Cognitive ability is associated with suspected reporting errors on food frequency questionnaires

Objective: To examine potential for bias in reported total energy intake on a Food Frequency Questionnaire (FFQ) among older adults. Design: Longitudinal cohort study. Subjects/setting: 2,706 Community-dwelling Black and White older adults, aged 70-79 years, enrolled in the Health, Aging, and Body Composition study. Multivariate logistic regression analyses were conducted with potential errors on reported total energy intake on the Food Frequency Questionnaire (FFQ) as the outcome variable and with cognitive ability, measured by the Modified Mini Mental State Exam (3MS) as the primary independent variable. The regression model controlled for site, race, gender, age, body size, and physical activity. Separate models were fit using 3MS as a continuous variable and for multiple 3MS cutpoints. All models revealed similar findings. Results: Cognitive ability was inversely associated with potential errors in reporting total energy intake, whereby a five-point increase in 3MS scores was associated with a 14% decreased likelihood of reporting errors (Odds Ratio=0.86, 95% Confidence Interval: 0.77, 0.95). Additionally, compared to White women, White men were 2 times more likely, and Black women and Black men were 3 times more likely, to have errors in reporting total energy intake. Conclusion: This study provides evidence that for older adults, lower cognition scores are associated with increased potential errors in reporting total energy intake. Applications: Dietary reporting from older adults may be inaccurate due to cognitive deficits. A brief assessment of cognitive function may assist clinicians in dietary evaluations and recommendation and may benefit studies using FFQ data where the measure of cognitive function could be utilized to stratify data analyses and conduct sensitivity analyses. The Journal of Nutrition, Health & Aging