Dietary intake assessment is often complicated by intrinsic bias. This study investigated whether food purchase data could constitute a valid indication of dietary intake, by evaluating the extent to which diet quality as measured by supermarket food purchases is correlated with diet quality as measured by reported dietary intake. We used data from the Supreme Nudge cluster-randomised controlled supermarket trial (n=227). Data were collected at baseline from supermarket purchases (loyalty cards) and a dietary questionnaire (short 40-item food frequency questionnaire (FFQ)) to compute two scores reflecting diet quality from purchasing data (purchased diet quality) and FFQs (consumed diet quality). Both scores constituted of 13 food groups and could theoretically range between 0 (low diet quality) to 130 (high diet quality). The relationship between purchased diet quality and consumed diet quality was assessed using correlation coefficients, and the Bland-Altman limits-of-agreement method. Multiple linear regression was fitted between purchased diet quality and consumed diet quality, adjusted for age, sex, waist circumference, educational level, and household size. Consumed and purchased diet qualities were modestly positively correlated (Pearson’s ρ = 0·31, 95% CI: 0·18 - 0·42). A positive association from linear regression was found after confouding adjustments (βbaseline = 0·22, 95%CI: 0·10 - 0·34). Purchased diet quality was systematically lower than the consumed diet quality. This study found that diet quality as measured by supermarket purchases provided a reasonable indication of diet quality as reported by short-FFQs, albeit modest.

The aim of this study was to explore the associations between diet quality, socio-demographic measures, smoking, and weight status in a large, cross-sectional cohort of adults living in Yorkshire and Humber, UK. Data from 43, 023 participants aged over 16 years in the Yorkshire Health Survey, 2nd wave (2013–2015) were collected on diet quality, socio-demographic measures, smoking, and weight status. Diet quality was assessed using a brief, validated tool. Associations between these variables were assessed using multiple regression methods. Split-sample cross-validation was utilised to establish model portability. Observed patterns in the sample showed that the greatest substantive differences in diet quality were between females and males (3.94 points; P < 0.001) and non-smokers vs smokers (4.24 points; P < 0.001), with higher diet quality scores observed in females and non-smokers. Deprivation, employment status, age, and weight status categories were also associated with diet quality. Greater diet quality scores were observed in those with lower levels of deprivation, those engaged in sedentary occupations, older people, and those in a healthy weight category. Cross-validation procedures revealed that the model exhibited good transferability properties. Inequalities in patterns of diet quality in the cohort were consistent with those indicated by the findings of other observational studies. The findings indicate population subgroups that are at higher risk of dietary-related ill health due to poor quality diet and provide evidence for the design of targeted national policy and interventions to prevent dietary-related ill health in these groups. The findings support further research exploring inequalities in diet quality in the population.

Reliable dietary assessment is key to our understanding of diet-health interactions(1). Most current dietary assessment methods are self-reported, making them prone to a range of biases(2). Objective markers, such as certain metabolite concentrations in human tissue, may prove a more reliable method of dietary assessment in the future while reducing participant burden(2,3). An example of these metabolite markers are alkylresorcinols, a family of compounds found in the bran layer of common grains. High concentrations of alkylresorcinols reflect high wholegrain intakes, very low levels indicate only refined grain intake, while no alkylresorcinols would indicate a gluten-free diet. We conducted a randomised crossover feeding study of three interventions (one standardisation day followed by a feeding day where known amounts were consumed under observation). Each feeding day differed by which food groups were provided (e.g., chicken, legumes, and fruit) or entirely absent (e.g., grains, fish, and dairy). Participants provided 24-hour urine samples on all six days (standardisation days and feeding days), as well as a 24-recalls and fasted blood samples the morning after each day. Known metabolite markers of dietary intake (approximately 70) were identified in blood and urine with LC-MSqToF, and semi-quantified with a known standard. Twenty-four hour urine sodium content was also measured as the current best known objective marker of dietary intake(4). Twenty-four participants (74% female, age (SD) 24.8 (6.1), BMI 24.1 (4.0)) commenced the study, with 21 (88%) completing all three interventions. Mean energy intake on feeding days (11720 kJ (2943.01)) was higher than self-reported energy intake on standardisation days (9243.57 kJ (3582.92)). Meals were well tolerated, with mean (range) intakes of 9.8 (6.3 – 16.1) serves whole grains, 2.4 (1.6 – 4.8) serves fish, 3.1 (1.9 – 5.5) serves dairy, 5.6 (4.5 – 9.1) serves chicken, 8.2 (7.0 – 14.1) serves legumes, 3.1 (1.3 – 4.6) serves fruit, 3.9 (2.6 – 6.4) serves red meat, 1.7 (1.35 – 2.6) serves nuts and seeds, or 13.4 (9.4 – 19.5) serves vegetables on their respective feeding days. The three feeding days provided clearly identifiable clusters when assessing the overall metabolic profile, both in terms of what was measured on the feeding days, and the difference in metabolite concentrations between standardisation day and feeding day. The relative correlations between self-reported intakes and individual metabolite concentrations reflecting specific foods or food groups with the known dietary intakes from feeding days will be presented first at the conference.

In nutritional epidemiological studies, it is imperative to collect high-quality data to ensure accurate dietary assessment. However, dietary data collection using traditional paper forms has several limitations that may compromise data quality. The aim of this study was to propose novel methods to design and develop software applications (Apps) for dietary data collection to assess the nutritional status of pregnant women and infants. This study is part of the M-SAKHI (Mobile-Solutions for Aiding Knowledge for Health Improvement) cluster randomised controlled trial (cRCT) implemented in central India. Three tablet-based software Apps were developed in this study: the ACEC (Automated Coding and Energy Calculation) App to establish a generic cooked food recipe database, the FFQ (Food Frequency Questionnaire), and the IDR (24 h Infant Dietary Recall) Apps to collect dietary data from pregnant women and their infants from rural area of Bhandara and Nagpur districts. Regional food lists, recipes, and portion resource kits were developed to support the data collection using the Apps. In conclusion, the Apps were user-friendly, required minimal prior training, had built-in validation checks for erroneous data entry and provided automated calculations. The Apps were successfully deployed in low-resource rural settings to accurately collect high-quality regional cooked food data and individual-level dietary data of pregnant women and their infants.

With the recent growing interest in improving fruit and vegetable intake for better health and limited research resources in many settings, simple-to-administer and low-priced indicators are essential tools for monitoring fruit and vegetable intake at the population level. A potential candidate indicator is the fruit and vegetable component of the Global Dietary Recommendation score (FV-GDR) based on data collected using the Diet Quality Questionnaire (DQQ). We investigated the relative validity of FV-GDR collected with the DQQ to measure fruit and vegetable intake by comparison with a 24-h recall (24hR) as a reference collected from 620 Vietnamese and 630 Nigerian adults in 2021. We found proportional differences in the prevalence of intake of vitamin A-rich vegetables, other vegetables and other fruits in Vietnam and all vegetable food groups in Nigeria. In both countries, we found a small difference in the total FV-GDR from DQQ compared with the 24hR, and the percentage of agreement between the two methods was quite high for the majority of the food groups. The FV-GDR calculated from the DQQ correlated with the actual intake, although less strongly than the FV-GDR from 24hR. The DQQ is a promising low-burden, low-cost and simple tool to calculate FV-GDR and to monitor fruit and vegetable consumption at the population level. This provides the possibility of evaluating an important aspect of diet quality in low-resource settings.

Technological innovations can standardise and minimise reporting errors in dietary assessment. This scoping review aimed to summarise the characteristics of technological tools used to assess children's food intake. The review followed the Joanna Briggs Institute's manual. The main inclusion criterion was studied that assessed the dietary intake of children 0–9 years of age using technology. We also considered articles on validation and calibration of technologies. We retrieved 15 119 studies and 279 articles were read in full, after which we selected 93 works that met the eligibility criteria. Forty-six technologies were identified, 37 % of which had been developed in Europe and 32⋅6 % in North America; 65⋅2 % were self-administered; 27 % were used exclusively at home; 37 % involved web-based software and more than 80 % were in children over 6 years of age. 24HR was the most widely used traditional method in the technologies (56⋅5 %), and 47⋅8 % of the tools were validated. The review summarised helpful information for studies on using existing tools or that intend to develop or validate tools with various innovations. It focused on places with a shortage of such technologies.

The European Food Safety Authority has suggested that EU countries implement the 2 × 24 h diet recall (2 × 24 h DR) method and physical activity (PA) measurements for national dietary surveys. Since 2000, Denmark has used 7 d food diaries (7 d FD) with PA questionnaires and measurements. The accuracy of the reported energy intakes (EI) from the two diet methods, pedometer-determined step counts and self-reported time spent in moderate-to-vigorous PA (MVPA) were compared with total energy expenditure measured by the doubly labelled water (TEEDLW) technique and with PA energy expenditure (PAEE), respectively. The study involved fifty-two male and sixty-eight female volunteers aged 18–60 years who were randomly assigned to start with either the 24 h DR or the web-based 7 d FD, and wore a pedometer for the first 7 d and filled in a step diary. The mean TEEDLW (11·5 MJ/d) was greater than the mean reported EI for the 7 d FD (9·5 MJ/d (P < 0·01)) but the same as the 2 × 24 h DR (11·5 MJ/d). The proportion of under-reporters was 34 % (7 d FD) and 4 % (2 × 24 h DR). Most participants preferred the 7 d DR as it was more flexible, despite altering their eating habits. Pearson’s correlation between steps corrected for cycling and PAEE was r = 0·44, P < 0·01. Spearman’s correlation for self-reported hours spent in MVPA and PAEE was r = 0·58, P < 0·01. The 2 × 24 h DR performs better than the existing 7 d FD method. Pedometer-determined steps and self-reported MVPA are good predictors of PAEE in adult Danes.

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.

All dietary assessment methods inevitably introduce measurement errors, which should ideally be considered during data analysis and interpretation. Methodological studies should be conducted to address how well a given assessment method captures dietary intake and to highlight the extent and direction of the measurement error. Within a subgroup of the Hordaland Health Study (HUSK3), we examined the relative validity of a web-based food frequency questionnaire (WebFFQ) by comparing its estimates of mean daily intake of nutrients and foods with estimated mean daily intakes from repeated administrations of 24-hour dietary recall interviews (24-HDRs). Men and women born between 1950 and 1951 were recruited from HUSK3. The participants (n = 67) completed a WebFFQ and three non-consecutive 24-HDRs over the course of a year. Relative validity was assessed using Spearman's rank correlation, crosstab analysis and Bland–Altman plots. Linear regression models were used to compute the calibration coefficients. The estimated correlation coefficients were acceptable or strong for all nutrients and foods except iodine (rs = 0⋅19). The highest correlation coefficient was found for juice (rs = 0⋅71), whereas the lowest correlation coefficient was found for iodine (rs = 0⋅19). Cross-classification by quartiles categorised more than 72 % of the participants into the same or adjacent quartiles using the two methods. Few data points fell outside the limits of agreement in the Bland–Altman plots. Calibration coefficients ranged from 0⋅10 (wholegrain) to 0⋅81 (alcohol). Our findings suggest that the WebFFQ has reasonable ranking abilities for all the included nutrients and foods, except for iodine.

The aim of this study was to examine the relative validity of the online Meal-based Diet History Questionnaire (MDHQ) for assessing the overall diet quality and quality of each meal type (breakfast, lunch, dinner and snacks). In total, 222 Japanese adults (111 for each sex) aged 30–76 years completed the online MDHQ and then the 4-non-consecutive-day weighed dietary record (DR). The diet quality was assessed using the Healthy Eating Index-2015 (HEI-2015) and Nutrient-Rich Food Index 9.3 (NRF9.3). For the HEI-2015, compared with the DR, the MDHQ provided high median values for breakfast (in women only) and dinner and low median values for snacks. There were no significant differences observed for overall diet and lunch. For the NRF9.3, the MDHQ provided higher median values for breakfast and dinner and a lower median value for overall diet than the DR in women, with no significant differences for lunch and snacks. In men, no significant difference was observed, except for overall diet (the MDHQ providing a lower median value). For the HEI-2015, median Spearman correlation coefficient was 0·43, with a range from 0·12 (snacks in women) to 0·68 (breakfast in men). For the NRF9.3, median Spearman correlation coefficient was 0·47, with a range from 0·26 (snacks in men) to 0·65 (breakfast in men). Bland–Altman plots showed wide limits of agreement and, in some cases, proportional bias. In conclusion, the online MDHQ showed an acceptable ability for ranking individuals according to the quality of overall diet, breakfast, lunch and dinner, but not snacks.

Traditional methods of dietary assessment are prone to measurement error, with energy intake often under-reported. The 24-h recall is widely used in dietary assessment, however, its reliance on self-report without verification of consumption can result in inaccuracies in true nutrient intake. Wearable cameras may provide a complementary approach to improve self-report accuracy by providing an objective and passive measure of food consumption. The purpose of the present study was to determine whether a wearable camera improves the accuracy of a 24-h recall compared with a 24-h recall alone in twenty adults aged 18–65 years. The study also explored limitations associated with wearable cameras. Participants wore the camera for 1 d and a 24-h recall was then conducted the following day, before and after viewing the camera images. Dietary data were analysed using Nutritics dietary analysis software, while eating habits were assessed by a self-report questionnaire. Energy and nutrient intakes were compared between the recall alone and the camera-assisted recall. Results showed a significant increase in mean energy intake with the camera-assisted recall compared with the recall alone (9677⋅8 ± 2708⋅0 kJ/d v. 9304⋅6 ± 2588⋅5 kJ/d, respectively, P = 0⋅003). Intakes of carbohydrates, total sugars and saturated fats were also significantly higher with the camera-assisted recall. In terms of challenges, there were occasionally technological issues such as proper positioning of the camera by the participants. In conclusion, reporting of energy and nutrient intake may be enhanced when a traditional method of dietary assessment, the 24-h recall, is assisted by a wearable camera.

The present study examined the association of contextual factors (social and food preparation location) with the energy density of meals and snacks consumed in a sample of young Australian adults (18–30 years old) identified using wearable camera technology. Over three consecutive days, a subsample of young adults wore a wearable camera that captured images in 30 s intervals. Eating episodes from 133 participants were annotated for preparation location and social context (covering social interaction and screen use). Over the same period, participants completed daily 24 h recalls. The nutritional composition of meals and snacks was calculated by matching the items identified in the camera to the 24 h recall using time and date stamps. Self-reported data (weight and height) was used to calculate body mass index and (residential postcode) to assign socio-economic status. The association of context and demographic factors with energy density was determined using a mixed linear regression model employing the bootstrap method with bias-corrected and accelerated. In total, 1817 eating episodes were included in the analysis (n 8 preparation unclear and n 15 food components could not be identified excluded). Food prepared within the home was 1⋅1 kJ/g less energy-dense than other preparation locations. Lunches (CI −1⋅7 to −0⋅3) and dinners (CI −1⋅6 to −0⋅5) were both 1⋅0 kJ/g lower in energy density than breakfasts. Snacks were 3⋅5 kJ/g (CI 2⋅8–4⋅1) more energy-dense than breakfasts. Food prepared outside the home and food consumption during snacking appear to be adversely contributing to energy-dense food intake.

Time to evaluate diet quality and give dietary advice is limited in clinical inflammatory bowel disease (IBD) practice. The Eetscore is a web-based tool that assesses diet quality according to the Dutch dietary guidelines and provides personalised dietary advice. We aimed to assess diet quality of IBD patients using the Eetscore and to study changes in diet quality, health-related quality of life (HRQoL) and clinical disease activity over time. A prospective cohort study was performed in 195 adult IBD patients. Participants were invited to fill out questionnaires (Eetscore-FFQ, short Inflammatory Bowel Disease Questionnaire and Patient Harvey Bradshaw Index/Patient Simple Clinical Colitis Activity Index) at baseline and after 1 and 4 months. The Eetscore calculates diet quality based on sixteen food components (ten points per component, total score 0–160; the higher the better) and provides dietary advice per component based on the assessment. At baseline, mean diet quality was 98 (sd 19). Diet quality was positively associated with age, female sex and level of education. Component scores were highest for red meat, wholegrain products and sweetened beverages, and lowest for legumes, nuts and processed meat. Over time, diet quality increased to 107 (sd 21) at 4 months (P < 0·001). Each ten-point improvement in diet quality was associated with an increase in HRQoL (β = 0·4 (95 % CI (0·02, 0·7), P = 0·04). Clinical disease activity did not change. In conclusion, diet quality of IBD patients significantly improved following personalised dietary advice of the Eetscore. Improvement of diet quality was associated with a slight improvement in HRQoL. The Eetscore is a practical and useful tool to monitor and support a healthy diet in IBD patients.

Diet quality indices (DQIs) are tools used to evaluate the overall diet quality against dietary guidelines or known healthy dietary patterns. This review aimed to evaluate DQIs and their validation processes to facilitate decision making in the selection of appropriate DQI for use in Australian contexts. A search of CINAHL, PubMed and Scopus electronic databases was conducted for studies published between January 2010 and May 2020, which validated a DQI, measuring > 1 dimension of diet quality (adequacy, balance, moderation, variety) and was applicable to the Australian context. Data on constructs, scoring, weighting and validation methods (construct validity, criterion validity, reliability and reproducibility) were extracted and summarised. The quality of the validation process was evaluated using COnsensus-based Standards for the selection of health Measurement INstruments Risk of Bias and Joanna Briggs Appraisal checklists. The review identified twenty-seven indices measuring adherence to: national dietary guidelines (n 13), Mediterranean Diet (n 8) and specific population recommendations and chronic disease risk (n 6). Extensiveness of the validation process varied widely across and within categories. Construct validity was the most strongly assessed measurement property, while evaluation of measurement error was frequently inadequate. DQIs should capture multiple dimensions of diet quality, possess a reliable scoring system and demonstrate adequate evidence in their validation framework to support use in the intended context. Researchers need to understand the limitations of newly developed DQIs and interpret results in view of the validation evidence. Future research on DQIs is indicated to improve evaluation of measurement error, reproducibility and reliability.

The validity of grocery purchase data as an indicator of food consumption is uncertain. This paper investigated (1) the associations between food consumption and grocery purchases using automatically accumulated purchase data and (2) whether the strength of the associations differed in certain sub-populations. The participants filled in a FFQ, and a major Finnish retailer issued us with their loyalty-card holders’ grocery purchase data covering the 1- and 12-month periods preceding the FFQ. We used gamma statistics to study the association between thirds/quarters of FFQ and grocery purchase data (frequency/amount) separately for eighteen food groups among the 11 983 participants. Stratified analyses were conducted for subgroups based on sex, family structure, educational level, household income and self-estimated share of purchases from the retailer. We also examined the proportion of participants classified into the same, adjacent, subsequent and opposite categories using the FFQ and purchase data. The gammas ranged from 0·12 (cooked vegetables) to 0·75 (margarines). Single households had stronger gammas than two-adult families, and participants with > 60 % of purchases from the retailer had stronger gammas. For most food groups, the proportion of participants classified into the same or adjacent category was > 70 %. Most discrepancies were observed for fresh/cooked vegetables, berries and vegetable oils. Even though the two methods did not categorise all food groups similarly, we conclude that grocery purchase data are able to describe food consumption in an adult population, and future studies should consider purchase data as a resource-saving and moderately valid measure in large samples.