Grant: Catalyst Grant: Analysis of CLSA Data
McMaster University
$70,000.00
Strategies for Handling Large-Scale Missing Blood Markers in Longitudinal Studies
FY 2024-25
Summary
Understanding the changes in blood markers over time is crucial for diagnosing diseases, evaluating treatments, and monitoring patient health. However, collecting these blood markers in large, long-term studies can be challenging, especially when circumstances such as the COVID-19 pandemic prevent in-person visits. In the Canadian Longitudinal Study on Aging (CLSA), a nationwide study of over 50,000 adults, the collection of blood samples was significantly affected during the six-year follow-up period (2018-2020). Due to pandemic-related restrictions, about two-thirds of participants were unable to provide blood samples, resulting in a large amount of missing data.
Typically, when there is missing data, researchers either exclude participants with incomplete information or use advanced methods to fill in the gaps. Excluding participants can lead to biased results, while certain methods for filling in missing data may not work well when there are many gaps. Recently, machine learning and deep learning techniques have shown promise in handling missing data, but they do not fully account for changes that happen over time, which are crucial in longitudinal studies like the CLSA.
Our study aims to: (1) identify the main reasons why blood data is missing; (2) develop deep learning-based imputation methods using deep learning techniques ; and (3) run simulations to compare these methods with traditional imputation methods and understand which works best in different scenarios. By doing this, we hope to provide clear guidelines for other researchers facing similar challenges with missing data in long-term health studies. This research will improve the accuracy of findings and ultimately help develop better health strategies and care for people as they age.
Department
Program
Catalyst Grant: Analysis of CLSA Data
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
CIHR Funding Decisions Database
As published
Id
512842
Name
Ma Jinhui
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Aging
Theme2
Biomedical
Prcname2
Catalyst Grant: Analysis of CLSA Data
Projecttitle
Strategies for Handling Large-Scale Missing Blood Markers in Longitudinal Studies
Programtype2
Operating Grants
Orgnameinp2
McMaster University
Cihramount2
$70,000
Cihrequipment2
$0
Programname2
Catalyst Grant: Analysis of CLSA Data
Keyworddelim
Aging; Blood Markers; Missing Data
Abstract
Understanding the changes in blood markers over time is crucial for diagnosing diseases, evaluating treatments, and monitoring patient health. However, collecting these blood markers in large, long-term studies can be challenging, especially when circumstances such as the COVID-19 pandemic prevent in-person visits. In the Canadian Longitudinal Study on Aging (CLSA), a nationwide study of over 50,000 adults, the collection of blood samples was significantly affected during the six-year follow-up period (2018-2020). Due to pandemic-related restrictions, about two-thirds of participants were unable to provide blood samples, resulting in a large amount of missing data.
Typically, when there is missing data, researchers either exclude participants with incomplete information or use advanced methods to fill in the gaps. Excluding participants can lead to biased results, while certain methods for filling in missing data may not work well when there are many gaps. Recently, machine learning and deep learning techniques have shown promise in handling missing data, but they do not fully account for changes that happen over time, which are crucial in longitudinal studies like the CLSA.
Our study aims to: (1) identify the main reasons why blood data is missing; (2) develop deep learning-based imputation methods using deep learning techniques ; and (3) run simulations to compare these methods with traditional imputation methods and understand which works best in different scenarios. By doing this, we hope to provide clear guidelines for other researchers facing similar challenges with missing data in long-term health studies. This research will improve the accuracy of findings and ultimately help develop better health strategies and care for people as they age.
Competitiondate
202410
Approvedterm2
1 yr 0 mth
Pinamesdelim
Ma, Jinhui; Raina, Parminder S
Approvedterm
1 an 0 mois
Cihrcontribution2
IA~ $70,000~ $0~