New R package ‘bdlnm’ released on CRAN to advance the study of climate impacts on health

Climate change is rapidly increasing the population's exposure to extreme environmental conditions. More frequent heatwaves, higher levels of air pollution, and more persistent extreme events are becoming part of everyday life in many regions.

However, environmental risks rarely act instantly. For example, a heatwave today may lead to an increase in mortality several days later or and exposure to high air pollution levels can have delayed and cumulative effects on health. Understanding these time-dependent relationships is essential for informed decision-making.

Distributed Lag Non-Linear Models (DLNMs) are the reference framework for studying these complex exposure-response relationships delayed in time. In a context of growing data availability and the need to analyse health impacts at finer geographical scales, DLNMs have been extended to a Bayesian framework (B-DLNMs). This Bayesian approach allows researchers to fit more complex and flexible models, while also better quantifying uncertainty and deriving more informative estimates through posterior distributions.

In this context, Pau Satorra, a PhD researcher in the Biostatistics Unit at IGTP, has developed and released the R package bdlnm as part of his doctoral research. The package makes this methodology accessible to researchers through a user-friendly implementation. The bdlnm package provides a complete set of tools to:

• Fit and estimate B-DLNMs
• Predict and visualize exposure-lag-response relationships
• Estimate optimal exposure values, such as the Minimum Mortality Temperature (MMT)
• Quantify health impacts through attributable fractions and numbers

This development represents a relevant milestone within his PhD and contributes to expanding the analytical tools available to study the impact of environmental exposures -such as temperature and air pollution- on health. These approaches are particularly relevant for assessing the effects of climate change, including heatwaves and other extreme events.

The bdlnm package is now available on CRAN, with documentation and tutorials designed for users with different levels of expertise.

More information in the bdlnm website: https://pasahe.github.io/bdlnm/

Its release marks an important step toward more flexible, accessible, and impactful tools for studying how environmental exposures affect health over time.