Estimating the Effects of U.S.-Mexico Migration on Elevated Depressive Symptoms: Evidence From Pooled National Cohorts

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

Estimates of the health effects of migration may be biased due to migration selectivity, whereby migrants systematically differ from their non-migrant counterparts on pre-migration characteristics (e.g., education) that also influence post-migration health outcomes. Time-dependent mediator-confounders, which influence migration probabilities but are also mechanisms by which migration may influence health, make it difficult to account for selection without controlling for mediators. Pooling data from the Mexican Health and Aging Study (N=17,628) and Mexican-born participants in the U.S. Health and Retirement Study (n= 898), we evaluated effects of migration (at any age and in childhood, adolescence, or adulthood) to the U.S. on elevated depressive symptoms. We used covariate-adjusted generalized estimating equations, with inverse probability weighting to account for age-specific migration selection. Migration to the U.S. was unrelated to depressive symptoms among men. Among women, there was some evidence that migration in early adulthood protected against depressive symptoms in late life.

See paper

 Presented in Session 133. Race, Ethnicity, and the Demography of Mental Health