Association of Wildfire Air Pollution and Health Care Use for Atopic Dermatitis and Itch

Key Points

Question  Is short-term exposure to wildfire-associated air pollution associated with the frequency of outpatient clinic visits for atopic dermatitis and itch?

Findings  In this cross-sectional time-series study of 8049 dermatology clinic visits by 4147 patients, short-term wildfire-associated air pollution exposure was significantly associated with increased rates of weekly appointments. The increases were seen in pediatric and adult appointments for atopic dermatitis and in pediatric appointments for itch.

Meaning  In this study, short-term exposure to poor air quality due to wildfire smoke appeared to be associated with exacerbations of patients’ skin conditions as measured by increased use of health care services for the management of atopic dermatitis and itch.

Importance  Air pollution is a worldwide public health issue that has been exacerbated by recent wildfires, but the relationship between wildfire-associated air pollution and inflammatory skin diseases is unknown.

Objective  To assess the associations between wildfire-associated air pollution and clinic visits for atopic dermatitis (AD) or itch and prescribed medications for AD management.

Design, Setting, and Participants  This cross-sectional time-series study assessed the associations of air pollution resulting from the California Camp Fire in November 2018 and 8049 dermatology clinic visits (4147 patients) at an academic tertiary care hospital system in San Francisco, 175 miles from the wildfire source. Participants included pediatric and adult patients with AD or itch from before, during, and after the time of the fire (October 2018 through February 2019), compared with those with visits in the same time frame of 2015 and 2016, when no large wildfires were near San Francisco. Data analysis was conducted from November 1, 2019, to May 30, 2020.

Exposures  Wildfire-associated air pollution was characterized using 3 metrics: fire status, concentration of particulate matter less than 2.5 μm in diameter (PM_2.5), and satellite-based smoke plume density scores.

Main Outcomes and Measures  Weekly clinic visit counts for AD or itch were the primary outcomes. Secondary outcomes were weekly numbers of topical and systemic medications prescribed for AD in adults.

Results  Visits corresponding to a total of 4147 patients (mean [SD] age, 44.6 [21.1] years; 2322 [56%] female) were analyzed. The rates of visits for AD during the Camp Fire for pediatric patients were 1.49 (95% CI, 1.07-2.07) and for adult patients were 1.15 (95% CI, 1.02-1.30) times the rate for nonfire weeks at lag 0, adjusted for temperature, relative humidity, patient age, and total patient volume at the clinics for pediatric patients. The adjusted rate ratios for itch clinic visits during the wildfire weeks were 1.82 (95% CI, 1.20-2.78) for the pediatric patients and 1.29 (95% CI, 0.96-1.75) for adult patients. A 10-μg/m³ increase in weekly mean PM_2.5 concentration was associated with a 7.7% (95% CI, 1.9%-13.7%) increase in weekly pediatric itch clinic visits. The adjusted rate ratio for prescribed systemic medications in adults during the Camp Fire at lag 0 was 1.45 (95% CI, 1.03-2.05).

Conclusions and Relevance  This cross-sectional study found that short-term exposure to air pollution due to the wildfire was associated with increased health care use for patients with AD and itch. These results may provide a better understanding of the association between poor air quality and skin health and guide health care professionals’ counseling of patients with skin disease and public health practice.

Air pollution is a worldwide public health issue: in 2015, exposure to ambient particulate matter less than 2.5 μm in diameter (PM_2.5) contributed to 4.2 million deaths and 103.1 million disability-adjusted life-years.¹ Air pollution is a complex mixture of gaseous molecules and suspended solid and liquid particles that contributes to the development and exacerbation of respiratory, cardiovascular, neurologic, and other diseases.^2-5 The association between air pollution and skin health, however, is relatively underexplored and less apparent to clinicians and researchers.

The incidence of atopic dermatitis (AD) has been increasing in industrialized countries, and it affects approximately 15% to 20% of children worldwide.^6,7 Pathways through which air pollution may contribute to the development and exacerbation of AD include activation of the aryl hydrocarbon receptor pathway,^8,9 generation of reactive oxygen species,^10-12 and induction of a proinflammatory response.^9,13-16 The epidemiologic evidence for an association between air pollution and AD is mixed; some studies report null findings,^17-21 but many others report a positive association.^22-30

Epidemiologic studies on skin diseases have primarily focused on chronic exposure to air pollution, but the association between acute exposure and common cutaneous diseases, such as AD, is becoming increasingly important to assess as environmental conditions change. An increase in the frequency and severity of wildfires is contributing to a recent increase in PM_2.5 concentrations in the western US after a decades-long reduction in air pollution.^4,31-34 Most of the largest wildfires in the US since the 1950s have occurred within the past decade.^35,36 For example, the Camp Fire in Northern California during November 2018 was one of the deadliest and most destructive wildfires in modern California history.³⁷ After 2018, record-breaking wildfires in the western US burned over 6 million acres of land, and the resulting smoke spread to the eastern US and even northern Europe.^38,39 During this time, severe wildfires also occurred in Brazil and Australia,^36,40 suggesting that wildfire smoke is becoming an increasingly pressing issue as the climate continues to warm and conditions become drier in some regions.

Increased outdoor air pollution attributable to wildfires is associated with greater health care services use and costs owing to a greater number of emergency department visits for urgent health conditions.^41-43 As wildfire frequency and intensity increase across the world, we sought to study the short-term outcomes of wildfire-associated air pollution on skin health and health care use to inform patient care and public health practice. We hypothesized that acute exposure to wildfire-associated air pollution would be associated with increased dermatology clinic visits for AD or itch. The California Camp Fire posed a unique opportunity to test this hypothesis, since particulate matter generated by the fire caused a significant spike in air pollution concentrations for 2 weeks in San Francisco, California, located 175 miles from the site of the wildfire.³⁷

We collected environmental data for San Francisco and clinical data for patients seen at the University of California, San Francisco’s dermatology clinics. Data were collected for 18 weeks encompassing the time before, during, and after the Camp Fire (October 2018-February 2019) as well as for the corresponding dates from previous years when there were no large wildfires close to San Francisco (October 2015-February 2016 and October 2016-February 2017). Fall 2017 was excluded because of the presence of regional wildfires during that time that affected air quality in San Francisco. Data analysis was conducted from November 1, 2019, to May 30, 2020. The study was approved by the University of California, San Francisco Institutional Review Board with waiver of informed consent.

We collected daily 24-hour mean PM_2.5 concentrations (micrograms per cubic meter) from the Bay Area Air Quality Management District’s air quality monitoring station in San Francisco.⁴⁴ For all zip codes in San Francisco, mean daily smoke plume density scores, ranging from 0 to 3 (none, light, medium, and heavy), were collected from the National Oceanic and Atmospheric Administration Hazard Mapping System for Fire and Smoke.⁴⁵ The Hazard Mapping System combines satellite-detected fire information with analysis of visible imagery. We found insignificant spatial variation in daily smoke plume scores across zip codes in San Francisco. Therefore, for our analysis, citywide population-weighted mean scores were calculated to characterize daily smoke plume density by using the number of residents in each zip code.⁴⁶ Daily mean temperature and relative humidity for San Francisco were obtained from the National Oceanic and Atmospheric Administration Local Climatological Data⁴⁷ because these factors may affect AD symptoms.^22,23

Using specific codes from the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), outpatients for whom a dermatologist indicated a diagnosis of AD or itch at a clinic visit were identified during the periods assessed. The ICD-10 codes used for eczema/AD were L30.9, L20.82, L20.84, L20.9, H01.139, H01.136 L, H01.133 R, and L28.0 and, for itch, the codes were L29.9 and L29.8. Patients were excluded if they resided outside San Francisco (approximately 20% of the original group). Patients’ age, sex, and race/ethnicity, prescribed AD medications for adult patients, and the total number of visits for any skin health–related concern were also recorded. The medication data for pediatric patients with AD were excluded because of low counts.

Exposure, outcome (counts), and covariate data were aggregated on a weekly basis for analysis. We calculated the weekly sum of clinic visits and prescribed medications and the weekly mean PM_2.5 concentration, smoke plume density score, temperature, relative humidity, and age of patients. The counts for weekly visits and medications displayed underdispersion; therefore, we used a generalized Poisson regression model.^48,49 Counts from pediatric and adult patients were analyzed separately. Each outcome was analyzed in models that included a specific exposure metric as the primary predictor: fire status (a binary indicator of whether a wildfire occurred during a week), plume density score, or PM_2.5. To account for potential delays in seeking care during a wildfire or developing symptoms following exposure, four 1-week cumulative exposure lags were constructed for each exposure metric and analyzed independently in separate models. The 1-week exposure lags estimate the delayed effects of exposure by representing a fixed amount of passing time. For example, cumulative lag 1 includes pollution exposure during a given week and from the week before, and cumulative lag 2 includes pollution exposure during a given week and from the previous 2 weeks. Covariates in the models included average temperature, relative humidity, patient age, and period (2015-2016, 2016-2017, or 2018-2019). In addition, each model included an offset variable, the logarithm of the weekly total number of dermatology clinic visits for any skin health–related concern, to account for changes in the number of practicing dermatologists and total weekly patient volume both within and across years. The models also included a holiday week indicator variable that accounted for weeks that had US national holidays, such as Thanksgiving. We compared the rates of patient visits and prescriptions during the exposure period vs the nonexposure period.

The primary outcomes of interest were the weekly total numbers of adult and pediatric clinics visits for AD or itch. We conducted sensitivity analyses using simple exposure lags and combined 2-week exposure lags for the fire status metric. For each outcome, we report rate ratios and 95% CIs. Two-tailed P values were calculated; P < .05 was considered statistically significant. Data management and statistical analyses were conducted using Stata, version 16 (StataCorp LLC) and R, version 3.6.2 (R Foundation for Statistical Computing).

The California Camp Fire started on November 8, 2018, and was contained 17 days later, on November 25.⁵⁰ Due to prevailing wind patterns, the burning caused a 9-fold increase in average weekly PM_2.5 concentration in San Francisco during a 2-week period (November 8-21, 2018), compared with baseline before and after; it peaked during the second week of the fire (Figure 1). The smoke plume density score, a measurement of the haziness of atmospheric conditions attributable to the wildfire, achieved its maximum value during the beginning of the 2 weeks. During this time, the average temperature in San Francisco was similar to baseline, and the relative humidity was reduced (eTable 1 in the Supplement). Compared with the preceding weeks, the total number of clinic appointments for AD or itch increased during the second week of the Camp Fire and the following 4 weeks, with the exception of the national Thanksgiving holiday week.

Patient demographic characteristics and the numbers of visits and medications are summarized in the Table and eTable 2 in the Supplement. The study population included 3448 unique adult patients and 699 unique pediatric patients, with a total of 6439 AD visits and 1610 itch visits across all 3 periods. There were more female (2322 [56%]) than male (1825 [44%]) patients seen overall. The mean (SD) age of the patients was 44.6 (21.1) years. During the Camp Fire, 89% of the adults who presented with itch were not previously diagnosed with AD, compared with patients with itch during the same 2 weeks in 2015 (59%) and 2016 (51%).

The results from the generalized Poisson regression, adjusting for temperature, humidity, and patient age, demonstrated statistically significant associations between exposure to wildfire-associated air pollution and weekly clinic visits for AD in both adult and pediatric patients (Figure 2). The rates of weekly AD clinic visits during the Camp Fire were 1.49 (95% CI, 1.07-2.07) times the rate for nonwildfire weeks at lag 0 for pediatric patients and 1.15 (95% CI, 1.02-1.30) times the rate for nonwildfire weeks at lag 0 for adults. Rate ratios for smoke plume density and PM_2.5 exposures corresponded to a 1-U increase in the weekly average National Oceanic and Atmospheric Administration smoke plume density score and a 10-μg/m³ increase in weekly average PM_2.5 concentration. A 1-U increase in the average weekly National Oceanic and Atmospheric Administration score was associated with 1.12 (95% CI, 1.04-1.22) times more average weekly adult AD clinic visits. A 10-μg/m³ increase in weekly average PM_2.5 concentration was associated with a 5.1% (95% CI, 0.8%-9.7%) increase in the average weekly pediatric AD clinic visits.

The associations were larger for the pediatric population compared with the adult population and largest at lag 0 for all 3 exposure metrics. The rate ratios generally decreased from lag 0 to lag 4 for all 3 exposure metrics.

The adjusted rate ratios for weekly pediatric itch visits were 1.82 (95% CI, 1.20-2.77) for fire status and 1.55 (95% CI, 1.15-2.09) for average weekly smoke plume density score. A 10-g/m³ increase in weekly mean PM_2.5 concentration was associated with a 7.7% (95% CI, 1.9%-13.7%) increase in weekly pediatric itch visits (Figure 3A). Similar to the results for pediatric AD visits, the rate ratios for pediatric itch visits generally were largest for lag 0 and decreased from lag 0 to lag 4 for all 3 exposure metrics.

The adjusted rate ratios for adult itch visits were greater than 1 across all lags in the 3 exposure metrics. At lag 0, the rate ratio for the fire status exposure was 1.29 (95% CI, 0.96-1.75), for the smoke plume density exposure was 1.16 (95% CI, 0.94-1.43), and for the PM_2.5 exposure was 1.03 (95% CI, 0.99-1.07). However, unlike the results for pediatric itch visits, these results did not achieve statistical significance (Figure 3B).

Medication data were assessed separately for topical and systemic agents; the latter are generally prescribed for the treatment of more severe AD. The results showed a statistically significant association between air pollution exposure and weekly number of systemic medications, including the immunosuppressant prednisone, prescribed for treatment of AD in adults for the early lags in all 3 exposure metrics (Figure 4). The adjusted rate ratio for prescribed systemic medications during the Camp Fire at lag 0 was 1.45 (95% CI, 1.03-2.05). For the topical medication outcome, the adjusted rate ratios were greater than 1 but did not reach statistical significance, except for lag 0 for the smoke plume density metric: 1.14 (95% CI, 1.01-1.28).

All adjusted and unadjusted results are reported in eTable 3 and eTable 4 in the Supplement. Sensitivity analyses were performed for all outcomes using simple exposure lags and 2-week combined lags for fire status exposure (eTable 5 in the Supplement). These analyses generated results that were overall similar to those from the primary analysis.

For this study, we assessed the effects of the California Camp Fire, which caused a short-term spike in air pollution in San Francisco, on health care use for AD and itch. We found that wildfire-associated air pollution was significantly associated with increased rates of weekly visits for AD for pediatric and adult patients at an academic medical center’s dermatology clinics located 175 miles from the fire, adjusting for temperature, relative humidity, patient volume, and patient age. In addition, wildfire smoke exposure was associated with a significantly increased rate of clinic visits for itch among pediatric patients and rate of prescribed systemic medications for treatment of adult patients. Overall, our findings indicated that both pediatric and adult patients experienced exacerbations of AD and itch symptoms, as measured by health care use rates, in association with short-term exposure to increased air pollution.

We used 3 different metrics to characterize pollution exposure—fire status, smoke plume density, and PM_2.5—and obtained consistent results across these metrics for each outcome. In addition, we constructed lagged versions of each exposure metric to investigate delayed effects possibly due to patients’ slow onset of symptoms, decision to wait to seek care, or difficulty in obtaining an appointment. The highest rate ratios for all outcomes were generally seen at exposure lags 0 to 2, which is consistent with short-term exposure to air pollution being associated with rapid effects on the rates of weekly clinic visits and prescribed medications.

Pruritus is a key symptom of AD; however, most of the adult patients seen for itch during the Camp Fire were not previously diagnosed with AD. This finding suggests that air pollution exposure may affect the skin health of a wider range of patients than those diagnosed with AD. Alternatively, these patients may have subclinical AD not previously diagnosed. In either case, clinicians can broadly counsel patients to use more articles of clothing to cover their skin and emollients to optimize their skin barrier to prevent risk of skin irritation during short-term episodes of poor air quality. In addition, the significant associations between air pollution and number of prescribed systemic medications for adults suggest that patients with more severe skin disease were particularly affected by poor air quality or that exposure to air pollution increased the severity of existing skin disease, requiring the initiation of systemic therapy. Small longitudinal studies of schoolchildren in Korea similarly found increased patient-reported AD symptom severity in association with exposure to particulate matter.^27,28 Systemic treatment for asthma flares triggered by wildfire smoke could additionally treat symptoms of AD, mitigating patients’ need to seek dermatologic care and possibly resulting in this study capturing underestimated associations between wildfires and skin health.

The roles of environmental factors in the development and exacerbation of AD have been of clinical concern because understanding them informs disease management. On a molecular level, air pollution contributes to AD pathogenesis through several mechanisms, including the aryl hydrocarbon receptor pathway, oxidative stress, and inflammation.^51-53 Tapinarof, a new topical aryl hydrocarbon receptor modulating agent, has shown promise for treatment of patients with AD^54,55 and may be useful in the setting of air pollution–related exacerbations of AD. In addition, environmental disasters cause psychosocial stress,⁵⁶ and this increased stress during a wildfire could contribute to AD symptoms.⁵⁷

The association between air pollution exposure and AD on a population level has been unclear because of conflicting epidemiologic evidence.^17-30 This study addresses gaps in previous epidemiology studies^18,21 that have examined PM₁₀ instead of PM_2.5, the latter of which is considered more harmful,⁵⁸ and relied on patient-reported outcome data rather than physician-confirmed diagnoses. Our findings also support studies on long-term PM_2.5 exposure and AD, including indoor air pollution and AD severity in children,⁵⁹ AD prevalence among adults,⁶⁰ and prenatal pollution exposure.²⁵ Regarding health care use, 2 studies^61,62 reported that outpatient visits for eczema and dermatitis in China were associated with air pollution in areas with continuously high background levels. To our knowledge, our study is the first to show a similar association for short-term exposure to increased air pollution in an area with relatively low background levels.

The current increased risk for wildfires in California and other areas around the world is anticipated to result in exposures to high concentrations of air pollution, even in regions where background levels are usually low.^63,64 Both patients with AD and the general public may benefit from understanding the negative skin health effects of air pollution. This study suggests that clinics serving affected communities may experience increased visits for certain skin disease exacerbations that decrease patient quality of life and increase health care expenditures. Future studies could assess associations of short-term air pollution exposure and other inflammatory skin diseases, such as psoriasis. Furthermore, study of environmental skin health disparities is needed, as people of color and from low-income communities experience disproportionately increased exposure to air pollution.^65,66 A better understanding of the skin health implications of wildfire-associated air pollution may help inform clinical management and public health policies on health care resource use.

This study has limitations. The population was restricted to patients living in San Francisco and seeking care at dermatology clinics affiliated with one health care system, which limits the generalizability of our results. However, this study found significant associations for visits at clinics located relatively far from the wildfire; observed associations may be greater at clinics closer to the origin. Air pollution from wildfires is chemically heterogeneous, and the composition of PM_2.5 during the Camp Fire may not match that of PM_2.5 from other wildfires, which may also limit generalizability of this study’s results. This study did not include environmental exposure data on an individual patient level; however, we analyzed 3 exposure metrics and found consistent results for all 3 metrics. We also focused on patients with AD or itch seen at dermatology clinics and did not include patients at primary care clinics.

In this cross-sectional time-series study, short-term exposure to wildfire-associated air pollution was associated with increased rates of clinic visits for AD and itch, a previously unexplored possible consequence of climate change. These results highlight an association between air quality and skin health, which may assist clinicians and public health practitioners to better understand the prevention and treatment of cutaneous diseases and the associated health care use rates.

Accepted for Publication: January 26, 2021.

Published Online: April 21, 2021. doi:10.1001/jamadermatol.2021.0179

Corresponding Author: Maria L. Wei, MD, PhD, Department of Dermatology, University of California, San Francisco, 1701 Divisadero St, San Francisco, CA 94115 (maria.wei@ucsf.edu).

Author Contributions: Mr Fadadu and Dr Wei had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Fadadu, Abuabara, Balmes, Wei.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Fadadu, Grimes, Jewell, Balmes, Wei.

Critical revision of the manuscript for important intellectual content: Fadadu, Jewell, Vargo, Young, Abuabara, Balmes, Wei.

Statistical analysis: Fadadu, Grimes, Jewell, Vargo, Young, Wei.

Obtained funding: Fadadu, Wei.

Administrative, technical, or material support: Fadadu, Balmes, Wei.

Supervision: Jewell, Abuabara, Balmes, Wei.

Conflict of Interest Disclosures: Mr Fadadu reported receiving grants from Alameda Contra-Costa Medical Association, UCSF Summer Explore Fellowship, and grants from UC Berkeley–UCSF Joint Medical Program during the conduct of the study. Dr Abuabara reported receiving personal fees from Target RWE Consultant for design of an observational cohort study on atopic dermatitis and grants from Pfizer during the conduct of the study. No other disclosures were reported.

Funding/Support: This study was supported by the UCSF Summer Explore Fellowship, the Alameda-Contra Costa Medical Association Summer Fellowship, and the UC Berkeley–UCSF Joint Medical Program Thesis Grant (Mr Fadadu).

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Federico Inting, BS (University of California, San Francisco), assisted with data collection; there was no financial compensation.