Location
Moultrie, GA
Start Date
17-4-2026 12:00 PM
End Date
17-4-2026 1:00 PM
Description
Introduction
Psoriasis is a chronic inflammatory skin condition affecting approximately 3% of U.S. adults. While environmental exposures such as fine particulate matter (PM₂.₅) have been implicated in systemic inflammation, their association with psoriasis at the population level remains unclear. This study modeled county level psoriasis prevalence across the United States and evaluated its ecological association with long term PM₂.₅ exposure from 2011 to 2014.
Methods
We used NHANES 2011-2014 data to fit a survey-weighted logistic regression model predicting psoriasis diagnosis based on sex, age group, race/ethnicity, education, and poverty level. These demographic coefficients were applied to U.S. Census-derived county populations to estimate local psoriasis prevalence using an ecological small area estimation approach. PM₂.₅ exposure was defined as the average annual number of days in which PM₂.₅ was the dominant pollutant, based on EPA Air Quality Index (AQI) data. Simulated prevalence and pollution data were merged at the county level and analyzed using unadjusted and adjusted linear regressions, generalized additive models (GAM), state-level scatterplots, and exposure quartile comparisons.
Results
Estimated psoriasis prevalence ranged from 1.8 - 4.2%, with higher rates in older,
predominantly non-Hispanic White counties. PM₂.₅ exposure varied from < 10 to >300 days per year. In unadjusted ecological regression, PM₂.₅ exposure was modestly associated with lower modeled psoriasis prevalence (β = -0.017, p = 0.001), likely due to demographic confounding. Adjusted models weakened this association (β = -0.011, p = 0.131). Quartile comparisons showed no clear monotonic trend, and highest prevalence was seen in the lowest exposure quartile. However, GAM analysis revealed a significant non-linear relationship (p < 0.001), suggesting slightly higher prevalence at elevated PM₂.₅ levels. State level scatter plots revealed no consistent trend.
Discussion
Ecologic analysis revealed no clear linear association between PM₂.₅ exposure and modeled psoriasis prevalence. The inverse trend seen in unadjusted models likely reflects underlying demographic distributions rather than a protective pollution effect. The significant non-linear relationship observed in GAM analysis warrants further exploration. Since PM₂.₅ was not included in the predictive model itself, these associations are exploratory. Future work will incorporate direct pollution measures into prevalence modeling using multilevel regression with post-stratification (MRP) and individual level exposure data.
Embargo Period
5-28-2026
Included in
Association between PM₂.₅ air pollution exposure and county level psoriasis prevalence: A nationwide ecological study
Moultrie, GA
Introduction
Psoriasis is a chronic inflammatory skin condition affecting approximately 3% of U.S. adults. While environmental exposures such as fine particulate matter (PM₂.₅) have been implicated in systemic inflammation, their association with psoriasis at the population level remains unclear. This study modeled county level psoriasis prevalence across the United States and evaluated its ecological association with long term PM₂.₅ exposure from 2011 to 2014.
Methods
We used NHANES 2011-2014 data to fit a survey-weighted logistic regression model predicting psoriasis diagnosis based on sex, age group, race/ethnicity, education, and poverty level. These demographic coefficients were applied to U.S. Census-derived county populations to estimate local psoriasis prevalence using an ecological small area estimation approach. PM₂.₅ exposure was defined as the average annual number of days in which PM₂.₅ was the dominant pollutant, based on EPA Air Quality Index (AQI) data. Simulated prevalence and pollution data were merged at the county level and analyzed using unadjusted and adjusted linear regressions, generalized additive models (GAM), state-level scatterplots, and exposure quartile comparisons.
Results
Estimated psoriasis prevalence ranged from 1.8 - 4.2%, with higher rates in older,
predominantly non-Hispanic White counties. PM₂.₅ exposure varied from < 10 to >300 days per year. In unadjusted ecological regression, PM₂.₅ exposure was modestly associated with lower modeled psoriasis prevalence (β = -0.017, p = 0.001), likely due to demographic confounding. Adjusted models weakened this association (β = -0.011, p = 0.131). Quartile comparisons showed no clear monotonic trend, and highest prevalence was seen in the lowest exposure quartile. However, GAM analysis revealed a significant non-linear relationship (p < 0.001), suggesting slightly higher prevalence at elevated PM₂.₅ levels. State level scatter plots revealed no consistent trend.
Discussion
Ecologic analysis revealed no clear linear association between PM₂.₅ exposure and modeled psoriasis prevalence. The inverse trend seen in unadjusted models likely reflects underlying demographic distributions rather than a protective pollution effect. The significant non-linear relationship observed in GAM analysis warrants further exploration. Since PM₂.₅ was not included in the predictive model itself, these associations are exploratory. Future work will incorporate direct pollution measures into prevalence modeling using multilevel regression with post-stratification (MRP) and individual level exposure data.