Part I of our series on ozone described how 2015 was a bad year for Houston ozone. Part II reviewed recent research from leading Houston scientists that explains why more ozone pollution is harmful to our health. Part III explains how faulty logic and erroneous assumptions had led to costly lawsuits and poor public health policy across the state. Part IV will identify some solutions to Houston’s ozone problem and suggest measures to protect the health of Houston area residents.
There has been quite a bit of activity related to the proposed U.S. ozone regulations in the past year. As part of a four part series on ozone in 2015, we’d like to take the time to rebuke some of the scientifically-flawed testimony provided by state environmental officials, including Dr. Michael Honeycutt, toxicologist for the Texas Commission on Environmental Quality (TCEQ), the state environmental agency. We feel that the agency has presented health information in a way that is misleading and contradicts the robust opinion of the medical health community on the issue.
First, a little context is important. We at Environmental Defense Fund (EDF) have participated in the public process involving the ozone standard and provided testimony to Congress on the health effects of ozone exposure. TCEQ has challenged the health-based standards in an aggressive way, and their efforts have been fodder for expensive and frivolous lawsuits filed by the state.
Prior to last year’s rule making, the ozone standard was 75 parts per billion (ppb), a level set in 2008 by the Bush administration, even though the Environmental Protection Agency’s (EPA) independent scientific advisory panel had recommended the standard be set at 60-70ppb, based on the scientific evidence. About a year ago, the Obama administration proposed an updated ozone standard in the range of 65-70ppb. After substantial discussion and public comment, the ozone standard was strengthened to 70ppb this fall.
Before EPA proposes new regulations, the agency performs an exhaustive, scientifically-rigorous review of the scientific literature. The Integrated Scientific Assessment (ISA), “2013 Final Report: Integrated Science Assessment of Ozone and Related Photochemical Oxidants”, was published in 2013. Its main findings related to respiratory effects of ozone and susceptible populations are that exposure to ozone likely causes respiratory health effects. As stated in the assessment:
- “Together, the evidence integrated across controlled human exposure, epidemiologic, and toxicological studies and across the spectrum of respiratory health endpoints continues to demonstrate that there is a causal relationship between short-term O3 exposure and respiratory health effects.”
- “the more recent epidemiologic evidence, combined with toxicological studies in rodents and nonhuman primates, provides biologically plausible evidence that there is likely to be a causal relationship between long term exposure to O3 and respiratory health effects.”
- “The populations and lifestages identified. that have “adequate” evidence for increased O3-related health effects are individuals with certain genotypes, individuals with asthma, younger and older age groups, individuals with reduced intake of certain nutrients, and outdoor workers, based on consistency in findings across studies and evidence of coherence in results from different scientific disciplines.”
The bottom line? Exposure to ozone is bad for your health.
In the most recent round of congressional hearings in October 2015, EDF provided testimony and evidence supporting the congressionally mandated and rigorous public process that EPA uses to review health-based standards for pollutants such as ozone.
In assessing Dr. Honeycutt’s testimony, it appears that his primary strategy was to identify one study among many, or one health-related effect among many, which were not consistent with the preponderance of scientific evidence. Dr. Honeycutt’s conclusion was that these exceptions to the overall findings of the scientific literature indicate that there is great uncertainty about the health effects of ozone.
Some variability across scientific studies is normal. And attempting to draw attention to cherry-picked outliers belies the fact there is near-unanimous agreement about the dangers of ozone pollution in the studies on the subject.
However, even if we consider the few exceptions in the scientific literature, which Dr. Honeycutt cited as evidence that ozone does not cause the health effects observed in the bulk of the scientific literature, these specific arguments presented in his testimony are also seriously flawed:
Honeycutt states that with regard to changes in lung function and asthma exacerbations, that(1) 8 out of 9 studies investigating lung function changes caused by ozone showed no difference between asthmatics and healthy individuals.
Here Dr. Honeycutt selects one outcome among many respiratory health outcomes on which to base his argument and fails to consider the concepts included in EPA’s decision making. The idea of susceptible populations is that, while we know that there are effects of pollutants on overall, general populations, there may be subgroups of general populations that are more at risk of the harmful effects of pollution. For ozone, those with asthma have been identified as one such group: “Overall, there is adequate evidence for asthmatics to be an at risk population based on the substantial, consistent evidence among controlled human exposure studies and coherence from epidemiologic and toxicological studies.” (from the ISA).
This increased susceptibility of those with asthma to air pollution is not necessarily based on differential effects of the pollutant on any single biologic parameter or health effect (of many) in the susceptible population. However, even if all of the evidence showed that those with asthma had similar changes in lung function and inflammation with ozone exposure to those without asthma (which, to be clear, the evidence does not show, as discussed below), it’s important to realize that people with asthma generally have more inflammation and worse lung function than those without asthma. Any decrease in their lung function or increase inflammation is going to make them much more likely to have respiratory problems than someone without asthma.
Interestingly, the studies that Dr. Honeycutt cites do not even support his argument. The major study cited showed that both those with asthma and the general population have worsened lung function with ozone exposure, but also showed that inflammation following ozone exposure was worse among those with asthma. Thus, the evidence does not even support this critique of asthma as a susceptible category.
Supporting this concept, in 2015 alone, there were 5 studies published linking worsening of asthma to ozone exposure in children and 1 systematic review and meta-analysis that came to the same conclusions (See references below).
(2) As we stated in our comments to EPA, the dose a person would be expected to receive at 75ppb is almost no different than at 70ppb, or even 65ppb – see Figure 1. Consistent with this finding, EPA does not predict that a decrease in the ozone standard will cause a statistically significant decrease in asthma exacerbations (attacks) – see Figure 2.
This argument, and the graphs that accompany it demonstrate a faulty understanding of what statistically significant means, and when it is a useful concept. To understand the flawed logic, we need to delve a little into some statistical and epidemiologic concepts. Determining whether something is statistically significantly different from something else is something we do when we’re trying to make inferences about something we can’t observe. In the cases where it’s a useful concept, we’re taking a sample of a population, and trying to use that sample to infer something about the population. So, we might take a sample of one population and a sample of another and try to figure out whether the average of one of their traits is statistically significantly different between the two populations. Statistical inference is an important tool when we only get to see a little of the world. One of the main things that determines whether a finding is statistically significant depends on how big that sample is, and how well we can measure it. For example, let’s say I decide to eat three slices of pizza for dinner instead of two. Is my three pieces of pizza statistically significantly different from two pieces of pizza? No, it’s not – you’d have to have a much, much bigger sample of my eating habits if you wanted to say that I eat more than 2 slices of pizza when I eat pizza. But does that mean that the extra piece of pizza makes no difference to my caloric intake? Or that we can’t tell that three pieces is bigger than two? Of course not.
This logical error is seen very clearly in figure 1 (below), where Dr. Honeycutt claims that the dose of ozone that a child would inhale would not differ between a 75 ppb exposure and a 70 ppb or 65 ppb exposure. It is indisputable fact that 70 is less than 75, and a decrease in ozone exposure between exposure to 70ppb ozone and 75 ppb ozone is 7% (70/75), and between 65 and 75 is 14%. We don’t need to do a statistical test to determine whether these ozone concentrations are different from one another.
In figure 2, a similar process is at work. Since there is no reference included for this figure in his testimony, it is hard to be sure how it was created, but it is very likely that the same flawed thinking that led to wide error bars in figure 1 are also at play here. Indeed, if you look closely at this table, you can see that the projected avoided morbidity here is about 100,000 or 200,000 exacerbations for a 70 ppb standard and more than 1 million for a 65 ppb standard. Thus the “almost no different” rate of asthma exacerbations is up to 1 million asthma exacerbations. Think about that: 1 million exacerbations!
(3) The basis for setting the standard at 70 ppb was to make it lower than the lowest exposure concentration where adverse effects were observed in human controlled exposure studies, which was 72 ppb . However, in order to observe any effects at this low ozone concentration, the study authors had to expose the human subjects to ozone while they were exercising at moderate to heavy exertion for 50 minutes out of every hour for 6.6 hours [25-28]. This is an unrealistic exposure scenario for the general public, much less for sensitive groups. Therefore, it would take higher concentrations to have the same effect noted in the study.
Here again Dr. Honeycutt’s logic is flawed. The ozone exposure model used in these controlled human exposure studies is “intended to simulate work performed during a day of heavy to severe manual labor in outdoor laborers.” This model is appropriate as regulation must be set to protect not just people who have more sedentary jobs in an indoor environment, but also those who are vulnerable because they have higher exposure. For example, children are more susceptible to respiratory effects of pollution exposure, at least in part because they inhale more air per body weight and tend to spend more time outdoors than adults. Along the same lines, people performing manual labor outdoors have greater exposure than those who do not. In his argument, Dr. Honeycutt suggests that the standard should only protect people whose health and sociodemographic characteristics may limit their susceptibility and exposure, respectively. This notion is counter to the Clean Air Act.
Dr. Honeycutt also ignores the observation that there was a clear subset of the study population (in: Schelegle et al and Adams, W.C., Comparison of chamber and face-mask 6.6-hour exposures to ozone on pulmonary function and symptoms responses. Inhal Toxicol, 2002. 14(7): p. 745-64) that had decrements in one measure of lung function (FEV1) of more than 10% with exposure to 60ppb of ozone, suggesting that there are populations who are susceptible to exposure to ozone at concentrations as low as 60ppb.
We are always open to debate on the science of air pollutants. But it seems clear that the TCEQ and state leaders fail to see the forest through the trees, especially when trying to assess the broader scientific literature. We encourage the agency to devote more time and resources toward cleaning up the air, rather than serving as obstructionists to the progress we need.
2015 References for Ozone and Respiratory Effects
Alhanti BA, Chang HH, Winquist A, Mulholland JA, Darrow LA, Sarnat SE. Ambient air pollution and emergency department visits for asthma: a multi-city assessment of effect modification by age. J Expo Sci Environ Epidemiol. 2015 Sep 9. doi: 10.1038/jes.2015.57. [Epub ahead of print] PubMed PMID: 26350981.
Gleason JA, Fagliano JA. Associations of daily pediatric asthma emergency department visits with air pollution in Newark, NJ: utilizing time-series and case-crossover study designs. J Asthma. 2015 Oct;52(8):815-22. doi: 10.3109/02770903.2015.1033726. Epub 2015 Jul 27. PubMed PMID: 26211997.
Ierodiakonou D, Zanobetti A, Coull BA, Melly S, Postma DS, Boezen HM, Vonk JM, Williams PV, Shapiro GG, McKone EF, Hallstrand TS, Koenig JQ, Schildcrout JS, Lumley T, Fuhlbrigge AN, Koutrakis P, Schwartz J, Weiss ST, Gold DR; Childhood Asthma Management Program Research Group. Ambient air pollution, lung function, and airway responsiveness in asthmatic children. J Allergy Clin Immunol. 2015 Jun 29. pii: S0091-6749(15)00769-1. doi: 10.1016/j.jaci.2015.05.028. [Epub ahead of print] PubMed PMID: 26187234.
Sheffield PE, Zhou J, Shmool JL, Clougherty JE. Ambient ozone exposure and children's acute asthma in New York City: a case-crossover analysis. Environ Health. 2015 Mar 18;14:25. doi: 10.1186/s12940-015-0010-2. PubMed PMID: 25889205; PubMed Central PMCID: PMC4373115.
Yamazaki S, Shima M, Yoda Y, Oka K, Kurosaka F, Shimizu S, Takahashi H, Nakatani Y, Nishikawa J, Fujiwara K, Mizumori Y, Mogami A, Yamada T, Yamamoto N. Exposure to air pollution and meteorological factors associated with children's primary care visits at night due to asthma attack: case-crossover design for 3-year pooled patients. BMJ Open. 2015 May 3;5(4):e005736. doi: 10.1136/bmjopen-2014-005736. PubMed PMID: 25941174; PubMed Central PMCID: PMC4420953.
Zheng XY, Ding H, Jiang LN, Chen SW, Zheng JP, Qiu M, Zhou YX, Chen Q, Guan WJ. Association between Air Pollutants and Asthma Emergency Room Visits and Hospital Admissions in Time Series Studies: A Systematic Review and Meta Analysis. PLoS One. 2015 Sep 18;10(9):e0138146. doi: 10.1371/journal.pone.0138146. eCollection 2015. PubMed PMID: 26382947; PubMed Central PMCID: PMC4575194.
Image Credit: Rick Kimpel/flickr.com