Not All Smoke Is the Same Smoke

Air quality is measured by weight. PM2.5 — fine particulate matter under 2.5 micrometres — is reported in micrograms per cubic metre, and the underlying assumption of nearly every air quality index is that a microgram is a microgram. Traffic exhaust, industrial emissions and wildfire smoke all get counted the same way.

That assumption has been tested, and it does not hold.

Researchers examined 14 years of hospital admissions across Southern California, covering roughly 1.6 million respiratory admissions, and separated the PM2.5 attributable to wildfire smoke from PM2.5 arising from other sources. The comparison is unusually clean: the same population, the same hospitals, the same measurement methods, differing only in where the particulate came from.

Per equivalent increase in concentration, wildfire-derived PM2.5 was associated with a substantially larger increase in respiratory hospitalisations than PM2.5 from other sources. The effect was not marginally larger. The published estimate places it up to roughly an order of magnitude higher, depending on the model specification.

The mechanistic explanation is plausible enough. Wildfire smoke is not simply carbon. Burning vegetation, and increasingly burning structures, produces a chemically complex aerosol carrying oxidised organic compounds, transition metals and, where the fire has reached the built environment, combustion products of plastics, treated timber and insulation. Particles of the same mass can differ enormously in what is bound to their surface and how aggressively they provoke oxidative stress and inflammation in airway tissue.

The honest limits matter here. This is observational data. Exposure was estimated from monitoring networks and smoke plume modelling rather than measured at each person's location, which introduces error. Behaviour changes during smoke events — people stay indoors, cancel activity, seek care differently — and those changes are difficult to separate fully from the exposure itself. The order-of-magnitude figure is the upper end of a range, not a settled constant, and it applies to respiratory admissions rather than to mortality or long-term outcomes.

What survives those caveats is the direction and the rough scale. An air quality reading of 40 µg/m³ during a wildfire episode does not carry the same biological meaning as 40 µg/m³ on a still day in heavy traffic, and treating the two as interchangeable understates the risk of the first.

The practical implication is about thresholds rather than absolutes. The behaviour most indices encourage — glance at the number, act if it crosses some line — is calibrated on urban background pollution. During a smoke event, the same number warrants a more conservative response: filtration indoors rather than ventilation, a properly fitted respirator rather than a surgical mask if outdoors is unavoidable, and genuine caution with strenuous outdoor exercise, which multiplies the inhaled dose several-fold through increased minute ventilation.

Why this matters for tracking: if you log environmental exposure alongside symptoms or training load, the source of the particulate is worth recording, not just the concentration. A week of wildfire smoke and a week of urban haze at identical readings are not equivalent inputs, and a symptom log that treats them as the same will find less signal than it should.