Reducing Air Pollution Is Always A Good Thing, But How Good Has It Really Been?

I recently wrote about mortality changes following inadvertent increases in air pollution exposure and found increases in short but not long-term rates. My underlying assumption was one of linearity: if increasing emissions hurts, then equivalent decreases must help. But have they?  

“There is consistent evidence that decreased air pollution levels following an intervention resulted in health benefits for the assessed population.” – The EPA

As I have written, statistical associations do not comprise “evidence.” Evidence requires that measured inputs to the system (emissions) be compared with measured outputs (health benefits), and studies rarely performed. However, there are multiple examples and discussions of various air pollution control programs to estimate putative benefits to public health. I chose premature mortality as the indicator of all public health outcomes because of its availability and data reliability.

Timing

Responses to air pollution are sensitive to timing and duration of exposures; ambient air quality standards have followed suit. Twenty-four-hour average standards are intended to protect against peak exposures and acute health effects such as asthma or heart attacks. Annual standards are designed to protect against the development of chronic diseases like cancer or heart disease that might have been exacerbated by exposures over years or decades.

But these assumptions comport poorly in the real world. Peak episodes tend to be distributed over a week or so, and acute responses should be similarly considered. By contrast, chronic diseases develop slowly in response to cumulative processes. Long-term health effects should be considered in terms of cumulative exposures, e.g., smoking has been characterized in terms of pack-years. Searching for current evidence of air pollution control benefits must recognize those realities. Today’s lung cancer began with carcinogens inhaled decades ago, and the development of COPD or congestive heart failure is similarly delayed. Seeking to characterize long-term health benefits in terms of today’s environmental controls is thus a fool’s errand.

Mechanisms of long-term relief from air pollution are multifaceted and include:

• Fuel changes, i.e., switching from coal to gas firing or reducing fuel sulfur content, may have little effect on nitrogen oxide emissions and, thus, on ozone production.
• Source shutdowns, i.e., retiring a power plant, may increase emissions from others to meet demands.
• COVID lockdowns reduced outdoor emissions but increased exposures to indoor pollutants.

Examples of long-term reductions

• The use of coal was banned in Ireland beginning in 1995, and effects on mortality were reported in several studies, of which the most comprehensive reported a modest decrease in respiratory mortality but not total or cardiovascular. 
• Hong Kong restricted the sulfur content of fuels in 1990, after which average ambient SO₂ dropped by 45%. Health effects were studied by time-series analysis and reported, “The results offered little hope for estimating effects on life expectancy…”
• Smoke from woodstoves was reduced in Launceston, Australia, in 2001, after which winter PM_2.5 dropped from 44 µg/m³ from 1994-2000 to 27 µg/m³ from 2001-2007. Annual mortality fell by 11%, but only for males.

Short-term abatement strategies

• Stricter air pollution controls were imposed to reduce athletes’ exposure during the 1996 Olympic Games in Atlanta, GA, the 2008 Olympic Games in Beijing, and the 2010 Asian Games in Guangzhou, China. Improvements in air quality were modest, and no public health benefits were reported.
• Several cities experimented with limiting vehicular traffic in congested areas. This strategy can offer better air quality for a few hours each workday over relatively small parts of metropolitan areas; no health benefits have been reported.
• COVID lockdowns reduced traffic-related air pollution, but the number of individuals exposed outdoors also decreased. Exposures to indoor sources and electricity demand increased, resulting in trade-offs. The reduced mortality benefit was estimated at 0.3%.

Temporary shutdowns

• A regional electrical blackout occurred in the Northeastern US from August 14-15. Mortality and respiratory hospital admissions increased 2-8 fold after accounting for heat effects, but cardiovascular and renal hospitalizations did not. “Increased deaths were advanced more than a few days, and mortality risk remained elevated through the month.”  Although power station emissions dropped to zero, there could have been local effects from private generators.

Summary

In today’s clean environment, it is challenging to improve air quality enough to detect associated long-term changes in public health. For example, 167 New Yorkers died each day in 2020. For a pollution-related mortality rate of 5%, 8 of them would have been at risk from air pollution, but which 8? Data on age, race, gender, smoking, income, and housing would be required for each decedent to analyze long-term benefits. This illustrates fundamental difficulties in the epidemiology of long-term weak effects.  

The overall message is that significant long-term mortality reductions have not been identified after implementing various emission reductions. There are several possible explanations:

• The reduced exposures were below the threshold of effect.
• Reduced exposures did not follow decreasing emissions.

• The most critical pollutants were not affected.
• Disease latency was not considered.
• Some pollutants may have increased while others decreased. 
• Resulting changes in behavior were more important.

I will leave the last word to the current curator of all knowledge, ChatGPT:

Evaluation is an integral part of the regulatory process, providing valuable insight into regulations' effectiveness, efficiency, and impact and informing decision-making to enhance regulatory outcomes over time.