https://orcid.org/0000-0002-0164-1112
-
Views
-
Cite
Cite
Kai Chen, Annette Peters, Alexandra Schneider, KORA Study Group , Burden of myocardial infarctions attributable to heat and cold, European Heart Journal, Volume 40, Issue 41, 1 November 2019, Pages 3440–3441, https://doi.org/10.1093/eurheartj/ehz612
Close - Share Icon Share
Extract
This commentary refers to ‘Changing trends and public health relevance of myocardial infarctions attributable to cold and heat’ by V. Čulić, 2019;40:3438--3439.
We thank for Dr Čulić’s interest in our article,1 and his comment on the public health relevance of our findings in terms of myocardial infarctions (MIs) attributable to cold and heat.2 We appreciate Dr Čulić’s endeavour of calculating the population attributable risks (PAR) based on the risk estimates of our study. However, it should be noted that the risk estimates for heat or cold effects were only point estimates for temperatures at the 97.5th percentile (heat) or 2.5th percentile (cold) relative to the minimum MI temperature (i.e. optimum temperature). Thus, the PAR for temperatures above the 97.5th percentile or below the 2.5th percentile only account for the MI burden attributable to extreme heat or extreme cold.
Temperature and health generally showed U-shaped associations, with increasing risks for both temperatures above and below the optimum temperature (see Figure 1 in our article2). A previous multicountry study showed that mortality risk attributable to extreme temperatures (both extreme heat and extreme cold; 0.86%) was substantially less than that attributable to milder but non-optimum temperatures (6.85%).3 To gain a complete picture of the attributable burden, we estimated the number and fraction of MI cases attributable to heat (above the optimum temperature), cold (below the optimum temperature), and total temperature (net change, the summed impacts of heat and cold) in Augsburg, Germany using a recently developed approach.3 , 4 Briefly, we applied the risk estimates corresponding to each day’s temperature in the previously estimated exposure-response function for total MI,2 and then used the daily series of temperature and MI to calculate the daily attributable number of MI cases. We then calculated the total attributable number by summing the contributions from all the days of the series and calculated the attributable fraction as the ratio of the total attributable number to the total number of MI cases.
