Ecological networks experiencing persistent biological invasions may exhibit distinct topological properties, complicating the understanding of how network topology affects disease transmission during invasion-driven community assembly. We developed a trait-based network model to assess the impact of network topology on disease transmission, measured as community- and species-level disease prevalence. We found that trait-based feeding interactions between host species determine the frequency distribution of the niche of co-occurring species in steady-state communities, being either bimodal or multimodal. The width of the growth kernel influences the degree-biomass relationship of species, being either weakly positive or strongly negative. When this relationship is weakly positive, species-level disease prevalence is primarily correlated with biomass. However, when the degree-biomass relationship is strongly negative, species-level disease prevalence is determined by the difference between a host species’ in-degree and out-degree closeness centrality. At the community level, disease prevalence is generally amplified by increasing host richness, community biomass, and the standard deviation of interaction generality, while it is diluted by higher network connectance. Our framework verifies the amplification effects of host richness during invasion-driven community assembly and offers valuable insights for estimating disease prevalence based on host network topology.