Objectives: To identify misreporting among older rural adults using a prediction algorithm and to compare dietary patterns of underreporters and plausible reporters.
Design: In this cross-sectional study, diet information was assessed by five 24-hour recalls collected over 10 months. All foods were classified into 24 food subgroups. Demographic, health, and anthropometric data were collected via home visit.
Setting: Rural Pennsylvania.
Subjects: One hundred seventy-nine community-dwelling adults, aged 66 to 87 years.
Statistical analysis: Cluster analysis.
Results: Underreporters (n=43) were more likely than plausible reporters (n=133) to be overweight and less educated but did not differ by sex. Underreporters consumed fewer servings across the majority of food groups. Two dietary patterns were determined for all and plausible reporters, in both cases one of higher and one of lower nutrient density. Using only plausible reporters to determine dietary patterns was very similar to using all reporters. The correlation between energy intake and weight status was improved for plausible-reporting women, but not men.
Conclusions: Dietary patterns of plausible reporters were generally similar to that of all reporters; however, correlations with energy intake and weight status improved for women using only plausible reporters. Individuals may not accurately report dietary intake. Those obtaining diet reports should be aware of reporting errors before making decisions about dietary adequacy.