Studies on health selectivity around the time of migration that account for different sources of heterogeneity in results, such as different types of migrants, internal and external migration, and places of origin and destination, are lacking.

In this population-based data linkage study in the province of Manitoba, Canada (1985-2023), we matched outmigrants and returnees with stayers within three birthplace groups of Manitoba-born (411,290), interprovincial immigrants (71,393) and international immigrants (101,192). We compared hospitalizations before outmigration and hospitalizations and mortality after return migration, stratified by the destination of leavers and the origin of returnees.

During the year before outmigration, compared to those who did not migrate (stayers) Manitobans who migrated to another country had half the risk of being hospitalized (Relative Risk (RR): 0.48; 95% confidence interval (CI): 0.40-0.58) but those who migrated to another province did not. Interprovincial immigrants who returned to their province of origin had higher hospitalizations than stayers (RR: 1.21; 95% CI: 1.12-1.32) but those who remigrated did not. In the three years after return migration, returnees had lower risk of death compared to stayers in all three groups. The association was stronger for Manitobans who returned from another country (RR: 0.46; 95% CI: 0.27-0.80) than for those who returned from another province (RR: 0.83; 95% CI: 0.74-0.93).

Health selectivity varies according to the birthplace of migrants and, within them, to the places of destination at outmigration and of origin at return migration. The more extreme the migration (international more extreme than interprovincial) the stronger the health selectivity.