A global attribution study on historical heat-related mortality impacts attributed to climate change.

On behalf of the Multi-Country Multi-City Collaborative (MCC) Research Network.

Background & Aim: Climate change is considered the most important environmental threat to human health. Substantial mortality and morbidity burden have been directly or indirectly attributed to climate-sensitive environmental stressors. However, limited quantitative evidence exists on how much of this burden can be attributed to man-made influences on climate. In this large health attribution study, we aimed at quantifying the proportion of excess heat-related mortality attributed to anthropogenic climate change in recent decades across 626 locations across 41 countries in various regions of the world included in MCC database.

Methods: We first estimated the location-specific heat-mortality associations through two-stage time-series analyses with quasi-Poisson regression with distributed lag non-linear models and multivariate multilevel meta-regression using observed data. We then quantified the heat-related excess mortality in each location using daily modelled series derived from historical (factual) and preindustrial control (counterfactual) simulations from 5 general circulation models (ISIMIP2b database) in the period between 1991 and 2019. We finally computed the proportion of heat-related excess mortality attributable to anthropogenic influences as the difference between the two scenarios, with associated measures of uncertainty.

Results: We found a steep increase in level of warming, expressed as the difference in annual average temperature between scenarios, with an average increase of 1.0°C (from 0.7°C  to 1.2°C) across the 626 locations between 1991 and 2019. Overall excess heat-mortality fractions of 1.92% [95% confidence interval: 0.41, 3.25] and 1.28% [0.20, 2.50] were estimated under the factual and counterfactual scenarios, respectively, with an overall difference of 0.76% [0.25,1.74]. This translates to 33% of historical heat-excess mortality that can be attributed to anthropogenic climate change. Larger proportions were found in North America (46%), Central America (47%), South America (43%), South Africa (48%), Middle-East Asia (61%), South East-Asia (50%) and Australia (42%), although highly imprecise in most of cases.

Conclusions: Our findings suggest that current warming driven by anthropogenic influences is already responsible for a considerable proportion of the heat-related mortality burden. These results stress the importance of strengthening current mitigation strategies to reduce further warming of the planet and related health impacts.