The Johns Hopkins Gazette: April 1, 2002
April 1, 2002
VOL. 31, NO. 28


Public Health Study Explores the Effect of Temperature on Mortality

By Ming Tai
Bloomberg School of Public Health

Johns Hopkins Gazette Online Edition

The relationship between extreme temperature and mortality in the United States varies by location, according to a study from researchers at the Bloomberg School of Public Health. Excessive heat is more likely to increase mortality risk in the North, while excessive cold is more likely to increase mortality risk in the South. The study appeared in the Jan. 1 issue of the American Journal of Epidemiology.

“Historically, we know that episodes of extremely hot or cold temperatures increase mortality. Global warming and other weather phenomena, such as El Nino, have warranted the need for further investigation into the weather-mortality relation,” says lead author Frank Curriero, an assistant scientist in the Department of Biostatistics. “We explored this relation in cities along the eastern United States to further characterize the effects of temperature as varied by latitudes as well as other factors particular to a specific city or area.”

The researchers chose 11 large metropolitan areas in the eastern United States and compared daily weather and mortality data for 1973 to 1994. The analysis was performed in two stages. First, the relationship between temperature and mortality risk from cardiovascular, respiratory and other diseases was estimated for each city, then the variation in risk across cities, due to difference in latitude and other variables, was examined.

Within the selected cities, mortality risk decreased as temperature increased from the coldest days; however, after a certain critical temperature threshold, referred to as the minimum mortality temperature, mortality risk increased in most of the cities as temperature increased. The minimum mortality temperature threshold ranged between approximately 65 and 70 degrees Fahrenheit for cities in the North (Boston, Chicago, New York City, Philadelphia, Baltimore and Washington, D.C.) and approximately 76 and 90 degrees Fahrenheit for cities in the South (Charlotte, Atlanta, Jacksonville, Tampa and Miami).

Results indicate the northern cities were more sensitive to heat, showing a larger increase in mortality risk for temperatures greater than the minimum mortality temperature threshold. Southern cities were more vulnerable to cold temperatures, showing a larger increase in mortality risk for temperatures less than the minimum mortality temperature threshold.

Colder temperatures produced higher mortality risk in cities with higher proportions of elderly, but not in cities with a large fraction of heating systems. Hotter temperatures produced higher mortality risk in cities with larger percentages of persons without a high school education and those living in poverty, two indicators of socioeconomic status. A higher percentage of homes with air-conditioning was associated with small but significant effects of hot temperature on mortality.

“Temperature at either extreme can be hazardous,” says co-author Jonathan Patz, an assistant professor in the Department of Environmental Health Sciences. “The results of our study, however, will help guide public health interventions aimed at preventing deaths due to extreme heat and cold. These findings tell us where and to whom such interventions should be directed.”

The models developed in this analysis also may be useful in estimating weather-related mortality as a consequence of future climate-change scenarios. Researchers at the school are now applying these methods to a larger number of U.S. cities to improve upon their predictive capability, according to Jonathan Samet, chair of the Department of Epidemiology, and Scott Zeger, chair of the Department of Biostatistics.

This study was supported in part by the Climate Policy and Assessment Division of the United States Environmental Protection Agency.