Ventilation rate and physical activity

[markotainio]

In this issue: Discussion on how to link physical activity and ventilation rate. The idea here is that we would have function that defines ventilation rate as a result of physical activity intensity (defined as mMETS). Current versions of ITHIM wont have this link.

Ventilation rate is critical for defining the dose of air pollution inhaled by people. Based on earlier models it is one of the key variables defining the risks caused by air pollution for cyclists and pedestrians.

@leandromtg

[leandromtg]

Some background: All physiological measures of PA intensity (oxygen consumption or uptake (VO2), kcal, heart rate, etc.) are strongly, positively, and linearly related among themselves. METs is just a way to express PA intensity standardized by people’s weight (i.e., resting metabolic rate) (1 MET = 3.5 ml O2.kg−1.min−1 = 1 kcal.kg−1·min−1). Minute ventilation (or ventilation rate (Ve), as we use in ITHIM) is strongly, positively, and exponentially related to intensity, but is not a measure of PA intensity.

Sources consulted:

• PubMed and Google Scholar using: ("ventilation rate" OR "ventilation minute" OR "minute ventilation" OR "pulmonary ventilation") AND ("oxygen uptake" OR exercise OR exertion OR "oxygen consumption" OR intensity)
• Exercise physiology books;
• Exercise physiologists;
• HEAT 4.0 documentation.

Findings:

• The relationship between Ve and VO2 is very well studied, but I did not find an equation to predict ventilation rate from intensity in healthy, non-athlete adults. In general, authors fit a function for each person in the sample, but they do not present the parameter values of every individual or considering the whole sample.
• Some books present a lookup table for broad PA categories, like rest and mild, moderate, heavy and maximal exercise, but it is hard to track sources and assumptions.
• Numbers in HEAT 4.0 are based on Audrey’s calculations using an algorithm presented in a 2002 EPA report on the agency’s exposure model (https://www.epa.gov/sites/production/files/2013-08/documents/report052202.pdf, Section 9, pages 9-1 to 9-23). More recent (2017) technical documentation about the latest model (APEX – https://www.epa.gov/fera/human-exposure-modeling-air-pollutants-exposure-model) is available at https://www.epa.gov/fera/apex-user-guides. However, accuracy and precision of estimates was not reported or commented, and the algorithm demands other individual-level data, such as body mass.

Suggestions (in order of my preference):

• Contact authors of one of (or both?) these two papers and ask for the required calculations or databases. • https://ehjournal.biomedcentral.com/articles/10.1186/1476-069X-8-48 (more recent and more transparent methods and results) • https://www.med.nagoya-u.ac.jp/medlib/nagoya_j_med_sci/6534/nj6533.pdf (focused on the question we have and performance of regression model seems better, but paper overall quality seems lower) • We could use Baba solution to Zuurbier data.
• Use/adapt APEX’s algorithm.
• Use lookup tables from books.

@markotainio, let's discuss that and see whether there are other options before taking the next steps?

[walkabilly]

This recent paper provides a potential solution. They have provided an equation for predicting minute ventilation from heart rate (https://www.sciencedirect.com/science/article/pii/S0966692317308451?via%3Dihub). I know the authors and an chatting with them about providing the data so we could do the same for METS and ventilation. Or we could convert HR to METS using Karvonen formula (solve for intensity - https://en.wikipedia.org/wiki/Heart_rate#Karvonen_method). Not sure anyone is checking this. I will repost in Slack!

[robj411]

The approach for this work has not been finalised, and its implementation in ITHIM-R remains as it was from the beginning.

The plan was to create a function to describe relationship between METs and inhalation rate.

Soren (Co-I on Metahit) might have data to share from a treadmill study, which could be used to create that function.

The other option would be to use e.g. EPA-guidelines and calculate ventilation rates from those (this is how Audrey estimated HEAT-ventilation rates).

Pros and cons of each approach: 1) using EPA-guidelines: equation is already published, so easy to implement. Has been used in similar context before. But uses more assumptions to go from MET to ventilation rates. 2) talking with Soren: prediction function created directly from energy expenditure, so cleaner and more transparent approach. But might lead to some delay to implement the solution.