Measles is a highly contagious viral infection, which is associated with a transient but profound immune suppression resulting in increased susceptibility to the secondary infection. Recently, it has been proposed that measles-induced immune suppression lasts for ~30 months and leads to increased mortality risk from subsequent non-measles infections. To verify this theory, we focus on parallel epidemiological incidence reports of measles and whooping cough, with a view to testing whether infection with measles affects the severity and susceptibility to subsequent infection with whooping cough. We first developed a seasonally-driven transmission model of two infectious diseases to establish the dynamical consequences of measles-induced immunomodulation. Next, using likelihood-based statistical inference we examined the detectability of the immune suppression hypothesis in simulated mortality data. Finally, we applied this methodology to both mortality and disease incidence reports from England & Wales during the 20th Century.
Results/Conclusions
Our bifurcation analyses indicated no qualitative effect of immunomodulation. The notable effects are increasing amplitude of whooping cough epidemics and a higher average incidence than would be predicted by single-disease theory. The results of our trajectory matching experiments revealed those model parameters that may be confidently estimated (such as the severity of measles-induced immune amnesia) and those that are less well identified (such as the duration of this phenomenon). We further computed confidence intervals of focal parameters via profile likelihoods. We will extend this work by confronting our stochastic transmission model to historical incidence reports using Maximization by Iterated particle Filtering (MIF). Results of this study seek to test the evidence for measles-associated immune amnesia and a quantification of its effect on whooping cough.