70 Chapter 3 Correlational analyses were repeatedly performed for each country and for all individuals, only children and only adults. Type B Victoria and type B Yamagata data were only reliably available for the Netherlands. These analyses were not performed for the NT2/IH group because of insufficient power per country. As a sensitivity analysis to test whether our correlations could have been due to chance, we shifted incidence rates of influenza strains one year backward and repeated the correlations with respectively NT1 incidence rates in case there was a significant relationship in the main analysis. We chose a correlation coefficient > 0.2 and P-value < 0.05 as a significant correlation. Results We included representative samples (Table 1) of NT1 (N = 981), NT2 (N = 189) and IH (N = 356). The 2010 and 2013 NT1 incidence peaks We identified two main incidence peaks in NT1 when including all age groups (Figure 1A). The increased incidence was also significantly found separately for children (2.02-fold increase) and adults (1.99-fold increase) in 2010 and for children (1.84-fold increase) and adults (1.69-fold increase) in 2013 (Figure 1B-C). Figure 1. The changes in incidence of narcolepsy type 1 (NT1) in all individuals (A), children (B) and adults (C). The predicted incidence rates and their 95% predictive confidence intervals are marked as green circles/lines and the actual values are in black circles with in red the relative significant increase in incidence rates with 95% predictive confidence intervals. EDS = excessive daytime sleepiness.
RkJQdWJsaXNoZXIy MjY0ODMw