Regan Bailey

Ph.D., M.P.H., R.D.

Purdue University

Associate Professor

IDATA (Learn more about this study)

IDATA-41

Feb 11, 2021

Levering Interactive Diet and Activity Tracking Study Data to Inform Precision Dietary Assessment

Accurately measuring dietary exposures through self-report are notoriously difficult to measure accurately and reliably due to measurement error. Traditional methods of dietary assessment include diet histories, food records (FR), food frequency questionnaires (FFQ), and 24-hour recalls (24HR). The 24HR has traditionally been interviewer administered over the phone or in person (Casey, Goolsby et al. 1999); but 24HR are also collected in person or online (e.g. Automated-Self-Administered (ASA-24) (Subar, Kirkpatrick et al. 2012)). The use of ASA-24 in general reduces interviewer burden and costs. Several screening tools and diet history methods also exist. Not all methods measure diet optimally across population subgroups and all methods are subject to measurement error. Random error can be mitigated with statistical models with replicate data. Systematic error cannot be mitigated While all methods of dietary assessment have systematic error that tend to be in the direction of energy underreporting, the 24HR is that least biased estimator of energy intake at present (Subar, Kipnis et al. 2003). Previous research indicates pervasive errors in self-reported energy intakes (Block and Hartman 1989, Bingham 1991, Black, Prentice et al. 1993, Schoeller 1995, Beaton, Burema et al. 1997, Kaaks and Riboli 1997). Under-reporting of energy intake ranges from 10 to 50% lower than estimated caloric needs in validation studies using doubly-labeled water (Schoeller 1990, Schoeller 1995, Jonnalagadda, Mitchell et al. 2000, Subar, Kipnis et al. 2003). The challenges with doubly-labeled water are that it is expensive and technically difficult to perform (Speakman 1998) and subjects must collect multiple urine specimens (Gibson 2005). Recovery biomarkers, like doubly-labeled water, exhibit a direct relationship with food components consumed, but are limited to energy, potassium, sodium, and protein (Subar, Freedman et al. 2015).
This Interactive Diet and Activity Tracking in AARP (IDATA) study was designed to evaluate the structure of measurement error in selected dietary assessment tools and has replicate measurements for dietary data as well as doubly-labeled water and urinary potassium, sodium, and protein. For the planed studies, this consortium of experts in dietary assessment and measurement error modeling will work collaboratively to leverage iData, along with other similar study data in more diverse populations, to gather preliminary data on strategies and tools that help to help provide data to advance precision dietary assessment.