An examination of health selection among U.S. immigrants using multi-national data.

While migrants are widely believed to be positively selected on health, there has been very little empirical exploration of the actual health differential between migrants and non-migrants. This paper explored: 1) the extent of health selection by comparing US immigrants from 19 sending countries to their non-migrating counterparts still residing in the countries of origin; 2) country-level correlates of health selection; and 3) whether country-level health selection accounted for differences in self-rated health between immigrants and US-born Whites. We combined nationally-representative international data with data from US immigrants from the 2003-2007 Current Population Survey. The health selectivity measure was the Net Difference Index (NDI), which compares the distribution of self-rated health between migrants and non-migrants. We calculated Spearman correlation and bivariate regression coefficients between the NDI and economic, health, distance, and migration characteristics of the sending countries. We used generalized estimating equation models to examine the association between country-level health selection and immigrants' current self-rated health. We found immigrants from South America to show the most positive health selection. Health selection was significantly correlated with visa mode of entry, where family networks decrease, but work-related networks increase health selection. There was little evidence that country-level health selection explained differences in the self-rated health of US immigrants relative to US-born Whites. Our findings do not support the idea that country-level health selection underlies the "healthy immigrant effect".