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.[1],[2],[3] 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.[4]
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!
[1] 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.
[2] Keatinge, W. R. “Winter mortality and its causes.” International Journal of Circumpolar Health 61.4 (2002).
[3] Guo, Yuming, et al. “Global variation in the effects of ambient temperature on mortality: a systematic evaluation.” Epidemiology 25.6 (2014): 781-789.
[4] 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.”
JKnapp at March 2, 2015 at 1:50 am
RESPONSE:
1. According to the Guo paper, “minimum mortality” tends to occur at higher temperatures in warmer areas.
2. I am not assuming that the curve would change shape, although I would expect it would because of acclimation and adaptation.
Jimbo March 2, 2015 at 5:48 am
RESPONSE: Agree on both counts. Check out, e.g., Have increases in population, affluence and technology worsened human and environmental well-being?. and Humanity Unbound: How Fossil Fuels Saved Humanity from Nature and Nature from Humanity
theBuckWheat March 2, 2015 at 7:32 am
RESPONSE: That’s what benefits-cost analysis ought to be doing in theory, but I am skeptical that it’s done properly, but that’s another matter.
Dudley Horscroft March 2, 2015 at 8:42 am
RESPONSE: There does seem to be a local minimum at ~ 25% for some countries, but I think it is more useful to think of this as a minor max at ~50%. This might be because temperature by itself is too one dimensional a measure. Humidity could be important too, and changes in humidity may be responsible for this minor max.
famadore March 2, 2015 at 9:01 am
RESPONSE:
1. Not quite since there is a increase in risk is higher at the cold end than the warm end.
2. (a) Don’t forget Brazil, and Australia and China have many areas that are quite warm. (b) Part of the problem is data availability. Many of the warm weather countries are also poorer and less well-developed, and probably don’t keep statistics as meticulously as the richer countries.
Barry March 2, 2015 at 9:50 am
RESPONSE: Indeed this posting was on a pretty narrow topic. Since you are keen on a broader view of the human condition (and a systems approach), check out Humanity Unbound: How Fossil Fuels Saved Humanity from Nature and Nature from Humanity and The Improving State of the World: Why We’re Living Longer, Healthier, More Comfortable Lives on a Cleaner Planet
Chip Knappenberger March 2, 2015 at 11:47 am
RESPONSE: I agree that the former doesn’t follow from the latter. The “if” is just there to signal that an assumption is about to be made. Perhaps I should have said, “If ONE ASSUMES THAT global warming…” Nor do I believe that this is likely. In fact, I agree that the mortality-temperature curve would change because of acclimation and adaptation. The curve would, moreover, be different for different populations because social, economic and other factors (e.g., wealth, nutrition and availability of food, etc.) vary from population to population.
Mortality lower during harvest season…
Possible this has to do with food availability?
For this there had to be a study?
Who paid for it, and why?
The comfortable (healthy and happy) temperature is whatever a person would set his home thermostat on if the heat was free — lets say 70 degrees F.in the winter and 75 degrees F. in the summer based on my own preferences. Now let’s move on to study something important. .
How would you adjust the temperature numbers for effects of humidity and wind speed — both are obviously important factors in addition to temperature.
Any doctor would have told you heart attacks are much more common in the winter.
Consider a man over 50 years old who has to drive to a job he doesn’t like on Monday morning but he is running late and his car is stuck in a pile of snow in his driveway — he’d be better off calling in sick and going back to bed, because everyone “knows” that’s heart attack time.
I have read that a nude person outdoors can not survive a full day without clothing and a fire if the temperature is below 70 degrees F. and there is at least a small wind — although I have not tried it — don’t want to scare the neighbors — we are a tropical species that tolerates hot weather much better than cold weather,