Variabilities in Childhood Cardiovascular Risk Factors and Incident Diabetes in Adulthood: The Bogalusa Heart Study

Abstract

OBJECTIVE Although emerging evidence indicates that increased variability in cardiovascular risk factors (CVRFs) among populations at midlife or later is a reliable predictor of adverse health outcomes, it is unknown whether intraindividual CVRF variability during childhood or adolescence is an independent predictor of later-life diabetes. We aimed to examine the association of CVRF variability during childhood with diabetes in later life. RESEARCH DESIGN AND METHODS We included 1,718 participants who participated in the Bogalusa Heart Study and had measures at least four times during childhood (aged 4–19 years). The mean follow-up period was 20.5 years. Intraindividual CVRF variabilities during childhood were calculated using SD, coefficient of variation, deviation from age-predicted values, and residual SD based upon four to eight serial measurements in childhood. RESULTS Increased variability in BMI or HDL cholesterol (HDL-C) during childhood, irrespective of the indices used, was significantly positively associated with later-life diabetes risk independent of their respective mean levels in childhood and other possible confounding factors. In combined analysis, the magnitude of the association with diabetes risk was similar for high childhood BMI variability and high childhood HDL-C variability. After adjustments for potential confounding variables, other CVRF variabilities including systolic/diastolic blood pressure, total cholesterol, triglycerides, and LDL cholesterol were not significantly associated with diabetes. CONCLUSIONS Increased BMI and HDL-C variabilities during childhood were significant risk factors for the development of diabetes independently of diverse risk factors, which may offer new insights into the childhood origin of adult-onset diabetes.