Healthy Weight, Healthy Children: Understanding Weight Inequity across the Socioeconomic Gradient

Background. Low-resourced communities typically lack equal access to healthy foods and safe places for physical activity, which contribute to a social injustice in these communities: disparities in obesity risk. Childhood obesity is a health issue associated with increased risk of long-term health consequences. Little is known about variations in individual, household, and environmental predictors of overweight across gradients of social class. This study addresses this gap using the ecological framework of childhood obesity to assess how characteristics vary across socioeconomic gradients (SES).

Methods. Child, parent, and community characteristics were examined within SES quintiles in a national sample (Early Childhood Longitudinal Study-Birth Cohort). Data were collected from 9,850 mothers/toddlers via in-home and self-administered assessments. The sample was stratified by SES quintile (Q1-low, medium-low, medium, medium-high, Q5-high). Taylor series linearization was applied to logistic regression models within quintiles using toddler weight status as the dependent variable (overweight/normal).

Results. Higher rates of overweight were observed in Q1 (29.2%), and reduced as SES increased (16.9% - Q5). Regression results showed child characteristics of race/ethnicity were significant in Q1 and Q4. Compared to Caucasians, African Americans had decreased odds of overweight (55%) in Q1, and Hispanics and those of an “other” race/ethnicity had decreased odds of overweight in Q4 (47% and 60%, respectively). Better motor development was associated with increased odds of overweight in Q1 (71%) and Q2 (43%); better mental development was associated with decreased odds of overweight in Q2 (22%). Both Q1 (32%) and Q5 (36%) showed decreased odds of overweight for females.

Parameter estimates for parent characteristics also varied across quintiles. In Q1, being unmarried predicted increased odds of overweight. Normal maternal BMI was associated with decreased odds in both Q1 (29%) and Q2 (51%). Lack of employment was associated with increased odds of overweight in Q2, whereas parental feeding practices were significant across 3 quintiles. In Q2 (55%), Q4 (169%), and Q 5 (77%), solid foods <4 months was associated with increased odds of overweight. Being put to bed with no bottle was associated with decreased odds of overweight in Q2. More children in the home was associated with increased odds of overweight in Q1, and having more adults in the household was associated with increased odds of overweight in Q5. At the community level, in Q1, not having WIC/SNAP predicted increased odds of overweight (81%), and higher community engagement reduced odds of overweight (17%) in Q2.

Conclusions. Using an ecological and social gradient of health framework, this study indicates there are distinct predictors of overweight within each SES quintile, which yields important implications for public policy, intervention developers, and researchers. By understanding the dynamics of obesity across SES and using a social justice lens, social workers can assist in the development of integrative, multifaceted interventions to address the social structures that perpetuate oppression and limit access to resources for optimal health.  Existing policies and practices should be reformed to address the specific social determinants of health for families at each level of SES.