Sociodemographic Correlates of High Cardiovascular Health Across Childhood and Adolescence: A Prospective Study Among 2 Cohorts in the ECHO Consortium

Abstract

Background: This study seeks to characterize cardiovascular health (CVH) from early childhood to late adolescence and identify sociodemographic correlates of high CVH that serve as levers for optimizing CVH across early life.

Methods and results: Among 1530 youth aged 3 to 20 years from 2 cohorts in the ECHO (Environmental Influences on Child Health Outcomes) consortium, we first derived CVH scores on the basis of the Life's Essential 8 construct comprising 4 behavioral (nicotine use/exposure, physical activity, sleep, and diet) and 4 health factors (body mass index, blood pressure, non-high-density lipoprotein cholesterol, and fasting glucose) during early childhood (mean age, 3.5 years), middle childhood (8.0 years), early adolescence (13.3 years), and late adolescence (17.8 years). Next, we used generalized regression to estimate the probability of high (versus not high) CVH with respect to sociodemographic characteristics. Overall CVH score was stable across life stages: 81.2±7.6, 83.3±8.0, and 81.7±8.9 of 100 possible points in early childhood, middle childhood, and early adolescence, respectively. Accordingly, during these life stages, most children (63.3%-71.5%) had high CVH (80 to <100). However, CVH declined by late adolescence, with an average score of 75.5±10.2 and 39.4% high CVH. No children had optimal CVH (score=100) at any time. Correlates of high CVH include non-Hispanic White race and ethnicity, maternal college education, and annual household income >$70 000. These associations were driven by behavioral factors.

Conclusions: Although most youth maintained high CVH across childhood, the decline by late adolescence indicates that cardiovascular disease prevention should occur before the early teen years. Disparities in high CVH over time with respect to sociodemographic characteristics were explained by behavioral factors, pointing toward prevention targets.

Keywords: Life's Essential 8; cardiovascular disease; cardiovascular health; epidemiology; primordial prevention.

Figure 1. Probability of high CVH (score ≥ 80 points) per Life's Essential 8 estimated from early childhood through late adolescence (3–18 y) among 1530 children in ECHO with respect to child sex (A), child race and ethnicity (B), maternal education (C), and annual household income (D).

Trajectories were derived from generalized linear mixed‐effects models with a logit link. In these models, the outcome was high CVH (yes/no) and the predictors included the cubic spline terms for continuous age, the sociodemographic characteristic of interest, and interactions of sociodemographic characteristic with each spline term for age. CVH indicates cardiovascular health; and ECHO, Environmental Influences on Child Health Outcomes.

Figure 2. Probability of a high score (≥80 points) for the behavioral factors component of CVH per Life's Essential 8, estimated from early childhood through late adolescence (3–18 years) among 1530 children in ECHO with respect to child sex (A), child race and ethnicity (B), maternal education (C), and annual household income (D).

Trajectories were derived from generalized linear mixed‐effects models with a logit link. In these models, the outcome was high CVH (yes/no) and the predictors included the cubic spline terms for continuous age, the sociodemographic characteristic of interest, and interactions of sociodemographic characteristic with each spline term for age. CVH indicates cardiovascular health; and ECHO, Environmental Influences on Child Health Outcomes.

Figure 3. Probability of a high score (≥80 points) for the health factors component of cardiovascular health (CVH) per Life's Essential 8, estimated from early childhood through late adolescence (3–18 years) among 1530 children in ECHO with respect to child sex (A), child race and ethnicity (B), maternal education (C), and annual household income (D).

Trajectories were derived from generalized linear mixed‐effects models with a logit link. In these models, the outcome was high CVH (yes/no) and the predictors included the cubic spline terms for continuous age, the sociodemographic characteristic of interest, and interactions of sociodemographic characteristic with each spline term for age.