Part III – Exposure to wildfire smoke and child health outcomes

The epidemiological evidence from studies investigating wildfire smoke exposure on child health is heterogeneous and limited. Most research has been conducted in high-income countries in Australia, Europe, and North America and assesses, most commonly, acute outcomes close to the time of exposure (hospital presentations or admissions, or respiratory symptoms), and uses a mix of methods to assess outcome and exposure assessment that make estimation of summary effects difficult. There have been some studies conducted to examine longer-term outcomes however these mostly focus on wildfire as a traumatic event and the impact on mental health outcomes and are not necessarily wildfire smoke exposure investigations. As a body of evidence there are many gaps in our understanding of the impact of wildfire smoke exposure and child health. However, most literature acknowledges that childhood is a recognized window of susceptibility, and that children have particular vulnerabilities to wildfire smoke due to their immature immune system, developmental stage and their behaviour (such as outside play), and that wildfire events are increasing and therefore more children will be impacted (Liu et al., 2016, Holm et al., 2021).

Vulnerability during childhood

Early childhood is a vulnerable period of growth and development, and children are particularly sensitive to environmental exposures (Sheffield and Landrigan, 2011). Young children’s smaller airways, smaller lung size, higher respiratory rates, increased metabolic rates, and their under-developed immune system compared with adults increases their exposure and vulnerability to respirable pollution (Leibel et al., 2020). Young children are also less likely to adopt risk mitigation behaviors, have less access to appropriately sized masks, and may spend a large proportion of the time during an exposure period where the air is insufficiently ventilated (in school), or outdoors (Sheffield and Landrigan, 2011, Holm et al., 2021) and for these reasons are at greater risk of exposure, and more vulnerable to the risks.
Particulate matter, one of the most frequently measured solid particles found in air, are common in both ambient air and wildfire smoke. Levels of particulate matter and adverse health outcomes are linearly related and the risk for adverse health increases as concentrations of particulate matter increase regardless of the source (Gan et al., 2017). Although the evidence from wildfire smoke is increasing (Reid et al., 2016a), much of the earlier epidemiological evidence, comes from studies of ambient air pollution (Stieb et al., 2012). Although these studies are useful for hypothesis development of the effect of wildfire smoke exposure there are some important differences between these exposures. For example, the chemical composition of wildfire smoke is similar but not the same as other causes of air pollution including traffic, agricultural burn-off, and household air pollution (Künzli et al., 2006) and within these, the composition may vary based on type of biomass burned, whether structures were also burned and the duration of the burning event. Wildfire smoke is more likely to have a shorter duration, whereas exposure to traffic pollution is more constant. Examining short-duration, unpredictable exposure on acute health outcomes is also challenging, as is examining the longer-term impacts into adolescence and adulthood. The impact on adult health after short-term exposure in childhood is also unknown (Shao et al., 2019). Animal models of early life exposure and longer-term morbidity are suggestive of immune dysregulation and changes to lung function that extend at least into adolescence (Black et al., 2017a). Currently, most of the epidemiological research investigating longer-term effects has investigated mental health, or psychological outcomes due to exposure to trauma (of wildfire threat), rather than other morbidity, although all health outcomes are important areas for research.

Epidemiological evidence of childhood health effects of wildfire smoke

Health Service Use

Health service use is a frequently assessed outcome for epidemiological studies of wildfire exposure on children. Health service use is defined mostly through hospital, or emergency department presentations or admissions, and or doctor or clinic visits in children during or related to a wildfire event or smoke exposure. These episodes of care are mostly related to respiratory causes, and health service use in general and specifically related to respiratory symptoms has been reported to increase during periods of wildfire smoke exposure and comes from a range of studies (Henry et al., 2021). A prospective cohort study of >6000 school-aged children, mostly aged 6 to 7 years, collected annual respiratory health data and included wildfire exposure following the 2003 Californian wildfires (Künzli et al., 2006). All respiratory symptoms and doctor visits were increased in children affected by smoke by up to 33% (Künzli et al., 2006). There were differences by asthma diagnosis and children’s age (e.g. high school children reported more eye symptoms than primary children, 44.0% versus 35.5%). Children with asthma reported more respiratory symptoms, but this was modified by adoption of preventative behaviours (mask wearing, staying indoors, and using air conditioners), and the number of reported days of fire smell indoors. In children who were the most exposed (>6 days of smoke, and without masks), sneezing more than doubled [OR=2.8 (95% CI: 2.3, 3.5)]. The results from this study are generalizable to a range of similar settings as it was population-based study. A recent ecological study also from California, reported increases in pediatric (<19 years old) emergency hospital presentations by 30% when the PM2.5attributable to wildfires increased by 10 ug/m3(Aguilera et al., 2021). During the exposure period, younger children, those between 0 to 5 years were the most likely to present to hospital (34.5% increase) (Aguilera et al., 2021). Respiratory presentations of cough increased 21.8% with a 10 ug/m3 increase in PM2.5 attributable to wildfire smoke (Aguilera et al., 2021). A study using medical insurance claims for episodes of care related to respiratory or cardiac outcomes for people aged 0 to 64 years during a six-month window that included the wildfire season in San Diego reported that children’s attendance at outpatient clinics for respiratory conditions was greater during peak fire-exposure periods. Overall children aged 0 to 4 years accounted for 14.4% of all inpatient hospitalizations, 15.1% of emergency presentations, and 28.8% of outpatient visits. Children younger than 2 years had higher risk of presenting, although the numbers of children in this age group were very small and varied by the specific respiratory condition for presentation (Hutchinson et al., 2018). A larger interrupted time series study also from San Diego, used hospital data for all children (<19 years old) to quantify excess daily hospital visits for respiratory symptoms during a 10-day fire exposure period. Daily emergency department visits increased by 16.03 daily visits per 10,000 children, during the exposure window. Younger children aged 0 to 5 years had the highest absolute daily visits of 7.3 daily visits. Children aged 6 to 12 years had the most change in visits, which changed by 3.4 (95% CI: 2.3, 4.6) daily visits (Leibel et al., 2020). In a study to assess different simulation models using hospital respiratory admissions in Washington the odds of respiratory hospitalizations in children (<15 years) increased by ~10% on days when the PM2.5increased by 10 ug/m3, which was found consistently in all three simulation models (Gan et al., 2017). In contrast, an older population-based study used hospital admission and mortality data over eight years, from Sydney, Australia that included 32 bushfire days and found bushfire derived PM10 was associated with moderate decreases in asthma hospital presentations for children three days after the peak event (Morgan et al., 2010).
In epidemiological studies of the association between wildfire smoke and health, PM2.5 is the most common measure of wildfire smoke exposure; however, increases in ozone also occur during wildfires (Pratt et al., 2019). Ozone is a secondary pollutant and fluctuates with temperature, time of day and weather conditions, and wildfires are an additional source of ozone (Pratt et al., 2019). A 2019 study used analytical models to estimate the burden of pediatric emergency department visits attributable to elevated ozone associated with wildfire smoke in the US, between 2005-2014. They found associations between ozone that was attributable to wildfire smoke and increases in emergency department visits of children <18 years presenting for asthma (Pratt et al., 2019). The overall median number of visits for the ten years of the study period was 29 emergency department visits per child [(interquartile range (IQR): 3-63)]. The number was higher for males, 18 visits (IQR: 2-40) than females, 10 (IQR: 1-20) visits (Pratt et al., 2019).

Respiratory Symptoms and Asthma Medication Use

Morbidity based on respiratory symptoms, exacerbation of existing disease, new disease diagnoses, and medicine use associated with wildfire exposure in children has mostly been examined through self-report of these outcomes. A 2016 Spanish birth cohort study of 496 children, whose parents completed a telephone interview two weeks after a wildfire event, reported an increase of respiratory symptoms for all children. Children with existing asthma were nine times more likely to report itchy or watery eyes [OR=9.26 (95% CI: 2.14, 40.12)], or 11 times more likely to report sneezing compared to children without asthma associated with smoke from wildfires [OR=11.40 (95% CI: 2.01-64.52)] (Vicedo-Cabrera et al., 2016). A 2009 longitudinal study in California of 465 non-asthmatic children aged 16 to 19 years who self-reported respiratory and eye-symptoms following a wildfire event, demonstrated a link between duration of smoke exposure and respiratory symptoms (Mirabelli et al., 2009). For children who reported smelling smoke for six or more days (the exposure variable) the prevalence of respiratory symptoms was 60%, for eye symptoms the prevalence was 80% (Mirabelli et al., 2009). The Californian study by Künzli et al.described above reported that respiratory morbidity including dry cough, wheeze and eye-symptoms were more common in children exposed to wildfire smoke than those not exposed and was higher in children with existing asthma (Künzli et al., 2006). A panel study of 235 people in Darwin, Australia of whom ~50% were <18 years old, asthma medication use both reliever medication and initiation of oral steroids were associated with low levels of PM10 and PM2.5 derived from wildfire smoke (Johnston FH et al., 2006). A 2015 retrospective study of two cohorts of >6000 Californian children where exposure to air pollution was determined by ZIP codes and distance to smoke plumes one year before and one year after wildfires analyzed several asthma outcome measures. These included short-acting β-agonists dispensing behavior. Dispensing behavior following wildfire exposure increased in all asthmatic children but the result was inconsistent between the two cohorts (Tse et al., 2015). In an older and very small study (n=32 children), asthmatics exposed to excessive PM10 from bushfire smoke for seven days did not find any association with peak expiratory flow rates; however, among a subgroup without bronchial hyperreactivity, there was an inverse association between PM10 levels and evening peak expiratory flow rates (Jalaludin et al., 2004). In more recent work, 1404 children aged 4 to 21 years completed a Forced Expiratory Volume in 1 second test (FEV1) in a three-day window after wildfire smoke exposure. Children aged 12 to 21 years had poorer lung function on day one which returned to normal function by day three. Children 4 to 11 years did not demonstrate any change in lung function (Lipner et al., 2019).
There is limited data of the effect of wildfire smoke exposure on child mortality. A recent study estimated the under-5 child mortality attributable to wildfire and managed landscape burn smoke exposure, and reported an estimate of 104,279 child deaths a year, which accounts for 39% of the global annual mortality burden attributable to wildfire and landscape burns smoke (677, 745 annual premature deaths) (Roberts and Wooster, 2021). This finding comes from a 2021 modelling study that used statistical methods to isolate the contribution of managed landscape burning to overall PM2.5 and then calculated the number of people exposed to estimate mortality by country. Most of these deaths occur in low- and middle-income countries, and the data suggest that landscape biomass burning is responsible for between 8-21% to the population weighted PM2.5 concentration in many countries, making it the biggest contributor to all ambient air pollution. More accurate data and further testing of the modelling will increase the accuracy and precision of these estimates.

Other vulnerabilities and indirect outcomes

Some studies have investigated potentially vulnerable population sub-groups, including in obese children (Tse, 2015). Most studies investigating body mass index (BMI) as a risk factor in children have examined other air pollutant exposures; such as tobacco smoke or PM2.5, and not specifically wildfire smoke exposure (Huang et al., 2019). Obesity is as a risk factor for asthmatic disease, both incident disease and severity (Sin and Sutherland, 2008), and the prevalence of childhood obesity is also increasing (WHO, 2018) which suggests that there is an urgent need to conduct these investigations to better understand risk factors.
In two retrospective cohorts of children aged 5 to 17 with asthma in Southern California included 1050 obese children, as described above (Tse et al., 2015). Overall, dispensing of short-acting ß-agonists was increased in the post-fire periods, and the greatest increase in dispensing was in children who were obese (16% increase, P<0.05); however, when this was restricted to children only in the fire ZIP codes, there was no difference in dispensing by BMI. Overall, there was no increase in hospital presentations or admissions, oral corticosteroid dispensing frequency, or new asthma diagnoses post-wildfire (Tse et al., 2015).
Impacts from wildfire or prescribed burning on the immune system and cardiovascular health were investigated in a small convenience sample of 220 children (mean age 7 years) in California, reported as a letter to the editor (Prunicki et al., 2019). Of the 32 children who had been exposed to wildfire pollution or prescribed burning (based on residence) three months prior compared to 18 children with no exposure demonstrated a difference in Th1 cells, but not in Th2 or T regulatory cells (Prunicki et al., 2019). This was a small study suggest potential immune changes, though the lack of details make the study difficult to evaluate and results need to be further investigated.
The effect on mental health from exposure to wildfire evidence has been assessed mostly through the lens of experiencing a traumatic event. Trauma related to wildfire exposure has been defined using the full range of experience, from living in the fire or smoke affected community, being evacuated, perception of threat, through to the death of a parent. Estimates of post-traumatic stress disorder, the most studied mental health outcome in this field, is estimated to be between 12-30%, although this varies considerably with the measures used, the age of the children, and definition of the trauma experience. In one cross-sectional study conducted in Greece, 1339 adolescents aged 12 to 17 years completed a self-report questionnaire of wildfire exposure and symptoms of threat, or perceived threat including anxiety and depression, six months after a wildfire event. Of those 432 adolescents (29.4%) reported levels above the cut-off of post-traumatic stress disorder symptoms, and perception of life threat was a greater predictor of these symptoms than actual threat (Papadatou et al., 2012).
Academic achievement was assessed four years after exposure to a bushfire disaster in a recent Victorian, Australian population-based study of 24,642 children who were highly or moderately exposed to a major bushfire. Exposure was assessed through school proximity to the bushfire. Academic achievement was measured using the standardized annual literacy and numeracy test scores at two time points two years apart, two and four years after the bushfire for each child. There were reductions in the children’s expected numeracy and literacy test scores based on the expected trajectory between the two test scores at approximately ages 9 and 13 years. The hypothesized impact pathway was through cognitive deficits related to impacts on working memory as a result of trauma including loss of parent(s) which may have resulted in disruption to schooling and lack of parental support for reading at home (Gibbs et al., 2019). Another study, whose primary study factor was health effects in children from wildfires used an existing cohort of >6000 children, 86% aged 6 to 7 years old in California (Künzli et al., 2006). Exposure was measured both subjectively and objectively in two analytical models. Although academic achievement was not assessed, the authors found that missed school attendance was associated with increased respiratory symptoms especially in children with an asthma diagnosis (21.7% compared to 11.55%) (Künzli et al., 2006). School closures is a common outcome of wildfire, and in 2018 in California more than a million children were impacted by school closures due to wildfires (Chalupka and Anderko, 2019).
During the bushfire event in New South Wales, Australia of 2019, a natural experiment was conducted in the context of an ongoing cluster randomized controlled trial of a school-based physical activity intervention in children aged 8 to10 years old (del Pozo Cruz et al., 2021). During the bushfire event the schools received targeted health advice from the government health department to avoid all outdoor physical activity. Exposure was measured using air quality data measurements located near the schools participating in the trial during the bushfire, and physical activity was measured using accelerometers that the children were wearing for the trial. The authors reported a minimal impact on physical activity up to a ‘tipping point’ in air quality when it was rated as ‘very poor’ and the physical activity dropped sharply. This suggests that the public health advice did not change children’s activity and was largely ignored until the air quality reached very high pollution levels. Further, if air quality is poor for an extended period due to wildfires, it may result in reduced physical activity and children may have detrimental impacts on health outcomes.
The impact on infant feeding has also been assessed through the pathway of exposure to emergency or disaster. A study from Canada of 115 women with infants or young children up to 36 months were recruited through social media after being evacuated during the 2016 Fort McMurray wildfire. They reported a 30% reduction in breastfeeding before versus after the evacuation (\(x^{2}\)=29.62, P=<0.001) (DeYoung et al., 2018). Responses from participants suggested that many factors resulted in reduced breastfeeding, including stress, perception of reduced milk supply and lack of privacy.

Summary of childhood studies

The impact of wildfire exposure on child health is an emerging area for research investigation. Wildfire smoke exposure is a unique composition of toxic chemicals that presents uniquely with shorter durations subject to climate and geography resulting in unpredictable exposure patterns (Black et al., 2017b). The effect on children is a research gap that needs investigation, and the current body of evidence is a mix of methodology and quality and has many of the same methodological issues to the pregnancy literature, that make drawing meaningful and generalizable conclusions difficult. Population-level data with accurate exposure assessment, inclusion of important health outcomes, especially those with children as the focus are needed. Evidence examined here suggest that respiratory disease, either symptoms or exacerbation of disease is a confirmed consequence from wildfire exposure in children, as is increased health-service use including hospitalizations. Evidence from ambient sources of PM2.5 not reviewed in this chapter has more extensive investigation and more definitive results, and the adult research as discussed in earlier chapters is also more advanced and extensive, and suggests that there is much more to learn in relation to the impact on children.