Connecting the pieces: Building a life course picture from Canada's existing data
Health in later life is shaped by what happens decades earlier — in childhood, adolescence, and through major transitions such as entering school or retiring. Studying those patterns requires data that follow people from early life into old age. In Canada, no such dataset exists.
A new study published in the International Journal of Population Data Science (IJPDS) asks whether existing Canadian data sources could be combined, or harmonized, to fill that gap.
Researchers from McMaster University and the World Health Organization carried out an environmental scan to map what data are already collected across the country, which life stages they cover, and where they might align.
"Canada has a wealth of data on health and well-being, but often a single study captures only one life stage," said lead author Dr. Vanessa De Rubeis. "We wanted to know whether those pieces could be connected to answer questions that span the entire life course."
The researchers drew on two curated catalogues of longitudinal studies: Maelstrom and the Atlas of Longitudinal Datasets. From 262 Canadian records, 14 studies met the inclusion criteria. Together these studies cover more than nine million participants, and six recruit from across Canada rather than a single province or territory.
The studies were mapped to seven life stages, from childhood through to the oldest adults aged 85 and over, and to the three components of the WHO Healthy Ageing framework.
These are intrinsic capacity — the physical and mental capabilities a person can draw on; the environment —the places where people live and conduct their lives; and functional ability —what a person is able to do given their capacities and their environment.
No single study followed the same participants across the full life course. The Manitoba Multigenerational Cohort came closest, including people at every life stage by linking family members across generations. Studies that recruit in early life have so far followed participants only into childhood, while most others begin in adulthood.
Almost all the included studies captured at least one measure of intrinsic capacity, functional ability and the environment, showing that Canadian data look beyond disease to broader aspects of health and well-being.
But bringing these existing datasets together is not as simple as identifying similar measures in each study.
The researchers found that a measure suitable at one life stage may not be suitable at another. Instead, what matters is the underlying construct being captured. Harmonization should therefore begin by establishing what each construct means at each life stage and then identifying valid measures appropriate to those stages.
Other challenges would also need to be addressed. Differences in sampling and design raise the risk of selection and attrition bias when data are pooled, and measures would need to be transformed onto common scales before they could be combined.
The study suggests harmonization may be possible and could offer a potential solution for conducting life course research in Canada. However, additional steps would be needed before this approach could be applied in practice.
"Canada has made significant investments in collecting these data. Leveraging that investment is an efficient path for life course research, and one that could shape how we support health at every stage of life," said senior author Dr. Parminder Raina.
Click here to read the full article
Dr. Vanessa De Rubeis, PhD, Department of Psychiatry & Behavioural Neurosciences, McMaster University and Research Consultant, Department of Sexual, Reproductive, Maternal, Newborn, Child and Adolescent Health, and Ageing, World Health Organization, and Dr. Parminder Raina, PhD; Scientific Director, McMaster Institute for Research on Aging (MIRA), McMaster University; Professor, Health Research Methods, Evidence, and Impact, Faculty of Health Sciences, Canada