Traffic-Related Air Pollution and Childhood Asthma

We congratulate Clark et al. (2010) for their interesting article concerning traffic-related air pollution and asthma in children. They examined early-life (in utero and during the first year of life) exposure to traffic-related air pollution in a large population-based study (a nested case–control study including nearly 3,500 children). The authors found an association between elevated early-life exposure to traffic-related air pollution and a higher risk of asthma in preschool-aged children. However, Clark et al. (2010) address aspects, raise questions, and give results that deserve further comment—both concerning specific items and general “structural” criteria used in epidemiological studies on adverse health effects from air pollution. 
 
Recent reports have suggested that individual susceptibility could play a previously unsuspected role in the occurrence of diseases (Cetta et al. 2009a), perhaps a role greater than that of the intrinsic toxicity of pollutants (Cetta et al. 2009b). This could explain, at least in part, why it is so difficult to determine a precise threshold concentration that is harmful or safe for each individual (Cetta et al. 2007). But this is just one side of the question. 
 
The main question is that, in the absence of adequate and specific markers of exposure, effect and susceptibility, the linear dose and effect model, and the concomitant pollutant concentration and disease occurrence relationship cannot explain the complexity of the phenomenon of host–particle interactions. In particular, initial cell alterations (e.g., oxidative stress, DNA adduct formation) rarely turn into permanent tissue damage and evident disease because of host repair and defence mechanisms. 
 
In their article Clark et al. (2010) noted another important aspect that should be considered when comparing pollutant concentrations with the burden of deleterious effects, both at the individual and population levels: acute effects of peak concentrations of pollutants that lead to acute admission to hospital and the chronic damage that causes long-term effects. In fact, we should consider both of these as separate entities. However, we also should consider the effect of air pollution on newborns, which greatly depends on individual susceptibility—either congenital or acquired. This could play a major role in future outcomes and shed new light on the peculiar pathophysiological mechanisms of most pollution-related diseases. Clark et al. (2010) correctly outlined the asynchronism and the delay (0–4 years) between the initial pathogenetic exposure to pollutants (in utero or during the first year of life) and the occurrence and detectability of clinically relevant asthma. This further adds to the complexity of host-particle interactions. 
 
There are three issues that should be taken into account in developmental epidemiology studies such as that by Clark et al. (2010). First, epidemiological studies that concomitantly evaluate pollutant concentration and detectable diseases or hospital admissions usually neglect the perinatal damage in fetuses and newborns, which is not immediately detectable but is a delayed manifestation. 
 
Second, perinatal damage from air pollution deserves further attention and detailed analysis because it includes fetal malformations, birth defects, and developmental alterations of newborns. Injury from perinatal air pollution exposure could also be responsible for the increased proportion of unsusceptible individuals who, because of their exposure to pollutants during the susceptibility window and because of epigenetic alterations due to environmental factors, will become susceptible. Epigenetic alterations could also transfer this susceptibility to future generations, leading to individuals developing not only asthma at 4 years of age but also respiratory, cardiovascular or systemic diseases ≥ 20 years later. 
 
Third, clinical and pathophysiological details are not “details” but basic issues and questions—still unsolved— that should be primary goals for future research. They are critical to improving design of epidemiologic studies and to selecting appropriate models, which should also include biological and pathophysiological parameters and variables because they significantly affect clinical outcomes.

The statement concerning chrysotile asbes tos being "implicated but not conclusively established as a cause of mesothelioma" is inconsistent with current scientific opinion. I refer you to the most recent evaluation by the International Agency for Research on Cancer in which Straif et al. (2009) stated, Epidemiological evidence has increasingly shown an association of all forms of asbestos (chryso tile, crocidolite, amosite, tremolite, actinolite, and anthophyllite) with an increased risk of lung cancer and mesothelioma. Although the potency differences with respect to lung cancer or meso thelioma for fibres of various types and dimen sions are debated, the fundamental conclusion is that all forms of asbestos are "carcinogenic to humans" (Group 1).
In addition, opinions such as that expressed in the Editor's Summary are advanced only by scientists with pro chrysotile industry bias.
When I wrote the draft for the first IARC Monograph on asbestos in 1976, which the expert committee accepted and published in 1977 as IARC Monograph Volume 14, a simi lar conclusion was stated: "Many pleural and peritoneal mesotheliomas have been observed after occupational exposure to crocidolite, amosite and chrysotile." Since then-more than 30 years-science has not changed its opinion that all forms of asbestos, including chrysotile, cause mesothelioma.
In fact, in the article that is the subject of the Editor's Summary, Tse et al. (2010) did not indicate that chrysotile is not a cause of mesothelioma; on the contrary, they stated the following: Although the mesothelioma incidence is antici pated to decline in the coming decades, it may not decrease to background risk levels given that chrysotile consumption has not been banned under the current legislation and that second ary asbestos exposure from the environment will likely continue. Nevertheless, the hypotheses generated from this ecologic study need further confirmation by subsequent analytic studies. The present study provides supportive evidence for an immediate and global ban on asbestos use. I hope that future Editor's Summaries will reflect the conclusions of the article and not put forth statements that are not sup ported by mainstream science. I also support the conclusion of Tse e al. (2010) for "an immediate and global ban on asbestos use." The author testifies in asbestos litigation on behalf of plaintiffs.  Attaran et al. 2000;Roberts 2001Roberts , 2007bRoberts et al. 2000Roberts et al. , 2004Tren 2009), these articles and letters reduce the complex issue of malaria control to a single, dichotomous choice between DDT and malaria. Framing the issue in this manner is a dangerous over simplification and an distraction from the critical dialog on how to effectively combat malaria around the world-particularly in African communities.

Richard A. Lemen
The question that AFM and malaria con trol experts must ask is not "Which is worse, malaria or DDT?" but rather "What are the best tools to deploy for malaria control in a given situation, taking into account the on theground challenges and needs, efficacy, cost, and collateral effects-both positive and negative-to human health and the environ ment, as well as the uncertainties associated with all these considerations?"  briefly acknowl edged that alternatives to DDT exist (while denigrating them as "supposed solutions"), but in typical fashion they focused most of their letter on the chemical, arguing that the health effects of malaria are much worse that those of DDT exposure. As malaria professionals we are well aware of the dire health consequences of malaria, but also of DDT. The challenge before us is therefore to determine how much weight to give to vec tor control within the broader context of a malaria control program; within vector con trol, how much weight to allot to nets versus indoor residual spraying (IRS); and within IRS, how much weight to give to DDT or some other chemical.
These decisions are indeed complex and location specific. In this regard, van den Berg's commentary, "Global Status of DDT and Its Alternatives for Use in Vector Control to Prevent Disease" (van den Berg 2009), is a most useful contribution. In contrast,  advice that "van den Berg's concerns should be ignored" strikes us as reckless and irresponsible.
In 2006, Allan Schapira, former coor dinator of vector control and prevention of World Health Organization's Global Malaria Programme, observed that malaria control discussions had become "polluted," and warned, "The renewed interest in indoor residual spraying could lead to interminable debates in countries about the pros and cons of DDT" (Schapira 2006). However well intentioned, -as with much of AFM's output-do more to fuel those "interminable debates" than to meaningfully inform decisions that will save people's lives. Herren and Mbogo's critique of our response  to van den Berg (2009) is lacking in substance. In their letter, they attack our work by characterizing our advocacy for using DDT to control malaria as a distraction from larger malaria control issues. These authors apparently discount the fact that some African countries are presently making highly effective use of DDT to reduce both malaria deaths and malaria infections. Countries that use DDT benefit from its spa tial repellent action that stops mosquitoes from entering houses and transmitting disease, and no alternative insecticide does this . In addition, Herren and Mbogo apparently do not understand that our advo cacy is consistent with that exhibited by the malaria control community, with hundreds signing a petition to prevent DDT elimination through Stockholm Convention negotiations. If DDT had been eliminated, countries pres ently using DDT would have been deprived of its benefits for protecting health and sav ing lives. Herren and Mbogo claim that our response to van den Berg's commentary (van den Berg 2009) was fixated on DDT, in lieu of addressing the larger issues of what should be done to control malaria. In our letter , we addressed what we con sidered to be an attack on DDT use. How could we have responded without addressing the issues in van den Berg's commentary?
Herren and Mbogo mischaracterize our position visàvis DDT and alternative insecticides by asserting that we are reduc ing the malaria control debate to a simplis tic equation of malaria or DDT. In fact, we have a public record of supporting the use of insecticidetreated nets and the use of alterna tive insecticides for malaria control. However, we have repeatedly emphasized that, for obvi ous reasons, insecticidetreated nets are not the only solution for malaria control. In fact, we object to a theme of nets and nets alone as much as we would object to a theme of DDT and DDT alone. Basically, there is no single solution approach to malaria control. All tools are needed-not just those that are currently in vogue.
Herren and Mbogo state that they are fully aware that malaria is a worse outcome than possible health effects of DDT. We agree with them and appreciate their willingness to admit this, because their admission opposes published specu lations that DDT might be causing more harm than good (Chen and Rogan 2003).
Herren and Mbogo conclude that we "do more to fuel those 'interminable debates' [DDT or no DDT for malaria control] than to meaningfully inform decisions that will save people's lives." It seems that these authors ignore the fundamental fact that we do not elaborate on alternative approaches to malaria control because the alternatives are not pres ently under threat of elimination. The alter natives are being used and should continue to be used, but the future is far less certain for DDT. Advocacy saved DDT from being eliminated during the original negotiations for the Stockholm Convention, and lives are being saved and diseases prevented as a consequence. The idea that the threat is over and that DDT is now available to those countries making effective use of it is wrong. The Stockholm Convention Secretariat is now planning to stop all production of DDT in 2017 and eliminate it entirely from use in malaria control programs in 2020 (UN Environment Program 2010).
The Stockholm Convention Secretariat plans to prevent future uses of DDT, even though there is no costeffective replacement for DDT. Given these circumstances, Herren and Mbogo should expect the interminable debates to become even more polemic in the future.
As for the big issues of what should be done to control malaria, our position is clear: Decisions should be based on scientific evidence of what actually works, on local circum stances, and on what proves to be the most costeffective in terms of reducing disease and preventing human deaths. Correspondence (a nested case-control study including nearly 3,500 children). The authors found an asso ciation between elevated earlylife exposure to trafficrelated air pollution and a higher risk of asthma in preschoolaged children. However, Clark et al. (2010) address aspects, raise questions, and give results that deserve further comment-both concerning specific items and general "structural" criteria used in epidemiological studies on adverse health effects from air pollution. Recent reports have suggested that indi vidual susceptibility could play a previously unsuspected role in the occurrence of diseases (Cetta et al. 2009a), perhaps a role greater than that of the intrinsic toxicity of pollut ants (Cetta et al. 2009b). This could explain, at least in part, why it is so difficult to deter mine a precise threshold concentration that is harmful or safe for each individual (Cetta et al. 2007). But this is just one side of the question.

R.T. runs a policy and advocacy group, Africa
The main question is that, in the absence of adequate and specific markers of exposure, effect and susceptibility, the linear dose and effect model, and the concomitant pollut ant concentration and disease occurrence relation ship cannot explain the complexity of the phenomenon of host-particle inter actions. In particular, initial cell alterations (e.g., oxidative stress, DNA adduct forma tion) rarely turn into permanent tissue dam age and evident disease because of host repair and defence mechanisms.
In their article Clark et al. (2010) noted another important aspect that should be con sidered when comparing pollutant concentra tions with the burden of deleterious effects, both at the individual and population levels: acute effects of peak concentrations of pollut ants that lead to acute admission to hospital and the chronic damage that causes long term effects. In fact, we should consider both of these as separate entities. However, we also should consider the effect of air pollution on newborns, which greatly depends on individual susceptibility-either congenital or acquired. This could play a major role in future outcomes and shed new light on the peculiar pathophysiological mechanisms of most pollutionrelated diseases. Clark et al. (2010) correctly outlined the asynchronism and the delay (0-4 years) between the initial pathogenetic exposure to pollutants (in utero or during the first year of life) and the occur rence and detectability of clinically relevant asthma. This further adds to the complexity of hostparticle interactions.
There are three issues that should be taken into account in developmental epide miology studies such as that by Clark et al. (2010). First, epidemiological studies that concomitantly evaluate pollutant concen tration and detectable diseases or hospital admissions usually neglect the perinatal damage in fetuses and newborns, which is not immediately detectable but is a delayed manifestation.
Second, perinatal damage from air pollu tion deserves further attention and detailed analysis because it includes fetal malforma tions, birth defects, and develop mental altera tions of newborns. Injury from perinatal air pollution exposure could also be responsible for the increased proportion of unsusceptible individuals who, because of their exposure to pollutants during the susceptibility window and because of epigenetic alterations due to environmental factors, will become suscep tible. Epigenetic alterations could also transfer this susceptibility to future generations, lead ing to individuals developing not only asthma at 4 years of age but also respiratory, cardio vascular or systemic diseases ≥ 20 years later.
Third, clinical and pathophysiological details are not "details" but basic issues and questions-still unsolved-that should be primary goals for future research. They are critical to improving design of epidemiologic studies and to selecting appropriate mod els, which should also include biological and pathophysiological parameters and vari ables because they significantly affect clinical outcomes.
This work was supported by The PROLIFE Project, Milan, Italy, and the Polaris Project from Fondazione Cariplo.