Against the backdrop of rapid global biodiversity loss, the impact of biodiversity change on disease risk has attracted considerable attention. Although previous studies have demonstrated an association between biodiversity and disease risk, the underlying mechanisms governing this relationship remain controversial. To identify the key ecological factors influencing the biodiversity-disease risk relationship, a cellular automata (CA) model to systematically investigate the effects of species spatial distribution patterns, community assembly patterns, host life-history trade-offs, and disease transmission distance was employed. The results show that species spatial distribution patterns, host life-history trade-offs and the disease transmission distance significantly affect the number of infected individuals, but do not alter the overall trend of the biodiversity-disease risk relationship. Among these factors, community assembly patterns are the key factor determining the overall trend of the biodiversity-disease risk relationship. Under additive communities assembly, increasing biodiversity will intensify disease risk, exhibiting an amplification effect. Under substitutive communities assembly, increasing biodiversity exhibits a composite effect characterized by an initial amplification effect followed by a dilution effect. Furthermore, the effects of species spatial distribution patterns and the disease transmission distance on disease risk exhibit pronounced spatial heterogeneity. By applying a CA model, this study provides a novel theoretical framework, deepens the understanding of the biodiversity–disease risk relationship, and provides theoretical support for developing more evidence-based disease prevention and control strategies.