Nutritional survey data are inaccurate

摘要

Micha and colleagues' paper is the latest of many articles attemptingto investigate the diets of nations.1 Its extended description of data analysis techniques may bemuse many readers. Surely authors of such statistical sophistication know what they are talking about? Butthis research suffers from the same fundamental flaw as all such studies-the poor quality of the primary data on individual food consumption. Although diet surveys vary in their methods, all rely on subjects telling researchers honestly what they eat. But such "self report" data are inaccurate. In the trade, the problem is called "under-reporting." In plain English, people lie.