Table 1

Recent Studies of High Ambient Temperature and All-Cause Daily Mortality*
Reference
 Study population
 Method
 Exposure
 Result: effect estimate (95% CI)
Baccini 2008 [12]
 15 European cities, April-September 1990-2000 (5-11 years depending on data availability for city)
 Time-series;
% change
 Maximum apparent temperature (threshold 29.4°C Mediterranean cities and 23.3°C north-continental cities)
 1°C increase above threshold 3.12 (0.60-5.72) in Mediterranean and 1.84 (0.06-3.64) in north-continental region
Lag: 3 days prior
Basu 2008 [6]
 9 California counties, May to September 1999-2003
 Time-series and case-crossover;
% change
 Daily apparent temperature (minimum, mean, maximum); daily mean O3, PM2.5, PM10, NO2, CO, SO2
 Per 10°F increase mean temperature, 2.3 (1.0-3.6), similar results for minimum and maximum temperatures
Lag: 0
Bell 2008 [15]
 Sao Paulo, Brazil, Santiago, Chile and Mexico City, Mexico, 1998-2002
 Case-crossover; % change
 Same day apparent temperature compared with days at 75th percentile, O3, PM10
 2.69 (-2.06, 7.88) for Santiago, 6.51% (3.57, 9.52) for Sao Paulo and 3.22% (0.93, 5.57) for Mexico City
Lag: 0
McMichael 2008 [45]
 Delhi, Monterrey, Mexico City, Chiang Mai, Bangkok, Salvador, Sao Paulo, Santiago, Cape Town, Ljubljana, Bucharest, Sofia, 2 to 5-year series (1991-1999)
 Time-series; % change
 Daily maximum threshold (16°C-31°C) temperature, relative humidity, precipitation data, PM10, BS, or TSP
 1°C increase above threshold increased death rates with increasing heat in all cities: (ranging from 0.77-18.8) except Chiang Mai 2.39 (-0.49-5.35) and Cape Town 0.47 (-0.31-1.24)
Lag: 2-day average
Vaneckova 2008a [46]
 Sydney, Australia, October to March 1993-2001
 Time-series; ratio of highest 10% mortality days within air mass and % frequency of air mass occurrence
 Temporal Synoptic Index (TSI)
 1.64 and 2.64 (both significant)for warmest TSIs, no CI provided
Zanobetti and Schwartz 2008 [7]
 9 U.S. counties, May to September 1999-2002
 Time-series and case-crossover;
% change
 Daily apparent temperature (minimum, mean, maximum); daily mean O3, PM2.5, PM10
 Per 10°F increase mean temperature, 1.8 (1.09-2.5) case-crossover and 2.7 (2.0-3.5) time-series; similar results for minimum and maximum temperatures
Lag: 0
Barnett 2007 [47]
 107 U.S cities using data from the National Morbidity and Mortality Study, 1987-2000
 Case-crossover;
% change
 Daily temperature
 Per 10°F, summer 1987 average increase in cardiovascular deaths was 4.7 (3.0-6.5). By summer 2000, the risk with higher temperature had disappeared (-0.4, -3.2-2.5)
Lag: 04
Medina-Ramon 2007 [21]
 50 US cities in cold (November to March) and warm (May to September) seasons
 Case-crossover;
% change
 Binary variable as extreme heat (range 22-32°C)
and continuous; O3
 5.74 (3.38-8.15) for extreme heat
Lag: 2-day average
Kolb 2007 [32]
 Montreal, Canada 1984-1993
 Case-crossover;
odds ratio
 Mean daily and maximum temperature, barometric pressure, relative humidity, adjusted for O3 and both NO2 and O3
 1.20 (1.14-1.38) for 25-30°C maximum temperature; strong nonlinear association with a threshold at 25°C
Lag: average 02; no association after 3 days
Carson 2006 [48]
 London, England,
4 time periods, winter: December-March; non-winter: April-November
 Time-series; ratio of winter to non-winter deaths
 Daily mean temperature
 1.24 (1.16-1.34) from 1900-10,; 1.54 (1.42, 1.68) from 1927-37, 1.48 (1.35,-1.64) from 1954-64, 1.22 (1.13-1.31) from 1986-96; heat deaths diminished overall in the century
Kim 2006 [40]
 6 cities in South Korea, summer 1994-2006
 Time-series;
% change
 Daily mean temperature thresholds
(27-29.7°C)
 1°C above threshold 16.3 (14.2, 18.4), 9.10 (5.12, 13.2), 7.01 (4.42, 9.66), 6.73 (2.47, 11.2) for Seoul, Daegu, Incheon and Gwangiu, respectively
Michelozzi 2006 [49]
 4 Italian cities, June to September 2003 & 2004 and reference period (Roma, Torino, Milano: 1995-2002 and Bologna:
1996-2002)
 Time-series;
% change
 Daily maximum apparent temperature thresholds (28-32°C)
 1°C above threshold 3.2 (1.9-4.6), 5.0 (3.8-6.1), 5.4 (4.3-6.5), 3.8 (2.5-5.0) for Bologna, Milano, Roma, and Torino, respectively
Stafoggia 2006 [16]
 Bologna, Milan, Rome, Turin, 1997-2003
 Case-crossover; odds ratio
 30°C mean apparent temperature relative to 20°C; odds ratio
 1.34 (1.27, 1.42)
Lag: 01
Basu 2005 [5]
 20 US metropolitan areas, seasonal analysis 1992
 Time series (relative risk) and case-crossover (odds ratio)
 Mean daily temperature per 10F adjusted for dew point temperature; daily O3
 Per 10°F, 1.15 (1.07-1.24), 1.10 (0.96-1.27), 1.08 (0.92-1.26), 1.08 (1.02-1.15), and 1.01 (0.92-1.11) in the Southwest, Southeast, Northwest, Northeast, and Midwest, respectively, in the summer from the time-stratified case-crossover
Lag: 0,1
El-Zein 2004 [34]
 Greater Beirut, Lebanon, 1997-1999
 Time-series;
% change
 Mean daily temperature, mean daily humidity, minimum mortality temperature (TMM) = 27.5°C
 1°C above TMM 12.3 (5.7, 19.4%) increase in annual mortality
Lag: 0
Goodman 2004 [26]
 Dublin, Ireland, April 1980 to December 1996
 Time-series;
% change
 Daily minimum temperature, daily mean relative humidity
 1°C increase 0.4 (0.3-0.6) increase
Lag: 0
Pattenden 2003 [50]
 Sofia, Bulgaria (1996-1999) and London, England (1993-1996)
 Time-series;
% change
 Daily mean temperature, relative humidity and PM (black smoke for London and total suspended particulates for Sofia)
 1°C increase above 90th % 1.9 (1.4 to 2.4) in London, and 3.5 (2.2 to 4.8) in Sofia
Lag: 2 day average
Curriero 2002 [39]
 11 Eastern US cities, 1973-1994
 Time-series;
% change
 Daily mean temperature, dew point temperature; minimum mortality temperature (MMT) range: 65.2-90.3
 Per 10°F above MMT range 1.4-6.7
Lag: 0
Braga 2001 [11]
 12 US cities. 1986-1993
 Time-series;
% increase
 Mean daily temperature, relative humidity
 4% increase (no CI given);
Lag: 0 or 1
Harvesting effect for hot temperatures

* Exceptions: El-Zein 2004 and Carson 2006 reported annual and weekly deaths, respectively.

Basu Environmental Health 2009 8:40   doi:10.1186/1476-069X-8-40

Open Data