The main objective of this study was to compare self-rated diet quality with a more comprehensive score of diet quality and to assess the ability of self-rated diet quality to predict adherence to healthy eating guidelines. This study also aimed to evaluate the influence of individual characteristics on the association between self-rated diet quality and the overall diet quality score. As part of the PRédicteurs Individuels, Sociaux et Environnementaux (PREDISE) study, 1045 participants (51% women) from the Province of Québec, Canada, self-rated their diet quality ("In general, would you say that your dietary habits are excellent, very good, good, fair, or poor?"). Three Web-based 24-h food recalls were completed, generating data for the calculation of the Canadian Healthy Eating Index (C-HEI) score, an overall diet quality indicator. Participants rated their diet quality as excellent (2.4%), very good (22.7%), good (49.5%), fair (20.3%), or poor (5.1%). C-HEI scores differed significantly between diet ratings, in the expected direction (p < 0.0001). Self-rated diet quality predicted adherence to healthy eating guidelines (i.e., C-HEI > 68) with a sensitivity of 44.5% and a specificity of 81.5% (C-statistic = 0.63). Sex significantly modified the association between self-rated diet quality and C-HEI score (p interaction = 0.0131); women had higher C-HEI scores than did men in the "good" and "fair" ratings. Self-rated diet quality can be useful in obtaining an overview of the diet quality of a population, but the results of this study suggest that such data should be used with caution given their poor ability to predict adherence to healthy eating guidelines. Individual characteristics may influence one's ability to appropriately self-evaluate diet quality.
Keywords: analyse du régime alimentaire; apport alimentaire; behavioural nutrition; comportement alimentaire; diet; diet analysis; dietary intake; nutrition; régime alimentaire.