Use of glycaemic and lipid variability to predict the risk for major adverse cardiovascular events in patients with type 2 diabetes mellitus

Lee, Sharen (2024) Use of glycaemic and lipid variability to predict the risk for major adverse cardiovascular events in patients with type 2 diabetes mellitus. Doctoral thesis, University of East Anglia.

Preview

PDF Download (4MB) | Preview

Abstract

Introduction: With the global trend of shifting towards personalised medicine, there is an increasing need for new parameters to individualise the disease-monitoring of type 2 diabetes mellitus (T2DM) beyond existing models such as the Reynolds Risk Score. Measures of glycaemic and lipid variability, defined as the extent of change in glycaemic/ lipid indices during follow-up, have attained academic interest as potential prognostic biomarkers in patients with T2DM and cardiovascular diseases. The present thesis aims to explore the use of glycaemic and lipid variability for predicting major adverse cardiovascular events amongst patients with T2DM.

Methods: A number of retrospective, population-based studies were included, which assessed the predictive values of glycaemic and lipid variability for various major adverse cardiovascular events. The study population included patients with T2DM attending the Hong Kong Hospital Authority between January 1st, 2009 till December 31st, 2009, with follow-up until December 31st, 2019. Demographic, clinical, biochemical and pharmacological data was extracted from a territory-wide, linked electronic database. Cox proportional hazards regression was applied with risk scores constructed from the hazard ratios. The models were further enhanced by machine-learning techniques.

Results: Up to 273 678 patients were analysed in the studies described herein. Glycaemic and lipid variability were found to be consistently predictive for major adverse cardiovascular events across the different studies (p < 0.05). HbA1c standard deviation (p<0.0001) and lipid indices (total cholesterol: p=0.033, high density lipoprotein: p=0.082) were found to be predictors of sudden cardiac death. Significant predictors of all-cause mortality were incorporated into a score-based predictive risk model that had a c-statistic of 0.73, which was
improved to 0.86 (random survival forest) and 0.87 (deep survival learning models).

Conclusion: In conclusion, glycaemic and lipid variability can predict cardiovascular adverse events amongst patients with T2DM, allowing early intervention and management upon initial clinic visits in high-risk groups.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	Publication
Faculty \ School:	Faculty of Medicine and Health Sciences > Norwich Medical School
Depositing User:	Jennifer Whitaker
Date Deposited:	04 Apr 2025 10:10
Last Modified:	04 Apr 2025 10:10
URI:	https://ueaeprints.uea.ac.uk/id/eprint/98933
DOI:

Downloads

Actions (login required)

View Item