Health Implications of Migration: Cross-Classified Multilevel Models to Disentangle Country of Origin and State of Resettlement Effects of Bodyweight (P04-106-19). (13th June 2019)

• Record Type: Journal Article
• Authors: Jones, Rebecca; Haardoerfer, Regine; Riosmena, Fernando; Cunningham, Solveig Argeseanu
• Is Part Of: Current developments in nutrition. Volume 3(2019)Supplement 1
• Publication Date: 2019-06-13
• Subjects: Nutrition -- Periodicals; Nutritional Physiological Phenomena; Nutrition; Periodicals; Periodicals; Fulltext; Internet Resources; Periodicals; 612.3
• DOI: 10.1093/cdn/nzz051.P04-106-19
• Languages: English
• ISSNs: 2475-2991

Abstract

Abstract: Objectives: Pre- and post-migration environments have differing demographic, social and economic characteristics which can affect health (Figure 1 ). What level of variance in bodyweight is attributable to individual-, country of origin-, and state of resettlement-level factors? Methods: We test what portion of the variance (as captured through intra-cluster correlations (ICCS)) in bodyweight is attributable to multi-level factors. Data come from the New Immigrant Survey (NIS), a nationally representative, longitudinal study of international migrants. The outcome of interest is BMI (kg/m 2 ) ( n = 7329). We utilize a cross-classified multi-level model approach (CCMM), where clustering in both country of origin and state of resettlement are modeled simultaneously using Bayes estimation. Results: Preliminary results are based upon the public version of the NIS, which condenses states of resettlement into 15 states/regions and countries of origin into 27 countries/regions. For the conference, access to the restricted-access dataset will allow for expansion of level-2 clusters to include specific states and countries. Table 1 presents results of initial models for country-only, state-only, and CCMM-predicted BMI. In the null model (Model 1 ), the between-level variance in BMI was driven largely by the country of origin (4.3) and not by the state of resettlement (0.04). In the CCMM, the country ICC was 7.5% and the state ICC was 0.09% indicating that state-level variance was minimal. Throughout, estimates for CCMM are closely aligned with estimates in the country-only model, further indicating that country-level variance is playing a much larger part than state-level variance in individual-level BMI after being in the country for eight years on average. Conclusions: The large share of the variance in BMI at the point of legal permanent residency in 2003 is attributable to individual-level factors. Some variance in this baseline BMI is also attributable to where an individual was born. This research helps contribute to our understanding of how environments shape health behaviors. Funding Sources: Research reported in this presentation was supported in part by the NIDDK of the NIH. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Supporting Tables, Images and/or Graphs: