Ambient air pollution—including particulate matter and gaseous pollutants—was associated with several common skin diseases, but the certainty of evidence was generally low because of a substantial risk of bias and heterogeneous study designs, according to a recent systematic review.
In the review, investigators synthesized epidemiologic evidence linking exposure to particulate matter with an aerodynamic diameter of 10 µm or smaller and 2.5 µm or smaller as well as the gaseous pollutants nitrogen dioxide, sulfur dioxide, ozone, and carbon monoxide with atopic dermatitis, psoriasis, urticaria, acne, melanoma, nonmelanoma skin cancer, and skin aging. Most of the studies reported positive associations, often strongest within days of exposure; however, the results varied in direction and magnitude across conditions and regions.
Methods and Scope
The investigators searched PubMed and Scopus for relevant observational studies published from January 1, 1990, to April 30, 2025, following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. After the screening of 1,393 eligible studies, 77 studies were selected. The investigators also included ecological, time series or time-stratified studies, cross-sectional, case-crossover, time-stratified–case-crossover, case-control, cross-sectional–case-crossover, longitudinal, and cohort or panel design studies.
The outcomes were outpatient, inpatient, and emergency department visits; longitudinal or cross-sectional surveys of participants; and health assessments of a general population because of dermatologic conditions . The risk of bias was assessed using a World Health Organization tool across six domains, including confounding, selection bias, exposure assessment, outcome measurement, missing data, and selective reporting.
Key Findings by Condition
Atopic dermatitis was the most studied outcome, represented in 43 studies. Most of the studies focused on school-aged children or adolescents (n = 27, 63%). Both short- and long-term exposures were associated with increased risk or symptom severity, with gaseous pollutants frequently showing stronger short-term associations compared with particulate matter. The effects were often immediate and persisted up to 1 week. Subgroup analyses by age, sex, and season produced inconsistent results. Because the high risk of bias was common—particularly for confounding and selection bias—the certainty of evidence for causality was rated very low.
Psoriasis was examined in 12 studies. In those analyzing multiple air pollutants, short-term associations with psoriasis were stronger for gaseous pollutants such as nitrogen dioxide, sulfur dioxide, and carbon monoxide; whereas long-term associations were more often stronger for particulate pollutants. A small number of long-term studies had a low risk of bias across domains, yielding moderate certainty for the findings, although the limited number of studies and high risk of bias in short-term analyses constrained overall inference.
For urticaria, 11 studies—mostly short-term analyses—reported generally positive associations with air pollution. Short-term associations were higher for gaseous pollutants, while both short- and long-term studies showed strong associations with particulate matter, with peak effects occurring several days after exposure. All short-term studies were rated as having a high risk of bias for selection, resulting in very low certainty for short-term evidence, whereas the limited long-term evidence was rated moderate certainty.
Evidence for acne was based on six studies, largely short-term analyses. Nitrogen dioxide showed the most consistent association with outpatient visits, but a pervasive high risk of bias and few long-term data led to very low certainty.
Five long-term studies assessed melanoma, with heterogeneous results across regions and pollutants. Confounding was a major concern, and certainty of evidence was rated very low.
Four long-term cross-sectional studies linked particulate matter and ozone with markers of skin aging, such as hyperpigmentation and wrinkles; however, a moderate to high risk of bias limited confidence, resulting in low certainty.
Limitations and Gaps
Across the dermatologic conditions, a high risk of bias was frequently attributed to incomplete adjustment for confounders, reliance on hospital-based samples, and coarse exposure assessment using area-level monitoring. Most of the studies originated from high-income or upper-middle–income countries, with limited data from low- and middle-income regions. Few studies evaluated disease severity or vulnerable subgroups, and interaction analyses with meteorologic or genetic factors were uncommon.
Conclusion
Despite a growing body of epidemiologic research, the investigators emphasized that evidence linking air pollution to skin diseases remains constrained by substantial methodologic limitations.
“[T]he overall quality of evidence is low. We encourage more longitudinal studies—such as cohort studies or panel studies—to support causality and study change in disease severity over time, ... improved exposure assessments, and adjustment for critical confounding factors. Importantly, more studies are needed from low- to middle-income countries and on susceptible groups who are most vulnerable to climate change," concluded lead study author Nidhi Singh, of the IUF-Leibniz Research Institute for Environmental Medicine, and colleagues.
The study authors reported no conflicts of interest.