Quadrivalent HPV vaccine and the risk of type 1 diabetes mellitus in grade 8 girls: A population based cohort study
MetadataShow full item record
Background: Vaccines have been hypothesized in the etiology of autoimmune diseases including type 1 diabetes. There are cases of diabetes reported in the Vaccine Adverse Event Reporting System (VAERS) following the administration of the human papillomavirus (HPV) vaccine, however this potential association has yet to be investigated. The objective of this thesis was to determine whether there is an association between immunization against HPV and the development of type 1 diabetes mellitus in grade 8 girls eligible for Ontario’s vaccination program. Methods: A retrospective, population-based, cohort study of girls residing within an Ontario health unit and eligible for the province’s publicly funded school-based HPV vaccination program between 2007 and 2010 was executed using provincial administrative health databases and the Immunization Recording Information System (IRIS) database. To control for known, unknown and unmeasured time-independent confounders, a self-controlled case series analysis was conducted. The relative incidence and 95% confidence interval were estimated using conditional Poisson regression. Results: The study cohort was comprised of 3465 girls with a mean age of 13.2 years at cohort entry (range 12.7 to 13.6 years). The mean duration of follow up was 2.7 years and ranged from 1.6 to 3.6 years. The proportion of girls who received at least one dose of the qHPV vaccine during the observation period was 58.3% (n=2020). During the study follow-up 15 cases of new onset type 1 diabetes were observed, six of which were classified as etiologically exposed to the qHPV vaccine. Using an indefinite risk window, immunization with the qHPV vaccine was not associated with an increased risk of developing type 1 diabetes (age and season-adjusted RI 0.15; 95% CI 0.02-1.32). Conclusions: The results of this thesis regarding the risk of type 1 diabetes following immunization with the qHPV vaccine are inconclusive as a consequence of the small number of cases identified. However, the random distribution of cases across time and across exposure status suggests that there is no association. Before a definitive conclusion is reached the analysis must be re-conducted on a larger cohort.