Reliability and validity of a healthy diet determinants questionnaire for adolescents

Abstract

Objective: To describe the development of a short questionnaire including a wide range of personal, social and environmental constructs in relation to a healthy diet in an adolescent population.

Design: The questionnaire was developed based on the literature, past experiences and feedback from a small pilot study (n 10). Test and retest (2 weeks later) of the questionnaire was done to investigate test-retest reliability. Data of four non-consecutive 24 h recalls were collected to investigate the predictive validity with food (fruit, vegetables, milk, snacks, soft drinks), nutrient (fibre, ascorbic acid, Ca, percentage energy from fat) and energy intakes.

Setting: At home, in the presence of a student.

Subjects: Convenience sample of fifty-five Belgian-Flemish adolescents approached by university students for course credits.

Results: Test-retest correlations of the constructs ranged between 0.51 and 0.78. Eleven of the eighteen final constructs were significantly associated with one or more of the five selected food items, all in the expected direction. Most significant associations with the food items were found for taste, perceived peers' behaviour and availability of soft drinks at home. Fresh fruit and soft drinks were correlated with most constructs (seven) followed by snacks (five). Concerning energy and the selected nutrient variables, the findings were less clear.

Conclusions: The results are promising: the test-retest stability was moderate to good; most of the psychosocial constructs were significantly associated with one or more of the selected dietary variables. Future work in a larger sample of European adolescents is warranted.