78 Chapter 3 onset of NT1 could have played a role in the 2010 peak in NT2 and IH with more individuals subsequently seeking medical attention [135, 174]. The absence of similar media coverage in 2013 suggests increased awareness alone cannot fully explain the peaks in NT2 and IH during that time. In our study, we cannot dismiss the possibility of influenza vaccination, especially Pandemrix, playing a delayed role in the incidence peaks. However, the link between the severity of the influenza season and the occurrence of central disorders of hypersomnolence also implies that influenza vaccination (non-Pandemrix) might offer protection against the development of hypersomnolence disorders. There are several limitations to our study. First, the observed effects in our study were strongest in the analyses combining data from all countries and similar results were not always observed in each individual country. We believe this is most likely because of the limited data points per country (often only nine years) and the large variability in influenza monitoring policies per country [176]. Many countries improved their influenza surveillance system during pH1N1 and since then mainly focused on subtyping of type A influenza strains [177]. The Netherlands is known for its precise governmental influenza monitoring program with representative national coverage and standard subtyping of all four influenza strains through a combination of sentinel, non-sentinel and modelling-based estimates of influenza infections [166, 167]. Together with the compactness of the Netherlands resulting in limited provincial influenza strain circulation differences, we believe that this likely explains the large effect sizes that we observed within the Netherlands. The immunological mechanisms responsible for the connections we identified between the onset of hypersomnolence disorder and type A H1N1, and NT1 onset with type B Victoria, remain unclear. We provide new epidemiological associations and this should not be mistaken for causality. Various underlying pathophysiological mechanisms have been proposed. While molecular mimicry may initiate a self-reactive T cell response related to hypocretin after influenza infection or Pandemrix vaccination, current supporting evidence is inconsistent. Autoaggressive bystander activation by T cells without direct cross-reactivity has also been proposed and warrants further studies [157, 178]. Immunological investigations are essential to better comprehend these processes and establish possible causal relationships between influenza infection and onset of central disorders of hypersomnolence. In such studies it is important to integrate other pathogens (such as streptococcus pyogenes and influenza type C) and different clades/sub-clades of type A and type B influenza. Our sample exhibits a minor overlap of approximately 25% with the EU-NN database, thus it cannot be regarded as entirely independent. Within
RkJQdWJsaXNoZXIy MjY0ODMw