Key findings of air pollution health effects and its application

Transcription

1 CAFE Steering Group (Brussels, May 2004) Key findings of air pollution health effects and its application Recent results from the WHO systematic review Jürgen Schneider Project Manger WHO ECEH, Bonn, Germany

2 CAFE Steering Group (Brussels, May 2004) Content Introduction Findings Outlook

3 Published reports: Health aspects of air pollution with particulate matter, ozone and nitrogen dioxide Health aspects of air pollution answers to follow-up questions form CAFE Meta analysis of time-series studies and panel studies op particulate matter and ozone In preparation: Introduction - Products Impact of air pollution on children s health (2 nd draft internally available) Short summary report (in preparation) Complementary are the reports on the 6th and 7th meetings of the joint WHO/UN ECE Task Force on Health

4 WHO Review - process Process: Background document on epi evidence Background document on tox evidence Draft answers incl rationale Review and revisions by SAC External peer review Finalisation, editing, print Revision and additional review WHO WG meeting

5 Key findings A full risk assessment comprises: Hazard identification: Review of relevant evidence (epidemiologic, toxicologic, etc. to determine wether the agent poses a hazard Exposure response function: Quantifying the relationship between exposure and adverse health effects Exposure assessment: Determination of the magnitude, duration and route of exposure Risk characterication: An integration of the above three leads to an estimation of the health budren of the hazard. The approach, assumptions and uncertainties should be highlighted

6 Key findings Hazard identification Short term versus long term Hot spot versus urban background Epi versus tox evidence (susceptibility) Thresholds for ozone WHO AQG for NO2 Coarse PM (Other pollutants)

7 Systematic Review key finings Relevance of exposure to peaks versus long term Peak versus long-term: Ozone: As there are usually many more days with mildly elevated ozone levels than days with very high peaks, the largest burden of disease might be expected with the many days with mildly elevated levels and not with the few days with very high levels PM: the public health significance of the long-term effects clearly outweighs the public health significance of the short term effects. This obviously does not diminish the significance of the short-term effects of PM, which consist of very large numbers of attributable deaths and cardiovascular and respiratory hospital admissions in Europe.

8 Systematic Review key findings Relevance of exposure at hot spots; exposure to peaks - Hot-spot versus urban background A policy which aims at a significant reduction of the overall health burden caused by air pollution will have to aim at a reduction of the exposure of the general population. This is in particular true for pollutants/health endpoints with (a) no threshold of effects and (b) a linear relationship between exposure and response.

9 Systematic Review key findings Relevance of exposure at hot spots; exposure to base case peaks - Hot-spot versus urban background No exceedance' General reduction Population exposed PM10 µg/m3, JMW

10 Systematic Review key findings Relevance of exposure at hot spots versus urban background However, some studies have documented that subjects living close to busy roads experience more short-term and long-term effects of air pollution than subjects living further away. In urban areas, up to 10% of the population may be living at such hot spots. The public health burden of such exposures is therefore significant. Unequal distribution of health risks over the population also raises concerns of environmental justice and equity.

11 Systematic Review key findings Consistency of epi and tox evidence in the defining thresholds - General remark For a uniform population with specific exposures, thresholds for certain pairs of pollutant/health endpoints may be detectable At a population level, susceptibility and exposure vary over a wide range. Large differences in individual exposure-response curves may lead to a disappearance of thresholds at a population level. Taking into account these differences, the evidence coming from the epidemiological and toxicological studies is not contradictory.

12 Systematic Review key findings Consistency of epi and tox evidence in the defining thresholds - PM Most epidemiological studies on large populations have been unable to identify a threshold concentration below which ambient PM has no effect on mortality and morbidity. It is likely that within any large human population, there is a wide range in susceptibility so that some subjects are at risk even at the low end of current concentrations.

13 Systematic Review key findings Consistency of epi and tox evidence in the defining thresholds - Ozone: Statistical significant associations between ozone and mortality have been demonstrated at places with low ozone levels Sophisticated statistical analysis applied to address the question of thresholds for PM have not been applied to the same degree for O 3 There remain uncertainties about the shape of the CR function, in particular at the lower end of the ambient range Although there is evidence that associations exist well below the current guideline value, the confidence in the existence of associations with health outcomes decreases as concentrations decrease.

14 Systematic Review key findings Consistency of epi and tox evidence in the defining thresholds - Ozone: A recent paper (Kim et al., 2004) applied a linear model, a natural spline model and a threshold model to a dataset in Seoul and found that the threshold model, with a threshold at 56 µg/m 3 (28 ppb) 1 hour average for effect son mortality, gave the best fit. However, the slope above the threshold was steeper than in the linear model. The authors conclude: that models that not take thresholds into consideration, could underestimate the true risk of ozone effects on mortality. Effects Effects Levels Levels

15 Systematic Review key findings Contribution of different sources to PM related health effects Epi studies: Combustion sources are particularly important for health effects. Tox studies: Primary, combustion-derived particles have a higher toxic potency. These are often rich in transition metals and organics, in addition to their relatively high surface area Other single components of the PM mix are lower in toxicity in laboratory studies e.g. ammonium salts, chlorides, sulphates, nitrates and windblown crustal dust such as silicate clays. Despite these differences among constituents studied under laboratory conditions, it is currently not possible to quantify the contributions from different sources and different PM components to health effects from exposure to ambient PM.

16 Systematic Review key findings Coarse particles and health There is limited evidence that coarse particles are associated independently of PM2.5 with mortality in time series studies. One study has investigated the effect of long-term exposure to coarse particles on life expectancy without producing evidence of altered survival. There is evidence that coarse particles are independently associated with morbidity endpoints such as respiratory hospitalizations in time-series studies. There is sufficient concern about the health effects of coarse particles to justify their control.

17 Systematic Review key findings Basis for maintaining the WHO AQG for NO 2 New toxicological studies show that LT exposure to NO 2 at concentrations higher than current ambient concentrations have adverse effects. Uncertainty remains over the significance of NO 2 as a pollutant with a direct impact on human health at current ambient air concentrations NO 2 has been associated with adverse health effects even when the annual average NO 2 concentration is within a range that includes 40 µg/m 3 It is recommend that the WHO annual specific guideline value of 40 µg/m 3 should be retained or lowered.

18 Systematic Review key findings Effects of air pollution on children s health There are several factors which potentially increase children s susceptibility to adverse effects of air pollution There is solid evidence for effects of air pollution (PM) on infant mortality Poor air quality effects lung development of children Air pollution is associated with increased upper and lower respiratory symptoms in children Air pollution may increase bronchitis, cough and aggravates asthma symptoms There are uncertainties whether ambient air pollution at current levels causes cancer in children

19 Systematic Review key findings The evidence is sufficient to recommend strongly further policy action to reduce levels of air pollutants including PM, NO 2 and ozone; it is reasonable to assume that a reduction of air pollution will lead to considerable health benefits.

20 Systematic Review key findings

21 Systematic Review key finings Meta-analysis of time-series studies Set-up of a Task Group upon recommendation of a WHO WG Development of a strict protocol Bibliographic database of time-series studies at St. George s Hospital/London Summary estimates for different health outcomes related to PM and ozone Outcome in RR per 10 µg/m3 increase in pollution level Analysis of possible publication bias

22 Systematic Review key finings European studies on ozone and and all-cause mortality 1.03 Barcelona Madrid Valencia Le Havre Lyon Paris Rouen Strasbourg Toulouse West Midlands London Netherlands Amsterdam Rome Turin Summary estimate RR for a 10 µg/m3 increase in daily ozone levels 0.97

23 Systematic Review key finings RR for all-cause mortality PM and ozone RR Relative risks (RR) for mortality endpoints related to a 10 µg/m 3 increase in pollution including 95 % confidence intervals; left part: PM10, black smoke (BS) and ozone from European studies; right part: PM2.5 from North American studies PM10 BS Ozone PM2.5 all cause respiratory cardiovascular all cause respiratory European studies cardiovascular all cause respiratory cardiovascular all cause respiratory American studies cardiovascular

24 Systematic Review key finings RR for all-cause mortality publication bias Relative Risk for all-cause mortaility, 10 µg/m 3 increase inpollution level, without and with correction for publication bias PM10 BS Ozone RR all cause all cause corr. all cause all-cause corr. all cause all-cause corr

25 Systematic Review key finings No indication for PM threshold Stockholm Birmingham, West Midlands Paris Helsinki London Rouen Basel Zurich Strasbourg Le Havre Geneva Madrid Amsterdam Koln Netherlands Lyon Athens Budapest Huelva Florence Bologna Teplice Seville Palermo Tel Aviv Czech Republic Milano Erfurt Cracow Rome Barcelona Torino Prague mean PM10 RR mean PM10 level in µg/m3 RR increase per 10 µg/m

26 Systematic Review key findings

27 7 th Meeting of the TF Health Develop recommendations on the approach to model health impacts within RAINS Health impact assessment: Science-based No cookbook approach yet Several assumptions and choices transparency is needed Conservative approach CBA complementory

28 7 th Meeting of the TF Health - Ozone Ozone effects: Health endpoint: attributable mortality Causal relationship between ozone exposure and mortality Independent of PM RR from WHO meta-analysis of European studies: The relative risk for all-cause mortality is (C.I , 1.004) for a 10 µg/m 3 increase in the daily maximum 8-hour mean.

29 7 th Meeting of the TF Health - Ozone Ozone effects: A linear concentration response function is assumed. The principle metric for assessing effects of ozone on mortality should be the daily maximum eight-hour mean. Current evidence is insufficient to derive a level below which ozone has no effect on mortality. The use of a cut-off for IAM at 35 ppb is recommeded

30 80 70 daily max 8-hr mean daily max 8-hr mean

31 Ozone Why use a cut-off? There remain uncertainties about the shape of the CR function, in particular at the lower end of the ambient range 20 to 40 ppb is the hemispheric background main purpose is to analyse European abatement options There are difficulties of state-of-the-art dispersion models to simulate low ozone situations in urban areas. Titration becomes important at low levels

32 Ozone This recommendation is based on the application of a very conservative approach for integrated assessment modelling and takes account of uncertainties in relation to the evidence in health studies and lower confidence in model performace at low ozone levels. It is recommended to make a sensitivity analysis applying no cut-off. This estimate would outline an upper estimate of the attributable effects of ozone on mortality. The Task Force also stressed that the approach did not yield an overall quantification of all effects related to exposure to ozone. Important effects which were currently not covered, but should be taken into account in CBA, including morbidity outcomes.

33 Ozone Effects Effects Levels Levels

34 PM effects: Particulate matter Use annual mean of PM2.5 as indicator for PM related mortality (this also implies the use of one factor for different PM constituents) RR from cohort study Linear CR function Health outcome: Reduction of life expectancy Effect for explained anthropogenic contribution only Most epidemiological studies on large populations have been unable to identify a threshold concentration below which ambient PM has no effect on mortality and morbidity. It is likely that within any large human population, there is a wide range in susceptibility so that some subjects are at risk even at the low end of current concentrations.

35 Follow-up actions Systematic Review outlook Finalisation of the reports Dissemination of main results Keep contact with CAFE Revision of WHO Air Quality Guidelines (PM, ozone)

36 CAFE Steering Group (Brussels, May 2004) For further information see: