Development and validation of a food frequency questionnaire (FFQ) for assessing dietary macronutrients and calcium intake in Cambodian school-aged children

Table 6 Validity of FFQ: Pearson’s correlation coefficient analysis of agreement and cross-classification analysis for energy and nutrient intakes as reported in FFQ1 and 24Hs among the school-aged children in Cambodia (n = 107)

Energy and nutrients	Pearson's correlation coefficient				% classified into same ±1 quartile^d	% Mis-classified^d
Energy and nutrients	Crude	Energy-adjusted^a	Transformed^b	De-attenuated transformed^c	% classified into same ±1 quartile^d	% Mis-classified^d
Energy	0.53^**	–	0.50^**	0.71	81	7
Protein	0.50^**	0.33^**	0.44^**	0.62	79	6
Fat	0.33^**	0.18	0.29^**	0.38	68	6
Carbohydrates	0.53^**	0.31^**	0.51^**	0.70	81	5
Calcium	0.50^**	0.46^**	0.45^**	0.59	81	6

*Significant (p < 0.05), **Highly significant (p < 0.01)
^aAdjustment for energy was done by the residual method
^bTransformed: values were log-transformed
^cDe-attenuated: values calculated as the ratio of within-person to between-person variation estimated in the two 24-h recall surveys. The formula for calculating de-attenuated values was (1+ (S²w/S²b) n) ^0.5, where S²w is within-person variation and S²b is between-person variation. n = 2
^dCross-classification analyses were completed to validate agreement between FFQ1 and 24Hs in terms of the proportions of participants classified into the same ±1 quantile or misclassified. FFQ food frequency questionnaire, FFQ1 the first FFQ, 24Hs the mean of the two 24-h recall surveys

ISSN: 1475-2891