I read with interest the recent paper by Kell et al.( 1 ) Nevertheless, I was surprised that these authors failed to address the well-established fact that the memory-based dietary assessment methods they used produce data that are frequently physiologically implausible (i.e. lack credibility or validity) and are often incompatible with life( 2 – 4 ).
Given the ubiquity of implausible dietary data( 2 – 4 ) and the existence of validated cut-off points to ascertain the credibility of dietary reports( 5 ), I question why Kell et al.( 1 ) failed to use the accepted empirical approach. This error is particularly problematic, given that these authors were examining patterns of diet. The well-established differential misreporting of particular foods and beverages suggests that dietary patterns (i.e. macronutrient and micronutrient consumption) are differentially and unpredictably misreported when total energy intake is physiologically implausible. This non-uniformity leads to non-quantifiable nutrient-specific errors.
Stated simply, implausible dietary data cannot be used to examine patterns of dietary consumption because it is impossible to ascertain what percentage of the reported foods and beverages are completely false memories( 6 , 7 ), intentional misreporting (i.e. lies( 8 )), grossly inaccurate estimates( 4 ) or somewhat congruent with actual consumption( 2 ). Despite the intuitive appeal of Kell et al.’s( 1 ) conclusions, it does not appear that they are supported by credible scientific evidence.
E. A. is funded by the National Institute of Diabetes and Digestive and Kidney Diseases, grant number T32DK062710. E. A. reports receiving honoraria (i.e. speaking fees) from the International Life Sciences Institute, The Coca Cola Company and the World Sugar Research Organization. The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health.