The association of highly processed food consumption with food choice values and food literacy in Japanese adults: a nationwide cross-sectional study

Background Highly processed food (HPF) consumption is increasing globally and has become a prominent public health concern. However, the relationship between HPF consumption and food choice values and food literacy is unknown. This study aimed to examine the association of HPF consumption with food choice values and food literacy. Methods This cross-sectional study used data from a nationwide questionnaire survey conducted in 2018 among 2232 Japanese adults aged 18–80 years. We assessed eight food choice values (accessibility, convenience, health/weight control, tradition, sensory appeal, organic, comfort, and safety) using a 25-item scale, and food literacy characterised by nutrition knowledge (using a validated 143-item questionnaire), cooking and food skills (using 14- and 19-item scales, respectively), and eight eating behaviours (hunger, food responsiveness, emotional overeating, enjoyment of food, satiety responsiveness, emotional undereating, food fussiness, and slowness in eating, using the 35-item Adult Eating Behavior Questionnaire). HPF consumption was estimated using a validated brief diet history questionnaire. The associations between HPF consumption and age, body mass index, energy intake, and each score on food choice values and food literacy were evaluated by multiple linear regression. Results In males, one standard deviation increase in scores for cooking skill and satiety responsiveness was associated with an increase in HPF consumption by 22.1 g/4184 kJ (95% confidence interval (CI): 6.6 to 37.5) and 15.4 g/4184 kJ (95% CI: 6.0 to 24.7), respectively. In females, one standard deviation increase in age and scores for safety and nutrition knowledge corresponded to a decrease in HPF consumption by − 16.4 g/4184 kJ (95% CI: − 23.4 to − 9.3), − 9.9 g/4184 kJ (95% CI: − 19.1 to − 0.7), and − 11.1 g/4184 kJ (95% CI: − 17.0 to − 5.3), whereas one standard deviation increase in the satiety responsiveness score corresponded to an increase in HPF consumption by 13.1 g/4184 kJ (95% CI: 6.8 to 19.4). Conclusions This cross-sectional study suggests that several aspects of food choice values and food literacy were associated with HPF consumption in Japanese adults. Further studies are needed to confirm our findings in a broader context. Supplementary Information The online version contains supplementary material available at 10.1186/s12966-023-01538-7.

All dishes identified from the dietary record excluding supplements (N=25,691) Home-prepared dishes a,b (N=14,066) Dishes prepared away from home a (N=11,625) Each food ingredient (N=46,504) was classified based on its food code For 149 dishes, 'home-made' or 'other' c food ingredients (N=197) were separated and classified based on its food code b Home-prepared dishes includes foods eaten in a raw state, such as fresh fruits and vegetables.c Other food ingredients included seasonings added during home cooking or used at the table at home (e.g., mayonnaise used to make sandwiches), and unprepared ingredients before cooking at home (e.g., fresh vegetables and milk), except for water.d Packaged food products with the names of products, brands, or manufacturers.e Dishes purchased from supermarkets, restaurant meals, and other miscellaneous dish items without information that identifies them as packaged food.All dishes (N=4,295) were classified after disaggregation into food ingredients (N=28,328) Step 1 Step 2 Step 4 Step 5 Step 3 b Differences between the two groups were analyzed using independent t-tests for age, height, weight, body mass index, and energy intake; chi-square test was used for smoking status; and Fisher's exact test for educational categories.Text S1.Calculation and validation of highly processed food consumption estimated from the brief diet history questionnaire

Dataset
The dietary record (DR) dataset used to calculate the weight ratio of highly processed food (HPF) was obtained from a nationwide survey conducted between February and March 2013 in 23 prefectures in Japan [1].Briefly, participants were asked to record all foods and beverages consumed on four nonconsecutive days (three working days and one non-working day, excluding night-shift days and days before and after a night shift).Research dietitians explained how to keep the DR on the participants and requested them to weigh foods and beverages with a digital scale (KD-812WH, Tanita, Tokyo, Japan) or measure with the spoon and cup provided.Body weight (in 0.1 kg) and height (in 0.1 cm) were measured in light clothes without shoes.Body mass index (BMI) was calculated as body weight (kg) divided by the square of height (m 2 ).In addition, information on sex, age, education, and smoking status (never, past, or current) was collected using a questionnaire.
The recording sheet included the following items: (i) dish names, (ii) whether dishes were prepared at home, away from home, or other (foods eaten in a raw state, such as fresh fruits and vegetables), (iii) food names (ingredients included in dishes), and (iv) approximate amounts or measured weights of foods consumed.Participants were also asked to record the names of products and manufacturers for store-bought products and the names of menus and restaurants when they dined out.
In addition, participants were asked to collect packaged food packages.
Recording sheets and packages were submitted to a research dietitian at each facility immediately after recording.The research dietitians reviewed the recording sheets as soon as possible and, if necessary, asked the participants to provide additional information to clarify the name or amount of food on the sheet.The research dietitian at each facility assigned each food item a food code from the Standard Tables of Food Composition in Japan (STFCJ) [2] in a uniform procedure.For packaged foods and dishes prepared away from home, each food ingredient and its consumed weight were estimated as precisely as possible based on the approximate amount of food, the website of the restaurant or manufacturer, ingredient labels, nutrition facts labels on food packages, and cooking books.
All foods items recorded in the column 'food names' were then classified into three categories by the research dietitian as follows: (1) 'home-made': foods cooked at home (e.g., rice cooked at home and bread baked at home); (2) 'store-bought': ingredients of dishes prepared away from home and foods processed by the manufacturer (e.g., ready-to-heat curry, processed meat, and chocolate); and ( 'other': unprocessed ingredients before cooking at home (e.g., fresh vegetables, meats, fish, and milk) or seasonings added during home cooking or used at the table at home (e.g., mayonnaise used to make sandwiches).In addition to this classification, all food codes and weights were reconfirmed by two other research dietitians at the central office of this study.
Classification of foods based on the degree of food processing in the DR We used the framework developed by researchers at the University of North Carolina at Chapel Hill (UNC) [3] to classify foods according to the level of processing.The UNC system, developed based on the most widely used classification system, NOVA [4], provides enhanced definitions of food categories [3].Although the UNC system classifies packaged food products with a barcode sold in the United States (US), it would be useful to classify unpackaged food products, as it provides broad and detailed examples of foods for each category, ranging from fresh vegetables (e.g., whole carrots) to refrigerated ready-made vegetable-based mixed dishes (e.g., coleslaw).A previous study showed that the UNC system had higher inter-rater reliability than the NOVA classification system [5].
According to the UNC system, the author classified all foods (except for dietary supplements) in the DR into one of the four groups:(1) unprocessed and minimally processed, ( moderately processed, and (4) highly processed.The food classification procedure is shown below and in Figure S1.
Step 1: Home-prepared dishes (including foods eaten raw) were disaggregated into component ingredients, each of which was classified based on its food code (e.g., miso, water, spinach, and egg in miso soup cooked at home, fresh fruits, and vegetables).
Step 2: For ready-made dishes consisting of a single food ingredient, each food ingredient was classified based on its food code (e.g., black tea, black coffee, and candy).
Step 3: If dishes prepared away from home contained 'home-made' or 'other' food ingredients (mostly seasonings, such as soy sauce added to Chinese dumplings at home), that ingredient was individually classified based on its food code.
Step 4: Packaged food products (i.e., dishes prepared away from home containing multiple food ingredients with the names of products, brands, or manufacturers of packaged foods) were classified without disaggregation.For instance, we classified packaged sandwiches from convenience stores as single items.
Step 5: Other dishes prepared away from home (e.g., dishes purchased from supermarkets, restaurant meals, and other miscellaneous dish items without information identifying them as packaged food) were classified after disaggregation into food ingredients, In Steps 1-3 and 5, each food item was classified based on the food code of the STFCJ [2], considering whether each food item was categorised as 'home-made' or 'store-bought' by the research dietitian.For instance, among tea with food code 16042 'oolong tea, infusion', those considered 'home-made' were classified into the unprocessed and minimally processed category.

Calculation of the weight ratio of HPF
To calculate the weight ratio of HPF for each code and assess the relative validity of HPF consumption estimated using them in the brief diet history questionnaire (BDHQ), we randomised participants into the calculation (N = 199) or validation (N = 193) groups using the PROC SURVEYSELECT procedure in the SAS statistical software (version 9.4, SAS Institute Inc., Cary, NC).Stratified randomisation based on age (10-year-age band) and sex was conducted to ensure that the participant's age and sex were not biased between the two datasets.No significant difference was observed between the two groups for participants' characteristics, including age, body mass index, education status, and smoking status, as well as energy intake and HPF consumption calculated from the DR (Table S2).
Using the DR data in the calculation group, the weight ratio of HPF was calculated for each food code in the BDHQ as follows: weight ratio of HPF for a food code = the sum of foods identified as 'highly processed' in the food code (g) / total food consumption in the food code (g) ×100 (Table S1).
For 11 food codes not appearing in the 4-d DR in the calculation group, the weight ratio of HPF was substituted by that of similar food.

Validation of HPF consumption estimated from the BDHQ
The weight ratio of HPF was then used to calculate HPF consumption from the BDHQ in the validation group.The BDHQ computes the intake of each food group (N = 58) as the sum of the weight of food items included in the food group.For instance, the 'mayonnaise and dressing' group intake is calculated as the sum of the weights of mayonnaise (food code 17043) and French dressing (food code 17040).
Therefore, we first calculated the estimated HPF consumption from each food item (e.g., mayonnaise) for each participant as the total intake of the food item (g) multiplied by its weight ratio of HPF.
Subsequently, the HPF consumption from each food group (e.g., 'mayonnaise and dressing' group) was calculated by summing the HPF consumption from each food item within that food group (e.g., mayonnaise and French dressing).Finally, the total HPF consumption was computed as the sum of the HPF consumption from all 58 food groups.
The validity of HPF consumption estimated based on the BDHQ using the weight ratio of HPF was assessed in the validation group using HPF consumption estimated from the DR as a reference.The Wilcoxon signed-rank test was used to compare the median HPF consumption (g/4184 kJ) estimated based on the BDHQ and those estimated based on the DR.The correlation between estimates from the BDHQ and DR was evaluated using Spearman correlation coefficients.In addition, a Bland-Altman plot was used to assess the agreement for estimated HPF consumption between the BDHQ and the DR [10,11].Statistical analyses were conducted separately for males and females, using the statistical software package SAS version 9.4.Two-sided P< 0•05 was considered statistically significant.

Figure S1 .
Figure S1.Flow chart of food classification in the 4-d dietary record obtained from 388 Japanese adults a Self-identified by the participants.b Home-prepared dishes includes foods eaten in a raw state, such as fresh fruits and vegetables.c Other food ingredients included seasonings added during home cooking or used at the table at home (e.g., mayonnaise used to make sandwiches), and unprepared ingredients before cooking at home (e.g., fresh vegetables and milk), except for water.d Packaged food products with the names of products, brands, or manufacturers.e Dishes purchased from supermarkets, restaurant meals, and other miscellaneous dish items without information that identifies them as packaged food.

ValidationFigure S2 .
Figure S2.Bland-Altman plots for HPF consumption for (a) males and (b) females.BDHQ, brief diet history questionnaire; DR, dietary record; HPF, highly processed food; SD, standard deviation.The solid line represents the mean difference, and the dotted line represents the lower and upper limits of agreement with an added regression line.
The median (P25-P75) HPF consumption per 4184 kJ estimated based on the BDHQ was significantly lower than those estimated based on the DR in males (170.8 (117.4-260.5)g vs 249.8 (193.2-378.5)g, respectively, P<0.0001) and females (137.2 (105.1-177.3)g vs 218.3 (145.9-298.8)g, respectively, P<0.0001).The Spearman correlation coefficient for HPF consumption per 4184 kJ was 0.59 for males (N = 97) and 0.46 for females (N = 96), indicating moderate correlation.The limits of agreement were wide for both sexes, mainly because of increased dispersion with larger serving sizes.The Bland-Altman plots are shown in FigureS2.The limits of agreement were wide for both sexes, mainly because of increased dispersion with larger HPF consumption.Linear regression analysis demonstrated that the slope of the mean bias for each food item was significantly different from 0 in females, indicating a proportional bias such that differences between the true and estimated serving size increased as the mean HPF consumption increased.

Table S1 .
Weight ratio (%) of HPFs in 147 food codes used to calculate food intake in the BDHQ.

Table S2 .
Participant characteristics in the groups in which HPF weight proportion was calculated and validated a .
a Values are mean ± standard deviation, unless otherwise indicated.c Calculated from 4-d dietary records.