Categorizations of migrants and ethnic minorities—are they useful for decisions on public health interventions?

Decisions regarding population-based preventive interventions require convincing measures of health and risks. Is it justified to initiate special community programs on diabetes prevention among migrants on the basis of a higher prevalence of diabetes than in the non-migrant population? Even as we know that diabetes is not a problem for the majority of the migrants? And that diabetes is also a problem for many non-migrants—however less prevalent? Relative risks and differences in prevalence of risks and diseases between groups are often used to justify such new programs for selected groups and communities based on certain characteristics such as ethnicity, migrant status, family situation or socio-economic position. Mulinari et al.1 question the use of broad categorizations as instruments for predicting individual health problems using the area of ethnicity, migration and health as an example, and warn against the practice of only including measures of association in the consideration of public health interventions. Instead, these kinds of measures should always be reported together with measures of discriminatory accuracy. Their study …