Salmon bias effect as hypothesis of the lower mortality rates among immigrants in Italy

Compared with natives, immigrants have lower all-cause mortality rates, despite their lower socioeconomic status, an epidemiological paradox generally explained by the healthy migrant effect. Another hypothesis is the so-called salmon bias effect: “statistically immortal” subjects return to their country of origin when they expect to die shortly, but their deaths are not registered in the statistics of the country of residence. This underestimation of deaths determines an artificially low immigrant mortality rate. We aimed to estimate the potential salmon bias effect on differences in mortality rates between Italians and immigrants. We used a national cohort of all Italians registered in the 2011 census and followed up for mortality from 2012 to 2016. Mortality data were retrieved from the Causes of Death Register, which included all deaths occurring in the country and the Resident Population Register, which collects also the deaths occurring abroad. We assumed as a possible salmon bias event the death of an immigrant resident in Italy that died in his/her country of origin. Considering the deaths occurring in the country of origin, we observed an 18.1% increase in the overall mortality rates for immigrants and an increase of 23.7% in the age-standardized mortality rate. Mortality rates of immigrants resident in Italy, calculated without taking into account the deaths occurring in the country of origin, are certainly underestimated. However, the salmon bias only partly explains the difference in mortality rates between immigrants and Italians.