Impact of memory bias in the dietary estimation of older adults measured through a food frequency questionnaire and weighed food records: A cross-sectional study in older adults and primary caregivers.

INTRODUCTION: changes in cognitive performance and memory of older adults (OA) can interfere in their reporting their diet.

OBJECTIVE: to evaluate the impact of memory bias in dietary estimation between OA and their primary caregivers (PC) through the food frequency questionnaire (FFQ) for Mexican OA and weighed food records (WFR).

METHODS: the present analysis uses the estimated dietary information based on the response provided by 51 older adults (OA) and their primary caregivers (PC) from the validation study of the FFQ for Mexicans OA was conducted during lockdowns for COVID-19. The personnel who applied FFQ and WFR were trained with standardized instruments and procedures. The Wilcoxon test was used to compare the intake per day of the foods and food groups, the Spearman correlation coefficient was used to evaluate the grams of intake per day of the food groups, and kappa coefficient was used to compare the level of food items and food groups between OA and PC.

RESULTS: in 11 of 14 food groups, no significant differences were observed between the amounts of intake reported by OA and PC. In the groups of dairy products, fruits, vegetables, and legumes, moderate agreement was observed (κ = 0.63 to 0.79), and in the rest of the groups was strong to perfect (κ ≥ 0.80). Correlation of the amount of intake between OA and PC was high in all food groups (r ≥ 0.87).

CONCLUSION: the high correlation and high agreement between the amounts and frequencies of the food groups consumed as reported by the older adults and primary caregivers indicate that the information from both respondents is reliable.