scholarly journals Development of quality index to classify meal healthiness through photos: first step for app of meal assessment using Machine Learning

2020 ◽  
Vol 79 (OCE2) ◽  
Author(s):  
Josiane Steluti ◽  
Jun Okamoto Junior ◽  
Dirce Maria Lobo Marchioni
Keyword(s):  
Food Intake ◽  
Quality Index ◽  
Processed Food ◽  
Cooking Method ◽  
Meat Intake ◽  
Significant Agreement ◽  
Meat Substitute ◽  
Meal Quality ◽  
Food Variety ◽  
Index Components

AbstractIntroduction:Currently, there are approximately 84 million smartphones in use in Brazil. The use of this technology facilitates the daily life of the individuals in the personal, social and professional scope. Therefore, we hypothesize that it can be used to assess and improve dietary intake. Thus, we intend to develop an application to assess meal quality from a photo taken by a smartphone, using artificial intelligence. For this, the machine should be trained to recognize which meal is healthy or unhealthy, and, as a first step, a meal quality index is necessary.Objective:This study aims to develop a meal quality index, to be applied to photos of dishes from a main meal.Methods:For the development of the index, it was considered the main recommendations established in the scientific literature and the “Dietary Guidelines for the Brazilian Population”. The index includes nine components: 1- meat intake and/or meat substitute, 2- cooking method, 3- vegetables intake, 4- whole food intake, 5- food variety, 6- ultra-processed food intake, 7- fruits intake, 8- carbohydrates intake, 9- fat recipes and/or food. First, questions were elaborated and scored as 0 point (unhealthy answer) or 1 point (healthy answer). After, the meal photo was classified as “needs improvement” (< 4 points), good (> 5 e < 7 points) and very good (> 8 points). Each photo was assessed by two experts. Then, statistical analyses were performed considering Kappa (k) statistic to evaluate the agreement between the assessments by experts.Results:Data from 154 meal photos were assessed. We analyzed the % of agreement, k-value and significant agreement (p-value) for all index components and final classification, respectively: 1- meat intake and/or meat substitute, 64.94% and k = 0.2759 (p < 0.001); 2- cooking method, 81.82% and k = 0.5915 (p = 0.000); 3- vegetables intake, 77.27% and k = 0.5353 (p = 0.000); 4- whole food intake, 98.05% and k = -0.0087 (p = 0.545); 5- food variety, 79.22% and k = 0.5899 (p = 0.000); 6- ultra processed food intake, 76.62% and k = 0.4515 (p = 0.000); 7- fruits intake, 94.81% and k = 0.8494 (p = 0.000); 8- carbohydrates intake, 65.58% and k = 0.2960 (p = 0.000); 9- fat recipes and/or food, 73.38% and k = 0.4654 (p = 0.000); and final classification, 58.44% and k = 0.3218 (p = 0.000).Conclusion:We verified a moderate and significant agreement for almost all index components using meal photo.

Abstract MP53: Consumption Of Ultra-processed Food Is Associated With Risk Of Coronary Heart Disease In The Atherosclerosis Risk In Communities Study

Circulation ◽  
2021 ◽  
Vol 143 (Suppl_1) ◽  
Author(s):  
Shutong Du ◽  
Hyunju Kim ◽  
Josef Coresh ◽  
Casey M Rebholz
Keyword(s):  
Coronary Heart Disease ◽  
Heart Disease ◽  
Food Intake ◽  
Proportional Hazards ◽  
Sociodemographic Factors ◽  
Incidence Rates ◽  
Cox Proportional Hazards ◽  
Processed Food ◽  
Cox Proportional Hazards Models ◽  
Food And Drink

Introduction: Ultra-processed food defined as food and drink products formulated through sequences of industrial processes, and generally contain non-culinary used additives. Previous studies have linked higher ultra-processed food intake with several cardiometabolic and cardiovascular diseases. However, longitudinal evidence from US populations remains scarce. Hypothesis: We hypothesized that higher intake of ultra-processed food is associated with higher risk of coronary heart disease (CHD). Methods: We selected 12,607 adults aged 44-66 years in 4 US communities from the ARIC study at baseline. Dietary intake data were collected through a validated 66-item food frequency questionnaire. Ultra-processed foods were defined using the NOVA classification and the level of intake was calculated for each participant. We conducted Cox proportional hazards models to study the association between quartiles of ultra-processed food intake and incident CHD. Nonlinearity was assessed by using restricted cubic spline regression. Results: There were 1,899 incident CHD cases documented after an median follow up of 27 years (291,285.2 person-years). Incidence rates were higher in the highest quartile of ultra-processed food intake (71.6 per 10,000 person-years; 95% CI, 65.8-78.0) compared to the lowest quartile (59.7 per 10,000 person-years; 95% CI, 54.3-65.7). Participants in the highest vs. lowest quartile were associated with a 18% higher risk of CHD (Hazard ratio 1.18 [95% CI, 1.04 - 1.34]; P-trend = 0.010) after adjusting for sociodemographic factors and health behaviors. An approximately linear relationship was observed between ultra-processed food intake and risk of CHD after 4 servings/day ( Figure ). Conclusion: In conclusion, higher ultra-processed food intake was associated with a higher risk of coronary heart disease among middle-aged US adults. Further prospective studies are needed to confirm these findings and to investigate the mechanisms by which ultra-processed food may affect health.

scholarly journals Consumption of ultra-processed foods and the risk of overweight and obesity, and weight trajectories in the French cohort NutriNet-Santé

2020 ◽  
Vol 79 (OCE2) ◽  
Author(s):  
Bernard Srour ◽  
Marie Beslay ◽  
Caroline Méjean ◽  
Benjamin Allès ◽  
Thibault Fiolet ◽  
...  
Keyword(s):  
Food Intake ◽  
Confidence Interval ◽  
Prospective Study ◽  
Food Consumption ◽  
Overweight And Obesity ◽  
Processed Foods ◽  
Processed Food ◽  
Weight Trajectories ◽  
Incident Cases

AbstractIntroductionPrevious epidemiological studies have found associations between the consumption of ultra-processed foods and the risk of obesity-related outcomes, such as post-menopausal breast cancer, cardiovascular diseases, hypertension and mortality. However, only one Spanish prospective study has explored the associations between the consumption of ultra-processed foods and the risk of overweight and obesity. The aim of this study is to investigate the associations between ultra-processed food consumption and the risk of overweight and obesity, as well as the associations between ultra-processed food consumption and weight trajectories, in middle-aged adults included in the French large scale NutriNet-Santé cohort.MethodsOverall, 110260 participants aged at least 18 years from the French NutriNet-Santé cohort (2009–2019) were included. Dietary intakes were collected using repeated 24 hour dietary records, merged with a food composition database of 3300 different products. These were categorized according to their degree of processing by the NOVA classification. Associations between ultra-processed food intake and risks of overweight and obesity were assessed using multivariable Cox proportional hazard models. Associations between ultra-processed food intake and weight trajectories were assessed using multivariable linear mixed models for repeated measures with random slope and intercept. Models were adjusted for known risk factors (sociodemographic, lifestyle, and nutritional factors).ResultsUltra-processed food intake was associated with a higher risk of overweight (n = 7063 incident cases; hazard ratio for an absolute increment of 10 in the percentage of ultra-processed foods in the diet = 1.11 (95% confidence interval 1.08 to 1.14); P < 0.0001, median follow-up: 4.1y, 260304 person-years) and obesity (n = 3066 incident cases; HR = 1.09 (95% confidence interval 1.05 to 1.13); P < 0.0001, median follow-up: 8.0y 365344 person-years). Higher consumers of ultra-processed foods (4th quartile) were more likely to present an increase in body mass index over time (change of BMI/time-unit in Q4 vs Q1 = 0.04, P < 0.0001). These results remained statistically significant after adjustment for several markers of the nutritional quality of the diet (fruits and vegetables and sugary drinks consumption, intakes of saturated fatty acids, sodium, sugar, dietary fiber or Healthy/Western patterns derived by principal component analysis) and after a large range of sensitivity analyses.ConclusionIn this large observational prospective study, higher consumption of ultra-processed foods in the diet was associated with a higher risk of overweight and obesity. Public health authorities in several countries recently started to recommend privileging unprocessed/minimally processed foods and limiting ultra-processed food consumption.

Highly processed food intake and immediate and future emotions in everyday life

Appetite ◽  
2021 ◽  
pp. 105868
Author(s):  
Jenna R. Cummings ◽  
Emma T. Schiestl ◽  
A. Janet Tomiyama ◽  
Tanvi Mamtora ◽  
Ashley N. Gearhardt
Keyword(s):  
Food Intake ◽  
Everyday Life ◽  
Processed Food

Abstract P181: Processed food intake is adversely associated with multiple indicators of cardiometabolic health in US adults: Findings from the National Health and Nutrition Examination Surveys (2005-2008)

Circulation ◽  
2014 ◽  
Vol 129 (suppl_1) ◽  
Author(s):  
Leah Lipsky ◽  
Tonja Nansel ◽  
Virginia Quick
Keyword(s):  
Waist Circumference ◽  
Food Intake ◽  
Weight Status ◽  
Cardiometabolic Health ◽  
Processed Foods ◽  
Processed Food ◽  
C Reactive Protein ◽  
Reactive Protein ◽  
Former Smokers ◽  
Never Smokers

Introduction: Reducing intake of ingredients characteristic of processed foods is vital to improving ideal cardiovascular dietary behaviors described in the American Heart Association 2020 Strategic Impact Goals. Hypothesis: We sought to develop an indicator of processed food intake and evaluate its hypothesized adverse relationships with biomarkers of cardiometabolic health in a nationally representative sample of US adults. Methods: Data from two 24 hour recalls were examined for US adults (>=18y) in NHANES (2005-2008). An index of processed food intake (PFI) was developed using the mean of the standardized (mean=0, standard deviation=1), energy-adjusted (per 1000 kcal) intakes of refined grains, processed meat, discretionary oils, discretionary solid fat, added sugar and sodium. We evaluated bivariate associations of PFI with demographics (sex, poverty-income ratio, education) and behavioral factors (smoking, nutritional supplement use). Multivariable linear regressions were used to examine associations of PFI with BMI (kg/m2), waist circumference (cm), and biomarkers for cardiometabolic health (total cholesterol, HDL-C, LDL-C, triglycerides, apo-b and c-reactive protein), adjusting for demographic and behavioral covariates. We tested for potential interactions between PFI and weight status, sex, and smoking. Results: PFI was higher in smokers than never smokers (p<.001). PFI was lower for those with at least a college degree than those with less education (p=.004) and for NH White vs. NH Black adults (p=.04). Adjusting for covariates, higher PFI was associated with greater BMI (p<.001) and waist circumference (p<.001), lower HDL-C (p<.001), and higher c-reactive protein (p=.01). Interactions (p<.05) were observed between PFI and sex for predicting BMI, and between PFI and smoking for predicting TC and HDL-C. The magnitude of associations was larger for female vs. male and for current and former smokers vs. non-smokers. No interactions were observed between PFI and weight status. Conclusion: Intake of components characteristic of processed foods is adversely associated with a variety of cardiometabolic biomarkers. Positive associations of PFI with BMI were greater for females vs. males, while associations of PFI with TC and HDL-C were greater for current and former smokers vs. never smokers. The nutritional value of dietary components of PFI is primarily restricted to energy, protein, and sodium, none of which are considered lacking in the diets of US adults. These findings underscore the rationale for encouraging replacing such components with foods that promote cardiovascular health including fruit, vegetables, whole grains, fish, legumes, nuts and seeds. Acknowledgment: This research was supported in part by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health.

scholarly journals Evaluation of a mobile application for estimation of food intake

2017 ◽  
Vol 30 (1) ◽  
pp. 91-98 ◽  
Author(s):  
Samantha Bittencourt MESCOLOTO ◽  
Simone CAIVANO ◽  
Semíramis Martins Álvares DOMENE
Keyword(s):  
Food Intake ◽  
University Students ◽  
Diet Quality ◽  
Mobile Application ◽  
Quality Index ◽  
Sao Paulo ◽  
São Paulo ◽  
Strong Correlations ◽  
Dietary Recall ◽  
Diet Quality Index

ABSTRACT Objective: Evaluate the use of the Nutrabem (São Paulo, Brasil) mobile application as a tool for measurement of food intake among university students. Methods: Cross-sectional study of a random sample of 40 undergraduate students at the Universidade Federal de São Paulo, Campus Baixada Santista. Food intake data were estimated using the Nutrabem app and the 24-hour dietary recall. Intakes of energy, carbohydrates, proteins, lipids, calcium, iron, and vitamin C were calculated. The intake of food groups and diet quality were evaluated by the Diet Quality Index associated with the Digital Food Guide. The agreement between the methods was assessed using the Pearson's correlation coefficient and the Student' t-test. Results: Strong correlations were observed between energy (0.77), carbohydrates (0.82) and protein (0.83). The groups: poultry, fish, and eggs; beef and pork; refined grains and breads; and fruits and legumes showed strong correlations (between 0.76 and 0.85). There were moderate correlations (0.59 and 0.71) between the groups sugars and sweets; whole grains, tubers and roots, milk and dairy products, animal fats, and the Diet Quality Index associated with the Digital Food Guide scores. Vegetables and leafy greens, nuts, and vegetable oils showed weak correlations (0.31 and 0.43). Homogeneity assessment revealed similarity between the results obtained by both methods (p>0.05) . Conclusion: The Nutrabem app can be used as a tool to assess dietary intake among university students since it produces results similar to those obtained by the 24-hour dietary recall method.

scholarly journals Contribution of Ultra-processed Food to the Daily Food Intake of HIV-positive and HIV-Negative Women during Pregnancy

2019 ◽  
Vol 41 (10) ◽  
pp. 588-596 ◽  
Author(s):  
Clarissa de Oliveira Agostini ◽  
Ester Zoche ◽  
Rafaela da Silveira Corrêa ◽  
Eunice Beatriz Martin Chaves ◽  
Helena von Eye Corleta ◽  
...  
Keyword(s):  
Food Intake ◽  
Demographic Data ◽  
Caloric Intake ◽  
Hiv Positive ◽  
Fisher Exact Test ◽  
Processed Foods ◽  
Processed Food ◽  
Trans Fat ◽  
Positive Group ◽  
Hiv Negative

Objective To assess the daily dietary intake and energy contribution of ultra-processed foods among women who are positive and negative for the human immunodeficiency virus (HIV) during pregnancy. Methods This case–control study included 77 HIV-positive and 79 HIV-negative puerperal women between 2015 and 2016. The socioeconomic and maternal demographic data were assessed, and a food frequency questionnaire (FFQ) adapted for pregnant women was applied. The Fisher exact test and the Mann-Whitney test were applied to detect differences between the groups. Linear regression was used to assess the associations between the intake of ultra-processed food and energy, macro- and micronutrients, with values of p < 0.05 considered significant. Results The HIV-positive group was older (p < 0.001) and had lower income (p = 0.016) and level of schooling (p < 0.001) than the HIV-negative group. Both groups presented similar average food intake: 4,082.99 Kcal/day and 4,369.24 Kcal/day for the HIV-positive and HIV-negative women respectively (p = 0.258).The HIV-positive group consumed less protein (p = 0.048), carbohydrates (p = 0.028) and calcium (p = 0.001), and more total fats (p = 0.003). Ultra-processed foods accounted for 39.80% and 40.10% of the HIV-positive and HIV-negative groups' caloric intake respectively (p = 0.893). The intake of these foods was associated with a higher consumption of carbohydrates (p < 0.001), trans fat (p = 0.013) and sodium (p < 0.001), as well as lower protein (p < 0.001) and fiber intake (p = 0.022). Conclusion These findings demonstrate that the energy consumption and ultra-processed food intake were similar in both groups, which reinforces the trend toward a high intake of ultra-processed food in the general population. The intake of ultra-processed food was positively associated with the consumption of carbohydrates, trans fat and sodium, and negatively associated with the consumption of protein and fiber.

scholarly journals Current Intake of Ultra-Processed Foods in the U.S. Adult Population According to Education-Level and Income

2021 ◽  
Vol 5 (Supplement_2) ◽  
pp. 418-418
Author(s):  
Filippa Juul ◽  
Niyati Parekh ◽  
Euridice Martinez-Steele ◽  
Carlos Augusto Monteiro ◽  
Virginia Chang
Keyword(s):  
High School ◽  
Food Intake ◽  
Family Income ◽  
Adult Population ◽  
Processed Foods ◽  
Processed Food ◽  
High School Degree ◽  
Poverty Threshold ◽  
Socioeconomic Strata ◽  
The U.S

Abstract Objectives Ultra-processed food have been associated with multiple chronic diseases, yet recent data regarding its consumption in the U.S. and potential differences in intake across population groups is lacking. We determined the intake of ultra-processed food across diverse socioeconomic strata in the U.S. adult population. Methods We performed cross-sectional analysis of dietary intake among adults (&gt;20y, N = 9759) in the National Health and Nutrition Examination Survey (NHANES) 2015–2018. Data on dietary intake was collected by 24h dietary recall. Foods were classified as ultra-processed/non ultra-processed according to the NOVA classification. We determined intake of ultra-processed food (%kcal) in the overall sample, and stratified by education (&lt;high school, high school degree, some college, college graduate of above) and family poverty income ratio, (&lt;130%, 130–349% and ≥ 350% of the federal poverty threshold). Multivariable linear regression was used to assess if education and income were independent predictors of ultra-processed food intake, controlling for age, sex and race/ethnicity. Results Ultra-processed foods provided 54% of energy among U.S. adults in 2015–2018. Compared to adults without a high school degree (52%kcal), high school graduates and adults with some college education consumed significantly more ultra-processed foods (57% kcal, P = 0.022 and 57.0% kcal, P = 0.009, resp.), while college graduates consumed significantly less ultra-processed foods (49% kcal, P &lt; 0.001). Adults with a family income of 130–349% of the federal poverty threshold consumed significantly more ultra-processed foods than adults with the lowest family income (56 vs. 54% kcal, P = 0.009). However, intake did not differ significantly between adults with low and high income (52% kcal, P = 0.817). Conclusions This study uniquely describes ultra-processed food consumption across socioeconomic groups in the U.S. population and may inform policies and intervention to reduce intakes of ultra-processed foods and prevent chronic disease outcomes. Although consumption differed across education- and income levels, ultra-processed food intake is high in all socioeconomic strata. Our results highlight the need for public health efforts to reduce ultra-processed food consumption in the U.S. Funding Sources None.

scholarly journals Reproducibility study of the meal quality index

2014 ◽  
Vol 41 (1) ◽  
pp. 72-76 ◽  
Author(s):  
Roseane Sampaio Moreira Barbosa ◽  
Haydée Serrão Lanzillotti ◽  
Patricia Henriques
Keyword(s):  
Quality Index ◽  
Reproducibility Study ◽  
Meal Quality
Sign in / Sign up

Export Citation Format

Share Document