Breathing Under Pressure? Air Pollution and Mortality in Pulmonary Hypertension

Since the passage of the Clean Air Act in 1970, the United States has seen a steady and dramatic improvement in air quality, with an almost 80% reduction in the most common pollutants in the first 50 years after the act was signed into law (1). However, in recent years, extreme heat and wildfires have threatened to reverse these gains; for the first time in decades, Americans are breathing increasingly more polluted air (2).

Under the Clean Air Act, the U.S. Environmental Protection Agency (EPA) regulates several key air pollutants, including particulate matter, ozone, and nitrogen dioxide, which have been shown to be especially harmful to respiratory health. Particulate matter with an aerodynamic diameter < 2.5 μm, or PM_2.5, is a mixture of particles and dust that is fine enough to penetrate the lungs, enter the bloodstream, and cause respiratory and cardiovascular damage. Studies have consistently shown that long-term exposure to PM_2.5 is associated with poorer pulmonary and vascular health, including reduced lung function, increased respiratory hospitalizations, cardiovascular events, and mortality (3–5). For the latter two, cardiovascular effects and mortality, the EPA designates the relationship between PM_2.5 and these outcomes as causal, their strongest determination for the exposure–health relationship. Similarly, ground-level ozone (O₃) and nitrogen dioxide (NO₂) have been associated with the development of pulmonary and cardiovascular disease and a greater risk of death as well (4, 6–12).

Despite the extensive body of evidence linking air pollution exposure to poorer pulmonary and systemic vascular health, much less is understood about its effect on diseases of the pulmonary vessels. Given the intimate relationship between the alveoli and the pulmonary vascular bed, it is plausible that individuals with pulmonary hypertension might be particularly vulnerable to the harmful effects of air pollution. However, prior studies on this topic have demonstrated conflicting results. For example, in a relatively small cohort of patients with pulmonary arterial hypertension (PAH) in the United Kingdom, greater PM_2.5 exposure was associated with poorer transplant-free survival, with a large magnitude of effect (13); in contrast, in a similarly sized cohort of individuals in Belgium, no relationship was seen (14).

In this issue of AnnalsATS (pp. 1351–1360), Huang and colleagues tackle the question of how air pollution might affect individuals living with PAH and chronic thromboembolic pulmonary hypertension, two understudied but devastating diseases of the pulmonary circulation (15). The authors leverage data from the Pulmonary Hypertension Association Registry, a large, national, multicenter, prospective, and ongoing registry of people with newly diagnosed PAH or chronic thromboembolic pulmonary hypertension, to examine whether long-term exposure to PM_2.5, O₃, and NO₂ is associated with poorer transplant-free survival. In limited models, the authors found that greater long-term PM_2.5 exposure was associated with a greater rate of death or lung transplantation after adjustment for demographic variables, pulmonary hypertension characteristics, and socioeconomic status, including a social vulnerability index score. However, this association was entirely attenuated in their fully adjusted models, which accounted for local and regional spatial confounders. A similar pattern was seen with lower O₃ exposure, whereas no associations were seen with NO₂. In subsequent stratified analyses, the authors found possible evidence that the relationship between PM_2.5 and pulmonary hypertension outcomes may differ somewhat based on region, albeit limited by smaller samples and multiple testing. Overall, these findings suggest that local and regional factors may play a key role in the relationship between air pollution and pulmonary hypertension–related outcomes—and perhaps that the regional variation in pulmonary hypertension outcomes may be mediated to some degree by differences in air pollution exposure.

There are important limitations to this study, as the authors point out. First, air pollution exposure was assigned based on the annual average concentrations from a single year, 2015, for all participants, despite recruiting participants between 2015 and 2024. This means, for example, that the PM_2.5 concentrations from 2015 would have been used as the exposure for a participant who would not be enrolled in the study until years later. This approach also cannot capture shorter-term fluctuations in air pollution, particularly wildfire events, as acknowledged by the authors. This is unfortunately increasingly relevant, as over the past decade there have been tremendous short-term spikes in PM_2.5 concentrations across the United States due to wildfire smoke, including from fires that originated internationally. These spikes almost certainly contribute to poorer outcomes in PAH (16) but will not be well-captured by an annual average. Relatedly, the authors note that 13% of participants moved during the study period (a particular concern for longer-term exposure studies), which calls into question the relevance of a single annual exposure. These factors may have introduced exposure misclassification and contributed to the overall null results of the study. The authors defend this choice based on similar methods in previously published studies (13) and the unavailability of more up-to-date exposure estimates. More importantly, the authors provide evidence that PM_2.5 concentrations (for one) did not vary substantially from year to year between 2015 to 2024, as shown in their supplemental materials, suggesting that the 2015 exposure may have been a reasonable proxy for the entire study period. It appears to us that annual average ozone and NO₂ concentrations have also been relatively stable since 2015 (17, 18). As spatially and temporally resolved exposure datasets are updated for more recent years, it will be important to confirm the findings by Huang and colleagues and others by replicating these analyses with more precise, time-varying exposure assignment. Such analyses will likely yield further insight into the regional differences suggested by this current paper.

One of the main strengths of the study is the large sample size (nearly 10 times larger than the U.K. or Belgian cohorts), with careful pulmonary hypertension phenotyping at centers of excellence across the United States. Despite a relatively short duration of follow-up, with a median of only 2 years, the event rate was notable, with almost a fifth of the cohort experiencing either death or lung transplant—highlighting the substantial morbidity suffered by patients with pulmonary hypertension and the urgent need to identify potentially modifiable factors for poor outcome, such as air pollution.

Although the overall results of the study are null, the suggestion that PM_2.5 may increase mortality in patients with pulmonary hypertension in certain regions and/or contexts merits further attention and should not dampen efforts to reverse the recent trend of worsening air quality in the United States. This is particularly true given the overall relatively low concentrations of ambient pollutants observed in this study. This is in line with prior work that has shown deleterious health effects at very low concentrations of PM_2.5 (10, 19), with no evidence that there is any clear threshold below which PM_2.5 exposure can be considered “safe.” For this reason, although the current EPA standard for annual PM_2.5 exposure was lowered last year from 12 μg/m³ to 9 μg/m³ (a change that is currently being reconsidered by the EPA under the current administration), the American Thoracic Society has long advocated for an annual PM_2.5 standard of <8 μg/m³.

Under the system of medieval humoralism, the essential humor of Blood (which perfuses the lungs) was influenced by the classical element of Air. Although our current understanding of medicine has mostly moved beyond humoral theory, we commend the authors of the current study for their efforts to advance the limited body of evidence related to the effects of air pollution exposure and outcomes in pulmonary hypertension, which remains incompletely understood. We look forward to their plans to follow up on these results in the future as additional data and improved methodology become available for the Pulmonary Hypertension Association Registry cohort.