Using race as a social construct when looking at differences in racial associations between non‑alcoholic fatty liver disease and type 2 diabetes. Reply to Rollin FG, Washington C, Fluker S-A [letter].