The Air Never Arrives One Pollutant at a Time, but the Rules Pretend It Does

For years I treated air pollution as a single number. The little colored tile in my weather app, one figure, good or amber or red, and I planned my morning run around it. What I never once asked was which pollutant that number was standing in for, or what my airways were quietly doing with everything it left out. That lazy habit turns out to be baked into the rules that are supposed to protect me, and into a whole body of asthma research that kept coming back with a shrug.

On September 21, a team writing in Environmental Health did something deceptively simple. Instead of asking whether one pollutant raises the risk of adult asthma, they treated the air as the mixture it actually is and looked at the combined, long-term load on adults with active asthma. Their starting complaint: most studies examine pollutants one at a time, which cannot capture what a real lung meets on a real street. Nobody stands at a bus stop inhaling pure fine particulate with the ozone politely subtracted out.

Go looking at the classic single-pollutant studies and the signal is oddly weak. The big European ESCAPE analysis followed 23,704 adults across six cohorts and counted 1,257 new asthma cases, and almost every pollutant it tested landed borderline or null. NO2 came in at an odds ratio of 1.10, confidence interval 0.99 to 1.21. PM2.5 sat at 1.04, interval 0.88 to 1.23. The authors’ own honest phrase was “suggestive but not firm evidence.” For pollutants we are sure damage lungs, that is a strangely quiet result. So why does the smoking gun keep looking like a water pistol?

Because the method was fighting the biology. A single-pollutant model statistically holds every other pollutant constant while it wiggles one up and down. But pollutants do not rise and fall independently in the real world. Traffic pours out particulates and nitrogen dioxide together, and a hot, stagnant day cooks up ozone on top of both. Freeze the others in place and you have measured a scenario that never happens on any street I have ever run down.

The mixture approach is built to close exactly that gap. The tool these newer studies reach for is called quantile g-computation, and once I understood what it does the whole thing clicked. Instead of isolating one chemical, it nudges the entire mixture up one notch together, the way real air gets dirtier, then apportions the blame, telling you how much of the combined hit each pollutant is carrying. Ask the question that way and the numbers wake up. A prospective cohort using the same method followed 8,490 adults aged 50 and up for 6.9 years and found that pushing PM2.5 and ozone up together raised the risk of asthmatic symptoms by 18 percent (hazard ratio 1.18, confidence interval 1.05 to 1.33), with PM2.5 shouldering 68 percent of that combined load and ozone the other 32 percent. Different cohort, different method, but the same two pollutants that looked limp on their own snapped into a clear signal the moment a model let them act together.

Wait, why would the same particle behave so differently depending on who else is in the room with it? PM2.5 is small enough to ride deep into the alveoli and sit there, driving oxidative stress and inflammation in the airway wall. Ozone is a reactive gas that oxidizes the lining directly, tearing at the same tissue by a different route. They are not two separate insults your lung tallies on separate ledgers. They converge on overlapping inflammatory pathways, so the damage compounds instead of simply adding up. And the fingerprint of that convergence may run deeper still: researchers have started tracing DNA methylation as a potential mediator between air pollution exposure and how well asthma stays controlled, which would mean the dirty air is not only irritating tissue but leaving marks on the genes that set how fiercely the airway inflames. So the mixture may not just hit harder. It may be tampering with the switchboard.

So where does that leave those of us who cannot relocate our lungs to a mountaintop? Food is not a substitute for clean air, and I will not pretend otherwise. But if the shared mechanism really is oxidative stress, then what you eat plausibly enters the fight, and the evidence here is not nothing: a body of work on diet and air pollution as interacting exposures suggests an antioxidant-rich diet may buffer some of the oxidative damage pollution drives through the airway. I am not selling you a supplement. The narrower, honest claim is that a diet dense in real antioxidant-bearing food gives an ordinary person one small lever to pull while the institutions lag.

And the lag is what ought to make you a little angry. Every air-quality standard we live under is written pollutant by pollutant. The EPA sets separate national limits for six criteria pollutants, particulates here, ozone there, nitrogen dioxide somewhere else, each policed on its own. That is precisely the single-pollutant framing the newer science says undercounts real harm. When the regulatory apparatus measures the mixture as a set of unrelated numbers, it produces the tidiest, smallest, most reassuring version of the risk, then builds the rules on top of that flattering picture. The mixture evidence raises the possibility that the combined burden on your airways is worse than any single one of those standards admits. That is not a reason to panic. It is a reason to stop letting an official number, siloed by design, tell you the air is fine when your chest disagrees.

So what do I actually do with this? I stopped trusting the single tile in the app. That headline number only ever reports one pollutant, whichever happens to be worst that hour, so I started reading the separate PM2.5 and ozone lines side by side, and on the hot, stagnant days when both climb at once I move my run indoors and let the filter carry the load instead of my alveoli. I would not wait for a regulator to catch up to a lung that already knows.

Sources

1. Environmental Health – Combined effects of long-term air pollution exposures on adult active asthma: a quantile g-computation analysis (2026)
2. Prospective cohort – joint effects of PM2.5 and ozone on asthmatic symptoms via quantile g-computation
3. ESCAPE – ambient air pollution and adult asthma incidence in six European cohorts
4. Clinical Epigenetics – DNA methylation as a potential mediator between air pollution exposures and asthma control (2025)
5. Current Allergy and Asthma Reports – Air Pollution and Diet: Potential Interacting Exposures in Asthma (2023)
6. U.S. EPA – National Ambient Air Quality Standards (NAAQS) table
7. AirNow – Air Quality Index basics