Epidemiology of hypoglycaemia: Trends, risk factors and outcomes

Background: Few data are available on the burden, risk factors, and outcomes of hospitalisation for hypoglycaemia. Newer glucose-lowering medications, such as glucagon-like peptide-1 receptor agonists (GLP-1RAs) and sodium–glucose cotransporter 2 (SGLT2) inhibitors, have been associated with a lower risk of hypoglycaemia in individual randomised controlled trials (RCTs); yet, they have not been systematically compared to older therapies. Lastly, recent observations have also suggested an association between hypoglycaemia and cardiovascular mortality. Methods: This research is structured in three parts. First, I used the NHS Hospital Episode Statistics data to: examine trends of admissions for hypoglycaemia in England between 2005 and 2014; define risk factors for admissions and differences in outcomes; develop and validate prognostic models to calculate risk of inpatient death and length of hospital stay. Second, I compared the risk of hypoglycaemia for once-weekly GLP-1RAs and SGLT2 inhibitors vs other medications with network meta-analyses of RCTs. Third, I investigated the relationship between fasting plasma glucose and risk of arrhythmias in a cohort study, aiming to clarify the pathophysiological mechanisms linking hypoglycaemia to cardiovascular disease. Results: Admissions for hypoglycaemia increased between 2005 and 2010, with more stable trends thereafter. Differences exist across regions in England for both trends and risk factors for admissions: these findings have been instrumental for the development of a tool to calculate individual risk of inpatient mortality and length of hospital stay. Meta-analyses indicated a lower risk of hypoglycaemia for GLP-1RAs and SGLT2 inhibitors compared to older glucose-lowering therapies. Lastly, in the cohort analysis, there was an inverse relationship between fasting plasma glucose and risk of arrhythmias. Conclusion: This thesis can broaden understanding of the burden of hospitalisation for hypoglycaemia and elucidate the link between hypoglycaemia and cardiovascular disease. These results could also assist decision makers in the adoption of individual- and population-level strategies.