Objective: To examine the comparative accuracy of telephone and in-person 24-hour dietary recall methods.
Subjects: One hundred eighty-five African-American females, aged 40 years and older, recruited from Sunday church services in Baltimore City, Md.
Methods: Participants were trained to estimate portion size with plastic food models and a 2-dimensional food recall booklet. Dietary intake was then assessed with 2 in-person 24-hour dietary recalls and 1 telephone 24-hour dietary recall, all using a computer-assisted, multiple pass approach. Results from the 2 in-person recalls were averaged and compared with the results from the telephone recall.
Statistical analyses: Cross-tabulation, paired t test, Pearson's correlation, chance-corrected agreement, and stepwise linear regression analyses were performed.
Results: There were no significant differences between the telephone and in-person methods for any nutrient. Agreement between methods was moderate for all major dietary components, with corrected correlations between methods ranging from 0.26 to 0.97 (P<.001), and kappas ranging from 0.155 to 0.372 (P<.01). Levels of low-energy reporting were high (88% telephone, 91% in-person), though there were no significant differences between methods.
Conclusions: The telephone 24-hour dietary recall method appears to be comparable to the standard in-person method among older African-American women. Portion-size training in person may make subsequent telephone dietary recalls acceptable in this population.