Validation and calibration of the Eating Assessment in Toddlers FFQ (EAT FFQ) for children, used in the Growing Up Milk – Lite (GUMLi) randomised controlled trial

Abstract

The Eating Assessment in Toddlers FFQ (EAT FFQ) has been shown to have good reliability and comparative validity for ranking nutrient intakes in young children. With the addition of food items (n 4), we aimed to re-assess the validity of the EAT FFQ and estimate calibration factors in a sub-sample of children (n 97) participating in the Growing Up Milk – Lite (GUMLi) randomised control trial (2015–2017). Participants completed the ninety-nine-item GUMLi EAT FFQ and record-assisted 24-h recalls (24HR) on two occasions. Energy and nutrient intakes were assessed at months 9 and 12 post-randomisation and calibration factors calculated to determine predicted estimates from the GUMLi EAT FFQ. Validity was assessed using Pearson correlation coefficients, weighted kappa (κ) and exact quartile categorisation. Calibration was calculated using linear regression models on 24HR, adjusted for sex and treatment group. Nutrient intakes were significantly correlated between the GUMLi EAT FFQ and 24HR at both time points. Energy-adjusted, de-attenuated Pearson correlations ranged from 0·3 (fibre) to 0·8 (Fe) at 9 months and from 0·3 (Ca) to 0·7 (Fe) at 12 months. Weighted κ for the quartiles ranged from 0·2 (Zn) to 0·6 (Fe) at 9 months and from 0·1 (total fat) to 0·5 (Fe) at 12 months. Exact agreement ranged from 30 to 74 %. Calibration factors predicted up to 56 % of the variation in the 24HR at 9 months and 44 % at 12 months. The GUMLi EAT FFQ remained a useful tool for ranking nutrient intakes with similar estimated validity compared with other FFQ used in children under 2 years.

Reproducibility and Relative Validity of a Short Food Frequency Questionnaire for Chinese Older Adults in Hong Kong

“…Pearson correlation coefficients ranging from 0.5 to 0.7 indicate a good correlation, while those larger than 0.7 indicate a very strong correlation between the two methods [ 42 ]. At the nutrient level, for most parameters, the correlations between the two methods ranged from being good to very strong.…”

Relative Validity of a Short Food Frequency Questionnaire for Disadvantaged Families in Hong Kong

“…Pearson correlation coefficients ranging from 0.5 to 0.7 reflected a good correlation while coefficients larger than 0.7 showed a very strong correlation between the two assessment methods [ 25 ]. From our results, the correlations between the two assessment methods were good to very strong for most parameters, except for iron, manganese, and sodium, for which only acceptable correlations were observed.…”

Which Milk during the Second Year of Life: A Personalized Choice for a Healthy Future?

“…Randomized clinical trials have been conducted in the last years regarding YCF consumption. An early development of overweight and obesity and higher concentrations of IGF-1 at 2.5 years of age have been associated with an excessive amount of protein intake [35], and some studies [19,60,62,64] have shown that children who consumed YCF had a lower protein intake than those who consumed cow's milk. Moreover, results showed that toddlers who consumed YCF with a lower protein content than cow's milk had a lower percentage of body fat mass [60,62], and that IGF-1 concentration at 2 years was positively associated with the total cow's milk intake [68].…”

“…In the GUMLi study, authors validated the Eating Assessment in Toddlers FFQ (EAT FFQ) [64], that estimate the content of protein, total fat, carbohydrate, fiber, Ca, Fe, Zn, and Vitamins B12, C, and D. Subsequently they administered the EAT FFQ [19], which describe dietary intake over previous 4 weeks at 3, 6, 9, and 12 months of the intervention.…”