Climate-driven inequalities in dengue transmission between rural and urban populations in western India

Background Dengue incidence in India is rising and expanding across diverse geographical terrains and climatic zones, with transmission increasingly reported beyond urban centres into peri-urban and rural regions. Methods Using a 12-year longitudinal data from Pune district in India, we investigate the spatiotemporal dynamics of dengue across rural and urban regions and developed a climate integrated- and artificial intelligence (AI)–driven framework to predict current and future dengue risk under climate change scenarios till 2100s. The relative environmental suitability for dengue transmission at 1 km ² resolution is also estimated using Bayesian hierarchical models. Results Between 2012 and 2023, the mean prevalence in urban Pune (181 cases per 10 million) was ~ 3.5-fold higher than in rural areas, with peak transmission observed in October in both settings. Males reported a 1.5 times higher prevalence than females, with maximum burden reported among individuals aged 10–50 years. Spatial analyses identified persistent transmission hotspots that periodically intensified during monsoon/post-monsoon season. The AI models identified the role of 2-month lagged temperature in urban areas and no delayed-temperature effect in rural areas. Under the SSP5 scenario, the projected cumulative burden in rural Pune between 2024–2100 reaches 194,219 cases, representing ~ 15% of the projected urban burden. Bayesian analyses indicate that largest number of neighbourhoods (n = 427) with suitability > 8.75% occurred in September. Conclusion Collectively, these findings reveal distinct urban–rural climatic sensitivities and demonstrate the utility of AI-enabled predictive modelling for early warning, preparedness and targeted intervention to mitigate dengue transmission under changing climate conditions.