The utility of childhood and adolescent obesity assessment in relation to adult health

Abstract

Background: High childhood obesity prevalence has raised concerns about future adult health, generating calls for obesity screening of young children.

Objective: To estimate how well childhood obesity predicts adult obesity and to forecast obesity-related health of future US adults.

Design: Longitudinal statistical analyses; microsimulations combining multiple data sets.

Data sources: National Longitudinal Survey of Youth, Population Study of Income Dynamics, and National Health and Nutrition Evaluation Surveys.

Methods: The authors estimated test characteristics and predictive values of childhood body mass index to identify 2-, 5-, 10-, and 15 year-olds who will become obese adults. The authors constructed models relating childhood body mass index to obesity-related diseases through middle age stratified by sex and race.

Results: Twelve percent of 18-year-olds were obese. While screening at age 5 would miss 50% of those who become obese adults, screening at age 15 would miss 9%. The predictive value of obesity screening below age 10 was low even when maternal obesity was included as a predictor. Obesity at age 5 was a substantially worse predictor of health in middle age than was obesity at age 15. For example, the relative risk of developing diabetes as adults for obese white male 15-year-olds was 4.5 versus otherwise similar nonobese 15-year-olds. For obese 5-year-olds, the relative risk was 1.6.

Limitation: Main results do not include Hispanics due to sample size. Past relationships between childhood and adult obesity and health may change in the future.

Conclusion: Early childhood obesity assessment adds limited information to later childhood assessment. Targeted later childhood approaches or universal strategies to prevent unhealthy weight gain should be considered.

Figure 1. Receiver Operating Characteristic curves for detecting obesity at age 18 at various age- and sex-specific BMI percentile thresholds

Panel A (males) and Panel B (females) show Receiver Operating Characteristic (ROC) curves comparing the true positive rate (sensitivity) and false positive rate (1 - specificity) for various cutoffs based on CDC standardized sex-specific BMI percentiles at ages 5 and 15 years. Thresholds used to generate the ROC curves include the 3^rd, 5^th, 25^th, 50^th, 75^th, 85^th, 95^th, and 97^th percentiles.

Figure 2. The ability of child and maternal obesity assessment to discriminate future obese 18 year-olds from those who will not become obese

Panels A and B show the proportion of obese 18 year-old males and females, respectively, whose BMIs were above the 85th percentile (black filled bars) at ages 2–17 years and whose BMIs were above the 85th percentile or had mothers who were overweight or obese (white filled bars). Panels C and D show the proportion of males and females, respectively, testing positive on childhood obesity screening at ages 2–17 who become obese 18 year-olds (black filled bars: BMI above the 85th percentile; white filled bars: BMI above the 85th percentile or had a mother who was overweight or obese). (Note that for comparability, all panels in Figure 2 include only the subsample of children who also had maternal BMI measurements).