Respiratory health outcomes of children and adolescents exposed to wildfire smoke: a systematic review

Abstract

Wildfire events are increasing in frequency and intensity globally, partly due to climate change. This emerging public health crisis will disproportionately impact vulnerable populations such as children. Epidemiological studies link wildfire smoke, especially fine particulate matter (PM2.5 and PM10), with adverse respiratory outcomes; yet few focus specifically on pediatric populations. This systematic review examines wildfire smoke impacts respiratory health in youth populations by analyzing studies identified through a comprehensive literature search in PubMed and Web of Science through 30 September 2024. Of 120 publications identified, five met the inclusion criteria: three retrospective cohorts, one cross-sectional, and one case-crossover study. Studies were conducted in the U.S. and Canada, using various exposure assessment methods including stationary monitors, satellite imagery, and surveys. Not all studies reported compatible effect measures, vote counting based on the direction of effect, and statistical significance was applied. All studies reported increases in respiratory symptoms, hospital visits, and medication use on days with significant wildfire smoke exposure. Differences in exposure measurement methods, health outcome definitions, and age stratifications limited cross-study comparability. Despite limitations, the review found consistent evidence linking wildfire smoke exposure to worsened respiratory health in children. Further research using standardized exposure assessments and age-specific analyses is needed.

Introduction

Due in part to the climate crisis, wildfires are increasing annually in number and severity, both in the U.S. and globally. Recent record-setting wildfire seasons in regions of North America have imposed immense costs on human health and lives (McFarla Nd et al. 2025). In addition to increased severity, the wildfire season has increased in length, spanning April through October in 2023 (Jain et al. 2024). Record breaking number of areas burned, extended wildfire seasons, and fire frequency can be attributed to several environmental factors including early snowmelt, reduced snowpack, multiannual drought conditions, increased temperatures and fuels (McFarla Nd et al. 2025; Jain et al. 2024; Wasserman and Mueller 2023). The onset of anthropogenic climate change has largely contributed to increased wildfire activity (Jain et al. 2024). For example, longstanding U.S. Forest Service fire suppression policies have resulted in an unnatural buildup of fuels, further intensifying wildfire behaviour (Kreider et al. 2024). Collectively, these climatic and anthropogenic factors have disrupted historical fire patterns and led to more frequent, severe wildfire events, in turn posing heightened risks to air quality, population exposure, and acute and long-term public health outcomes.

Climatic events, including wildfires, place significant burdens on human health, especially vulnerable populations. Because wildfire smoke is associated with human health, it has become an increasing public health concern. Epidemiological studies across a wide range of populations have established associations between poor air quality and human health. However, few studies have looked at the impact of wildfire smoke exposures on children (Aguilera et al. 2021; Künzli et al. 2006; Lipner et al. 2019; Moore et al. 2023; Stowell et al. 2019). Due to our warming climate, exposure to wildfire smoke is projected to increase into the future. With more children exposed to wildfire smoke annually, there is a clear need to better understand the impacts on this sensitive population.

Wildfires emit particulate matter (PM) and air pollutants into the air, which can lead to adverse health effects following exposure. Fine particulate matter (PM2.5, airborne particles < 2.5 in aerodynamic diameter) is of special concern due to its ability to enter the human respiratory system and enter the blood-stream (US EPA 2024). PM₁₀ and PM_2.5 are both concerns, however, PM₁₀ is generally less worrisome due to its larger particle size. PM₁₀ typically originates from mechanical processes like mining, road dust, and street sanding, whereas fine particulate matter (PM_2.5) is more often produced by combustion sources such as wildfires (Davies et al. 2018). Particulate matter contains microscopic solids or liquid droplets that are small enough to be inhaled and causes serious health problems due to short- and long-term exposure (Reid et al. 2016). Fine particles found in wildfire smoke are respiratory irritants, and exposure can cause persistent coughing, phlegm, wheezing, and difficulty breathing. Globally, wildfire smoke is estimated to cause over 339,000 premature deaths a year (Black et al. 2017; Reid et al. 2016). Wildfire smoke exposure is a concern for vulnerable populations, including children, the elderly, or those with pre-existing respiratory and cardiovascular diseases (US EPA 2024). Although PM_2.5 exposure can lead to more severe impacts such as cancer, and even death, this review will solely focus on exposure to wildfire smoke PM_2.5.

Children are especially vulnerable to the effects of wildfire smoke. However, very few studies have intentionally focused on pediatrics as a target population (Liu et al. 2015). Risk is generally higher in pediatric populations due to increased exposure as a result of breathing more air relative to their body weight as well as still developing respiratory systems. Wildfire smoke can have short- and long-term health effects. It is expected that smoke can lead to worsened respiratory health, increasing hospitalizations, and lowering overall lung function. Additionally, a growing body of literature suggests that exposure to particulate matter may have neuropsychological and growth impacts to children (Oliveira et al. 2019; Suades-González et al. 2015).

There are a range of studies that have explored the health impacts of wildfire smoke on the general population (Black et al. 2017; Reid et al. 2016). Yet, few studies have looked at the impacts of wildfire smoke on youth respiratory health outcomes, especially acute and chronic respiratory health outcomes. This study aims to systematically review and summarize the current peer-reviewed literature on the impact of wildfire smoke exposure on respiratory health in children, less than 21 years old.

The review will include the literature that currently exists within the field. The goal of this review is to effectively highlight gaps in the literature. Additionally, the systematic review aims to determine whether there is a shared understanding of the impacts of wildfire smoke on respiratory health outcomes in children. This systematic review will answer the research question, “How does wildfire smoke impact respiratory health outcomes in youth populations?” In conducting this review, we developed a PECO (Population, Exposure, Comparator, and Outcomes) statement to better guide the research and selection strategies (Morgan et al. 2018).

Population:

Children, youth, adolescents, pediatrics (Age < 21 years)

Exposure:

Particulate matter (PM_2.5 and PM_10), Wildfire smoke assessments; satellite data, temporal and spatial patterns, survey and questionnaires

Comparator:

Populations not exposed to wildfire smoke or baseline levels of air quality (e.g. seasonal or regional air quality without wildfire events)

Outcome:

Respiratory health, acute or chronic

Methods

Literature search strategy

PubMed and Web of Science databases were searched for relevant primary epidemiological studies published through 30 September 2024. Detailed information on the search terms is provided in Table 1. The systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al. 2009). The PRISMA flow chart is presented in Figure 1. The search combined a list of keyword synonyms for wildfires and respiratory health with a modified search for pediatric studies. No date or study design limits were used.

Table 1.

Keywords for systematic search

PubMed Search	Search Terms
PubMed 1	(Wildfire smoke AND Wildfire exposure AND Respiratory health)
PubMed 2	OR (Particulate matter OR Respiratory outcomes)
PubMed 3	AND (allchild AND Wildfire)
PubMed 4	AND (allchild)
Web of Science 1	(Wildfire smoke OR Wildfire OR Wildfire exposure)
Web of Science 2	AND (respiratory health OR respiratory outcomes OR respiratory)
Web of Science 3	AND (respiratory health OR respiratory outcomes OR respiratory)
Web of Science 4	AND (pediatric OR child OR children)

Figure 1.

Prisma flow diagram of the selection of studies.

Source: Page MJ, et al. BMJ 2021;372:n71. doi: 10.1136/bmj.n71.

This work is licensed under CC BY 4.0. To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/

Study eligibility criteria

Peer-reviewed primary research publications on wildfire smoke and pediatric respiratory health published through 30 September 2024 were systematically reviewed. The search was limited to peer-reviewed English-language and excluded papers written in other languages. Articles reporting research on the topic of human health and wildfire smoke were reviewed. As this review focuses on wildfire smoke, the research team also excluded studies that investigated non-smoke-related morbidities. To be included in the evaluation, studies must have included and independently described a population of children between 0–21 years old. Thus, researchers focused on wildfire smoke and its respiratory health impacts on the pediatric and adolescent population. Exposures were characterized as specific exposure to smoke produced by wildfires; the search was not limited by vegetation burned or combustion conditions. Measures of exposure were direct, utilizing existing data from local air sampling devices, or indirect through satellite imagery (i.e. visibility of smoke plumes). The exposed populations were determined by postal or zip code, county, or address of residence. Outcomes included emergency department and urgent care visits, self-reported medical questionnaires, measures of lung function and asthma control, non-urgent clinic visits, and respiratory symptoms. Studies that exclusively reported pregnancy and birth-related outcomes were excluded in this review, as well as studies outside of the U. S. and Canada. Studies outside of the U.S. and Canada were excluded due to differences in wildfire exposure patterns, healthcare systems, and environmental monitoring practices, which may limit comparability and generalizability to the North American context.

Additionally, studies that focused on non-smoke-related outcomes were excluded, given that our aim was to report respiratory health outcomes following wildfire smoke exposures. Included study designs were observational studies (retrospective cohort, cross-sectional, and case-crossover). All studies underwent title and abstract screening for relevance to the topic, followed by a full-text review. If duplicate studies were presented, the most recent article was used, and the remainders excluded.

Study selection

As part of our methodology, we independently reviewed the studies included in the systematic review; all records were screened using predetermined criteria. Primary epidemiological studies that examined the association between wildfire smoke and respiratory health outcomes in children were eligible for inclusion in the review. Information regarding study characteristics (study date, location), population characteristics (age), exposure characteristics (exposure measures including number of smoke days, satellite imagery, air quality data sources, PM_2.5, and PM₁₀ sources), and outcome assessment (symptoms, outpatient clinic, and ED visits or hospitalizations for respiratory-related illness) were collected. The research team did not exclude studies based on the type or diversity of vegetation burned. A wide array of study designs and geographies were included. Studies were excluded if they did not have methods for tracking PM_2.5 and PM₁₀ air quality sources. Exposure scenarios were excluded if they did not track particulate matter concentrations (PM_2.5 and PM₁₀₎ and exposures to wildfire smoke (i.e. satellite imagery, questionnaires).

Study selection took place in three stages, by an independent reviewer. All studies underwent 1) title and abstract screening for relevance, 2) full-text screening, and 3) exposure assessment screening for wildfire smoke and PM_2.5 and PM₁₀ air quality measures. Study screening results are presented in the PRISMA flow diagram (Figure 1). No automation tools were used in the selection process.

Study quality

Risk of bias was evaluated by the primary investigator. Using the Cochrane guidance to inform judgment on bias, risk was defined as either low, probably low, probably high, or high risk of bias. In addition to the Cochrane guidance, the Navigation Guide methodology was applied, involving the following four steps: Specify the study question, Select the evidence, Rate the quality and strength of the evidence, Grade the strength of the recommendations (Woodruff and Sutton 2014). Criteria for risk, low, probably low, probably high, or high risk, are defined in Supplementary File 1 (Lam et al. 2016).

The following eight risk of bias domains were evaluated for this systematic review: 1) recruitment strategy, 2) blinding, 3) exposure assessment, 4) outcome assessment, 5) confounding, 6) incomplete outcome data, 7) selective outcome reporting, and 8) conflict of interest. The risk of bias assessment criteria used in Lam et al. (2015) were implemented to answer the research question. For this review, ratings were low, probably low, probably high, or high risk of bias for the following 9 parameters: 1) recruitment strategy, 2) blinding, 3) exposure assessment, 4) outcome assessment, 5) incomplete outcome data, 6) potential confounding factors, 7) selective outcome reporting, 8) conflicts of interest, and 9) other problems that could put our review at risk of bias. Study characteristics and risk of bias are summarized in table format. Risk of bias is reported in tabular format (Table 2).

Table 2.

Characteristics of included studies on wildfire smoke and respiratory health outcomes in children, by publication and study design

Study	Study Design	Study Population & Location	Sample Size	Exposure Assessment	Outcome Assessment
Moore et al., 2023	Retrospective cohort study	Aged 1–17 years, Calgary, Canada	n = 57,375	Postal codes in the health record at the time of emergency department visits. Hourly PM2.5 levels from ground-level monitoring sites. Wildfires were identified by overlaying information from the Government of Alberta “Access Air Quality and Deposition Data” website, news reporting on wildfire smoke coverages, and satellite imagery.	Emergency department (ED) visits for asthma were obtained from ambulatory data.
Aguilera et al., 2021	Retrospective cohort study	Aged ≤19 years, San Diego County, California	n = 884,471	Zip code–specific concentrations of PM2.5 using 24-hour daily means sampled and analyzed by the US EPA Air Quality System. Wildfires were identified with smoke plume data sets from NOAA’s Hazard Mapping System (HMS) to identify zip code days exposed to wildfire smoke.	Visits to emergency and urgent care facilities of Rady’s Children Hospital for the following respiratory conditions: difficulty breathing, respiratory distress, wheezing, asthma, or cough.
Nino Künzli et al., 2006	Cross-sectional	Aged 17–18 years and 6–7 years, Southern California	n = 873 (17–18 yr); n = 5,551 (6–7 yr)	Objective smoke measurements estimating concentrations of PM10 during the 5 days with the highest fire activity. Fire questionnaires were administered on exposure to fire smoke and personal measures taken to modify this exposure.	Questionnaire-based investigation of wildfire smoke exposure and occurrence of symptoms.
Lipner et al., 2019	Retrospective cohort study	Aged ≤21 years, Denver, Colorado	n = 1,799	Wildfire smoke–related PM2.5 at patients’ residential ZIP codes characterized using satellite-derived smoke polygons from NOAA’s Hazard Mapping System combined with ground-based US EPA monitors.	Forced Expiratory Volume in 1 Second (FEV1) and the Asthma Control Test (ACT) and Children’s Asthma Control Test (CACT) scores, during non-urgent clinic visits.
Stowell et al., 2019	Case-crossover	Aged, 0–65+ years; Children (0–18 years) Colorado, United States	n = 490,368 total n = 94,022 (0–18 years)	PM2.5 from ambient sources estimated from high-resolution satellite aerosol optical depth (AOD) and ground measurements obtained from the US EPA.	Emergency department visits and hospitalizations for acute cardiorespiratory outcomes.

Analysis

To summarize the overall direction and statistical associations across studies, we employed vote counting methods. For each study, we recorded whether the reported effect of wildfire smoke exposure on pediatric respiratory outcomes was positive, negative, or not significant for each variable. Effect estimates included odds ratios, relative risks, incidence rate ratios, or percent change per unit PM_2.5/PM₁₀, and statistical significance was assessed using reported confidence intervals. Studies were categorized by exposure type (PM _2.5, PM ₁₀, wildfire smoke days) and outcome type (e.g. asthma exacerbations, ED visits, cough, wheezing). For each variable, analyses cast a “vote” for a positive, negative, or non-significant association, with significance levels documented. Vote counts were then tallied to provide a descriptive summary of the consistency in reported effects. The analysis is repeated for all health outcomes of interest. Vote counting provides an insightful starting point in regard to systematic assessment of studies within this research area, this methodology has been popular among public health research (McKenzie and Brennan 2024).

All analyses were conducted with R v4.5.1 and RStudio.

Results

Search results

The electronic literature search identified 120 studies and 105 after duplicate removal. The detailed study selection process is outlined in Figure 1. After title and abstract screening, 94 of the 105 identified studies were excluded. Of the 13 included in full-text screening, one was excluded due to no respiratory or healthcare outcomes, three due to no pediatric data, one was excluded due to studying two variables on respiratory health outcomes, one due to seasonal extreme temperatures (not wildfire smoke), and two solely used particulate matter to define wildfire smoke. The five remaining studies included in this review are summarized in Table 3.

Table 3.

Characteristics of included studies on wildfire smoke and respiratory health outcomes in children, by publication and study design.

Study characteristics

These five studies we focused on were published between 2006 and 2021, and include three retrospective cohort studies, one case crossover, and one cross-sectional study. All studies were conducted in North America, with the majority of the studies focusing on a single city or county (Calgary, Alberta, Canada (Moore et al. 2023), San Diego County (Aguilera et al. 2021), Denver, Colorado (Lipner et al. 2019)) or region (Southern California (Künzli et al. 2006), Colorado, USA (Stowell et al. 2019)). These five studies investigated both urban and rural populations.

Data for 1,044,091 children under the age of 21 years were reported for the five studies. None of these studies evaluated preschool-aged children separately. One study evaluated an existing pediatric cohort for children between the ages of 6–7 and 17–18 in Southern California regarding symptoms during wildfire smoke exposure (Künzli et al. 2006). The remaining studies were population-level, utilizing medical care provider or government databases focusing on respiratory causes for hospitalization or emergency department visits (Aguilera et al. 2021; Lipner et al. 2019; Moore et al. 2023; Stowell et al. 2019). A single study relied on self-reported data through the use of surveys to identify health outcomes (Künzli et al. 2006).

While two studies utilized the International Classification of Diseases and Related Health Problems, 9th/10th Revision (ICD-9/ICD-10) codes to identify outcomes (Moore et al. 2023; Stowell et al. 2019). One study relied on the visit’s chief complaint compared to using the International Classification of Diseases, Ninth Revision and International Classification of Diseases, 10th Revision diagnostic codes (Aguilera et al. 2021). The final study included in the systematic review relied on measures of lung function and asthma control through the Forced Expiratory Volume in 1 Second (FEV1) and the Asthma Control Test (ACT) and Children’s Asthma Control Test (CACT) test scores (Lipner et al. 2019). The most reported outcomes were ED visits in three of the studies (Aguilera et al. 2021; Moore et al. 2023; Stowell et al. 2019), hospitalizations in one (Lipner et al. 2019), and one which relied on self-reported symptoms (Künzli et al. 2006). Symptoms were reported specifically as diagnostic codes, including but not limited to asthma, bronchiolitis, bronchitis, pneumonia, and upper respiratory tract infection (Aguilera et al. 2021; Künzli et al. 2006; Lipner et al. 2019; Moore et al. 2023; Stowell et al. 2019).

Comparison groups differed between these studies. Most were conducted before and after analysing healthcare visits during the timeframes directly preceding and/or following wildfire events (Aguilera et al. 2021; Künzli et al. 2006; Moore et al. 2023; Stowell et al. 2019). However, one study compared periods with exposure to wildfire smoke-related PM2.5 and periods without such exposure for the same pediatric asthma patients (Lipner et al. 2019).

Exposure assessment

Studies differed in how wildfire smoke exposure was reported. Particulate matter was most common, utilizing either PM_2.5 (Lipner et al. 2019; Stowell et al. 2019; Aguilera et al. 2021; Moore et al. 2023) or PM₁₀ data (Künzli et al. 2006). Particulate matter was measured using local sampling devices, often through access to government or agency-based air quality monitoring programs.

Most studies compared periods of wildfire smoke exposure to daily PM_2.5 concentrations from non-wildfire sources (Aguilera et al. 2021; Moore et al. 2023). The studies examined pediatric respiratory visits associated with wildfire-specific PM_2.5 and contrasted these with visits related to PM_2.5 from other sources, enabling a comparative analysis of the health impacts of different PM_2.5 origins. Each of the five studies included a second measuring system, specifically focusing on particulate matter. Several studies added satellite data associated with wildfires (Lipner et al. 2019; Stowell et al. 2019; Aguilera et al. 2021; Moore et al. 2023). Other measures of exposure included the report of “smell of fire smoke indoors” (Künzli et al. 2006), meteorological data (Stowell et al. 2019), and fire count data (Stowell et al. 2019). Exposures were reported as a daily average in all studies (Künzli et al. 2006; Lipner et al. 2019; Stowell et al. 2019; Aguilera et al. 2021; Moore et al. 2023). Lag times between exposure and visit varied between studies, with some as same-day results (Aguilera et al. 2021) and others varying between 0–33 days after smoke events (Künzli et al. 2006; Lipner et al. 2019; Stowell et al. 2019; Aguilera et al. 2021; Moore et al. 2023). One study reported only a lag time of three days for respiratory health outcomes (Stowell et al. 2019). Additionally, a single study reviewed cases up to three days before a smoke event (Moore et al. 2023).

To enhance clarity and comparability across studies, Table 4 summarizes the statistical modeling approaches used in each of the five studies. The most frequently investigated outcomes across studies were emergency department (ED) visits or hospitalizations for respiratory conditions (three studies) (Stowell et al. 2019; Aguilera et al. 2021; Moore et al. 2023). Several studies additionally examined pediatric respiratory visits or asthma-related exacerbations (Lipner et al. 2019; Aguilera et al. 2021) or respiratory symptoms based on self-reported smoke exposure (Künzli et al. 2006). One study focused specifically on lung function (FEV1) and asthma control (ACT/CACT) measures (Lipner et al. 2019). Across studies, a variety of modeling approaches were used, including Poisson regression, mixed-effects models, panel linear regression, multilevel logistic regression, and conditional logistic regression, with most studies incorporating meteorological variables, temporal trends, and fixed or random effects to control for confounding.

Table 4.

Analytical approaches of included studies on wildfire smoke and respiratory health outcomes in children, by publication.

Study	Model Type	Covariates	Estimated Associations/Effect Estimates
Moore et al., 2023	Poisson regression	CAAQS PM2.5 zone categories (green, yellow, orange, red); smoke days	Incidence rate ratios (IRRs) and 95% CIs for ED visit rates compared to green zone (reference). Daily ED visit rates were calculated per CAAQS zone.
Aguilera et al., 2021	Ordinary Least Squares (OLS) linear panel regression with zip code fixed effects	Wildfire-specific PM2.5 concentrations; day-of-week effects; month-of-year effects; linear time trend; zip code fixed effects. Sensitivity analyses removed fixed effects and examined lags (1–3 days) and age strata (<6, 6–12, 13–19).	Rates of pediatric respiratory visits (per 100,000) were regressed on wildfire-specific PM2.5. Analyses also conducted for total and nonsmoke PM2.5. Fixed effects controlled for baseline risk and population composition.
Künzli et al., 2006	Mixed-effects logistic regression (multilevel approach)	Sex, ethnicity, parental education, pre-fire asthma status, cohort (elementary vs. high school). Analyses stratified by asthma status.	Associations between reported indoor fire smoke exposure and respiratory symptoms were estimated by separating within-community and between-community components. Results compared across exposure groups: no fire smoke, 1–5 days, ≥6 days. Secondary models replaced exposure with 5-day mean PM10.
Lipner et al., 2019	Mixed-effects models with random intercept by subject (Gaussian for FEV1, binary for ACT/CACT)	Calendar month; nonlinear terms for age (natural spline, 3 df); mean daily temperature (lags 0–1, spline); precipitation (lags 0–1); ozone and nonsmoke PM2.5 (lags 0–3 for FEV1; averaged over lags 0–33 for ACT/CACT).	Associations between wildfire smoke exposure and both FEV1 and asthma control (ACT/CACT) were estimated. Statistical significance assessed via 95% CIs compared to null (0.0). Models accounted for repeated measures within individuals.
Stowell et al., 2019	Conditional logistic regression	3–day mean PM2.5 (smoke, nonsmoke, total); 3–day mean temperature; spline for day of year (2 nodes/year)	Associations between PM2.5 exposure and ED visits/hospitalizations estimated for respiratory outcomes (3–day avg) and cardiovascular outcomes (2–day avg).

Outcomes

Reported outcomes and effect direction for respiratory health outcome and type of visit are summarized in Table 5 (Boon and Thomson 2021). The most frequently reported outcome was Emergency Department/Urgent Care visits for asthma (four studies) (Lipner et al. 2019; Stowell et al. 2019; Aguilera et al. 2021; Moore et al. 2023), or for any respiratory symptom (three studies) (Künzli et al. 2006; Stowell et al. 2019; Aguilera et al. 2021). Several studies looked at both asthma and respiratory health outcomes (Künzli et al. 2006; Stowell et al. 2019; Aguilera et al. 2021). Outpatient medical clinic visits were noted for one study (Lipner et al. 2019). Despite all studies investigating the effects of wildfire smoke exposure on asthma outcomes, only one study had a positive association to emergency department/clinic visits (Stowell et al. 2019). However, when wildfire smoke days were not compared to air quality monitoring standards (green, yellow, orange, red) Moore et al. 2023., identified that people were 13% (95% CI 1.02–1.24) more likely to go to the ED for an asthma exacerbation compared to the green zone, or clean air days.

Table 5.

Reported respiratory outcomes of included studies: Effect direction

					Respiratory Outcomes: Effect Direction
Study	Study Design	Risk of Bias	Wildfire Measure	Age (Years)	ED or Clinic Visits	Symptoms
Moore et al. (2023)	Retrospective cohort study	NA	Satellite, PM2.5	1–17	ED	↔ AV
Aguilera et al. (2021)	Retrospective cohort study	NA	Satellite, PM2.5	≤19	ED	↔ AV, ↑ RV
Künzli et al. (2006)	Cross-sectional	Attrition, outcome assessment	Smell of fire smoke, PM10	6–7 17–18	Clinic Visits	↔ AV, ↑ RV
Lipner et al. (2019)	Retrospective cohort study	NA	Satellite, PM2.5	≤21	Clinic Visits	↔ AV
Stowell et al. (2019)	Case-crossover study	NA	Satellite, PM2.5	0–18	ED	↑ AV, ↔ RV

NA = not applicable, ED = emergency department, RV = respiratory visits, AV = asthma visits, ↑ = positive association, ↓ = negative association, ↔ = no significant association

Following count-based analyses the five included studies suggest consistent evidence for positive associations between worsened respiratory health outcomes and wildfire smoke exposure (Figure 2). Across the included literature, strength of association varied with exposure and specific health outcomes, but consistent patterns emerged. Künzli et al. (2006) reported the highest number of statistically significant associations between wildfire smoke exposure, deemed as “fire smoke smelled” and respiratory health outcomes. Out of 30 outcomes examined, 20 demonstrated significant positive associations (see Table 6). Künzli et al. (2006) reported outcomes for each of the 3 lag periods (1–5 days, > 6 days, and scaled to the contrast in PM₁₀ between the communities with the highest and lowest levels, respectively (∼ 210 vs. 30 μg/m³).

Figure 2.

Vote count summary of statistical associations by respiratory health outcome and wildfire exposure by study.

Table 6.

Summary of reported associations between wildfire smoke exposure and pediatric respiratory outcomes across studies.

Study	Outcome	Exposure	Lag	Effect Estimate	95% CI	Association
Aguilera et al., 2021	Asthma	Wildfire smoke PM2.5	per 10	0.21%	−2.11 – 2.53	No Association
Aguilera et al., 2021	Cough	Wildfire smoke PM2.5	per 10	21.80%	3.6 – 40.1	Positive
Aguilera et al., 2021	Difficulty breathing	Wildfire smoke PM2.5	per 10	7.30%	−3.8 – 18.4	No Association
Aguilera et al., 2021	Respiratory distress	Wildfire smoke PM2.5	per 10	0.44%	−1.3 – 2.2	No Association
Aguilera et al., 2021	Wheezing	Wildfire smoke PM2.5	per 10	−0.14%	−3.8 – 3.5	No Association
Künzli et al., 2006	Asthma attack	Smoke smell days	1–2 d	1.32	0.84 – 2.07	No Association
Künzli et al., 2006	Asthma attack	Smoke smell days	3–5 d	1.63	1.00 – 2.67	No Association
Künzli et al., 2006	Asthma attack	Smoke smell days	210 vs. 30 μg/m3	1.03	0.58 – 1.8	No Association
Künzli et al., 2006	Bronchitis	Smoke smell days	1–5 d	1.33	0.87 – 2.02	No Association
Künzli et al., 2006	Bronchitis	Smoke smell days	≤6 d	2.23	1.45 – 3.43	Positive
Künzli et al., 2006	Bronchitis	Smoke smell days	210 vs. 30 μg/m3	0.79	0.39 – 1.59	No Association
Künzli et al., 2006	Cold	Smoke smell days	1–5 d	1.5	1.25 – 1.81	Positive
Künzli et al., 2006	Cold	Smoke smell days	≤6 d	2.13	1.73 – 2.63	Positive
Künzli et al., 2006	Cold	Smoke smell days	210 vs. 30 μg/m3	0.92	0.67 – 1.25	No Association
Künzli et al., 2006	Dry cough at night	Smoke smell days	1–5 d	2.25	1.87 – 2.71	Positive
Künzli et al., 2006	Dry cough at night	Smoke smell days	≤6 d	3.35	2.71 – 4.15	Positive
Künzli et al., 2006	Dry cough at night	Smoke smell days	210 vs. 30 μg/m3	1.92	1.38 – 2.67	Positive
Künzli et al., 2006	Dry cough first thing morning	Smoke smell days	1–5 d	2.24	1.85 – 2.72	Positive
Künzli et al., 2006	Dry cough first thing morning	Smoke smell days	≤6 d	2.91	2.33 – 3.63	Positive
Künzli et al., 2006	Dry cough first thing morning	Smoke smell days	210 vs. 30 μg/m3	1.93	1.36 – 2.73	Positive
Künzli et al., 2006	Dry cough other times	Smoke smell days	1–5 d	2.67	2.2 – 3.24	Positive
Künzli et al., 2006	Dry cough other times	Smoke smell days	≤6 d	3.27	2.61 – 4.09	Positive
Künzli et al., 2006	Dry cough other times	Smoke smell days	210 vs. 30 μg/m3	2.49	1.86 – 3.33	Positive
Künzli et al., 2006	Wet cough	Smoke smell days	1–5 d	1.42	1.13 – 1.79	Positive
Künzli et al., 2006	Wet cough	Smoke smell days	≤6 d	2.15	1.67 – 2.77	Positive
Künzli et al., 2006	Wet cough	Smoke smell days	210 vs. 30 μg/m3	1.01	0.72 – 1.41	No Association
Künzli et al., 2006	Wheeze/disturbed sleep	Smoke smell days	1–5 d	2.29	1.56 – 3.37	Positive
Künzli et al., 2006	Wheeze/disturbed sleep	Smoke smell days	≤6 d	4.94	3.33 – 7.33	Positive
Künzli et al., 2006	Wheeze/disturbed sleep	Smoke smell days	210 vs. 30 μg/m3	0.89	0.56 – 1.42	No Association
Künzli et al., 2006	Wheeze/limited speech	Smoke smell days	1–5 d	2.23	1.03 – 4.83	Positive
Künzli et al., 2006	Wheeze/limited speech	Smoke smell days	≤6 d	5.49	2.63 – 11.48	Positive
Künzli et al., 2006	Wheeze/limited speech	Smoke smell days	210 vs. 30 μg/m3	0.78	0.29 – 2.1	No Association
Künzli et al., 2006	Wheezing or whistling	Smoke smell days	1–5 d	2.15	1.63 – 2.83	Positive
Künzli et al., 2006	Wheezing or whistling	Smoke smell days	≤6 d	3.53	2.62 – 4.75	Positive
Künzli et al., 2006	Wheezing or whistling	Smoke smell days	210 vs. 30 μg/m3	1.37	0.86 – 2.2	No Association
Lipner et al., 2019	Asthma control (WC vs NWC+VPC)	Wildfire PM2.5	All lags	OR 0.5	−0.2 – 2.97	No Association
Moore et al., 2023	Asthma ED visits	Wildfire smoke days	Wildfire day indicator	RR 1.13	1.02 – 1.24	Positive
Stowell et al., 2019	Asthma	Fire PM (1 <b5>g/m<b3>) 3-day avg	3-day avg	OR 1.08	1.04 – 1.11	Positive
Stowell et al., 2019	Bronchitis	Fire PM (1 <b5>g/m<b3>) 3-day avg	3-day avg	OR 0.97	0.89 – 1.06	No Association
Stowell et al., 2019	Respiratory disease (composite)	Fire PM (1 <b5>g/m<b3>) 3-day avg	3-day avg	OR 1.02	1.00 – 1.03	Positive
Stowell et al., 2019	Upper respiratory infection (URI)	Fire PM (1 <b5>g/m<b3>) 3-day avg	3-day avg	OR 1.01	0.99 – 1.03	No Association

Aguilera et al. (2021) reported a statistically significant increase in cough associated with wildfire PM_2.5 exposure, with an effect estimate of 21.8 (95% CI: 3.6–40.1) per 10 µg/m³ increase. Other outcomes including asthma (0.21; 95% CI: −2.11 to 2.53), difficulty breathing (7.3; 95% CI: −3.8 to 18.4), respiratory distress (0.44; 95% CI: −1.3 to 2.2), and wheezing (0.14; 95% CI: −3.8 to 3.5) did not reach statistical significance. Aguilera et al. (2021), reported data for both aggregated PM_2.5, but our analysis only considers the primary exposure outcome, specific PM_2.5 (wildfire smoke).

Lipner et al. (2019) investigated associations between ambient wildfire smoke exposure and measures of lung function and asthma control. Results indicate that the odds ratio of 0.505 (95% CI: −1.96 to 2.97) compared well-controlled asthma (WC) to not well-controlled and very poorly controlled asthma (NWC + VPC) across all lag periods suggests that individuals exposed to wildfire smoke had lower odds of well-controlled asthma, although the association was not statistically significant due to the wide confidence interval. Lastly, Stowell et al. (2019) identified positive associations for asthma (1.07; 95% CI: 1.03–1.11) and composite respiratory disease (1.02; 95% CI: 1.00–1.03) with fire PM exposure, while bronchitis (0.97; 95% CI: 0.89–1.06) and upper respiratory infection (1.01; 95% CI: 0.99–1.03) were not statistically significant.

These findings suggest that wildfire smoke exposure is most consistently associated with acute respiratory symptoms particularly cough, wheezing, and asthma exacerbations. Chronic conditions such as asthma and bronchitis show more variability, likely reflecting differences in study design, exposure assessment, and population susceptibility.

Risk of bias assessment

A summary of the risk of bias assessment is presented in Figure 1. The risk of bias assessment was done independently, by the primary investigator. Of the five studies systematically reviewed, no articles were excluded from the final analysis due to risk of bias; however, several methodological concerns were identified.

Two studies were assessed as having a high risk of source population representation bias. One study (Lipner et al. 2019) was conducted in a tertiary treatment center, which may limit generalizability to the broader asthmatic population. Another study (Künzli et al. 2006) reported low participation rates in certain communities following wildfire exposure, potentially introducing selection bias.

Blinding-related bias was identified in one study (Künzli et al. 2006) that relied on survey measures without efforts to prevent knowledge of exposure status, raising concerns about differential reporting. This same study (Künzli et al. 2006) also exhibited a high risk of exposure misclassification, as it relied on self-reported exposure measures rather than objective data, which may have affected the accuracy of exposure assessment. An additional study (Lipner et al. 2019) similarly showed potential bias in exposure measurement due to limited precision in estimating wildfire-related PM_2.5 concentrations.

For outcome assessment, one study (Künzli et al. 2006) was judged to have a high risk of bias due to its reliance on self-reported asthma symptoms without the use of validated instruments or clinical verification, introducing the possibility of recall and reporting biases. Three studies were assessed as having a high risk of bias due to unaddressed or unmeasured confounding factors. One study (Moore et al. 2023) noted seasonal influences on asthma exacerbations (e.g. return to school, viral infections, and allergen exposure), the impact of COVID-19, and pollution from other sources. Another did not account for preventative behaviours or spatial variability in exposure, both of which may have influenced reported outcomes (Künzli et al. 2006). A third study lacked information on asthma medication use, a key factor influencing symptom severity (Lipner et al. 2019). Lastly, one study exhibited attrition bias due to a low response rate to the follow-up wildfire symptom survey, which may limit the reliability of the findings (Künzli et al. 2006).

To reduce risk of bias, ensure a causal inference and strengthen an exposure-outcome relationship, four (80%) of the studies reported for confounders that were controlled in the analysis. Across the included studies, a variety of strategies were employed to address potential confounding, reflecting differences in study design, data sources, and analytic approaches. Stowell et al. (2019) used a case-crossover design with matching on geographic location, day of week, and calendar month, and further adjusted for temperature and seasonal trends, while assessing but excluding additional meteorological variables that did not influence results. Lipner et al. (2019) controlled for meteorological conditions and co-pollutants in time-series models examining wildfire-related PM_2.5 and pulmonary function, and conducted sensitivity analyses excluding these covariates to evaluate model robustness. Künzli et al. (2006) incorporated key demographic and health-related factors including sex, ethnicity, parental education, pre-existing asthma, and cohort, in multilevel mixed-effects models and performed stratified analyses by asthma status. Aguilera et al. (2021) used fixed-effects panel regression to account for time-invariant confounders such as baseline risk and population composition, while also adjusting for day of week, month, and linear time trends; additional analyses examined lagged exposures and age-stratified associations. Lastly, Moore et al. (2023) addressed several potential confounders, including seasonal variation, viral respiratory infections, meteorological factors, and co-pollutants, while acknowledging the potential for residual confounding related to asthma management interventions, COVID-19 public health measures, and exposure misclassification. However, this sole article did not report for adjusting for confounders in the statistical analysis. The studies included demonstrate consistency on controlling for temporal patterns, meteorological influences, and underlying population characteristics when examining the respiratory health impacts of wildfire smoke exposure in children.

Discussion

Summary of findings

This review employed a descriptive approach to evaluate and summarize existing evidence on the impact of wildfire smoke on respiratory health outcomes in children. Overall, wildfire smoke exposures, as measured in each study, were consistently associated with increased respiratory health symptoms. Respiratory-related health outcomes of wildfire smoke included increases in risk of emergency department and urgent care visits, hospitalizations, use of respiratory medication, coughing, wheezing, and respiratory distress (for non-asthmatic children). However, there was no significant association specifically between asthma and wildfire smoke exposure. Similarly, there was no significant association found between wildfire smoke exposure and pediatric ED/urgent care visits or hospitalizations specifically for asthmatic children. Children with asthma were more likely to take precautionary measures (proactively use medication, spend time indoors, mask-wearing, etc.) than their counterparts when air quality was poor (Lipner et al. 2019).

Across the included studies, wildfire smoke exposure was consistently associated with increased respiratory symptoms, particularly cough, wheezing, and asthma exacerbations. Künzli et al. (2006) found the strongest evidence of association, with 20 of 30 respiratory outcomes showing statistically significant positive associations, including dry cough (e.g. OR = 3.35; 95% CI: 2.71–4.15) and wheezing (e.g. OR = 3.53; 95% CI: 2.62–4.75). However, these findings rely on self-reported data and should therefore be interpreted with caution. Aguilera et al. (2021) found a significant increase in cough (β = 21.8; 95% CI: 3.6–40.1), while other outcomes were not statistically significant. Moore et al. (2023) and Stowell et al. (2019) identified elevated risks for asthma-related outcomes, including ED visits (OR = 1.13; 95% CI: 1.02–1.24) and composite respiratory disease (OR = 1.02; 95% CI: 1.00–1.03).

Findings suggest a positive association between wildfire smoke exposure and poorer respiratory health outcomes, however, statistical uncertainty varied widely between studies. The results indicate limited precision and variability. Due to wide confidence intervals and inclusion of the null in many of the findings may indicate limited precision and potential variability in effect estimates. Results suggest differences in exposure measurements, sample sizes, and heterogeneity of the study populations. Though many studies found statistical significance, this does not necessarily translate to clinical significance. Conversely, studies with null or mixed findings may still suggest clinically relevant effects that could not be detected due to insufficient power or exposure misclassification.

Four of the five studies solely looked at pediatric cohorts; the one outlier was a population-level study that included separately reported pediatric data. The outcomes of this study included both pediatric and adult outcomes (Stowell et al. 2019). One study separated age-specific subgroups within the pediatric population, while the remaining studies reported results < 21 years of age together. The single study that specifically separated age groups found no significant association among younger children between wildfire smoke and lung function at any lag. However, older patients experienced below‐normal lung function on the day after wildfire smoke exposure. Additionally, a single study recruited participants using an existing pediatric cohort. The age groups and outcomes were limited for this study, making it difficult for conclusions as well as increasing the risk for selection bias (Künzli et al. 2006).

The populations studied provide a wide variety of airway sizes, developmental stages, and day-to-day activities, specifically between pediatrics in comparison with older children and adult-sized teens. It is expected that results will vary between age groups; however, only one study in this review analyzed results at different age groups, limiting the conclusions that can be made. This review highlights the need for research on the response to wildfire smoke exposure in varying youth age groups.

Most studies included in the review used PM_2.5 and modeling as measures of exposure. PM₁₀, another commonly measured component of air pollution, was employed by a single study, which also used surveying as a secondary exposure measure. Despite using similar measures of exposure in the studies, the length of time of exposure and lag time between exposure and outcome measurement varied considerably. While using two exposure measures might have improved accuracy in estimating exposure impacts on youth, it is difficult to compare between the studies. This review highlights the need for a universal modeling system, focusing on PM_2.5 for evaluating air pollution impacts on population health.

Despite the cumulative sample size across studies exceeding one million, the included studies varied widely in their individual sample sizes and analytic approaches. This review did not assess the adequacy of statistical power within each study. Consequently, some non-significant or inconsistent findings especially in subgroup or lagged analyses should be interpreted with caution, as limited power may have reduced the likelihood of detecting smaller but meaningful effects (Friedman 2001).

Strengths and limitations

The authors followed the PRISMA guidelines for conducting a systematic review. A detailed protocol outlining the methodology was included in the systematic review. Comprehensive searches conducted by an independent reviewer identified literature highlighting pediatric respiratory health outcomes associated with wildfire smoke. One independent reviewer conducted the screening and risk of bias assessments. These strengths enhanced the study by ensuring a thorough approach to identifying eligible research and applying rigorous criteria for their selection.

Our study presents several limitations. This review is limited by utilizing only two scientific databases and the limited number of studies (n = 5) included. The five studies focused on North America, potentially limiting the applicability of findings to other regions with different wildfire characteristics, healthcare systems, and population demographics. Additionally, including a second reviewer for risk of bias could have benefited this review.

The analysis presents significant heterogeneity observed among included studies. Variation is likely due to potential influencing factors including differences in study populations, diagnostic methods, geography, and time frames. Additionally, a key limitation across studies is the presence of wide confidence intervals and varying levels of statistical uncertainty, which may obscure the distinction between statistically and clinically meaningful effects. This review only includes vote counting based on the direction of effect and statistical significance. A limitation of this statistical methodology is the lack of account for differences in the relative sizes of the studies included (Borenstein et al. 2009). Pooling data and a meta-analysis were not feasible in this review due to time constraints. This review focuses specifically on pediatric populations, but few studies broke down data by age group. There was little comparison between different stages of childhood (infants, preschool children, and teenagers) despite these groups being physically and developmentally unique. However, few studies report solely pediatric outcomes; therefore, addressing the limited research done on this topic. Lastly, the majority of studies focused on acute outcomes, potentially limiting insights into long-term respiratory impacts or chronic conditions resulting from wildfire smoke exposure in the population.

Conclusions

There are many challenges in understanding the impact of wildfire smoke on human health, however, a main priority lies in improving health outcomes during and after wildfire event. There is limited literature on observational studies pertaining to children’s risk of respiratory-related health outcomes associated with wildfire smoke exposure. More studies are needed to better understand the health impact of wildfire smoke on youth health. With an increase in wildfire frequency and severity, it is a pressing topic that requires further investigation. Future studies should prioritize observational research to assess the long-term effects of wildfire smoke exposure on children while also analyzing how these impacts vary by age. Moving forward future reviews should consider a full meta-analysis on the impacts of wildfire smoke on youth respiratory health outcomes, considering the impact of bias on weight of each study in addition to the studies sample size. Findings from future studies can also inform interventions that reduce wildfire smoke exposures to this sensitive population.

The studies included in the review were based on local regions, with few studies at national or other large geographic scales. Understanding larger geographies and additional areas both within and outside of the United States and Canada is necessary to note communities at the greatest risk for wildfire smoke-related respiratory health outcomes. Additionally, large-scale studies can positively influence policymakers’ engagement in the topic.

Nearly all studies, with the exception of one, used PM_2.5 as the primary measure of exposure. As an inclusion requirement, all studies were required to use a second measure, most utilizing modeling to account for more detail in localizing which sites were impacted by wildfire smoke. While these two forms of exposure measurements may improve accuracy, they were not universal across all studies and, therefore, not comparable. Given the importance of air quality monitoring to understand the importance of air pollution on health, there is a need for a universal system of exposure reporting. Inconsistencies in exposure reporting allowed for high variability between wildfire exposures, often making it difficult to compare studies. Future epidemiological studies should consider more consistent and thorough reporting of air quality measures. In more developed countries PM_2.5 data is becoming more readily available, as a result reporting on baseline air quality periods and wildfire smoke periods could allow for more meaningful between-studies comparisons (Barkjohn et al. 2022). In addition to outdoor exposures, indoor air quality monitoring during smoke events should also be considered, given majority of individuals spend their time indoors.

The emergence of PurpleAir particulate matter sensors has provided the public, media, and air quality agencies with real-time air quality information, particularly during wildfire smoke events. These sensors have helped fill gaps in traditional air monitoring networks. While the data provided by PurpleAir and similar low-cost air quality sensors (LCAQS) is valuable during smoke impacts, their accuracy and performance often do not match that of regulatory-grade equipment (Zhang and Srinivasan 2020). Despite ongoing national efforts to improve these technologies, there is a growing need for standardized LCAQS systems. Technical uncertainties limited cross-validation, and a lack of consistent verification protocols present challenges when comparing LCAQS to each other or to more sophisticated instruments. These issues underscore the importance of establishing universal equipment standards and monitoring strategies. As interest in both indoor and outdoor air quality monitoring grows, the proliferation of diverse approaches further emphasizes the need for consistent and validated methods. While LCAQS can provide meaningful insights into air pollution, without a standardized monitoring and reporting framework, it remains difficult to assess and ensure their performance. A universal system for data reporting would promote integration and collaboration between researchers and fields to understand the true impacts of exposure from wildfire smoke and other air pollutants.

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09603123.2025.2579084