Estimating Alveolar Ventilation for Use in Physiological-based Exposure Models

Abstract

The United States Environmental Protection Agency (EPA) utilizes the Air Pollutants Exposure (APEX) model to estimate exposure to criteria air pollutants such as carbon monoxide (CO), to inform the setting of air quality standards (e.g., US EPA, 2010). The data for estimating alveolar ventilation rates (𝑉 ̇𝐴) in this model used for modeling CO exposures and dose was last updated over 20 years ago and could benefit from additional information extracted from new studies as well as performing additional data analysis. Specifically, the existing constant value used in the equation to estimate 𝑉 ̇𝐴, 19.63, has been raised by public reviewers of the most recent CO human health risk and exposure assessment as potentially limited in its applicability, particularly at when simulated individuals are high breathing rates (US EPA, 2011). This new study was conducted to increase our knowledgebase of the physiological aspects of ventilation and to further develop an internal alveolar ventilation database that could be used to either support the existing quantitative linear relationship or, where possible, improve the algorithm. This evaluation also explored the relationship of how alveolar ventilation with might be affected by key demographic attributes such as body mass, age, and sex. Two approaches were explored to estimate 𝑉 ̇𝐴 and are linked to key respiratory variables already modeled by APEX. The first, used a direct relationship between 𝑉 ̇𝐴 and oxygen consumption rates (𝑉 ̇𝑂2). The second used the relationship of dead space to tidal volume (VD/VT) to 𝑉 ̇𝑂2 and expected to be used with total ventilation rate (𝑉 ̇𝐸) to indirectly estimate 𝑉 ̇𝐴. This study increased the amount of useful data within the internal database by over two-fold, albeit sample size limitations for important population groups remain an issue (e.g., adults > 35 years old). New data analyses using the updated database suggest a linear relationship between 𝑉 ̇𝐴 and 𝑉 ̇𝑂2 appears appropriate across all population groups and breathing rates, while differences in the VD/VT to 𝑉 ̇𝑂2 relationship across 3 population groups suggests adoption of this latter approach is less likely due to the sample size issues. It is hoped study results assist in the development of a newly refined algorithm to for estimating CO exposure and dose estimations, reduce uncertainties in exposure models that would use such an approach, and provide sound support to air quality regulations.