1. Heat-Related Mortality in Washington State: Past and Future The Washington Climate Change Impacts Assessment Conference February 12, 2009 J. Elizabeth Jackson University of Washington

2. Heat: Core Public Health Concern  Heat waves linked with hundreds of deaths in the United States annually Chicago, 1995: estimated 700 deaths Europe, 2003: estimated 30,000 deaths UK, 2003: estimated 2,000 deaths

3. Heat: Core Public Health Concern  Heat waves linked with hundreds of deaths in the United States annually  Heat waves projected to increase in frequency, duration and intensity

4. Heat: Core Public Health Concern  Heat waves linked with hundreds of deaths in the United States annually  Heat waves projected to increase in frequency, duration and intensity  Heat-related mortality is likely under- reported

5. Heat: Core Public Health Concern  Heat waves linked with hundreds of deaths in the United States annually  Heat waves projected to increase in frequency, duration and intensity  Heat-related mortality is likely under- reported  What about Washington State?

6. Thermal Stress  Hyperthermia: the body cannot dissipate heat absorbed from the environment (CDC, 2005)

7. Thermal Stress  Hyperthermia: the body cannot dissipate heat absorbed from the environment (CDC, 2005)  Vulnerable Groups Elderly Urban residents Those with chronic or mental illnesses The poor, the socially isolated Outdoor laborers

8. Other Vulnerabilities Major cities of Washington State are particularly at risk for high mortality during heat waves Milder summers = less adaptation to heat Little residential air conditioning Heat Island Effect

9. Analytic Goals  Establish historical relationship between heat events and mortality (1980-2006)

10. Analytic Goals  Establish historical relationship between heat events and mortality (1980-2006)  Estimate future mortality due to heat, i.e., “excess deaths” (2025, 2045, 2085)

11. Analytic Goals  Establish historical relationship between heat events and mortality (1980-2006)  Estimate future mortality due to heat, i.e., “excess deaths” (2025, 2045, 2085)  Customize estimates for 4 study areas: Greater Seattle Area (King, Pierce, Snohomish) Spokane Tri-cities (Benton, Franklin) Yakima

12. Measuring Heat Events  Humidex: combined effects of heat and humidity  Heat event threshold: hottest 1% of all days (99 th percentile humidex)  Heat Events: Counted the number of heat events and the duration of each heat event  Three climate change scenarios: high, moderate and low summer warming

13. Measuring Mortality  Outcome: daily mortality rate, May- September (= deaths/population)  Age groups: 45 and older; 65 and older; 85 and older  Causes of death All non-traumatic causes Circulatory, Cardiovascular, Respiratory (not presented)

14. Method: Historical Relationship Relative Risk of death during a heat event = Mean Daily Mortality Rate (heat event) Mean Daily Mortality Rate (non-event) If RR > 1, then risk of death is greater during heat events

15. Method: Future Excess Mortality Excess deaths during future heat events are calculated from: Baseline Mortality Rate Risk of death during heat event Future population Climate change Mean Daily Mortality Rate (non-event) *(Relative Risk – 1)* Projected Population*Projected Heat Events

16. Baseline Population Parameters Greater Seattle AreaEastern Washington

17. Baseline Climate Parameters Heat Events, 1980-2006

18. Relative Risk (non-traumatic), Seattle

19. Relative Risk (non-traumatic), East

20. Projected Climate Parameters Number of Heat Events Greater SeattleEastside High Moderate Low

21. Projected Climate Parameters Average Duration of Heat Events Greater SeattleEastside High Moderate Low

22. Projected Excess Deaths, Seattle † † Population held constant at 2025 projection

23. Projected Excess Deaths, East † † Population held constant at 2025 projection

24. Discussion  Why the difference between East and West?

25. Discussion  Why the difference between East and West?  Our estimates of excess deaths due to future heat are conservative Population growth not factored in Analysis of past heat events may underestimate effect of longer, hotter events

26. Discussion  Why the difference between East and West?  Our estimates of excess deaths due to heat are conservative Population growth not factored in Analysis of past heat events may underestimate effect of longer, hotter events  Consider the possibility of higher-order failures

27. Limitations  Use of county as geographic level linking heat events and mortality  Reliability of climate and population projections  Change in cause-of-death coding during historical study period  Method does not allow for analysis of smaller, dispersed populations

28. Conclusions  Heat Stress is already a significant factor in Washington mortality  The number of heat-related deaths will increase because of climate change

29. Acknowledgements Co-Authors Michael G. Yost PhD, UW Env & Occ Health Sciences Catherine Karr MD PhD, UW Env & Occ Health Sciences, Pediatrics Cole Fitzpatrick MA, UW Env & Occ Health Sciences Brian K. Lamb PhD, WSU Lab for Atmos Research, Civil & Env Engineering Serena H. Chung PhD, WSU Lab for Atmos Research, Civil & Env Engineering Jack Chen PhD, National Research Council Canada, Ottawa, ON, Canada Jeremy Avise PhD, California Air Resources Board, Sacramento, CA Roger A. Rosenblatt MD, UW Family Medicine Richard A. Fenske PhD, UW Env & Occ Health Sciences CIG/JISAO Phil Mote PhD Eric Salathe PhD Alan Hamlet PhD Marketa McGuire Elser PhD Lara Whitely Binder