Obesity in the context of migration and socio-economic risk factors – a multivariate epidemiologic analysis

Abstract

Purpose: While the existence of a healthy migrant effect remains controversial, overweight and obesity are considered a global pandemic. Migrants seem to be affected more often, however most of the few existing studies did not differentiate a first-generation from a second-generation migration background and/or did not control common socio-demographic confounders. This study aims at examining the influence of first- and second-generation migration on body mass index (BMI) and obesity in Germany. Subjects and Methods: We conducted a controlled observational study based on a survey of n = 64,089 participants of the German Socio-Economic-Panel. Missing values were multivariate imputed via chained equations (MICE). The influence of migration on BMI and obesity was assessed by comparing first-generation, second-generation and non-migrants to each other. Pairwise statistical testing was done by t tests and Fisher's exact tests. For the multivariate analysis, OLS and logistic regression models and its coefficients (beta, odds ratio) were used. Targeting multiple testing, Holm-correction was utilized. Results: Within the bivariate analysis, all three group-specific mean BMI-values differ significantly from each other (P < .001). The pairwise differences regarding the obesity risk are also significant (P < .001). Within the multivariate analysis, only second-generation migration reveals an influence on BMI compared to first-generation migration (ß = 0.297; 97.5% CI: 0.127–0.467) and non-migrants (ß = 0.366; 98.33% CI: 0.103–0.628). This is equivalent to its influence on obesity versus first-generation (odds ratios = 1.220; 98.33% CI: 1.045–1.423) and non-migrants (odds ratios = 1.134; 97.5% CI: 1.018–1.262). Conclusions: After controlling socio-demographic confounders, a second-generation migration background but not a first-generation migration background is associated with a higher BMI and obesity.