Food combination questionnaire for Japanese: relative validity regarding food and nutrient intake and overall diet quality against the 4-day weighed dietary record

Abstract

The aim of this cross-sectional study was to examine the relative validity of food and nutrient intakes and overall diet quality scores derived using a newly developed dietary assessment questionnaire (food combination questionnaire, FCQ). Dietary data were collected from 222 Japanese adults (111 for each sex) aged 30-76 years using the online FCQ and then the 4-non-consective-day weighed dietary record (DR). The median of Spearman correlation coefficients for sixteen food groups was 0⋅32 among women and 0⋅38 among men. The median of Pearson correlation coefficients for forty-six nutrients was 0⋅34 among women and 0⋅31 among men. The Pearson correlation coefficient between the total scores of Healthy Eating Index-2015 (HEI-2015) derived from the DR and FCQ was 0⋅37 among women and 0⋅39 among men. The corresponding value for the Nutrient-Rich Food Index 9.3 (NRF9.3) total scores was 0⋅39 among women and 0⋅46 among men. Bland-Altman plots for these diet quality scores showed poor agreement at the individual level, although mean difference was small for the HEI-2015 (but not NRF9.3). Similar results were obtained using the paper version of FCQ, which was answered after conducting the DR, except for somewhat high Pearson correlation coefficients for the total scores of HEI-2015 (0⋅50 among both women and men) and NRF9.3 (0⋅37 among women and 0⋅53 among men). In conclusion, this analysis may lend support to the possible use of the FCQ as a rapid dietary assessment tool in large-scale epidemiologic studies in Japan, but further refinement of this tool should be pursued.

Keywords: DR, dietary record; Dietary assessment; FCQ, food combination questionnaire; HEI-2015, Healthy Eating Index-2015; Japan; NRF9.3, Nutrient-Rich Food Index 9.3; Overall diet; Rapid assessment; Validity.

Fig. 1.

Structure of the food combination questionnaire (FCQ). In the FCQ, consumption frequency of each staple food in each meal type was enquired about in terms of the number of days with consumption per week during the preceding month; for snacks, consumption frequency was similarly enquired about without specifying any staple foods. For accompanying foods for each staple food, relative consumption frequency was enquired about, namely how often the food was consumed with the staple food, with the possible answers of ‘always’, ‘sometimes’ and ‘never’. For snacks, relative consumption frequency of selected foods was similarly enquired about. The food group ‘fish’ includes shellfish; the food group ‘pulses’ includes nuts.

Fig. 2.

Bland–Altman plots assessing the agreement between estimates of the Healthy Eating Index-2015 (HEI-2015) total score and the Nutrient-Rich Food Index 9.3 (NRF9.3) total score derived from the 4-day weighed dietary record (DR) and those derived from the web version of the food combination questionnaire (FCQ) in 111 Japanese women (a: HEI-2015; c: NRF9.3) and 111 Japanese men (b: HEI-2015; d: NRF9.3). sd, standard deviation.