Dataset about the Life Cycle Assessment of new fermented food products mixing cow milk and pea protein sources.

In recent years, the food industry has expended considerable effort to design novel products that replace animal proteins with legumes; however, the actual environmental benefits of such products are often not quantified. Here, we performed Life Cycle Assessments (LCA) to evaluate the environmental performance of four new fermented food products based on different mixtures of animal (cow milk) and plant (pea) protein sources (100% Pea, 75% Pea-25% Milk, 50% Pea-50% Milk, 25% Pea-75% Milk). The system perimeter encompassed all stages from agricultural production of the ingredients to the creation of the final ready-to-eat products. Impacts were calculated for all environmental indicators included in the EF 3.0 Method in SimaPro software based on a functional unit of 1 kg of ready-to-eat product. Life cycle inventories included all of the flows analyzed by the LCA (raw materials, energy, water, cleaning products, packaging, transport, waste). Foreground data were acquired directly on the manufacturing site; background data were taken from the Ecoinvent 3.6 database. The dataset contains details on the products, processes, equipment, and infrastructure considered; mass and energy flows; Life Cycle Inventories (LCI); and Life Cycle Impact Assessment (LCIA). These data improve our understanding of the environmental impact of plant-based alternatives to dairy products, which is currently poorly documented.

Data on the Life Cycle Assessment of pizzas cooked and consumed at home taking into account the variability of consumer practices.

Human food consumption is responsible for significant environmental impacts, which in recent years have been the focus of an increasing amount of research. One of the major results of these efforts has been an appreciation for the ways in which impacts can differ among products. To date, though, relatively little is known about possible differences in the environmental performance of a single food product that is made or produced in different contexts. Furthermore, the influence of consumer practices, such as cooking time or cleaning method, has not yet been investigated. The goals of the study were therefore (i) to compare the environmental impacts of a single food product-in this case, pizza-that is produced in different contexts (industrial, homemade, and assembled at home) and (ii) to investigate the influence of real-world consumer practices on these impacts. Two study models were used: a ham-and-cheese pizza and a mixed-cheese pizza. The functional units (FU) examined were one pizza and 1 kg of ready-to-eat pizza. The system boundaries extended from the agricultural production of ingredients to the consumption of the pizza at home. All inventory data related to the steps occurring before purchase (including storage at the supermarket) came from databases or the literature, while inventory data related to the steps occurring after the sale were obtained from questionnaires answered by 69 consumers who prepared and consumed the six pizza preparation pathways (two recipes multiply three methods of preparation) at home. Background data were selected in the AGRIBALYSE 3.0 and Ecoinvent 3.6 databases. The environmental impacts of the six pizza preparation pathways were calculated by Life Cycle Assessment (LCA) using the characterization method "EF 3.0 Method (adapted) V1.00 / EF 3.0 normalization and weighting set" in SimaPro software. To compare the environmental impacts of the six pizza preparation pathways, 69 LCAs were performed for each; to compensate for missing data from incomplete questionnaires, we performed random draws from the available data to generate the life cycle inventory for each assessment. The data obtained in this study can be used to make recommendations to consumers regarding more environmentally friendly food choices and practices.

How Different Are Industrial, Artisanal and Homemade Soft Breads?

Soft bread has a significant relevance in modern diets, and its nutritional impact on human health can be substantial. Within this product category, there is an extensive range of ingredients, formulations, and processing methods, which all contribute to the vast diversity found in the final products. This work compared the impact of three different processing methods (industrial, artisanal, and homemade preparation) on the technological (formulation and processing, as they are interconnected in real-life conditions), nutritional, and physicochemical properties of soft bread. In total, 24 types of soft bread were analyzed: 10 industrial, 6 artisanal, and 8 homemade. Although production diagrams were similar among the three methods, industrial recipes contained on average more ingredients and more additives. Industrial bread was lower in saturated fat compared to the other two groups, but contained more sugar than homemade bread. The physical properties of all loaves were comparable, with the exception of higher crumb elasticity in industrial bread compared to homemade. An analysis of volatile molecules revealed more lipid oxidation markers in industrial bread, more fermentation markers in artisanal bread, and fewer markers of Maillard reactions in homemade bread. Chemical reactions during processing seem to be the principal criterion making possible to discriminate the different processing methods. These results offer a quantitative assessment of the differences within a single product category, reflecting the real-world choices for consumers.

Modulation of Metabolome and Overall Perception of Pea Protein-Based Gels Fermented with Various Synthetic Microbial Consortia.

Moving to a more sustainable food system requires increasing the proportion of plant protein in our diet. Fermentation of plant product could thus be used to develop innovative and tasty food products. We investigated the impact of fermentation by synthetic microbial consortia (SMC) on the perception of pea protein-based gels, giving possible keys to better understand the origin of sensory perception (e.g., beany, bitter). Two types of pea gels, containing (i) 100% pea proteins and (ii) 50% pea proteins/50% milk proteins, were fermented with three different SMC. Major species developing in both types of gels were Geotrichum candidum, Lactococcus lactis, and Lactobacillus rhamnosus. In pea gels, sensory analyses revealed that bitterness increased after fermentation, which could be due to hydrophobic amino acids resulting from protein hydrolysis, but also decreased pea note intensity in pea gels. In mixed gels, pea perception was similar whatever the SMC, whereas cheesy perception increased. Olfactometry experiments revealed that some specific "green" aroma compounds, responsible for green off-note, were suppressed/reduced by fermentation. The data presented investigated to which extent the design of SMC, together with gels composition (pea gels versus mixed gels), could modulate sensorial perception and drive consumer acceptability.

Life cycle inventory and assessment data for quantifying the environmental impacts of a wide range of food products belonging to the same food category: A case study of 80 pizzas representatives of the French retail market.

Most of the time, Life Cycle Assessments (LCA) of food products are performed only on one representative of a food category. This doesn't allow us to understand the possible variations of environmental impacts within a food product category and the responsible factors for these variations. For this reason, LCAs were conducted for 80 different industrial pizzas representative of the French retail market. The LCAs were performed using the "EF 3.0 Method (adapted) V1.00/EF 3.0 normalization and weighting set" on SimaPro software. Most of the data used were taken from the AGRIBALYSE 3.0 and Ecoinvent 3.6 databases. The system perimeter goes from the production of the ingredients to the pizza consumption. The functional unit used was 1 kg of ready-to-eat pizza. Life cycle inventories were made to include the different flows in the LCA (materials, transport, energy, water, waste, etc.). The dataset contains details on products, life-cycle inventories (LCI) and LCIA results. These data can enrich the discussion on the need to study the environmental impacts of different products belonging to the same food category and not only one representative in order to avoid erroneous conclusions.

Ultra-processed foods: how functional is the NOVA system?

BACKGROUND: In the NOVA classification system, descriptive criteria are used to assign foods to one of four groups based on processing-related criteria. Although NOVA is widely used, its robustness and functionality remain largely unexplored. We determined whether this system leads to consistent food assignments by users. METHODS: French food and nutrition specialists completed an online survey in which they assigned foods to NOVA groups. The survey comprised two lists: one with 120 marketed food products with ingredient information and one with 111 generic food items without ingredient information. We quantified assignment consistency among evaluators using Fleiss' κ (range: 0-1, where 1 = 100% agreement). Hierarchical clustering on principal components identified clusters of foods with similar distributions of NOVA assignments. RESULTS: Fleiss' κ was 0.32 and 0.34 for the marketed foods (n = 159 evaluators) and generic foods (n = 177 evaluators), respectively. There were three clusters within the marketed foods: one contained 90 foods largely assigned to NOVA4 (91% of assignments), while the two others displayed greater assignment heterogeneity. There were four clusters within the generic foods: three clusters contained foods mostly assigned to a single NOVA group (69-79% of assignments), and the fourth cluster comprised 28 foods whose assignments were more evenly distributed across the four NOVA groups. CONCLUSIONS: Although assignments were more consistent for some foods than others, overall consistency among evaluators was low, even when ingredient information was available. These results suggest current NOVA criteria do not allow for robust and functional food assignments.

The contribution of gastric digestion of starch to the glycaemic index of breads with different composition or structure.

Breads of higher density exhibit lower glycaemic index (GI) both in vivo and in vitro, a phenomenon generally attributed to a slower intestinal starch digestion. The aim of this work was to gain a better understanding of the relationship between bread density, oral and gastric digestion, and GI. Three breads were studied: industrial-style and traditional-style French baguettes (similar composition, different densities), and whole-wheat baguette. In vitro GI predictions confirmed that, for an identical composition, higher bread density was associated with a lower GI. Subsequent oro-gastric digestions, using the dynamic system DIDGI®, showed extensive starch digestion at the gastric stage by salivary α-amylase, in line with recently published data. They further showed that higher bread density led to a lower hydrolysis rate. The concurrence of these results with those of in vivo studies, suggests a mediating role for gastric digestion in the relationship between bread density and GI, possibly via the repercussions on the starch proportion that remains to be hydrolysed in the small intestine. This study therefore adds to the scientific knowledge of the importance of salivary α-amylase to starch digestion, and draws special attention to the possible role of the gastric phase in determining the GI.

Identification and characterization of the main peptides in pea protein isolates using ultra high-performance liquid chromatography coupled with mass spectrometry and bioinformatics tools.

Pea protein isolates are a source of high-quality plant proteins. However, from a sensory perspective, they are usually described as having strong beany and bitter notes, which arise from a complex mixture of volatiles, phytochemicals, and peptides. The aim of this study was to identify the main peptides in isolates and examine their correlations with sensory perceptions. Thus, 28 solutions containing different mixtures of pea protein fractions were assessed. Any peptides present were identified and characterized using ultra high-performance liquid chromatography-mass spectrometry. There were a total of 3,005 unique peptides representing various protein families; 1,640 and 275 peptides were correlated with broth and bitter attributes, respectively. In particular, 14 peptides with short sequences (<8 residues) were correlated with bitterness. These results show how key peptides in isolates may cause sensory perceptions.

Impact of intra-category food substitutions on the risk of type 2 diabetes: a modelling study on the pizza category.

Advice on replacing unhealthy foods with healthier alternatives within the same food category may be more acceptable and might ease the transition towards a healthy diet. Here, we studied the potential impact of substitutions within the pizza category on the risk of type 2 diabetes (T2D). The study sample consisted of 2510 adults from the INCA2 French national survey. Based on their nutritional characteristics, the 353 pizzas marketed in France were grouped into 100 clusters that were used to run various scenarios of pizza substitutions, which were either isoenergetic (IE) or non-isoenergetic (NIE). We then used a model structurally similar to the Preventable Risk Integrated ModEl to assess the expected rate of change in risk of T2D. Pizzas characterised by a low energy, high vegetable content and whole grain dough were associated with a greater reduction in the risk of T2D. The rates of change in risk of T2D were markedly stronger in men and for NIE substitutions. When the rates of change were estimated in the subsample of pizza consumers, replacing the observed pizzas with the best pizza resulted in a T2D risk reduction of -6·7 % (-8·4 %; -4·9 %, IE) and -8·9 % (-11·2 %; -6·3 %, NIE), assuming that this is their usual diets. The greatest risk reduction induced by an IE substitution of the observed pizza with a mixed dish was similar to that observed with the best pizzas. Overall, this modelling study suggests that healthy swaps within a category can effectively supplement broader dietary changes towards a healthier diet.

Texture contrast: Ultrasonic characterization of stacked gels' deformation during compression on a biomimicking tongue.

When undergoing compression during oral processing, stacked gels display different mechanical properties that shape perceptions of texture contrasts (Santagiuliana et al., 2018). However, to date, characterizing the mechanical responses of individual gel layers has been impossible. In this study, an ultrasound (US) technique was developed, that allowed such deformation dynamics to be visualized in real time. Stacked gels were created using layers (height: 5 mm) of brittle agar and elastic gelatin in different combinations. In a series of experimental tests, different stacked gel combinations were placed on a rough, deformable artificial tongue model (ATM) made of polyvinyl alcohol; a texture analyzer was used to apply uniaxial force, and deformation was monitored by an US transducer (5 MHz) located under the ATM. From the obtained results, it was observed that the deformation of ATM surface during compression was in accordance with the force recorded by the texture analyzer, suggesting a collaborative response of different layers under compression. Moreover, US imaging revealed that differences in Young's modulus values between layers led to heterogeneous strain distributions, which were more pronounced for the agar layers. Biopolymer elasticity was also a key factor. Regardless of combination type, the gelatin layers never fractured; such was not the case for the agar layers, especially those with lower Young's modulus values. The results of this US study have thus paved the way for a better understanding of the mechanical deformation that occurs in heterogeneous foods, a phenomenon that has been difficult to examine because of the limitations of conventional techniques.

Using a mixture design and fraction-based formulation to better understand perceptions of plant-protein-based solutions.

The food industry is focused on developing plant-based foods that incorporate pea protein isolates. However, these ingredients are often described as having persistent beany, bitter, and astringent notes, which can decrease the desirability of the resulting foods. These perceptions are rooted in the complex composition of volatile and non-volatile compounds in foods. The aim of our study was to better understand how the volatile and non-volatile fractions of pea protein isolates influence the perception of pea-protein-based foods. To this end, a mixture design was used. First, we obtained three fractions (the pellet, permeate, and retentate) from two pea protein isolates, resulting in a total of six fractions. Second, we used various combinations of the six fractions to create a set of 46 pea-protein-based solutions via various processes (solubilization, centrifugation, filtration, and mixing). Each fraction was specifically representative of the following constituent groups: insoluble proteins (the pellet); soluble compounds, such as volatiles, peptides, and phenolics (the permeate); and soluble proteins interacting with volatiles (the retentate). Factor levels were chosen with two aims: to explore the widest possible range of combinations and to realistically represent protein concentrations so as to build optimal mixture models. Third, 17 trained panelists were asked to score the attributes of the solutions using sensory profiling. Model performance was assessed using analysis of variance; results were significant for 18/18 attributes, and there was no significant lack-of-fit for 17/18 attributes. It was also assessed using the results of trials conducted with six supplementary solutions. These results clarified the origin of the perceived beany, bitter, and astringent notes. Beaniness was mainly influenced by the retentate and permeate fractions and was strongly affected by hexanal levels. Bitterness was mainly influenced by the retentate fraction, whereas astringency was influenced by the retentate and pellet fractions. Additionally, perception of these latter two attributes was affected by caffeic acid levels. This study has increased understanding of the relationship between pea protein fractions and the undesirable sensory attributes of pea protein isolates. It has also revealed how fraction-based formulation could be used to reduce the beaniness, bitterness, and astringency of pea-protein-based foods.

True ileal amino acid digestibility and digestible indispensable amino acid scores (DIAASs) of plant-based protein foods.

Plant-based protein foods are increasingly common, but data on their nutritional protein quality are scarce. This study evaluated it for seitan (wheat-based food), tofu (soya-based food), soya milk, and a pea emulsion. The true ileal digestibility (TID) of their amino acids was determined in minipigs, to calculate the digestible indispensable amino acid score (DIAAS). The TID of the proteins was high and not significantly different between the foods tested: 97% for seitan, 95% for tofu, 92% for soya milk and 94% for pea emulsion. There were only minor differences in individual amino acid TIDs. DIAAS ranking was thus essentially driven by the amino acid composition of the food: soya-based food > pea emulsion > seitan. Nevertheless, the lower TID of sulphur-containing amino acids in tofu than in soya milk induced a significant decrease in DIAAS (from 117% to 97%), highlighting the importance of the matrix effect on nutritional protein quality.

"How to Select a Representative Product Set From Market Inventory?" A Multicriteria Approach as a Base for Future Reformulation of Cookies.

Consuming too much fat, sugar, and salt is associated with adverse health outcomes. Food reformulation is one possible strategy to enhance the food environment by improving the nutritional quality of commercial products. However, food reformulation faces many hindrances. One way to alleviate some of these hindrances is to embrace a multicriteria approach that is based on a market inventory. In this objective, additional sensory screening and water content analyses allow going beyond nutrition and composition information on the packaging. However, due to feasibility reasons for later in-depth analyses, it is necessary to work with several reduced and manageable products. To the best of the authors' knowledge, in the literature, there is no sample selection approach taking into account multiple criteria as a base for future food reformulation. The overall aim of this paper is to propose a method to select the best representative products from the market base, for future reformulation by going beyond nutrition and composition information on the packaging. This approach considered therefore nutrition, composition, economic, water content, and sensory information with the example of the cookies market. The first step is the creation of an extensive cookie database including sensory and water content information. In total 178 cookies among the French market were identified, then focus was placed on 62 chocolate chip cookies only. Sensory screening and water content analyses of all 62 products were conducted. The second step is to make an informed subset selection, thanks to a cluster analysis based on 11 nutrition, composition, and water content variables. A representative subset of 18 cookies could be derived from the obtained clusters. The representativity was evaluated with statistical uni- and multivariate analyses. Results showed a broad variety of chocolate chips cookies with a large nutritional, compositional, water content, and sensory differences. These results highlighted the first paths for future reformulation in this product category and showed the importance to include physical product information beyond the information on the packaging. This complete database on the selected cookies constituted a solid base for identifying future reformulation levers, in order to improve the nutritional quality and health.

Food-dependent set-up of the DiDGI® dynamic in vitro system: Correlation with the porcine model for protein digestion of soya-based food.

The present study compared in vivo protein digestion in a miniature pig model with the dynamic in vitro system DiDGI®, using three digestive compartments (stomach, duodenum, and jejunum + ileum). Two soya-based meals-commercial soya milk and tofu-were studied, each with the same macronutrient content but different macrostructures. Our aim was to first deduce from the in vivo experiments in pigs key digestive parameters such as gastric pH, stomach emptying kinetics, and intestinal transit time, in order to design a relevant set-up for the dynamic in vitro system. Then, we compared digestive samples collected at fixed sampling times from both in vivo and in vitro models regarding different values related to proteolysis. We observed similar evolutions of gastric peptide distribution and duodenal proteolysis between models. Overall, apparent ileal digestibility of nitrogen was similar in vitro and in vivo and the differences between the two meals were conserved between models.

Using Multiple Sensory Profiling Methods to Gain Insight into Temporal Perceptions of Pea Protein-Based Formulated Foods.

The food industry is focused on creating plant-based foods that incorporate pea protein isolates. However, pea protein isolates are often described as having persistent beany, bitter, and astringent notes that can decrease the desirability of the resulting foods and make static sensory profiling difficult. To obtain more realistic descriptions of the sensory experiences associated with this category of products, researchers should consider using temporal methods and multi-intake methods, which allow consumers to evaluate whole food portions. This study aimed to understand better how product composition affected the sensory perception of pea protein-based beverages using three different sensory profiling methods. Particular focus was placed on beany, bitter, and astringent notes. Twelve pea protein-based beverages were formulated; they varied in pea protein type (pellet vs. isolate) and their content of gellan gum, salt, sunflower oil, sugar, and soy lecithin. They were evaluated by 16 trained panelists using three sensory profiling methods: static block profiling, mono-intake temporal dominance of sensations (TDS) profiling, and multi-intake TDS profiling. The static block and mono-intake TDS profiling methods yielded complementary results about the impact of beverage composition on attribute perceptions. Static block profiling revealed that beaniness was mainly affected by gellan gum and oil content and that bitterness and astringency were mainly affected by protein type and gellan gum content. Mono-intake TDS profiling highlighted the dynamics of beaniness and the strong persistence of astringency, and its results suggested that higher gellan gum and salt contents could limit this persistence. Multi-intake TDS profiling found that, throughout the consumption of a full product portion, beaniness and bitterness decreased, indicating an adaptation effect, while fattiness increased, indicating a build-up effect. This study has increased the understanding of how pea protein-based beverages are perceived under conditions that more closely resemble those associated with real-life consumption. It has also revealed how product formulation can reduce bitterness and astringency.

Hydrolysis of plant proteins at the molecular and supra-molecular scales during in vitro digestion.

The digestion of plant protein is highly dependent on multiple factors, with two of the most important being the protein source and the food matrix. The present study investigated the effects of these two factors on the digestion of seitan (a wheat-based food), tofu, soya juice, and a homemade emulsion of soy oil and water that was stabilised with pea protein. The four plant matrices and their respective protein isolates/concentrates (wheat gluten, soya protein, pea protein) were subjected to in vitro static digestion following the INFOGEST consensus protocol. We monitored the release of α-amino groups during digestion. We found that food matrix had a strong influence on protein digestion: soya juice was more hydrolysed than fresh tofu (51.1% versus 33.1%; P = 0.0087), but fresh tofu was more hydrolysed than soya protein isolate (33.1% versus 17.9%; P < 0.0001). Likewise, the pea-protein emulsion was better hydrolysed than the pea-protein isolate (P = 0.0033). Differences were also detected between the two solid foods investigated here: a higher degree of hydrolysis was found for tofu compared to seitan (33.1% versus 11.8%), which was perhaps a function of the presence of numerous dense protein aggregates in the latter but not the former. Furthermore, freeze-drying more than doubled the final degree of hydrolysis of seitan (P < 0.0001), but had no effect on tofu (P = 1.0000). Confocal microscopy revealed that protein networks in freeze-dried seitan were strongly altered with respect to the fresh product; instead, protein networks in freeze-dried and fresh tofu were largely similar. Finally, we found that the protease:protein ratio had a strong effect on the kinetics of proteolysis: a 3.7-fold increase in the concentration of the soya protein isolate with respect to that of the soya juice decreased the final degree of hydrolysis from 50.3 to 17.9% (P = 0.0988).

Using ultrasound to characterize the tongue-food interface: An in vitro study examining the impact of surface roughness and lubrication.

We measured the apparent reflection coefficient of a 1-MHz ultrasound compressional wave at the interface between rough and lubricated tongue mimicking surfaces and various food gels, composed of agar or gelatin. For the smoothest mimicking surface, when a lubricating layer was present, the apparent reflection coefficient was fairly similar for the different food gels (33.6% on average). The apparent reflection coefficient was significantly larger in the following situations: (i) tongue asperities were high and dense; (ii) lubrication levels were low; and (iii) gels were less rigid (range for the different gels-45.9-84.3%). The apparent reflection coefficient conveys the ability of food gels to mold themselves to surface asperities or to form a coupling film of liquid at the interface. This study demonstrates that ultrasound methods can and should be used to explore the physical phenomena that underlie the texture perceptions resulting from tongue-palate interactions.

Temporal changes in postprandial intragastric pH: Comparing measurement methods, food structure effects, and kinetic modelling.

Intragastric pH greatly affects food disintegration and the release of nutrients in the gut. Here, the behaviour of two liquid meals (soymilk, pea emulsion) and two solid meals (tofu, seitan) was tested in miniature pigs fitted with gastric cannula. For 5 h, intragastric pH was recorded using one of three methods: ex vivo measurements of chyme samples, in situ measurements using pH catheters, or in situ measurements using wireless pH capsules, both inserted through a pig's cannula. The pH values obtained with the two in situ methods were highly correlated. The liquid and solid foods yielded distinct pH kinetics. For the solids, pH simply decreased exponentially. For the liquids, pH increased rapidly and then plateaued for 2 h before dropping Food macrostructure and, to a lesser extent, food buffering capacity clearly had an impact on intragastric pH. We modelled changes in intragastric pH over time with food-dependent nonlinear equations.

INFOGEST static in vitro simulation of gastrointestinal food digestion.

Developing a mechanistic understanding of the impact of food structure and composition on human health has increasingly involved simulating digestion in the upper gastrointestinal tract. These simulations have used a wide range of different conditions that often have very little physiological relevance, and this impedes the meaningful comparison of results. The standardized protocol presented here is based on an international consensus developed by the COST INFOGEST network. The method is designed to be used with standard laboratory equipment and requires limited experience to encourage a wide range of researchers to adopt it. It is a static digestion method that uses constant ratios of meal to digestive fluids and a constant pH for each step of digestion. This makes the method simple to use but not suitable for simulating digestion kinetics. Using this method, food samples are subjected to sequential oral, gastric and intestinal digestion while parameters such as electrolytes, enzymes, bile, dilution, pH and time of digestion are based on available physiological data. This amended and improved digestion method (INFOGEST 2.0) avoids challenges associated with the original method, such as the inclusion of the oral phase and the use of gastric lipase. The method can be used to assess the endpoints resulting from digestion of foods by analyzing the digestion products (e.g., peptides/amino acids, fatty acids, simple sugars) and evaluating the release of micronutrients from the food matrix. The whole protocol can be completed in ~7 d, including ~5 d required for the determination of enzyme activities.

Design of microbial consortia for the fermentation of pea-protein-enriched emulsions.

In order to encourage Western populations to increase their consumption of vegetables, we suggest turning legumes into novel, healthy foods by applying an old, previously widespread method of food preservation: fermentation. In the present study, a total of 55 strains from different microbial species (isolated from cheese or plants) were investigated for their ability to: (i) grow on a emulsion containing 100% pea proteins and no carbohydrates or on a 50:50 pea:milk protein emulsion containing lactose, (ii) increase aroma quality and reduce sensory off-flavors; and (iii) compete against endogenous micro organisms. The presence of carbohydrates in the mixed pea:milk emulsion markedly influenced the fermentation by strongly reducing the pH through lactic fermentation, whereas the absence of carbohydrates in the pea emulsion promoted alkaline or neutral fermentation. Lactic acid bacteria assigned to Lactobacillus plantarum, Lactobacillus rhamnosus, Lactococcus lactis and Lactobacillus casei species grew well in both the pea and pea:milk emulsions. Most of the fungal strains tested (particularly those belonging to the Mucor and Geotrichum genera) were also able to grow on both emulsions. Although most Actinobacteria and Proteobacteria did not compete with endogenous microbiota (Bacillus), some species such as Hafnia alvei, Acinetobacter johnsonii and Glutamicibacter arilaitensis grew strongly and appeared to restrict the development of the endogenous microbiota when the pea emulsion was inoculated with a combination of three to nine strains. In the mixed emulsions, lactic fermentation inhibited Actinobacteria and Proteobacteria (e.g. Brevibacterium casei, Corynebacterium casei, Staphylococcus lentus) to the greatest extent but also inhibited Bacillus (e.g. Bacillus subtilis and Bacillus licheniformis). Overall, this procedure enabled us to select two microbial consortia able to colonize pea-based products and positively influence the release of volatile compounds by generating a roasted/grilled aroma for the 100% pea emulsion, and a fruity, lactic aroma for the 50:50 pea:milk emulsion. Moreover, the fermentation in the pea-based emulsions reduced the level of hexanal, which otherwise leads to an undesired green pea aroma. Our present results show how the assembly of multiple microbial cultures can help to develop an innovative food product.