Validation of Toddler NutriSTEP®

The objective was to establish criterion and construct validity for Toddler NutriSTEP®, a 17‐item, community‐based parent‐administered nutrition risk screening questionnaire (total score = 68), adapted from the original preschooler version. Parents of 200 toddlers (18–35 mo), recruited from community settings, completed Toddler NutriSTEP® and a demographic questionnaire; an assessment by a registered dietitian (RD) included medical and nutritional history, weight, height, 3‐d food records. The RD rated nutritional risk on a scale of 1 (low) to 10 (high). Receiver operating curves (ROC) were used to determine sensitivity and specificity for various cut‐points on the scale. Differences in Toddler NutriSTEP® scores for selected demographic variables were determined by ANOVA. Scores on Toddler NutriSTEP® and the RD rating were correlated (r=0.66, p<0.000). Areas under the ROC curves for the moderate risk (score 5–7) and the high risk (score 8+) RD ratings were 0.87 and 0.82 with cut points of >;20 & >;25, respectively. Scores on Toddler NutriSTEP® were significantly higher (more risk) for parents with low incomes, low education and for immigrants to Canada. The Toddler NutriSTEP® questionnaire demonstrates both criterion and construct validity in children 18 – 35 months. Funding: Canadian Institutes of Health Research