Glycemic and lipid variability for predicting complications and mortality in diabetes mellitus using machine learning
The researches in the past have shown the usefulness of HbA1c and lipid variability in risk stratification in diabetes mellitus. Hence, Sharen Lee and colleagues have conducted a study under the title “ Glycemic and lipid variability for predicting complications and mortality in diabetes mellitus using machine learning” published in the BMC Endocrine Disorder Journal. The summary of this study is given below:
Objective:
To investigate the variability of HbA1c and lipids for adverse outcomes.
Method:
This is a retrospective cohort study conducted with type 1 and type 2 diabetic patients. This patient has prescribed insulin between 1st January to 31st December 2009 at outpatient clinics of Hong Kong public hospitals. The variability of HbA1c, total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and triglyceride were measured by the standard deviation (SD) and coefficient of variation. The main outcome was all-cause mortality followed by diabetes-related complications.
Findings:
The study reported that an increase in lipid variability and HbA1c is associated with increased risk for adverse outcomes in diabetes across different organ systems. Additionally, the relationship between hypoglycemic episodes and baseline neutrophil-lymphocyte ratio (NLR) with HbA1c and lipid variability suggests that intermittent hypoglycemia and chronic inflammation can affect the mechanism responsible for the pathogenic effect of changing glycated hemoglobin and lipid levels.
Limitation:
Firstly, there was under-coding, missing data, and coding error similar to other observational studies. Secondly, as it is an observational study it was only capable to establish correlation, not causation. Investigators didn’t note the year of diagnosis of diabetes. The variables that affect lipid levels such as smoking, alcoholism, and diet were not recorded.
Image Credit : Background photo created by xb100 – www.freepik.com