Prenatal exposure to wildfire-related PM2.5 and respiratory infections by age 1 year: A population-based case-control analysis of critical developmental windows

Qingyi Lan, Kate Weinberger, Sabrina Luke, Eric Lavigne, Scott Weichenthal, Sarah B. Henderson
doi: https://doi.org/10.1101/2023.09.09.23295304

Abstract

Background The 2017 and 2018 wildfire seasons in British Columbia (BC), Canada were unprecedented. Among all the pollutants in wildfire smoke, fine particulate matter (PM2.5) poses the most significant risk to human health. There is limited research on prenatal wildfire-related PM2.5 exposure and its impacts on infant health.

Objectives We assessed the effect of prenatal wildfire-related PM2.5 exposure on respiratory infections in the first year of life. We explored exposure windows based on prenatal respiratory tract development and gestational trimester.

Methods All live singleton infants in utero during the wildfire seasons (July to September) from 2016 to 2019 were identified using the BC Perinatal Data Registry (BCPDR). The Canadian Optimized Statistical Smoke Exposure Model (CanOSSEM) was used to estimate daily PM2.5 exposures throughout pregnancy and the first year of life. Infants who were diagnosed with otitis media or a lower respiratory infection by age 1 year were identified, including infections associated with dispensations of the antibiotic amoxicillin. The odds ratios (OR) for exposures during specific developmental windows and gestational trimesters were estimated with logistic regression, conditioned on epidemiologic week of birth to control for seasonal trends. Analyses were adjusted for multiple environmental and individual-level covariates.

Results There were 134,475 infants included in the cohort, of which 22,093 were diagnosed with at least one outcome of interest. For every 10 μg/m3 increase in wildfire-related PM2.5, the OR [95% confidence interval] for otitis media associated with the fourth window of eustachian tube development (weeks 19-28) was 1.08 [1.02, 1.16]. Results were similar for infections associated with amoxicillin dispensations and during the second trimester (weeks 14-27) of development. Similarly, the canalicular stage of lower respiratory tract development (weeks 18-27) and the second trimester were associated with increased odds of lower respiratory tract infections.

Conclusion Specific windows of prenatal exposure to wildfire-related PM2.5 during the second trimester were associated with increased odds of otitis media and lower respiratory infections by age 1 year.

Study questions Is prenatal exposure to wildfire-related PM2.5 associated with risk of respiratory infections in the first year of life? If so, what are the critical prenatal exposure windows?

What’s already known Few studies have focused on prenatal exposure to wildfire-related PM2.5 and respiratory outcomes in the very early stages of life, nor the critical windows for gestational exposures. Evidence suggests that wildfire-related PM2.5 may impact the foetus directly or indirectly, potentially leading to adverse consequences for foetal respiratory tract development and vulnerability of the respiratory system in early life.

What this study adds Prenatal exposure to wildfire-related PM2.5 during specific periods of respiratory tract development were associated with increased risk of otitis media and lower respiratory infections by age 1 year. Results were similar for all infections and those associated with antibiotic dispensations. This study suggests that wildfire-related PM2.5 exposure during specific developmental windows can affect respiratory health in early life. Public health practitioners and healthcare providers should work to protect pregnant people and their children from the detrimental effects of wildfire smoke.

Background

Exposure to air pollution during pregnancy has been linked to negative effects in early life, such as cardiovascular and metabolic outcomes, respiratory and allergic responses, and neurodevelopmental impairments 1. Air pollution can directly impact pregnancy through dispersion across the tissue barriers and permeation across the cellular membrane of the placenta2. Indirectly, oxidative stress and inflammatory reactions in mothers can reduce foetal nutrient and oxygen supply 2,3. Interactions between air pollution and the epigenome can lead to adverse birth outcomes, including decreased respiratory function, though the mechanisms are not completely understood3.

Wildfire smoke is a complex form of air pollution, for which PM2.5 is generally accepted to be the most health-relevant marker. Compared with other PM2.5 sources such as traffic and industry, wildfire-related PM2.5 is far more episodic and the magnitudes of the exposure can be very high. Studies have shown that PM2.5 from wildfires may be more harmful to human respiratory health than ambient PM2.5 from other sources4,5. Although some studies have examined the relationship between wildfire smoke exposure and respiratory outcomes in adults and school-age children6–8, very few have looked at the effects of prenatal exposure on respiratory outcomes in early life.

A recent study in Australia found that a small cohort of infants prenatally exposed to smoke from a coal mine had more respiratory symptoms and diagnoses of respiratory tract infections in early life than an unexposed group9. Research from a birth cohort in Poland suggested that exposure to wildfire-related PM2.5 during pregnancy may increase the frequency of wheezing attacks in the first two years of life10. However, neither of these studies have assessed the relationship between prenatal exposure to wildfire-related PM2.5 during specific developmental windows.

Acute respiratory infections are the leading cause of morbidity and mortality in early childhood, even in high-income countries11. According to one study conducted in Canada, respiratory infections can impose substantial strain on acute care infrastructure and the medical system, and the burden is expected to increase with population growth12. Respiratory infections are also the primary reason for antibiotic use in early life, which can have long-lasting impacts on the infant microbiome and childhood respiratory health13. The most prevalent infections involving antibiotic treatment are acute otitis media, bronchitis, and upper respiratory tract infections14.

Exposure to wildfire smoke during pregnancy, or certain windows of pregnancy, may increase the risk of infant respiratory infections in early life. It follows that limiting exposure to wildfire smoke during pregnancy could protect infant health. To address this important public health question, our study focusses on the critical stages of gestational development for parts of the upper and lower respiratory tract associated with the two most diagnosed respiratory infections: otitis media (i.e., middle ear infections) and lower respiratory tract infections.

Otitis media is an upper respiratory infection that frequently requires extensive paediatric care in the early stages of life. It is characterized by acute or chronic inflammation of the middle ear, resulting from Eustachian tube dysfunction15,16. Lower respiratory tract infections include bronchitis, bronchiolitis, and pneumonia, all of which can lead to serious complications. Most lower respiratory infections in children are caused by bacteria, but they can also be caused by viruses and fungi17. The complex and protracted process of lower respiratory tract development, commonly referred to as foetal lung development, begins during the embryonic stage. Damage and changes to the upper and lower respiratory tracts during foetal development may lead to higher risk of respiratory diseases, including acute and chronic infections 18–20.

Our study aimed to investigate the prenatal impact of wildfire-related PM2.5 on respiratory infections during the first year of life using a population-based case-control design. We focused on specific stages of prenatal respiratory tract development and gestational trimesters to examine their association with respiratory infections in early life. Evidence from this study will be used to articulate preventive public health strategies to protect pregnant people and their infants.

Methods

Study setting and population

This study is set in the province of British Columbia (BC), on the west coast of Canada. Like other parts of western North America, BC is heavily forested and prone to seasonal wildfires21, especially from July through September22. The total area of BC is almost 1 million square kilometres. The 2016 population was approximately 5 million residents, with most living in large urban centres in the southern coastal areas and the rest living in smaller urban, rural, or remote areas throughout the interior and the north (Figure 1A).

Figure 1.
Figure 1.

Population of British Columbia by 2016 Census Divisions (panel A) and daily population-weighted average fine particulate matter (PM2.5) concentrations during the study period (panel B). The dashed lines show the wildfire seasons from July to September.

In this study we focussed on the wildfire seasons from 2016-2019 because of the contrasts between these four years. The 2017 and 2018 wildfire seasons were both record-setting, with more than 1.2 million hectares burned in 2017 and more than 1.3 million burned in 2018 22,23. In both summers, the smoke impacts were widespread and prolonged (Figure 1B). By contrast, the 2016 and 2019 wildfire seasons were well below the 10-average, with less than 0.13 million hectares burned in both years, and the smoke impacts were limited23. The study included all singleton births that were prenatally exposed to the peak wildfire seasons (July – September) of 2016 to 2019 during any stage of pregnancy. Given an average gestation period of 38 weeks, this includes most infants born in BC between June 2016 and February 2020.

Health data sources

All health data were accessed through the Population Data BC research platform. The population-based cohort was identified using the BC Perinatal Data Registry (BCPDR)24. The BCPDR compiles extensive data on mothers, foetuses, and newborns, covering almost all births in the province. Most infants enter the registry at approximately 15 weeks of gestation. The BCPDR provides comprehensive information on delivery, the postpartum period, and includes many maternal covariates, such as BMI and smoking status, which are described in more detail in the covariates section to follow. The BCPDR also includes infant birthdates and estimated dates of conception, which facilitates precise exposure assessment for critical developmental windows25. Finally, the BCPDR includes the residential forward sortation area (FSA) of the mother, which is the first three digits of the 6-digit postal code.

BC has a single-payer healthcare insurance program called the Medical Services Plan (MSP), which covers most people in the province26,27. The MSP client roster was used to create a weekly residential history for mothers during their pregnancy and for infants during their first year of life28. The residential histories were created using 6-digit postal codes. In urban areas, each 6-digit postal code typically covers one building or one city block, but in rural settings they can cover much larger areas. In all cases, we used the geographic coordinates of the single link indicator, which indicates the location of most of the population within the postal code.

We used the MSP outpatient physician billings file to identify all infants who were diagnosed with otitis media or a lower respiratory infection by age 1 year. All prescription medications dispensed by community pharmacies in BC are recorded in the PharmaNet database29. By linking the MSP billings with PharmaNet records for each infant, we were able to identify infections that were associated with a subsequent antibiotic dispensation, as described in further detail in the outcomes section to follow.

Exposure

The Canadian Optimized Statistical Smoke Exposure Model (CanOSSEM) is an empirical framework that estimates daily population exposure to PM2.5 from biomass burning across Canada from 2010 to 2022 at a resolution of 5 km × 5 km for all inhabited areas 30. The model was built using PM2.5 values from over 300 air quality monitoring stations in the National Air Pollution Surveillance (NAPS) program, which encompasses both rural and urban areas. The predictive variables include daily smoke plumes from remote sensing images, measurements of Aerosol Optical Depth, measurements of fire radiative power, and meteorological parameters such as temperature, humidity, and wind speed from satellite products 30. CanOSSEM demonstrates excellent performance, with approximately 98% of estimated concentrations within 5 µg/m3 of the observed PM2.5 value for the validation dataset30.

Daily estimated PM2.5 concentrations for each 6-digit postal code were derived from CanOSSEM. The estimates were then matched with the residential histories created from the MSP client roster and BCPDR records, resulting in daily maternal (from conception to the birthdate) and infant (from the birthdate to age 1 year) PM2.5 exposure history. The maternal and infant residential histories were created by identifying a continuous set of 6-digit postal codes covering every day of pregnancy or the first year of life. If there were duplicated records for any date, we always chose the 6-digit postal code that matched the 3-digit FSA in the BCPDR for the main analysis. As a sensitivity analysis, we always chose the 6-digit postal code that did not match the 3-digit postal code in the FSA, which resulted in different exposure estimates.

Outcomes

Diagnoses of otitis media and lower respiratory infections were identified from the MSP outpatient physician billings filed using the International Classification of Diseases, 9th Revision (ICD-9) codes. We used MSP billings for specific codes to identify otitis media (ICD-9 codes 381-382) and lower respiratory infections (ICD-9 codes 466, 480-487) in the first year after birth. Subsequent dispensations of amoxicillin were identified from PharmaNet. We only included dispensations of amoxicillin, which was the most prescribed antibiotic in the first year of life and it is also the preferred medication for treating infant otitis media and lower respiratory infections in Canada based on clinical guidelines 31,32. We included any amoxicillin dispensation within seven days of a physician visit for a respiratory infection.

Covariates

Analyses were adjusted for several potentially confounding variables from the BCPDR and MSP, reflecting maternal and infant health. Maternal covariates included: age at delivery; BMI category; maternal smoking status; and neighbourhood socioeconomic status (SES) quintile, all of which have been identified as potential confounders in prior research 33–35. Infant variables included indicators for: biological sex; preterm birth; and low birthweight34,36,37. Preterm birth refers to a newborn with a gestational age of less than 37 weeks37. Low birthweight is defined as an absolute weight less than 2500 g, regardless of gestational age 36. Missing values of categorical variables were addressed by including an unknown category.

Analyses were also adjusted for temperature and postnatal exposure to wildfire-related PM2.5 during the first year of life, which are also associated with an increased risk of respiratory illness 38–40. Exposure over the whole year was used for all infants, regardless of outcome status. We used the same temperature data included in CanOSSEM, which were derived from the NASA Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2) 30.

Exclusion criteria

We excluded infants missing 25% or more days of prenatal wildfire-related PM2.5 or temperature exposure data due to incomplete residential histories or missing CanOSSEM estimates. The proportion of missing data was calculated for the whole pregnancy, and for each critical exposure window. Similarly, infants missing more than 25% of postnatal exposure in the first year of life were also excluded. We also excluded infants who could not be matched with the MSP and PharmaNet datasets, and those missing values for maternal age.

Critical developmental windows

For otitis media, the critical windows of Eustachian tube development were defined as: 0-9 weeks (stage 1); 10-12 weeks (stage 2); 13-18 weeks (stage 3); 19-28 weeks (stage 4); and >28 weeks (stage 5). For the lower respiratory tract the developmental windows were defined as: 0-7 weeks (embryonic stage); 8-17 weeks (pseudoglandular stage); 18-27 weeks (canalicular stage); 28-36 weeks (saccular stage); and >37 (alveolar stage)41,42,43. We also examined the effects of each gestational trimester 0-13 weeks (Trimester 1); 14-27 weeks (Trimester 2) ; >27 weeks (Trimester 3)44,45. These critical time windows were identified from the existing literature based on significant physiological changes in both pregnant individuals and their foetuses (eTable 1).

View this table:
Table 1. Characteristics of infants included in the population-based cohort and their mothers matched on each unique birth

Statistical analysis

All data preparation and statistical analysis were done using R 4.0.5 46. We used conditional logistic regression to quantify the association between prenatal wildfire-related PM2.5 exposure and respiratory infections by age 1 year. Models were conditioned on the epidemiologic week of birth to address confounding by seasonality, which is associated with both wildfire smoke exposures (higher in summer months) and respiratory infections (higher in winter months). Epidemiologic weeks allow comparison of population health trends between years, with the first epidemiologic week of the year beginning on the first Sunday in January. Some studies suggest matching by conception date rather than birthdate to address seasonality 47, and we did this as a sensitivity analysis.

We used the daily average estimates of wildfire-related PM2.5 concentration during each prenatal exposure window as the main predictor variable. The dependent outcome variables were all otitis media, all lower respiratory tract infections, and the otitis media or lower respiratory tract infections associated with amoxicillin dispensations. All estimates were adjusted for exposure during all stages of development, and the covariates described in the previous section (Eq 1). Embedded Image Embedded Image

Where Infection is a binary indicator of infection status by age 1 year; PM2.5-first and PM2.5-last are the average exposure in the first and last developmental windows or trimesters, respectively, and … indicates all windows between; postnatal PM2.5 is the average exposure during the first year of life; Temp-first and Temp-last are the average temperature the first and last developmental windows or trimesters, respectively, and … indicates all windows between; sex categorizes the biological sex of the infant as male, female or unknown; preterm is a binary indicator of birth with gestational week less than 37; low birthweight is a binary indicator of birth weight less than 2500 g; age indicates maternal age at delivery; smoking categorizes maternal smoking status as never smoker, former smoker, or current smoker; BMI indicates whether the mother was underweight, normal, overweight or obese; and SES represents maternal neighbourhood income quintiles (1 = most deprived to 5 = least deprived) during pregnancy.

Ethical approval

This work received approval from the UBC Research Ethics Board (Certificate number: H20-01077) and Population Data BC (approval number: 20-197).

Results

Descriptive summary

There were 164,415 singleton births in BC that were prenatally exposed to the peak wildfire seasons (July – September) of 2016 to 2019. Of these, 134,475 were matched to 121,280 unique mothers and included in the study (Figure 2). Maternal characteristics (Table 1) were calculated based on each pregnancy. The following 29,940 infants were excluded from the study (Figure 2): stillbirths (n=698); infants with more than 25% missing PM2.5 exposure data or more than 25% missing temperature data, and infants with missing PM2.5 exposure and temperature data at any critical developmental windows (n=27,848); infants that could not be matched to the MSP outpatient physician billings or PharmaNet files (n=1,393); Infants with missing maternal age (n=1).

Figure 2.
Figure 2. Flowchart for the inclusion and exclusion of infants in the population-based cohort

Among all infants in the study, 16.4% (n=22,093) were diagnosed with one or more of the outcomes of interest (i.e., cases) and 83.6% (n=112,382) did not have any outcomes of interest (i.e., controls). There were 9,508 cases of otitis media, of which 6,395 were associated with amoxicillin dispensations. There were 14,464 cases of lower respiratory infections, of which 3,170 were associated with amoxicillin dispensations.

Significant differences (p-value <0.05) were observed in the distributions of all covariates and between cases and controls (Table 1). The biggest differences were male sex (57% cases, 50% controls), preterm births (3.7% cases, 2.7% controls), an ever smokers (13.4% cases, 11.6% controls). Small but statistically significant differences between groups were also observed in the PM2.5 and temperature exposures. The median (IQR) daily wildfire-related PM2.5 exposure during pregnancy was 7.5 (2.0) μg/m3 for cases and 7.4 (2.0) μg/m3 for controls (Table 1).

Wildfire-related PM2.5 exposure, temperature, and respiratory infections all showed clear seasonal trends by epidemiologic week across the study period (Figure 3). The wildfire-related PM2.5 and temperature peak in mid-summer, while the number of infections peak in the winter months. Furthermore, there were clear differences in wildfire-related PM2.5 exposure and outcomes among the infants prenatally exposed during the severe 2017 and 2018 wildfire seasons compared with those exposed during the 2016 and 2019 seasons (Figure 3).

Figure 3.
Figure 3.

The time trends of exposure and outcomes of interests by epidemiologic week (Epiweek) for infants prenatally exposed during the severe 2017 and 2018 wildfire seasons (red) and the 2016 and 2019 seasons (blue).

Prenatal exposure to wildfire-related PM2.5 and otitis media

Exposure to wildfire-related PM2.5 during the fourth stage (19-28 weeks) of Eustachian tube development was significantly associated with an increased odds of otitis media diagnosis and associated amoxicillin dispensations. The odds ratio (OR) [95% confidence interval] for a 10 μg/m3 increase in average PM2.5 exposure during the fourth window was 1.08 [1.02, 1.16] for all diagnoses and 1.11 [1.02,1.20] for those associated with amoxicillin dispensations. A protective effect of wildfire-related PM2.5 was observed in both the first (0-9 weeks) and second (10-12 weeks) stages of Eustachian tube development (Figure 4).

Figure 4.
Figure 4.

Adjusted odds ratios for prenatal exposure to wildfire-related PM2.5 and all diagnoses of otitis media in the first year of life (red), as well as otitis media diagnoses with associated amoxicillin dispensations (blue). Plot shows the effects of exposures during critical windows of Eustachian tube development (left) and trimesters (right).

Similarly, the second trimester (14-27 weeks) was significantly associated with increased odds of otitis media diagnosis and associated amoxicillin dispensations by age 1 year. The OR for a 10 μg/m3 increase in wildfire-related PM2.5 was 1.10 [1.03, 1.20] for otitis media diagnosis and 1.12 [1.03, 1.21] for diagnoses with associated amoxicillin dispensations. Once again, exposure during the first trimester (0 – 13 weeks) was associated with a protective effect (Figure 4). The sensitivity analyses using exposure estimates from alternate residential histories and conditioned on epidemiologic week of conception produced very similar results (eFigure 2).

Prenatal exposure to wildfire-related PM2.5 and lower respiratory infections

Exposure to wildfire-related PM2.5 during the canalicular stage (18-27 weeks) was significantly associated with an increased odds of lower respiratory infection diagnosis and associated amoxicillin dispensations. The odds ratio (OR) [95% confidence interval] for a 10 μg/m3 increase in average PM2.5 exposure during the canalicular stage was 1.08 [1.03, 1.14] for all diagnoses and 1.12 [1.01,1.25] for those associated with amoxicillin dispensations. Exposure during the pseudoglandular stage (8-17 weeks) was associated with a protective effect for lower respiratory infections associated with an amoxicillin dispensation. The second trimester (14-27 weeks) was significantly associated with increased odds of lower respiratory tract infection diagnosis by age 1 year with an OR of 1.06 [1.00, 1.12]. There were no significant associations observed for amoxicillin dispensations related to lower respiratory tract infection diagnosis for the analyses by trimester (Figure 5). Once again, the sensitivity analyses showed very similar results (eFigure 3).

Figure 5.
Figure 5.

Adjusted odds ratios for prenatal exposure to wildfire-related PM2.5 and all diagnoses of lower respiratory infection in the first year of life (red), as well as lower respiratory infection diagnoses with associated amoxicillin dispensations (blue). Plot shows the effects of exposures during critical windows of lower respiratory tract development (left) and trimesters (right).

Comments

Principle finding

This population-based case-control study examined wildfire-related PM2.5 exposure during critical windows of respiratory prenatal development and associations with respiratory infections by age 1 year. We found that wildfire-related PM2.5 exposure during the second trimester was associated with increased odds of otitis media and lower respiratory infections, including those treated with amoxicillin. We also consistently observed that wildfire-related PM2.5 exposure during the first trimester was associated with protective effects. Analyses based on critical windows of organ system development may provide more information than those based on trimester when examining the effects of environmental exposures during pregnancy.

Strengths

Our study has several strengths. First, we used the rich, population-based BCPDR to include most infants in BC who were prenatally exposed to wildfire smoke from 2016-2019. Because wildfire smoke was highly variable across these four years, we have excellent exposure contrasts within the cohort. Second, we were able to link BCPDR data with population-based MSP and PharmaNet data to identify all infections and those with associated amoxicillin dispensations. Third, we matched the analyses on epidemiologic week of birth to address confounding by seasonality, which is associated with both the exposure and the outcome. Fourth, we used a high-quality model of wildfire-related PM2.5 at high spatial resolution to assign daily residential exposures during pregnancy and the first year of life. Finally, we adjusted analyses for postnatal exposure to wildfire-related PM2.5, which can also be associated with the outcomes of interest.

Limitations

Our study used residential history as an indicator for exposure measurements, but we cannot account for mobility or errors in the residential history data, which leads to exposure misclassification and potential attenuation of the true effects. Second, we did not have individual-level SES information, and we used neighbourhood-level SES as approximation. Neighbourhood-level SES can even out the distribution of outcomes across income decile, and cause the estimates to be biased towards the null48. Third, we were unable to incorporate critical covariates identified in literature including maternal education and exposure to second-hand smoke because this information is not reliably captured in the BCPDR. Finally, all analyses were based on administrative data that are not collected for health research purposes. We lacked clinical details of each case, and we assume that physician billings for respiratory infections are a robust indicator of a true infection, especially when followed by an antibiotic dispensation.

Interpretation

Our findings are consistent with Hsu. et al., who reported that the mid-pregnancy period (16–25 weeks) is a critical window for the relationship between PM2.5 exposure and diagnosis with asthma by the age of six 49. Hazlehurst et al. also reported that higher PM2.5 exposure during the canalicular stage of lower respiratory tract development was linked to an increased risk of asthma 50. Another study identified the pseudoglandular and canalicular stages as the most critical periods for prenatal PM2.5 exposure and the subsequent development of asthma and wheezing after birth51. The critical developmental windows identified for respiratory outcomes are consistent with the sensitive exposure windows we report.

One other observational study specific to wildfire smoke found that the second and third trimesters were the most sensitive windows for use of upper respiratory medications before age 8 years52. We also observed the increased risk of otitis media associated with exposure during the second trimester. Another study using a mouse model found that maternal exposure to PM2.5 had a greater impact in the later stages of pregnancy, which could be related to the duration of exposure and accumulation of PM2.5.

Different from earlier studies, we consistently observed that wildfire-related PM2.5 exposure during the first stages of pregnancy was associated with protective effects for respiratory infections. We do not have an explanation for this observation, but we hypothesize that it could be related to a healthy survivor effect. Infants do not enter the BCPDR until 15 weeks’ gestation, and thus we have no information regarding the loss of earlier pregnancies. However, increased prenatal exposure to PM2.5 has been associated with increased risk of pregnancy loss53, and population exposure to smoke has been associated with a significant reductions in the size of birth cohorts54. There are also several studies indicating that elevated exposure to air pollution during preconception and early pregnancy is associated with spontaneous abortion prior to 20 weeks 55–57. If wildfire smoke exposure in early pregnancy leads to pregnancy loss, the foetuses that survive may develop into healthier infants and children 58.

Our results suggest that prenatal exposure to wildfire-related PM2.5 may interfere with foetal development. During development of the Eustachian tubes, the fourth stage (19-28 weeks) emerges as a pivotal period characterized by differentiation of certain seromucous glandular extensions? within the lumens, which is associated with maturation of submucosal gland 41. The glandular mucus secreted by the submucosal gland aids in the removal of inhaled particles and provides protection to the respiratory airway by combating bacterial threats59.

We hypothesize that prenatal exposure to wildfire-related PM2.5 during the fourth stage of Eustachian tube development may be associated with inadequate submucosal gland development. The compromise in the protective functioning of the Eustachian tube potentially increases susceptibility to upper respiratory infections, such as otitis media, during early life. Similarly, the canalicular stage of foetal lower respiratory tract development is associated with the differentiation of type I and type II cells, as well as formation of the alveolar-capillary barrier. Insufficient maturation during the canalicular stage may result in underdeveloped peripheral airways, hindering efficient gas exchange and respiratory balance. This, in turn, can contribute to the proliferation of alveoli populated by inflammatory cells 19,60, which might make infants more prone to lower respiratory infections.

We identified a significant association between prenatal exposure to wildfire-related PM2.5 and respiratory infections in infants, as evidenced by increased odds of infection diagnoses and related antibiotic dispensations. To mitigate the detrimental impact of wildfire smoke on infant respiratory health, clear communication regarding the risks and protective measures is needed for those who are pregnant. For example, use of indoor air cleaning technologies should be strongly recommended for this population 61. Future studies should focus on the biological mechanisms underlying the effects of prenatal exposure to wildfire smoke, including epigenetics and markers of inflammation. Additionally, extending the follow-up period of the cohort will enable us to examine the longer-term effects associated with prenatal exposure to wildfire smoke.

Data Availability

Access to data provided by the Data Stewards is subject to approval but can be requested for research projects through the Data Stewards or their designated service providers. All inferences, opinions, and conclusions drawn in this publication are those of the author(s), and do not reflect the opinions or policies of the Data Stewards (the British Columbia Ministry of Health, the Perinatal Services BC).

Author Contributions

Qingyi Lan: Conceptualization; formal analysis; writing-original draft preparation; writing-review and editing. Kate Weinberger: Conceptualization; writing-review and editing. Sabrina Luke: Conceptualization; writing-review and editing. Eric Lavigne: Conceptualization; writing-review and editing. Scott Weichenthal: Conceptualization; writing-review and editing. Sarah B. Henderson: Supervision; conceptualization; writing-reviewing and editing.

Conflict of Interest disclosure

None.

Disclaimer

Access to data provided by the Data Stewards is subject to approval but can be requested for research projects through the Data Stewards or their designated service providers. All inferences, opinions, and conclusions drawn in this publication are those of the author(s), and do not reflect the opinions or policies of the Data Stewards (the British Columbia Ministry of Health, the Perinatal Services BC).

Acknowledgements

This study was funded by Canadian Institutes of Health Research Grant (CIHR). Qingyi Lan received Graduate Research Stipend from the CIHR, International Tuition Award, President’s Academic Excellence Initiative PhD Award, and Faculty of Medicine Graduate Award.

References

  1. 1.↵
    Gheissari R, Liao J, Garcia E, Pavlovic N, Gilliland FD, Xiang AH, et al. Health Outcomes in Children Associated with Prenatal and Early-Life Exposures to Air Pollution: A Narrative Review. Toxics (Basel) 2022;10:458.
    OpenUrl
  2. 2.↵
    Proietti E, Röösli M, Frey U, Latzin P. Air pollution during pregnancy and neonatal outcome: a review. Journal of aerosol medicine and pulmonary drug delivery 2013;26:9.
    OpenUrl
  3. 3.↵
    Korten I, Ramsey K, Latzin P. Air pollution during pregnancy and lung development in the child. Paediatric respiratory reviews 2016;21:38–46.
    OpenUrlPubMed
  4. 4.↵
    Aguilera R, Corringham T, Gershunov A, Benmarhnia T. Wildfire smoke impacts respiratory health more than fine particles from other sources: observational evidence from Southern California. Nature communications 2021;12:1493–8.
    OpenUrlPubMed
  5. 5.↵
    Franzi LM, Bratt JM, Williams KM, Last JA. Why is particulate matter produced by wildfires toxic to lung macrophages? Toxicology and applied pharmacology 2011;257:182– 188.
    OpenUrlCrossRefPubMed
  6. 6.↵
    Guo C, Hoek G, Chang L-Y, Bo Y, Lin C, Huang B, et al. Long-Term Exposure to Ambient Fine Particulate Matter (P M 2.5) and Lung Function in Children, Adolescents, and Young Adults: A Longitudinal Cohort Study. Environmental health perspectives 2019;127:127008.
    OpenUrlPubMed
  7. 7.
    Haikerwal A, Akram M, Sim MR, Meyer M, Abramson MJ, Dennekamp M. Fine particulate matter (PM2.5) exposure during a prolonged wildfire period and emergency department visits for asthma. Respirology (Carlton, Vic.) 2016;21:88–94.
    OpenUrlCrossRefPubMed
  8. 8.↵
    Reid CE, Brauer M, Johnston FH, Jerrett M, Balmes JR, Elliott CT. Critical Review of Health Impacts of Wildfire Smoke Exposure. Environmental health perspectives 2016;124:1334–1343.
    OpenUrlCrossRefPubMed
  9. 9.↵
    Willis GA, Chappell K, Williams S, Melody SM, Wheeler A, Dalton M, et al. Respiratory and atopic conditions in children two to four years after the 2014 Hazelwood coalmine fire. Medical journal of Australia 2020;213:269–275.
    OpenUrlPubMed
  10. 10.↵
    Jedrychowski WA, Perera FP, Maugeri U, Mrozek-Budzyn D, Mroz E, Klimaszewska-Rembiasz M, et al. Intrauterine exposure to polycyclic aromatic hydrocarbons, fine particulate matter and early wheeze. Prospective birth cohort study in 4-year olds. Pediatric allergy and immunology 2010;21:e723–e732.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Jamison DT,
    2. Breman JG,
    3. Measham AR,
    4. Alleyne G,
    5. Claeson M,
    6. Evans DB
    Simoes EAF, Cherian T, Chow J, Shahid-Salles SA, Laxminarayan R, John TJ. Acute Respiratory Infections in Children. In: Disease Control Priorities in Developing Countries. Editors: Jamison DT, Breman JG, Measham AR, Alleyne G, Claeson M, Evans DB, et al: 2nd ed. Washington (DC): World Bank, 2006.
  12. 12.↵
    Santibanez P, Gooch K, Vo P, Lorimer M, Sandino Y. Acute care utilization due to hospitalizations for pediatric lower respiratory tract infections in British Columbia, Canada. BMC health services research 2012;12:451–451.
    OpenUrlPubMed
  13. 13.↵
    Patrick DM, Sbihi H, Dai DLY, Al Mamun A, Rasali D, Rose C, et al. Decreasing antibiotic use, the gut microbiota, and asthma incidence in children: evidence from population-based and prospective cohort studies. The lancet respiratory medicine 2020;8:1094–1105.
    OpenUrlPubMed
  14. 14.↵
    Sur DKC, Plesa ML. Antibiotic Use in Acute Upper Respiratory Tract Infections. American family physician 2022;106:628–636.
    OpenUrlPubMed
  15. 15.↵
    Fireman P. Otitis media and eustachian tube dysfunction: Connection to allergic rhinitis. Journal of allergy and clinical immunology 1997;99:s787–s797.
    OpenUrl
  16. 16.↵
    Mazer BD. Otitis Media. In: Pediatric Allergy: Principles and Practice. Editors: Leung DYM MD PhD Faaaai, Szefler SJ MD, Bonilla FA MD PhD, Akdis CA MD, Sampson HA MD: Third., 2016; pp. 219–227.e3.
  17. 17.↵
    Dasaraju PV, Liu C. Infections of the Respiratory System. In: Medical Microbiology. Editor: Baron S: 4th ed. Galveston (TX): University of Texas Medical Branch at Galveston, 1996.
  18. 18.↵
    Biyyam DR, Chapman T, Ferguson MR, Deutsch G, Dighe MK. Congenital lung abnormalities: embryologic features, prenatal diagnosis, and postnatal radiologic-pathologic correlation. Radiographics 2010;30:1721–1738.
    OpenUrlCrossRefPubMed
  19. 19.↵
    Kotecha S. Lung growth: implications for the newborn infant. Archives of Disease in Childhood - Fetal and Neonatal Edition 2000;82:F69–F74.
    OpenUrlCrossRef
  20. 20.↵
    Shiny Sherlie V, Varghese A. ENT Changes of Pregnancy and Its Management. Indian Journal of Otolaryngology and Head & Neck Surgery 2014;66:6–9.
    OpenUrl
  21. 21.↵
    British Columbia, Ministry of Forests and Range. The state of British Columbia’s forests 2006. Victoria, B.C.: Ministry of Forests and Range; 2007.
  22. 22.↵
    BC Wildfire Service. Wildfire Season Summary - Province of British Columbia. https://www2.gov.bc.ca/gov/content/safety/wildfire-status/about-bcws/wildfire-history/wildfire-season-summary (last accessed February 2022).
  23. 23.↵
    BC Wildfire Service. Wildfire Averages - Province of British Columbia. https://www2.gov.bc.ca/gov/content/safety/wildfire-status/about-bcws/wildfire-statistics/wildfire-averages (last accessed March 2022).
  24. 24.↵
    Perinatal Services BC(2021): British Columbia Perinatal Data Registry. V2. Population Data BC. Data Extract. PSBC (2021). http://www.popdata.bc.ca/data
  25. 25.↵
    BC Perinatal Data Registry | Population Data BC. https://www.popdata.bc.ca/data/health/perinatal (last accessed July 2023).
  26. 26.↵
    Ark TK, Kesselring S, Hills B, McGrail KM. Population Data BC: Supporting population data science in British Columbia. International journal of population data science 2020;4:1133.
    OpenUrlPubMed
  27. 27.↵
    British Columbia Ministry of Health [creator](2020): Medical Services Plan (MSP) Payment Information File. V2. Population Data BC [publisher]. Data Extract. MOH (2021). http://www.popdata.bc.ca/data.
  28. 28.↵
    British Columbia Ministry of Health [creator](2021): Consolidation File (MSP Registration & Premium Billing). V2. Population Data BC [publisher]. Data Extract. MOH (2021). http://www.popdata.bc.ca/data.
  29. 29.↵
    British Columbia Ministry of Health [creator](2021): PharmaNet. V2. Population Data BC [publisher]. Data Extract. MOH (2021). http://www.popdata.bc.ca/data
  30. 30.↵
    Paul N, Yao J, McLean KE, Stieb DM, Henderson SB. The Canadian Optimized Statistical Smoke Exposure Model (CanOSSEM): A machine learning approach to estimate national daily fine particulate matter (PM2.5) exposure. The Science of the total environment 2022;850:157956–157956.
    OpenUrlCrossRefPubMed
  31. 31.↵
    Le Saux N, Robinson J. Pneumonia in healthy Canadian children and youth: Practice points for management. Paediatrics & child health 2011;16:417–420.
    OpenUrlPubMed
  32. 32.↵
    Le Saux N, Robinson JL, Canadian Paediatric Society ID and IC. Management of acute otitis media in children six months of age and older. Paediatrics & child health 2016;21:39– 50.
    OpenUrlPubMed
  33. 33.↵
    Harskamp-van Ginkel MW, London SJ, Magnus MC, Gademan MG, Vrijkotte TG. A Study on Mediation by Offspring BMI in the Association between Maternal Obesity and Child Respiratory Outcomes in the Amsterdam Born and Their Development Study Cohort. . Editor: Dewan A: PloS one 2015;10:e0140641.
    OpenUrlPubMed
  34. 34.↵
    Lee A, Leon Hsu H-H, Mathilda Chiu Y-H, Bose S, Rosa MJ, Kloog I, et al. Perinatal air pollution exposure. Journal of allergy and clinical immunology 2018;141:1880–1886.
    OpenUrl
  35. 35.↵
    Torchin H, Le Lous M, Houdouin V. In Utero Exposure to Maternal Smoking: Impact on the Child from Birth to Adulthood - CNGOF-SFT Expert Report and Guidelines for Smoking Management during Pregnancy. Gynecologie, obstetrique, fertilite & senologie 2020;48:567.
    OpenUrlPubMed
  36. 36.↵
    Sbihi H, Tamburic L, Koehoorn M, Brauer M. Perinatal air pollution exposure and development of asthma from birth to age 10 years. The European respiratory journal 2016;47:1062–1071.
    OpenUrlPubMed
  37. 37.↵
    Decrue F, Gorlanova O, Salem Y, Vienneau D, de Hoogh K, Gisler A, et al. Increased Impact of Air Pollution on Lung Function in Preterm versus Term Infants: The BILD Study. American journal of respiratory and critical care medicine 2022;205:99–107.
    OpenUrlPubMed
  38. 38.↵
    Jung C-R, Chen W-T, Tang Y-H, Hwang B-F. Fine particulate matter exposure during pregnancy and infancy and incident asthma. Journal of allergy and clinical immunology 2019;143:2254–2262.e5.
    OpenUrl
  39. 39.
    Lavigne É, Talarico R, van Donkelaar A, Martin RV, Stieb DM, Crighton E, et al. Fine particulate matter concentration and composition and the incidence of childhood asthma. Environment international 2021;152:106486.
    OpenUrlPubMed
  40. 40.↵
    Ostro B, Roth L, Malig B, Marty M. The Effects of Fine Particle Components on Respiratory Hospital Admissions in Children. Environmental health perspectives 2009;117:475–480.
    OpenUrlCrossRefPubMedWeb of Science
  41. 41.↵
    Dornhoffer JL, Leuwer R, Schwager K, Wenzel S. A Practical Guide to the Eustachian Tube. Berlin, Heidelberg: Springer Berlin Heidelberg; 2014.
  42. 42.↵
    McEvoy CT, Spindel ER. Environmental Effects on Lung Morphogenesis and Function:: Tobacco Products, Combustion Products, and Other Sources of Pollution. In: Fetal and Neonatal Lung Development: Clinical Correlates and Technologies for the Future. Editors: Jobe AH, Whitsett JA, Abman SH: Cambridge: Cambridge University Press, 2016; pp. 77–93.
  43. 43.↵
    Stocks J, Hislop A, Sonnappa S. Early lung development: lifelong effect on respiratory health and disease. The lancet respiratory medicine 2013;1:728–742.
    OpenUrlPubMed
  44. 44.↵
    Maternity. First, Second & Third Trimester | Pregnancy Milestones and Stages. https://maternity.jacksonhealth.org/pregnancy-milestones/first-second-third-trimester/ (last accessed April 2023).
  45. 45.↵
  46. 46.↵
    R: a language and environment for statistical computing. https://www.gbif.org/tool/81287/r-a-language-and-environment-for-statistical-computing (last accessed June 2023).
  47. 47.↵
    Huang M, Strickland MJ, Richards M, Warren JL, Chang HH, Darrow LA. Confounding by Conception Seasonality in Studies of Temperature and Preterm Birth: A Simulation Study. Epidemiology (Cambridge, Mass.) 2023;34:439–449.
    OpenUrlPubMed
  48. 48.↵
    Hanley GE, Morgan S. On the validity of area-based income measures to proxy household income. BMC health services research 2008;8:79–79.
    OpenUrlPubMed
  49. 49.↵
    Hsu H-HL, Chiu Y-HM, Coull BA, Kloog I, Schwartz J, Lee A, et al. Prenatal Particulate Air Pollution and Asthma Onset in Urban Children. Identifying Sensitive Windows and Sex Differences. American journal of respiratory and critical care medicine 2015;192:1052–1059.
    OpenUrlCrossRefPubMed
  50. 50.↵
    Hazlehurst MF, Carroll KN, Loftus CT, Szpiro AA, Moore PE, Kaufman JD, et al. Maternal exposure to PM2.5 during pregnancy and asthma risk in early childhood. Environmental epidemiology 2021;5:e130.
    OpenUrl
  51. 51.↵
    Chen G, Zhou H, He G, Zhu S, Sun X, Ye Y, et al. Effect of early-life exposure to PM2.5 on childhood asthma/wheezing: a birth cohort study. Pediatric Allergy and Immunology 2022;33.
  52. 52.↵
    Dhingra R, Keeler C, Staley BS, Jardel HV, Ward-Caviness C, Rebuli ME, et al. Wildfire smoke exposure and early childhood respiratory health: a study of prescription claims data. Environmental health 2023;22:1–48.
    OpenUrlPubMed
  53. 53.↵
    Ha S, Ghimire S, Martinez V. Outdoor Air Pollution and Pregnancy Loss: a Review of Recent Literature. Current epidemiology reports 2022;9:387–405.
    OpenUrl
  54. 54.↵
    Jayachandran S. Air Quality and Early-Life Mortality: Evidence from Indonesia’s Wildfires. The Journal of human resources 2009;44:916–954.
    OpenUrlAbstract/FREE Full Text
  55. 55.↵
    Enkhmaa D, Warburton N, Javzandulam B, Uyanga J, Khishigsuren Y, Lodoysamba S, et al. Seasonal ambient air pollution correlates strongly with spontaneous abortion in Mongolia. BMC pregnancy and childbirth 2014;14:146–146.
    OpenUrl
  56. 56.
    Green RS, Malig B, Windham GC, Fenster L, Ostro B, Swan S. Residential Exposure to Traffic and Spontaneous Abortion. Environmental health perspectives 2009;117:1939–1944.
    OpenUrlCrossRefPubMedWeb of Science
  57. 57.↵
    Zhang L, Liu W, Hou K, Lin J, Zhou C, Tong X, et al. Air pollution-induced missed abortion risk for pregnancies. Nature sustainability 2019;2:1011–1017.
    OpenUrl
  58. 58.↵
    Williams AD, Kanner J, Grantz KL, Ouidir M, Sheehy S, Sherman S, et al. Air pollution exposure and risk of adverse obstetric and neonatal outcomes among women with type 1 diabetes. Environmental research 2021;197:111152–111152.
    OpenUrl
  59. 59.↵
    Widdicombe JH, Wine JJ. Airway Gland Structure and Function. Physiological reviews 2015;95:1241–1319.
    OpenUrlCrossRefPubMed
  60. 60.↵
    Gurka DP, Balk RA. 36 - Acute Respiratory Failure. In: Critical Care Medicine Fifth. Elsevier Inc, 2019; pp. 591–604.e8.
  61. 61.↵
    Barn PK, Elliott CT, Allen RW, Kosatsky T, Rideout K, Henderson SB. Portable air cleaners should be at the forefront of the public health response to landscape fire smoke. Environmental health 2016;15:116–116.
    OpenUrlPubMed
Back to top
Posted September 10, 2023.
Download PDF

Supplementary Material

Data/Code
Email
Prenatal exposure to wildfire-related PM2.5 and respiratory infections by age 1 year: A population-based case-control analysis of critical developmental windows
Qingyi Lan, Kate Weinberger, Sabrina Luke, Eric Lavigne, Scott Weichenthal, Sarah B. Henderson
medRxiv 2023.09.09.23295304; doi: https://doi.org/10.1101/2023.09.09.23295304
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area