Clustering and Tracking the Stability of Biological CVD Risk Factors in Adolescents: The Malaysian Health and Adolescents Longitudinal Research Team Study (MyHeARTs)

Background: Cardiovascular disease (CVD) risk factors tend to cluster and progress from adolescence to young adulthood. Reliable and meaningful clustering of CVD risk factors is essential to circumvent loss of information. Tracking adverse and high-risk profiles of adolescents is hoped to curb CVD progression later in life. The study aims to investigate the clustering of biological CVD risk factor among adolescents in Malaysia and the transitions between clusters over time. Method: The Malaysian Health and Adolescents Longitudinal Research Team study (MyHeARTs) examined school students aged 13 in 2012 and re-examined them in 2014 and 2016. In a two-stage stratified cluster sampling, 1,361 students were recruited, of which, 1,320 had complete data. In the follow-up, there were 881 and 637 students in 2014 and in 2016, respectively. Pearson's correlation coefficients were used to identify and remove highly correlated CVD risk factors. All risk factors were standardized into z-scores. The hierarchical and non-hierarchical (k-means) cluster analyses were used to classify students into high, medium and low risk clusters in each screening year. The tracking and stability of cluster transitions through cross-classification were enumerated with Pearson's inter-age correlations and percentages. Results: Three significant clusters of high, medium and low risk groups were derived from the clustering of eight biological CVD risk factors. The transitions between risk clusters from one screening year to the other were categorized as either stagnant, improved or adverse. The number of students who had adverse transitions increased from 15.5% (13–15 year) to 19.5% (15–17 year), 13.8 to 18.2% among the girls and 19.9 to 22.8% among the boys. For girls, the number of them who remained at high risk over the two transition periods were about the same (13.6 vs. 13.8%) whereas for boys, the percentage reduced from 14.6 to 12.3%. Conclusion: Over time, more than 12% of adolescents remained in the high risk cluster. There were sizable adverse transitions over time as more adolescents appear to be shifting toward an increased risk of having CVD. Collaborative and constant measures should be taken by parents, school, health promotion boards and policy makers to curb the multiplicative effect of clustering CVD risk factors among adolescents.