Dengue, climate and marginalization in Mexico, 2021–2025

This repository contains the data and R code needed to reproduce a Bayesian spatiotemporal analysis of weekly confirmed dengue cases in Mexican municipalities from 2021 to 2025. Case data come from the national epidemiological surveillance system (SINAVE, Dirección General de Epidemiología), climate exposures (dewpoint temperature, air temperature and precipitation) from the ERA5-Land reanalysis, and municipal marginalization from the 2020 CONAPO index. The main model is a negative binomial model fitted with R-INLA, with a BYM2 spatial effect, distributed lag effects of climate over 0 to 10 weeks and state-specific seasonality. Because many municipality-weeks are not reported to surveillance, the analysis treats unreported weeks as missing rather than as zero cases, restricts the main sample by weekly reporting coverage, and assesses reporting bias with several sensitivity analyses, including multiple imputation. The numbered R pipelines run from sample construction to the final figures and tables. A single README file describes all datasets and scripts.