Wildfire smoke particulates were linked to higher mortality after a lung cancer diagnosis, unit for unit more than other PM2.5

Design

Retrospective registry-based cohort, time-varying Cox model

Population

414,016 US patients aged ≥65 with lung cancer, SEER-Medicare 2006–2019

Primary outcome

All-cause mortality

Effect

Per 1 µg/m³: wildfire PM2.5 HR 1.078 (1.064–1.092); other PM2.5 HR 1.017 (1.015–1.019)

This retrospective cohort used SEER-Medicare data on 414,016 patients aged 65 or over with lung cancer diagnosed from 2006 to 2019, assigning each a three-year moving average of wildfire-related and other PM2.5 at their residential zip code.

Exposures were low by global standards — median 0.28 µg/m³ for wildfire PM2.5 and 8.11 µg/m³ for the rest. Each 1 µg/m³ rise in wildfire PM2.5 was associated with a 7.8% higher risk of death (HR 1.078, 1.064–1.092), against 1.7% for other PM2.5 (HR 1.017). The authors attribute about 700 excess deaths a year in this cohort to wildfire PM2.5 and about 3,500 to other PM2.5.

This is ecological exposure assigned by postcode, so individual exposure, indoor air and smoking history are imperfectly captured. It does add to evidence that air pollution matters after diagnosis, not only before. Many Indian cities have PM2.5 levels several times those studied here; whether the per-unit estimates hold at those levels is not known from this study.

• Advise patients with lung cancer to limit outdoor exposure on days of heavy smoke or very poor air quality.
• Well-fitting N95-type masks and indoor air filtration reduce personal particulate exposure.
• Ask about home cooking fuel and indoor smoke, which may outweigh ambient exposure for some patients.
• Air quality is a modifiable exposure worth including in survivorship advice.

Why it matters

It treats air pollution as a factor in survival after diagnosis, not only in who gets lung cancer.

Don't overread it

Exposure was assigned by postcode; the study cannot show that reducing an individual's exposure improves their survival.

The statistics, in plain English

A hazard ratio of 1.078 per µg/m³ is small per unit but applies to a whole population over years. Because exposure was estimated by postcode rather than measured per person, the true individual effect could be larger or smaller.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.