Infant BMI or Weight-for-Length and Obesity Risk in Early Childhood

Abstract

Background: Weight-for-length (WFL) is currently used to assess adiposity under 2 years. We assessed WFL- versus BMI-based estimates of adiposity in healthy infants in determining risk for early obesity.

Methods: Anthropometrics were extracted from electronic medical records for well-child visits for 73 949 full-term infants from a large pediatric network. World Health Organization WFL and BMI z scores (WFL-z and BMI-z, respectively) were calculated up to age 24 months. Correlation analyses assessed the agreement between WFL-z and BMI-z and within-subject tracking over time. Logistic regression determined odds of obesity at 2 years on the basis of adiposity classification at 2 months.

Results: Agreement between WFL-z and BMI-z increased from birth to 6 months and remained high thereafter. BMI-z at 2 months was more consistent with measurements at older ages than WFL-z at 2 months. Infants with high BMI (≥85th percentile) and reference WFL (5th-85th percentiles) at 2 months had greater odds of obesity at 2 years than those with high WFL (≥85th percentile) and reference BMI (5th-85th percentiles; odds ratio, 5.49 vs 1.40; P < .001). At 2 months, BMI had a higher positive predictive value than WFL for obesity at 2 years using cut-points of either the 85th percentile (31% vs 23%) or 97.7th percentile (47% vs 29%).

Conclusions: High BMI in early infancy is more strongly associated with early childhood obesity than high WFL. Forty-seven percent of infants with BMI ≥97.7th percentile at 2 months (versus 29% of infants with WFL ≥97.7th percentile at 2 months) were obese at 2 years. Epidemiologic studies focused on assessing childhood obesity risk should consider using BMI in early infancy.

FIGURE 1

Bar graph indicates the proportion of discordant measurements between WFL and BMI at each visit age, with 95% CI indicated by error bars.

FIGURE 2

A, Raw growth data were plotted by gender and adiposity classification group at age 2 months (gray box), and mean BMI for each group is shown over time. Number of subjects per gender and adiposity classification group on the basis of 2-month classification is shown in the figure text and varies across time. B, Raw growth data were plotted by gender and adiposity classification group at age 2 months (gray box), and mean length for each group is shown over time. Number of subjects per gender and adiposity classification group on the basis of 2-month classification is shown in the figure text and varies across time. C, Raw growth data were plotted by gender and adiposity classification group at age 2 months (gray box), and mean weight for each group is shown over time. Number of subjects per gender and adiposity classification group on the basis of 2-month classification is shown in the figure text and varies across time.