Exploring the shared multi-omic architecture of late-life depression and cognitive decline: Discovering and predicting biologically-defined patient subtypes using machine learning
Description: Late-life depression (LLD) in adults over the age of 60, is a highly debilitating mental illness affecting up to 16% of older adults in Canada. More than 50% of patients do not achieve remission which is associated with progressive cognitive decline and a 2-3 times increased risk for dementia. In addition, antidepressant treatment is highly variable and might also depend on cognitive functioning in LLD. Given the aging population with 937,000 Canadians expected to develop dementia in the next 15 years, we need to understand the joint biology underlying depression and cognitive decline to inform treatment and reduce the public health burden currently totalling $10.4B dollars/year. The clinical complexity of depression, cognitive decline and dementia requires that advanced statistical methods, such as machine learning, be applied to develop predictive biomarkers based on biologically-informed subgroups of patients. The proposed investigation will have three major aims which will use machine learning models and multiple types of genomic data, including: (1) exploring the existence of cognitively-different subgroups of depressed patients; (2) predicting the discovered subgroups using genomic data; and, (3) understanding whether the cognitive subgroups respond differently to antidepressant treatment. For these aims, I will be using publicly-available data from large studies of older adults including, as well as, well-characterized clinical sample of older adults treated with antidepressants. By leveraging genomic data and large samples, we will be able to discover subgroups of patients which will have different risks for cognitive decline and dementia, which may affect their response to antidepressants. Furthermore, these findings will allow us to develop genetic tests which could be used to personalize and optimize treatment of depression, as well as, to slow the progression of cognitive decline
Amount: CA$105,000
Funding type: fellowship
Funder scheme: Doctoral Award - Frederick Banting and Charles Best Canada Graduate Scholarships (CGS-D)
Start date: 2018-05-01
End date: 2021-04-30
Investigator: Victoria Marshe (lead)
Institutions: Centre for Addiction and Mental Health, Centre for Addiction and Mental Health (Toronto)
Subfield: Pharmacology (medical)
Field: Medicine
Domain: Health Sciences
Funded outputs count: 0
Funder award ID: 157926_1
OpenAlex ID: G2563448401
