The 24-hour recall for Senegalese weanlings: a validation exercise

Abstract

Objective: In Africa, where growth retardation is highly prevalent, the use of expensive and time-consuming diet assessment techniques is a major obstacle to the collection of dietary data on large samples of children. The 24-h recall could be a valuable alternative. Its validity and reproducibility were assessed by comparison with the precise weighing technique.

Design & subjects: Intakes of complementary foods of 45 Senegalese weanlings, aged 11-18 months, were estimated with both techniques on the same days.

Results: There was neither a level-dependent nor a systematic bias with the 24-h recall for energy and macronutrient intakes. Mean differences between techniques were < 11% of mean intakes. Rank order correlations (r) ranged from 0.60 to 0.81 for energy and macronutrients. However, the analysis by food group showed that the 24-h recall was less precise than the reference, especially for foods from the household common pot (r ranged from 0.31 to 0.61 for rice, oil and fish). In particular, measuring rice intake as a number of 'handfuls' was not satisfactory. The technique needs further improvement for these foods.

Conclusion: Since it provides unbiased estimates of weanlings' intakes, the 24-h recall can be used for diet surveys of groups of children. Its lack of precision, which could cause attenuation in epidemiological studies, can be compensated by increasing the number of days of survey. This study suggests that a precision equivalent to that of a 1-day weighed survey could be obtained with two 24-h recalls, at a considerably lower cost.

PIP: Growth retardation is highly prevalent in Africa. Given the limited available financial resources across most of the continent for the collection of dietary data on large samples of children, alternatives to expensive and time-consuming diet assessment techniques are called for. 24-hour recall may be one such alternative. The authors report their findings from an exploration of the validity and reproducibility of this approach compared to the precise weighing technique. The intakes of complementary foods were estimated for 45 Senegalese weanlings aged 11-18 months with both techniques on the same days. For each child, the study lasted four days spread over one day of qualitative observation, two days of precise weighing, and three days of 24-hour interviews. The study found the 24-hour recall approach to have neither a level-dependent nor systematic bias with regard to energy and macronutrient intakes. Mean differences between techniques were 11% of mean intakes. Analysis by food group found recall to be less precise than the reference, especially for food from the household common pot such as rice, oil, and fish. The authors conclude that the recall approach's ability to provide unbiased estimates of weanling's intakes makes it a viable assessment tool in diet surveys of groups of children. Its lack of precision can be compensated by increasing the number of days of the survey. Findings suggest that the precision of a one-day weighed survey could be obtained with two 24-hour recalls at considerably lower cost.