Impact of projected climate and socioeconomic scenarios on state-wise annual dengue incidence in India using ensemble models

India, the world's most populous country, has reported over 1 million dengue cases and ~3,000 deaths between 2007 and 2022. With the annual state-wise data, we examined the spatiotemporal distribution of dengue in 28 states and eight union territories for 16 years across India. Using state-wise data on climatic variables, socio-economic inequities and land-use land-cover changes, potential determinants for the state-wise transmission were identified through a supervised regression model. The identified determinants were then mapped to various novel developmental scenarios, which were designed based on the existing shared socio-economic pathways. To estimate the dengue burden for each scenario, ensemble models of XGBoost and Gradient Boosting regression algorithms were developed. We note that 73% of the cases occurred between 2016 and 2022, highlighting a significant increase in dengue outbreaks across the country. All Himalayan states, which witness colder temperatures, have witnessed a growth in cases: Himachal Pradesh reported 168 times more cases between 2016 and 2022 than those observed between 2007 and 2015. The models suggest that dengue incidences may potentially change under future socioeconomic burden, although projections are associated with substantial uncertainty and should be interpreted as potential trajectories rather than definitive forecasts. We estimate that development focused on sustainability (874.2 per 10 million; 95% CI: 535.4, 1212.9) and fossil fuels (888.02 per 10 million; 95% CI: 521.2, 1254.9) will relatively cause a lesser burden across the country by the 2030s. Southern states are projected to have higher dengue outbreaks, while Jharkhand, a historically malaria-endemic state, is estimated to report twice as many cases in 2050 as what was reported in 2022. Given the uncertainty associated with long-term projections, public health strategies may benefit from adaptive approaches which are backed by climate- and socioeconomic-data integrated early warning systems that can respond to evolving climatic and socioeconomic conditions influencing dengue transmission. Our study provides insights into how the spread of dengue will change with varying models of socio-economic development, which highlights the spatial heterogeneity in potential future dengue risk, suggesting that resource allocation and surveillance efforts may benefit from region-specific prioritisation instead of a uniform policy.Author summary: We study the impact of climate change on dengue cases in Indian states using machine learning models. With modified socio-economic development scenarios, we projected the increase in dengue cases for each state in a near (2030) and mid-scenario future (2050). Historically, dengue cases reported during 2016–2022 were three times higher than those reported between 2007–2015. Our modelling results suggest that dengue incidence could change under different future socioeconomic pathways. However, the magnitude and direction of these changes vary across scenarios and are associated with considerable uncertainty, highlighting the need for cautious interpretation of long-term projections. In the immediate future, development focused on sustainability will likely have a relatively lesser increase, while a regional rivalry scenario may witness the highest increase. In the southern India, increased outbreaks are estimated in the future. In contrast, dengue is likely to decrease significantly in the densely populated states of the Gangetic Plains. The projected spatial heterogeneity in dengue incidence suggests that targeted surveillance and early detection systems may be particularly important in states where future risk is projected to change. Further, integrating climate and socioeconomic information into routine monitoring could improve preparedness for potential changes in transmission patterns.