Evaluation of the Ability of Diet-Tracking Mobile Applications to Estimate Energy and Nutrient Intake in Japan

Evaluating the Quality and Comparative Validity of Manual Food Logging and Artificial Intelligence-Enabled Food Image Recognition in Apps for Nutrition Care

For artificial intelligence (AI) to support nutrition care, high quality and accuracy of its features within smartphone applications (apps) are essential. This study evaluated popular apps' features, quality, behaviour change potential, and comparative validity of dietary assessment via manual logging and AI. The top 200 free and paid nutrition-related apps from Australia's Apple App and Google Play stores were screened (n = 800). Apps were assessed using MARS (quality) and ABACUS (behaviour change potential). Nutritional outputs from manual food logging and AI-enabled food-image recognition apps were compared with food records for Western, Asian, and Recommended diets. Among 18 apps, Noom scored highest on MARS (mean = 4.44) and ABACUS (21/21). From 16 manual food-logging apps, energy was overestimated for Western (mean: 1040 kJ) but underestimated for Asian (mean: -1520 kJ) diets. MyFitnessPal and Fastic had the highest accuracy (97% and 92%, respectively) out of seven AI-enabled food image recognition apps. Apps with more AI integration demonstrated better functionality, but automatic energy estimations from AI-enabled food image recognition were inaccurate. To enhance the integration of apps into nutrition care, collaborating with dietitians is essential for improving their credibility and comparative validity by expanding food databases. Moreover, training AI models are needed to improve AI-enabled food recognition, especially for mixed dishes and culturally diverse foods.

A Study on the Compatibility of a Food-Recording Application with Questionnaire-Based Methods in Healthy Japanese Individuals

In Japan, nutritional guidance based on food-recording apps and food frequency questionnaires (FFQs) is becoming popular. However, it is not always recognized that different dietary assessment methods have different nutritional values. Here, we compared the compatibility of dietary intake data obtained from an app with those obtained from FFQs in 59 healthy individuals who recorded information regarding their diet for at least 7 days per month using an app developed by Asken (Tokyo, Japan). The diurnal coefficient of variation in total energy and protein intake was 20%, but those for vitamins B12 and D were >80%, reflecting the importance of 7 days of recording rather than a single day of recording for dietary intake analyses. Then, we compared the results of two FFQs-one based on food groups and one based on a brief self-administered diet history questionnaire-for 7 days, as recorded by the app. There was a correlation coefficient of >0.4 for all the items except salt. Regarding the compatibility between the app and FFQs, the percentage errors for total energy and nutrients were >40-50%, suggesting no agreement between the app and the two FFQs. In conclusion, careful attention should be paid to the impact of different dietary assessment methods on nutrient assessment.

Differences in Portion Sizes in Brazil, France, and the USA

Portion size is recognized as a major determinant of food intake, at least over the short term, and could be related to overconsumption and obesity. In this study, we developed and evaluated a new visual measure of portion size (PS), examined whether the PS of chicken, ice cream, and soda varied among people in Brazil, France, and the USA, and tested whether PS was related to gender, body mass index, body weight, and socioeconomic status. We conducted a cross-sectional study using online convenience samples of university students (total N = 1391). Across all three foods, French personal and country PSs were significantly smaller than the other three countries. Estimated country PS was reliably larger than personal PS. Women's personal PSs were smaller than men's, but women's and men's estimates for country PS were similar. French personal and country PSs were the lowest. Some PSs had a weak but significant correlation with SES but were not significantly related to either weight or BMI. The study confirms French-American differences in personal PS and demonstrates that perceived norms correspond to individual PS.

Mitigating errors in mobile-based dietary assessments: Effects of a data modification process on the validity of an image-assisted food and nutrition app

OBJECTIVE

Mobile nutrition applications (apps) provide a simple way for individuals to record their diet, but the validity and inherent errors need to be carefully evaluated. The aim of this study was to assess the validity and clarify the sources of measurement errors of image-assisted mobile nutrition apps.

METHODS

This was a cross-sectional study with 98 students recruited from School of Nutrition and Health Sciences, Taipei Medical University. A 3-d nutrient intake record by Formosa Food and Nutrient Recording App (FoodApp) was compared with a 24-h dietary recall (24-HDR). A two-stage data modification process, manual data cleaning, and reanalyzing of prepackaged foods were employed to address inherent errors. Nutrient intake levels obtained by the two methods were compared with the recommended daily intake (DRI), Taiwan. Paired t test, Spearman's correlation coefficients, and Bland-Altman plots were used to assess agreement between the FoodApp and 24-HDR.

RESULTS

Manual data cleaning identified 166 food coding errors (12%; stage 1), and 426 food codes with missing micronutrients (32%) were reanalyzed (stage 2). Positive linear trends were observed for total energy and micronutrient intake (all Ptrend < 0.05) after the two stages of data modification, but not for dietary fat, carbohydrates, or vitamin D. There were no statistical differences in mean energy and macronutrient intake between the FoodApp and 24-HDR, and this agreement was confirmed by Bland-Altman plots. Spearman's correlation analyses showed strong to moderate correlations (r = 0.834 ∼ 0.386) between the two methods. Participants' nutrient intake tended to be lower than the DRI, but no differences in proportions of adequacy/inadequacy for DRI values were observed between the two methods.

CONCLUSIONS

Mitigating errors significantly improved the accuracy of the Formosa FoodApp, indicating its validity and reliability as a self-reporting mobile-based dietary assessment tool. Dietitians and health professionals should be mindful of potential errors associated with self-reporting nutrition apps, and manual data cleaning is vital to obtain reliable nutrient intake data.

Association of Eating Pattern, Chronotype, and Social Jetlag: A Cross-Sectional Study Using Data Accumulated in a Japanese Food-Logging Mobile Health Application

Chronotype (morningness-eveningness) and social jetlag (SJL; discrepancy in the sleep pattern between the weekday and weekend) are related to eating behavior and health. The association between sleep behavior and the daily macro- and micronutrient eating pattern of each meal (breakfast, lunch, and dinner) have not been discussed well and need more evidence. Here, meal pattern datasets of Japanese participants aged 20-59 years were obtained as averages over 1 month from the data stored in the food-logging app "Asken". We allocated three groups for each chronotype and SJL. Multiple regression analyses revealed that morning chronotype and small SJL were associated with higher total daily intake of potassium, fiber, magnesium, phosphorus, and vitamin K. Breakfast energy intake and consumption of nutrients, including protein, lipid, carbohydrate, and minerals, were higher in the morning chronotype or small SJL. Lunch intake of potassium, cholesterol, fiber, magnesium, and vitamin K was also higher in the morning chronotype or small SJL. Dinner energy intake and nutrient intake of proteins, lipids, carbohydrates, sodium, and saturated fatty acids were lower in the morning chronotype or small SJL. The current data would help to establish a detailed reference for dietary intake which considers eating patterns over a day.

Effects of a smartphone app-based diet and physical activity program for men living with HIV who have dyslipidemia: A pilot randomized controlled trial

AIMS

People living with HIV are at a high risk for cardiovascular disease owing to antiretroviral therapy use and chronic inflammation. There is limited evidence on the evaluation of serum lipid levels through lifestyle modification. This study aims to evaluate the effect of a smartphone application-based diet and exercise improvement program on men living with HIV and dyslipidemia.

METHODS

This was a randomized controlled trial recruiting Japanese men living with HIV who have dyslipidemia: intervention group (n = 19) and control group (n = 19). The intervention group received a third individual guidance session during the 6-month intervention and was encouraged to record their diet on a smartphone application. An intention-to-treat analysis of the results was conducted.

RESULTS

The intervention group showed significantly reduced change in low-density lipoprotein levels compared to the control group (-4.00 ± 20.2 mg/dL vs. 10.11 ± 21.1 mg/dL) (p = .042) from baseline to 6 months post-intervention. No significant differences were found in other serum lipid levels. Abdominal circumference decreased significantly in the intervention group (p = .048) from baseline to 6 months post-intervention. Total energy, protein, carbohydrate, fat, and salt intake, dietary and physical activity behavior change stages and social support, dietary self-efficacy, and loneliness significantly improved in the intervention group (p < .05) from baseline to 6 months post-intervention.

CONCLUSIONS

A diet and physical activity improvement program using a smartphone application based on Japanese-specific health guidance may reduce low-density lipoprotein cholesterol levels in this population. Further sample expansion and examination of long-term effects are needed.

Validation of Dietary Intake Estimated by Web-Based Dietary Assessment Methods and Usability Using Dietary Records or 24-h Dietary Recalls: A Scoping Review

The goal was to summarize studies comparing the accuracy of web-based dietary assessments with those of conventional face-to-face or paper-based assessments using 24-h dietary recall or dietary record methods in the general population. Using two databases, mean differences and correlation coefficients (CCs) for intakes of energy, macronutrients, sodium, vegetables, and fruits were extracted from each study independently by the authors. We also collected information regarding usability from articles reporting this. From 17 articles included in this review, the mean dietary intake differences in the web-based dietary assessment compared to conventional methods, were -11.5-16.1% for energy, -12.1-14.9% for protein, -16.7-17.6% for fat, -10.8-8.0% for carbohydrates, -11.2-9.6% for sodium, -27.4-3.9% for vegetables, and -5.1-47.6% for fruits. The CC was 0.17-0.88 for energy, protein, fat, carbohydrates, and sodium, and 0.23-0.85 for vegetables and fruits. In three out of four studies reporting usability, more than half of the participants preferred the web-based dietary assessment. In conclusion, % difference and CC of dietary intake were acceptable in both web-based dietary records and 24-h dietary recalls. The findings from this review highlight the possibility of wide-spread application of the web-based dietary assessment in the future.

Relationship between intra-individual variability in nutrition-related lifestyle behaviors and blood glucose outcomes under free-living conditions in adults without type 2 diabetes

AIMS

This study determined the relationship between intra-individual variability in day-to-day nutrition-related lifestyle behaviors (meal timing, eating window, food intake, movement behaviors, sleep conditions, and body weight) and glycemic outcomes under free-living conditions in adults without type 2 diabetes.

METHODS

We analyzed 104 adults without type 2 diabetes. During the 7-day measurement period, dietary intake, movement behaviors, sleep conditions, and glucose outcomes were assessed. Daily food intake was assessed using a mobile-based health application. Movement behaviors and sleep conditions were assessed using a tri-axial accelerometer. Meal timing was assessed from the participant's daily life record. Blood glucose levels were measured continuously using a glucose monitor. Statistical analyses were conducted using a linear mixed-effects model, with mealtime, food intake, body weight, movement behaviors, and sleep conditions as fixed effects and participants as a random effect.

RESULTS

Dinner time and eating window were positively significantly correlated with mean (dinner time, p = 0.003; eating window, p = 0.001), standard deviation (SD; both at p < 0.001), and maximum (both at p < 0.001) blood glucose levels. Breakfast time was negatively associated with glucose outcomes (p < 0.01). Sedentary time was positively significantly associated with blood glucose SD (p = 0.040). Total sleep time was negatively significantly correlated with SD (p = 0.035) and maximum (p = 0.032) blood glucose levels. Total daily energy intake (p = 0.001), carbohydrate intake (p < 0.001), and body weight (p < 0.05) were positively associated with mean blood glucose levels.

CONCLUSION

Intra-individual variations in nutrition-related lifestyle behaviors, especially morning and evening body weight, and food intake, were associated with mean blood glucose levels, and a long sedentary time and total sleep time were associated with glucose variability. Earlier dinner times and shorter eating windows per day resulted in better glucose control.

Nutrition-related mobile applications - Should they be used for dietary prevention and treatment of cardiovascular diseases?

BACKGROUND AND AIMS

There is no prior research on the usefulness that popular nutrition-related mobile applications would have in assessing fatty acids intake. In this study, we examine these applications through their utilization in the assessment of consumption of saturated (SFAs) and polyunsaturated (PUFAs) fatty acids against the Polish reference method (RM, Dieta 6.0). This report does also include the information about monounsaturated fatty acids and cholesterol intake.

METHODS AND RESULTS

SFAs and PUFAs intake was assessed using two-day dietary recalls obtained from 120 individuals by 3 selected mobile applications (App1 = Yazio, App2 = MyFitnessPal, App3 = Fitatu) and compared with RM. Despite strong (SFAs by App1 and App3) and moderate (SFAs by App2 and PUFAs by App1, App2, App3) correlations with RM, Bland-Altman analyses showed relevant biases and wide range between limits of agreement. Considering SFAs and MUFAs intake, App1 had the best agreement. App1 had high sensitivity (94.6%) in recognition of subjects with SFAs intake >10% with moderate specificity (67.9%), while App2 had poor sensitivity (27.2%) and high specificity (100%). App3 showed moderate sensitivity and specificity (77.2% and 75%, respectively).

CONCLUSIONS

Mobile applications are not accurate tools in SFAs and PUFAs assessment when compared to the RM. Nonetheless, their ability to recognize SFAs intake >10% energy intake may suggest that further development of mobile applications could potentially become an attractive tool in clinical practice.

Dietary counseling based on artificial intelligence for patients with nonalcoholic fatty liver disease

BACKGROUND

About 25% of the general population in Japan are reported to have nonalcoholic fatty liver disease (NAFLD). NAFLD and nonalcoholic steatohepatitis carry a risk of progressing further to hepatocellular carcinoma. The primary treatment for NAFLD is dietary therapy. Dietary counseling plays an essential role in dietary therapy. Although artificial intelligence (AI)-based nutrition management software applications have been developed and put into practical use in recent years, the majority focus on weight loss or muscle strengthening, and no software has been developed for patient use in clinical practice.

AIM

To examine whether effective dietary counseling is possible using AI-based nutrition management software.

METHODS

NAFLD patients who had been assessed using an AI-based nutrition management software application (Calomeal) that automatically analyzed images of meals photographed by patients and agreed to receive dietary counseling were given dietary counseling. Blood biochemistry tests were performed before (baseline) and 6 mo after (6M follow-up) dietary counseling. After the dietary counseling, the patients were asked to complete a questionnaire survey.

RESULTS

A total of 29 patients diagnosed with NAFLD between August 2020 and March 2022 were included. There were significant decreases in liver enzyme and triglyceride levels at the 6M follow-up compared to baseline. The food analysis capability of the AI used by Calomeal in this study was 75.1%. Patient satisfaction with the AI-based dietary counselling was high.

CONCLUSION

AI-based nutrition management appeared to raise awareness of dietary habits among NAFLD patients. However, it did not directly alleviate the burden of registered dietitians, and improvements are much anticipated.

Identification of Factors Necessary for Enabling Technology-Based Dietary Record Surveys: A Qualitative Focus Group Interview with Japanese Dietitians

Weighed food records together with an in-person interview approach constitute the most basic methods used to estimate energy and nutrient intakes in dietary surveys. In the background of the coronavirus disease-2019 pandemic, the need for non-face-to-face dietary surveys using information and communication technology (ICT) is increasing. We aimed to evaluate ICT-based dietary record surveys and identify factors that may enable this survey method to become more widely used in the future. We conducted a non-face-to-face survey of dietary records of 44 Japanese individuals, maintained by dietitians using dietary photography and video conferencing services. We conducted a focus group interview with the six dietitians who conducted that survey. Their opinions on the factors necessary to popularize ICT-based dietary survey method were analyzed. In the focus group interview, dietitians highlighted fewer restrictions on time and place as positive aspects. Negative aspects included insufficient skills to operate computers, difficulty in hearing, and understanding facial expressions using ICT. We identified three main factors for enabling widespread use of ICT-based dietary record survey: individual skill, device and technology, and social environmental factors. This suggests that a comprehensive approach is necessary for popularizing the use of ICT in dietary surveys.

Nutritional Assessment of Hospital Meals by Food-Recording Applications

Mobile food records are currently used to determine the nutrition of healthy subjects. To determine the accuracy of such records, we evaluated the nutritional composition of a test meal (noodles and fruit juice) and a hospital meal (Japanese set meal) using two types of mobile food records. Eighteen healthy subjects (2 males and 16 females) were enrolled. Using these diets and validated nutrient-composition information, we evaluated the accuracy of the dietary assessments made by two dietary-record applications, Asken^® and Calomeal^®, over 5 days. For the test meal, the values provided by the two applications were close to the actual values. In contrast, for the hospital meal, the values provided by the two applications were approximately 1.5 times higher than the actual values. A linear-mixed-model analysis showed that the total energy, carbohydrate, and salt contents were significantly overestimated in the hospital meal. Protein also tended to be overestimated, while the fat content was not significantly overestimated. Furthermore, the total energy and fat contents increased significantly over time. No association with age was observed. A comparison of the coefficients of variation (CVs) for each nutrient in the hospital meal indicated that the fat levels were significantly higher than those in the test meal. In conclusion, the accuracy of mobile food records depends on the type of meal. Our data will provide lessons for the use of meal-recording applications in special cases, such as hospital food.

Association of Japanese Breakfast Intake with Macro- and Micronutrients and Morning Chronotype

(1) Background: Dietary intake may have a remarkable effect on sleep because skipping breakfast and having a late dinner affects many sleep parameters. Breakfast is the most important meal of the day for children and adults to maintain morning chronotype. We examine whether breakfast style is associated with nutrient intake and sleep factors. (2) Methods: This cross-sectional analysis, with a large sample size of 2671 (766 men and 1805 women aged 20–60 years after data brush-up), was based on data obtained from an online survey. Correlation analysis was performed using Spearman’s rank correlation test. The Kruskal–Wallis’s test followed by post hoc Dunn’s multiple comparison test was used to evaluate the interaction between sleep factors and breakfast categories. Multiple regression analyses were performed to identify variables associated with multiple confounding factors. Dietary data were analyzed using approximately one-month average dietary records from the application. The basic characteristics of the participants (age, sex, and BMI) and other lifestyle-related factors (sleep and physical activity) were obtained accordingly. Sleep parameters including the timing of weekday sleep onset, weekday wake-up, weekend (free day) sleep onset, weekend wake-up, sleep, and midpoints of sleep phase were calculated for each participant. We categorized participants’ breakfast types into five groups: (1) Japanese meal, where breakfast may contain Japanese ingredients such as rice; (2) Western meal, where breakfast may contain bread; (3) alternating eating patterns of Japanese and Western meals; (4) cereals and supplements, where breakfast may contain cereals or supplements and energy bars; and (5) skipped breakfast (no breakfast). (3) Results: The midpoint values of the sleep phase on weekends adjusted for sleep debt on work days (MSFsc) related to chronotype were higher in women, suggesting that they may prefer eveningness. Participants with obesity, young age, and low physical activity preferred eveningness with longer sleep durations. Intake of Japanese-style breakfast was significantly associated with early wake-up time on both weekdays and weekends. Cereal-style breakfast intake was significantly associated with late wake-up on both weekdays and weekends. Intake of macronutrients such as protein, fat, carbohydrate, and sodium at breakfast time was positively and strongly associated with the intake of Japanese breakfast, whereas macronutrients were negatively associated with the intake of cereal breakfast. Among micronutrients, vitamin K was positively correlated with Japanese breakfast and negatively correlated with cereal breakfast; (4) Conclusions: Japanese-style breakfast is associated not only with morning preference but also with high intake of macro- and micronutrients.

Validating Accuracy of an Internet-Based Application against USDA Computerized Nutrition Data System for Research on Essential Nutrients among Social-Ethnic Diets for the E-Health Era

Internet-based applications (apps) are rapidly developing in the e-Health era to assess the dietary intake of essential macro-and micro-nutrients for precision nutrition. We, therefore, validated the accuracy of an internet-based app against the Nutrition Data System for Research (NDSR), assessing these essential nutrients among various social-ethnic diet types. The agreement between the two measures using intraclass correlation coefficients was good (0.85) for total calories, but moderate for caloric ranges outside of <1000 (0.75) and >2000 (0.57); and good (>0.75) for most macro- (average: 0.85) and micro-nutrients (average: 0.83) except cobalamin (0.73) and calcium (0.51). The app underestimated nutrients that are associated with protein and fat (protein: −5.82%, fat: −12.78%, vitamin B12: −13.59%, methionine: −8.76%, zinc: −12.49%), while overestimated nutrients that are associated with carbohydrate (fiber: 6.7%, B9: 9.06%). Using artificial intelligence analytics, we confirmed the factors that could contribute to the differences between the two measures for various essential nutrients, and they included caloric ranges; the differences between the two measures for carbohydrates, protein, and fat; and diet types. For total calories, as an example, the source factors that contributed to the differences between the two measures included caloric range (<1000 versus others), fat, and protein; for cobalamin: protein, American, and Japanese diets; and for folate: caloric range (<1000 versus others), carbohydrate, and Italian diet. In the e-Health era, the internet-based app has the capacity to enhance precision nutrition. By identifying and integrating the effects of potential contributing factors in the algorithm of output readings, the accuracy of new app measures could be improved.

Relationship Between Protein Intake in Each Traditional Meal and Physical Activity: Cross-sectional Study

Background

Protein intake plays an important role in the synthesis and maintenance of skeletal muscles for the prevention of health risks. It is also widely known that physical activity influences muscle function. However, no large-scale studies have examined the relationship between daily dietary habits, especially the timing of protein intake, and daily physical activity.

Objective

The purpose of this cross-sectional study was to investigate how protein intake and composition (involving the 3 major nutrients protein, fat, and carbohydrate) in the 3 traditional meals (breakfast, lunch, and dinner) are associated with physical activity.

Methods

Using daily dietary data accumulated in the smartphone food log app “Asken” and a web-based cross-sectional survey involving Asken users (N=8458), we analyzed nutrient intake and composition, as well as daily activity levels. As very few individuals skipped breakfast (1102/19,319 responses, 5.7%), we analyzed data for 3 meals per day.

Results

Spearman rank correlation analysis revealed that breakfast and lunch protein intakes had higher positive correlations with daily physical activity among the 3 major macronutrients (P<.001). These findings were confirmed by multivariate logistic regression analysis with confounding factors. Moreover, participants with higher protein intake and composition at breakfast or lunch tended to exhibit significantly greater physical activity than those with higher protein intake at dinner (P<.001).

Conclusions

Among the 3 macronutrients, protein intake during breakfast and lunch was closely associated with daily physical activity.

The role of body mass index, healthy eating-related apps and educational activities on eating motives and behaviours among women during the COVID-19 pandemic: A cross sectional study

The COVID-19 related lockdown made it much more difficult for people to control their eating behaviours and body weight with the methods and means they had used before. This is reflected in reports that show that eating behaviours deteriorated significantly during the COVID-19 pandemic (including in Poland). Therefore, it is important to determine what factors may be conducive to healthy eating behaviours among people with different BMI. As previous studies show, the use of healthy eating related-apps and training programs may be a protective factor against the development of unhealthy eating behaviours. Therefore, it is worth checking whether their action will be a protective factor during COVID-19. The aim of this cross sectional study was to analyse whether the current use of healthy eating-related apps and previous participation in training in this field (educational activities) as well as body mass index may play a role in eating motives and behaviours among women during COVID-19. Our final sample included 1,447 women (age: M = 31.34 ± 11.05). Participants completed: the Eating Motivation Survey, the Emotional Overeating Questionnaire, the Mindful Eating Questionnaire, socio-demographic survey and questions about healthy eating-related apps and training (educational activities). Referring to the selected significant results, our study shows that during COVID-19, the use of healthy eating-related apps alone, as well as the use of apps and prior training participation promote healthy eating motives and behaviours. It suggests that promoting the use of healthy eating applications and the acquisition of knowledge and skills in this field could be one way of shaping resources that can be effectively used to deal with crisis situations.

Assessing initial validity of the PortionSize app to estimate dietary intake among adults: A pilot and feasibility study (Preprint)

Background

Accurately assessing dietary intake can promote improved nutrition. The PortionSize app (Pennington Biomedical Research Center) was designed to quantify and provide real-time feedback on the intake of energy, food groups, saturated fat, and added sugar.

Objective

This study aimed to assess the preliminary feasibility and validity of estimating food intake via the PortionSize app among adults.

Methods

A total of 15 adults (aged 18-65 years) were recruited and trained to quantify the food intake from a simulated meal by using PortionSize. Trained personnel prepared 15 simulated meals and covertly weighed (weigh back) the amount of food provided to participants as well as food waste. Equivalence tests (±25% bounds) were performed to compare PortionSize to the weigh back method.

Results

Participants were aged a mean of 28 (SD 12) years, and 11 were female. The mean energy intake estimated with PortionSize was 742.9 (SD 328.2) kcal, and that estimated via weigh back was 659.3 (SD 190.7) kcal (energy intake difference: mean 83.5, SD 287.5 kcal). The methods were not equivalent in estimating energy intake (P=.18), and PortionSize overestimated energy intake by 83.5 kcal (12.7%) at the meal level. Estimates of portion sizes (gram weight; P=.01), total sugar (P=.049), fruit servings (P=.01), and dairy servings (P=.047) from PortionSize were equivalent to those estimated via weigh back. PortionSize was not equivalent to weigh back with regard to estimates for carbohydrate (P=.10), fat (P=.32), vegetable (P=.37), grain (P=.31), and protein servings (P=.87).

Conclusions

Due to power limitations, the equivalence tests had large equivalence bounds. Though preliminary, the results of this small pilot study warrant the further adaptation, development, and validation of PortionSize as a means to estimate energy intake and provide users with real-time and actionable dietary feedback.

Association Between Na, K, and Lipid Intake in Each Meal and Blood Pressure

Cardiovascular diseases (CVDs) are one of the leading causes of death worldwide, and one of the most significant risk factors for CVDs is high blood pressure. Blood pressure is associated with various nutrients, such as sodium, potassium, and cholesterol. However, research focusing on the timing of intake of these nutrients and blood pressure has not been conducted. In this study, we used dietary data and a questionnaire asking about the sleep, physical activity, and blood pressure, collected from the food-log app “Asken” (total N = 2,402), to investigate the relationship between the dietary data of nutrient intake in the breakfast, lunch, and dinner and blood pressure. Daily total intake of various nutrients such as sodium, sodium-to-potassium ratio, total energy, lipid, carbohydrate, and saturated fat showed a significant association with blood pressure depending on the meal timing. From multiple regression analysis, eliminating the confounding factors, lunch sodium-to-potassium ratio, dinner energy, lipid, cholesterol, saturated fat, and alcohol intake were positively associated with blood pressure, whereas breakfast protein and lunch fiber intake showed a negative association with blood pressure. Our results suggest that nutrient intake timing is also an important factor in the prevention of high blood pressure. Our study provides possibilities to prevent hypertension by changing the timing of nutrient intake, especially sodium, together with potassium and lipids. However, because our research was limited to food-log app users, broader research regarding the general population needs to be conducted.

Performance of the Digital Dietary Assessment Tool MyFoodRepo

Digital dietary assessment devices could help overcome the limitations of traditional tools to assess dietary intake in clinical and/or epidemiological studies. We evaluated the accuracy of the automated dietary app MyFoodRepo (MFR) against controlled reference values from weighted food diaries (WFD). MFR’s capability to identify, classify and analyze the content of 189 different records was assessed using Cohen and uniform kappa coefficients and linear regressions. MFR identified 98.0% ± 1.5 of all edible components and was not affected by increasing numbers of ingredients. Linear regression analysis showed wide limits of agreement between MFR and WFD methods to estimate energy, carbohydrates, fat, proteins, fiber and alcohol contents of all records and a constant overestimation of proteins, likely reflecting the overestimation of portion sizes for meat, fish and seafood. The MFR mean portion size error was 9.2% ± 48.1 with individual errors ranging between −88.5% and +242.5% compared to true values. Beverages were impacted by the app’s difficulty in correctly identifying the nature of liquids (41.9% ± 17.7 of composed beverages correctly classified). Fair estimations of portion size by MFR, along with its strong segmentation and classification capabilities, resulted in a generally good agreement between MFR and WFD which would be suited for the identification of dietary patterns, eating habits and regime types.

Nutrient and Food Group Prediction as Orchestrated by an Automated Image Recognition System in a Smartphone App (CALO mama): Validation Study

Background

A smartphone image recognition app is expected to be a novel tool for measuring nutrients and food intake, but its performance has not been well evaluated.

Objective

We assessed the accuracy of the performance of an image recognition app called CALO mama in terms of the nutrient and food group contents automatically estimated by the app.

Methods

We prepared 120 meal samples for which the nutrients and food groups were calculated. Next, we predicted the nutrients and food groups included in the meals from their photographs by using (1) automated image recognition only and (2) manual modification after automatic identification.

Results

Predictions generated using only image recognition were similar to the actual data on the weight of meals and were accurate for 11 out of 30 nutrients and 4 out of 15 food groups. The app underestimated energy, 19 nutrients, and 9 food groups, while it overestimated dairy products and confectioneries. After manual modification, the predictions were similar for energy, accurately capturing the nutrients for 29 out of 30 of meals and the food groups for 10 out of 15 meals. The app underestimated pulses, fruits, and meats, while it overestimated weight, vitamin C, vegetables, and confectioneries.

Conclusions

The results of this study suggest that manual modification after prediction using image recognition improves the performance of the app in assessing the nutrients and food groups of meals. Our findings suggest that image recognition has the potential to achieve a description of the dietary intakes of populations by using “precision nutrition” (a comprehensive and dynamic approach to developing tailored nutritional recommendations) for individuals.

An Italian Case Study for Assessing Nutrient Intake through Nutrition-Related Mobile Apps

National food consumption surveys are crucial for monitoring the nutritional status of individuals, defining nutrition policies, estimating dietary exposure, and assessing the environmental impact of the diet. The methods for conducting them are time and resource-consuming, so they are usually carried out after extended periods of time, which does not allow for timely monitoring of any changes in the population’s dietary patterns. This study aims to compare the results of nutrition-related mobile apps that are most popular in Italy, with data obtained with the dietary software Foodsoft 1.0, which was recently used in the Italian national dietary survey IV SCAI. The apps considered in this study were selected according to criteria, such as popularity (downloads > 10,000); Italian language; input characteristics (daily dietary recording ability); output features (calculation of energy and macronutrients associated with consumption), etc. 415 apps in Google Play and 226 in the iTunes Store were examined, then the following five apps were selected: YAZIO, Lifesum, Oreegano, Macro and Fitatu. Twenty 24-hour recalls were extracted from the IV SCAI database and inputted into the apps. Energy and macronutrient intake data were compared with Foodsoft 1.0 output. Good agreement was found between the selected apps and Foodsoft 1.0 (high correlation index), and no significant differences were found in the mean values of energy and macronutrients, except for fat intakes. In conclusion, the selected apps could be a suitable tool for assessing dietary intake.

Low Comparability of Nutrition-Related Mobile Apps against the Polish Reference Method-A Validity Study

Nutrition-related mobile applications (apps) are commonly used to provide information about the user’s dietary intake, however, limited research has been carried out to assess to what extent their results agree with those from the reference method (RM). The main aim of this study was to evaluate the agreement of popular nutrition-related apps with the Polish RM (Dieta 6.0). The dietary data from two days of dietary records previously obtained from adults (60 males and 60 females) were compared with values calculated in five selected apps (FatSecret, YAZIO, Fitatu, MyFitnessPal, and Dine4Fit). The selection of apps was performed between January and February 2021 and based on developed criteria (e.g., availability in the Polish language, access to the food composition database, and the number of downloads). The data was entered by experienced clinical dietitians and checked by one more researcher. The mean age of study participants was 41.7 ± 14.8. We observed that all the apps tended to overestimate the energy intake, however, when considering the macronutrient intake, over- and underestimation were observed. According to our assumed criterion (±5% as perfect agreement, ±10% as sufficient agreement), none of the apps can be recommended as a replacement for the reference method both for scientific as well as clinical use. According to the Bland-Altman analysis, the smallest bias was observed in Dine4Fit in relation to energy, protein, and fat intake (respectively: −23 kcal; −0.7 g, 3 g), however, a wide range between the upper and lower limits of agreement were reported. According to the carbohydrate intake, the lowest bias was observed when FatSecret and Fitatu were used. These results indicate that the leading nutrition-related apps present a critical issue in the assessment of energy and macronutrient intake. Therefore, the implementation of validation studies for quality assessment is crucial to develop apps with satisfying quality.