Evaluation Of Dietary Assessment Tools: Does ‘Validated’ Mean What We Think It Means?

Abstract

Evaluation of the extent to which self‐report dietary assessment tools accurately capture true intake is critical to informing the collection of dietary intake data that can advance our understanding of diet and health. Such evaluation is challenging due to the lack of unbiased reference measures for dietary intake, forcing reliance on self‐report (and thus error‐prone) measures as comparators. Depending on the type of instrument and its intended use, evaluation is also complicated by the need to test tools in different settings and populations. The objective of this work is to examine challenges in evaluating dietary assessment tools and the ways in which the findings of validation studies are interpreted. We draw upon a scoping review of Canadian studies of free‐living adult populations that included an assessment of dietary intake to illustrate the extent and ways in which validity of tools are addressed. Of 58 studies that used a frequency questionnaire or screener to assess diet, tools were reported to have been validated for the given study sample in a small subgroup. In close to two‐thirds, authors reported that the tool has been validated, in some cases with few details on how it was evaluated and in what population(s). In close to a third, there was little attention to whether tools used have been tested. Given that self‐report instruments cannot be expected to perfectly capture true intake, more nuanced discussions of what it means for a tool to be ‘valid’ as well as the implications of limitations in tools for study results appear warranted. Such discussions are key to improving the quality of self‐reported dietary intake data.Support or Funding InformationThis scoping review was supported by the Canadian Institutes of Health Research (137238) and Canadian Cancer Society Research Institute (702855).