Combination of toxicological and epidemiological approaches for estimating the health impact of atmospheric pollutants. A proof of concept for NO.

Background: Regular monitoring of the air pollutant nitrogen dioxide (NO), an indicator for traffic-related emissions, is a priority in urban environments. The health impacts associated with NO exposure are the result of a combination of factors, including concentration, duration of exposure, and interactions with other pollutants. WHO has established air quality guidelines based on epidemiological studies.

Objective: This study develops a new concept "Health Impact Pathways (HIPs)" using adversity as a probabilistic indicator of health effects. For this purpose, it integrates available toxicological and epidemiological information, using Adverse Outcome Pathways (AOPs), in order to understand chemical-biological interactions and their consequences on health.

Methods: Literature review and meta-analysis of toxicological data supported by expert judgment were performed to establish: a) adversity pathways, b) quantitative criteria for scoring the observed toxicological effects (adversity indicators), c) NO exposure - adversity relationship for both long-term (1-36 months) and shortterm (1-7 days). The NO daily concentrations from January 2001 to December 2022, were obtained from Madrid city Air Quality network monitoring database. Adversity levels were compared with relative risk levels for all-cause and respiratory mortality estimated using linear equations from WHO 2021 guidelines.

Results: Non-linear relations were obtained for all long- and short-term NO related adversity indicators; for long-term effects, the best fitting was obtained with a modified Haber's law model with an exponential coefficient for the exposure time of 0.25. Estimations are presented for a set of case studies for Madrid city, covering temporal and spatial variability. A clear improvement trend along the two decades was observed, as well as high inter- and intra-station variability; the adversity indicators provided integrated information on the temporal and spatial evolution of population level risk.

Discussion: The proposed HIP conceptual approach offers promising advances for integrating experimental and epidemiological data. The next step is linking the concentration-adversity relationship with population health impacts through probability estimations, the preliminary estimations confirm the need for assessing independently different population groups.