Causality Analysis From Grey Haze And Smoking To Lung Cancer

In recent years,the incidence of lung cancer and the mortality rates of lung cancer are showing a rapid growth trend,the cause of the lung cancer has also attracted wide attention.Currently smoking and lung cancer are considered to have a strong correlation,and 80% of lung cancer patients have a habit of smoking.It is noted that the rate of smoking throughout the world has not increased over the years or even been declining in many years,but the incidence of lung cancer has continued to soar in recent 30 years,there are also about 20% of lung cancer patients not having a history of smoking.So many researchers start to focus on the increasingly worsening environmental problems.Currently the most serious environmental pollution is undoubtedly the haze problem.In our thesis,we mainly discuss the effects of the two factors(grey haze and smoking)on the lung cancer.Causality relationship between two variables is an important concept and can reflects cause and effect between two variables,it is different from correlation.In our thesis we mainly apply the widely used Granger causality method and our recently proposed new causality to analyze the effects of the grey haze and smoking on lung cancer.Our results demonstrate that new causality method can reveal true causality between time series compared to Granger causality method.In this thesis we first study causality from grey haze to lung cancer in Guangzhou city and further indentify the lag relationship from grey haze to lung cancer.To do so,we estimate a linear regression model for grey haze and lung cancer mortality,calculate Granger causality value and new causality value between two variables,take significance test for the two causality values.New causality result confirms the fact that grey haze has important causal influence on lung cancer,which cannot be derived from Granger causality result.Then,we calculate all values of different lag parts in the joint linear regression model for lung cancer and take the lag part of the largest value as lag relationship between lung cancer and haze.Averagely lung cancer mortality and grey haze have causal lag relationship of 8 years.We consider grey haze of three different concentration levels and divide the grey haze data into three different parts,we confirm that grey haze at low concentration can cause lung cancerwith longer lag time,grey haze at high concentration can cause lung cancer with shorter lag time.We then study causal relationship between smoking and mortality of lung cancer based on the data of 12 countries such as Australia,the United States and the United Kingdom,etc.To do so,for each country we select two time series data: smoking of different lag years and the lung cancer mortality and estimate a linear regression model.Then calculate Granger causality value and new causality value.For each causality method the highest value of causal value is taken as the lag year of smoking’s causal effect on the country’s lung cancer mortality.New causality results demonstrate that the lag year of smoking’s causal effect on the country’s lung cancer mortality in 12 countries follows a normal distribution with mean value of 44 years.However,Granger causality results are not true because in one country the lag year of smoking’s causal effect on the country’s lung cancer is only 2,which is definitely not true.