PC1156: Association of Birth Weight and Weight Trajectory with Severe Obesity: A Case-control Study

Background: The prevalence of obesity has steadily increased over the past 40 years in the United States. A few studies have shown early weight trajectory is associated with future risk for obesity. However, most studies were not conducted in the United States, did not measure weight trajectory prior to age 5 years, did not adjust for socioeconomic status (SES), or did not evaluate obesity status as an adult.

Methods: This study used a nested case-control design matched 1:1 on age and gender to assess the relationship between early childhood weight-for-age trajectory and severe adult obesity, defined as a body mass index (BMI) ≥ 40 after 18 years of age. Cases and controls (BMI < 40) were from a 1976-1982 birth cohort for a single county. Maternal demographic and birth characteristics were obtained from birth certificate data. The HOUSES index (categorized into 4 quartiles) was used as a proxy for SES. Weight and height data from birth through 5.5 years was abstracted from the medical records and sex-specific weight-for-age percentiles were obtained from the 2000 Centers for Disease Control and Prevention growth charts. A nonparametric hill-climbing algorithm and the Calinski Harabatz criterion were used to identify distinct homogeneous clusters of weight-for-age trajectories among all cases and controls combined. Conditional logistic regression models were fit to assess associations with case-control status.

Results: Among the 785 cases with severe adult obesity, the mean age at first adult documented BMI ≥ 40 kg/m2 was 30.3 (SD 6.3) years and 39.4% were male. Univariate analysis showed that the average maternal age at birth was significantly younger for the cases who had severe adult obesity (25.0 years) compared with controls (26.4 years). In addition, the odds of a maternal education less than a college degree (vs. a college degree or higher) was significantly higher for cases compared to controls, as was the odds of being in the HOUSES quartiles with the lowest SES (Table). Birth weight was not associated with adult obesity.

A total of 737 matched sets of cases and controls had at least two weight measurements per person for the trajectory analysis. A 2-cluster trajectory solution was identified using weight-for-age percentiles (Figure); the odds of being in cluster 1 was significantly higher for cases compared to controls (OR 1.99, 95% CI 1.60-2.47) with 61.9% of the cases in cluster 1 compared to 45.3% of the controls. The association between cluster membership and case-control status persisted after adjusting for maternal age and education (adjusted OR 2.08, 95% CI 1.66-2.61).

Conclusion: Our data suggests that early childhood weight-for-age trajectories are associated with severe BMI status in adult life. Our results add to growing evidence that is critical to prevent excess early childhood weight gain.