Health Effects of Nitrogen Dioxide and the EU policy Francesco Forastiere Roma, 16 October 2012

Policy makers for environmental issues Local councils Regional environmental authorities Ministry of Environment and Industry EU DG Environment Regional and national Health Authorities

Types of policy makers Informed/uninformed Honest / not honest Conflicts of interests / no conflicts Equality / inequality Environmental sustainability Yes/ No Long-term planner / short-term planner Ability to manage stakeholders Yes/ No Take a decision or wait for the next term

What kind of work we do? Provide a single effect estimate (one point with bars) Provide a pooled effect estimate (one BIG point with bars) Provide a narrative synthesis of the evidence (with/without uncertainty) Provide health impact assessment Provide integrated environmental and health impact assessment

What is the most relevant evidence? Evidence Causality Latency period Dose-response Threshold What if... Impact assessment Qualitative / quantitative Qualitative Quantitative Qualitative Quantitative

The Nitrogen dioxide case

Description of the problem Nitrogen dioxide is one the classical outdoor air pollutants A wide range of effects on health have been found by epidemiological studies to be associated with NO2 WHO has recommended both short- and long-term guidelines which have been translated into standards The annual average Limit Value (40μg/m3, annual average) is difficult to meet, at least at some sites

Description of the problem And yet, NO2 has faded a little, from current thinking about the effects of air pollutants on health. We have mainly focused on particles and on ozone. Sulphur dioxide and carbon monoxide are now less of an outdoor problem than they were.

Description of the problem In the EU: policies are cost-benefit tested This requires coefficients that are firmly based on associations regarded as causal. And here is a problem: how many of the coefficients linking NO2 and effects on health do we regard as causal?

Description of the problem Nitrogen dioxide may have: –Primary effects: the effects of exposure to NO2, per se. –Secondary effects: the effects of ozone and nitrate particles: formation dependant on NO2; –And may be associated with “effects” by virtue of its acting as a surrogate, an index or marker, for other pollutants. Here is the central problem: What evidence do we have for these primary, secondary and surrogacy effects?

Description of the problem Primary effects: if true, we should control emissions of NO2 or, perhaps, exposure to NO2 Secondary: if true, we have options: there may be better ways of controlling the production of ozone and fine particles than by controlling emissions of NO2. This conceals a further problem: to what extent do nitrate particles contribute to the toxicity of the aerosol we monitor as PM2.5 Surrogacy: if NO2 is a non-toxic surrogate for some other pollutant (or pollutants), then controlling NO2 will only lead to a predictable reduction in effects on health if the ratio of NO2 to the active component(s) remains constant.

The EC- WHO project “Evidence on health aspects of air pollution to review EU policies – REVIHAAP” Questions from EU to WHO

REVIHAAP timeline June: SAC Meeting Aug: REVIHAAP Expert Meeting 23 Aug - 21 Sept: Finalization of rationales by authors 1 Oct - 9 Nov: External review 19 Nov - 18 Dec: Preparation of final version by authors 18 Dec: Final version ready & sent to meeting participants Jan 2013: Adoption of REVIHAAP material during REVIHAAP/HRAPIE Meeting April 2013: Final REVIHAAP report

Question C2 Is there any new evidence on the health effects of nitrogen dioxide (NO2) that impact upon the current limit values ? Are long-term or short-term limit values justified on the grounds that NO2 affects human health directly, or is it linked to other co-emitted pollutants for which NO2 is an indicator substance?

Question C4 Based on currently available health evidence, what NO2 metrics, health outcomes and concentration-response functions can be used for health impact assessment?

Per se….or…. not per se

Understanding the importance The question of the independence of effects of NO 2 and PM 2.5 is important for policy, as the effectiveness of NO 2 standards for health protection is being questioned, so this should be considered carefully in the development of the answers.

Approach Short Answers, longer Rationales Evidence summaries: –Epi studies of short -term exposure –Epi studies of long -term exposure –Experimental studies – etc. Search criteria set out, for a set of answers; with detailed references – clearly a huge amount of work Initially the questions were appreciated because they were specific about pollutants – and they do capture important specific aspects. BUT they don’t make it easy to present the necessary content in a clear way. … Can we change the Questions?

Style The questions are clear and the answers should directly address the questions posed. The technical level of the answers should be the same as the questions. The language used should be understandable to the intended audience, in particular policy makers who are not necessarily technical experts. The rationales should relate directly and to and clearly support the answers provided. The rationales should be concise. More extensive reference to existing reviews may help.

Methods The process of reviewing the new evidence (since 2005) from epidemiology and experimental studies has been straightforward. For epidemiologic studies (short and long term) evidence of associations has been considered together with the following aspects: the levels at which health effects have been noted, the shape of the concentration response curve, and the stability of the effects after adjustments for other pollutants. For experimental studies (in animals and in humans), the evidence of effects under controlled and consequently limited sets of conditions has been evaluated. The overall main difficulty is in drawing all these aspects together for an understanding and assessment of causality and the key discussion is on the role of toxicological evidence.

NO2 controlled human studies -summary - In healthy subjects, changes in pulmonary function, ↑ airway responsiveness, mild inflammation& ↓host defences at concentrations (>1800 μg/m3+/-) in excess to those outdoors Asthmatics more susceptible to acute effects In mild asthmatics, lowest concentration to change pulmonary function: 500 μg/m3and to enhance effect of allergens: 200 μg/m3

Uncertainty IPCC AR4 Summary for policy makers: -The changing climate is very likely (more than 90%) due to the observed increase in anthropogenic greenhouse gases -IPCC have a heirarchy of probabilities ranging with associated language

Impact assessment Least uncertainty Respiratory hospital admissions (short term) Increased degree of uncertainty All cause mortality (short-term). Increased degree of uncertainty for cardiovascular mortality where there is an absence of a solid body of supporting chamber study and toxicological evidence. Bronchitic symptoms in asthmatic children, long-term (McConnell et al., 2003). Most uncertainty [For the pollutant-outcomes there was increased difficulty of controlling for confounding by other pollutants, less data or less available supporting clinical or toxicological evidence]. Cardiovascular admissions (short-term). Asthma prevalence All cause mortality (long-term)

Broader issue  to consider what package of guideline recommendations will maximise human protection, rather than set of individual guidelines

Key points  Difficult to formulate policy relevant questions  Difficult to provide answers to those questions  Difficult to fit a rationale for the answers  When the answer is provided, how useful it will be for the policy setting?