Risk calculation of type 2 diabetes

Abstract

The paper presents an analysis of the risk of developing Type 2 diabetes according to family history and anthropometric variables. The age of diabetes onset was analysed in 2024 diabetics. We obtained several groups according to family history. In each group taken separately, the data describing the cumulative percentage of diabetes onset was fitted by logistic curve F(x) = p1/(1 + p2 ∗ p3 ((x/10)−p4) ). Comparing thesecurves we see that cumulative age-dependent risk increases from the group of randomly chosen persons through the group of first degree relatives to the children of diabetics. The highest risk of diabetes onset is determined by the curve representing the group of known diabetics. Another analysis was performed in a different group of 390 obese subjects (34 diabetics among them). Male diabetics had significantly higher body mass index (BMI) and weight than male non-diabetics. Female diabetics showed significantly higher weight, body mass index, waist to hip ratio (WHR) and age than female non-diabetics. Elimination of factors with randomization and matching showed a complicated relationship between diabetes, age and anthropometric variables. Using stepwise logistic regression we obtained the model for prediction of diabetes risk based on age, BMI, WHR: probability of diabetes + exp( u)), where u = −13.9 + 0.05431 ∗ age + 6.789 ∗ WHR + 0.07881 ∗ BMI for obese women, u = −11.84 + 10.01 ∗ WHR for obese men. In conclusion, genetic factors are the most important and can be exactly quantified in Type 2 diabetes. The importance of anthropometric variables for prediction of diabetes risk is also presented.