Diabetes Mellitus has been an on-growing epidemic all over the world. It costs diverse tolls from human life, causes damage to hearing, seeing and different body organs. Moreover, it holds back the usual protection system after the body skin is cut off. Comprehending such its devastating outbreak and authority Bangladesh Diabetic Association was formed in 1956, which, by time turned into Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders (BIRDEM) in 1980. This study concerns an analysis of a diabetic patients’ data set, with size (n)=850, collected from BIRDEM that contains demographic variables such as age, sex, blood-glucose level etc. accompanied by various records of complications arose. These patients were under the BIRDEM record in between 1973 and 1996. To analyze the data the Cox’s Proportional Hazards Model was applied because of its being distribution free and its significance to interpret the parameters. However, the analysis reveals that age is an important factor positively associated with long term effects arose due to Diabetes – though in most cases it could not satisfy the proportionality assumption. Hence, a stratified Cox model was used later on. And to check the proportionality assumption the graphical approach is employed. The other factors sex, weight also influence largely the resulted complications. Here the used computer programming was S-plus, and the packages SPSS and Excel.