Associations Between Short-Term Pollution Exposures and Childhood Autoimmune Diseases

By Andrew S. Zeft, MD, MPH

Autoimmune diseases are an understudied area in the field of pollution-related health effects research. By examining representative childhood autoimmune diseases, Cleveland Clinic Children’s pediatric rheumatologists hope to further understand the effects of acute, short-term pollution exposures on the clinical presentation of autoimmune diseases.

What triggers the clinical onset and exacerbation of autoimmune diseases? The trigger may be driven by certain environmental exposures that precipitate immune activation in genetically susceptible individuals. In fact, twin studies indicate that environmental factors play a significant role in disease development.

Testing short-term ambient fine particulate matter exposures

One hypothesis is that short-term pollution exposures that stimulate pulmonary-mediated, systemic inflammation trigger the clinical presentation of pro-inflammatory autoimmune diseases. This idea is testable, since the Environmental Protection Agency coordinates pollution-monitoring sites in and around urban areas throughout the United States.

Particulate matter concentration measurements from the monitors can be spatially modeled (correlated), and when climatic data from the National Weather Service are factored in, day-today particulate matter can be used as an exposure measurement for environmental epidemiology study. Fine particulate matter (diameter ≤ 2.5 mm, PM2.5) is a measurable component of ambient pollution.

With a three-year grant from the EPA, the research team examined whether short-term ambient fine particulate matter exposures are associated with the clinical presentation of two childhood autoimmune diseases; systemic-onset juvenile idiopathic arthritis (SJIA) and Kawasaki disease (KD).

These diseases typically have an acute presentation, so they are ideal for a time-series (temporally dependent) case-crossover epidemiologic study design. Cases have been identified in different metropolitan regions of North America where ambient particulate matter is a significant public health concern.

The team previously examined the effect of short-term ambient fine particulate matter on the clinical presentation of JIA in a regional cohort with Brigham Young University research collaborator, C. Arden Pope III, PhD. Researchers demonstrated that short-term ambient fine particulate matter may influence the presentation of JIA in young children, particularly those having the systemic-onset subtype.

A look at study findings for SJIA

In this current study, a case-crossover study design was used to analyze associations of short-term PM2.5 exposures with the clinical event of SJIA symptom onset in children residing in five metropolitan regions.

Time trends, seasonality, month and weekday were controlled for by matching. Selected exposure windows (to 14 days) of PM2.5 were examined. Positive, statistically significant associations between PM2.5 concentrations and elevated risk of SJIA were not observed. The most positive associations between short-term PM2.5 exposure and SJIA were in children < 5.5 years (RR 1.75, 95% CI, 0.85-3.62). However, an ad hoc extended pooled analysis including previously reported cases from Utah’s metropolitan areas identified an increased risk of SJIA for children < 5.5 years (RR = 1.76, 95% CI, 1.07-2.89 per 10 μg/m3 increase in 3-day lagged moving average PM2.5). Small, statistically insignificant PM2.5-SJIA associations were observed in this multicity, multiperiod study. However, as found in a prior study, the PM2.5-SJIA association is most suggestive in preschool-aged children. Larger numbers of SJIA cases spatially located in geographic areas that experience a greater day-to-day ambient particulate burden may be required by the analysis to demonstrate effects.

Conclusions further understanding of KD

The study’s second objective was to analyze associations between short-term fine particulate matter exposures (PM2.5) and the event date of fever onset in KD cases in seven metropolitan regions. A case-crossover study design was used. Time trends, seasonality, month and weekday were controlled for by matching. Researchers assembled PM2.5 exposure measurements from urban monitors and imputed PM2.5 to provide day-to-day temporal variability and resolution for a time-series index of exposures. Selected exposure windows (to 14 days) of PM2.5 were examined. The data included 3,009 KD events in which the subjects resided in a study metropolitan area and the event date occurred during years with available PM2.5. The estimated odds ratios (95% CIs) of an event of KD associated with a 10 μg/m3 PM2.5 lagged moving average concentration of lagged exposure period [concurrent, preceding day(s)] revealed no evidence of a consistent, statistically significant, positive association between elevated PM2.5 exposure and elevated risk of KD.

Extended analysis with stratification by city, sex, age, ethnic origin, incomplete or complete clinical manifestations, the presence of coronary aneurysms and IVIG resistance did not provide evidence of a consistent, statistically significant, positive association between elevated exposure to PM2.5 and elevated risk of KD for any of the strata.

The results of this relatively large, multicity study failed to establish a risk of the event of KD with short-term fine particulate exposure. The negative findings add to the growing field of environmental epidemiology research of KD.

Figure 1. Relative risks (and 95% CIs) of an onset event of JIA associated with a 10μg/m3 of PM2.5 for ages 0–5.5 and 6–16.5 and for selected subgroups.

Figure 2. ORs (and 95% CIs) of onset event of KD associated with 10 mg/m3 PM2.5 stratified by city, sex, age, race, and disease attributes with central monitor data plus imputed data from nearby sites. Coronary, presence of coronary aneurysm ≥8 mm; IVIG resistance, received repeat IVIG dose.