Beef flavor vegetable hamburger patties with high moisture meat analogs (HMMA) with pulse proteins‐peas, lentils, and faba beans

Abstract Pulses have been an excellence source of foods due to their nutritional profile including high protein content. In addition, they have less concerns about allergens, gluten, and genetically modified organisms. In this study, high moisture meat analogs (HMMA) that contained commercial pea protein (55.4% protein), lentil protein (55.4% protein), or faba bean protein (61.5% protein) mixed with other constant ingredients (pea isolates, and wheat gluten and canola oil) were produced using a twin‐screw extruder (TX‐52) with an attached cooling chamber. For consumer sensory tests and texture profile analysis, vegetable hamburger patties with HMMA were produced with the addition of spices, binders, and so on. Trained panelists reported that HMMA with pulses had higher scores on bean‐like, sweet, and cohesiveness of mass compared to HMMA with soy that had higher soy, cardboardy, hardness, and springiness scores. Compared to the control, consumer panelists indicated that samples containing pulse proteins had no differences in consumers’ liking for the cooked appearance and overall flavor, but had a lower overall texture liking score, and samples with faba bean proteins (FP) had a lower overall liking score. Cooked patties containing pulse proteins were redder and had more cooking yield. Patties with FP required less cooking time. Therefore, vegetable patties with pulse proteins are competitive with soy‐based samples.


| INTRODUC TI ON
High moisture meat analogs (HMMA) are protein products produced by an extrusion process with the addition of moisture (40% to 80%) during the process to prevent expansion of the product in a cooling die attached to the end of the extruder. Unlike low moisture extruded protein products, HMMAs have well defined fiber formations, resemble chicken or turkey breast meat, and therefore have an enhanced visual appearance and taste sensation s (Malav et al., 2015;Sadler, 2004). These meat analogs are also called meat substitutes, mock meat, faux meat, or imitation meat (Sadler, 2004).
The key ingredients used during the preparation of meat analogs are soy protein, mushrooms, wheat gluten, egg albumin, carbohydrates, gum, and flavoring and other miscellaneous compounds such as fiber, caseinate, or carrageenan, as needed (Kumar et al., 2017).
Pulses are the dry edible seeds of plants in the legume family including field peas, dry beans, lentils, chickpeas, and faba beans (Tyler et al., 2017).
Pulses have a high protein content (about 20%-40%) and are abundant in dietary fiber, resistant starch, vitamins, and minerals (Sozer et al., 2017). In addition, the Frost and Sullivan Analysis found that pulses are considered nonallergenic, non-GMO, and appeal to vegans (Crane, 2015). Therefore, pulses can be excellent alternative meat sources.
The objectives of this study were to understand the trained panelist's perceptions of HMMA, the consumers' perceptions of vegetable hamburger patties with HMMA formulated with the addition of spices, binders, and so on, and the relationships between the trained panelists' tests and the consumers' tests. The hypothesis was that the qualities of the HMMA used in the trained panelists' tests and the consumers' tests would not be significantly different from the control (soy), and both tests would have a strong relationship in evaluating the quality of the products. Therefore, this study will contribute to developing a vegetable patty as a meat substitute, for which the consumer demand is rising steadily worldwide.

| Sample preparations
The HMMA samples produced in the research (Kim, Riaz, Awika, & Teferra, 2021) were used in this research. The HMMA samples are identified as C1 with soy concentrate (control), T1 with pea protein, T2 with lentil protein, and T3 with faba bean protein. The recipe, C1, containing soy concentrate and soy isolates and constant ingredients (6% canola oil and 15% wheat gluten) served as a control. There recipes (T1, T2, and T3) with pea isolates and pulse protein were mixed with the constant ingredients (6% canola oil and 15% wheat gluten) for the HMMA. Frozen HMMA (C1, T1, T2, and T3) were thawed in the refrigeration for 24 hrs.
Samples for each treatment were randomly selected. Before being served to trained panelists, they were boiled for 2 mins and stored in an oven at 80ºC covered with aluminum foil. The samples in the oven were cut into 2 cm cubes and placed in randomized plastic cups to be served.

| Sensory evaluation with trained panelists
The treated samples were evaluated by nine trained panelists from Texas A&M University who have been trained to evaluate beef flavor descriptive attributes. They were also trained for 3 days to help them become familiar and understand the attributes of a vegetable patty.

Panelist training and testing was approved by the Institutional Review
Board for the Protection of Human Subjects in Research (IRB) protocol IRB2017-0362 M. Each trained panelist had a packet including the attributes of the food, double-distilled deionized water, sparkling water, and saltless saltines. On the first day, panelists learned about basic tastes, cardboardy, grainy, musty-earth/hummus, malt-like, hay-like, buttery, and heated oil. On the second day, the panelists learned about greens, lentils, vegetable IDs, celery, carrots, roots, starches, faba beans, peas, and soy. Next day, the panelists learned texture including cohesiveness, hardness, springiness, particle size, and slipperiness. On the last day, the training on the third day will be repeated to help in understanding.
At the end of each training day, each sample in a randomized plastic cube was given to the trained panelists to determine the appropriate lexicons applicable to describe the characteristics of the samples.
Panelists received a warm-up sample to calibrate each sensory day, and the warm-up was individually evaluated by each panelist and discussed. Panelists came to consensus for all attributes prior to testing. Each sample was served in a plastic cup marked with a random three-digit code. Double-distilled deionized water was prepared as a mouth cleanser between samples. Each panelist was given a tablet (iPad Air 1, Apple Inc., Cupertino, CA) to record their individual data using an electronic spreadsheet (Microsoft Excel, One Drive, Microsoft Corporation, Redmond, WA), and samples were evaluated independently. Four random samples over the course of a two-hour session were evaluated each sensory day.  Frozen HMMA samples were stored for 24 hrs in the refrigerator at 4ºC. The HMMA samples were boiled for 2 mins as was done in the trained panelists test, resized using a size reducer (Comitrol 3,500, Urschel Laboratories, Inc., Valparaiso, IN) with a 9-mm blade, and stored in the refrigerator for further experiments. Table 1 shows the recipe used to produce a vegetable hamburger patty for the consumer test. The spice mixture including minced onion, dried egg white, beef flavor, carrageenan, and flavor enhancers was prepared in the mixer, and lactic acid and citric acid were mixed to form a homogenous dry ingredient mixture. Shortening was chilled, ground through a 3.18 mm grinder plate and frozen. The frozen strings were broken into fat pellets and stored in the freezer until they were added to the mixture to make a vegetable patty.

| Making patties
For the consumer test, the HMMA (0-2ºC) chilled with ice in a container were mixed in a Hobart mixer (Hobart mixer, Model N50, Canada) at 20 rpm controlled by a rheostat (Type 3PN1010, Staco Energy Products Co., Dayton, OH) as water (0-2ºC) chilled with ice in a container was added slowly. During mixing, methylcellulose was sprinkled in slowly for long enough to ensure a uniform methylcellulose coating of the HMMA particles. The dry ingredient mixture was added slowly to ensure uniform distribution of all ingredients followed by the addition of the shortening until the mix was uniform.
Patties for each treatment were formed with a patty maker (Supermodel 54 Food Portioning Machine, Hollymatic Corporation, Countryside, IL) with a 2.54 cm plate. The patties were placed with patty paper on top and bottom in a single layer on trays, placed in a −40℃ freezer, crust frozen for 20 mins, vacuum packaged, and stored in the −40℃ freezer until the sensory test.

| Cooking protocols
Approximately 24 hrs prior to testing, samples were removed from the freezer and placed on racks in a single layer to thaw in a cooler (4℃). One hour before testing, patties were organized by cooking order on the trays. Their vacuum packaged bags and patty paper were removed, and patty trays were covered with plastic wrap and held in the cooler until time to cook. Prior to cooking, five temperature readings of the surface of the grill were checked using an infrared temperature reader (MS6530H Infrared Thermometer, Commercial Electric Products Corporation, Cleveland, OH) with a target temperature of 162°C. As seen in APPENDIX A, the weights and temperatures of the raw samples and the time they were put on the grill were recorded, along with the end temperature, time they were taken off the grill, and final cooked weights.
Samples were cooked on a commercial flat-top grill to an end temperature of 71°C, with a flip temperature at 27°C. Internal temperatures were monitored using thermocouple probes (Model SCPSS-040 U-6, Type T, 0.040 Sheath Diameter, 15.24 cm length Ungrounded Junction Thermocouple, Omega Engineering, Stamford, CT) by inserting them into the geometric center of each vegetable patty periodically during cooking.
The temperature was displayed using a thermometer (Omega HH501BT Type T, Omega Engineering, Stanford, CT). Each sample was prepared on a clear plastic plate (clear 15.88 cm plastic plates premium quality, Members Mark, Sam's Club, Bentonville, AR) marked with a random three-digit code. Each patty was cut into four equal pieces, and a quarter of a patty was served to each consumer. Consumers were given a new transparent plastic fork and transparent plastic knife to use for each sample as well.
After patties were taken off the grill and weighed, they were wrapped in foil and placed in a holding oven (Model 750-TH-II, Alto-Shaam, Menomonee Falls, WI) for no longer than 20 mins, until served.

| Sensory test
In advance, 80 consumers were recruited by emails and advertisements. They provided demographic information and signed a consent form through a survey website (www.tamuag.az1.qualt rics.com).
Depending on their answers of available time for the test, they were assigned to one of four different sessions (20 consumer panelists each) for a 1 hr interval. In each session, they were divided into five groups since four wedges were cut from each patty. Four consumer panelists in each randomized group had the same treatment in the same order (APPENDIX A). Before the test, a consent form from each panelist was collected again. When they were seated in the booth under a red light, they were given a packet containing testing procedures, palate cleansers of distilled water and saltless saltine crackers, a demographic ballot, and five individual sample ballots. Demographic information from each panelist including gender, age, ethnicity, household income, household population, employment level, protein sources consumed, and location consumed, frequency of protein consumption, preferred cooking method for ground beef, degree of doneness desired for ground beef, type of ground beef typically purchased, desired fat percentage of ground beef, and types of cuisines consumed was received (APPENDIX B). Cooked appearance, overall appearance, overall flavor, and overall texture were evaluated by the panelists on each sample ballot utilizing a 9-point hedonic scale. Open-ended questions, "Please write any words that describe what you LIKE about this meat patty" and "Please write any words that describe what you DISLIKE about this meat patty," were included on each ballot (APPENDIX C).

| Cooking yield and cooking time
Cooking yield was calculated by using Equation.1.
Cooking time of each patty in minutes was measured.

| Color measurement
Frozen vegetable hamburger patties were thawed for 24 hrs in a cooler (4°C) and remained at room temperature about 20 min after their (1) Cooking yield ( % ) = Cooked patty weight ( g ) Raw patty weight (g) × 100 TA B L E 1 Recipe for producing a vegetable hamburger patty with a high moisture meat analog  Bourne (1978). All analyses were performed with five samples for each treatment.

| Trained Descriptive Flavor and Texture Perception
The trained panelists' perceptions are reported in Table 2. Protein sources in vegetable patties containing HMMA did not significantly affect flavor attributes for starchy (p = .58), grainy (p = .87), green (p = .65), and buttery (p = .41). Trained panelists could not perceive green and buttery in all samples.
Compared to the control, C1, the samples (T1, T2, and T3) containing pulse proteins were scored higher for flavor attributes that were bean-like, salty, sweet, umami, musty-earthy, and malt-like and heated oil indicated extremely small values (p < .0001). In contrast, these samples were lower for soy (p = .01) and cardboardy (p < .0001). However, these pulses did not have significantly different flavor attributes from each other except the sweetness attribute for T1 was higher than T3.
Pulse proteins in vegetable patties containing HMMA significantly affected texture attributes of cohesiveness of mass, hardness, and springiness resulting in extremely small values (p < .0001) compared to C1. The samples including pules proteins (T1, T2, and T3) were significantly higher for cohesiveness of mass and lower for hardness and springiness compared to the control. T1 was higher for hardness compared to T3.
Pulse proteins in the samples did not affect the flavor including starchy, grain, green, and buttery attributes compared to the control.

| Consumer Demographics
Demographic information for consumers (n = 80) participating in this study is reported in Consumers were asked to report how many times a week they consumed each protein source (data not presented). The majority of consumers reported consuming beef (steaks) 1 to 2 times per week (60.5%), followed by 0 times per week (26.3%) and 3 to 4 times per week (10.5%). For ground beef consumption, the majority of consumers reported eating it 1 to 2 times per week (66.2%) followed by 0 times per week (18.2%) and 3 to 4 times per week (15.6%). For pork consumption, consumers reported eating it 1 to 2 times per week (56.2%) followed by 0 times per week (38.4%). For lamb consumption, the majority of consumers reported 0 times per week (92.8%) followed by 1 to 2 times (7.2%).
For chicken consumption, the majority of consumers consumed chicken 3 to 4 times per week (50.0%), followed by 1 to 2 times per week (23.1%) and 5 to 6 times per week (21.8%). For fish consumption, the majority of consumers reported eating fish 1 to 2 time per week (64.5%) followed by zero times per week (27.6%).
When asked for preferences on degree of doneness, consumers reported fairly evenly distributed between medium rare to well (data not presented). They reported medium (26.3%), medium well (22.5%), and both medium rare (20%) and well done (20%). Few consumers preferred the extremes with only 2.5% reporting rare and 8.8% for very well done.

| Consumer perception of vegetable patties containing HMMA
Consumer perception scores are reported in Table 4. Protein sources in meat patties containing HMMA did not significantly affect the number of consumers who liked the cooked appearance (p = .89) and overall flavor (p = .24). However, C1 was more desirable for overall liking and overall texture than T3.
Word clouds were created using the comments from consumer TA B L E 2 Trained descriptive flavor and texture perception of vegetable patties containing HMMA with pulse proteins they liked or disliked vegetable patties containing HMMA. Figure 1 to 4 demonstrate the consumer's responses separated by vegetable patties containing HMMA with different protein sources (C1, T1, T2, and T3). The size of the word illustrates how often the consumers used the words. For C1, the most commonly used words for liking were texture, flavor, good, like, and taste ( Figure 1) and for disliking most commonly used words were texture, flavor, bad, little, dry, and bland. Flavor and texture were the most frequently used words for the like and dislike descriptors. For T1, the most frequently used words for liking were flavor, good, and texture ( Figure 2) and for disliking the most commonly used words were texture, taste, and flavor. As for the like descriptors, the most commonly used for T2 were flavor, good, and texture, while for dislike descriptors, the most commonly used word was texture. More positive words to describe the quality of the patties were used for T2 than negative words. As for like descriptors, the most commonly used words for T3 were flavor, texture, good, like, and the most frequently used word for disliking was texture. More positive words to describe the quality of the patties were used for T3 than were negative words. (Figure 3,4) For all vegetable patty samples, more descriptive words were used when the consumer panelists responded to describe liking points of the sample compared to dislike descriptors. Across all the words clouds, texture was most consistently used for describing whether or not a consumer liked a sample except for T1 for the like descriptors. Table 5 shows the color of raw and cooked vegetable patties with HMMA. The protein source did not affect the color of raw and cooked patties for lightness (L*, p = .09) and cooked patties for yellowness (p = .42) compared to C1. However, the protein source significantly affected the redness of raw patties (p = .0012) and yellowness (p = .0044) and redness in cooked patties (p = .05).

| Color of raw and cooked vegetable patties
Redness and yellowness were higher for these samples of raw patties containing pulse proteins compared to the control, and redness was similar for T2 to T1 and T3, but higher for T1 compared to T3. Yellowness was similar for these samples of raw patties to each other containing PLP. Redness was similar for these samples to each other.

CO N FLI C T O F I NTE R E S T
The authors, whose names are listed above, swear that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; or expert testimony or patentlicensing arrangements) or nonfinancial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.

S TU D I E S I N VO LV I N G H U M A N S U BJ EC TS
Human subjects for this study were approved by the Institutional

Review Board for the Protection of Human Subjects in Research
(IRB) protocol IRB2017-0362 M.

S TU D I E S I N VO LV I N G A N I M A L O R H U M A N S U BJ EC T S
Panelist training and testing was approved by the Institutional Review

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated during and/or analyzed during the current study are available in the Texas A&M University Libraries repository, a Means within a row and effect followed by the same letter are not significantly different (p >.05).
b RMSE = root mean square error.

TA B L E 6
Cooking yield, cooking time, and texture of cooked HMMA vegetable patties with PLP