Multi City Air

Majority Survey on Air Pollution and Mortality in Latin America (The ESCALA Study)

A first multi-city assessment of the effects of short-term exposure to PM10 and tropospheric oxygen oxides (ozone) on death in nine towns in the region of Cameroon. Under the direction of Dr. Isabelle Romieu in Mexico, the scientists, in cooperation with Dr. Nelson Gouveia in Brazil and Dr. Luis Cifuentes in Chile, have investigated morbidity from all causes and in different ages using a joint analytical frame.

The researchers analysed morbidity in each city and area as a whole and investigated two pollution model in different towns. In addition, they used two meta-analytical statistic methods to further analyse the effect of different towns.

Majority survey on air contamination and morbidity in Latin America (ESCALA study). - Circulation

INSTRUCTION: The ESCALA* Estudio de Salud y Contaminación del Aire en Latinoamérica (Project ESCALA*) is a university-funded research aimed at studying the relationship between air contamination exposures and deaths in nine towns in the region, using a shared analytical methodology to obtain comparative and up-to-date information on the impact of air contamination on multiple causes of deaths in different ages.

MODES: The ESCALA exercise comprised three investigation groups in charge of collecting and analysing city air quality and death statistics from three different states. Five different standard isation logs for collecting and analysing information in order to assess the impact of air pollutants on morbidity (see Annexes B-F).

Using the same protocol, researchers were able to directly benchmark the results between towns. In the first phase, analysis of the purely physical cause and cause-specific morbidity was performed in relation to particles < or = 22 o'clock in Aerodynamic Particle Size (PM10) and in relation to Ocean Opens (O3) in Brazilian, Chilean and Mexican city. In order to evaluate the relationship between air contamination exposures and the danger of deaths, the PM10 and Q3 values were adjusted using disseminated tag model (DLM) techniques.

Every DLM measure the accumulated effect of a contaminant on a given date (day 0) and the effect of that date on the effect of that contaminant on several consecutive (delayed) dates. Only the results of the OLMs with a 3-day delay (DLM 0-3) are presented in this paper, however, as we found a declining correlation with death rates in different ages to increase the delay effect from 3 to 5 for PM10 and O3.

Potentially modifying effects of socio-economic state ( "SES") on the relationship between PM10 or Q3 concentrations and morbidity were also investigated in four cities: Mexican City, Rio de Janeiro, São Paulo and Santiago. In this phase, the relationships between urban deaths and air pollutants were benchmarked in order to assess the existence of urban diversity and to investigate urban scale variable that could account for this diversity.

Meteorological analysis was performed to aggregate urban sterility effect estimations and assess the existence of urban outcome disparities, while meta-regression modelling was used to investigate variable that could account for urban disparity in death rates associated with exposure to PM10 (but not to O3). RESULTS: The results of the morbidity analysis are presented as percentage changes in exposure to disease (RPC) with a 95% CI.

The RPC is the rise in morbidity associated with an elevation of 10 microg/m3 in the 24-hour mean PM10 or in the 8-hour mean O3 peak per day. The majority of results for PM10 were positively and significantly statistical and showed an elevated death rate at elevated environmental levels.

The results for O3 also showed a statistical significant rise in urban deaths with available evidence. Environmental levels with the DLM 0-3, PM10 distribution layer were associated with an elevated death rate in all towns except Concepci6n and Temuco. The RPC and 95% CI were 1 in Mexico City and Santiago.

The PM10 was also significantly associated with elevated death from heart attacks, respiatory, cardio-vascular, cerebrovascular strokes and COPD in most urban areas. In general, the few insignificant impacts were found in the smallest towns (Concepción, Temuco and Toluca). Proportional increase in environmental morbidity associated with environmental concentration of 3O was lower than for PM10.

Naturally induced deaths were significantly associated with O3 in Mexico City, Monterrey, São Paulo and Rio de Janeiro. Also in these towns and in Santiago, elevated risk of death was noted for certain causes. Seasonally layered analysis showed different pattern of death and O3 for hot and colder periods of the year.

In São Paulo and Rio de Janeiro, the estimated risks for the hot seasons were greater and significant for several causes of deaths. Estimated risks for the winter months were higher and significant for some causes of deaths in Mexico City, Monterrey and Toluca. However, we found that the risks of airway -related deaths were greater in those with low SES, while the risks of heart and brain -related deaths were greater in those with moderate or high SES.

São Paulo had a higher completely normal death rate for those with high SES, while Rio de Janeiro had a higher completely normal death rate for those with low SES. Life-threatening diseases were more common in both towns, especially in the low and high seas.

At Santiago, the death rate did not fluctuate from the severity of SES due to inherent causes, but those with low SES had a higher airway death rate, especially for COPD. Individuals with moderate SES had a higher death rate due to heart and brain strokes. Surrounding PM10 levels were only investigated in Mexico City, Santiago and São Paulo for the effects on baby and adolescent deaths due to airway diseases and lower airway infections (LRI).

A significantly higher death rate due to these causes was found both in Santiago (infants and older children) and in Mexico City (infants only). O3 was associated with an elevated morbidity in Mexico City (infants and older children) and in São Paulo (only babies during the hot season).

Results of the meta-analyses confirm the affirmative and statistical significant relationship between PM10 and pure nature death (RPC = 0.77%[95% CI: 0.60 to 1.00]) using the chance effect models. Concerning death from specified causes, the percent elevation in death rate varied from 0.72% (0.54 to 0.89) in the case of cardio-vascular diseases to 2.44% (1.36 to 3.59) in the case of COPD, also using the chance effect mode.

Significant beneficial association of some causes with the chance effect models was found for occupational oestrogen (O3), but not for all naturally occurring causes or for airway disease in humans 65 years of life and older (> or = 65 years), and not for COPD and cerebral circulatory disorders in ages > or = 65.

Purely normal morbidity increased by 0.16% (-0.02 to 0.33). Mean mean temperatures in the hot seasons, infant populations (< 1 year), child populations of at least 1 year but < 5 years (i.e. 1-4 years), human populations > or = 65 years, geographical densities of PM10 screens, mean PM10 concentration per year, and death rate from pulmonary oncology.