In-unit particulate matter (PM2.5) in a public housing complex and the importance of tobacco and cannabis as indoor pollutants

Fine particulate matter (PM 2.5 ) is a ubiquitous air pollutant with known socioeconomic disparities in exposure related to both ambient and indoor sources. Understanding how common indoor behaviors such as smoking, cooking, candle or incense burning, and air freshener use contribute to indoor PM 2.5 exposure in public housing can inform interventions to protect residents’ health. PM 2.5 data were collected via PurpleAir monitors for one week within 138 units of a public housing development in New England, United States and corrected via an additive relative humidity model. Study participants completed a pre-monitoring survey and a post-monitoring survey of in-unit PM 2.5 -related behaviors. Participants’ weekly average in-unit PM 2.5 levels were variable, ranging from 3.6 µg/m 3 to 142.0 µg/m 3 , with a mean of 19.7 µg/m 3 . Two-thirds (65.9%) of households’ average levels exceeded the EPA annual outdoor standard of 9 µg/m 3 , a heuristic reference given the lack of national indoor guidelines. A multivariable model predicting unit-level average PM 2.5 based on resident behaviors found significant associations between elevated PM 2.5 and past-week tobacco smoking (121% higher PM 2.5 ), weekly or more cannabis use (56% higher), and living in an apartment building rather than a townhome (26% higher). Cannabis use, which is studied here for the first time in an uncontrolled residential setting, may represent an underappreciated risk factor for exposure. Interventions directed at mitigating indoor smoking of both tobacco and cannabis may be effective at reducing average in-unit PM 2.5 exposure in multifamily housing, as compared to combatting other known indoor sources of PM 2.5 . • Weekly PM 2.5 concentration varied between 138 public housing units with a mean of 19.7 µg/m³ • Two thirds of households’ PM 2.5 exposure exceeded the EPA annual outdoor standard of 9 µg/m³ • Tobacco smoking significantly associated with a 121% increase in average PM 2.5 after adjustment • Cannabis use significantly associated with a 59% increase in average PM 2.5 after adjustment