Moderator: Adrián Fernández, Instituto Nacional de Ecología
Speakers: Gustavo Olaiz, Secretaría de Salud and John Spengler, Harvard

Dr. Adrián Fernández opened the session by remarking that the 1990 PICCA plan had more than 30 actions, only one of which had a health componentan epidemiological surveillance program that in the end was not implemented. In fact, at the beginning of the present presidential administration, according to Dr. Olaiz, the official position of the government's Health Ministry's was that air pollution in Mexico City had no effects at all on health.

Presentation by Dr. Gustavo Olaiz

Dr. Gustavo Olaiz began his talk by noting that the air pollution-health connection was long in coming. He recalled the time not long ago when researchers studied the benefits of ozone, drawing their support from the fact that bacteria die from ozone.

Surveillance Studies

According to Dr. Olaiz, the principal causes of Mexico City's air pollution are:

• the roughly 4 million vehicles, of which many are old;
• emissions from industry, particularly in the northern section;
• overpopulation (which he considers to be the most important factor); and
• deforestation.

As Under-Secretary of Environmental Health, Dr. Olaiz stated that the basic objectives of his Directorate were to (1) promote public participation to reduce health effects, (2) create coordination mechanisms for dealing with environmental emergencies, and (3) avoid exposures in the population.

To demonstrate to other authorities that this is a serious problem, health researchers began an epidemiological surveillance program to monitor daily acute health effects associated with air pollution. Currently, they are also conducting controlled perspective studies: for example, monitoring school-age children to as sess chronic effects. This research is being conducted in coordination with Johns Hopkins and other institutions.

Describing the surveillance system, Dr. Olaiz said they are able to capture information from all health clinics and hospitals thanks to Mexico's comprehensive national health system. They are also conducting ongoing surveillance of households, schools, etc. To establish a geographic program, they established surveillance systems in selected areas in the vicinity of environmental monitors (see Figure 3-1).¹ They survey hospitals and collect information on everyone inside a two-kilometer area surrounding the monitor, using a census of homes to conduct actual visits to the houses in order to evaluate respiratory symptoms. According to Dr. Olaiz this is the biggest surveillance program in the world, collecting information from 600,000 people in 180,000 households.

Figure 3-1. Environmental health surveillance areas

Ozone

Many of ozone's more serious health effects come from its impact on lung membranes. Ozone also affects lung secretions, reduces tracheobroncheal diameter, and is thought to impair immune responses. Exposure results in a significant reduction in breathing capacity, which can cause severe problems for those with asthma or severe or chronic respiratory disease. Also at risk are people who work outdoors.

Describing some of their findings from their health studies, Dr. Olaiz reported that at base ozone levels, about 8% of the population experience symptoms. At 240 IMECAS, this increases to about 15%. At 240-260 IMECAS, 25% of the population experience symptoms. As ozone levels increase, symptoms increase from eye irritation to tear secretion, headache, sore throats, and respiratory difficulty.

Based on the cost of direct health problems alone, Dr. Olaiz estimated that $5 million is spent on direct health costs for each contingency episode above 240 IMECAS.

Figure 3-2. Symptoms at different ozone levels

Particulates

Dr. Olaiz went on to discuss data for the health impacts of PM₁₀ (Figure 3-3). These studies are fairly new, he said. For PM₁₀ they found similar results to ozone, but with a lag of 7-10 days between exposure and health effects (3-5 days for children), making particulates more difficult to study. Though the media are focused on ozone, particulates are also a significant problem. The worst situation, he pointed out, is when ozone and PM₁₀ levels are elevated at the same time (Figure 3-4). He said that combined high levels of ozone and particles is new.

Figure 3-3. Symptoms at different ozone levels

Figure 3-4: Symptoms for combined levels of ozone and particulates

Initial studies show that a 10 mg/m³ increase in the ambient concentration of particulates results in a 0.5% increase in mortality. The problem with the models, however, is that they are linear. Dr. Olaiz said there is a need to work on better understanding the dose-response relationship.

Finally, he said that PM₁₀ and ozone aren't all we need to be concerned about. There are unpublished findings that when ozone levels start to rise, symptoms start earlyeven by a full day. There are many more poorly understood pollutants in the air. Health researchers need to understand what is happening, beginning with the toxicity of different pollutants.

Dr. Olaiz concluded his presentation by asking what's needed by this group. Health is a good motivator for getting something to happen, he said, so the information from these health studies is important. There is a need to assess exposure better; ambient monitoring is useful to understand what a typical day looks like, he said, but personal exposures are different. They are currently doing various long-term studies centered on personal exposure monitoring, including a three-year study of healthy adults, a five-year study of asthmatic children, and another study looking at school-age children, which has been in progress for three and a half years.

Presentation by Professor John Spengler

Professor John Spengler noted that Mexicans have long-term monitoring data, and a sense of the trends and spatial distributions of pollution, which is more than the case for many cities. Ozone in Mexico City is the highest he has seen in the world, he said, while particulates, though also a problem, are not as severe as in Beijing or Delhi. He pointed out that Mexico experiences forest fires and warm dry spells where wind speeds drop by a factor of two for a whole year. These environmental and meteorological events need to be incorporated into de-trending the data.

With respect to monitoring, he asked whether the right compounds are being measured. Instrumentation also plays a crucial role in the validity of the data. He asked if the right instruments were being used, and used correctly. In the case of PM, for example, we don't even know what component of particles we should be measuring. All measurements are approximations, he pointed out, and researchers should look for consistencies in the dose-response relationship. He referred to the Los Alamos/IMP/DRI study examining particle composition and source factors, and asked whether this is enough. Perhaps there is a need to institutionalize these studies and expand what's monitored, he suggested.

Discussing population exposures, Prof. Spengler explained why exposure data are needed even if there is so much ambient monitoring. There is a need to be comprehensive with respect to risk assessment and exposure analysis, as well as a strong need to disentangle signals on the health effects side. There may be conflicting sources masking which substances are actually causing the effects. For instance, indoor (cooking) and outdoor (transportation) combustion both put out NO₂, but exposure scenarios aren't well understood yet. There is a need to measure exposures in order to get at status and trends in subpopulations. About ten years ago, Adrián Fernández did a CO exposure study

on various transportation routes and compared his results to ambient monitoring data. Paulina Serrano, a Harvard Ph.D. student, is conducting a similar study now using personal monitors to track exposure to VOCs and CO. They have found that during the commute period, CO exposures are much higher than would be anticipated by the ambient monitoring data alone. Therefore, the epidemiological impacts might not be directly correlated to ambient concentrations.

These studies are just a beginning, as they only look at respiratory symptoms and cardio-pulmonary impacts. An expanded analysis should consider exposures in a more coordinated way. Shifts in behavior patterns (e.g. by commuters) have an impact on exposure to different sources as well as the production of emissions. This needs to be understood from an economic and sociopolitical point of view, not just a scientific one. For example, a study might focus on combi (taxi) driver exposures. Such a study could also be helpful in influencing their behavior and generating political will. Another study might consider exposure on schoolteachers. In general, Prof. Spengler said, it is important to bring in other sectors, particularly agriculture and construction. The forest fires illustrate the impact agricultural policy has had when combined with dry weather. In the construction sector, design and building codes and the use of air conditioning are also relevant.

Initial studies indicate a 6% change in mortality rates given a 10 mg/m³ change in PM_2.5. This corresponds to roughly 8 respiratory deaths per day in the neighborhood being studied. Research is beginning to look at mechanisms of cardiovascular effects. Changes in blood viscosity and heart arrhythmias are being linked to particulates. Asthma, an immunological disease, is also an important indicator. Recent evidence elsewhere shows that ozone/NO exposure worsens the effects of exposure to allergens.

For an epidemiological study in Mexico City, Prof. Spengler proposed a longitudinal study of freshman coming to UNAM from other parts of the country. He said this group could be used to study housing exposures, track illness, development of allergies, etc. in a group newly exposed to Mexico City pollution.

Referring to the cost figures presented by Dr. Olaiz, Prof. Spengler said that $5 million per episode is a huge amount, but if costs were integrated under the rest of the curve (capturing health effects from exposures at lower levels as well), the amount would be enormous. Generating this information could add a great deal to the political discussion.

Finally, Prof. Spengler described the experience of the Netherlands. In the mid-1980s the Dutch government commissioned a study of the state of the environment 25 years into the future given then-current trends. The image presented served as shock therapy to the nation, he explained. Everyone became informed with the help of a strong image of what they didn't want. This strengthened political forces sufficiently to generate powerful legislation across all sectors. Scientific integrity and enforceability were essential. Professor Spengler said that the Dutch model could possibly serve as an example of what could be done to inform policy makers and the public in Mexico.

Discussion Summary

Four years ago, public authorities were reviewing what level should be considered an "episode," with a strong interest in setting the standard so as to limit the number of episodes. Different institutions were involved, but the complicated discussions only came to a close when health officials weighed in with data showing the effects and costs of air pollution on human health. Though the influence of health officials on air pollution policies is real, it was noted 240 IMECAS was chosen as the contingency level, while the goal of health officials was 100 IMECAS.

Health costs incurred at contingency levels of pollution focus on short-term solutions; the context of this project, however, is to consider the long-term. For this, more comprehensive cost figuresshowing costs from health impacts experienced at concentrations below contingency levelswould be very useful. It was noted that not all people do the same calculus; some focus on economic issues, some on politics. Also, determining how much a life is worth for economic analyses is a very difficult issue.

Public opinion and the role of education was discussed. In Mexico City, the environmental agenda competes with other agendas (security, unemployment, etc.) A recent Mexico City newspaper survey found that the most pressing problem for the population had shifted from security to pollution. There was a suggestion that we have been too successful in raising the alarm against air pollution, at the expense of other environmental problems like water quality. Nevertheless, there was general agreement about the need to educate people that air quality is part of their well-being. Bad public relations in this area have led to a defeatist attitude, as evidenced by the article in that day's New York Times, "A Fatal Case of Fatalism," which discussed air pollution in Mexico City. The example of cancer was raised to suggest alternative strategies. The American Cancer Society recently started a program of "good, better, and best news" to overcome the defeatist attitude by providing people with hope and focusing their actions.

It was also noted that educational efforts should attempt to condition people to the value of sacrifice. In this respect, it's helpful to recognize two different kinds of democracy: one that looks for the best solution for society as a whole (the Scandinavian design) and individualistic democracy, where each individual's stake is equally valid and the result is often paralysis.

The issue of exposures vs. ambient pollution concentrations was further discussed. Participants agreed with the presenters that we should be looking at how to reduce human exposures over the long term, not just reducing source emissions or ambient concentrations. Risk assessment is the link needed to target reducing human exposures.

The topic of assessing exposures raised the issue of confounding factors, which generated a few questions. Indoor ozone exposures are dramatically lower than outdoors, and this can confound effects. This is why exposure assessment is important. Researchers are able to control for meteorological variables and other pollutants on the "dose" side of the equation; they were asked how they control for other causes of respiratory diseases, such as ordinary colds, on the "response" side. There is a need to disentangle various health problems, especially considering that most schools and homes in Mexico City are not heated and people blame any symptoms they have on ozone. (It was noted that the city is "ozonized;" everyone knows when ozone levels are elevated and when they "should" feel sick.)

Though particles are routinely monitored, and contingency measures triggered at certain levels, this approach doesn't account for varying PM composition in different areas of the city. Other complicating factors in using health research to set standards are (1) standards are set for one pollutant at a time, which doesn't necessarily address the effects of multiple exposures; and (2) standards are set to protect the majority, without considering subgroups. (Children breathe more air and at a faster rate than adults, for example, so their exposures are typically higher.)

Participants discussed how to reach decision makers, suggesting we focus on those issues with opportunities for the biggest impact. We should first educate the publicperhaps focusing on a study of combi drivers or infants, as suggested earlierin order to mobilize popular demands for action, then target a small but powerful group of decision makers (the President, the finance minister, regulators including Congress).

Footnote

1. The monitoring stations and study areas in the city are: Xalostoc (northeast); Tlanepantla (northwest); Iztapalapa (southeast); Centro (downtown); Plateros (southwest); Pedregal (southwest); and Netzahualcoyotl (east).