Oral Presentation ANZOS-OSSANZ-AOCO Joint Annual Scientific Meeting 2017

Understanding the limitations of dietary assessment methods and how they impact on reported dietary intake (#16)

Gilly Hendrie 1
  1. CSIRO, Adelaide, SA, Australia

Accurately measuring what we eat and drink is difficult.  Traditional methods of dietary assessment, such as 24-hour recall, have a high participant burden and are resource intensive.  Shorter assessment instruments are usually less detailed, and while some quantitative accuracy is sacrificed, in certain situations they are an appropriate and appealing option.  Technology is slowly being incorporated into existing dietary assessment tools, and new technologies may even change how we measure diet in the future.

The retrospective, self-reported nature of dietary assessment methods can introduce some degree of error or bias in terms of their accuracy to measure “true” intake.  The delivery of 24-hour recalls (telephone/face-to-face) and the sampling framework (number of weekday/weekend days) of 24-hour recall surveys can also impact on what foods and beverages are reported. 

Removing the face-to-face element of traditional methods, and delivering surveys online, may allow data to be collected from larger samples.  The CSIRO Healthy Diet Score Survey is freely available online and has been completed by over 200,000 Australians.  A survey such as this provides useful information about reported food intake, however this application of technology does not usually overcome measurement biases.  Short surveys tend to result in an overestimation of healthy foods and underestimation of discretionary foods.  However, the anonymity of answering in an online environment can reduce misreporting associated with weight status – which is problematic in data collected by face-to-face recalls.

Advances in technology means new methods, such as using photography to capture intake instead of self-report or physiological measurements as a proxy to intake, may overcome some of these limitations. However, these applications are generally not ready for use in large population studies. 

It is important to understand the limitations, and potential measurement error of dietary assessment methods, and how to deal with this error in analysis, or when working towards improving current methods to measure dietary intake.