Pandemic influenza and subsequent risk of type 1 diabetes: a nationwide cohort study

Pandemic influenza and subsequent risk of type 1 diabetes: a nationwide cohort study Aims/hypothesis Case reports have linked influenza infections to the development of type 1 diabetes. We investigated whether pandemic and seasonal influenza infections were associated with subsequent increased risk of type 1 diabetes. Methods In this population-based registry study, we linked individual-level data from national health registries for the entire Norwegian population under the age of 30 years for the years 2006–2014 (2.5 million individuals). Data were obtained from the National Registry (population data), the Norwegian Patient Registry (data on inpatient and outpatient specialist care), the Primary Care Database, the Norwegian Prescription Database and the Norwegian Surveillance System for Communicable Diseases. Pandemic influenza was defined as either a clinical influenza diagnosis during the main pandemic period or a laboratory-confirmed test. Seasonal influenza was defined by a clinical diagnosis of influenza between 2006 and 2014. We used Cox regression to estimate HRs for new- onset type 1 diabetes after an influenza infection, adjusted for year of birth, sex, place of birth and education. Results The adjusted HR for type 1 diabetes after pandemic influenza infection was 1.19 (95% CI 0.97, 1.46). In the subgroup with laboratory-confirmed influenza A (H1N1), influenza was associated with a twofold higher risk of subsequent type 1 diabetes before age 30 years (adjusted HR: 2.26, 95% CI 1.51, 3.38). Conclusions/interpretation Overall, we could not demonstrate a clear association between clinically reported pandemic influenza infection and incident type 1 diabetes. However, we found a twofold excess of incident diabetes in the subgroup with laboratory- confirmed pandemic influenza A (H1N1). . . . . . Keywords Incidence Infections Influenza Influenza H1N1 Register-based study Type 1 diabetes Abbreviations ICPC-2 International Classification of Primary Hanne L. Gulseth and Lars C. Stene are joint senior authors. Care, Second Edition Electronic supplementary material The online version of this article MSIS Norwegian Surveillance System for (https://doi.org/10.1007/s00125-018-4662-7) contains peer-reviewed but unedited supplementary material, which is available to authorised users. Communicable Diseases * Paz L. D. Ruiz Paz.Lopez-Doriga.Ruiz@fhi.no Introduction Department of Chronic Diseases and Ageing, Norwegian Institute of Type 1 diabetes is a chronic autoimmune disease with both Public Health, Postbox 4404, Nydalen, 0403 Oslo, Norway genetic and environmental contributions. Viruses may influ- Department of Endocrinology, Morbid Obesity and Preventive ence susceptibility and trigger autoimmunity in individuals Medicine, Oslo University Hospital, Oslo, Norway genetically predisposed to diabetes [1, 2]. Enteroviruses and Institute of Clinical Medicine, University of Oslo, Oslo, Norway other viruses have been most frequently studied in relation to type 1 diabetes [3]. Recently, respiratory virus infections have Centre for Fertility and Health, Norwegian Institute of Public Health, Oslo, Norway also been associated with the development of islet autoimmu- nity and the first manifestations of clinical symptoms of type 1 Department of Influenza, Norwegian Institute of Public Health, Oslo, Norway diabetes [4–8]. Influenza virus can infect human pancreatic Diabetologia cell lines, and can cause pancreatitis and hyperglycaemia in 2006–2014 and subsequent type 1 diabetes in a nationwide animal models [9]. Influenza A (H1N1) virus infection has register-based cohort study from Norway. also been associated with acute pancreatitis [10–12] and type 1diabetes [13, 14] in human case-report studies. Influenza spreads yearly across the continents, and when a new influenza virus emerges and transmits among humans, an Methods influenza pandemic can occur [15]. In April 2009, the World Health Organization detected an outbreak of a new influenza Participants and design In this open cohort study, we linked virus [Influenza A(H1N1)pdm09] in Mexico. This influenza, individual-level data from seven national registers with prospec- the swine flu, was declared a pandemic in June 2009, and this tively collected data for more than 2.5 million residents in lasted until August 2010 [16]. The virus has subsequently con- Norway aged 30 years and younger, and followed them from tinued to circulate as one of the seasonal influenza strains. 2006 to 30 June 2014 (Fig. 1). Two small retrospective studies have suggested that pandemic The study population consisted of all residents in Norway as influenza may be associated with type 1 diabetes [17, 18]. These registered in the National Registry. Statistics Norway provided showed a concomitant increase in type 1 diabetes during the data on education and place of birth. Further, we obtained in- pandemic influenza period among children. However, as the typ- formation on use of glucose-lowering drugs from the ical time from induction of islet autoimmunity to clinical onset of Norwegian Prescription Database, and diabetes diagnosis type 1 diabetes is several years [19], studies with longer follow- codes and dates of diagnoses from the Primary Care Database up after influenza are necessary to elucidate the role of pandemic and the Norwegian Patient Registry. Laboratory-confirmed in- influenza in development of diabetes. fluenza cases were registered in the Norwegian Surveillance No previously published studies have addressed whether System for Communicable Diseases (MSIS). (Details of each pandemic influenza diagnosis is associated with the develop- register is outlined in the electronic supplementary material ment of type 1 diabetes. We have recently reported that there [ESM].) Information in each database is registered with the was no association between vaccination against the 2009 personal identification number that is given to all Norwegian H1N1 pandemic influenza virus with the AS03 adjuvanted residents, which enables linking on an individual level. Pandemrix vaccine and type 1 diabetes [20]. In Norway, the The study was approved by the Norwegian Data Protection vaccination campaign and the main influenza wave occurred Authority (approval number 10/00910-12) and the Regional simultaneously [21, 22]. Here, we investigate a potential rela- Committee for Medical and Health Research Ethics (approval tionship between pandemic or seasonal influenza in the years number 2010/2583). Diabetologia 1 Jan 2006 Follow-up seasonal influenza Follow-up pandemic influenza 1 Jun 2009 30 Jun 2014 SYSVAK MSIS NPR: Norwegian Patient Registry ICD-10: E10 J09, J10, J11 KUHR: Primary care database ICPC-2: T89 R80 NorPD: Norwegian Prescription Database ATC: A10A Statistics Norway Education, place of birth The National Registry Demographic data of Norwegian residents 2004 2005 2014 2015 2006 2007 2009 2010 2011 2012 2013 Fig. 1 Overview of nationwide registers linked via personal identification that records an individual’s vaccination status and vaccination coverage in numbers assigned to all Norwegian residents. Laboratory-confirmed in- Norway (ICD-10, ICPC-2, the Anatomical Therapeutic Chemical). Blue fluenza data are from the MSIS register, May 2009 to April 2011. The brackets represent the seasonal influenza period each year (for 2006, data Norwegian Patient Registry collects data on individuals receiving special- were available from 1 January). ATC, Anatomical Therapeutic Chemical; ist healthcare (inpatient and outpatient) (see ESM Table 1 for codes and Jan, January; Jun, June abbreviations). SYSVAK is a national electronic immunisation registry Diagnosis of type 1 diabetes before age 30 years We defined season, from around October to mid-May each year, in the years incident cases of type 1 diabetes as registration of dispensed 2006 to 2012–2013, with the exception of the 2009–2010 sea- insulin for at least 6 months and at least one registration of a son, which was the pandemic influenza period defined as above. type 1 diabetes diagnosis from specialist or primary care (see The pandemic season was included in the seasonal analysis. ESM Table 1 for included classification codes and abbreviations). To ensure inclusion of type 1 diabetes only, Subgroups/stratification We additionally performed analyses we excluded individuals who received oral glucose-lowering separately for males and females, and for people under agents within 1 year after diagnosis. To avoid any prevalent 15 years of age. We did subgroup analysis for those registered cases of diabetes at baseline, we excluded individuals who had with a laboratory-confirmed influenza diagnosis, those regis- a diagnosis of any type of diabetes or used any glucose- tered in primary care and those registered in specialist care. lowering medication before the start of the study (Fig. 1). Covariates Information on sex, date of birth, emigration, im- Exposures The primary exposure was pandemic influenza in- migration and death were obtained from the National Registry. fection, and people were defined as having pandemic influen- Information on vaccination with Pandemrix, an AS03- za either by a diagnosis of influenza registered in the national adjuvanted influenza A(H1N1)pdm09 vaccine, was obtained primary care database (using the International Classification from the Norwegian Immunisation Register [23], and Statistics of Primary Care, Second Edition [ICPC-2] code R80) or in the Norway provided data on place of birth and education (Fig. 1). Norwegian Patient Registry (specialist care, coding according We used the highest education in year 2013 for the participant to the ICD-10: J09, J10, J11 [www.who.int/classifications/icd/ or his/her parents for all the individuals in the analyses. Place en/]) during the pandemic period or by laboratory-confirmed of birth was classified into three categories (‘Norway’, ‘Europe pandemic influenza registered in the Norwegian Surveillance except Norway’ and ‘outside Europe’). Information about sea- System for Communicable Diseases (ESM Table 1). We de- sonal influenza vaccinations or laboratory-confirmed influenza fined the pandemic influenza period as starting in June 2009 outside the pandemic was not available. and lasting until May 2010. The secondary exposure was seasonal influenza infection, Sensitivity analysis During the pandemic period (June 2009 defined as any influenza diagnosis occurring in the surveillance until May 2010) other viruses may have caused influenza-like symptoms [24]. In a sensitivity analysis we restricted to the periods for influenza in Norway. We analysed each influenza Diabetologia peak pandemic period, October 2009 to December 2009, confirmed pandemic influenza (7.4% of influenza diagnoses where there were no other influenza viruses in circulation. in primary care and 48.4% of influenza diagnoses registered in The main influenza wave and the vaccination campaign oc- specialist healthcare). When restricting analyses to those with curred at the same time in Norway, and vaccinated individuals a laboratory-confirmed pandemic influenza, there was a two- could also be registered with influenza diagnosis. In a sensi- fold higher risk of type 1 diabetes (adjusted HR 2.26, 95% CI tivity analysis we restricted to those who were not vaccinated 1.51, 3.38, Fig. 2). Those who developed type 1 diabetes after against pandemic influenza. a laboratory-confirmed influenza A infection were, on aver- age, 10 years old at the time of influenza, and were diagnosed Statistical analyses We used Cox regression with months as with type 1 diabetes on average 2.2 years later (ESM Fig. 1). the time metric and influenza as time-dependent exposure In the analysis of those with pandemic influenza diagnosed in variable to estimate HRs with 95% CI for type 1 diabetes, both specialist healthcare, the adjusted HR was 2.83, 95% CI 1.18, unadjusted and adjusted for year of birth (in groups of 3 years), 6.81 (Fig. 2; of those, four of five new cases of type 1 diabetes sex, place of birth and education. The study population was were laboratory confirmed). followed from birth, 1 year after immigration or start of follow-up (January 2006), whichever occurred last, until type Seasonal influenza and type 1 diabetes During the study pe- 1 diabetes diagnosis, emigration, death, 30 years of age or end riod for seasonal influenza, from 2006 to mid-2014, 3700 of follow-up (July 2014), whichever occurred first. individuals under 30 years of age were diagnosed with type Immigrants were included for follow-up 1 year after immigra- 1 diabetes during 15,583,847 person-years of follow-up (in- tion to avoid prevalent cases of type 1 diabetes being cluding the 2009–2010 pandemic influenza period). The num- misclassified as incident and to ensure that influenza exposure ber of individuals registered with a seasonal influenza diagno- could be registered. sis included in the analysis varied from 19,691 in the season In the model where we estimated risk of type 1 diabetes 2007–2008 to 39,179 in the 2012–2013 season (ESM Table 2 after pandemic influenza, follow-up started in June 2009. We and ESM Fig. 2). Higher risk of type 1 diabetes after a sea- performed separate analyses for each influenza season, with sonal influenza diagnosis was observed in all seasons between start of follow-up in October each year for the same calendar 2007 and 2011, but only the season 2010–2011 was statisti- year, except for the 2005–2006 season, when follow-up cally significantly associated in both the total population and started in January 2006 (the beginning of our observation those aged below 15 years (Fig. 4). period). Influenza was included as time-varying exposure for which individuals contributed with unexposed person-time Sensitivity analyses We assessed whether the pandemic influ- until their first month of influenza diagnosis, and were enza association with type 1 diabetes was different when we restricted to influenza diagnoses made during the peak pan- regarded as exposed afterwards. Data handling and analyses were done using Stata version demic period which, in Norway, occurred from 1 October 14 (Stata Statistical Software: Release 14, StataCorp, College 2009 to 31 December 2009 [25]. This restriction resulted in Station, TX, USA). a slightly smaller number of exposed individuals, but yielded very similar results (adjusted HR 1.17, 95% CI 0.95, 1.45). Second, to be able to analyse the association of the different Results seasons of influenza and have similar follow-up times after infection (for better comparison of the association with type 1 Pandemic influenza and risk of type 1 diabetes Among the diabetes across seasons), we defined the time-dependent ex- 2,286,650 individuals in the study population, 76,173 (3.3%) posure (seasonal influenza) to last for a maximum of 2 years were diagnosed with pandemic influenza. From June 2009 to after influenza diagnosis. For instance, a person exposed to December 2014, 2376 individuals (0.1%) were diagnosed with seasonal influenza diagnosis in 2006 would be regarded as new-onset type 1 diabetes (Table 1). New-onset type 1 diabetes exposed in the following 2 years, but after 2008 would con- was registered at a slightly younger age for those registered tribute unexposed person-time unless he/she developed type 1 with pandemic influenza diagnosis compared with those with- diabetes before this time). These analyses gave similar results out such registration (12.9 years compared with 13.3 years). as the main analyses (ESM Fig. 3). Pandemic influenza diagnosis was associated with an ap- In the analyses with pandemic influenza as the exposure, proximately 20% higher risk of type 1 diabetes, though this we performed an analysis where we adjusted for vaccination was not statistically significant (Fig. 2). Thecumulativeinci- with Pandemrix, and the results were not modified (adjusted dence of being diagnosed with type 1 diabetes during the HR [aHR] 1.19, 95% CI 0.97, 1.46 for all diagnosed with study period for those with and without a registered influenza pandemic influenza). Finally, when we restricted the analyses diagnosis is shown in Fig. 3. Among those diagnosed with to those who were not vaccinated against pandemic influenza, pandemic influenza was associated with an increased risk of influenza during the pandemic, 11.4% had laboratory- Diabetologia Table 1 Characteristics of the Characteristics Individuals study population aged <30 years in the pandemic analysis All With pandemic influenza With incident type 1 diabetes n = 2,286,650 n = 76,173 n =2376 Sex Male 1,169,485 (51) 37,344 (49) 1382 (58) Female 1,117,165 (49) 38,829 (51) 994 (42) Year of birth 1979–1989 713,963 (31) 25,563 (34) 339 (14) 1990–1999 665,198 (29) 24,904 (33) 934 (39) 2000–2009 628,606 (27) 25,694 (34) 1020 (43) >2010 278,883 (12) 12 (0) 83 (4) Education level ≤10 years 200,040 (9) 7838 (10) 165 (7) 11–13 years 779,682 (34) 30,289 (40) 953 (40) ≥14 years 1,214,276 (53) 37,076 (49) 1244 (52) No information 92,652 (4) 970 (1) 15 (1) Place of birth Norway 1,973,332 (86) 67,667 (89) 2260 (95) Europe (except Norway) 166,130 (7) 3582 (5) 63 (3) Outside Europe 147,188 (6) 4924 (6) 53 (2) Data shown are n (%) See Methods for details of influenza diagnosis. Note that individuals with pandemic influenza and individuals with incident type 1 diabetes were not mutually exclusive (numbers in the two columns therefore do not add to the total) The highest education level the individual achieved up to 2013 or the highest attained education level of their parents T1D cases Exp. Unexp. aHR (95% CI) Pandemic influenza 96 2280 1.19 (0.97, 1.46) 1.19 (0.97, 1.46) Male 60 1322 1.31 (1.01, 1.70) 1.31 (1.01, 1.70) Female 36 958 1.05 (0.75, 1.47) 1.05 (0.75, 1.47) Pandemic influenza, T1D <15 years 64 1443 1.25 (0.97, 1.61) 1.25 (0.97, 1.61) Lab-confirmed influenza 24 2352 2.26 (1.51, 3.38) 2.26 (1.51, 3.38) Pandemic influenza primary care 85 2291 1.12 (0.90, 1.39) 1.12 (0.90, 1.39) Pandemic influenza specialist care 5 2371 2.83 (1.18, 6.81) 2.83 (1.18, 6.81) 0.8 1.0 1.5 2.0 2.5 3.0 aHR (95% CI) Fig. 2 Association between pandemic influenza diagnosis and risk of confirmed pandemic influenza (during the pandemic period). HRs were type 1 diabetes in up to 2.28 million Norwegian residents under 30 years adjusted for year of birth, sex, place of birth, education and pandemic of age, overall and in subgroups. Incident cases of type 1 diabetes defined influenza vaccination (except analysis stratified for sex, which was ad- as registration of dispensed insulin for at least 6 months and at least one justed for year of birth, place of birth, education and pandemic influenza registration of a type 1 diabetes diagnosis from specialist or primary care. vaccination). Exp., exposed; Lab, laboratory; T1D, type 1 diabetes; Pandemic influenza was defined as a clinical diagnosis of influenza reg- Unexp., unexposed istered in the primary care database, specialist care, or a laboratory- Diabetologia 0.0020 Discussion In this nationwide study of all residents in Norway below 0.0015 30 years of age, we found a twofold higher risk of developing type 1 diabetes in individuals who had a specialist care diag- nosis or a laboratory-confirmed infection with pandemic 0.0010 influenza. This is the first study using national registries to address the 0.0005 long-term risk of type 1 diabetes after a pandemic influenza diagnosis. A main strength of the study is the large sample size and the complete inclusion of all residents in Norway. Nearly all persons with type 1 diabetes are diagnosed and treated in 0 1 2 3 4 5 the public health system in Norway, and consultations and Years from study start (1 Oct 2009) medications are free for children with type 1 diabetes until Fig. 3 Cumulative incidence and 95% CI of type 1 diabetes for pandemic 16 years of age. Dispensing of insulin registered in the influenza (blue line and grey shaded area) and for no pandemic influenza Norwegian Prescription Database is likely to detect nearly (red line and light red shaded area). Logrank test p =0.049 all cases of type 1 diabetes [26], and our algorithm for diag- type 1 diabetes (aHR 1.29, 95% CI 1.01, 1.65, based on 67 nosis of type 1 diabetes combining registers is likely to ensure cases of type 1 diabetes registered after pandemic influenza true type 1 diabetes in the vast majority of cases [27]. By among unvaccinated people). restricting the study population to those below 30 years of Fig. 4 Association between T1D cases seasonal influenza diagnosis and Season Exp. Unexp. aHR (95% CI) risk of type 1 diabetes in more than 2.5 million individuals under 2006 44 3656 1.07 (0.80, 1.45) 30 years of age (a), and under 2006/07 43 3521 1.02 (0.75, 1.38) 15 years of age (b). HRs were adjusted foryearofbirth,sex, 2007/08 30 3211 1.37 (0.95, 1.96) place of birth and education. 2008/09 29 2685 1.36 (0.94, 1.96) Seasonal influenza from 1 January 2006 to 30 June 2014 (the 2009/10 96 2280 1.19 (0.97, 1.46) pandemic season, 2009–2010, 2010/11 35 1687 1.41 (1.01, 1.97) shows the same data as in Fig. 2; 2011/12 12 1241 0.94 (0.53, 1.65) during this season, pandemic influenza was defined as an 2012/13 11 805 0.99 (0.55, 1.80) influenza registration in primary care or in specialist care or laboratory-confirmed pandemic influenza). Exp., exposed; 0.4 1.0 1.5 2.0 2.5 Unexp., unexposed aHR (95% CI) T1D cases Season Exp. Unexp. aHR (95% CI) 2006 23 2282 1.04 (0.69, 1.57) 2006/07 22 2213 1.06 (0.70, 1.61) 2007/08 15 2012 1.52 (0.92, 2.53) 2008/09 16 1686 1.60 (0.96, 2.66) 2009/10 64 1443 1.25 (0.97, 1.61) 2010/11 21 1068 1.67 (1.08, 2.57) 2011/12 7 780 1.06 (0.50, 2.22) 2012/13 5 502 0.85 (0.35, 2.06) 0.4 1.0 1.5 2.0 2.5 aHR (95% CI) Cumulative incidence of type 1 diabetes Diabetologia age, we reduced the risk of misclassifying type 2 diabetes as Previous studies of influenza and type 1 diabetes have been type 1 diabetes. However, as misclassification of diabetes may limited to a few small retrospective studies, showing a temporal have occurred in a small number of cases, we performed sen- relationship between H1N1 influenza infection and increased sitivity analyses in which we restricted our sample to those type 1 diabetes incidence [17, 18]. In one study, influenza A below 15 years of age. In this age group, such misclassifica- antibodies were not associated with initiation of islet autoim- tion is highly unlikely [27]. munity in children, but this study did not investigate pandemic The main limitation of this study is that we did not capture all influenza A (H1N1), included asymptomatic influenza infec- cases of influenza, as we only have data on those who sought tions and did not study clinical type 1 diabetes as outcome [33]. healthcare and received an influenza diagnosis. It is probable In theory, viral infections may affect the progression from islet that many people infected with influenza did not seek healthcare, autoimmunity to clinical diabetes in the small proportion of especially those with milder illness. Furthermore, males and individuals who are positive for islet autoantibodies [34]. females, or different age groups, might seek medical help fol- Any association with type 1 diabetes could in theory be due lowing different patterns. Administration of antivirals without to non-specific immunological mechanisms associated with prescription during the pandemic may also have contributed to infections. Respiratory infections in early life and type 1 dia- fewer people visiting a physician for treatment of influenza in- beteshavebeenlinked[4–8], and we can consider influenza- fection [28]. Probably those with more severe symptoms were like illness as a respiratory viral infection caused by influenza morelikelytoseekhealthcare[29], and our estimates are there- or other viruses. These, and other common viruses causing forelikelytorelatetomoresevereinfluenza infections. Infection infection with fever, could be important as cytokine inducer severity or host response could be important, as influenza diag- and T cell activators [3]. Our finding that pandemic influenza nosed in primary care did not show a clear association. A pos- diagnosis in specialist healthcare was more strongly associat- sible explanation is that this group contains misclassified cases ed with type 1 diabetes may possibly indicate that an associ- that might not have been infected with influenza, or that suscep- ation with type 1 diabetes is stronger with severe infections tible individuals had more severe symptoms. In our study pop- (needing hospitalisation or other type of specialist care). It is ulation, 3% were diagnosed with influenza during the pandemic. plausible that those with more severe illness had their infec- Likewise, a report from Sweden suggested that around 6% of the tion confirmed by laboratory test. Also, it may be that individ- population was diagnosed with H1N1 [30], an estimate probably uals with preclinical diabetes have an underlying higher risk restricted to those with symptomatic infection. It has been esti- of developing severe influenza. It is likely that most individ- matedthataround20–30% of the Norwegian population were uals with type 1 diabetes experience a short period of infected during the 2009–2010 pandemic [31]. Serum A H1N1 hyperglycaemia before clinical diagnosis of diabetes. The du- antibody positivity can occur after clinical influenza, after mild/ ration of this period is unlikely to last for much more than a quiescent non-clinical influenza or after vaccination with few months [35], a period shorter than the average time be- Pandemrix, making it difficult to estimate the true proportion tween influenza and diagnosis of type 1 diabetes in our current affected by clinical H1N1 infection in a population. A study analysis. The increased risk of diabetes ascertained from from Norway conducted in January 2010 showed influenza A laboratory-confirmed cases could also have occurred by (H1N1) serum antibody positivity in up to 65% of younger age chance. Even though we have a very large dataset, the actual groups [32]. Many of these are probably positive because of numbers of laboratory-confirmed cases were small. vaccination with Pandemrix, as around 40% of the total popula- The dominating circulating influenza types usually differ tion were vaccinated during the pandemic. by season, and it is difficult to discern from our data whether Most individuals were not tested for pandemic influenza as any specific strain tends to be more strongly associated with the capacity in laboratories was stretched, and at the point type 1 diabetes. We found the strongest evidence for pandemic when the pandemic strain was considered to be the primary influenza. However, there was also increased risk of type 1 cause of influenza-like illness, it was regarded as unnecessary diabetes after influenza in the following season (2010–2011), to test individuals in primary care when making a diagnosis. where influenza A (H1N1)pdm09 virus and influenza B co- Therefore, we do not have laboratory confirmation for the dominated (albeit with slightly more influenza B) [36]. It is majority of the pandemic influenza diagnoses in our study. also possible that the 2009 pandemic influenza strain has During the pandemic period in Norway, there were periods stronger tropism for pancreatic cells than other influenza in which non-influenza viruses may have given influenza- strains [37]. The immunological response to influenza infec- like symptoms and resulted in an influenza diagnosis. tion with different severity and with different virus strains may However, when restricting the exposure period to the pandem- differ and is not well understood [38]. More studies are needed ic peak period (October to December 2009), when no other to conclude on the role of different seasonal influenza viruses influenza strains were circulating and most individuals under in type 1 diabetes aetiology. age 30 with influenza symptoms were likely to have pandemic Our large register cohort does not include information on influenza, we found similar results. pre-diagnostic diabetes associated autoantibodies. Hence, we Diabetologia analysis or interpretation of data. PLDR wrote the first draft of the man- could not investigate whether influenza infections induced or uscript and subsequent revisions. PLDR, LCS and IJB performed the accelerated autoimmunity. Viruses could contribute to the de- statistical analyses. SEH and IJB provided administrative, technical or velopment of clinical diabetes through stress and inflamma- material support. All authors critically revised the paper for important tion in individuals with autoimmunity (non-specific effect of intellectual content and approved the final version. virus infections) [39]. Open Access This article is distributed under the terms of the Creative We could speculate that preventing viral infections, for ex- Commons Attribution 4.0 International License (http:// ample through influenza vaccination, could help reduce the creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appro- incidence of type 1 diabetes. In a recent paper from The priate credit to the original author(s) and the source, provide a link to the Environmental Determinants of Diabetes in the Young Creative Commons license, and indicate if changes were made. (TEDDY) study, the Pandemrix vaccination was associated with a lower risk of islet autoimmunity in children at increased genetic risk in Finland, whereas no difference was seen in References Sweden [40]. Unfortunately, we do not have access to data on autoimmunity in our register-based study. In line with the 1. Hyöty H (2016) Viruses in type 1 diabetes. Pediatr Diabetes 17:56– Swedish data we did not find any association with Pandemrix and type 1 diabetes in our study [20]. In Norway, many of 2. Op de Beeck A, Eizirik DL (2016) Viral infections in type 1 diabe- those who were vaccinated with Pandemrix had already been tes mellitus–why the beta cells? Nat Rev Endocrinol 12:263–273 3. Rodriguez-Calvo T, Sabouri S, Anquetil F, von Herrath MG (2016) infected, or were infected with influenza A (H1N1) after vac- The viral paradigm in type 1 diabetes: who are the main suspects? cination, but before effective protective antibodies had been Autoimmun Rev 15:964–969 induced [21]. 4. Beyerlein A, Wehweck F, Ziegler A-G, Pflueger M (2013) In conclusion, we could not demonstrate a clear association Respiratory infections in early life and the development of islet autoimmunity in children at increased type 1 diabetes risk: evidence between clinically reported pandemic influenza infection and from the BABYDIET study. JAMA Pediatr 167:800–807 incident type 1 diabetes in this register-based cohort study. We 5. Rasmussen T, Witsø E, Tapia G, Stene LC, Rønningen KS (2011) did, however, find a twofold excess of incident diabetes in the Self-reported lower respiratory tract infections and development of subgroups with laboratory-confirmed pandemic influenza A islet autoimmunity in children with the type 1 diabetes high-risk (H1N1) or pandemic influenza diagnosed in specialist HLA genotype: the MIDIA study. Diabetes Metab Res Rev 27: 834–837 healthcare. This suggests that respiratory infections may play 6. Beyerlein A, Donnachie E, Jergens S, Ziegler A (2016) Infections a role in the aetiology of type 1 diabetes, but more studies are in early life and development of type 1 diabetes. JAMA 315:1899– warranted. 7. Lonnrot M, Lynch KF, Elding Larsson H et al (2017) Respiratory infections are temporally associated with initiation of type 1 diabe- Acknowledgements Data from the Norwegian Patient Register have tes autoimmunity: the TEDDY study. Diabetologia 60:1931–1940 been used in this publication. The interpretation and reporting of these 8. Lee H-Y, Lu C-L, Chen H-F, Su H-F, Li C-Y (2015) Perinatal and data are the sole responsibility of the authors, and no endorsement by the childhood risk factors for early-onset type 1 diabetes: a population- Norwegian patient register is intended nor should be inferred. based case-control study in Taiwan. Eur J Pub Health 25:1024– A part of the study in abstract form was presented as a poster at the 53rd EASD Annual Meeting in Lisbon, 11–15 September 2017. 9. Capua I, Mercalli A, Pizzuto MS et al (2013) Influenza A viruses grow in human pancreatic cells and cause pancreatitis and diabetes Data availability The datasets generated during and/or analysed during in an animal model. J Virol 87:597–610 the current study are not publicly available because of data protection 10. Blum A, Podvitzky O, Shalabi R, Simsolo C (2010) Acute pancre- regulations. atitis may be caused by H1N1 influenza A virus infection. Isr Med Assoc J 12:640–641 Funding This research was partly supported by the Norwegian Institute 11. Habib A, Jain A, Singh B, Jamshed N (2016) H1N1 influenza of Public Health and a grant from the South-Eastern Norway Regional presenting as severe acute pancreatitis and multiorgan dysfunction. Health Authority. Am J Emerg Med 34:1911.e1911–1911.e1912 The sponsors had no role in the: design and conduct of the study; collection, management, analysis and interpretation of the data; prepara- 12. Baran B, Karaca C, Soyer OM et al (2012) Acute pancreatitis as- tion, review or approval of the manuscript; or decision to submit the sociated with H1N1 influenza during 2009 pandemic: a case report. manuscript for publication. Clin Res Hepatol Gastroenterol 36:e69–e70 13. Watanabe N (2011) Conversion to type 1 diabetes after H1N1 in- Duality of interest HLG has received grants for lectures or consulting fluenza infection: a case report. J Diabetes 3:103 from AstraZeneca, Novo Nordisk, Sanofi, Boehringer Ingelheim and 14. Larcombe PJ, Moloney SE, Schmidt PA (2011) Pandemic (H1N1) Merck Sharp & Dohme. All other authors declare that there is no duality 2009: a clinical spectrum in the general paediatric population. Arch of interest associated with their contribution to this manuscript. Dis Child 96:96–98 15. World Health Organization (2014) Influenza virus infections in Contribution statement LCS, PLDR and IJB had full access to all of the humans [Internet]. Available from http://www.who.int/influenza/ data in the study and take responsibility for the integrity of the data and human_animal_interface/virology_laboratories_and_vaccines/ the accuracy of the data analysis. LCS, HLG and SEH contributed to the influenza_virus_infections_humans_feb14.pdf?ua=1. Accessed 9 study concept and design. All authors contributed to the acquisition, Oct 2017 Diabetologia 16. DG Statement following the meeting of the Emergency Committee, 29. Simonsen KA, Hunskaar S, Sandvik H, Rortveit G (2013) Capacity and adaptations of general practice during an influenza pandemic. WHO Available from http://www.who.int/csr/disease/swineflu/ 4th_meeting_ihr/en/. Accessed 9 Oct 2017 PLoS One 8:e69408 17. Valdes C, Unanue N, Hernandez M et al (2013) Is there a link 30. Swedish WHO National Influenza Centre (2011) The influenza between influenza and type I diabetes? Increased incidence of A(H1N1)2009 pandemic in Sweden, 2009–2010. Available from TID during the pandemic H1N1 influenza of 2009 in Chile. https://www.folkhalsomyndigheten.se/pagefiles/15030/influenza- Pediatr Endocrinol Rev 11:161–166 in-sweden-2009-2010.pdf. Accessed 9 Oct 2017 18. Nenna R, Papoff P, Moretti C et al (2011) Detection of respiratory 31. (2010) Rapport. Ny influensa A (H1N1) 2009. Gjennomgang av viruses in the 2009 winter season in Rome: 2009 influenza A erfaringene i Norge. Direktoratet for samfunnssikkerhet og (H1N1) complications in children and concomitant type 1 diabetes beredskap (DSB), Oslo [In Norwegian]. Available from http:// onset. Int J Immunopathol Pharmacol 24:651–659 www.dsbinfo.no/DSBno/2010/Rapport/Pandemirapport/. 19. Ziegler AG, Rewers M, Simell O et al (2013) Seroconversion to Accessed 9 Oct 2017 multiple islet autoantibodies and risk of progression to diabetes in 32. Waalen K, Kilander A, Dudman S, Krogh G, Aune T, Hungnes O children. JAMA 309:2473–2479 (2010) High prevalence of antibodies to the 2009 pandemic influ- 20. Ruiz PLD, Stene LC, Gulseth HL et al (2018) Pandemic influenza enza A (H1N1) virus in the Norwegian population following a A H1N1 vaccination and subsequent risk of type 1 diabetes in major epidemic and a large vaccination campaign in autumn Norway. Epidemiology 29:e6-e8 2009. Euro Surveill 15:19633 21. Van Effelterre T, Dos Santos G, Shinde V (2016) Twin peaks: A/ 33. Kondrashova A, Nurminen N, Patrikainen M et al (2015) Influenza H1N1 pandemic influenza virus infection and vaccination in A virus antibodies show no association with pancreatic islet auto- Norway, 2009–2010. PLoS One 11:e0151575 antibodies in children genetically predisposed to type 1 diabetes. 22. Freiesleben de Blasio B, Iversen BG, Scalia Tomba G, Barry MA Diabetologia 58:2592–2595 (2012) Effect of vaccines and antivirals during the major 2009 A 34. Christen U, Bender C, von Herrath MG (2012) Infection as a cause (H1N1) pandemic wave in Norway–and the influence of vaccina- of type 1 diabetes? Curr Opin Rheumatol 24:417 tion timing. PLoS One 7:e30018 35. Stene LC, Barriga K, Hoffman M et al (2006) Normal but increas- 23. Trogstad L, Ung G, Hagerup-Jenssen M, Cappelen I, Haugen IL, ing hemoglobin A1c levels predict progression from islet autoim- Feiring B (2012) The Norwegian immunisation register— munity to overt type 1 diabetes: Diabetes Autoimmunity Study in SYSVAK. Euro Surveill 17:pii:20147 the Young (DAISY). Pediatr Diabetes 7:247–253 24. Anestad G, Nordbo SA (2011) Virus interference. Did rhinoviruses 36. Cuesta JG, Aavitsland P, Englund H et al (2016) Pandemic vacci- activity hamper the progress of the 2009 influenza A (H1N1) pan- nation strategies and influenza severe outcomes during the influen- demic in Norway? Med Hypotheses 77:1132–1134 za A (H1N1) pdm09 pandemic and the post-pandemic influenza 25. Håberg SE, Trogstad L, Gunnes N et al (2013) Risk of fetal death season: the Nordic experience. Euro Surveill 21 https://doi.org/10. after pandemic influenza virus infection or vaccination. N Engl J 2807/1560-7917.ES.2016.21.16.30208 Med 368:333–340 37. Qi Z, Hu H, Wang Z et al (2018) Antibodies against H1N1 influ- 26. Furu K (2008) Establishment of the nationwide Norwegian enza virus cross-react with alpha-cells of pancreatic islets. Prescription Database (NorPD)—new opportunities for research JDiabetes Investig 9:265–269 in pharmacoepidemiology in Norway. Norsk Epidemiologi 18 38. Monsalvo AC, Batalle JP, Lopez MF et al (2011) Severe pandemic https://doi.org/10.5324/nje.v18i2.23 2009 H1N1 influenza disease due to pathogenic immune com- 27. Skrivarhaug T, Stene LC, Drivvoll AK, Strøm H, Joner G, Group plexes. Nat Med 17:195–199 NCDS (2014) Incidence of type 1 diabetes in Norway among chil- dren aged 0–14 years between 1989 and 2012: has the incidence 39. Richardson SJ, Horwitz MS (2014) Is type 1 diabetes “going viral”? stopped rising? Results from the Norwegian Childhood Diabetes Diabetes 63:2203–2205 Registry. Diabetologia 57:57–62 40. Elding Larsson H, Lynch KF, Lonnrot M et al (2018) Pandemrix(R) 28. Simonsen KA, Hunskaar S, Wensaas K-A et al (2012) Influenza- vaccination is not associated with increased risk of islet autoimmu- like illness in Norway: clinical course, attitudes towards vaccination nity or type 1 diabetes in the TEDDY study children. Diabetologia and preventive measures during the 2009 pandemic. Fam Pract 29: 61:193–202 139–146 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diabetologia Springer Journals

Pandemic influenza and subsequent risk of type 1 diabetes: a nationwide cohort study

Free
9 pages

Loading next page...
 
/lp/springer-journals/pandemic-influenza-and-subsequent-risk-of-type-1-diabetes-a-nationwide-pKK8jEhiEv
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Internal Medicine; Metabolic Diseases; Human Physiology
ISSN
0012-186X
eISSN
1432-0428
D.O.I.
10.1007/s00125-018-4662-7
Publisher site
See Article on Publisher Site

Abstract

Aims/hypothesis Case reports have linked influenza infections to the development of type 1 diabetes. We investigated whether pandemic and seasonal influenza infections were associated with subsequent increased risk of type 1 diabetes. Methods In this population-based registry study, we linked individual-level data from national health registries for the entire Norwegian population under the age of 30 years for the years 2006–2014 (2.5 million individuals). Data were obtained from the National Registry (population data), the Norwegian Patient Registry (data on inpatient and outpatient specialist care), the Primary Care Database, the Norwegian Prescription Database and the Norwegian Surveillance System for Communicable Diseases. Pandemic influenza was defined as either a clinical influenza diagnosis during the main pandemic period or a laboratory-confirmed test. Seasonal influenza was defined by a clinical diagnosis of influenza between 2006 and 2014. We used Cox regression to estimate HRs for new- onset type 1 diabetes after an influenza infection, adjusted for year of birth, sex, place of birth and education. Results The adjusted HR for type 1 diabetes after pandemic influenza infection was 1.19 (95% CI 0.97, 1.46). In the subgroup with laboratory-confirmed influenza A (H1N1), influenza was associated with a twofold higher risk of subsequent type 1 diabetes before age 30 years (adjusted HR: 2.26, 95% CI 1.51, 3.38). Conclusions/interpretation Overall, we could not demonstrate a clear association between clinically reported pandemic influenza infection and incident type 1 diabetes. However, we found a twofold excess of incident diabetes in the subgroup with laboratory- confirmed pandemic influenza A (H1N1). . . . . . Keywords Incidence Infections Influenza Influenza H1N1 Register-based study Type 1 diabetes Abbreviations ICPC-2 International Classification of Primary Hanne L. Gulseth and Lars C. Stene are joint senior authors. Care, Second Edition Electronic supplementary material The online version of this article MSIS Norwegian Surveillance System for (https://doi.org/10.1007/s00125-018-4662-7) contains peer-reviewed but unedited supplementary material, which is available to authorised users. Communicable Diseases * Paz L. D. Ruiz Paz.Lopez-Doriga.Ruiz@fhi.no Introduction Department of Chronic Diseases and Ageing, Norwegian Institute of Type 1 diabetes is a chronic autoimmune disease with both Public Health, Postbox 4404, Nydalen, 0403 Oslo, Norway genetic and environmental contributions. Viruses may influ- Department of Endocrinology, Morbid Obesity and Preventive ence susceptibility and trigger autoimmunity in individuals Medicine, Oslo University Hospital, Oslo, Norway genetically predisposed to diabetes [1, 2]. Enteroviruses and Institute of Clinical Medicine, University of Oslo, Oslo, Norway other viruses have been most frequently studied in relation to type 1 diabetes [3]. Recently, respiratory virus infections have Centre for Fertility and Health, Norwegian Institute of Public Health, Oslo, Norway also been associated with the development of islet autoimmu- nity and the first manifestations of clinical symptoms of type 1 Department of Influenza, Norwegian Institute of Public Health, Oslo, Norway diabetes [4–8]. Influenza virus can infect human pancreatic Diabetologia cell lines, and can cause pancreatitis and hyperglycaemia in 2006–2014 and subsequent type 1 diabetes in a nationwide animal models [9]. Influenza A (H1N1) virus infection has register-based cohort study from Norway. also been associated with acute pancreatitis [10–12] and type 1diabetes [13, 14] in human case-report studies. Influenza spreads yearly across the continents, and when a new influenza virus emerges and transmits among humans, an Methods influenza pandemic can occur [15]. In April 2009, the World Health Organization detected an outbreak of a new influenza Participants and design In this open cohort study, we linked virus [Influenza A(H1N1)pdm09] in Mexico. This influenza, individual-level data from seven national registers with prospec- the swine flu, was declared a pandemic in June 2009, and this tively collected data for more than 2.5 million residents in lasted until August 2010 [16]. The virus has subsequently con- Norway aged 30 years and younger, and followed them from tinued to circulate as one of the seasonal influenza strains. 2006 to 30 June 2014 (Fig. 1). Two small retrospective studies have suggested that pandemic The study population consisted of all residents in Norway as influenza may be associated with type 1 diabetes [17, 18]. These registered in the National Registry. Statistics Norway provided showed a concomitant increase in type 1 diabetes during the data on education and place of birth. Further, we obtained in- pandemic influenza period among children. However, as the typ- formation on use of glucose-lowering drugs from the ical time from induction of islet autoimmunity to clinical onset of Norwegian Prescription Database, and diabetes diagnosis type 1 diabetes is several years [19], studies with longer follow- codes and dates of diagnoses from the Primary Care Database up after influenza are necessary to elucidate the role of pandemic and the Norwegian Patient Registry. Laboratory-confirmed in- influenza in development of diabetes. fluenza cases were registered in the Norwegian Surveillance No previously published studies have addressed whether System for Communicable Diseases (MSIS). (Details of each pandemic influenza diagnosis is associated with the develop- register is outlined in the electronic supplementary material ment of type 1 diabetes. We have recently reported that there [ESM].) Information in each database is registered with the was no association between vaccination against the 2009 personal identification number that is given to all Norwegian H1N1 pandemic influenza virus with the AS03 adjuvanted residents, which enables linking on an individual level. Pandemrix vaccine and type 1 diabetes [20]. In Norway, the The study was approved by the Norwegian Data Protection vaccination campaign and the main influenza wave occurred Authority (approval number 10/00910-12) and the Regional simultaneously [21, 22]. Here, we investigate a potential rela- Committee for Medical and Health Research Ethics (approval tionship between pandemic or seasonal influenza in the years number 2010/2583). Diabetologia 1 Jan 2006 Follow-up seasonal influenza Follow-up pandemic influenza 1 Jun 2009 30 Jun 2014 SYSVAK MSIS NPR: Norwegian Patient Registry ICD-10: E10 J09, J10, J11 KUHR: Primary care database ICPC-2: T89 R80 NorPD: Norwegian Prescription Database ATC: A10A Statistics Norway Education, place of birth The National Registry Demographic data of Norwegian residents 2004 2005 2014 2015 2006 2007 2009 2010 2011 2012 2013 Fig. 1 Overview of nationwide registers linked via personal identification that records an individual’s vaccination status and vaccination coverage in numbers assigned to all Norwegian residents. Laboratory-confirmed in- Norway (ICD-10, ICPC-2, the Anatomical Therapeutic Chemical). Blue fluenza data are from the MSIS register, May 2009 to April 2011. The brackets represent the seasonal influenza period each year (for 2006, data Norwegian Patient Registry collects data on individuals receiving special- were available from 1 January). ATC, Anatomical Therapeutic Chemical; ist healthcare (inpatient and outpatient) (see ESM Table 1 for codes and Jan, January; Jun, June abbreviations). SYSVAK is a national electronic immunisation registry Diagnosis of type 1 diabetes before age 30 years We defined season, from around October to mid-May each year, in the years incident cases of type 1 diabetes as registration of dispensed 2006 to 2012–2013, with the exception of the 2009–2010 sea- insulin for at least 6 months and at least one registration of a son, which was the pandemic influenza period defined as above. type 1 diabetes diagnosis from specialist or primary care (see The pandemic season was included in the seasonal analysis. ESM Table 1 for included classification codes and abbreviations). To ensure inclusion of type 1 diabetes only, Subgroups/stratification We additionally performed analyses we excluded individuals who received oral glucose-lowering separately for males and females, and for people under agents within 1 year after diagnosis. To avoid any prevalent 15 years of age. We did subgroup analysis for those registered cases of diabetes at baseline, we excluded individuals who had with a laboratory-confirmed influenza diagnosis, those regis- a diagnosis of any type of diabetes or used any glucose- tered in primary care and those registered in specialist care. lowering medication before the start of the study (Fig. 1). Covariates Information on sex, date of birth, emigration, im- Exposures The primary exposure was pandemic influenza in- migration and death were obtained from the National Registry. fection, and people were defined as having pandemic influen- Information on vaccination with Pandemrix, an AS03- za either by a diagnosis of influenza registered in the national adjuvanted influenza A(H1N1)pdm09 vaccine, was obtained primary care database (using the International Classification from the Norwegian Immunisation Register [23], and Statistics of Primary Care, Second Edition [ICPC-2] code R80) or in the Norway provided data on place of birth and education (Fig. 1). Norwegian Patient Registry (specialist care, coding according We used the highest education in year 2013 for the participant to the ICD-10: J09, J10, J11 [www.who.int/classifications/icd/ or his/her parents for all the individuals in the analyses. Place en/]) during the pandemic period or by laboratory-confirmed of birth was classified into three categories (‘Norway’, ‘Europe pandemic influenza registered in the Norwegian Surveillance except Norway’ and ‘outside Europe’). Information about sea- System for Communicable Diseases (ESM Table 1). We de- sonal influenza vaccinations or laboratory-confirmed influenza fined the pandemic influenza period as starting in June 2009 outside the pandemic was not available. and lasting until May 2010. The secondary exposure was seasonal influenza infection, Sensitivity analysis During the pandemic period (June 2009 defined as any influenza diagnosis occurring in the surveillance until May 2010) other viruses may have caused influenza-like symptoms [24]. In a sensitivity analysis we restricted to the periods for influenza in Norway. We analysed each influenza Diabetologia peak pandemic period, October 2009 to December 2009, confirmed pandemic influenza (7.4% of influenza diagnoses where there were no other influenza viruses in circulation. in primary care and 48.4% of influenza diagnoses registered in The main influenza wave and the vaccination campaign oc- specialist healthcare). When restricting analyses to those with curred at the same time in Norway, and vaccinated individuals a laboratory-confirmed pandemic influenza, there was a two- could also be registered with influenza diagnosis. In a sensi- fold higher risk of type 1 diabetes (adjusted HR 2.26, 95% CI tivity analysis we restricted to those who were not vaccinated 1.51, 3.38, Fig. 2). Those who developed type 1 diabetes after against pandemic influenza. a laboratory-confirmed influenza A infection were, on aver- age, 10 years old at the time of influenza, and were diagnosed Statistical analyses We used Cox regression with months as with type 1 diabetes on average 2.2 years later (ESM Fig. 1). the time metric and influenza as time-dependent exposure In the analysis of those with pandemic influenza diagnosed in variable to estimate HRs with 95% CI for type 1 diabetes, both specialist healthcare, the adjusted HR was 2.83, 95% CI 1.18, unadjusted and adjusted for year of birth (in groups of 3 years), 6.81 (Fig. 2; of those, four of five new cases of type 1 diabetes sex, place of birth and education. The study population was were laboratory confirmed). followed from birth, 1 year after immigration or start of follow-up (January 2006), whichever occurred last, until type Seasonal influenza and type 1 diabetes During the study pe- 1 diabetes diagnosis, emigration, death, 30 years of age or end riod for seasonal influenza, from 2006 to mid-2014, 3700 of follow-up (July 2014), whichever occurred first. individuals under 30 years of age were diagnosed with type Immigrants were included for follow-up 1 year after immigra- 1 diabetes during 15,583,847 person-years of follow-up (in- tion to avoid prevalent cases of type 1 diabetes being cluding the 2009–2010 pandemic influenza period). The num- misclassified as incident and to ensure that influenza exposure ber of individuals registered with a seasonal influenza diagno- could be registered. sis included in the analysis varied from 19,691 in the season In the model where we estimated risk of type 1 diabetes 2007–2008 to 39,179 in the 2012–2013 season (ESM Table 2 after pandemic influenza, follow-up started in June 2009. We and ESM Fig. 2). Higher risk of type 1 diabetes after a sea- performed separate analyses for each influenza season, with sonal influenza diagnosis was observed in all seasons between start of follow-up in October each year for the same calendar 2007 and 2011, but only the season 2010–2011 was statisti- year, except for the 2005–2006 season, when follow-up cally significantly associated in both the total population and started in January 2006 (the beginning of our observation those aged below 15 years (Fig. 4). period). Influenza was included as time-varying exposure for which individuals contributed with unexposed person-time Sensitivity analyses We assessed whether the pandemic influ- until their first month of influenza diagnosis, and were enza association with type 1 diabetes was different when we restricted to influenza diagnoses made during the peak pan- regarded as exposed afterwards. Data handling and analyses were done using Stata version demic period which, in Norway, occurred from 1 October 14 (Stata Statistical Software: Release 14, StataCorp, College 2009 to 31 December 2009 [25]. This restriction resulted in Station, TX, USA). a slightly smaller number of exposed individuals, but yielded very similar results (adjusted HR 1.17, 95% CI 0.95, 1.45). Second, to be able to analyse the association of the different Results seasons of influenza and have similar follow-up times after infection (for better comparison of the association with type 1 Pandemic influenza and risk of type 1 diabetes Among the diabetes across seasons), we defined the time-dependent ex- 2,286,650 individuals in the study population, 76,173 (3.3%) posure (seasonal influenza) to last for a maximum of 2 years were diagnosed with pandemic influenza. From June 2009 to after influenza diagnosis. For instance, a person exposed to December 2014, 2376 individuals (0.1%) were diagnosed with seasonal influenza diagnosis in 2006 would be regarded as new-onset type 1 diabetes (Table 1). New-onset type 1 diabetes exposed in the following 2 years, but after 2008 would con- was registered at a slightly younger age for those registered tribute unexposed person-time unless he/she developed type 1 with pandemic influenza diagnosis compared with those with- diabetes before this time). These analyses gave similar results out such registration (12.9 years compared with 13.3 years). as the main analyses (ESM Fig. 3). Pandemic influenza diagnosis was associated with an ap- In the analyses with pandemic influenza as the exposure, proximately 20% higher risk of type 1 diabetes, though this we performed an analysis where we adjusted for vaccination was not statistically significant (Fig. 2). Thecumulativeinci- with Pandemrix, and the results were not modified (adjusted dence of being diagnosed with type 1 diabetes during the HR [aHR] 1.19, 95% CI 0.97, 1.46 for all diagnosed with study period for those with and without a registered influenza pandemic influenza). Finally, when we restricted the analyses diagnosis is shown in Fig. 3. Among those diagnosed with to those who were not vaccinated against pandemic influenza, pandemic influenza was associated with an increased risk of influenza during the pandemic, 11.4% had laboratory- Diabetologia Table 1 Characteristics of the Characteristics Individuals study population aged <30 years in the pandemic analysis All With pandemic influenza With incident type 1 diabetes n = 2,286,650 n = 76,173 n =2376 Sex Male 1,169,485 (51) 37,344 (49) 1382 (58) Female 1,117,165 (49) 38,829 (51) 994 (42) Year of birth 1979–1989 713,963 (31) 25,563 (34) 339 (14) 1990–1999 665,198 (29) 24,904 (33) 934 (39) 2000–2009 628,606 (27) 25,694 (34) 1020 (43) >2010 278,883 (12) 12 (0) 83 (4) Education level ≤10 years 200,040 (9) 7838 (10) 165 (7) 11–13 years 779,682 (34) 30,289 (40) 953 (40) ≥14 years 1,214,276 (53) 37,076 (49) 1244 (52) No information 92,652 (4) 970 (1) 15 (1) Place of birth Norway 1,973,332 (86) 67,667 (89) 2260 (95) Europe (except Norway) 166,130 (7) 3582 (5) 63 (3) Outside Europe 147,188 (6) 4924 (6) 53 (2) Data shown are n (%) See Methods for details of influenza diagnosis. Note that individuals with pandemic influenza and individuals with incident type 1 diabetes were not mutually exclusive (numbers in the two columns therefore do not add to the total) The highest education level the individual achieved up to 2013 or the highest attained education level of their parents T1D cases Exp. Unexp. aHR (95% CI) Pandemic influenza 96 2280 1.19 (0.97, 1.46) 1.19 (0.97, 1.46) Male 60 1322 1.31 (1.01, 1.70) 1.31 (1.01, 1.70) Female 36 958 1.05 (0.75, 1.47) 1.05 (0.75, 1.47) Pandemic influenza, T1D <15 years 64 1443 1.25 (0.97, 1.61) 1.25 (0.97, 1.61) Lab-confirmed influenza 24 2352 2.26 (1.51, 3.38) 2.26 (1.51, 3.38) Pandemic influenza primary care 85 2291 1.12 (0.90, 1.39) 1.12 (0.90, 1.39) Pandemic influenza specialist care 5 2371 2.83 (1.18, 6.81) 2.83 (1.18, 6.81) 0.8 1.0 1.5 2.0 2.5 3.0 aHR (95% CI) Fig. 2 Association between pandemic influenza diagnosis and risk of confirmed pandemic influenza (during the pandemic period). HRs were type 1 diabetes in up to 2.28 million Norwegian residents under 30 years adjusted for year of birth, sex, place of birth, education and pandemic of age, overall and in subgroups. Incident cases of type 1 diabetes defined influenza vaccination (except analysis stratified for sex, which was ad- as registration of dispensed insulin for at least 6 months and at least one justed for year of birth, place of birth, education and pandemic influenza registration of a type 1 diabetes diagnosis from specialist or primary care. vaccination). Exp., exposed; Lab, laboratory; T1D, type 1 diabetes; Pandemic influenza was defined as a clinical diagnosis of influenza reg- Unexp., unexposed istered in the primary care database, specialist care, or a laboratory- Diabetologia 0.0020 Discussion In this nationwide study of all residents in Norway below 0.0015 30 years of age, we found a twofold higher risk of developing type 1 diabetes in individuals who had a specialist care diag- nosis or a laboratory-confirmed infection with pandemic 0.0010 influenza. This is the first study using national registries to address the 0.0005 long-term risk of type 1 diabetes after a pandemic influenza diagnosis. A main strength of the study is the large sample size and the complete inclusion of all residents in Norway. Nearly all persons with type 1 diabetes are diagnosed and treated in 0 1 2 3 4 5 the public health system in Norway, and consultations and Years from study start (1 Oct 2009) medications are free for children with type 1 diabetes until Fig. 3 Cumulative incidence and 95% CI of type 1 diabetes for pandemic 16 years of age. Dispensing of insulin registered in the influenza (blue line and grey shaded area) and for no pandemic influenza Norwegian Prescription Database is likely to detect nearly (red line and light red shaded area). Logrank test p =0.049 all cases of type 1 diabetes [26], and our algorithm for diag- type 1 diabetes (aHR 1.29, 95% CI 1.01, 1.65, based on 67 nosis of type 1 diabetes combining registers is likely to ensure cases of type 1 diabetes registered after pandemic influenza true type 1 diabetes in the vast majority of cases [27]. By among unvaccinated people). restricting the study population to those below 30 years of Fig. 4 Association between T1D cases seasonal influenza diagnosis and Season Exp. Unexp. aHR (95% CI) risk of type 1 diabetes in more than 2.5 million individuals under 2006 44 3656 1.07 (0.80, 1.45) 30 years of age (a), and under 2006/07 43 3521 1.02 (0.75, 1.38) 15 years of age (b). HRs were adjusted foryearofbirth,sex, 2007/08 30 3211 1.37 (0.95, 1.96) place of birth and education. 2008/09 29 2685 1.36 (0.94, 1.96) Seasonal influenza from 1 January 2006 to 30 June 2014 (the 2009/10 96 2280 1.19 (0.97, 1.46) pandemic season, 2009–2010, 2010/11 35 1687 1.41 (1.01, 1.97) shows the same data as in Fig. 2; 2011/12 12 1241 0.94 (0.53, 1.65) during this season, pandemic influenza was defined as an 2012/13 11 805 0.99 (0.55, 1.80) influenza registration in primary care or in specialist care or laboratory-confirmed pandemic influenza). Exp., exposed; 0.4 1.0 1.5 2.0 2.5 Unexp., unexposed aHR (95% CI) T1D cases Season Exp. Unexp. aHR (95% CI) 2006 23 2282 1.04 (0.69, 1.57) 2006/07 22 2213 1.06 (0.70, 1.61) 2007/08 15 2012 1.52 (0.92, 2.53) 2008/09 16 1686 1.60 (0.96, 2.66) 2009/10 64 1443 1.25 (0.97, 1.61) 2010/11 21 1068 1.67 (1.08, 2.57) 2011/12 7 780 1.06 (0.50, 2.22) 2012/13 5 502 0.85 (0.35, 2.06) 0.4 1.0 1.5 2.0 2.5 aHR (95% CI) Cumulative incidence of type 1 diabetes Diabetologia age, we reduced the risk of misclassifying type 2 diabetes as Previous studies of influenza and type 1 diabetes have been type 1 diabetes. However, as misclassification of diabetes may limited to a few small retrospective studies, showing a temporal have occurred in a small number of cases, we performed sen- relationship between H1N1 influenza infection and increased sitivity analyses in which we restricted our sample to those type 1 diabetes incidence [17, 18]. In one study, influenza A below 15 years of age. In this age group, such misclassifica- antibodies were not associated with initiation of islet autoim- tion is highly unlikely [27]. munity in children, but this study did not investigate pandemic The main limitation of this study is that we did not capture all influenza A (H1N1), included asymptomatic influenza infec- cases of influenza, as we only have data on those who sought tions and did not study clinical type 1 diabetes as outcome [33]. healthcare and received an influenza diagnosis. It is probable In theory, viral infections may affect the progression from islet that many people infected with influenza did not seek healthcare, autoimmunity to clinical diabetes in the small proportion of especially those with milder illness. Furthermore, males and individuals who are positive for islet autoantibodies [34]. females, or different age groups, might seek medical help fol- Any association with type 1 diabetes could in theory be due lowing different patterns. Administration of antivirals without to non-specific immunological mechanisms associated with prescription during the pandemic may also have contributed to infections. Respiratory infections in early life and type 1 dia- fewer people visiting a physician for treatment of influenza in- beteshavebeenlinked[4–8], and we can consider influenza- fection [28]. Probably those with more severe symptoms were like illness as a respiratory viral infection caused by influenza morelikelytoseekhealthcare[29], and our estimates are there- or other viruses. These, and other common viruses causing forelikelytorelatetomoresevereinfluenza infections. Infection infection with fever, could be important as cytokine inducer severity or host response could be important, as influenza diag- and T cell activators [3]. Our finding that pandemic influenza nosed in primary care did not show a clear association. A pos- diagnosis in specialist healthcare was more strongly associat- sible explanation is that this group contains misclassified cases ed with type 1 diabetes may possibly indicate that an associ- that might not have been infected with influenza, or that suscep- ation with type 1 diabetes is stronger with severe infections tible individuals had more severe symptoms. In our study pop- (needing hospitalisation or other type of specialist care). It is ulation, 3% were diagnosed with influenza during the pandemic. plausible that those with more severe illness had their infec- Likewise, a report from Sweden suggested that around 6% of the tion confirmed by laboratory test. Also, it may be that individ- population was diagnosed with H1N1 [30], an estimate probably uals with preclinical diabetes have an underlying higher risk restricted to those with symptomatic infection. It has been esti- of developing severe influenza. It is likely that most individ- matedthataround20–30% of the Norwegian population were uals with type 1 diabetes experience a short period of infected during the 2009–2010 pandemic [31]. Serum A H1N1 hyperglycaemia before clinical diagnosis of diabetes. The du- antibody positivity can occur after clinical influenza, after mild/ ration of this period is unlikely to last for much more than a quiescent non-clinical influenza or after vaccination with few months [35], a period shorter than the average time be- Pandemrix, making it difficult to estimate the true proportion tween influenza and diagnosis of type 1 diabetes in our current affected by clinical H1N1 infection in a population. A study analysis. The increased risk of diabetes ascertained from from Norway conducted in January 2010 showed influenza A laboratory-confirmed cases could also have occurred by (H1N1) serum antibody positivity in up to 65% of younger age chance. Even though we have a very large dataset, the actual groups [32]. Many of these are probably positive because of numbers of laboratory-confirmed cases were small. vaccination with Pandemrix, as around 40% of the total popula- The dominating circulating influenza types usually differ tion were vaccinated during the pandemic. by season, and it is difficult to discern from our data whether Most individuals were not tested for pandemic influenza as any specific strain tends to be more strongly associated with the capacity in laboratories was stretched, and at the point type 1 diabetes. We found the strongest evidence for pandemic when the pandemic strain was considered to be the primary influenza. However, there was also increased risk of type 1 cause of influenza-like illness, it was regarded as unnecessary diabetes after influenza in the following season (2010–2011), to test individuals in primary care when making a diagnosis. where influenza A (H1N1)pdm09 virus and influenza B co- Therefore, we do not have laboratory confirmation for the dominated (albeit with slightly more influenza B) [36]. It is majority of the pandemic influenza diagnoses in our study. also possible that the 2009 pandemic influenza strain has During the pandemic period in Norway, there were periods stronger tropism for pancreatic cells than other influenza in which non-influenza viruses may have given influenza- strains [37]. The immunological response to influenza infec- like symptoms and resulted in an influenza diagnosis. tion with different severity and with different virus strains may However, when restricting the exposure period to the pandem- differ and is not well understood [38]. More studies are needed ic peak period (October to December 2009), when no other to conclude on the role of different seasonal influenza viruses influenza strains were circulating and most individuals under in type 1 diabetes aetiology. age 30 with influenza symptoms were likely to have pandemic Our large register cohort does not include information on influenza, we found similar results. pre-diagnostic diabetes associated autoantibodies. Hence, we Diabetologia analysis or interpretation of data. PLDR wrote the first draft of the man- could not investigate whether influenza infections induced or uscript and subsequent revisions. PLDR, LCS and IJB performed the accelerated autoimmunity. Viruses could contribute to the de- statistical analyses. SEH and IJB provided administrative, technical or velopment of clinical diabetes through stress and inflamma- material support. All authors critically revised the paper for important tion in individuals with autoimmunity (non-specific effect of intellectual content and approved the final version. virus infections) [39]. Open Access This article is distributed under the terms of the Creative We could speculate that preventing viral infections, for ex- Commons Attribution 4.0 International License (http:// ample through influenza vaccination, could help reduce the creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appro- incidence of type 1 diabetes. In a recent paper from The priate credit to the original author(s) and the source, provide a link to the Environmental Determinants of Diabetes in the Young Creative Commons license, and indicate if changes were made. (TEDDY) study, the Pandemrix vaccination was associated with a lower risk of islet autoimmunity in children at increased genetic risk in Finland, whereas no difference was seen in References Sweden [40]. Unfortunately, we do not have access to data on autoimmunity in our register-based study. In line with the 1. Hyöty H (2016) Viruses in type 1 diabetes. Pediatr Diabetes 17:56– Swedish data we did not find any association with Pandemrix and type 1 diabetes in our study [20]. In Norway, many of 2. Op de Beeck A, Eizirik DL (2016) Viral infections in type 1 diabe- those who were vaccinated with Pandemrix had already been tes mellitus–why the beta cells? Nat Rev Endocrinol 12:263–273 3. Rodriguez-Calvo T, Sabouri S, Anquetil F, von Herrath MG (2016) infected, or were infected with influenza A (H1N1) after vac- The viral paradigm in type 1 diabetes: who are the main suspects? cination, but before effective protective antibodies had been Autoimmun Rev 15:964–969 induced [21]. 4. Beyerlein A, Wehweck F, Ziegler A-G, Pflueger M (2013) In conclusion, we could not demonstrate a clear association Respiratory infections in early life and the development of islet autoimmunity in children at increased type 1 diabetes risk: evidence between clinically reported pandemic influenza infection and from the BABYDIET study. JAMA Pediatr 167:800–807 incident type 1 diabetes in this register-based cohort study. We 5. Rasmussen T, Witsø E, Tapia G, Stene LC, Rønningen KS (2011) did, however, find a twofold excess of incident diabetes in the Self-reported lower respiratory tract infections and development of subgroups with laboratory-confirmed pandemic influenza A islet autoimmunity in children with the type 1 diabetes high-risk (H1N1) or pandemic influenza diagnosed in specialist HLA genotype: the MIDIA study. Diabetes Metab Res Rev 27: 834–837 healthcare. This suggests that respiratory infections may play 6. Beyerlein A, Donnachie E, Jergens S, Ziegler A (2016) Infections a role in the aetiology of type 1 diabetes, but more studies are in early life and development of type 1 diabetes. JAMA 315:1899– warranted. 7. Lonnrot M, Lynch KF, Elding Larsson H et al (2017) Respiratory infections are temporally associated with initiation of type 1 diabe- Acknowledgements Data from the Norwegian Patient Register have tes autoimmunity: the TEDDY study. Diabetologia 60:1931–1940 been used in this publication. The interpretation and reporting of these 8. Lee H-Y, Lu C-L, Chen H-F, Su H-F, Li C-Y (2015) Perinatal and data are the sole responsibility of the authors, and no endorsement by the childhood risk factors for early-onset type 1 diabetes: a population- Norwegian patient register is intended nor should be inferred. based case-control study in Taiwan. Eur J Pub Health 25:1024– A part of the study in abstract form was presented as a poster at the 53rd EASD Annual Meeting in Lisbon, 11–15 September 2017. 9. Capua I, Mercalli A, Pizzuto MS et al (2013) Influenza A viruses grow in human pancreatic cells and cause pancreatitis and diabetes Data availability The datasets generated during and/or analysed during in an animal model. J Virol 87:597–610 the current study are not publicly available because of data protection 10. Blum A, Podvitzky O, Shalabi R, Simsolo C (2010) Acute pancre- regulations. atitis may be caused by H1N1 influenza A virus infection. Isr Med Assoc J 12:640–641 Funding This research was partly supported by the Norwegian Institute 11. Habib A, Jain A, Singh B, Jamshed N (2016) H1N1 influenza of Public Health and a grant from the South-Eastern Norway Regional presenting as severe acute pancreatitis and multiorgan dysfunction. Health Authority. Am J Emerg Med 34:1911.e1911–1911.e1912 The sponsors had no role in the: design and conduct of the study; collection, management, analysis and interpretation of the data; prepara- 12. Baran B, Karaca C, Soyer OM et al (2012) Acute pancreatitis as- tion, review or approval of the manuscript; or decision to submit the sociated with H1N1 influenza during 2009 pandemic: a case report. manuscript for publication. Clin Res Hepatol Gastroenterol 36:e69–e70 13. Watanabe N (2011) Conversion to type 1 diabetes after H1N1 in- Duality of interest HLG has received grants for lectures or consulting fluenza infection: a case report. J Diabetes 3:103 from AstraZeneca, Novo Nordisk, Sanofi, Boehringer Ingelheim and 14. Larcombe PJ, Moloney SE, Schmidt PA (2011) Pandemic (H1N1) Merck Sharp & Dohme. All other authors declare that there is no duality 2009: a clinical spectrum in the general paediatric population. Arch of interest associated with their contribution to this manuscript. Dis Child 96:96–98 15. World Health Organization (2014) Influenza virus infections in Contribution statement LCS, PLDR and IJB had full access to all of the humans [Internet]. Available from http://www.who.int/influenza/ data in the study and take responsibility for the integrity of the data and human_animal_interface/virology_laboratories_and_vaccines/ the accuracy of the data analysis. LCS, HLG and SEH contributed to the influenza_virus_infections_humans_feb14.pdf?ua=1. Accessed 9 study concept and design. All authors contributed to the acquisition, Oct 2017 Diabetologia 16. DG Statement following the meeting of the Emergency Committee, 29. Simonsen KA, Hunskaar S, Sandvik H, Rortveit G (2013) Capacity and adaptations of general practice during an influenza pandemic. WHO Available from http://www.who.int/csr/disease/swineflu/ 4th_meeting_ihr/en/. Accessed 9 Oct 2017 PLoS One 8:e69408 17. Valdes C, Unanue N, Hernandez M et al (2013) Is there a link 30. Swedish WHO National Influenza Centre (2011) The influenza between influenza and type I diabetes? Increased incidence of A(H1N1)2009 pandemic in Sweden, 2009–2010. Available from TID during the pandemic H1N1 influenza of 2009 in Chile. https://www.folkhalsomyndigheten.se/pagefiles/15030/influenza- Pediatr Endocrinol Rev 11:161–166 in-sweden-2009-2010.pdf. Accessed 9 Oct 2017 18. Nenna R, Papoff P, Moretti C et al (2011) Detection of respiratory 31. (2010) Rapport. Ny influensa A (H1N1) 2009. Gjennomgang av viruses in the 2009 winter season in Rome: 2009 influenza A erfaringene i Norge. Direktoratet for samfunnssikkerhet og (H1N1) complications in children and concomitant type 1 diabetes beredskap (DSB), Oslo [In Norwegian]. Available from http:// onset. Int J Immunopathol Pharmacol 24:651–659 www.dsbinfo.no/DSBno/2010/Rapport/Pandemirapport/. 19. Ziegler AG, Rewers M, Simell O et al (2013) Seroconversion to Accessed 9 Oct 2017 multiple islet autoantibodies and risk of progression to diabetes in 32. Waalen K, Kilander A, Dudman S, Krogh G, Aune T, Hungnes O children. JAMA 309:2473–2479 (2010) High prevalence of antibodies to the 2009 pandemic influ- 20. Ruiz PLD, Stene LC, Gulseth HL et al (2018) Pandemic influenza enza A (H1N1) virus in the Norwegian population following a A H1N1 vaccination and subsequent risk of type 1 diabetes in major epidemic and a large vaccination campaign in autumn Norway. Epidemiology 29:e6-e8 2009. Euro Surveill 15:19633 21. Van Effelterre T, Dos Santos G, Shinde V (2016) Twin peaks: A/ 33. Kondrashova A, Nurminen N, Patrikainen M et al (2015) Influenza H1N1 pandemic influenza virus infection and vaccination in A virus antibodies show no association with pancreatic islet auto- Norway, 2009–2010. PLoS One 11:e0151575 antibodies in children genetically predisposed to type 1 diabetes. 22. Freiesleben de Blasio B, Iversen BG, Scalia Tomba G, Barry MA Diabetologia 58:2592–2595 (2012) Effect of vaccines and antivirals during the major 2009 A 34. Christen U, Bender C, von Herrath MG (2012) Infection as a cause (H1N1) pandemic wave in Norway–and the influence of vaccina- of type 1 diabetes? Curr Opin Rheumatol 24:417 tion timing. PLoS One 7:e30018 35. Stene LC, Barriga K, Hoffman M et al (2006) Normal but increas- 23. Trogstad L, Ung G, Hagerup-Jenssen M, Cappelen I, Haugen IL, ing hemoglobin A1c levels predict progression from islet autoim- Feiring B (2012) The Norwegian immunisation register— munity to overt type 1 diabetes: Diabetes Autoimmunity Study in SYSVAK. Euro Surveill 17:pii:20147 the Young (DAISY). Pediatr Diabetes 7:247–253 24. Anestad G, Nordbo SA (2011) Virus interference. Did rhinoviruses 36. Cuesta JG, Aavitsland P, Englund H et al (2016) Pandemic vacci- activity hamper the progress of the 2009 influenza A (H1N1) pan- nation strategies and influenza severe outcomes during the influen- demic in Norway? Med Hypotheses 77:1132–1134 za A (H1N1) pdm09 pandemic and the post-pandemic influenza 25. Håberg SE, Trogstad L, Gunnes N et al (2013) Risk of fetal death season: the Nordic experience. Euro Surveill 21 https://doi.org/10. after pandemic influenza virus infection or vaccination. N Engl J 2807/1560-7917.ES.2016.21.16.30208 Med 368:333–340 37. Qi Z, Hu H, Wang Z et al (2018) Antibodies against H1N1 influ- 26. Furu K (2008) Establishment of the nationwide Norwegian enza virus cross-react with alpha-cells of pancreatic islets. Prescription Database (NorPD)—new opportunities for research JDiabetes Investig 9:265–269 in pharmacoepidemiology in Norway. Norsk Epidemiologi 18 38. Monsalvo AC, Batalle JP, Lopez MF et al (2011) Severe pandemic https://doi.org/10.5324/nje.v18i2.23 2009 H1N1 influenza disease due to pathogenic immune com- 27. Skrivarhaug T, Stene LC, Drivvoll AK, Strøm H, Joner G, Group plexes. Nat Med 17:195–199 NCDS (2014) Incidence of type 1 diabetes in Norway among chil- dren aged 0–14 years between 1989 and 2012: has the incidence 39. Richardson SJ, Horwitz MS (2014) Is type 1 diabetes “going viral”? stopped rising? Results from the Norwegian Childhood Diabetes Diabetes 63:2203–2205 Registry. Diabetologia 57:57–62 40. Elding Larsson H, Lynch KF, Lonnrot M et al (2018) Pandemrix(R) 28. Simonsen KA, Hunskaar S, Wensaas K-A et al (2012) Influenza- vaccination is not associated with increased risk of islet autoimmu- like illness in Norway: clinical course, attitudes towards vaccination nity or type 1 diabetes in the TEDDY study children. Diabetologia and preventive measures during the 2009 pandemic. Fam Pract 29: 61:193–202 139–146

Journal

DiabetologiaSpringer Journals

Published: Jun 22, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off