MESA Air Progress Report MESA Steering Committee Meeting

MESA Air Progress Report MESA Steering Committee Meeting March 18, 2009

Study Timeline: Years 1 -5 Year 1 08/04 - 7/05 Year 2 8/05 – 7/06 Year 3 8/06 – 7/07 Year 4 8/07 – 7/08 Year 5 8/08 – 7/09 Subject Recruitment; Baseline Exam MESA Exam 3 MESA Exam 4 Exposure Monitoring for Air and Ancillary Studies F/U Calls including Air (starting F/U 8) Surveillance for Clinical Events

Exposure Monitoring: Years 1 -5 Year 1 08/04 - 7/05 Year 2 8/05 – 7/06 Year 3 8/06 – 7/07 Year 4 8/07 – 7/08 07/05 Year 5 8/08 – 7/09 07/31/09 Fixed Site Monitoring: On-going through this summer 02/06 07/31/08 Residential Monitoring: Complete 10/06 07/31/08 Personal Monitoring: Complete 05/06 07/31/07 NOx Snapshots: Complete

Study Timeline: Years 6 -10 Year 6 08/09 - 7/10 Year 7 8/10 – 7/11 Year 8 8/11 – 7/12 Data Analysis and Manuscript Preparation MESA Exam 5 Follow Up Calls Surveillance for Clinical Events Year 9 8/12 – 7/13 Year 10 8/13 – 7/14

MESA Air “Baseline Exam” Complete, Follow Up Underway o Full recruitment into our Subclinical Cohort o Follow-Up 8 complete; F/U 9 in progress o Planning for MESA Air Exam 2 (MESA Exam 5) is underway

Health Data Available for Analysis o February 2009: n n n o MESA Exam 4 CT Baseline Air questionnaire New Recruit dataset (excluding IMT) Q 2 or Q 3 2009: n IMT

Exposure Assessment: Big Picture o o o Over-arching goal: Most accurate estimate possible of each individual study participant’s exposure to air pollutants of interest over the period of study Focus is on ambient air pollutants Concentration vs. Exposure

Individual Exposure Estimates EA = CA = [f o +(1 - f o)Finf] CA EA Ambient exposure Attenuation factor fo Time outdoors (1 - f o) Time indoors Finf Infiltration factor CA Outdoor concentration

Exposure Assessment Overview Integrating Information from: n n n Spatio-temporal outdoor concentration model (for CA) residential characteristics including pollutants’ infiltration efficiencies (for Finf ) personal time-activity records (for f o )

Indoor Pollutant Measurements Outdoor Pollutant Measurements Geographic Data Reported Housing Characteristics Observed Housing Characteristics Deterministic Models Spatio-temporal Hierarchical Modeling Infiltration Modeling Predicted Outdoor Concentrations at Homes Predicted Indoor Concentrations at Homes Reported Time/Location Information Weighted Average Measurements Personal Exposure Predictions for Each Subject Questionnaires Predictions

MESA Air Monitoring Campaigns o o Fixed Sites Home-based monitoring n n o o Outdoor Infiltration Traffic gradient and community-based NOx snapshots Personal monitoring

Exposure Monitoring: Sample Counts Paired indoor-outdoor residential locations (collected twice) NOx snapshot monitoring campaign locations (collected 3 x) * * Number of biweekly fixed site samples collected (2 -7 fixed sites per city) *LA and NY include New Recruit areas of Coastal LA, Riverside, CA, and Rockland NY

MESA Air Exposure Monitoring Concludes July 31, 2009 o Includes ancillary studies (HEI, Coarse*) o Exposure technicians in some sites being transferred to interview duties or other projects *with the possible exception of an additional month of Coarse monitoring in Chicago

Exposure Modeling In Progress o o o First iteration (“pragmatic”) of spatio-temporal modeling estimates complete, incorporating AQS data, early MESA Air monitoring data, land use and certain GIS covariates Infiltration modeling underway: first set of results available based on early data Incorporation of time-location data underway: first set of personal-level estimates

Indoor Pollutant Measurements Outdoor Pollutant Measurements Geographic Data Reported Housing Characteristics Observed Housing Characteristics Deterministic Models Spatio-temporal Hierarchical Modeling Infiltration Modeling Predicted Outdoor Concentrations at Homes Predicted Indoor Concentrations at Homes Reported Time/Location Information Weighted Average Measurements Personal Exposure Predictions for Each Subject Questionnaires Predictions

Individual Exposure Estimates EA = CA = [f o +(1 - f o)Finf] CA EA Ambient exposure Attenuation factor fo Time outdoors (1 - f o) Time indoors Finf Infiltration factor CA Outdoor concentration

Interim Long-Term Avg PM 2. 5 ( g/m 3) Estimates at Participant Homes 25 20 15 10 LA LA Riverside LA NYC Coastal Chi. Balt. W-S NY St. Paul Rockland

MESA and Air Pollution: Between-City Exposure Heterogeneity L=Low, M=Medium, H=High, VH=Very High e. g. , for PM 2. 5: L= (~12 g/m 3), M= (~16 g/m 3), H = (~20 g/m 3), VH= (~24 g/m 3) annual averages Blue = Areas of New Recruitment for MESA Air

Interim Long-Term Avg PM 2. 5 ( g/m 3) Estimates at Participant Homes 25 20 15 10 LA LA Riverside LA NYC Coastal Chi. Balt. W-S NY St. Paul Rockland

Outdoor Contributions to Indoor Air Homes Undergoing Indoor-Outdoor Monitoring CHI LA NY SP SUMMER WINTER Mean Indoor PM 2. 5 = 10. 8 ug/m 3 Mean Indoor PM 2. 5 = 8. 9 ug/m 3 79% 83% Mean Indoor PM 2. 5 = 15. 0 ug/m 3 Mean Indoor PM 2. 5 = 13. 6 ug/m 3 87% 81% Mean Indoor PM 2. 5 = 17. 2 ug/m 3 Mean Indoor PM 2. 5 = 15. 3 ug/m 3 82% 74% Mean Indoor PM 2. 5 = 7. 6 ug/m 3 Mean Indoor PM 2. 5 = 6. 0 ug/m 3 67% 75% *Indoor/outdoor monitoring only in non-smoking homes.

Finf Models N Summer 18 0. 63 – (0. 27 x Temp > 25) + (0. 27 x July Open Win) – (0. 15 x Dbl Pane) 0. 69 Winter 19 0. 51 – (0. 18 x Air Clnr) + (0. 10 x Jan Open Win) 0. 38 Summer 20 0. 13 + (0. 02 x Temp) + (0. 10 x Free Stndg) – (0. 10 x Air Clnr) + (0. 23 x Jul Open Win) 0. 83 Winter 41 0. 77 + (0. 07 x Free Stndg) – (0. 13 x Central AC) 0. 28 Summer 33 0. 85 – (0. 01 x Floor) + (0. 14 x Jul Open Win) – (0. 04 x Dbl Pane) 0. 48 Winter 34 0. 60 + (0. 01 x Temp) – (0. 10 x Above 3 rd Flr) + (0. 16 x Jan Open Win) 0. 43 Summer 17 0. 55 – (0. 24 x Air Clnr) + 0. 27 (Jul Open Win) 0. 59 Winter 24 0. 43 + (0. 02 x Temp) – (0. 17 x Air Clnr) 0. 59 CHI LA NY SP LOO = “Leave one out” Model Season R 2 City Season Overall R 2 “LOO” R 2 0. 71 0. 62 0. 54 0. 43 0. 67 0. 57 0. 60 0. 45

Reported Average Hrs/Day o Spent Outdoors (f ) Summer Winter

Ambient Outdoor PM 2. 5 Concentration Ambient Infiltrated PM 2. 5 Concentration Personal Exposure to Ambient PM 2. 5

PM 2. 5 Exposure Modeling Estimates: Dataset Timepoints Teleatlas Road Network Emissions Sources MESA AIR Census Road Network “Land Use” Variables AQS 2000 -06 MESA Air Fixed & Homes 2005 -06 2008 “Pragmatic” Interim Estimates UNIVERSITY OF MICHIGAN Nearest Monitor/ Distance to Nearest Road Long-Term PM 10 2006 AQS 1982 – 2002 Road Network Data “Land Use” Variables Meteorology Long-Term PM 2. 5 Traffic Counts Deterministic Model Outputs Measurement Error Considerations AQS 2000 -09 MESA Air Fixed & Homes 2005 -09 2009 Air Quality Model Integration AQS 1982 -2011 MESA Air Fixed & Home 2005 -09 Flexible Long-term Residential Histories 2011 Spatio-Temporal Interim Estimates Spatio-Temporal Final Estimates Infiltration & Time-Location Information Interim Exposure Predictions for Each Subject Final Exposure Predictions for Each Subject

Interim analyses of exposure and health o o o Microvascular abnormalities (Adar et al. ) Left ventricular mass (Van Hee et al. ) Serum and plasma biomarkers of endothelial activation (Van Hee et al. ) Systemic vascular response (Krishnan et al. ) Quantitative markers of emphysema (Barr et al. )

Health Model Building o Current focus on Agatston CAC o Model selection underway n n n Identification of confounders Treatment of cross-sectional and longitudinal components Coordinating with main MESA investigators

Exam 5 o o o Air components included in proposed schedule Working with Participant Relations Committee to ensure minimum participant burden Electronic Data Capture Spot Urine 5 Blood Pressure 15 ECG 30 Anthropometry 10 Phlebotomy 15 Demographics & Personal History 5 Psychosocial 10 Medical History & Medications 20 Diet Assessment 60 Air Questionnaire 20 Physical Activity Assessment 15 ABI 20 Cognitive Assessment 30 Sleep Questionnaire 5 Retinal Photography 35 Cardiac MRI (n=4000) 45 Cardiac CT (n=3600) 40 Cardiac Ultrasound (n = 3600) 50

Minimizing Participant (& Interviewer) Burden o Faster administration of Air Questionnaire with Electronic Data Capture n n o o Include shortcuts, like allowing participant to state that their activity patterns don’t change by season Reanalysis of actual time required for administration Reiterate to interviewers primary importance of retaining participants Review with interviewers the purpose of MESA Air, and provide a list of FAQs

Transition Away From the Paper MESA Air Questionnaire

EDC Interface - Air Questionnaire

Paper MESA Air Questionnaire – Time/Activity Info

EDC Interface Air Questionnaire Time/Activity Info

Exam 5 Ultrasound and CT/CAC Responsibilities Fall to MESA Air o Sub-contracts with Reading Centers n n n Imaging Protocols Training Data Transfer Protocols Reading Procedures QA/QC Processes

MESA Environmental Health Working Group o Meeting this evening 6: 30 pm n n n o Scientific opportunities Writing opportunities Funding opportunities 7: 15 pm Scientific Presentation n RG Barr et al, PM 2. 5 and Emphysema in MESA

Summary o On target with regards to schedule: n n n o o Baseline exam complete Exposure monitoring nearly complete “Pragmatic” estimates of outdoor concentrations at participants’ residences available First iterations of infiltration estimation and timelocation data integration underway Initial cross-sectional health analyses and health modeling underway Positive response from On-Site Peer Review Planning for smooth integration with MESA Exam 5

MESA Air o o o EPA Grant RD 831697 Project Period 8/1/04 – 7/31/13 Although the research described in this presentation has been funded wholly or in part by the United States Environmental Protection Agency through RD 831697 to the University of Washington, it has not been subjected to the Agency's required peer and policy review and therefore does not necessarily reflect the views of the Agency and no official endorsement should be inferred.