# Reliability of a Modified 24 h Dietary Recall and Veggie Meter to Assess Fruit and Vegetable Intake in New Zealand Children

## TL;DR

The study tested a modified 24-hour dietary recall and a non-invasive device to assess fruit and vegetable intake in New Zealand children, finding the device more reliable than the recall method.

## Contribution

The study introduces a modified 24-hour dietary recall and evaluates the reliability of the Veggie Meter® for assessing carotenoid intake in children.

## Key findings

- The modified 24-hour recall was valid for estimating fruit and vegetable servings but unreliable for carotenoid intake.
- The Veggie Meter® showed high reliability with low measurement error for skin carotenoid scores.
- Children in the study did not meet recommended fruit and vegetable intake levels.

## Abstract

Adequate intake of fruits and vegetables (F + V) supports healthy growth and development in children, yet many New Zealand children do not meet national dietary recommendations, and methods to evaluate intake require good reliability. Objectives: To establish the validity and reliability of a modified 24 h multiple pass recall (MPR) for evaluating F + V and carotenoid intakes in children aged 9–13 years. The reliability of the Veggie Meter® (VM®), a non-invasive reflection spectrometer to estimate skin carotenoid scores and derive blood carotenoid concentrations, was also examined. Methods: Thirty-two children (20 boys, 12 girls) completed three 24 h MPRs and parent-assisted weighed food diaries (WFDs) on randomised weekdays and weekends. Skin carotenoid scores were assessed using the VM®. The validity of the MPR was evaluated against WFDs using log-transformed Pearson correlations and mean x-axis bias. The reliability was assessed by the coefficient of variation (CV) and Pearson correlations. Results: Participants did not meet recommended F + V intakes (5–5.5 servings/day): MPR (mean fruit 1.3 servings/day; vegetables 2.0), WFD (fruit 1.3; vegetables 1.9). The MPR was a valid tool to estimate fruit and vegetable daily servings (combined-day Pearson coefficients > 0.71) with only trivial–small standardized mean bias-offset vs. WFD; however, the reliability was poor for the MPR-estimated carotenoid intake (CV 126%) and F + V intake. In contrast, the VM® was reliable (Pearson correlation 0.97–0.99) with low measurement error (CV 4.0–5.2%). Conclusions: The modified 24 h MPR was valid but unreliable for estimating F + V and carotenoid intake. The VM® demonstrated high reliability as a biomarker of skin carotenoid status in children.