Estimating the Effects of Migration From Mexico to the United States on Self-rated Health and Mortality: The Hispanic Paradox Revisited

Adina Zeki Al Hazzouri
Lanyu Zhang, University of Miami
Audrey Murchland, University of California, San Francisco
Leslie Grasset
Jacqueline Torres, RWFJ Health & Society Scholars, University of California, San Francisco / University of California
Richard Jones, Brown University
Maria Glymour, University of California, San Francisco

Much of the available evidence on the health effects of migration to the US is derived indirectly by comparing migrants residing in the US with US-born individuals, but few studies compare migrants to non-migrants who remained in the country of origin. Merging harmonized and nationally representative datasets from Mexico and the US – the Mexican Health and Aging Study (MHAS,N=18,302) and the Health and Retirement Study (HRS,N=924), we estimated cumulative probability of migration and calculated stabilized inverse probability treatment weights (IPTW). We estimated effects of migration on self-rated health at study enrollment (mean age=60.8), using logistic regression, and on incident mortality, using Cox proportional hazards regressions, with and without IPTW. In demographic-adjusted models, migration predicted less fair/poor self-rated health (OR=0.87, 95% CI: 0.80, 0.96) but the association with mortality was inconclusive (HR=0.92, 95%CI: 0.83, 1.02). After accounting for migration selection forces, migration appeared to benefit self-rated health but not mortality.

See extended abstract

 Presented in Session 7. Migration & Urbanization