By Indur M. Goklany
It is well known that the risk of mortality increases at both the high and low ends of the temperature range experienced by a particular population.,, Therefore, there should be a temperature at which that population’s risk of mortality is at a minimum. [There, however, may be more than one “local” minimum in a graph of mortality risk versus temperature as one proceeds from the lowest to highest temperatures.]
Recently, Guo et al. (2014) undertook a systematic evaluation of the variation in the risk of mortality from non-accidental causes as a function of daily mean temperature in 12 countries. The figures below display their results. They used mortality data for multiple years (ranging from 10 years for Thailand to 38 years for Japan) for 306 communities in the 12 countries, and pooled the data for the communities in each country to derive these figures.
Note that the temperature on the x-axis for each graph is measured in terms of the percentile of the temperature range rather than the actual temperature (in °F or °C). Also, their methodology was designed to account for deaths that occurred over the following 21 days, since additional deaths from exposure to hot or cold temperatures are known to occur for several days subsequent to actual exposure. [The period over which these deaths occur is longer for cold temperatures than for hot.] Their methodology also apparently accounted for “mortality displacement” or “harvesting,” which is the concept that temperature-related deaths that occur in a vulnerable population immediately following the temperature exposure would be partially offset by fewer deaths in that population over the following weeks.
These graphs show that:
· The relative mortality risk for each country is at a minimum between the 66th and 80th percentile of mean temperature. Nine of the twelve countries have an “optimum” temperature between the 72nd and 76th percentiles.
· For each country the relative mortality risk is substantially higher at the 1 percentile temperature (cold end) than at the 99th percentile (hot end).
· Remarkably, the above bullet points hold not only for relatively cold countries such as Canada and South Korea but also the relatively warm ones such as Brazil and Thailand.
The study also reports that, “The minimum-mortality temperatures were higher in countries with high temperature or in countries close to equator.”
What all this means is that, first, because (a) there are more days during the year that are cooler than the optimum, and (b) relative risk is higher at the cold end than the warm end, more deaths should be associated with temperatures that are colder than optimum than those that are warmer. Hence, if global warming merely slides each curve to the right wholesale, total mortality during the year should drop. But, in fact, global warming is supposed to warm winters more than summers — even so-called Skeptical Science acknowledges this! Therefore, we should get a double dividend from global warming in terms of reduced global mortality.
Figure 1: Relative risk of mortality (y-axis) as a function of mean daily temperature plotted as the percentile of the entire temperature data. Data for each country was pooled. Source Guo et al. (2014).
In summary, there is an optimum temperature which minimizes mortality for any given population, and it is toward the warmer end of what that population generally experiences. Specifically, it is at about the 70th–75th of the mean temperature to which that population is exposed. Finally, if there is any doubt about it, there is a good health-based rationale for:
· The general preference for warm temperatures,
· For taking winter vacations in warm places and summer vacations in cold places,
· For retiring to warmer climes!
 McMichael, Anthony J., et al. “International study of temperature, heat and urban mortality: the ‘ISOTHURM’project.” International journal of epidemiology 37.5 (2008): 1121-1131.
 Keatinge, W. R. “Winter mortality and its causes.” International Journal of Circumpolar Health 61.4 (2002).
 Guo, Yuming, et al. “Global variation in the effects of ambient temperature on mortality: a systematic evaluation.” Epidemiology 25.6 (2014): 781-789.
 Deschenes, Olivier. “Temperature, human health, and adaptation: A review of the empirical literature.” Energy Economics 46 (2014): 606-619.
Note: an earlier version of this essay rfereenced the”y-axis” corrrected to “x-axis”. h/t to “joelobryan on March 2, 2015 at 2:49 am.”