Sources of bias in a dietary survey of children

OBJECTIVE: To compare non-responders and responders to a dietary survey with respect to demographic variables and intention to choose selected breakfast foods, and to examine if there was any systematic change in number of food items reported during a 7 d recording period. DESIGN: Cross-sectional survey. SETTING: M?lndal, Sweden. SUBJECTS: All pupils in 5th, 7th and 9th grades in the municipality were asked to complete a questionnaire during school hours. All those present (n = 1584, 92% of total) answered questions about lifestyle factors and about intentions, attitudes and beliefs concerning high-fibre bread and milk with varying fat content. All subjects in the initial sample were asked to fill in a 7 d record of food consumed. Acceptable food records were completed by 69% of the initial participants. RESULTS: Subjects not completing the food record differed significantly from participants with respect to demographic, lifestyle and dietary factors. Dropout was more common among those who reported not usually eating breakfast and among those intending to drink whole milk for breakfast. A decline in reported food items during the recording period was also observed. CONCLUSIONS: Two sources of bias were observed here, one indicating significant differences between non-participants and participants, the other suggesting the presence of a time-dependent trend in number of recorded foods. It is likely such biases are present in other dietary surveys involving schoolchildren, and should be taken into consideration in the design, analysis and interpretation of such studies.