The Use of Pictorial Recall Aids Modifies Dietary Assessment Results: Experiences from Quantitative 24-hour Dietary Recalls of Young Children in Nepal and Senegal

Recall bias or memory error is a known limitation of recall-based dietary assessment methods and can lead to substantial misrepresentation of diets. Pictorial recall aids have been used by various studies to mitigate recall bias, but their effect on quantitative, 4-pass 24-h dietary recall (24HR) results has not yet been measured. This article examines uptake of pictorial recall aids among caregivers of young children in Kathmandu Valley, Nepal, and Guédiawaye Department, Senegal; describes what foods/beverages are subject to recall bias in these contexts; and assesses the effect of these recall aids on estimated dietary outcomes. We used cross-sectional data from 24HR surveys that provided a recall aid to reduce caregiver recall bias. Young children’s dietary outcomes were calculated from the 24HR data. Logistic regressions were used to compare characteristics and estimated dietary outcomes of children by recall aid uptake. McNemar’s test and paired t-tests were used to compare estimated dietary outcomes in initial 24HR data with revised 24HR data that incorporated omitted items identified by the recall aid. Pictorial recall aids had relatively high uptake by caregivers in both contexts. Beverages, unhealthy snacks, and fruit were most subject to recall bias. Adding omitted items identified by recall aids resulted in statistically significant changes in most dietary outcomes assessed. Children of caregivers who chose to use the recall aid had significantly different estimated dietary outcomes than children of caregivers who did not use the recall aid. The use of pictorial recall aids modifies 24HR results in these 2 contexts. Rigorous research is needed to fully understand the impact of recall aids on 24HR methodology and results, with an aim to improve the accuracy of dietary assessment and ultimately inform evidence-based nutrition programming, policy, and recommendations.