| AUTHORS | Melania Michetti (1) ,Maurizio Gualtieri (1) ,Alessandro Anav (2) ,Mario Adani (1) ,Barbara Benassi (2) ,Claudia Dalmastri (2) ,Ilaria D'elia (2) ,Antonio Piersanti (1) ,Gianmaria Sannino (2) ,Gabriele Zanini (1) ,Raffaella Uccelli (2) |
| ABSTRACT | High temperatures can cause excess mortality in exposed populations, particularly among the elderly. Temperature effects can be amplified further by the simultaneous presence of air pollution, which can act synergistically on the same target organs increasing the risks to human health. Mortality related to high and low temperatures and air pollution in EU will be significantly influenced by future climate and demographic changes. In existing multi-country and multi-city studies published on the association between temperature and mortality, little attention has been given to the combined influence exerted by temperature and air pollution, especially in Europe and particularly in perspective terms. Using nonlinear time series approaches with delayed effects (Distributed Lag Nonlinear Models, DLNM), by means of a Poisson model with overdispersion, we assess the combined short-term impact of temperature and air pollution on mortality from natural causes, for the two most populated cities in Italy, Rome and Milan. The short-term relative risk (RR) is estimated for the decade 2004-2015 and for the future (2050), assuming climate and air pollution scenarios consistent with the representative concentration pathways RCP2.6 and RCP8.5. For the exposure variables we used, for the temperature, the daily mean (T) and the daily apparent mean (AT), the SOMO35 indicator for ozone (sum of the surpluses from the threshold of 35 parts per billion of the maximum daily average on 8-h) and the maximum daily value for PM10. The RRs for mortality are estimated for the overall population and the most vulnerable age group (over 85 years of age; 85+), also defining the "minimum mortality temperature" (Tmm) for each city, conceived as an indicator of human adaptability to the local climate. For each of the exposure-response relationships considered (temperature-mortality, O3-mortality and PM10-mortality) a specific and different functional form is assumed after testing different models, controlling for seasonality and long-term effects. The results show that in the period 2004-2015 the two cities recorded a similar mortality pattern with no significant differences in terms of overall death numbers, in proportion to their reference population (about 1000 people per 100,000 inhabitants die every year from natural causes). On the other hand, the relative risk coefficients, the RRs, are higher in Milan than in Rome, both at warm and cold temperatures. The exposure-response associations for each variable – temperature, O3 and PM10 – show comparable patterns for both cities and a greater influence of temperature on mortality risk than that of pollutants. In particular, the number of attributable deaths is mainly associated with low temperature conditions, rather than temperatures higher than the optimum one (Tmm), which is lower in Rome and increases if we consider only the older age group. With regard to pollutants, their seasonal effect is appreciated: while PM10 affects the risk values mainly related to Milan, and above all, in association with the values of low temperatures, O3 - which affects both cities but with greater effects in Rome - concerns a wider temperature range, although it exerts its greatest influence especially in conjunction with high temperatures. Projecting the RRs to 2050, the results show that when standardizing deaths according to the reference population, the largest fractions of total attributable deaths are recorded in Milan. For both cities and climate scenarios, projections to the year 2050 do not show a significant difference in mortality load compared to historical values. However, they capture an increase in attributable mortality associated with warmer temperatures. Although cold mortality fractions (including the effects of air pollution) decrease in all cases, projected global warming in 2050 partially offsets the benefit of this reduction, increasing mortality fractions for temperatures above Tmm in both RCP scenarios, both cities and age categories. This generalized 'substitution effect' between hot and cold mortality is most evident in the RCP8.5 scenario and for the 85+ age group. In conclusion, the results suggest that a more rigorous and internationally coordinated climate policy, implying more stringent emission scenarios, can lead to significant co-benefits in terms of reducing the future health burden, reducing air pollution, achieving climate objectives, containing global warming below the 2 °C recently recalled at COP26 in Glasgow. Indeed, when the effects of climate change and air pollution are contained with climate policy action (RCP2.6), the number of fatalities (absolute cases) could decrease by 8 times in Rome and by 1.4 times in Milan, compared to the historical values. |