1、Ambient Ozone Health EffectsNESCAUM Health Effects Workshop Joann Held July 29, 2008With thanks to: Gregg Recer, PhD Center for Environmental Health New York State Department of Health,Ozone Health EffectsExperimental Studies Human Chamber StudiesEpidemiological Studies Panel/Field Studies Populatio
2、n Based Time Series Studies,Topics To Cover,Ozone Health Effects,Acute & Chronic Respiratory Effects Cardiovascular Effects Daily MortalityData Sources: Human clinical chamber studies Epidemiology studies Also: Animal toxicology studies In vitro/mechanistic studies,Experimental Studies,Strengths Exp
3、osures are controlled Almost all sources of variation excluded Provides strong evidence of causality Prospective Studies Limitations (Usually) Very high dose (Usually) Not humans (Species to species extrapolations required) (Usually) Inbred animals Restrictive exposure conditions,Epidemiological Stu
4、dies,Strengths No need to extrapolate across species, provides direct evidence of human effects Limitations Exposures are complex and always changing Exposures are to many chemicals Difficult to control for bias and confounding factors (correlation vs. causality) Retrospective evaluations Conclusion
5、s require weight-of-evidence judgements based on entire body of evidence vs. causality criteria,Ozone Background,Reactive gas “Category 1 gas” Acute effects - Respiratory irritant Chronic effects Associated with accumulated effects of repeated acute insults (?),History of the Ozone NAAQS,Original: 0
6、.08 ppm for 1-hour(as total oxidants)1979: 0.12 ppm for 1-hour (revoked in 2005)1997: 0.08 ppm for 8-hours2008: 0.075 ppm for 8-hours,1996 Ozone EPA AQCD,Most conclusive evidence: Controlled acute human exposures Acute lung function deficits (e.g., reduced FEV1) Field/Panel studies Acute lung functi
7、on deficits (e.g., reduced PEF) Highly suggestive evidence: Associations w/ respiratory Emergency Departments (ED) visits, hospitalizations Co-pollutant interactions & chronic effects unclear,AQCD = Air Quality Criteria Document,2005 Ozone AQCD,Further evidence from controlled exposures suggestion o
8、f sensitive individuals below 0.08 ppm Much larger epidemiology database: Acute respiratory effects considered causal Strong evidence of association w/ daily mortality Cardiovascular effects suggestive, but evidence inconclusive Chronic respiratory effects inconclusive Lack of conclusive evidence fo
9、r several other endpoints (chronic mortality, lung cancer, developmental effects),Human Chamber Studies,Observed effects at near-ambient concentrations for one several hours: Spirometry effects (e.g., FEV1) Bronchoconstriction Airway hyper-responsiveness Lung inflammation Airway cell damage Shallow/
10、rapid breathing pattern Airway irritation cough/pain on inspiration,Near-ambient chamber data - Adams & McDonnell et al,Exposure conditions: ozone or filtered air 6.6 hr total exposure time intermittent exercise (50 min/hr) Main conclusions: Significant effects at 0.080 ppm - FEV1 decrement Increase
11、d symptom score & PDI No significant group effects at 0.080 ppm,Near-ambient chamber data,B.,Limitations of Chamber Studies,At lowest levels, mainly healthy adults Only exposed to ozone; not ambient pollutant mix Spirometry changes may not be most sensitive indicator of relevant effects Small sample
12、 sizes limit statistical power,Epidemiology Evidence - Respiratory effects,Panel/Field studies Time series studiesCommon effect estimate metrics: Relative risk (RR): 1 = increased risk % change in risk: 0 = increased risk scaled to change (increment) in pollutant levelOdds Ratio (OR): 1 = increased
13、risk,Field/Panel Studies,Repeated observations of lung function and concurrent exposure levels Individual-level data on health endpoint and (sometimes) exposure Relatively small sample sizes,Field/Panel Studies,80s/90s camp studies: 600 children, mostly healthy, non-asthmatic; multiple measurements
14、per child significant FEV1 with increased ozone ozone predominant ambient pollutant associated with health endpoint mean levels - 0.053 - 0.123ppm maximum levels - 0.095 -0.245 ppm,Field/Panel Studies,Mortimer et al (2002) (NCICAS): 8 cities 800 asthmatic children daily diaries of peak expiratory fl
15、ow rate (PEF) and asthma symptoms during June - August 1993 significant morning PEF with increased ozone significant symptoms with increased ozone ozone effect on PEF largely robust in multi-pollutant models ozone symptom incidence effect reduced somewhat in multi-pollutant models but trend remains
16、max 8-hour 99th %-ile = 0.066 ppm significant result remains after removing days 0.08 ppm,Field/Panel Studies,Korrick et al (1998): Adult hikers at Mt. Washington; multi-hour exposures Results, all hikers: significant FEV1 post- v. pre-hike strength of effect similar after controlling for PM2.5 and
17、acidity (P0.05) max 8-hour 99th %-ile = 0.089 ppm,Korrick et al results,Population-based Studies,Acute endpoints assessed via daily time-series analysis Large sample sizes Consistent statistical design across studies Area/Regional monitor exposure data More severe endpoints (e.g., hospitalizations)
18、Confounding important consideration, e.g.: temperature, season, day of week, co-pollutants, local exposures, socioeconomic status (SES),Population-based Studies,Respiratory outcomes: emergency department visits unscheduled hospitalizations mortality,Population-based Studies - summary,% increase in r
19、isk per standard increment,Acute cardiovascular effects,Panel studies: Heart rhythm variables (ARIC, NAS, HSPH): ozone associated with reduced HRV & ventricular arrhythmias PM effects on HRV generally larger and more consistent for different outcome measures ozone effect on arrhythmias robust in two
20、-pollutant models,Acute cardiovascular effects,Panel studies: Myocardial Infarction - MONICA: current-day ozone associated with acute MI no PM co-pollutant analysis Myocardial Infarction - Peters et al (2001): non-significant trend toward increased risk with increased ozone in previous 2 hours PM ef
21、fects were stronger,Cardiovascular effects - hospitalizations,Acute cardiovascular effects,Overall summary: Results from field/panel studies are suggestive of an association between ozone and cardiac outcome measures Results from hospitalization time-series studies are inconclusive,Chronic respirato
22、ry effects,1996 AQCD: Repeated exposure studies in animals show irreversible lung damage Human epidemiology data too limited to conclude whether chronic effects occur More recent longitudinal & retrospective epidemiology studies focus on asthma development & lung-function development,Chronic respira
23、tory effects,Seasonal studies: Lung-function growth in children inversely related to summer ozone concentrations Limited evidence of recovery in lung-function parameters after ozone season,Chronic respiratory effects,Childrens Health Study Long-term cohort from 12 SoCal communities Cross-sectional a
24、nalyses: self-reported asthma prevalence, cough, bronchitis & wheeze unrelated to max 1hr ozone limited evidence of reduced baseline lung function in females related to annual average ozone limited evidence of reduced lung function growth related to ozone levels, but weaker than PM10 effects (subset
25、 who moved from area),Chronic respiratory effects,Childrens Health Study Longitudinal analysis (first cohort): Annual average 8hr or 24hr ozone unrelated to lung function growth measures Annual average 8hr or 24hr ozone unrelated to reduced FEV1 Significant effects on FEV1 and lung function growth d
26、etected for other pollutants,Chronic respiratory effects,Development of asthma: CHS - No increased risk comparing high ozone to low ozone communities Within high ozone communities RR = 3.3 for those playing 3+ sports vs. no sports No relationship to sports activity in low ozone communities (effect m
27、odification) No effects of other pollutants on asthma incidence,Chronic respiratory effects,Development of asthma: AHSMOG - 15-year prospective follow up of new-onset asthma in non-smoking white adults Increased risk of asthma in males, but not females No effects of other pollutants on asthma incide
28、nce Ozone effect robust to co-pollutants Robust effect in 2 follow-up periods, but limited generalizability,Chronic respiratory effects,Retrospective studies: Yale freshmen - Reduce lung function in cohort from high ozone communities vs. low ozone communities Effect stronger in males than females No
29、 co-pollutant analysis UC Berkley freshman - Reduced lung function comparing student from LA area to students from San Francisco area when adjusting for intrinsic airway diameter Robust effect in multipollutant models,Chronic respiratory effects,Summary Some suggestive evidence of increased asthma i
30、ncidence & decreased lung function growth Large prospective cohort study found little relationship to ozone compared to co-pollutants Ecological designs harder to control for exposure mis-classification & confounding Overall evidence for chronic respiratory morbidity inconclusive,Daily Mortality Stu
31、dies,Population Based Time Series Studies: Estimating associations between mortality and daily variations in population average exposures,Daily mortality - single-day lag,Daily mortality - multi-day lag,Daily mortality - PM adjustment,Bell et al/NMMAPS: 24-hour ozone 90th %-ile values range from 0.0
32、25 - 0.060 ppmPooled effect estimate robust to: PM excluding high temperature days excluding high ozone days ( 0.060ppm 24-hour average ozone),Daily mortality - meta-analysis summary,Daily mortality - cause-specific,Bell et al - cardiopulmonary RR all cause RR Bell et al - cardiovascular RR total RR
33、 respiratory RR (NS) Single city & smaller meta-analyses: cardiovascular effects more consistently positive and significant respiratory RRs vary from larger to negative (fewer cases?),Mortality,Summary Robust association between daily mortality and increased ozone Not sensitive to co-pollutants, daily weather Larger risks in analyses limited to warm season Cardiovascular mortality somewhat stronger and more consistent than respiratory,
