Rx Food App: A Proof-of-Concept Study of an Image-Based Dietary Assessment Mobile Application

Abstract Objectives To determine if Rx Food, an image-based dietary assessment app powered by artificial intelligence, can derive comparable nutritional composition estimates compared to calculated methods. Sub-group analyses assessed differences between composite (i.e., multiple ingredients) and single item foods. Methods Food items were selected for testing based on their frequency of consumption among patients attending a weight management clinic. Food photos were uploaded, and serving sizes entered, into the app which generated estimated nutrient data. The nutritional composition of foods was also analyzed with ESHA Food Processor software. Nutrient estimates between the methods were compared using paired t-tests, Pearson correlation coefficients, and Bland-Altman plots for energy, carbohydrates, protein, total fat, fibre, total sugar and sodium. Results Thirty-nine food items were analyzed [n = 10 (27%) composite items and n = 29 (73%) single item foods]. There were no statistically significant differences in the mean differences in estimates from Rx Food and calculated values for all nutrients: −4.3 ± 29.2 kcal for energy, −0.4 ± 2.6 g for carbohydrates, −0.1 ± 1.9 g for protein, −0.3 ± 1.7 g for fat, −0.2 ± 2.3 g for fibre, 0.01 ± 1.4 g for sugar, and −33 ± 135 mg for sodium. Among all food items, a strong, significant correlation (r > 0.80; P < 0.05) was observed for all nutrients except fibre (r = 0.552; P < 0.001). In the Bland-Altman plots for all foods, significant bias was found for fibre (r = 0.562; P < 0.001), fat (r = 0.562; P = 0.025), and sodium (r = 0.359; P = 0.025), suggesting that Rx Food may underestimate nutrient composition at higher levels. Subgroup analyses of composite items showed significant strong correlations for energy, carbohydrates, protein, and sugar (r > 0.80; P < 0.05), significant moderate correlations (r = 0.60–0.79; P < 0.05) for fat and fibre, but not for sodium (r = 0.591; P = 0.072). Single item analysis showed significant correlations for all nutrients (r > 0.80; P < 0.05). Conclusions This preliminary data shows that Rx Food has the potential to be an accurate, image-based, low burden tool to calculate nutrient composition of foods. These findings justify further research to determine the validity of Rx Food in its ability to generate accurate nutrient intake data as a dietary assessment tool. Funding Sources N/A.

Download Full-text

Concurrent validity of skin carotenoid status as a concentration biomarker of vegetable and fruit intake compared to multiple 24-h recalls and plasma carotenoid concentrations across one year: a cohort study

Nutrition Journal ◽

10.1186/s12937-019-0500-0 ◽

2019 ◽

Vol 18 (1) ◽

Cited By ~ 3

Author(s):

Lisa Jahns ◽

LuAnn K. Johnson ◽

Zach Conrad ◽

Michael Bukowski ◽

Susan K. Raatz ◽

...

Keyword(s):

Concurrent Validity ◽

Linear Models ◽

Assessment Tool ◽

Pearson Correlation ◽

Dietary Assessment ◽

Resonance Raman Spectroscopy ◽

Correlation Coefficients ◽

Strongly Correlated ◽

Clinical Trial Registry ◽

One Year

Abstract Background Biological markers of vegetable and fruit (VF) intake are needed both for nutrition surveillance and for the evaluation of nutrition interventions. Optically assessed skin carotenoid status (SCS) has been proposed as a marker of intake but there are few published validity studies to date. Therefore, the objective of the study was to examine the concurrent validity of multiple methods of assessing VF intake cross-sectionally and seasonally over one year and to discuss the relative merits and limitations of each method. Methods Fifty-two 40–60 y old women completed a 1-year longitudinal study that included 1) SCS assessment using resonance Raman spectroscopy (RRS) and using pressure-mediated reflection spectroscopy (RS) at 12 timepoints, 2) thirty-six 24-h recalls using the Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA24; total 1866 recalls), and 3) plasma carotenoid concentrations measured every 3 months. Pearson correlation coefficients and mixed linear models were used to estimate pairwise correlations between RRS, RS, ASA24, and plasma carotenoids. Results RS and RRS were strongly correlated at baseline and over the year (r = 0.86 and 0.76; respectively, P < 0.001). RS was strongly correlated with plasma carotenoids at baseline (r = 0.70) and moderately across the year (r = 0.65), as was RRS (r = 0.77 and 0.69, respectively, all P < 0.001). At baseline, self-reported VF was weakly correlated with RRS (r = 0.33; P = 0.016), but not with RS or plasma carotenoids. Across the year, self-reported VF intake was weakly correlated with both RS (r = 0.37; P = 0.008), RRS (r = 0.37; P = 0.007), and with plasma carotenoids (r = 0.36; P < 0.008). Conclusions SCS as measured by RS and RRS is moderately to strongly correlated with plasma carotenoid concentrations both cross-sectionally and longitudinally, indicating that it can be a powerful tool to assess carotenoid-rich VF intake in populations. Clinical trial registry This trial was registered at ClinicalTrials.gov as NCT01674296.

Download Full-text

Effect of Manual Data Cleaning on Nutrient Intakes using the Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA24)

Current Developments in Nutrition ◽

10.1093/cdn/nzab005 ◽

2021 ◽

Author(s):

Yasmine Y Bouzid ◽

Joanne E Arsenault ◽

Ellen L Bonnel ◽

Eduardo Cervantes ◽

Annie Kan ◽

...

Keyword(s):

Energy Intake ◽

Assessment Tool ◽

Data Cleaning ◽

Dietary Assessment ◽

Assessment Tools ◽

Energy Requirements ◽

Nutrient Intakes ◽

Significant Difference ◽

Before And After ◽

Total Fat

Abstract Background Automated dietary assessment tools such as ASA24® are useful for collecting 24-hour recall data in large-scale studies. Modifications made during manual data cleaning may affect nutrient intakes. Objectives We evaluated the effects of modifications made during manual data cleaning on nutrients intakes of interest: energy, carbohydrate, total fat, protein, and fiber. Methods Differences in mean intake before and after data cleaning modifications for all recalls and average intakes per subject were analyzed by paired t-tests. Chi-squared test was used to determine whether unsupervised recalls had more open-ended text responses that required modification than supervised recalls. We characterized food types of text response modifications. Correlations between predictive energy requirements, measured total energy expenditure (TEE), and mean energy intake from raw and modified data were examined. Results After excluding 11 recalls with invalidating technical errors, 1499 valid recalls completed by 393 subjects were included in this analysis. We found significant differences before and after modifications for energy, carbohydrate, total fat, and protein intakes for all recalls (p < 0.05). Limiting to modified recalls, there were significant differences for all nutrients of interest, including fiber (p < 0.02). There was not a significantly greater proportion of text responses requiring modification for home compared to supervised recalls (p = 0.271). Predicted energy requirements correlated highly with TEE. There was no significant difference in correlation of mean energy intake with TEE for modified compared to raw data. Mean intake for individual subjects was significantly different for energy, protein, and fat intakes following cleaning modifications (p < 0.001). Conclusions Manual modifications can change mean nutrient intakes for an entire cohort and individuals. However, modifications did not significantly affect correlation of energy intake with predictive requirements and measured expenditure. Investigators can consider their research question and nutrients of interest when deciding to make cleaning modifications.

Download Full-text

Assessment of Dietary Intake Among South Asian Adults in the United States

Current Developments in Nutrition ◽

10.1093/cdn/nzab035_032 ◽

2021 ◽

Vol 5 (Supplement_2) ◽

pp. 124-124

Author(s):

Nurgul Fitzgerald ◽

Shailja Mathur

Keyword(s):

United States ◽

Dietary Intake ◽

Energy Intake ◽

South Asian ◽

Assessment Tool ◽

Dietary Assessment ◽

The United States ◽

Management Program ◽

Convenience Sample ◽

Total Fat

Abstract Objectives To examine the dietary intake patterns of South Asian adults by using three different assessment methods. Methods The participants were a convenience sample of 62 adults from South Asian descent, who lived in the United States and participated in a community-based diabetes self-management program. Dietary intake data were collected through Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA24), self-administered Diet History Questionnaire III (self-DHQ), and researcher-administered DHQ III (res-DHQ) (National Cancer Institute). Thirty-seven participants completed ASA24 and self-DHQ back-to-back during in-person sessions, and 25 participants completed res-DHQ through video conferencing sessions with the researcher. Group level data were examined using IBM SPSS Statistics software. Results On average, participants’ daily energy intake levels were estimated to be 805.8 ± 551.3, 1686.4 ± 985.9 and 1469.7 ± 887.5 kcal/d by self-DHQ, ASA24, and res-DHQ, respectively. Self-DHQ produced the lowest of the estimates (mean ± SD) for daily protein (28.9 ± 18.8 vs 63.1 ± 35.2, and 53.1 ± 27.9 g/d), carbohydrate (106.4 ± 68.0 vs 224.9 ± 128.4 and 199.9 ± 119.7 g/d), and total fat (31.7 ± 29.2 vs. 63.5 ± 46.5 and 56.2 ± 40.9 g/d) intakes in comparison to ASA24 and res-DHQ, respectively. Conclusions In this study, self-administered DHQ produced substantially lower estimates of daily macronutrient and energy intake levels. The ASA24 or researcher-administered DHQ were relatively more reliable methods of dietary assessment in this sample of South Asian adults. Funding Sources NJ Department of Health, Office of Minority and Multicultural Health.

Download Full-text

Relative validity of dietary patterns derived from a self-administered diet history questionnaire using factor analysis among Japanese adults

Public Health Nutrition ◽

10.1017/s1368980009993211 ◽

2010 ◽

Vol 13 (7) ◽

pp. 1080-1089 ◽

Cited By ~ 39

Author(s):

Hitomi Okubo ◽

Kentaro Murakami ◽

Satoshi Sasaki ◽

Mi Kyung Kim ◽

Naoko Hirota ◽

...

Keyword(s):

Factor Analysis ◽

Dietary Patterns ◽

Pearson Correlation ◽

Dietary Assessment ◽

Correlation Coefficients ◽

Food Groups ◽

Relative Validity ◽

Japanese Adults ◽

Diet History Questionnaire ◽

Diet History

AbstractObjectiveAlthough dietary pattern approaches derived from dietary assessment questionnaires are widely used, only a few studies in Western countries have reported the validity of this approach. We examined the relative validity of dietary patterns derived from a self-administered diet history questionnaire (DHQ) among Japanese adults.DesignThe DHQ, assessing diet during the preceding month, and 4 d dietary records (DR) were collected in each season over one year. To derive dietary patterns, 145 food items in the DHQ and 1259 in the DR were classified into thirty-three predefined food groups, and entered into a factor analysis.SettingThree areas in Japan; Osaka (urban), Nagano (rural inland) and Tottori (rural coastal).SubjectsA total of ninety-two Japanese women and ninety-two Japanese men aged 31–76 years.ResultsWe identified three dietary patterns (‘healthy’, ‘Western’ and ‘Japanese traditional’) in women and two (‘healthy’ and ‘Western’) in men, which showed a relatively similar direction and magnitude of factor loadings of food groups across the first and mean of four DHQ (DHQ1 and mDHQ, respectively) and 16 d DR. The Pearson correlation coefficients between DHQ1 and 16 d DR for the healthy, Western and Japanese traditional patterns in women were 0·57, 0·36 and 0·44, and for the healthy and Western patterns in men were 0·62 and 0·56, respectively. When mDHQ was examined, the correlation coefficients improved for women (0·45–0·69).ConclusionsDietary patterns derived from the DHQ could be used for epidemiological studies as surrogates of those derived from DR.

Download Full-text

Reproducibility and Validity of a Food Frequency Questionnaire for Dietary Assessment in Adolescents in a Self-Reported Way

Nutrients ◽

10.3390/nu12072081 ◽

2020 ◽

Vol 12 (7) ◽

pp. 2081

Author(s):

Leyre Notario-Barandiaran ◽

Carmen Freire ◽

Manuela García-de-la-Hera ◽

Laura Mª Compañ-Gabucio ◽

Laura Torres-Collado ◽

...

Keyword(s):

Food Frequency Questionnaire ◽

Vegetable Intake ◽

Pearson Correlation ◽

Dietary Assessment ◽

Correlation Coefficients ◽

Food Group ◽

Birth Cohort Study ◽

Food Groups ◽

Food Frequency ◽

Male Adolescents

Tools to assess diet in a reliable and efficient way are needed, particularly in children and adolescents. In this study, we assess the reproducibility and validity of a semiquantitative food frequency questionnaire (FFQ) among adolescents in Spain. We analyzed data of 51 male adolescents aged 15–17 years from a prospective birth cohort study. Participants answered the FFQ twice in a self-administered way over a 12-month period. Reproducibility was assessed by comparing nutrient and food intakes from the FFQs, and validity by comparing nutrient intakes from the average of two FFQs and the average of two 24-Hour Dietary Recalls obtained in the period. Pearson correlation coefficients were calculated. The average of reproducibility correlation coefficients for food group intakes was 0.33, with the highest correlation for vegetable intake (r = 0.81); and the average for nutrient intake was 0.32, with the highest coefficients for α- and β-carotene (r = 0.65). Validity correlation coefficients ranged from 0.07 for carbohydrates to 0.53 for dietary fiber. The average of the validity correlation coefficients was r = 0.32. This study suggests that our FFQ may be a useful tool for assessing dietary intake of most nutrient and food groups among Spanish male adolescents in a self-administered way, despite reproducibility and, particularly validity, being low for some nutrients and food groups.

Download Full-text

Acculturation and dietary habits of Korean Americans

British Journal Of Nutrition ◽

10.1079/bjn20031068 ◽

2004 ◽

Vol 91 (3) ◽

pp. 469-478 ◽

Cited By ~ 31

Author(s):

Jeongseon Kim ◽

Mabel M. Chan

Keyword(s):

New York ◽

Dietary Habits ◽

Assessment Tool ◽

Dietary Assessment ◽

Korean Americans ◽

Food Groups ◽

Validity And Reliability ◽

Nutrient Intakes ◽

Brown Sugar ◽

Total Fat

This is the first study to investigate how food and nutrient intakes vary with the levels of acculturation of Korean Americans using a dietary assessment tool tested for validity and reliability. The respondents were Korean Americans (n 486) resident in the Greater New York metropolitan area, USA. They were divided into two groups according to the total score of acculturation: low- and high-acculturated groups. Using a food-frequency questionnaire and a modified Suinn-Lew Asian self-identity acculturation scale, differences in the frequencies of food items, food groups and nutrient intakes consumed were analysed by level of acculturation. The low-acculturated group tended to consume significantly more rice, mixed rice, soyabean paste chigae (pot stew), saengsun (fish) chigae, kimchi chigae, other fish broiled or baked, eggs, kimchi, spinach, persimmons, and white or brown sugar in coffee or tea. The high-acculturated group demonstrated a statistically greater tendency to consume bread, cereal, spaghetti, ham, green salad, sweetcorn, chocolate, candies and diet soft drinks. The more acculturated an individual, the more significant was the tendency to consume sweets. The more acculturated group had significantly (P<0·05) higher intakes of % energy from total fat, thiamin, vitamin E and folate, while the low-acculturated group consumed greater amounts of Na, niacin and dietary fibre. The information from the present study can be used to describe dietary habits according to various aspects of acculturation, and allows a better understanding of the dynamics of acculturation and dietary habits.

Download Full-text

Relative validity of an FFQ to estimate daily food and nutrient intakes for Chilean adults

Public Health Nutrition ◽

10.1017/s1368980012004107 ◽

2012 ◽

Vol 16 (10) ◽

pp. 1782-1788 ◽

Cited By ~ 18

Author(s):

Mahshid Dehghan ◽

Solange Martinez ◽

Xiaohe Zhang ◽

Pamela Seron ◽

Fernando Lanas ◽

...

Keyword(s):

Pearson Correlation ◽

Dietary Assessment ◽

Correlation Coefficients ◽

Weighted Kappa ◽

Epidemiological Studies ◽

Food Groups ◽

Nutrient Intakes ◽

Dietary Assessment Methods ◽

Daily Food ◽

Almost All

AbstractObjectiveFFQ are commonly used to rank individuals by their food and nutrient intakes in large epidemiological studies. The purpose of the present study was to develop and validate an FFQ to rank individuals participating in an ongoing Prospective Urban and Rural Epidemiological (PURE) study in Chile.DesignAn FFQ and four 24 h dietary recalls were completed over 1 year. Pearson correlation coefficients, energy-adjusted and de-attenuated correlations and weighted kappa were computed between the dietary recalls and the FFQ. The level of agreement between the two dietary assessment methods was evaluated by Bland–Altman analysis.SettingTemuco, Chile.SubjectsOverall, 166 women and men enrolled in the present study. One hundred men and women participated in FFQ development and sixty-six individuals participated in FFQ validation.ResultsThe FFQ consisted of 109 food items. For nutrients, the crude correlation coefficients between the dietary recalls and FFQ varied from 0·14 (protein) to 0·44 (fat). Energy adjustment and de-attenuation improved correlation coefficients and almost all correlation coefficients exceeded 0·40. Similar correlation coefficients were observed for food groups; the highest de-attenuated energy-adjusted correlation coefficient was found for margarine and butter (0·75) and the lowest for potatoes (0·12).ConclusionsThe FFQ showed moderate to high agreement for most nutrients and food groups, and can be used to rank individuals based on energy, nutrient and food intakes. The validation study was conducted in a unique setting and indicated that the tool is valid for use by adults in Chile.

Download Full-text

Agreement between an Image-Based Dietary Assessment Method and a Written Food Diary among Adolescents with Type 1 Diabetes

Nutrients ◽

10.3390/nu13041319 ◽

2021 ◽

Vol 13 (4) ◽

pp. 1319

Author(s):

Laura Heikkilä ◽

Marja Vanhala ◽

Raija Korpelainen ◽

Päivi Tossavainen

Keyword(s):

Type 1 Diabetes ◽

Dietary Intake ◽

Assessment Tool ◽

Dietary Assessment ◽

Intraclass Correlation ◽

Correlation Coefficients ◽

Assessment Method ◽

Food Diary ◽

Total Carbohydrate

Valid and useful dietary assessment methods for adolescents with type 1 diabetes (T1D) are needed. In this study, we compared an image-based method with a written food diary for dietary intake estimation among adolescents with T1D and evaluated the adolescents’ experiences of the methods. Adolescents with T1D aged 13 to 18 years (n = 13) photographed their meals (n = 264) with a mobile phone camera and simultaneously kept a written food diary for four consecutive days. The participants filled out electronic background and feedback questionnaires. The agreement between the methods was evaluated using intraclass correlation coefficients (ICCs) and Bland–Altman plot analyses. The agreement between the methods was moderate to excellent for the energy intake (ICC = 0.91, 95% confidence interval (CI): 0.66 to 0.97, p < 0.001) and good to excellent for total carbohydrate intake (ICC = 0.95, 95% CI: 0.84 to 0.99, p < 0.001). The adolescents considered photographing easier and faster than keeping a food diary. In conclusion, the image-based method appeared comparable to the food diary for dietary intake estimation among adolescents with T1D. The photographing of meals may become a useful dietary assessment tool for adolescents with T1D, but must be further developed and validated.

Download Full-text

Development and evaluation of image-series for portion size estimation in dietary assessment among adults

Journal of Nutritional Science ◽

10.1017/jns.2020.58 ◽

2021 ◽

Vol 10 ◽

Author(s):

Lorentz Salvesen ◽

Dagrun Engeset ◽

Nina C. Øverby ◽

Anine C. Medin

Keyword(s):

Assessment Tool ◽

Tertiary Education ◽

Dietary Assessment ◽

Portion Size ◽

Estimation Accuracy ◽

Size Estimation ◽

Web Based ◽

Food Items ◽

Image Series ◽

Portion Size Estimation

Abstract Portion size images are advantageous in dietary assessment. The aim of the present study was to develop and validate new culturally specific image-series for portion size estimation to be used in a new Norwegian version of a British web-based dietary assessment tool (myfood24). Twenty-three image-series of different foods, each containing seven portion size images, were created and validated in a group of adults (n 41, 58 % female) aged 19–44 (median 23), out of which 63 % had higher (tertiary) education. The participants compared 46 portions of pre-weighed foods to the portion size images (1886 comparisons in total). Portion size estimations were either classified as correct, adjacent or misclassified. The weight discrepancy in percentage between the chosen and the correct portion size image was also calculated. Mann–Whitney U tests were used to explore if portion size estimation accuracy differed across sample characteristics, or if it depended on how the foods were presented. For thirty-eight of the forty-six presented food items, the participants selected the correct or adjacent portion size image 98 % on average. The remaining eight food items were on average misclassified by 27 % of the participants. Overall, a mean weight discrepancy of 2⋅5 % was observed between the chosen and the correct portion size images. Females estimated portion size more accurately than males (P = 0⋅019). No other significant differences in estimation accuracy were observed. In conclusion, the new image-series performed satisfactorily, except for the image-series depicting bread, caviar spread and marzipan cake, which will be altered. The present study demonstrates the importance of validating portion size estimation tools.

Download Full-text

The Youth Throwing Score: Validating Injury Assessment in Young Baseball Players

The American Journal of Sports Medicine ◽

10.1177/0363546516667503 ◽

2016 ◽

Vol 45 (2) ◽

pp. 317-324 ◽

Cited By ~ 12

Author(s):

Christopher S. Ahmad ◽

Ajay S. Padaki ◽

Manish S. Noticewala ◽

Eric C. Makhni ◽

Charles A. Popkin

Keyword(s):

Upper Extremity ◽

Assessment Tool ◽

Pearson Correlation ◽

Intraclass Correlation ◽

Correlation Coefficients ◽

Importance Rating ◽

Extremity Injury ◽

Level Of Evidence ◽

Baseball Players ◽

Upper Extremity Injury

Background: Epidemic levels of shoulder and elbow injuries have been reported recently in youth and adolescent baseball players. Despite the concerning frequency of these injuries, no instrument has been validated to assess upper extremity injury in this patient population. Purpose/Hypothesis: The purpose of this study was to validate an upper extremity assessment tool specifically designed for young baseball players. We hypothesized that this tool will be both reliable and valid. Study Design: Cohort study (diagnosis); Level of evidence, 2. Methods: The Youth Throwing Score (YTS) was constructed by an interdisciplinary team of providers and coaches as a tool to assess upper extremity injury in youth and adolescent baseball players (age range, 10-18 years). The psychometric properties of the test were then determined. Results: A total of 223 players completed the final survey. The players’ mean age was 14.3 ± 2.7 years. Pilot analysis showed that none of the 14 questions received a mean athlete importance rating less than 3 of 5, and the final survey read at a Flesch-Kincaid level of 4.1, which is appropriate for patients aged 9 years and older. The players self-assigned their injury status, resulting in a mean instrument score of 59.7 ± 8.4 for the 148 players “playing without pain,” 42.0 ± 11.5 for the 60 players “playing with pain,” and 40.4 ± 10.5 for the 15 players “not playing due to pain.” Players playing without pain scored significantly higher than those playing with pain and those not playing due to pain ( P < .001). Psychometric analysis showed a test-retest intraclass correlation coefficient of 0.90 and a Cronbach alpha intra-item reliability coefficient of 0.93, indicating excellent reliability and internal consistency. Pearson correlation coefficients of 0.65, 0.62, and 0.31 were calculated between the YTS and the Pediatric Outcomes Data Collection Instrument sports/physical functioning module, the Kerlan-Jobe Orthopaedic Clinic Shoulder and Elbow score, and the Quick Disabilities of the Arm, Shoulder, and Hand (QuickDASH) score, respectively. Injured players scored a mean of 9.4 points higher after treatment ( P < .001), and players who improved in their self-assigned pain categorization scored 16.5 points higher ( P < .001). Conclusion: The YTS is the first valid and reliable instrument for assessing young baseball players’ upper extremity health.

Download Full-text