Associations of cardiometabolic outcomes with indices of obesity in children aged 5 years and younger.

BackgroundChildhood obesity is a world-wide concern due to its growing prevalence and association with cardiometabolic risk factors in childhood and subsequent adult cardiovascular disease. In young pre-school children, there is uncertainty regarding which of the commonly used anthropometric measures of childhood obesity is best associated with cardiometabolic risk factors. This study compared the utility of common measures used in identifying obesity in these young children.MethodsThe four commonly used metrics for identifying obesity in children: body fat percentage ≥ 90th percentile, waist circumference ≥ 90th percentile, BMI z score > 2 SD and waist-to-height ratio (WHtR) ≥ 0.5, were measured in a cohort of children born singleton, at full term and followed from birth (n = 761) to 5 years of age (n = 513). The utility of each in identifying cardiometabolic risk factors (fasting lipid profile, fasting blood glucose and blood pressure) was examined.ResultsAt age 5 years, children with percent body fat ≥ 90th percentile or waist circumference ≥ 90th percentile, were associated with higher levels of triglycerides, glucose, and systolic and diastolic blood pressures than those 2 SD was associated with higher levels of triglycerides and systolic and diastolic blood pressure but not glucose at age 5 years. Differences in HDL cholesterol, fasting glucose and systolic blood pressure were observed in children with BMI z score > 2 SD at age 3 years but not with the other indices of obesity. As almost all children had WHtR ≥ 0.5 at birth, ages 1 and 3 years, this measure could not differentiate increased cardiometabolic risk. At age 5 years, the differences were much more obvious, with significant differences in triglycerides and systolic and diastolic blood pressures between those with WHtR ≥ 0.5 and those with ConclusionEach of the four commonly used measures of childhood obesity shows moderate associations with cardiometabolic risk factors at 5 years, with no advantage of one measure over the other. These associations were less consistent at 3 years of age or younger. These observations have not been reported previously