There is evidence for a role of infections within the pathogenesis of islet autoimmunity and type 1 diabetes mellitus (T1D), but previous studies did not allow assessment of potential critical time windows in this context.
To examine whether early, short-term, or cumulative exposures to episodes of infection and fever during the first 3 years of life were associated with the initiation of persistent islet autoimmunity in children at increased T1D risk.
Prospective cohort study with daily infection records and regular assessment of islet autoimmunity.
Diabetes Research Institute, Munich, Germany.
A total of 148 children at high T1D risk with documentation of 1680 infectious events in 90 750 person-days during their first 3 years of life.
Main Outcomes and Measures
Hazard ratios (HRs) for seroconversion to persistent islet autoantibodies were assessed in Cox regression models with numbers of respiratory, gastrointestinal, and other infections, adjusting for sex, delivery mode, intervention group, season of birth, and antibiotic use.
An increased HR of islet autoantibody seroconversion was associated with respiratory infections during the first 6 months of life (HR = 2.27; 95% CI, 1.32-3.91) and ages 6.0 to 11.9 months (HR = 1.32; 95% CI, 1.08-1.61). During the second year of life, no meaningful effects were detected for any infectious category. A higher number of respiratory infections in the 6 months prior to islet autoantibody seroconversion was also associated with an increased HR (HR = 1.42; 95% CI, 1.12-1.80).
Conclusions and Relevance
Respiratory infections in early childhood are a potential risk factor for the development of T1D.
The incidence of type 1 diabetes mellitus (T1D) is increasing worldwide,1 but its etiology is still not well understood. Infections have been discussed as important environmental determinants in the pathogenesis of T1D.2-4 Retrospective case-control studies showed that patients with newly diagnosed T1D had higher titers of enterovirus antibodies5 and were more likely to be positive for enterovirus RNA6,7 compared with healthy controls. Prospective studies from Finland confirmed the potential association between enterovirus infections and islet autoimmunity,8,9 while others from Germany and the United States did not.10-12
These studies were based on a nested case-control design, which has only a limited ability to control for potential confounding factors. Furthermore, the approaches used did not allow detailed analyses on whether there are critical time windows in which infections might have a particularly strong influence on the development of islet autoimmunity. Moreover, infectious agents other than enteroviruses, eg, rotaviruses, have been reported to be related to T1D.13-15
The occurrence of fever is also of interest in this context, as fever indicates a strong immune response and therefore a potentially sufficient and fast removal of the possibly autoimmunity-triggering pathogen.16,17 However, fever might also indicate a severe infection and excessive immune response leading to tissue damage and activation of autoreactive cells.18,19 Therefore, it is not yet known whether fever is a favorable or unfavorable prognostic factor in the pathogenesis of autoimmune diseases.
The aim of this study was to examine whether infections and fever episodes during the first 3 years of life were associated with the risk of developing islet autoimmunity. We hypothesized 3 scenarios to explain how infectious diseases might be involved in this context. First, early exposure, eg, in the first months of life, may be instrumental in inducing a state of susceptibility for future seroconversion. Alternatively, infections may have a short-term effect, causing islet autoantibody seroconversion in the following few months. Third, infections might have a cumulative effect such that the frequency of infectious events would increase the likelihood of islet autoimmunity irrespective of when they occur. To study this, we took advantage of data from children at high risk for T1D who had been followed up from 3 months of age within the BABYDIET study.
The BABYDIET study is an intensively monitored German dietary intervention study testing the potential effect of delayed gluten exposure on the development of islet autoimmunity in children at increased risk for diabetes. The study cohort is still followed up and has been described in detail elsewhere.20,21 In brief, 150 children younger than 3 months with at least 1 first-degree relative with T1D and 1 of 3 specific T1D-associated HLA genotypes were recruited between August 2000 and August 2006 (participation rate, 88.8%) and randomized to exposure to dietary gluten in the first year of life. After inclusion, children were followed up in 3-monthly intervals until age 3 years and yearly thereafter for efficacy (persistent islet autoantibodies) and safety assessment, including intensive monitoring with 3-monthly sample collections of venous blood, urine, and stool.
The BABYDIET study was conducted at the Diabetes Research Institute, Munich, Germany, and approved by the ethics committee of the Ludwig-Maximilian University, Munich.
Assessment of Infectious Episodes
At the child’s age of 3 months, parents completed a detailed questionnaire on their children’s history of infections, fever, and medication. Specifically, they were asked about fever, infectious symptoms (such as diarrhea, vomiting, constipation, and allergies), and the names of administered pharmaceutical agents or their active ingredients with starting date and duration of infections and medication. The questionnaire additionally addressed the occurrence of diabetes and other autoimmune diseases in the child or his or her family and lifestyle habits such as smoking during pregnancy. All families then received a BABYDIET book in which they were asked to record all infections and fever events (temperature >38°C) together with medication and health visits of their children on a daily basis until age 3 years. Parents were instructed to report disease-free days in the BABYDIET book as well. The diseases were classified using the International Statistical Classification of Diseases, 10th Revision (ICD-10) code from 2011 based on the physicians’ diagnosis (if medical care had been sought) or on parental reports of symptoms.
Infectious disease was defined as an acute event according to the ICD-10 code or by a symptom indicating an infectious genesis. Infectious events were assigned to a specific time interval by their date of onset. We defined 3 categories of infectious diseases: (1) infections of the respiratory tract, of the ear, nose, and throat, and of the eye (if inflammatory symptoms of the respiratory tract were reported); (2) gastrointestinal tract infections (if the main symptoms were diarrhea and/or vomiting); and (3) other infections (eg, with symptoms of skin or mucosa lesions). Other disease events such as allergies or accidents were not considered as infectious diseases. Separate infectious diseases of 1 category were defined as 1 infectious event if there were less than 6 days of potential remission between the respective infections, as these seemed likely to be caused by the same infectious agent. Additionally, we defined a category of “any infections” in which infectious diseases of different categories were considered as 1 infectious event if their time intervals overlapped.
Assessment of Islet Autoantibody Seroconversion
Seroconversion was defined as the development of persistent autoantibodies to 1 or more of the following antigens: insulin, glutamic acid decarboxylase 65, insulinoma antigen 2, or zinc transporter 8. Only the first event of seroconversion in each subject was taken into account. Measurement of autoantibodies has been described elsewhere.21 The upper limits of normal were determined using quantile-quantile plots, corresponded to the 99th percentile of control children, and were 1.5 local units/mL for insulin autoantibodies, 35 World Health Organization units/mL for glutamic acid decarboxylase antibodies, 5 World Health Organization units/mL for insulinoma antigen 2 antibodies, 16 units/mL for zinc transporter 8–arginine antibodies, and 30 units/mL for zinc transporter 8–tryptophan antibodies. Using these thresholds for positivity, the assays had respective sensitivities and specificities of 70% and 99% (insulin autoantibodies), 86% and 93% (glutamic acid decarboxylase antibodies), 72% and 100% (insulinoma antigen 2 antibodies), and 84% and 100% for multiple islet autoantibodies in the Diabetes Antibody Standardization Program proficiency workshop.22 Persistence was defined as being positive in at least 2 consecutive samples and in the last available sample. Islet autoantibodies were measured in venous blood samples from all scheduled visits and therefore every 3 months. Diabetes development was monitored and diagnosed according to the American Diabetes Association Expert Committee on the Diagnosis and Classification of Diabetes Mellitus criteria.23
From the daily records in the BABYDIET books, which also covered disease-free days, it was possible to distinguish between documented days (with or without disease) and undocumented days. To avoid potential bias due to variation in the number of documented days between subjects and time periods, we calculated the numbers of infectious disease and fever events per 100 documented days for each child as a measure of incidence. Due to the detailed questionnaire at the child’s age of 3 months, we considered all days from birth to this visit as being documented. The analyses were restricted to observations based on at least 20 documented days in the respective time interval.
To address the early-exposure hypothesis, we calculated hazard ratios (HRs) of time until seroconversion to islet autoantibodies with respective 95% CIs by the number of infectious events per 100 documented days in the intervals of 0 to 5.9 months, 6.0 to 11.9 months, and 12.0 to 23.9 months of age using Cox proportional hazard regression model. For any time point, we calculated 4 models, using total number of infectious events as predictor in the first model, numbers of respiratory, gastrointestinal, and other infections as separate predictors in the second model, number of any fever events (recorded with or without an infectious episode) as predictor in the third model, and number of fever events without infection in the fourth model. In total, the data set included 26 events of islet autoantibody seroconversion, which occurred at ages 0.59 to 7.90 years. There were no observations with reported islet autoantibody seroconversion in the first 6 months of life; those with seroconversion at ages 6.0 to 11.9 months were excluded from the analysis of ages 12.0 to 23.9 months.
To examine the short-term effect hypothesis, we calculated Cox regression models with numbers of infectious and fever events per 100 documented days in the last 6 months before autoantibody seroconversion within the first 3 years of life as time-varying predictors.24
To disentangle potential early-exposure and short-term effects, we performed 2 sensitivity analyses. In the first, we restricted the models with infectious or fever events during the first half year of life to those seroconversion events that occurred after the first year of life. In the second, we restricted the models with time-varying predictors to those seroconversion events that happened between ages 1.5 and 3 years.
All Cox regression models were adjusted for sex, delivery mode (cesarean or vaginal), intervention group, season of birth (March-May; June-August; September-November; December-February), and frequency of antibiotic use and were tested for interaction with the virus-sensing interferon induced with helicase C domain 1 (IFIH1) gene polymorphism.25
With respect to the cumulative-exposure hypothesis, we compared the distributions of the recorded numbers of infectious events in children with and without autoantibody seroconversion during their first 2 years of life in 3-monthly intervals.
One subject dropped out of the study after the first visit, and another one had less than 20 documented days in every age interval examined. Both children were therefore excluded from the sample, leaving a final sample size of 148 subjects.
The level of significance was set to .05 in all analyses. All calculations were carried out with SAS version 9.2 statistical software (SAS Institute, Inc). Plotting was done using R version 2.14.1 statistical software (http://cran.r-project.org).
In total, our data comprised 90 750 documented days (Table 1). Overall, 1680 events (11.35 per child) and 690 events (4.66 per child) had been reported with respect to infections and fever, respectively. Children with islet autoantibody seroconversion during follow-up were significantly more exposed to respiratory and gastrointestinal infections as well as to fever, both during the first year of life and overall (Table 2). The median number of documented days decreased from 350.5 in the first year of life to 177.5 in the second year and 12.5 in the third year. The median number of documented days and median follow-up time per child were similar in both groups.
While there were no islet autoantibody seroconversion events observed in the first 6 months of life, the incidence rates of seroconversion per 100 person-years were 8.51 in the second half year, 4.07 in the second year, and 3.67 in the third year of life. The mean incidences of the 3 infection categories also increased considerably after the first 6 months of life and remained relatively constant thereafter, with a slight decline in the third year of life (Figure 1). The same tendency was observed for the incidence of fever. Respiratory infections constituted the majority of the infections recorded.
Addressing the Early-Exposure Hypothesis
Despite the overall lower incidence of infections in the first 6 months of life, the number of infections during this age period per 100 documented days was associated with an increase in the HR for islet autoantibody seroconversion in both crude and adjusted analyses (Table 3). The effect size was particularly high for respiratory infections (adjusted HR = 2.27; 95% CI, 1.32-3.91), while there were no significant associations with gastrointestinal and other infections in this early period. The results were similar when we excluded all seroconversion events that happened during the first year of life (data not shown).
In the following 6 months of life, any infections and respiratory infections were again significantly associated with islet autoantibody seroconversion. However, the effect size was much lower than in the preceding period (eg, respiratory infections: adjusted HR = 1.32; 95% CI, 1.08-1.61). There were no significant associations with fever events in any of the 2 periods. During the second year of life, no meaningful effects were detected for infections or fever (data not shown). We did not detect significant interactions with IFIH1 genotype in any model (data not shown).
When we explored the associations observed in the Cox regression analyses in further detail (Figure 2), we found that the rate of islet autoantibody seroconversion was highest in children with more than 5 recorded events of respiratory infections in the first year of life. For all seroconverters, at least 2 infections had been reported in this period, including respiratory infections for all except 1 case.
Addressing the Short-term Effect Hypothesis
Calculations with time-dependent covariates indicated that the HR for seroconversion was significantly increased by a higher number of respiratory infections per 100 documented days within the preceding 6 months (adjusted HR = 1.42; 95% CI, 1.12-1.80) (Table 4). Again, gastrointestinal and other infections did not have a positive association with the seroconversion rate. However, we detected significant associations with any fever events and fever events without infection. When we restricted these analyses to seroconversion events after 1.5 years of life to disentangle potential effects of the first year of life and the preceding 6 months, the results were similar but not significant, with the exception of any fever events. Again, we found no interactions with IFIH1 genotype (data not shown). For all seroconverters, at least 1 infection had been reported within 6 months prior to the seroconversion event, including respiratory infections for all except 2 cases (data not shown).
Addressing the Cumulative-Exposure Hypothesis
Between 9 and 18 months of life, subjects with later islet autoantibody seroconversion had on average a higher total number of infectious events recorded than their peers without seroconversion as determined by Mann-Whitney U tests (data not shown). The respective interquartile ranges of their number of cumulative infections differed accordingly, although the highest exposures were observed in children without seroconversion (Figure 3). The upper quartiles (75th percentiles) of the distributions of the cumulative numbers of respiratory infections increased considerably faster in seroconverters compared with healthy peers during the first year of life, but this difference remained relatively constant thereafter (Figure 4). There were no meaningful results with respect to gastrointestinal and other infections. Therefore, the average higher cumulative infection number in children with later seroconversion (as concluded from higher interquartile ranges) reflected mainly a higher exposure to respiratory infections during the first year of life. All these findings were similar when we compared numbers of events per 100 documented days (data not shown).
In an attempt to identify specific infectious diseases that might be mainly responsible for the increased rate of seroconversion by respiratory infections, we compared infection rates between children with and without seroconversion. We observed significantly higher rates of infections of the upper respiratory tract (Mann-Whitney U test, P = .005) and among these of acute rhinopharyngitis (ICD-10 code J00, P = .001). When we used infections of the upper respiratory tract as predictor variable in the Cox regression models, we found significant associations at ages 0 to 5.9 months, at ages 6.0 to 11.9 months, and for 6 months before seroconversion (Table 5). We observed no significant associations with infections of the lower respiratory tract (Table 5), of ear, nose, and throat, and of the eye (data not shown).
Through the daily documentation of illnesses and fever in children with a family history of T1D during the first 3 years of life, we could demonstrate that exposure to infectious diseases, and in particular to respiratory infections, in early life predicted the development of islet autoimmunity. Primarily, we found strong associations with infections occurring in the first year of life, confirming the early-exposure hypothesis.
There was also some evidence for the short-term effect hypothesis, as infections and fever events within 6 months prior to islet autoantibody seroconversion were also associated with an increased risk of islet autoimmunity. Notably, all children who developed islet autoimmunity had at least 2 infections in the first year of life and at least 1 infection within 6 months before islet autoantibody seroconversion. However, cumulative exposure alone seemed not to be instrumental, as potential differences in the cumulative disease numbers between children with and without islet autoantibody seroconversion mainly reflected different exposure to infections in the first year of life only. Fever seemed to be associated with a detrimental rather than protective effect, particularly with respect to short-term effects, but these findings were not conclusive.
These results are novel. To our knowledge, this study for the first time prospectively assessed infectious diseases together with their starting date and duration in children at risk for T1D. Overall, only few data exist about the incidence of infections and fever in infancy. Our finding that the infection rates increased largely after the first 6 months of life and remained relatively stable thereafter is comparable to results from another study26 and appears likely to reflect protection by maternal passive immunity, which is known to decline during the first year of life.27-29 Extrapolated to yearly rates, our incidences of respiratory infections were comparable to those observed in another German population.30 Other strengths of our study lie in the frequent 3-monthly measurement of all 4 relevant islet autoantibodies and in the fact that parents reported disease-free days as well, so that we were able to estimate disease frequencies without reporting bias. Accompanied collection of medication data even enabled us to rule out that our results were confounded by antibiotic use.
A potential limitation is the sample size, but such frequent follow-up of study subjects, detailed phenotyping, and daily documentation of infectious events constitute a challenge for both investigators and participants, so that it appears difficult to obtain a larger population size with such a study design. As a consequence, our data did not allow us to finally exclude that infections other than upper respiratory tract infections might also have a potential effect on development of islet autoimmunity, as the statistical power might simply have been too low to detect such associations. Apparently for the same reason, we were not able to finally exclude that potential short-term associations observed in our data might be due to early-exposure effects. Although we performed a number of sensitivity analyses, it appears unlikely that this has resulted in a multiple-testing problem. The results of these analyses were only used to assess the validity of the main analyses (which had been performed beforehand) and were not interpreted otherwise.
Various mechanisms have been discussed to explain how infectious diseases might induce autoreactivity in T1D. Results from animal studies indicate that different viruses can affect the development of islet autoimmunity via various mechanisms such as direct beta-cell lysis, bystander activation of autoreactive T cells, loss of regulatory T cells, and molecular mimicry,31,32 which might therefore also be relevant in humans.33-36
The period between ages 6 months and 3 years has been identified as the peak incidence period of islet autoimmunity in the BABYDIET data as well as in other studies,37-39 implying that most of the relevant causative events occur at or before this time. Our results indicate that early exposure to infections increases the susceptibility for developing islet autoantibodies. Although it might appear plausible that such a state of susceptibility would soon lead to islet autoantibody seroconversion, we found that exposure to infections in the first 6 months of life was also associated with seroconversion after the first year of life, thus also indicating longer-term effects by early exposure.
Potential prevention strategies against T1D derived from studies like this might address early vaccination against specific infectious agents. Unfortunately, we were not able to identify a single infectious agent that might be instrumental in the development of T1D. Our results point to a potential role of infections in the upper respiratory tract and specifically of acute rhinopharyngitis. These results are partly supported by the MIDIA study,40 which also indicated that respiratory infections are associated with the development of islet autoimmunity. In general, our findings are in accordance with other previous studies that showed associations with enterovirus infections,5,6,8,9 as rhinoviruses—an enterovirus species—are known to be the major causative agents of respiratory infections.41-43
In conclusion, our study identified respiratory infections in early childhood, especially in the first year of life, as a risk factor for the development of T1D. We also found some evidence for short-term effects of infectious events on development of autoimmunity, while cumulative exposure alone seemed not to be causative.
Correction: This article was corrected online November 21, 2016, to correct numbers in the Abstract, text, and Table 1.
Corresponding Author: Anette-Gabriele Ziegler, MD, Institute of Diabetes Research, Helmholtz Zentrum München, Ingolstädter Landstraße 1, Neuherberg, D-85764, Germany (email@example.com).
Published Online: July 1, 2013. doi:10.1001/jamapediatrics.2013.158
Author Contributions: Beyerlein had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Ziegler is the principal investigator of the BABYDIET study.
Study concept and design: Beyerlein, Ziegler.
Acquisition of data: Ziegler, Pflueger.
Analysis and interpretation of data: Beyerlein, Wehweck.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: Ziegler.
Statistical analysis: Beyerlein, Wehweck.
Obtained funding: Ziegler.
Administrative, technical, and material support: Pflueger.
Study supervision: Ziegler.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grants ZI 310/14-1, ZI 310/14-2, and BE 4682/1-1from Deutsche Forschungsgemeinschaft and by Stiftung “Das zuckerkranke Kind.”
Role of the Sponsors: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.
Additional Contributions: Sandra Hummel, PhD, coordinated the study at the start, Christiane Winkler, PhD, provided database management, Kerstin Adler, PhD, and Ramona Puff, PhD, provided laboratory management, and Annette Knopff, Steffi Krause, Anita Gavrisan, and Marlon Scholz provided expert technical assistance.
Additional Information: This work is partly based on the dissertation of Wehweck at the Technische Universität München.
et al. Trends in childhood type 1 diabetes incidence in Europe during 1989-2008: evidence of non-uniformity over time in rates of increase. Diabetologia
. 2012;55(8):2142-2147.PubMedGoogle ScholarCrossref
M. The epidemic of type 1 diabetes: what is it telling us? Curr Opin Endocrinol Diabetes Obes
. 2011;18(4):248-251.PubMedGoogle ScholarCrossref
et al. Cellular tropism of human enterovirus D species serotypes EV-94, EV-70, and EV-68 in vitro: implications for pathogenesis. J Med Virol
. 2010;82(11):1940-1949.PubMedGoogle ScholarCrossref
KW. Viral antibodies in diabetes mellitus. Br Med J
. 1969;3(5671):627-630.PubMedGoogle ScholarCrossref
KW. Coxsackie B virus infection and onset of childhood diabetes. Lancet
. 1995;346(8969):221-223.PubMedGoogle ScholarCrossref
ME. Enterovirus infection and type 1 diabetes mellitus: systematic review and meta-analysis of observational molecular studies. BMJ
. 2011;342:d35.PubMedGoogle ScholarCrossref
et al; Finnish TRIGR Study Group. Enterovirus infections as a risk factor for type I diabetes: virus analyses in a dietary intervention trial. Clin Exp Immunol
. 2003;132(2):271-277.PubMedGoogle ScholarCrossref
et al. Enterovirus infections are associated with the induction of beta-cell autoimmunity in a prospective birth cohort study. J Med Virol
. 2003;69(1):91-98.PubMedGoogle ScholarCrossref
AG. No evidence for an association of coxsackie virus infections during pregnancy and early childhood with development of islet autoantibodies in offspring of mothers or fathers with type 1 diabetes. J Autoimmun
. 2001;17(4):333-340.PubMedGoogle ScholarCrossref
et al. Prospective study of enteroviral infections and development of beta-cell autoimmunity: Diabetes Autoimmunity Study in the Young (DAISY). Diabetes Res Clin Pract
. 2003;59(1):51-61.PubMedGoogle ScholarCrossref
et al. Human enterovirus infections in children at increased risk for type 1 diabetes: the Babydiet study. Diabetologia
. 2011;54(12):2995-3002.PubMedGoogle ScholarCrossref
et al. Association between rotavirus infection and pancreatic islet autoimmunity in children at risk of developing type 1 diabetes. Diabetes
. 2000;49(8):1319-1324.PubMedGoogle ScholarCrossref
M. How robust is the evidence for viruses in the induction of type 1 diabetes? Curr Opin Immunol
. 2005;17(6):616-623.PubMedGoogle ScholarCrossref
T, von Herrath
M. Virus infections in type 1 diabetes. Cold Spring Harb Perspect Med
. 2012;2(1):a007682.PubMedGoogle ScholarCrossref
BA. Fever myths and misconceptions: the beneficial effects of fever as a critical component of host defenses against infection. Heart Lung
. 2012;41(1):99-101.PubMedGoogle ScholarCrossref
JD. Febrile core temperature is essential for optimal host defense in bacterial peritonitis. Infect Immun
. 2000;68(3):1265-1270.PubMedGoogle ScholarCrossref
N. Coxsackieviral-mediated diabetes: induction requires antigen-presenting cells and is accompanied by phagocytosis of beta cells. Clin Immunol
. 2004;110(2):134-144.PubMedGoogle ScholarCrossref
SL. Host responses in mediating sepsis and adult respiratory distress syndrome. Semin Respir Infect
. 1990;5(3):233-247.PubMedGoogle Scholar
AG. BABYDIET, a feasibility study to prevent the appearance of islet autoantibodies in relatives of patients with type 1 diabetes by delaying exposure to gluten. Diabetologia
. 2004;47(6):1130-1131.PubMedGoogle ScholarCrossref
AG. Primary dietary intervention study to reduce the risk of islet autoimmunity in children at increased risk for type 1 diabetes: the BABYDIET study. Diabetes Care
. 2011;34(6):1301-1305.PubMedGoogle ScholarCrossref
PJ; Participating Laboratories. Diabetes Antibody Standardization Program: evaluation of assays for insulin autoantibodies. Diabetologia
. 2010;53(12):2611-2620.PubMedGoogle ScholarCrossref
Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care
. 2003;26(suppl 1):S5-S20.PubMedGoogle ScholarCrossref
JB. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York, NY: Oxford University Press; 2003.
et al. An interferon-induced helicase (IFIH1
) gene polymorphism associates with different rates of progression from autoimmunity to type 1 diabetes. Diabetes
. 2011;60(2):685-690.PubMedGoogle ScholarCrossref
et al. Population-based study of acute respiratory infections in children, Greenland. Emerg Infect Dis
. 2002;8(6):586-593.PubMedGoogle ScholarCrossref
N. Helicobacter pylori
acquisition in infancy after decline of maternal passive immunity. Pediatr Res
. 1997;41(5):641-646.PubMedGoogle ScholarCrossref
U. Seroprevalence of IgG antibodies against measles, mumps and rubella in Swiss children during the first 16 months of life [in German]. Schweiz Med Wochenschr
. 2000;130(41):1479-1486.PubMedGoogle Scholar
Y. Decline of maternal hepatitis a antibodies during the first 2 years of life in infants born in Turkey. Am J Trop Med Hyg
. 2005;73(2):457-459.PubMedGoogle Scholar
V; MAS-90 Study Group. History of respiratory infections in the first 12 yr among children from a birth cohort. Pediatr Allergy Immunol
. 2008;19(6):505-512.PubMedGoogle ScholarCrossref
van der Werf
JL. Viral infections as potential triggers of type 1 diabetes. Diabetes Metab Res Rev
. 2007;23(3):169-183.PubMedGoogle ScholarCrossref
et al. Exposure to the viral by-product dsRNA or Coxsackievirus B5 triggers pancreatic beta cell apoptosis via a Bim/Mcl-1 imbalance. PLoS Pathog
. 2011;7(9):e1002267.PubMedGoogle ScholarCrossref
AK. Immunopathology of the human pancreas in type-I diabetes. Semin Immunopathol
. 2011;33(1):9-21.PubMedGoogle ScholarCrossref
et al. Enterovirus-induced gene expression profile is critical for human pancreatic islet destruction. Diabetologia
. 2012;55(12):3273-3283.PubMedGoogle ScholarCrossref
PJ. Worth the wait: type 1 diabetes prospective birth cohort studies enter adolescence. Diabetologia
. 2012;55(7):1873-1876.PubMedGoogle ScholarCrossref
E; BABYDIAB-BABYDIET Study Group. Age-related islet autoantibody incidence in offspring of patients with type 1 diabetes. Diabetologia
. 2012;55(7):1937-1943.PubMedGoogle ScholarCrossref
et al. Early seroconversion and rapidly increasing autoantibody concentrations predict prepubertal manifestation of type 1 diabetes in children at genetic risk. Diabetologia
. 2012;55(7):1926-1936.PubMedGoogle ScholarCrossref
KS. Self-reported lower respiratory tract infections and development of islet autoimmunity in children with the type 1 diabetes high-risk HLA genotype: the MIDIA study. Diabetes Metab Res Rev
. 2011;27(8):834-837.PubMedGoogle ScholarCrossref
et al. Viruses and bacteria in the etiology of the common cold. J Clin Microbiol
. 1998;36(2):539-542.PubMedGoogle Scholar
M. Detection of rhinovirus and enterovirus in upper respiratory tract samples using a multiplex nested PCR. J Virol Methods
. 2003;108(2):223-228.PubMedGoogle ScholarCrossref