Cross-cultural measurement equivalence of the Healthy Eating Index adapted version for children aged 1–2 years

• Published in: British journal of nutrition Vol. 126; no. 5; pp. 782 - 789
• Main Authors: Ribas, S. A., Santos, D. M. G., Rosa, G. P. C., Teixeira, M. T., Rodrigues, L. G., Marques, E. S.
• Format: Journal Article
• Language: English
• Published: Cambridge, UK Cambridge University Press 14-09-2021
• Subjects: Adaptation; Age groups; Babies; Birth weight; Brazil; Cereals; Children; Children & youth; Correlation coefficient; Correlation coefficients; Cross-Cultural Comparison; Cross-Sectional Studies; Dairy products; Diet; Diet, Healthy; Dietary Surveys and Nutritional Epidemiology; Eating; Eating behavior; Energy consumption; Equivalence; Evaluation; Food; Food consumption; Food intake; Food quality; Humans; Infant; Infant, Premature; Infants; Legumes; Meat; Milk; Neonates; Newborn babies; Nutrition; Nutrition Policy; Obesity; Pediatrics; Premature birth; Principal components analysis; Psychometrics; Reliability analysis; Reproducibility of Results; Surveys and Questionnaires; Brazil; Validity; Children; Diet; Reliability; Quality
• ISSN: 0007-1145; 1475-2662
• DOI: 10.1017/S0007114520004729

The objectives of this study were to evaluate the cross-cultural measurement equivalence of the Healthy Eating Index (HEI) for children aged 1–2 years and to analyse the quality of nutrition of preterm infants. This was a cross-sectional study with 106 premature infants attended in two specialised outpatient clinics of university hospitals. The quality of the diet was analysed through an adapted HEI to meet the dietary recommendations of Brazilian children aged 1–2 years. Food consumption was measured by 24-h recalls. The reliability of the instrument was evaluated by internal consistency analysis and inter-observer reliability using Cronbach’s α coefficient and κ with quadratic ponderation. The construct validity was evaluated by principal component analysis and by Spearman’s correlation coefficient with total energy and consumption of some groups’ food. The diet quality was considered adequate when the total HEI score was over 80 points. Cronbach’s α was 0·54. Regarding inter-observer reliability, ten items showed strong agreement (κ > 0·8). The item scores had low correlations with energy consumed (r ≤ 0·30), and positive and moderate correlation of fruit (r 0·67), meat (r 0·60) and variety of diet (r 0·57) with total scores. When analysing the overall quality of the diet, most patients need improvement (median 78·7 points), which can be attributed to low total vegetable intake and the presence of ultraprocessed foods in the diet. The instrument showed auspicious psychometric properties, being promising to evaluate the quality of the diet in children aged 1–2 years.