Rotavirus Vaccination and Type 1 Diabetes Risk Among US Children With Commercial Insurance

Rotavirus may precipitate type 1 diabetes (T1D).^1,2 Thus, rotavirus vaccination may be associated with T1D. For example, incidence of T1D in Australian children younger than 5 years declined after rotavirus vaccine introduction.³ However, 2 Finnish studies^4,5 showed no association of rotavirus vaccination with T1D risk in children 10 years and younger, although statistical power in these studies was limited. We compared T1D risk by rotavirus vaccination status among US children with commercial insurance.

Data for January 2006 through December 2017 were analyzed from the IBM MarketScan Commercial Database (formerly known as the Truven Health MarketScan Commercial Claims and Encounters database), which contains deidentified data from more than 20 million individuals in employer-sponsored commercial health insurance plans. Because all data were fully deidentified and no interaction occurred with humans, this analysis was deemed exempt from review and informed consent by the US Centers for Disease Control and Prevention institutional review board.

Children were included if they were age eligible to receive rotavirus vaccine (with a birthdate of January 1, 2006, or later) and continuously enrolled in their insurance plan from birth. Current Procedural Terminology codes (90680 [for RotaTeq vaccine (Merck Sharp & Dohme Corporation)] and 90681 [for Rotarix vaccine (GlaxoSmithKline)]) were used to determine rotavirus vaccination status and age at each dose, as previously described.⁶ To increase specificity, T1D was defined as 2 or more separate claims listing International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes 250.X1 or 250.X3 or International Statistical Classification of Diseases, Tenth Revision, Clinical Modification codes E10.X in any care setting. Children with a single T1D claim were excluded, as were those diagnosed those diagnosed before 27 weeks of age (survival curve) or before 12 months of age (measures of association).

For the survival curve, follow-up began at 27 weeks of age. For measures of association, follow-up began at 12 months of age. Crude rate ratios were calculated using person-time accumulated in each stratum of rotavirus vaccination status. Extended Cox regression models with age in weeks as the time scale included rotavirus vaccination status as the time-varying exposure and were controlled for birth year. To mitigate exposure misclassification from missing vaccination claims and control for confounding by medical conditions possibly associated with missed vaccinations and risk of T1D, only children who received at least 1 dose of diphtheria-tetanus-pertussis–containing vaccine were included. This restriction was lifted in a sensitivity analysis. In a second sensitivity analysis, children in states with universal vaccine purchase programs were excluded, since they could have been vaccinated without a vaccination-associated insurance claim. Results were stratified by birth year, since loss to follow-up increased with increasing age. Data were censored at the end of continuous enrollment, the end of the study period (December 31, 2017), or diagnosis of T1D, whichever came first.

Analyses used SAS version 9.4 (SAS Institute) and R version 3.4.2 (the R Foundation for Statistical Computing). Data analysis occurred from January 2019 to April 2019; α was set to .05.

Overall, 843 of 2 854 571 eligible children were diagnosed with T1D; the 10-year cumulative incidence was 0.25% (95% CI, 0.22%-0.28%). Among 1 563 540 children followed up until age 12 months or older, 1 035 198 (66.2%) were fully vaccinated (2 doses of Rotarix or 3 doses of RotaTeq), 210 057 (13.4%) were partially vaccinated, and 318 285 (20.4%) were unvaccinated against rotavirus.

Unadjusted T1D-free survival curves were stratified by rotavirus vaccination status (Figure). Adjusted hazard ratios comparing fully and partially vaccinated children with unvaccinated children were nonsignificant in primary and sensitivity analyses (Table).

We found no association between rotavirus vaccination and T1D among US children with commercial insurance who were followed up until a maximum of 12 years of age. This is consistent with results from 2 Finnish cohorts.^4,5

Some limitations should be considered. First, the MarketScan database, although comprehensive, is not nationally representative. Second, neither vaccination nor T1D were confirmed by medical record review, and 138 751 children (those with no recorded diphtheria-tetanus-pertussis doses; 8.2% of 1 702 291 children who would otherwise have been eligible based on date of birth, length of follow-up, and continuous enrollment) were excluded. However, sensitivity analyses suggest that vaccine status misclassification had minimal association with outcomes, and ICD-9-CM codes have a strong association with T1D among children. Last, loss to follow-up is substantial, especially after 5 years, and this contributed to wide 95% CIs. However, sensitivity analyses indicated similar conclusions by birth cohort.

Our results are consistent with a favorable safety profile for rotavirus vaccines. They support continued universal rotavirus vaccination in the United States.

Accepted for Publication: July 24, 2019.

Corresponding Author: Rachel M. Burke, PhD, MPH, US Centers for Disease Control and Prevention, 1600 Clifton Rd NE, Atlanta, GA 30329 (rburke@cdc.gov).

Published Online: January 21, 2020. doi:10.1001/jamapediatrics.2019.5513

Author Contributions: Dr Burke 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.

Concept and design: Burke, Tate, Gregg, Parashar.

Acquisition, analysis, or interpretation of data: Burke, Tate, Dahl, Saydah, Imperatore, Parashar.

Drafting of the manuscript: Burke.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Burke, Dahl, Imperatore.

Administrative, technical, or material support: Saydah, Parashar.

Supervision: Tate, Parashar.

Conflict of Interest Disclosures: None reported.

Disclaimer: The findings and conclusions of this report are those of the authors and do not necessarily represent the official position of the US Centers for Disease Control and Prevention.