Only five studies qualified for a review of wildfire smoke and children's lungs

Wildfire smoke is one of the few climate-linked exposures that reaches tens of millions of people at once, and children are routinely named among the groups most at risk from it. A systematic review published this month set out to find the epidemiological studies supporting that claim specifically in children and adolescents. It found five [s1].

The search

The reviewers searched PubMed and Web of Science for studies through 30 September 2024, identifying 120 publications [s1]. Five met the inclusion criteria: three retrospective cohorts, one cross-sectional study and one case-crossover study [s1]. All five were conducted in the United States and Canada [s1]. Exposure was assessed in different ways across them — stationary monitors, satellite imagery and surveys [s1].

The included studies did not report effect measures in compatible units, so the reviewers could not pool them. Instead they used vote counting based on the direction of effect and statistical significance [s1]. That is a weak synthesis method, and the review says so: differences in exposure measurement, outcome definitions and age stratification limited cross-study comparability [s1].

What the five studies agreed on

Every included study reported increases in respiratory symptoms, hospital visits and medication use on days with significant wildfire smoke exposure [s1]. Consistency of direction across five independent studies using different exposure methods is meaningful — it is harder to explain away than a single result — but consistency of direction is not a magnitude. The review cannot say how much a given smoke day raises a child's odds of an asthma exacerbation, because the underlying papers do not permit that calculation in common units.

The review's conclusion is a call for standardised exposure assessment and age-specific analyses [s1]. This is the ordinary shape of an evidence base that is younger than the problem it is meant to describe.

The other half of the question: who, exactly

A second paper published a fortnight later, in Environment International, takes a different route to the vulnerability question, and its results show why age alone is an incomplete answer [s2].

The researchers used daily hospital admission and emergency department data for cardio-respiratory disease in California from 2006 to 2019, applying a time-stratified case-crossover design to estimate the effect of wildfire-specific fine particulate matter, then a generalized random forest to look for heterogeneity across 27 potential effect modifiers [s2]. Those modifiers spanned individual-level factors — age, sex, race and ethnicity — and area-level ones including poverty and racial and ethnic segregation [s2].

The headline association was modest: a 10 µg/m³ increase in the two-day moving average of wildfire PM2.5 was associated with an odds ratio of 1.014 (95% CI 1.012–1.016) for respiratory hospital admissions and emergency visits [s2]. No association was found for cardiovascular disease [s2].

The variation around that average was the point of the exercise. Associations between wildfire PM2.5 and respiratory disease varied strongly by individual and area-level factors, including air-conditioning prevalence and the Black/White dissimilarity index, a standard measure of residential segregation [s2]. Crucially, which modifiers mattered most changed across California's Air Basins — regions grouped by shared meteorology and geography [s2].

What that combination implies

Read together, the two papers describe a field that knows the direction of the harm and not much else with precision. The systematic review establishes that wildfire smoke days coincide with worse respiratory outcomes in children wherever anyone has looked, and that almost nobody has looked outside North America [s1]. The California analysis establishes that the size of the effect depends on a combination of individual and neighbourhood characteristics that does not generalise even between regions of a single state [s2].

That has a practical consequence for how smoke advisories and school closure policies are built. A single national threshold applied uniformly assumes a uniform exposure–response relationship. The California work suggests the modifiers driving vulnerability — cooling access, segregation, age structure — are locally specific enough that the same PM2.5 concentration does not carry the same risk in two different basins [s2].

Limits worth stating plainly

Neither paper is a trial, and neither can attribute an individual illness to smoke. The California study is a case-crossover analysis, which controls well for stable individual characteristics but relies on modelled wildfire-specific PM2.5 separated from other particulate sources — an inference, not a measurement. Its cardiovascular null result is a null in this dataset, not evidence of no cardiovascular effect; other studies have reported associations, and the paper does not settle that literature.

The systematic review's five studies are a small foundation for a global claim about children. The review's own framing is that wildfire events are increasing in frequency and intensity globally and will disproportionately affect vulnerable populations such as children [s1] — a reasonable expectation, but one that currently rests on evidence from two countries.

Sources

• Respiratory health outcomes of children and adolescents exposed to wildfire smoke: a systematic review — International Journal of Environmental Health Research, 2025-11-06
• Using generalized random forests to characterize vulnerability to adverse health outcomes following wildfire smoke exposure in California — Environment International, 2025-11-21