| Sumario: | Introduction: One of the major risk factors for cardiovascular disease in diabetes mellitus is dyslipidaemia. The characteristic of diabetic dyslipidaemia includes high plasma triglycerides concentration, low HDL cholesterol concentration and increased the concentration of small dense LDL-cholesterol particles. Most dyslipidaemias are hyperlipidemias; which is an increase in lipids in the blood that occur in developed countries. This is often caused by diet and lifestyle. Prolonged levels of insulin can also cause dyslipidaemia. Objective: The aim of this paper is to determine the factor that contributes to dyslipidaemia disease among Type 2 Diabetes Mellitus attending the outpatient clinic in Hospital Universiti Sains Malaysia. Method: The first method approach for the factor determination is through binary logistic regression (BLR). After obtaining the list of factors, the validation process will take the place. The validation of the factor will be obtained through multilayer perceptron (MLP) procedure where the training and testing procedure will be applied. Results: The accuracy and the significant factor will be determined and used for educational purposes and for the decision maker. Through the both methodology it was found that gender (β3 = -2.643, p < 0.25, 95% CI: 0.004, 1.353), sodium (β2 = -0.661, p < 0.25, 95% CI: 0.313, 0.850), and creatinine, (β1 = 0.043, p < 0.25, 95% CI: 0.997, 1.093), were the factor that contribute most to the dyslipidaemia disease. Conclusion: From these two methodologies, it was found that creatinine, sodium and gender factor are the most dominating factor which contributes to dyslipidaemia disease. These three factors were validated through the multilayer perceptron (MLP) procedure.
|
|---|