Purpose
Develop a parsimonious model of individuals at heightened-risk for 3-year cardiometabolic multimorbidity (CM) onset.
Design
An observational, secondary analysis of Canadian Longitudinal Study on Aging (CLSA) data.
Setting
CLSA is a national cohort study in Canada. Baseline data were collected between 2010-2015, and follow-up data were collected between 2015-2018.
Subjects
CLSA included community-dwelling adults aged 45-85 at recruitment from across Canada.
Measures
Health conditions: stroke, heart disease or heart attack and diabetes. Personal factors: age, sex, marital status, household income, education, and ethnicity. Environmental factors: social support, personal assistance, and location of residence. CM cases: at least two of stroke, heart disease and diabetes at follow-up assessment.
Analysis
Hierarchical logistic regression analyses with backwards elimination procedures were used to develop a parsimonious prediction model.
Results
The sample consisted of 41 841 individuals, representing a weighted population of 13 741 119. The population had a mean age of 62.3 years (SD = 10.1), was 53% female, predominantly married or in common-law relationships (77%), post-secondary graduates (61%), white (95%), and lived in an urban area (81%). Males (OR:1.93, 95%CI:1.65-2.25, P < 0.001), ≥65 years (OR:1.51, 95%CI:1.29-1.76), P < 0.001), who had stroke (OR:20.09, 95%CI:12.88-30.35, P < 0.001), heart disease (OR:15.55, 95%CI:12.60-19.26, P < 0.001), or diabetes (OR:12.57, 95%CI:10.37-15.31, P < 0.001), not completed post-secondary (OR:1.30, 95%CI:1.04-1.61, P = 0.017), income of <50k (OR:1.29, 95%CI:1.10-1.52, P = 0.002), and received home care (OR:1.56, 95%CI:1.17-2.04, P = 0.002) were at heightened risk of CM.
Conclusions
Developing a profile of high-risk individuals may enhance the efficiency of CM prevention and reduce disease onset. Critical limitations include the CLSA exclusion criteria, and the small proportion of minoritized individuals that restrict generalizability in these populations.