Ambient temperature and deaths from homicide in Brazil during 2010-2019: A nationwide space-time-stratified case-crossover study.

Research question: Previous evidence suggests a positive association between temperature and homicide, but the association was less clear in Brazil where homicide is one of the leading causes of death. This study aimed to quantify the association between ambient daily temperature and homicides in Brazil with potential lag effects and to quantify the temperature attributed fractions of homicides in Brazil.

Methods: A space-time-stratified case-crossover design with distributed lag models was used to evaluate the temperature-homicide association from 1·1·2010 to 31·12·2019 in Brazil. The odds ratios (OR), attributable fractions and their confidence intervals (CI) were calculated.

Results: Overall every 5°C increase in daily mean temperature was associated with a 10·6% (OR=1·106, 95% CI: 1·085-1·127) increase in homicidal deaths at lag 0-8 days. The temperature-homicide association is stronger for females and elderly, homicides by fights, sharp objects or firearm, and in North region. During the study period, 1·8% (95% CI: 1·1%-2·7%) of homicides could be attributed to temperature above immediate-region-specific median temperature corresponding to 10,921 additional deaths (95% CI: 6,350-15,372).

Conclusion: Our nationwide study suggests that the homicides in Brazil may increase with temperature and recommends targeted preventions for certain risk groups to high temperature, considering future climate change circumstances.