İstanbul'da 2007-2012 Yılları Arasında Hava Kirliliğinin Ölümler Üzerindeki Etkisinin Modellenmesi

Abstract

Hava kirliliğinin insan sağlığı üzerindeki etkileri, günümüzde bütün dünyanın üzerinde durduğu önemli bir sorundur. 1952 yılında, İngiltere nin başkenti Londra da meydana gelen ve 5.000 kişinin hayatını kaybetmesine neden olan hava kirliliği felaketi gibi ciddi vakalar, hava kirliliği ile ilişkili sağlık sorunlarının ve konuyla ilgili önlemlerin dünyanın gündemine getirilmesine neden olmuştur. Yapılan pek çok çalışma sonucunda, hava kirliliğinin solunum ve kalp hastalıklarına bağlı ölüm ve hastalık sayılarıyla yakından ilişkili olduğu ortaya konulmuştur.Dünya Sağlık Örgütü nün verilerine göre, hava kirliliği nedeniyle yılda 1,3 milyon kişi hayatını kaybetmektedir. Ölüm miktarlarının hava kirliliği ile olan ilişkisinin araştırılması son yıllarda dünya çapında önem kazanmaktadır.Araştırmalar,kabul edilmiş olan hava kirliliği sınır değerlerinin çok altındaki kirlilik seviyelerinde dahi ölüm ve hastalık oranlarında artışların olduğunu göstermektedir. 1980 lerden beri kamuoyunun dikkatini çekmekte olan İstanbul un hava kirliliği problemi, Türkiye nin çevre sorunları arasında ön sıralarda yer almaktadır. Dünyanın pek çok büyük şehrinde olduğu gibi İstanbul da da, artan sanayi, araç sayısı ve plansız yapılaşmaya bağlı olarak hava kirliliği problemleri meydana gelmektedir. Günümüzde, gerek sabit nokta kaynaklı (evler, sanayi tesisleri, vb.) ve gerekse motorlu araç kirliliğine bağlı hava kirliliği olaylarında artışlar yaşanmaktadır. Buna bağlı olarak, İstanbul daki sağlık sorunları ve hava kirliliği arasındaki ilişkilerin belirlenmesi, konuyla ilgili koruyucu önlemlerin alınması açısından önemlidir. Bu çalışmada, İstanbul da 2007-2012 yılları arasında ölçülen kirletici değerleri ve bu dönem boyunca meydana gelen ölüm vakaları arasındaki ilişki araştırılmıştır. Çalışmamızda, Aksaray, Alibeyköy, Beşiktaş, Esenler, Kadıköy, Kartal, Saraçhane, Sarıyer, Ümraniye, Üsküdar ve Yenibosna hava kalitesi ölçüm istasyonlarından alınmış olan ve 2007-2012 yılları boyunca ölçülen kirletici (PM10, SO2, NO2, NOx, NO ve CO) değerleri kullanılmıştır. Kirletici miktarları ve meydana gelen ölüm vakaları arasındaki ilişkiyi araştırmak amacıyla, İstanbul Okmeydanı Eğitim ve Araştırma Hastanesi, Ümraniye Eğitim ve Araştırma Hastanesi, Haydarpaşa Numune Eğitim ve Araştırma Hastanesi ve İstanbul Şişli Etfal Eğitim ve Araştırma Hastanelerine ait ölüm verilerden yararlanılmıştır. Hava koşullarının ölümler üzerindeki etkilerini kontrol etmek amacıyla, Meteoroloji Genel Müdürlüğünden İstanbul daki 3 adet meteoroloji istasyonuna (Sarıyer, Florya ve Göztepe) ait2007-2012 yılları boyunca ölçülen saatlik ortalama sıcaklık, bağıl nem ve basınç verilerinden de yararlanılmıştır. Çalışmada, PM10, NO2 ve SO2 hava kirliliği ölçüm değerleri ile ölüm sayıları arasındaki ilişki, bir regresyon yöntemi olan genelleştirilmiş lineer model (GLM) ile analiz edilmiştir. Ayrıca, sıcaklık ve bağıl nem verileri de, meteorolojik unsurların karıştırıcı etkisini ortadan kaldırmak amacıyla modelde kullanılmıştır. Diğer taraftan mevsimlere bağlı olarak ortaya çıkan trendlerin düzeltilmesi amacıyla modelde doğal kübik splayn kullanılmıştır. Çalışma sonucunda, günümüzde İstanbul da yaşanan hava kirliliğinin,solunum sistemi hastalıklarına bağlı ölümler, kalp-damar hastalıklarına bağlı ölümler ve kaza dış tüm hastalıklar nedeniyle meydana gelen ölümler üzerinde anlamlı bir etkisinin olduğu ortaya çıkmıştır.

Today, because of increasing numbers of human population and pollutant sources, air pollution problem is a major concern throughout the world. Air pollution includes both gaseous and particle pollutants. Pollutants can originate as primary pollutants (those exhausted directly into the air by pollution sources) or as secondary pollutants, (those formed in the atmosphere, largely from the primary pollutants). The primary pollutants include particulate matter (PM), which includes carbonaceous black soot. Gaseous pollutants include sulfur oxides (suchas sulfur dioxide, SO2), nitrogen oxides (such as nitric oxide, NO, and nitrogen dioxide, NO2), and carbon oxides (such as carbon monoxide, CO). Gaseous ozone (a major component of photochemical smog) formed from nitrogen oxides and hydrocarbons, and particulate sulfate (e.g., sulfuric acid) and nitrate (e.g., ammonium nitrate) aerosols created in the atmosphere from sulfur and nitrogen oxide gasesare considered secondary pollutants, respectively. According to the results of the epidemiological studies, short-term exposure to outdoor air pollution is linked to increased mortality, increased rates of hospital admissions and emergency department visits. Mortality is the most important health effect of outdoor air pollution. According to World Health Organization, due to the air pollution, 3 million people lost their lives every year. The amount of studies investigating the relationship between air pollution and mortality is gaining importancein recent years. Many epidemiological studies of air pollutions show that, even pollution levels below the limits could increase the rates of mortality. Air pollution problem in Istanbul has received wide public attention since the 1980s. The city has experienced significant particulate matter and sulfur dioxide episodes due to the fossil fuel burning for domestic heating and industry in the late 1980s and beginning of 1990s. After switching from coal to natural gas, particulate matter concentrations and sulfur dioxide levels were gradually decreased in Istanbul. However, nowadays the city is facing especially particulate matter (PM10 and PM2.5) and NO2 problems depending on fixed point-source (houses, industrial facilities, etc.) as well as motor vehicle-related air pollution. To assess air pollution impact on human health, studies often use statistical methods that are very useful tools to summarize and interpret data. According to the health effects (acute or chronic), type of exposure (short or long term), the nature of the response (binary or continuous) and data structure, model to be applied and the effects to be estimated are determined. The time series studies are the most used as they demand simply data such as the amount of hospital admission or mortality in a given day, being easy to obtain on health government departments. As they are useful tools to assess the relationship between one or more explanatory variables (independent, predictor variables or covariates) (x1, x2,…,xn) and a single response variable (dependent or predictedvariable) (y), regression models are usually chosen in time series studies. Because of the non-linearity of the response variable in time series studies of air pollution impacts on human health, the Generalized Linear Models (GLM) with parametric splines (e.g. natural cubic splines) and the Generalized Additive Models (GAM) with non-parametric splines (such as smoothing splines or lowess smoothers) are now widely applied. In this study, Generalized Linear Model (GLM) is applied and the steps to apply the GLM to air pollution impact on mortality are presented in details. The results have shown that the GLM with Poisson regression fitted well to the database of the case studies considered. In the study, the relationship between the pollutant concentrations measured from January 1, 2007 to December 31, 2012 in Istanbul and the deaths that occurred during this period were investigated. Death records of İstanbul Okmeydanı Education and Research Hospital, Ümraniye Education and Research Hospital, Haydarpaşa Numune Education and Research Hospital and İstanbul Şişli Etfal Education and Research Hospital were obtained from Istanbul Provincial Health Directorate from January 1, 2007 to December 31, 2012. Mortality data were classified into deaths due to total non-accidental causes (ICD-10:A00-R99), cardiovascular diseases (ICD-10: I00-I99), and respiratorydiseases (ICD-10: J00-J98). Air quality indicators in our study include particulate matter with aerodynamic diameter of &#8804;10 &#956;m (PM10), sulfur dioxide (SO2) and nitrogen dioxide (NO2). We obtained hourly air quality data from Aksaray, Alibeyköy, Beşiktaş, Esenler, Kadıköy, Kartal, Saraçhane, Sarıyer, Ümraniye, Üsküdar and Yenibosna air quality stations of Istanbul Metropolitan Municipality. To allow adjustment for the effect of weather on mortality, meteorological data (daily mean temperature and humidity) were obtained from Florya, Sarıyer and Göztepe meteorological stations of General Directorate of Meteorology. In order to generate Generalized Linear Model (GLM) and make calculations, R statistical computing program was used. R was created by Ross Ihaka and Robert Gentleman at the University of Auckland, New Zealand, in 1993 and is currently developed by the R Development Core Team. R is highly extensible through the use of user-submitted packages for specific functions and specific areas of study. In our study dlnm and spline packages were used. After entering data in the appropriate format; R generates GLM, calculates output values and creates graphs. To apply the GLM with Poisson regression, four main steps were followed: development of the database; adjustment of the temporal trends; choosing best model (Akaike information criterion and ANOVA test) and results analysis. The details of each step are given. Seasonality is the long-term trend usually included in time series studies of air pollution impact on human health and in this case study it was considered a natural cubic spline, the most used parametric smooth in GLMs. To apply the natural cubic spline in GLM with Poisson regression, an explanatory variable for the days is added to the model, consisting of values from 1 to 2192, comprising all six years of data. The short-term trends usually considered in epidemiological studies of air pollution are the day of the week and holiday indicator. The day of the week variable was considered as a qualitative variable, which varies from one to seven, starting at Mondays. The holiday indicator wasn t included, because deaths can also occur at the weekends. After adjusting the GLM with Poisson regression including all the time trends and explanatory variables and choosing the degrees of freedom that better fits the data; the fitted model was tested using Akaike information criterion and ANOVA test to assure that it is the right one to be applied to the case-study. In order to estimate the impact of air pollution on human health, making some predictions according to pollutants concentration, the relative risk (called rate ratio by statisticians) is used. It is a measure of the association between an explanatory variable (e.g. air pollutant concentration) and the risk of a given result (e.g. the number of respiratory mortality). To analyze and estimate the PM10, NO2 and SO2 impacts on the health ofİstanbul s population, the relative risks for the models considering the effects of ten lag days was calculated. According to the results, the relative risks per 10 &#956;g/m3 increases in PM10 were 1.022 (1.0, 1.05), 1.004 (0.96, 1.05) and 1.024 (1.01,1.04) for cardiovascular, respiratory and all-cause mortality, respectively. The relative risks per 10 &#956;g/m3 increases in NO2was 1.033 (1.011, 1.055) for cardiovascular mortality. Because the p values for NO2 obtained from ANOVA test was p < 0,05 for respiratory and total non-accidental deaths, HA hypothesis was rejected. According to this result, NO2 concentrations do not have significant effects on the respiratory and total non-accidental deaths. The relative risks per 10 &#956;g/m3 increases in SO2 were 1.164 (1.028, 1.32), 1.15 (0.94, 1.41) and 1.114 (1.043,1.2) for cardiovascular, respiratory and all-cause mortality, respectively. Through the modelassessment, it was possible to conclude that air pollution is an important factor affecting the mortality in Istanbul. Considering the results achieved and the proven effects of air pollution on mortality, strategies should be defined to reduce itand consequently protect public health. These strategies could start by reviewing the legislated limits now in force in Turkey, considering those established by the World Health Organization and Europe, and also through the application of measures to reduce industrial, traffic and ship emissions and unplanned construction of buildings.