Ethnic Enclaves, Socioeconomic Status and Associations with Obesity and Breast Density among Latinas

Obesity and high breast density(BD) are two important modifiable breast cancer risk factors that disproportionately affect Latinx women in the US. There are mixed findings regarding ethnic enclaves and neighborhood socioeconomic status(nSES) with obesity in Latinxs; literature in BD is limited. We used data on n=13,815 Latinxs > 40-year-old with at least one mammogram from the Metro Chicago Breast Cancer Registry (1980–2017), from which n=864 were linked to their residential history. Multinomial logistic regressions were estimated to evaluate associations of ethnic enclaves and nSES with obesity and BD. Further, we used Latent Profile Analysis (LPA) to characterize neighborhood composition based on tract indicators of ethnic enclaves, disadvantage, and affluence and quantified their single point in time and cumulative associations with obesity and BD. In single point-in-time analyses, residing in an enclave was associated with lower prevalence of obesity, but only for women at greater tract disadvantage. Based on our single point-in-time LPA, we identified four profiles, which were labeled as “middling”, “disadvantaged”, “ethnic enclaves”, and “affluent”. Overall, compared to the middling profile, women assigned to the disadvantaged and the enclaves’ profile showed higher mean BMI, whereas women assigned to the affluent profile had lower mean BMI. Finally, based on residential history-based LPA, cumulative exposure to enclaves was not associated with BMI, cumulative exposure to tract disadvantage was associated with higher mean BMI. However, cumulative exposure to tract affluence was associated with lower BMI. For BD, residence in an ethnic enclave measured at a single point-in-time was associated with greater prevalence of extremely dense breasts, in models adjusting for BMI, disadvantage and affluence. Enclaves’ profiles (both single point-in-time and residential history-based) were not associated with BD. These findings highlight the complex relationships among neighborhood factors, obesity, and BD. We identified opposing results in the effect of enclaves in BMI based on the approach taken (logistic regression vs. point-in-time LPA vs. cumulative LPA). Our profiles were predictive of high BMI (at both single point-in-time and residential history-based). Lastly, describing profiles that predict differences in obesity and BD provides directions in terms of health policy development to address contextually rooted determinants.