A discussion about the climate change impacts on elderly mortality rates in the metropolitan region of São Paulo – Brazil: what to expect?

As consequence of the age-structural transitions observed in Brazil over the past years the elderly population (age ≥ 60 years) of the metropolitan region of São Paulo has increased from 1.2 million in 1996 to 2.1 millions in 2011. One of the major concerns about global warming is related to the impacts on vulnerable populations living in urban areas. The all-cause mortality rate of elderly people in São Paulo presents a seasonal pattern shaped by climate variations where the highest mortality rates are observed in winter and the lowest in summer. Since previous studies involving regional climate model projections for this region points to a warming trend scenario the following question arise: The expected global climate change impacts on the elderly population are going to be exclusively bad? Or there may be something positive about it? A preliminary data analysis indicate that all-cause mortality rates for both elderly men and women are decreasing while the mean annual temperatures are increasing. It is also notice that this mortality decrease is more pronounced in winter months. An observed/expected analysis of the daily mortality rate was also conducted from 1996 to 2011 period. Mortality data was associated with maximum, mean and minimum temperatures (Tmax, Tmea and Tmin respectively) obtained from a meteorological weather station. It is shown that excessive deaths per day significantly increases when Tmax ≥ 30ºC, Tmea ≥ 24ºC and when 14ºC > Tmin ≥ 21ºC. On the other hand a decrease in mortality rate is perceived when Tmax < 15ºC, Tmea < 11ºC and when 14ºC ≤ Tmin < 20ºC. It is also shown that the deficit or excess in daily mortality rates are mainly conditioned by low frequency temperature intervals, i.e., those represented by temperatures that are not common to be recorded in this region. These are key intervals regarding the impacts of climate change over the elderly people.