| Tweet | Follow @co2science |
Paper Reviewed
Allen, M.J and Sheridan, S.C. 2018. Mortality risks during extreme temperature events (ETEs) using a distributed lag non-linear model. International Journal of Biometeorology 62: 57-67.
Much research has examined the impacts of both hot and cold weather events on human mortality (see, for example, the many reviews we have posted on this topic under the heading Health Effects (Tempeature: Hot vs Cold Weather) on this page). One of the most recent of such studies was that of Allen and Sheridan (2018).
Publishing their work in the International Journal of Biometeorology, the two scientists investigated the relationship between all-cause mortality and extreme temperature events for 50 metropolitan areas (with populations from 1 to 19.5 million) in the United States for the period 1975 to 2004. More specifically, they used a distributed lag non-linear model to assess mortality impacts in response to 3-day means of daily apparent temperature (apparent temperature incorporates air temperature, vapor pressure and wind speed into a single variable) for a cumulative 14-day period following the hot and cold events. Additionally, they subdivided their calculations to discern such effects from both heat and cold events (95th and 5th percentiles) and extreme heat and cold events (97.5th and 2.5th percentiles) that occurred both early and late in the summer and winter seasons, respectively. They also determined the impact of the length of heat and cold spells (short = events lasting 2 days or less, long = events lasting for 3 or more days) on mortality.
Averaged for the 50 locations they studied, the mean cumulative relative risks for heat, extreme heat, cold and extreme cold are shown in the table below, based on the duration (short or long) and time (early or late in the season) of the event. As indicated there, the highest relative risk of mortality is for extreme cold events that occur early in the winter season and which last for 2 days or less, which risk is double that observed for such events that occur in the summer. Thus, cold weather is far more deadly than hot weather.
Another interesting observation is that the cumulative relative risk of mortality values from both short and long temperature events decline over the course of the winter and summer seasons, which declines suggest there is a seasonal human adaptation to extreme weather events occurring at both ends of the temperature spectrum.
