EpiBio Seminar Series | PhD Thesis Proposal Defense Public Lectures

"Understanding frailty in type 2 diabetes mellitus: From epidemiological evidence to a longitudinal analysis in a national cohort"

Presented by Aayushi Joshi, PhD Candidate.

Abstract: Frailty is highly prevalent among individuals with type 2 diabetes mellitus (T2DM) and independently associated with mortality, hospitalization, and disability, yet existing evidence is largely cross-sectional and frailty is rarely assessed in routine T2DM care. This dissertation addresses this gap through four integrated chapters. The first is a scoping review mapping longitudinal evidence on T2DM and frailty and predictors of frailty progression. The second and third are original analyses using three waves of the Canadian Longitudinal Study on Aging (CLSA); one examining the prospective association between T2DM and incident frailty and frailty severity over six years, and the other investigating osteoporosis as a mediator of the T2DM–frailty relationship. The fourth is a review article articulating the clinical rationale for integrating frailty assessment into T2DM and cardiometabolic care. Together, this work will generate key Canadian longitudinal evidence on T2DM and frailty and inform more individualized care for older adults with T2DM.

"Risk stratification algorithms to predict dementia"

Presented by Patricia Nistor, PhD Candidate.

Abstract: Dementia burden continues to increase as populations age, particularity in Canada. Therefore, it is important to identify individuals with high risk of developing dementia early, so that targeted resources can be directed towards the most appropriate preventative interventions, reducing illness burden individually and systemically by preventing progression to dementia. Three groups of risk algorithms will be developed, genetics only, lifestyle/clinical risk factors only, and a combination of genetics and lifestyle/clinical risk factors. From these models a simplified algorithm suitable for application in Canadian primary care will be derived and evaluated. CanPath data will be used for model development and internal validation. Model performance will be assessed used using the c-statistic, calibration, and sensitivity and specificity analyses. Internal validation will be conducted with a sample of data set aside for internal validation, along with bootstrapping, with 100 repetitions, will be employed, alongside the AUC, the F score, and the Brier Score. Development of a risk stratification algorithm for dementia for the Canadian context applicable at a stage when preventive interventions are most effective may help reduce the burden of dementia in later life for individuals, their caregivers, and the healthcare system.

Attend in person: PHFM 3015
Attend online: please request the Zoom link at epibio@uwo.ca