Long-Term Exposure to Ozone and Cardiopulmonary Mortality: Epidemiology Strikes Again

For many years ozone was the most studied air pollutant, primarily because it caused acute, mostly respiratory, symptoms during “smog” episodes in Southern California that generated considerable public concern. In 1966, California enacted regulations requiring smog controls on cars to limit emissions that promoted the formation of ozone, before the Clean Air Act of 1970 led to the establishment of the U.S. Environmental Protection Agency (EPA). The Clean Air Act mandated the EPA to set National Ambient Air Quality Standards (NAAQS) to control the levels of pollutants that are considered harmful to public health based on a review of the scientific literature. Ozone was one of the initial so-called “criteria” pollutants for which NAAQS were promulgated, but these standards were based solely on adverse health effects related to short-term exposures. Over time, as accumulating evidence indicated that acute respiratory effects (airway inflammation and lung function decrements) occurred in people at lower and lower exposure concentrations, the NAAQS for ozone has been periodically modified, but the standard has always been a short-term standard and is currently 70 ppb over an 8-hour averaging time.

The evidence for adverse health effects as a consequence of long-term exposure to ozone is much less robust than the evidence for effects of short-term exposures. The criteria pollutant for which there is strong evidence that long-term exposure is associated with a risk of premature mortality is particulate matter ≤10 μm in aerodynamic diameter (PM_2.5) (1). That said, there is an increasing body of studies that report an increased risk of cardiopulmonary mortality in association with long-term exposure to ozone. Perhaps the first important study to report a mortality effect with long-term exposure was conducted by Jerrett and colleagues using data from the large (n = 448,850) American Cancer Society Cancer Prevention Study II (CPS-II) (2). The authors found a significant increase in the risk of death from respiratory causes associated with increased annual ozone concentrations. They were not able, however, to detect an effect of ozone on the risk of death from cardiovascular causes when they adjusted their analysis for concentrations of PM_2.5. Subsequently, the analysis of this cohort was updated with a longer follow-up period and improved exposure estimates (3). The updated analysis confirmed the association of long-term ozone exposure with an increased risk of death owing to respiratory causes, but it also found an association with cardiovascular mortality that remained significant in a two-pollutant model that included PM_2.5.

In this issue of the Journal, Lim and colleagues (pp. 1022–1031) report the results of another large study of long-term ozone and cause-specific mortality (4). The authors prospectively studied 548,780 participants in the NIH-AARP Diet and Health Study over 17 years of follow-up and found that the long-term annual average concentration of ozone was significantly associated with deaths due to cardiovascular disease (per 10 ppb; hazard ratio [HR], 1.03; 95% confidence interval [CI], 1.01–1.06), ischemic heart disease (HR, 1.06; 95% CI, 1.02–1.09), respiratory disease (HR, 1.04; 95% CI, 1.00–1.09), and chronic obstructive pulmonary disease (HR, 1.09; 95% CI, 1.03–1.15) in single-pollutant models. In models adjusting for coexposure to PM_2.5 and NO₂, these associations remained significant. The authors also found a significantly elevated respiratory disease mortality risk associated with long-term ozone exposure for NIH-AARP participants living in locations with warmer climates.

The interaction with higher temperature is not really a surprise, as ozone concentrations increase as temperatures rise (5). Ozone is generated from motor vehicle emissions and other sources by photochemistry, so urban areas with warm, sunny afternoon weather after the morning rush hour (e.g., Denver, Houston, Las Vegas, Los Angeles, Philadelphia, Phoenix, Salt Lake City, and Washington-Baltimore) typically have the highest ozone concentrations (6). High temperature has an independent effect on mortality (7), but when these exposures occur together, the risk greatly increases. At least 35,000 deaths were attributed to a severe heat wave in Europe in August 2003 that was complicated by high ozone levels (8). With increasing global warming, coexposures to high heat and ozone will be occurring with greater frequency and longer duration (9).

An increased risk of mortality due to cardiovascular causes has now been reported from two large and well-conducted prospective cohort studies that adjusted for multiple potentially confounding covariates (3, 4). Although residual confounding always remains a possibility in observational epidemiological studies, the quantitative levels of risk were remarkably similar in the two studies, with an HR of 1.03 in the NIH-AARP cohort, as noted above, and an HR of 1.03 (95% CI, 1.01–1.05) in the CPS-II cohort. It is difficult to dismiss these reported associations when considering regulatory policies to protect public health. Because a causal relationship between long-term PM_2.5 exposure and mortality has been recognized by the EPA, the NAAQS for PM_2.5 includes an annual standard (10). Lim and colleagues suggest that based on the findings of the NIH-AARP and CPS-II cohort studies, policies to reduce long-term exposures to ozone, such as the development of an annual ozone NAAQS in addition to the current 8-hour standard, should be considered.

The Clean Air Act specifically states that the EPA must address “air pollution which may reasonably be anticipated to endanger public health or welfare” when setting a NAAQS (11). The act also requires a review of the body of scientific literature by an independent panel when a NAAQS for a criteria pollutant is being considered. This panel, named the Clean Air Scientific Advisory Committee, is currently chaired by a statistician who does not support the use of observational epidemiology for environmental regulatory purposes, arguing that such studies can never prove causation because they do not adequately adjust for potential confounders such as weather and demographic and socioeconomic variables (12). Both Lim and colleagues’ study published in this issue and Turner and colleagues’ study published in 2016 (3) attempted to adjust for potential confounding covariates.

More than a third of the nation’s population lives in areas that exceed the current ozone standard (6). As healthcare providers and scientists, the members of the American Thoracic Society should urge the EPA to carefully consider the increasingly strong epidemiologic evidence that long-term exposure increases the risk of cardiopulmonary mortality when the agency next evaluates the current NAAQS for ozone.