Abstracts: ISEE 21st Annual Conference, Dublin, Ireland, August 25-29, 2009: Oral Presentations
Background and Objective:
Hot and cold temperatures significantly increase the risk of death in many regions of the world. Different measures of temperature, including minimum, maximum and apparent temperature, have been used in previous research. Which temperature measure is the best predictor of mortality is not known.
We used mortality data from 106 cities in the US NMMAPS study (years 1987-2000). We examined the association between temperature and mortality using Poisson regression and fitted a non-linear spline for temperature. We examined five measures of temperature, the effect of including relative humidity, and various degrees of freedom for the temperature spline. The best model was defined as that with the minimum absolute residual. The residuals were calculated using cross-validation.
Maximum temperature was selected as the best temperature measure the most often (40 cities in the ≥65-year age group), and apparent temperature the least often (8 cities in the <65-year age group). Maximum temperature was the best measure in 10 out of 12 months in both age groups. Geographically, maximum temperature was the best measure in cold regions, and minimum temperature in warm regions. Humidity was important in almost every city in the ≥65 year age group. The seasonal variation in humidity showed a surprising peak in usefulness in winter.
Apparent temperature is no better than standard measures of temperature in predicting mortality. Maximum temperature was generally the best measure in cold climates and minimum temperature in warm climates. Humidity is an important predictor of mortality in the elderly and its effect should be estimated separately from temperature.