PM2.5, mortality
and responsible claims

What PM2.5 means

PM2.5 refers to particulate matter with an aerodynamic diameter of 2.5 micrometers or smaller. These particles can come directly from combustion and can also form in the atmosphere from precursor pollutants. Their small size allows them to penetrate deeply into the respiratory system.

A data-center air permit may regulate particulate matter and pollutants that contribute to secondary particle formation. But the presence of a permitted engine or turbine does not by itself tell us the resulting ambient concentration at a home, school or workplace.

EPA strengthened the annual standard in 2024

On February 7, 2024, U.S. EPA strengthened the primary annual PM2.5 standard from 12.0 to 9.0 micrograms per cubic meter. EPA said the revision reflected the available health science and was intended to provide greater protection against effects including heart attacks and premature death.

The annual standard is based on an area’s monitored concentration over time. It is not an emission limit for one generator, and it cannot be compared directly with a permit’s tons-per-year value without atmospheric modeling and monitoring context.

Sources for this sectionU.S. EPA final PM2.5 standard ↗

What the Dominici research contributes

Francesca Dominici and collaborators have used national Medicare data, exposure models and statistical methods to study associations between air pollution and mortality, including at concentrations below earlier national standards. A major research program included about 61 million Medicare enrollees from 2000 through 2012.

The importance of this work is methodological and population-wide: very large datasets can detect changes in risk that would be difficult to observe in a small local sample. Harvard identifies Dominici as a leader in causal-inference methods for air pollution and health, and EPA has publicly recognized the broader contribution of her research to the evidence base.

Sources for this sectionHealth Effects Institute report in PubMed Central ↗Harvard Dominici Lab ↗

Association, causal inference and facility attribution

Large observational studies must address confounding, exposure error, population differences and model choice. Causal-inference methods can strengthen the interpretation, but no national cohort study automatically proves the effect of a particular facility.

Facility-level analysis requires an additional chain: equipment and operating data, emissions estimates, atmospheric transport, incremental concentration, exposed population and an appropriate concentration-response function. Each step should remain inspectable.

• General PM2.5 evidence establishes plausibility and population risk.
• Facility records establish authorized equipment and conditions.
• Dispersion and exposure analysis connect a source to a modeled local effect.

Language that stays inside the evidence

A supported statement is: “Published population studies associate additional PM2.5 exposure with increased health risk, and EPA strengthened the annual standard in 2024.” A separate facility-specific analysis may estimate how much additional exposure a project could create.

An unsupported shortcut is: “This permit will cause a specific number of deaths.” That claim collapses authorization, operation, emissions, exposure and health outcomes into one sentence and removes the uncertainty that the analysis is supposed to illuminate.