Development and Validation of a Short Sport Nutrition Knowledge Questionnaire for Athletes

Abstract

Weak evidence exists on the relationship between nutritional knowledge and diet quality. Many researchers claim that this could be in part because of inadequate validation of the questionnaires used. The aim of this study was to develop a compact reliable questionnaire on nutrition knowledge for young and adult athletes (NUKYA). Researchers and the sport clubs medical staff developed the questionnaire by taking into consideration the latest athlete dietary guidelines. The questionnaire content was validated by a panel of 12 nutrition experts, and finally tested by 445 participants including athletes (n = 264), nutrition students (n = 49) and non-athletes with no formal nutrition knowledge (n = 132). After consulting the experts, 59 of the 64 initial items remained in the questionnaire. To collect the evaluation of experts, we used the content validity index, obtaining high indices for relevance and ambiguity (0.99) as well as for clarity and simplicity (0.98). The final questionnaire included 24 questions with 59 items. We ensured construct validity and reliability through psychometric validation based on the Classical Test Theory and the Item-Response Theory (Rasch model). We found significant statistical differences comparing the groups of nutrition knowledgeable participants with the rest of the groups (ANOVA p < 0.001). We verified the questionnaire for test-retest reliability (R = 0.895, p < 0.001) and internal consistency (Cronbach's α=0.849). We successfully fit the questionnaire data to a rating scale model (global separation reliability of 0.861) and examined discrimination and difficulty indices for items. Finally, we validated the NUKYA questionnaire as an effective tool to appraise nutrition knowledge in athletes. This questionnaire can be used for guiding in educational interventions, studying the influence of nutrition knowledge on nutrient intake and assessing/monitoring sport nutritional knowledge in large groups.

Keywords: Rasch model; athletes; classical test theory; item response theory; nutrition knowledge; reliability; sports nutrition; validity.

Figure 1

(Left) Item difficulty indices for each questionnaire item. Ninety percent of the items are within the range $(0.2,0.8)$. Items outside of that range are highlighted in red. (Right) Item discrimination indices for the same items. All questions positively discriminate high-scored respondents from low-scored ones.

Figure 2

Person–item maps for the fitted rating scales model. (Top) The questionnaire items (y-axis) sorted according to the value they take in the latent dimension (x-axis, negative to positive means easier to more difficult). (Bottom) The person measure distribution. Items should ideally be located along the whole scale to meaningfully measure the ‘ability’ of all persons.

Figure 3

Scatter plot of questionnaire scores by group.