Their data is measuring the PM5 levels. What I’m asking about is how that difference gets translated to actual health risk. It’s related, but not in their data set (as far as I can tell on mobile).

There’s a distinction between measuring the difference in a plausible mechanism (PM5)and an outcome of interest (health). It’s easy to get caught up in the “mechanism porn” of scientific studies and extrapolate too far in terms of assumed outcomes.