Principles and Validations of an Artificial Intelligence-Based Recommender System Suggesting Acceptable Food Changes.

BACKGROUND: Along with the popularity of smartphones, artificial intelligence-based personalized suggestions can be seen as promising ways to change eating habits toward more desirable diets. OBJECTIVES: Two issues raised by such technologies were addressed in this study. The first hypothesis tested is a recommender system based on automatically learning simple association rules between dishes of the same meal that would make it possible to identify plausible substitutions for the consumer. The second hypothesis tested is that for an identical set of dietary-swaps suggestions, the more the user is-or thinks to be-involved in the process of identifying the suggestion, the higher is their probability of accepting the suggestion. METHODS: Three studies are presented in this article, first, we present the principles of an algorithm to mine plausible substitutions from a large food consumption database. Second, we evaluate the plausibility of these automatically mined suggestions through the results of online tests conducted for a group of 255 adult participants. Afterward, we investigated the persuasiveness of 3 suggestion methods of such recommendations in a population of 27 healthy adult volunteers through a custom designed smartphone application. RESULTS: The results firstly indicated that a method based on automatic learning of substitution rules between foods performed relatively well identifying plausible swaps suggestions. Regarding the form that should be used to suggest, we found that when users are involved in selecting the most appropriate recommendation for them, the resulting suggestions were more accepted (OR = 3.168; P < 0.0004). CONCLUSIONS: This work indicates that food recommendation algorithms can gain efficiency by taking into account the consumption context and user engagement in the recommendation process. Further research is warranted to identify nutritionally relevant suggestions.

The contribution of milks and formulae to micronutrient intake in 1-3 years old children in urban China: a simulation study.

BACKGROUND AND OBJECTIVES: A recent dietary survey in 5 big cities in China provided information on various milk options consumed by 1-3 years old children. To investigate the nutritional role of these milks (young-child formula (YCF), cow's milk, others), simulation analyses based on this survey were performed. METHODS AND STUDY DESIGN: We studied daily intakes of calcium, iron, zinc, vitamins A, B-1, B-2, C and E and compared these to the Chinese DRIs. In Scenario 1, consumption of cow's milk, kid's milk and/or soy milk was replaced with matching amounts of YCF (n=66 children). In Scenario 2, where 348 children exclusively consumed YCF, YCF was replaced with matching amounts of cow's milk. RESULTS: Scenario 1 revealed significant increases in total dietary intakes of iron, vitamins A, B-1, C and E upon substitution of the various milks with YCF. The proportions of children not meeting the Estimated Average Requirement (EAR) for these nutrients dropped from 29, 26, 61, 53 and 54 % to 12, 11, 50, 27 and 24%, respectively. In Scenario 2, the hypothetical substitution of YCF by cow's milk increased the proportions of children not meeting the EAR for these nutrients, calcium and zinc from 11, 6, 49, 15, 28, 42, and 8 to 45, 24, 78, 69, 59, 44, and 20, respectively. Execution of Scenario 2 in subgroups of 1-2- and 2-3 years old children revealed similar results. CONCLUSIONS: YCF may help to reduce the risk of insufficient intake of several key micronutrients for toddlers, independent of age.

Dietary changes needed to reach nutritional adequacy without increasing diet cost according to income: An analysis among French adults.

OBJECTIVE: To explore the dietary changes needed to achieve nutritional adequacy across income levels at constant energy and diet cost. MATERIALS AND METHODS: Individual diet modelling was used to design iso-caloric, nutritionally adequate optimised diets for each observed diet in a sample of adult normo-reporters aged ≥20 years (n = 1,719) from the Individual and National Dietary Survey (INCA2), 2006-2007. Diet cost was estimated from mean national food prices (2006-2007). A first set of free-cost models explored the impact of optimisation on the variation of diet cost. A second set of iso-cost models explored the dietary changes induced by the optimisation with cost set equal to the observed one. Analyses of dietary changes were conducted by income quintiles, adjusting for energy intake, sociodemographic and socioeconomic variables, and smoking status. RESULTS: The cost of observed diets increased with increasing income quintiles. In free-cost models, the optimisation increased diet cost on average (+0.22 ± 1.03 euros/d) and within each income quintile, with no significant difference between quintiles, but with systematic increases for observed costs lower than 3.85 euros/d. In iso-cost models, it was possible to design nutritionally adequate diets whatever the initial observed cost. On average, the optimisation at iso-cost increased fruits and vegetables (+171 g/day), starchy foods (+121 g/d), water and beverages (+91 g/d), and dairy products (+20 g/d), and decreased the other food groups (e.g. mixed dishes and salted snacks), leading to increased total diet weight (+300 g/d). Those changes were mostly similar across income quintiles, but lower-income individuals needed to introduce significantly more fruit and vegetables than higher-income ones. CONCLUSIONS: In France, the dietary changes needed to reach nutritional adequacy without increasing cost are similar regardless of income, but may be more difficult to implement when the budget for food is lower than 3.85 euros/d.

Individual Diet Modeling Shows How to Balance the Diet of French Adults with or without Excessive Free Sugar Intakes.

Dietary changes needed to achieve nutritional adequacy for 33 nutrients were determined for 1719 adults from a representative French national dietary survey. For each individual, an iso-energy nutritionally adequate diet was generated using diet modeling, staying as close as possible to the observed diet. The French food composition table was completed with free sugar (FS) content. Results were analyzed separately for individuals with FS intakes in their observed diets ≤10% or >10% of their energy intake (named below FS-ACCEPTABLE and FS-EXCESS, respectively). The FS-EXCESS group represented 41% of the total population (average energy intake of 14.2% from FS). Compared with FS-ACCEPTABLE individuals, FS-EXCESS individuals had diets of lower nutritional quality and consumed more energy (2192 vs. 2123 kcal/day), particularly during snacking occasions (258 vs. 131 kcal/day) (all p-values < 0.01). In order to meet nutritional targets, for both FS-ACCEPTABLE and FS-EXCESS individuals, the main dietary changes in optimized diets were significant increases in fresh fruits, starchy foods, water, hot beverages and plain yogurts; and significant decreases in mixed dishes/sandwiches, meat/eggs/fish and cheese. For FS-EXCESS individuals only, the optimization process significantly increased vegetables and significantly decreased sugar-sweetened beverages, sweet products and fruit juices. The diets of French adults with excessive intakes of FS are of lower nutritional quality, but can be optimized via specific dietary changes.

Dietary Assessment in the MetaCardis Study: Development and Relative Validity of an Online Food Frequency Questionnaire.

BACKGROUND: The European study MetaCardis aims to investigate the role of the gut microbiota in health and cardiometabolic diseases in France, Germany, and Denmark. To evaluate long-term diet-disease relationships, a food frequency questionnaire (FFQ) was found to be the most relevant dietary assessment method for the MetaCardis study. OBJECTIVE: The objectives of this study were to describe the development of three semiquantitative online FFQs used in the MetaCardis study-one FFQ per country-and to assess the relative validity of the French MetaCardis FFQ. DESIGN: The layout and format of the MetaCardis FFQ was based on the European Prospective Investigation of Cancer (EPIC)-Norfolk FFQ and the content was based on relevant European FFQs. Portion size and nutrient composition were derived from national food consumption surveys and food composition databases. To assess the validity of the French MetaCardis FFQ, a cross-sectional study design was utilized. PARTICIPANTS/SETTING: The validation study included 324 adults recruited between September 2013 and June 2015 from different hospitals in Paris, France. MAIN OUTCOME MEASURES: Food intakes were measured with both the French MetaCardis FFQ and 3 consecutive self-administered web-based 24-hour dietary recalls (DRs). STATISTICAL ANALYSES PERFORMED: Several measures of validity of the French MetaCardis FFQ were evaluated: estimations of food groups, energy, and nutrient intakes from the DRs and the FFQ, Spearman and Pearson correlations, cross-classification, and Bland-Altman analyses. RESULTS: The French MetaCardis FFQ tended to report higher food, energy, and nutrient intakes compared with the DRs. Mean correlation coefficient was 0.429 for food, 0.460 for energy, 0.544 for macronutrients, 0.640 for alcohol, and 0.503 for micronutrient intakes. Almost half of participants (44.4%) were correctly classified within tertiles of consumption, whereas 12.9% were misclassified in the opposite tertile. Performance of the FFQ was relatively similar after stratification by sex. CONCLUSIONS: The French MetaCardis FFQ was found to have an acceptable level of validity and may be a useful instrument to rank individuals based on their food and nutrient intakes.

Theoretical Impact of Replacing Whole Cow's Milk by Young-Child Formula on Nutrient Intakes of UK Young Children: Results of a Simulation Study.

BACKGROUND: Research into the role of young-child formulae (YCF) in a child's diet is limited and there is no consensual recommendation on its use. We evaluated the theoretical nutritional impact of replacing the existing practice of consuming cow's milk by YCF. METHODS: From the UK Diet and Nutrition Survey of Infants and Young Children, whole cow's milk consumers, aged 12-18 months (n = 591) were selected for simulation scenarios. In Scenario 1, we tested the replacement of all whole cow's milk (434 ± 187 ml/day) by a matching volume of YCF, and in Scenario 2, all whole cow's milk was replaced by the on-pack recommended daily intake of 300 ml. Nutrient intakes before and after simulation scenarios were compared and evaluated against nutrient recommendations. RESULTS: Intakes of protein and saturated fatty acids were significantly decreased, whereas essential fatty acid intakes were increased. The prevalence of nutrient inadequacy before simulation was 95.2% for vitamin D and 53.8% for iron. After simulation, inadequacy decreased to 4.9% (Scenario 1) and 0% (Scenario 2) for vitamin D and to 2.7% (Scenario 1) and 1.1% (Scenario 2) for iron. CONCLUSIONS: Replacement of habitual cow's milk intake by a matching volume or 300 ml of YCF may lead to nutritional intakes more in line with recommendations in young children.

Portal glucose influences the sensory, cortical and reward systems in rats.

The detection of glucose in the hepatoportal area is a simple but crucial peripheral cue initiating a nervous signal that ultimately leads to a wide array of metabolic and behavioural responses, such as decreased food intake, tighter control of glucose homeostasis, or appearance of food preference. This signal has been suggested to mediate the effects of high-protein diets, as opposed to high-fat/high-sucrose diets. Nevertheless, the central targets of the signal originating from the hepatoportal area remain largely undocumented. Using immunohistochemistry on the brain of male rats, we show here that portal glucose increases c-Fos expression in the brainstem, in the hypothalamus (in particular in neurons expressing pro-opiomelanocortin) and also in olfactory and other limbic and cortical areas, including those functionally implicated in reward (Experiment 1). In similar postabsorptive conditions, a high-protein diet induced similar effects in the hypothalamus and the granular cells of the main olfactory bulb, whereas the high-fat/high-sucrose diet actually reduced the basal expression of c-Fos in cortical layers. Both diets also decreased the number of neurons expressing c-Fos in the amygdala and gustatory areas (Experiment 2). Altogether, these findings suggest that the peripheral signal primed by portal glucose sensing may influence behavioural adaptation such as food preference via a network including the olfactory pathway, central amygdala, nucleus accumbens and orbitofrontal cortex, in addition to satiety and metabolic effects primarily implicating the hypothalamic response.

Mu-opioid receptors and dietary protein stimulate a gut-brain neural circuitry limiting food intake.

Intestinal gluconeogenesis is involved in the control of food intake. We show that mu-opioid receptors (MORs) present in nerves in the portal vein walls respond to peptides to regulate a gut-brain neural circuit that controls intestinal gluconeogenesis and satiety. In vitro, peptides and protein digests behave as MOR antagonists in competition experiments. In vivo, they stimulate MOR-dependent induction of intestinal gluconeogenesis via activation of brain areas receiving inputs from gastrointestinal ascending nerves. MOR-knockout mice do not carry out intestinal gluconeogenesis in response to peptides and are insensitive to the satiety effect induced by protein-enriched diets. Portal infusions of MOR modulators have no effect on food intake in mice deficient for intestinal gluconeogenesis. Thus, the regulation of portal MORs by peptides triggering signals to and from the brain to induce intestinal gluconeogenesis are links in the satiety phenomenon associated with alimentary protein assimilation.

The role of sodium-coupled glucose co-transporter 3 in the satiety effect of portal glucose sensing.

Portal vein glucose sensors detect variations in glycemia to induce a nervous signal that influences food intake and glucose homeostasis. Previous experiments using high infusions of glucose suggested a metabolic sensing involving glucose transporter 2 (GLUT2). Here we evaluated the afferent route for the signal and candidate molecules for detecting low glucose fluxes. Common hepatic branch vagotomy did not abolish the anorectic effect of portal glucose, indicating dorsal transmission. GLUT2-null mice reduced their food intake in response to portal glucose signal initiated by protein-enriched diet. A similar response of Trpm5-null mice and portal infusions of sweeteners also excluded sugar taste receptors. Conversely, infusions of alpha-methylglucose, but not 3-O-methylglucose, decreased food intake, while phlorizin prevented the effect of glucose. This suggested sensing through SGLT3, which was expressed in the portal area. From these results we propose a finely tuned dual mechanism for portal glucose sensing that responds to different physiological conditions.