An Economic Analysis of the Environmental Impact of PM2.5 Exposure on Health Status in Three Northwestern Mexican Cities

36 Background: This study provides an assessment of the health effects due to exposure to 37 particulate matter PM 2.5 (diameter < 2.5 µm that cause the air to appear hazy when levels are 38 elevated) in three medium size cities of north western Mexico (Los Mochis, Culiacan and 39 Mazatlan). We computed the total avoidable premature mortality, avoidable cardiovascular 40 disease, gains in life expectancy, as well as the economic costs by air contamination corresponding 41 to PM 2.5 . To achieve this goal, the Health Impacts Assessment (HIA) methodology provided by 42 the European Aphekom Project was used. People in these cities are exposed to high PM 2.5 43 concentrations that exceed limits implemented in domestic and international guidelines. 44 Results: We determined the cost of the PM 2.5 pollutant associated with health outcomes under two 45 different scenarios: Official Mexican Standard (NOM, Spanish acronym) and World Health 46 Organization (WHO). The mean PM 2.5 concentrations in 2017 were 22.8, 22.4 y 14.1 µg/m 3 for 47 the cities Los Mochis, Mazatlan and Culiacan, respectively. The mean avoidable mortality for all 48 causes associated to the exposure to PM 2.5 in the three cities was 638 for the NOM scenario (i.e., 49 with a reduction to 12 µg/m 3 ) compared to 739 for the WHO scenario (reduction to 10 µg/m 3 ). 50 Complying with the WHO guideline of 10 µg/m 3 in annual PM 2.5 mean would add up to 15 months 51 of life expectancy at age 30, depending on the city. 52 Conclusions: The mean economic cost per year of the PM 2.5 effects on human life in these three 53 cities was $600 (NOM scenario) and $695 million dollars (WHO scenario). Effective public health 54 and industrial policy interventions are socially advantageous and cost-saving to promote better health.

Atmospheric emissions records were obtained from the State of Sinaloa's environmental 163 monitoring system, which has a fixed station in each of the selected cities measuring air quality 164 and in particular PM2.5. To measure these PM2.5, this monitoring system uses the Tapered Element 165 Oscillating Microbalances technology (TEOM), which is a widely used method to collect data on 166 10), selecting code A00-R99 to classify total mortality (external included) and code I00-I99 to 176 classify cardiovascular diseases only. Additionally, population data was retrieved from the 2010 177 National Population and Housing Census and updated to 2017 using the population growth rates 178 estimates by the Mexico's National Council for Population [14]. 179

The Concentration-Response Function (CRF) 180
The relationship between the concentration of air pollutants with total mortality and cardiovascular 181 diseases helps quantify public health risks. Our sample of cities were chosen on the basis of CRF's 182 data availability according to the American Cancer Society (ACS) [56], the largest research 183 assessing the effects of air pollution on human health. Our CRF's metrics are based on the long-184 run average PM2.5 concentrations recorded from these sites. We also used the Harvard Six Cities 9 cohort re-analysis [46] as well as the ACS follow up and space analysis, which relates 186 environmental pollution by particulate matter to mortality rates [45]. These studies provide reliable 187 estimation of uncertainty and decrease the variability among the exposed populations regarding 188 individual susceptibility, PM2.5 composition, minimum and maximum pollution concentrations 189 and exposition frequency. Pope  Where y represents the change in avoidable premature deaths in the health result associated with 202 the decrease in pollution concentrations at the baseline; yo is the base health outcome (deaths 203 according to ICD-10); e is the Napier constant; x is the change of pollution concentrations in a 204 given scenario; is the concentration-response function (CRF) coefficient, which is calculated 10 Where CRF is the concentration-response function; ln (RR) is the natural logarithm of the relative 208 risk (RR), which was selected from previous epidemiological studies to estimate the long-term 209 effects of PM2.5 on overall and cardiovascular mortalities; C captures deviations in pollution 210 concentration from 10 μg/m 3 [56, 57]. The value was computed when variables in equation (2), 211 representing the relationship between changes in the particulate matter concentration and changes 212 in the impact on the exposed population. The model restrictions were set at PM2.5 < 12 µg/m 3 for the NOM and PM2.5 < 10 µg/m 3 in the 235 WHO simulation. In both cases, the reduction in pollution concentration is equal to zero, assuming 236 that total current deaths would not be attributed to the abovementioned pollutant. Furthermore, we 237 used the same value in all age groups in which the population was classified. Aphekom shows 238 the results as the "number of proposed premature deaths", understood as avoidable and "years of 239 life lost" understood as gains in life expectancy (see Appendix A). 240

Mexico's value of a statistical life (VSL) 241
Our economic evaluation requires a method capable to quantify the money value of total avoidable 242 deaths and other effects derived from exposure to poor air quality due to high levels of pollutants. 243 There are different types of methodologies aiming to estimate the value of life, cost of disability 244 and losses in productivity caused by air pollution. Independently of the approach, we multiplied 245 the Value of a Statistical Life (VSL) by the number of avoidable deaths. In our study, the VSL was 246 required to better understand the money cost of the number of deaths by total and cardiovascular 247 mortality. The VSL estimator compares, under a common unit of value, the cost and the benefits 248 of introducing public policy aiming to reduce mortality rates due to air pollution [18]. Although 249 VSL is debatable for ascribing money value to human life, this approach quantifies life for 250 statistical and resource allocation purposes, allowing to explain the cost of death. The VSL index 251 has been extensively applied in several economic studies to assess the cost of events. The 252 Organization for Economic Co-operation and Development (OECD) [51] provides a useful 12 framework to estimate the VSL for Mexico, which was applied to the estimated avoidable deaths 254 in our sample of three cities. To compute these values, we used the following formula: 255

Avoidable deaths and life expectancy 267
The analyzed population was stratified by age groups of five years in each class. We accounted 268 for people 30 years and older exposed to air pollution cause by PM2.5 during 2017 in a sample of 269 three cities. Table 1 depicts the results per city and the total of the sample in which we can observe 270 that as we move along of the stratified age groups, the number of people potentially affected by 271 air pollution decreases, which is explained by the natural tendency of biological age of the 272 population. The median age in the state of Sinaloa is 29 years and although the state is mainly 273 made out of young people, the effects of PM2.5 on older populations can be noticeable. As the 274 person ages, the risk associated to pollution exposure increases due to the presence of multiple 275 13 comorbidities, loss of mental abilities and the deterioration of physical functioning. Older adults 276 become the most susceptible to exposure to pollution as they age. 277 278 279    (Table 2). 365 Table 2 also shows result for the WHO scenario. In Culiacan, the total avoidable mortality ranged      This research did not receive any specific grant from funding agencies in the public, commercial, 531 or not-for-profit sectors. 532 533 Availability of data and materials 534 The datasets used and/or analyzed during the current study are available from the corresponding 535 author on request. 536 537

538
Ethics approval and consent to participate 539 Not applicable 540 541 Consent for publication 542 Not applicable 543 544 Competing interests 545 The authors declare no conflict of interest. 546 547 Figure 1 Location of the three cities examined: Culiacan, Los Mochis and Mazatlan [38] Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. Avoidable total and cardiovascular mortality due to the reduction in PM2.5 concentration levels (a) (cases per 100,000 inhabitants) (b), for population older than 30 years in the three cities examined Supplementary Files