Understanding how host dynamics, including spatiotemporal variations of population size and dispersal, may affect the epidemiology of infectious diseases is an active research area. Population dynamics drive neutral and adaptive micro-evolutionary processes that in turn, influence pathogens persistence, their distribution and evolution. This study focuses on a bank vole (Myodes glareolus) metapopulation surveyed in Finland between 2005 and 2009. Bank vole is the reservoir of Puumala hantavirus (PUUV), the agent of nephropathia epidemica (NE, a mild form of hemorrhagic fever with renal symptom) in humans. M glareolus populations experience multiannual density fluctuations that have previously been related to the level of genetic diversity maintained in bank voles, variations in PUUV prevalence and NE occurence in humans. Here, we examine the potential impacts of vole metapopulation genetics on PUUV epidemiology. Using microsatellite markers, we evaluate the impact of density cycles on genetic drift and host gene flow. We show that genetic drift slightly and transiently affects neutral and adaptive genetic variability within the metapopulation. Gene flow seemed to counterbalance its effects during the multiannual density fluctuations. The low abundance phase may therefore be too short to impact genetic variation in the host, and consequently viral genetic diversity. We did not detect any environmental heterogeneity affecting vole gene flow, which might explain the absence of spatial structure previously detected in PUUV in this area. We found evidence for the role of vole dispersal on PUUV circulation through sex-specific and density dependent movements. We further analyzed how multiannual dynamic cycles affected selection acting on immune related genes involved in susceptibility to PUUV. We detected associations between Mhc-Drb haplotypes and PUUV serology, and density related patterns of departures from neutrality on Tnf promoter, Tlr-4 and Mx2 genes. They could reflect signatures of positive and balancing selection acting on these genes, which might influence PUUV microevolution. Altogether, this study provides an original framework for population genetics research developed in epidemiological contexts.