Development of a Population-Level Dichotomous Indicator of Minimum Dietary Diversity as a Proxy for Micronutrient Adequacy in the United States Adolescents Aged 10–19 Y

Abstract

BackgroundDiversity is a key component of diet quality and health, but no indicator exists for adolescents under the age of 15 y. ObjectivesTo establish a dichotomous indicator for population-level assessment of adolescent dietary diversity as a proxy for micronutrient adequacy. MethodsWe used the probability approach to construct the mean probability of adequacy (MPA) of 11 micronutrients from 2 d of 24-h dietary recall data from NHANES, 2007–2018. For each micronutrient, the probability of adequacy (PA) was calculated using the best linear unbiased predictor of usual intake. Adolescent dietary diversity score (ADDS) was derived with a maximum score of 10 food groups. Generalized linear mixed models were used to examine associations between ADDS and MPA. Receiver operating characteristic analysis was used to establish a cutoff for minimum dietary diversity for adolescents (MDD-A), using an energy-adjusted logistic model with ADDS predicting MPA > 0.6. ResultsThe PA was >80% for all nutrients except vitamin C (42.1%), folate (65.7%), and calcium (23.8%). Population MPA was 79.4%, and nearly 92% of adolescents had an MPA > 0.6. ADDS was positively associated with MPA, and energy was a significant confounder. The area under the curve was >0.8 on both days with sensitivity and specificity ranging from 0.71 to 0.80. The MDD-A cutoff was calculated as 5.12 and 5.10 food groups on days 1 and 2, respectively. ConclusionsIn the United States adolescents, the best cutoff for a dichotomous indicator of dietary diversity as a proxy for micronutrient adequacy is 6 food groups in a given day. Future research could validate MDD-A and its associated cutoff for use across country contexts.