Abstract C077: What do different modeling approaches tell us about the association between migration history and midlife body size and obesity in Latina women?

Abstract

Abstract General and central obesity, excess weight and body fat constitute potentially modifiable risk factors for several cancer sites and are important targets for cancer prevention efforts. The prevalence of obesity continues to increase in the United States (U.S.) and many parts of the world. In Latino populations, immigration to the U.S. and subsequent acculturation are accompanied by an increase in body size and obesity risk; however, evidence for Latinos of non-Mexican origin is limited. We examined the associations of birthplace, migration age, and percent of life in the U.S. with body size measures in a sample of 787 self-identified Latina women of predominantly Caribbean heritage (79% born in the Dominican Republic or Puerto Rico), recruited from a mammography clinic in New York City (ages 40-65 years, 53% monolingual Spanish speakers, 31% immigrated to the U.S. before age 20). We collected in-person interview data and measured women’s height, weight and waist circumference, which were used both as continuous and categorical variables to assess general obesity/body size using BMI (kg/m2, general obesity ≥ 30 vs. < 30) and central body fat composition/central obesity using waist circumference (in cm, ≥ 88 vs. < 88). We conducted multivariable regression analysis using linear regression and quantile regression methods for continuous body size measures and relative risk regression for risk of obesity. The results of linear and relative risk regression models showed lower BMI and waist circumference as well as lower risk of general and central obesity associated with foreign birthplace, later age at migration and less time in the U.S. Results from quantile regression revealed differences by percentiles with statistically significant inverse associations that were limited to the upper quantiles (≥ 75th percentile) of BMI and waist circumference. For instance, as compared with U.S.-born Latina women, BMI was lower for women who had spent < 50 % of life in the U.S., migrated to the U.S. at ages ≥ 30 years and were born in the Dominican Republic at the 75th percentile of BMI (β= -4.10, 95% CI: -6.75, -0.81; β= -4.34, 95% CI: -7.07, -0.88; β= -3.72, 95% CI: -6.41, -0.43, respectively), but no BMI differences by migration history were observed at the 25th percentile of BMI (β= 0.04, 95% CI: -1.01, 0.96; β= 0.04, 95% CI: -0.87, 1.08; β= 0.08, 95% CI: - 0.94, 1.14, respectively). In conclusion, our results using different modeling approaches provide support that migration to the U.S. is associated with larger body size and risk of obesity in midlife Latina women from Caribbean heritage, as observed for other Latina subgroups. Through the use of quantile regression methods, we provide new data highlighting that migration influences on body size may be stronger and most consistent at the higher distribution of BMI and waist circumference. Citation Format: Carmen B Rodriguez, Shweta Athilat, Parisa Tehranifar. What do different modeling approaches tell us about the association between migration history and midlife body size and obesity in Latina women? [abstract]. In: Proceedings of the Twelfth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2019 Sep 20-23; San Francisco, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl_2):Abstract nr C077.