Food-related environmental, behavioral, and personal factors associated with body mass index among urban, low-income African-American, American Indian, and Caucasian women.

PURPOSE: To examine racial/ethnic differences in relationships between food-related environmental, behavioral and personal factors and low-income women's weight status using Social Cognitive Theory (SCT) as a framework.
DESIGN: Cross-sectional survey.
SETTING: Community sites and low-income housing developments in the Twin Cities metropolitan area.
SUBJECTS: Low-income African-American, American Indian, and Caucasian women ≥18 years old (n  =  367). MEASURES: Participants completed a survey including demographic, food security, and theoretically based questions. Heights and weights were measured to determine body mass index (BMI). ANALYSIS: Data were split by race/ethnicity and reduced by examining Pearson coefficients for SCT survey questions significantly correlated with BMI (p < .05). Separate environmental, behavioral, and personal multiple linear regression models for each racial/ethnic group were run to explore the proportion of variance in BMI explained by each SCT construct and which questions were significant predictors.
RESULTS: All regression models were statistically significant, although the personal regression models predicted the greatest proportion of the variance in BMI for African-American (15% of the variance), American Indian (22% of the variance), and Caucasian women (37% of the variance).
CONCLUSION: Effective nutrition education and intervention efforts to control the obesity epidemic among urban, low-income women may call for a tailored approach with noted consideration of their racial/ethnic identity. Although broader changes to the food environment are necessary, the importance of addressing personal factors such as nutrition knowledge, self-efficacy, and emotional coping responses to stress, in the context of income constraints, food insecurity, and health beliefs, is also implicated.