmol/mol (0.24%), 95% CI 0.5-4.4 mmol/mol (0.05-0.44%) than in individuals with other genotypes. Additionally, meta-analysis of both cohorts showed that GG homozygous individuals had higher odds of reaching the 48 mmol/mol (6.5%) diagnostic threshold of diabetes; (odds ratio 1.90; 95% CI 1.07-3.36; P = 0.03). No such effects were observed for age at diagnosis of diabetes. CONCLUSIONS: Variation at the rs560887 locus of G6PC2 is associated with worse glycated haemoglobin levels in individuals with GCK mutations; GG homozygotes are more likely to meet diagnostic criteria for diabetes based on HbA(1c) level.

Boys with haemophilia have low trabecular bone mineral density and sarcopenia, but normal bone strength at the radius.

Although a decreased areal bone mineral density (BMD) has been reported in patients with haemophilia, data are lacking that would reflect the three-dimensional structure of the bone and the muscle-bone relationship. We aimed to assess volumetric BMD, bone geometry and muscle-bone phenotype in boys with haemophilia, and to describe the association between clinical characteristics of haemophilia and bone quality and structure. A cross-sectional study was conducted in 41 boys with haemophilia (mean age 12.4, range 6.6-19.8 years) using peripheral quantitative CT (pQCT) at the nondominant forearm. Results were transformed into Z-scores using previously published reference data. Significant differences were tested by one-sample t-test or sign test. Two-sample t-test and anova were used to compare results between subgroups of patients divided according to the severity of the disease, the fracture history and the number of joint and muscle bleedings. Boys with haemophilia had a decreased trabecular volumetric BMD (mean Z-score -0.5, P < 0.01), while their cortical volumetric BMD was increased (mean Z-score 0.4, P < 0.05). The volumetric bone mineral content and the bone geometry at the radial diaphysis were normal when adjusted for patients' shorter body height. Muscle area was decreased (mean Z-score -1.0, P < 0.001), irrespective of age. No association was observed of bone quality parameters and bone geometry with the disease severity, fracture history or number of bleedings. Bone strength measured at the diaphysis of the radius is not impaired in boys with haemophilia. The finding of the decreased trabecular bone density can be most likely attributed to their sarcopenia.

Identification of germline susceptibility loci in ETV6-RUNX1-rearranged childhood acute lymphoblastic leukemia.

Acute lymphoblastic leukemia (ALL) is a malignant disease of the white blood cells. The etiology of ALL is believed to be multifactorial and likely to involve an interplay of environmental and genetic variables. We performed a genome-wide association study of 355 750 single-nucleotide polymorphisms (SNPs) in 474 controls and 419 childhood ALL cases characterized by a t(12;21)(p13;q22) - the most common chromosomal translocation observed in childhood ALL - which leads to an ETV6-RUNX1 gene fusion. The eight most strongly associated SNPs were followed-up in 951 ETV6-RUNX1-positive cases and 3061 controls from Germany/Austria and Italy, respectively. We identified a novel, genome-wide significant risk locus at 3q28 (TP63, rs17505102, P(CMH)=8.94 × 10(-9), OR=0.65). The separate analysis of the combined German/Austrian sample only, revealed additional genome-wide significant associations at 11q11 (OR8U8, rs1945213, P=9.14 × 10(-11), OR=0.69) and 8p21.3 (near INTS10, rs920590, P=6.12 × 10(-9), OR=1.36). These associations and another association at 11p11.2 (PTPRJ, rs3942852, P=4.95 × 10(-7), OR=0.72) remained significant in the German/Austrian replication panel after correction for multiple testing. Our findings demonstrate that germline genetic variation can specifically contribute to the risk of ETV6-RUNX1-positive childhood ALL. The identification of TP63 and PTPRJ as susceptibility genes emphasize the role of the TP53 gene family and the importance of proteins regulating cellular processes in connection with tumorigenesis.

Birthweight and the risk of childhood-onset type 1 diabetes: a meta-analysis of observational studies using individual patient data.

AIMS/HYPOTHESIS: We investigated whether children who are heavier at birth have an increased risk of type 1 diabetes. METHODS: Relevant studies published before February 2009 were identified from literature searches using MEDLINE, Web of Science and EMBASE. Authors of all studies containing relevant data were contacted and asked to provide individual patient data or conduct pre-specified analyses. Risk estimates of type 1 diabetes by category of birthweight were calculated for each study, before and after adjustment for potential confounders.Meta-analysis techniques were then used to derive combined ORs and investigate heterogeneity between studies. RESULTS: Data were available for 29 predominantly European studies (five cohort, 24 case-control studies), including 12,807 cases of type 1 diabetes. Overall, studies consistently demonstrated that children with birthweight from 3.5 to 4 kg had an increased risk of diabetes of 6% (OR 1.06 [95% CI 1.01-1.11]; p=0.02) and children with birthweight over 4 kg had an increased risk of 10% (OR 1.10 [95% CI 1.04-1.19]; p=0.003), compared with children weighing 3.0 to 3.5 kg at birth. This corresponded to a linear increase in diabetes risk of 3% per 500 g increase in birthweight (OR 1.03 [95% CI 1.00-1.06]; p=0.03). Adjustments for potential confounders such as gestational age, maternal age, birth order, Caesarean section, breastfeeding and maternal diabetes had little effect on these findings. CONCLUSIONS/INTERPRETATION: Children who are heavier at birth have a significant and consistent, but relatively small increase in risk of type 1 diabetes.

Variants of CARD15, TNFA and PTPN22 and susceptibility to Crohn's disease in the Czech population: high frequency of the CARD15 1007fs.

Crohn's disease (CD) has been shown to be associated with the variants in the CARD15 gene as well as in other genes involved in the immune response. The frequencies of the variants profoundly differ among populations and so does the associated risk. We examined the associations of variants in the CARD15, TNFA and PTPN22 genes with pediatric-onset and adult-onset CD in the Czech population. Genotype, phenotype and allelic frequencies were compared between 345 patients with CD (136 pediatric-onset and 209 adult-onset patients) and 501 unrelated healthy controls. At least one minor allele of the CARD15 gene was carried by 46% patients and only 21% control subjects (OR = 3.2, 95% CI 2.4-4.4). In a multiple logistic regression model, the strongest association with CD was found for the 1007fs variant (OR = 4.6, 95% CI 3.0-7.0), followed by p.G908R (OR = 2.9, 95% CI 1.5-5.7) and p.R702W (OR = 1.7, 95% CI 1.0-2.9), while no independent association was found for the remaining variants in the CARD15 gene (p.268S, p.955I and p.289S), for the p.R620W variant in the PTPN22 gene or for the g.-308G>A variant in the TNFA gene. The age at CD onset was strongly modified by positivity for the 1007fs allele: it was present in 42% pediatric-onset and only 25% adult-onset patients. In conclusion, we report a high frequency of the minor allele of the CARD15 1007fs polymorphism in the Czech population and a strong effect of this allele on the age at disease onset.

Caesarean section is associated with an increased risk of childhood-onset type 1 diabetes mellitus: a meta-analysis of observational studies.

AIMS/HYPOTHESIS: The aim of this study was to investigate the evidence of an increased risk of childhood-onset type 1 diabetes in children born by Caesarean section by systematically reviewing the published literature and performing a meta-analysis with adjustment for recognised confounders. METHODS: After MEDLINE, Web of Science and EMBASE searches, crude ORs and 95% CIs for type 1 diabetes in children born by Caesarean section were calculated from the data reported in each study. Authors were contacted to facilitate adjustments for potential confounders, either by supplying raw data or calculating adjusted estimates. Meta-analysis techniques were then used to derive combined ORs and to investigate heterogeneity between studies. RESULTS: Twenty studies were identified. Overall, there was a significant increase in the risk of type 1 diabetes in children born by Caesarean section (OR 1.23, 95% CI 1.15-1.32, p < 0.001). There was little evidence of heterogeneity between studies (p = 0.54). Seventeen authors provided raw data or adjusted estimates to facilitate adjustments for potential confounders. In these studies, there was evidence of an increase in diabetes risk with greater birthweight, shorter gestation and greater maternal age. The increased risk of type 1 diabetes after Caesarean section was little altered after adjustment for gestational age, birth weight, maternal age, birth order, breast-feeding and maternal diabetes (adjusted OR 1.19, 95% CI 1.04-1.36, p = 0.01). CONCLUSIONS/INTERPRETATION: This analysis demonstrates a 20% increase in the risk of childhood-onset type 1 diabetes after Caesarean section delivery that cannot be explained by known confounders.

Failure of multiple antivirals to affect high HHV-6 DNAaemia resulting from viral chromosomal integration in case of severe aplastic anaemia.

We report a fifty-year-old woman presenting with severe aplastic anaemia (SAA) and prolonged high Human Herpesvirus 6 (HHV6) variant A DNAeamia detected by quantitative PCR. Multiple antiviral treatments failed to affect the HHV6 DNAemia and subsequent immunosuppressive treatment reached only partial improvement as judged by bone marrow examinations. The patient remained dependent on thrombocyte transfusions and G-CSF treatment. After one year of steady high HHV6 DNA load in blood, viral chromosomal integration was proved by demonstrating the viral DNA in hair follicles. This condition appeared to be unconnected with, and to have no effect, on the original SAA.

Thyroid autoimmunity in children with coexisting type 1 diabetes mellitus and celiac disease: a multicenter study.

BACKGROUND: Children with type 1 diabetes mellitus (DM1) are more prone to developing thyroid autoimmunity (TAI); TAI also occurs more frequently in patients with celiac disease (CD). AIM: To determine whether TAI occurs more frequently in children with coexisting DM1 and CD compared to children with DM1 only, and whether the clinical course of DM1 is influenced by concomitant TAI. PATIENTS AND METHODS: We performed a multicenter retrospective case-control study comparing data from 84 diabetic children with CD (group 1) to 167 diabetic children without CD (group 2), matched by age at DM1 onset, duration of DM1 and center. Markers of TAI, thyroid function and HbA1c were recorded. The TAI follow-up lasted 4.9 +/- 2.8 years. RESULTS: TAI was diagnosed in 13% of children in group 1 and 19% of children in group 2 (ns). Diabetes control was not influenced by TAI in either group. CONCLUSIONS: Occurrence of TAI in diabetic children is not related to coexisting CD. TAI does not lead to worsening of metabolic control in children with DM1.

[EBV quantification in children after allogeneic hematopoietic stem cell transplantation]. / Kvantifikace EBV u detí po alogenní transplantaci hematopoetických kmenových bunek.

BACKGROUND: Patients undergoing allogeneic hematopoietic stem cell transplantation (AHSCT) are endangered by developing Epstein-Barr virus-related post-transplant lymfoprolipherative disease (EBV-LPD). The aims of the study were to retrospectively characterise the viral loads in four patients who died of this complication, and to test possible risk factors for EBV reactivation in a prospectively observed cohort of children after AHSCT. METHODS AND RESULTS: Serial DNA samples extracted from whole blood from four patients who died of post-transplant EBV-LPD in year 2000 were retrospectively analysed for EBV load using quantitative real-time PCR. First detection of EBV activation preceded death by 24-91 days. All four patients exceeded a viral load of one million EBV copies per 100,000 human genome equivalents. A cohort of 72 children undergoing AHSCT between 2001-2004 was prospectively followed-on using the same quantification method from regularly obtained samples of whole blood, and clinical and laboratory data were recorded on a weekly basis, totalling at 3,896 person-weeks of observation. Approximately one half of the cohort experienced at least one episode of EBV reactivation during the first 100 days after AHSCT, four of the episodes being accompanied with viral loads higher than our provisional threshold of 10,000 copies per 100,000 human genome equivalents. Three of the four patients developed EBV-LPD and were successfully treated by intravenous administration of anti-CD20 antibody. Testing of possible clinical and laboratory predictors of EBV reactivation did not reveal any clinically useful association. CONCLUSIONS: The cornerstone of predicting EBV-LPD in AHSCT is a regular monitoring of EBV viral load using quantitative methods. Using this strategy with a threshold of 10,000 EBV copies per 100,000 human genome equivalents was proved to be effective, as shown by no death of EBV for the study period, compared to four cases in the year before the quantitative monitoring.

Longitudinal observation of enterovirus and adenovirus in stool samples from Norwegian infants with the highest genetic risk of type 1 diabetes.

BACKGROUND: Enterovirus and adenovirus are common in infancy, causing mostly asymptomatic infections. However, even an asymptomatic infection may be associated with increased risk of development of certain chronic non-infectious diseases, as has been suggested for enterovirus and type 1 diabetes. Data on occurrence and course of the infections in infancy are therefore important for designing effective approaches towards study of the association. OBJECTIVES: To estimate the frequency of enterovirus and adenovirus infections in Norwegian infants, to evaluate the duration of the infections, to investigate their association with symptoms, and to establish a robust procedure that will be used to study the association between these viruses and the development of auto-immunity leading to type 1 diabetes. STUDY DESIGN: Parents of infants, recruited for a study on environmental triggers of type 1 diabetes, submitted monthly samples of infant faeces, as well as information on symptoms of infection. The samples were analysed for enterovirus and adenovirus using quantitative real-time PCR, and enterovirus-positive samples were sequenced. RESULTS: Enteroviruses were found in 142/1,255 (11.3%), and adenoviruses in 138/1,255 (11.0%) of stool samples. Approximately half of the infants were exposed to these viruses at least once during the first year of observation (period 3-14 months of age). The presence of adenovirus was associated with fever and with symptoms of cold but not with diarrhoea and vomiting. The enterovirus positivity was not associated with any symptoms. CONCLUSIONS: The prevalence of enterovirus and adenovirus in longitudinally obtained faecal samples from infants is sufficiently high to enable studies of their association with chronic diseases. The present protocol for evaluating exposure to these viruses is well suited for large-scale efforts aimed at assessing possible long-term consequences, particularly in relation to type 1 diabetes.

Interferon-gamma and cortisol levels in cerebrospinal fluid and its relationship to the etiology of aseptic meningoencephalitis.

The aim of the study was to analyze the concentrations of Th1/Th2 cytokines and cortisol in the cerebrospinal fluid (CSF) from patients with aseptic meningoencephalitis (AM). The study enrolled 37 patients with AM and 11 control subjects. CSF concentrations of IL-2, IL-4, IL-5, IL-10, IFN-gamma, and TNF-alpha were analyzed using cytokine bead array and flow cytometry; CSF cortisol concentrations were measured by a RIA method. Cortisol was detected in 37 CSF samples (100%) from patients with AM, and it was significantly elevated in comparison to control subjects. IFN-gamma was detected in 32 CSF samples (86.5%) and IL-10 was detectable in 9 CSF samples (24.3%). The CSF cortisol levels correlated negatively with the duration of AM. The intrathecal concentration of IFN-gamma correlated positively with CSF numbers of leukocytes and lymphocytes, and negatively with the duration of AM. The etiology of AM influenced the CSF cortisol concentration, which was significantly higher in patients with tick-borne encephalitis when compared to persons with AM of unknown origin and control subjects. The results indicate that the prevailing intrathecal immune reaction during AM is shifted to a Th1-like response, whereas anti-inflammatory response in the brain is executed by the effect of cortisol.