Predicting cardiovascular disease risk factors in midadulthood from childhood body mass index: utility of different cutoffs for childhood body mass index

Abstract

Identifying adults at increased risk of cardiovascular disease (CVD) on the basis of childhood body mass index (BMI) could be informative for disease prevention but depends on the utility of childhood BMI cutoffs. We aimed to establish how well the International Obesity Task Force (IOTF) and population-specific cutoffs for childhood BMI predict CVD risk factors in midadulthood. We used the 1958 British birth cohort, whose BMI measures were collected at 7, 11, and 16 y and whose CVD risk factors (obesity, hypertension, adverse lipid concentrations, and type 2 diabetes risk) were collected at 45 y. The sensitivity and specificity of IOTF and population-specific cutoffs for childhood BMI were calculated for each CVD risk factor. The prevalence of overweight or obesity was low in childhood (<11%, IOTF cutoffs) compared with that in adulthood (75% men, 56% women). The IOTF cutoffs had high specificities (91.6-97.9%) but low sensitivities (7.1-31.5%) for predicting adult outcomes. In comparison, population-specific cutoffs identified large groups of children (eg, >38% for predicting adult obesity) who had improved sensitivities (17.3-67.3%) but lower specificities (52.9-84.6%) compared with IOTF cutoffs. Accelerated BMI gains in childhood predicted adult obesity and type 2 diabetes risk, but prediction was no greater than that for childhood BMI at one age (area under the curve: 0.55-0.65 compared with 0.59-0.75). Childhood BMI and BMI gain were weak predictors of adult hypertension and adverse lipid concentrations. Neither the IOTF cutoffs nor our population-specific cutoffs for childhood BMI are adequate diagnostic tools for adult CVD risk factors in a population experiencing rapid changes in obesity prevalence over their lifetime.