Not really. That's not how we did this study, nor how science works. Glad to discuss further if you're actually curious.

Lol. That's not how science works, actually. We (I'm a co-author on the paper being discussed) actually do this work because we find it interesting! It has essentially no impacts on our salaries or office space.

On the other hand, the public audience is quickly bored with the dry details of scientific research. So the main reason these stories get glossied up with eye-grabbing phrases is that's the way media reports science, and in fact most things (right?). If you want the straight facts, you can just read the scientific literature.

Co-author of the paper here...
You are correct that biological systems do not respond linearly. The statistical association of premature deaths with exposure to particulate matter is log-linear. This association has been known since the 90's and is quite robust. I don't expect you to read the actual paper, or to be an expert in air-pollution epidemiology. But don't worry, we are not idiots! Glad to answer further questions if you're interested in learning more. I agree the notion of "premature deaths" is complex.

We mention that specifically in the paper (I'm a co-author). In Europe, the human health impact of the excess diesel NOx emissions is estimated to be 1/4th that of the total impact of all land-transportation sources of pollution (as estimated in Silva et al., Environ. Health Perspect., 2016). I think that's a pretty significant portion of the problem to be coming just from this particular source!

Co-author of the paper checking in here, to clarify...

It's more that there is a very strong, robust statistical association between exposure to fine particular matter (to which emissions of NOx can ultimately contribute, by way of particulate nitrate) and premature mortality. This is based on regressions (not necessarily linear, but you're on the right track) of health data and ambient pollutant concentrations, when other co-founders are accounted for.

This association was first noticed in some famous studies (aka the "6 cities studies") in the US in the 90's and has been repeated 100's of times since then. Essentially, exposure to 10 ug/m3 of fine particulate matter over the course of a year can increase your chances of death owing to heart disease and lung cancer by several percent. This information was used in the current paper being discussed, but not something that we contributed to here (i.e., we used results from the literature). What we did do, however, was a lot of super-computer simulations to estimate how much particulate nitrate (and ozone) are formed from the diesel NOx emissions, and where they are formed, in order to calculate (given previously identified concentration-response relationships) impacts on human health.

Hi there, I'm a co-author on the paper being discussed, here to clear up some confusion. bluefoxlucid is correct that the concentrations of NOx owing to diesel emission are not high enough to be very harmful, and that the main health concern associated with air pollution is the formation of fine particulates (smaller than 2.5 micron). However, you're missing an understanding of atmospheric chemistry -- NOx reacts with other compound in the atmosphere to form nitric acid (HNO3), which lead to particulate nitrate (NO3-). So it is the formation of these particles from the NOx that is of concern. NOx also leads to the formation of O3 (which is a compound formed in the atmosphere -- it is not directly emitted by cars or any other human activity), which is also of concern for human health (to a lesser extent than particles) and crops (also discussed in our article). Hope that helps makes sense. I think there's more of a case of bad arm-chair science review happening here than bad science.