Positively interacting strains that circulate in a network structured population induce cycling epidemics of Mycoplasma Pneumoniae
In many countries Mycoplasma pneumoniae (MP) epidemics last approximately one to two years and occur every three to seven years. Poor understanding of the drivers of recurrent MP epidemics limits the predictability of and dynamic responses to the outbreak. Taking into account network structured contacts among people and co-circulating strains of MP, we propose a multi-strain SIRS network model of epidemics of MP where different strains interact during re-infection and within secondary infection. Simulations show that although strain interactions and network-mediated spatial correlations are two separate mechanisms for MP epidemics cycling, each requires very restricted model parameter values such as strong strain interactions and strong network contacts, respectively. When both mechanisms work collectively, MP recurrent epidemics become feasible within the plausible ranges of model parameters. This indicates that positively interacting strains that co-circulate within network contacts induce periodicity and dominant strain shift in observed MP incidence.