Portal de Periódicos da CAPES

Specific weights of metabolic syndrome risk factors in patients attending the Hospital Integral "Jose Maria Morelos", according to the Binary Logistic Regression Model – study of cases and controls

2024; Volume: 5; Issue: 7 Linguagem: Inglês

10.46932/sfjdv5n7-013

2675-5459

José Franco-Monsreal, Javier Jesús Flores–Abuxapqui, Mario Ramón Heredia-Navarrete, Lidia Esther del Socorro Serralta–Peraza, José Ricardo Hernández-Gómez, Eréndira Guadalupe Peralta-Martín, Angeles Yarintzi Tapia-Castro, María Selene Sánchez–Uluac,

Artificial Intelligence in Healthcare

Introduction. Metabolic syndrome is a group of disorders that occur at the same time and increase the risk of heart disease, stroke and type 2 diabetes mellitus. These disorders include high blood pressure, hyperglycemia, excess body fat around the waist, and abnormal cholesterol and triglyceride levels. Having just one of these conditions does not mean you have metabolic syndrome, but it does mean you are at increased risk for serious disease. And if you develop more of these conditions, the risk of complications such as type 2 diabetes mellitus and heart disease increases even more. Metabolic syndrome is increasingly common and up to one third of Mexican adults have it. If you have metabolic syndrome, or any of its components, radical lifestyle changes may be delayed. Binary logistic regression is one of the most expressive and versatile statistical tools available for data analysis in both clinical and epidemiology. Its origin dates back to the 1960s with the transcendental work of Cornfield, Gordon & Smith on the risk of suffering from coronary heart disease and, in the form we know it today, with the contribution of Walter & Duncan in which the subject of estimating the probability of occurrence of a certain event as a function of several variables is addressed. Its use has been universalized and expanded since the early eighties, mainly due to the computer facilities available since then. Objective. To evaluate multivariate specific weights of five risk factors (Abdominal obesity, Fasting blood glucose, Triglycerides, HDL–Cholesterol and LDL–Cholesterol). Material and methods. The epistemological approach corresponds to the quantitative, probabilistic or positivist approach. The study design corresponds to that of an analytical observational epidemiological case→control study with directionality effect→risk factors and prospective temporality. The criteria of the International Diabetes Federation (FDI) were used in the present work. Five hundred patients [100 (20.00%) cases and 400 (80.00%) controls] were studied. Any patient with Abdominal obesity ≥ 102 cm in the male gender and ≥ 88 cm in the female gender was labeled as a case. Any patient with Abdominal obesity ≤ 101 cm in the male gender and ≤ 87 cm in the female gender was labeled as control. Multiple controls from the same population base such as two or more controls per case can be used to increase the statistical power of the study. However, it is accepted that increased statistical power is gained only up to a rate of one case for every four controls (Gordis, 1996). In the evaluation of the specific weights of the five risk factors, the values of the Exponents β or Odds Ratios (OR) of the binary logistic regression model were used. Results. OR> 1 indicated the positive contribution of the risk factors Abdominal obesity, Fasting blood glucose, Triglycerides, HDL–Cholesterol and LDL–Cholesterol. Conclusion. The obtained values of the Exponents β or ORs point to the positive contribution in ascending numerical order of the risk factors HDL–Cholesterol (1.5966); Fasting Hematic Glucose (4.0735); Abdominal Obesity (4.6475); LDL–Cholesterol (13.4475); and Triglycerides (17.7779).