Electronic food records among middle-aged and older people: A comparison of self-reported and dietitian-assisted information

Aim: Nutrition-based applications ("apps") offer enormous research potential, however evidence of their use and acceptability among older adults is limited. We compared self-reported and dietitian-adjusted dietary intake records among adults aged 55 to 75 years using the Research Food Diary (RFD) app.

Methods: Participants were recruited from the 45 and Up Study and completed a 3-day food record using the RFD. A follow-up dietetic telephone interview was performed to confirm the electronic dietary data. Independent of these interviews, a set of adjustments based on dietetic skills, nutritional database knowledge, food composition and dietary assessment was established to resolve probable reporting errors. The "adjusted" and "dietitian-assisted" records were compared to self-reported records for nutrient intakes and serves of The Five Food Groups using one-way repeated measures analysis of variance.

Results: Sixty-two participants were recruited, with 48 using the RFD app which included eight records without any identified errors. Reporting errors contained in the raw self-reported records included: food items with missing/implausible quantities or insufficient descriptions to allow automatic coding. After removal of unusable records, 44 records were analysed. Differences were found between the self-reported and adjusted records for protein, calcium, vitamin B₁₂ , zinc and dairy food serves (all P < .001; differences up to 8%). No significant differences were found between the adjusted and dietitian-assisted measures.

Conclusions: Similarities between adjusted and dietitian-assisted records suggest carefully applied dietetic assumptions are likely to improve accuracy of self-reported intake data where dietitian interviews are not possible. We provide four key recommendations to guide this process.