a project being led by the World Wildlife Fund and Zoologic Society of London in which hundreds of academic institutions are tracking the total populations of more than 4,000 animal species selected to represent the diversity of animals worldwide. According to Brown, this project reports the ongoing collapse of selected populations—which include mammals, birds, reptiles, amphibians, and fish—with a 68 percent decrease in total population occurring between 1970 and 2016 (WWF, 2022). He added that recent reports have documented similar collapses in populations of insects, particularly flying insects, that may be even greater in magnitude. Brown described the decline in global biodiversity as having catastrophic effects on the viability of Earth’s biosphere.

The use of animals in the food system is driving this biodiversity collapse, Brown maintained, adding that exploitation, habitat destruction, and habitat degradation account for more than 80 percent of biodiversity decline (WWF, 2014). He explained that exploitation takes place primarily in the form of overfishing, while animal agriculture destroys and degrades habitats via land clearing and fencing. Brown described the scale of animal agriculture by comparing it with other animal life. He reported that the total biomass of farmed cattle outweighs that of all wild land mammals, birds, reptiles, and amphibians by 10-fold. Farmed pigs and farmed goats/sheep outweigh all wild land vertebrates by 50 percent and 25 percent, respectively. Farmed poultry outweighs all remaining wild birds by three-fold. Additionally the meat consumed by humans each year outweighs all living wild land vertebrates by a factor of more than six (WWF, 2014). Brown remarked that if humans met the current global demand for meat via hunting instead of farming, all land vertebrates would be eliminated within 2 months.

Brown went on to report that according to the International Livestock Research Institute, a group that advocates for the use of livestock to improve food security and reduce poverty, animal agriculture currently exploits 45 percent of Earth’s ice-free land surface for grazing or growing feed crops for livestock. This figure equates to more than 80 percent of humanity’s total land footprint, he added. In comparison, he said, the total land area of all cities worldwide occupies less than 1 percent of the ice-free surface of Earth, while all cities, suburbs, and infrastructure occupy less than 3 percent. Each day, he continued, the amount of land cleared to accommodate livestock—whether for grazing cattle or raising feed crops—averages twice the land area of San Francisco, and demand for meat drives more than 90 percent of the deforestation of the Amazon rainforest. Brown described the smoke rising from the Amazon as “secondhand smoke from steak.”

Brown highlighted the effects of cattle farming on biodiversity by presenting photos of Point Reyes National Seashore Park and a neighboring cattle farm. The park features wilderness and native coastal vegetation

typical of California. Brown remarked that the farm refers to itself as a “renewable, sustainable, organic, grass-fed beef farm,” yet its grazing lands are devoid of the abundant vegetation present in the park site at its border. He noted that the farm has vastly less biomass for storing carbon and less biodiversity in terms of plants and animals than the park. Brown referenced recent research finding that the total amount of biomass released into the atmosphere as carbon dioxide (CO₂) in the course of converting land to use for animal agriculture is the equivalent of 22 years’ worth of fossil fuel emissions at the 2018 rate (Hayek et al., 2021).

Environmental Implications of Plant-Based Proteins

According to Brown, eliminating the need for animals as a food technology would allow vast amounts of land to recover their original biomass and biodiversity, reducing CO₂ emissions on a massive scale and effectively reversing environmental damage. He and a colleague conducted research on the effect on atmospheric greenhouse gas levels and global warming of phasing out animal agriculture (Eisen and Brown, 2022). Using data from the Intergovernmental Panel on Climate Change, the National Oceanic and Atmospheric Administration, and the Food and Agriculture Organization (FAO) of the United Nations, they modeled the complete phaseout of animal agriculture over a period of 15 years.

Brown explained that, according to this model, native biomass and ecosystems could be restored on land previously used for grazing and growing livestock feed within 40 years. He added that, whereas the CO₂ emitted from fossil fuels cannot be turned back into coal, the CO₂ from animal agriculture can revert back to use by trees and plant biomass. In addition to CO₂, Brown noted that animal agriculture also contributes methane and nitrous oxide (major greenhouse gases); livestock emissions are responsible for about one-third of atmospheric methane and more than half of atmospheric nitrous oxide. Brown explained that because those gases spontaneously decay, ending their release would immediately result in negative emissions. In contrast, he said, the only method for achieving negative emissions for CO₂ at any scale is via photosynthesis and biomass recovery. Brown’s modeling indicates that neutrality of greenhouse gas emissions would begin 10 years after the end of animal agriculture. He maintained that eliminating the need for animal agriculture would generate negative emissions at a rate sufficient to negate ongoing fossil fuel emissions for a period of 30 years. However, he argued, both the magnitude and speed of such a shift in negative emissions are needed. He posited that phasing out animal agriculture over the next 15 years could offset more than half of total human contributions to global warming through the remainder of the 21st century.

animal-products industry in the global marketplace and provided consumers with options.

Brown surmised that, based on historical examples of technology adoption, the objective of replacing animals in the food system by 2035 is achievable. Moreover, he suggested that improved technology can rapidly replace even deeply entrenched technology. He illustrated this point by noting that sales of film cameras steadily increased throughout the 1970s, 1980s, and 1990s until the advent of the digital camera. Although the first digital camera cost around $1,000 and featured memory capacity limited to eight photos, the new digital technology quickly replaced film cameras. Within 8 years of the release of the first digital camera, Kodak went bankrupt. Thus, Brown asserted, change can happen quickly when driven by consumer demand and competition in the marketplace.

Benefits of Novel Alternative Proteins

Brown remarked that the technology platform created by Impossible Foods has delivered plant-based beef, pork, and chicken products whose taste mainstream consumers prefer over that of leading animal-based competitors. He described these novel products as having fewer calories; less total fat and saturated fat; and more protein, iron, and micronutrients than their animal-sourced counterparts. In addition, he reported, these alternative proteins are more sustainable according to independent life-cycle analyses. Brown added that in contrast to the traditional meat industry, plant-based technology and the products it enables continue to expand and improve. He noted that this technology is still in the early stages of discovery, with an entirely new platform and ingredient supply chain currently being built, both of which will continue undergoing an iterative process of improvement.

Brown acknowledged that existing plant-based products tend to be more expensive than their animal-based competitors. He stressed, however, that the plant-based food industry is being built from scratch, without existing infrastructure, production capability, and supply chains, whereas the infrastructure for the animal agriculture system has been in place for decades or centuries. Thus, the current economics of Impossible Foods are those of a growing startup company. Brown noted that the company is quickly moving toward scale, and the economics become inherently better at a larger scale, with lower prices for consumers and greater access to highly nutritious foods for people living in poverty worldwide. According to Brown, these economics at scale are related to the difference in resources consumed by novel alternative protein products compared with their animal-sourced counterparts. As an example, he remarked that the Impossible Foods beef product uses 4 percent of the land, 12 percent of

the water, and 8 percent of the fertilizer and agrochemicals needed to produce beef from cows. He added that fewer plant crops are required to produce plant-based meat than are needed to feed livestock, requiring less farm labor to raise crops and eliminating the farm labor needed to manage animals. Moreover, he said, manufacturing and packaging plant-based foods are far more labor efficient than disassembling cows to put into packages. Brown summarized by stating that every aspect of the economics of plant-based proteins is favorable compared with the economics of traditional meat once scale has been achieved.

Opportunities for Reinventing the Agricultural Supply Chain

Brown maintained that as technology continues to improve, the replacement of animals in the food system with plant-based technologies is inevitable and will create opportunities to reinvent the agricultural supply chain. To illustrate this point, he stated that much of the current agriculture system and many of the major stable crops are optimized for the use of livestock in producing meat; the crops grown for meat production are optimized for bulk yield of nutrients, polymerized amino acids, and starches to be fed to livestock. He characterized the current system—which would be rendered obsolete once replaced by a plant-based system—as inefficient and inherently destructive. Additionally, Brown suggested that consideration of the raw materials needed to convert proteins and fats from plant crops directly into human food altogether redefines the merits of and value proposition for crops and agriculture. He predicted, for example, that farmers and landowners will have the opportunity to increase profits through recovery of biomass. Given the 2030 carbon prices projected by The World Bank for biomass recovery on land currently used for animal agriculture, the development of this new agricultural supply chain stands to increase profits for farmers, he contended.

PLANT-BASED ALTERNATIVES AS TRANSITION AND MAINTENANCE FOODS

Mark Messina, Soy Nutrition Institute Global, discussed the global protein demand, the protein and nutritional content of various plant-based proteins, and the role of novel plant-based meat alternatives in transitioning to and maintaining a diet with a higher plant-to-animal ratio. The global population is projected to reach nearly 10 billion people by 2050, he observed, raising the issue of how best to meet the increasing calorie and protein demand over the next 30 years (PRB, 2020). He cited research indicating that meeting the demand for protein within environmental limits is

one of the greatest challenges for the global food system in the 21st century (Porritt and McCarthy, 2015).

In an effort to anticipate the worldwide global protein demand in 2050, Messina continued, researchers developed five potential scenarios based on a range of protein intake (Henchion et al., 2017). At the low end of the spectrum, he elaborated, if each person consumed 50 grams of protein per day, the result would be a worldwide protein demand 13 percent below its current level. At the high end of protein intake, if each person consumed 100 grams of protein per day, global protein demand would increase by 78 percent. According to Messina, evidence suggests that the latter scenario is more likely in the next 30 years because populations around the world currently derive about 16 percent of their calories from protein (Lieberman et al., 2020). He added that although predominant protein types differ dramatically among regions, the percentage of total calories derived from protein is relatively consistent. Messina suggested that this observation supports the protein leverage hypothesis proposed in 2005, which maintains that biological control mechanisms tightly regulate protein intake (Simpson and Raubenheimer, 2005). He surmised that populations would shift toward a plant-based diet in response to concerns about the environment, personal health, and animal welfare.

Establishing an Optimal Animal-to-Plant Protein Ratio

An optimal dietary ratio of animal-to-plant protein has not yet been established, observed Messina. The EAT-Lancet Commission Report suggests that most dietary protein should come from plants, while the Dutch Food-Based Dietary Guidelines recommend an animal-to-plant protein ratio of about 1 to 1—a recommendation that according to Messina aligns with the U.S. diet that was typical in 1909. The U.S. Department of Agriculture (USDA), he explained, used disappearance data to make crude estimates of protein intake (Gortner, 1975), and these estimates indicate that in 1909, approximately half of protein intake came from plants and half from animals. Over the past 100 years, Messina noted, the U.S. ratio changed dramatically in favor of animal protein, and a recent analysis found that in 2015, the animal-to-plant protein ratio was about 1.84 to 1 (Shan et al., 2019). Similarly, he added, most high-income countries around the world obtain about twice as much protein from animals as from plants. He pointed out, moreover, that 46 percent of the plant protein consumed in the United States comes from refined grains, which do not offer the nutritional benefits that other sources of plant protein do.

emulate the orosensory properties of meat; he cited research suggesting that the success of meat substitutes will be contingent on their resemblance to meat in both taste and texture. To support this point, he noted that household scanner data collected in the United States indicate that 86 percent of purchasers of novel plant-based meat alternatives in 2018–2020 also purchased beef (Neuhofer and Lusk, 2022).

Messina remarked that this new generation of alternatives are transition and maintenance foods in that they ease the transition from an animal-based diet to a diet featuring more plant protein and make it easier to maintain such a diet. He added that processed meat alternatives have long been considered transition foods and thus can be part of a strategy for supporting healthier eating patterns (Gastaldello et al., 2022). In his view, although cooked beans and tofu feature desirable nutritional attributes, transitioning from eating beef burgers to eating plant-based burgers is likely to be a more acceptable approach to incorporating plant protein into the diet. Messina acknowledged, however, that even this transition may be difficult because the preparation and consumption of meat dishes are linked to cultural traditions and norms as well as to collective and individual identities, such that meat has a central role in American society and in many societies throughout the world. In addition to aiding in this transition, Messina stressed the benefit of plant-based meat alternatives in making it easier to maintain a plant-based diet. He argued that this is an important consideration given the limited research suggesting that many vegetarians do not remain on a plant-based diet indefinitely.

Messina went on to remark that despite the favorable attributes of plant-based meat alternatives, these products have received pushback. He cited a paper published in 2021 in which the author contends that the current trend toward increased consumption of highly processed plant-based convenience foods poses concerns regarding public health and reduction of greenhouse gas emissions (Macdiarmid, 2022). Additionally, Messina noted that these products (namely plant-based meat and dairy alternatives, such as burgers, sausages, and cheese) are classified by the NOVA food classification system as ultraprocessed foods—foods that should be discouraged in the diet (Monteiro et al., 2019).

Messina explained that a meta-analysis of 43 studies found that intake of ultraprocessed foods was linked to a wide range of adverse health outcomes, including obesity, breast cancer, and metabolic syndrome (Lane et al., 2021). He noted, however, that these associations are based on observational not clinical trial data. Messina co-authored a paper reporting that the common criticisms of ultraprocessed foods do not apply to soy-based meat and dairy alternatives more so than to their animal-based counterparts, beef and cow’s milk, which are classified by NOVA as unprocessed or minimally processed—foods to be encouraged in the diet (Messina et al., 2022). The

specific concerns addressed in the paper include the hyperpalatability of ultraprocessed foods; their high energy density and high energy intake rate; and their tendency to have low satiety, snackability, and convenience, which are attributes that can lead to overconsumption and, potentially, obesity.

Messina also pointed out that numerous nutrient profiling systems used around the world have rated plant-based meat alternatives quite highly. For example, the Australian Health Star Rating System uses a five-point scale for foods, with 5 representing the healthiest score. Plant-based burgers received a Health Star rating of 3.9, whereas meat burgers received a score of 2.9 (Curtain and Grafenauer, 2019).

Comparative Nutritional Quality of Plant-Based Alternatives

Messina highlighted the relevance of comparing the nutritional content of plant-based meat alternatives with that of meat and of legumes. Animal products are nutrient dense and provide numerous micronutrients, he observed, thus playing a role in meeting human nutritional needs (Agarwal and Fulgoni, 2022). According to data from the National Health and Nutrition Examination Survey, for example, beef consumption contributes significantly to U.S. caloric and nutrient intake, including 21 percent of zinc, 20 percent of vitamin B12, and substantial amounts of other B vitamins and highly bioavailable iron. The authors of this analysis conclude that eliminating beef from dietary recommendations would require taking steps to replace the nutrients it provides. Messina noted that vegan diets can have nutrient shortfalls of zinc; vitamin B12; calcium; and, in some cases, B6, choline, iron, and vitamin A. He stated that these shortfalls are not likely to be present in a shift from a 2:1 to a 1:1 animal-to-plant protein intake ratio.

Messina noted that the Study With Appetizing Plantfood-Meat Eating Alternative Trial (SWAP-MEAT) found several beneficial effects and no adverse effects from the consumption of plant-based meats (Crimarco et al., 2020). In this crossover study, participants consumed plant-based meat alternatives or analogous animal products for 8 weeks, then alternated those for an additional 8 weeks. The meat-alternative products used in the study provided about 25 percent of total calories and 50 percent of total protein, and when on the plant-based meat-alternative diet, participants consumed approximately 2.5 servings of those products per day. Messina reported that at the end of the plant-based phase of the SWAP-MEAT trial, statistically significant reductions in trimethylamine N-oxide, low-density lipoprotein cholesterol, and body weight were observed. He emphasized that researchers found no adverse effects of these products. He added that limited research suggests that plant-based meat alternatives may have a positive effect on the gut microbiome, a finding that he believes warrants additional study (Toribio-Mateas et al., 2021).

$88 billion across fresh, frozen, and canned varieties. Plant-based foods and beverages make up the majority of plant-based products sold, she continued, with $9.3 billion in annual sales. Alternative plant-based meats—including both fresh and fully cooked products—total $971 million in sales and constitute 1.2 percent of total meat sales. Frey noted that alternative meat is currently at a price premium of about 38 percent above the cost of conventional meat. She reported further that alternative milk products total $2.6 billion in annual sales; plant-based foods in this category with the highest sales volume are alternative milk, creamers, and beverages. Fully cooked meat alternatives are the fourth highest–selling plant-based food category, Frey observed, and plant-based fully cooked, processed, and fresh meat alternatives have all seen sales growth of 39 percent or higher over the past 3 years, although sales of processed and fresh meat alternatives have declined over the past year. Frey commented that although the phenomenal growth in sales of alternative meats is now slowing, sales of alternative milk, cream, and beverages are continuing to increase at a consistent rate.

Consumer Preferences for Plant-Based Products

Frey next observed that NielsenIQ has studied consumer preferences and motivations for purchasing plant-based products, finding that the most commonly cited preference is for products that taste, look, and feel like their analogs. Other preferences expressed are for products that have few ingredients, are as natural as possible, are minimally processed, offer organic options, and are labeled as being sustainable or environmentally friendly. Frey reported that preferences for new and existing brands were nearly split in 2021, accounting for 45 percent and 55 percent, respectively, of growth in sales of plant-based protein. Market expansion and innovation are reflected in the brands available, she noted, as the NielsenIQ study identified 126 existing brands and 3,281 new brands.

Frey then highlighted research conducted on the types of plant-based protein experiencing sales growth. That research, she reported, showed that protein from bean, soy, and oat can be considered proven trends, accounting for the largest market shares of all plant-based proteins; peanut, almond, and textured vegetable protein are growing trends, with lower volume in sales than bean, soy, and oat but demonstrating substantial sales growth; and protein from flaxseed, cashew, and pea is a developing trend that accounts for the smallest volume of sales but also reflects growth.

Frey also described research conducted on the reasons consumers purchase meat alternatives on a regular or occasional basis. This research revealed that taste, overall health, and nutritional benefits are the top reasons for consuming meat alternatives, regardless of whether they are made with beans, vegetables, and grains or with isolated proteins and starches.

Frey noted that animal welfare was cited by 21 percent of consumers for both categories of meat alternatives. Consumers were also surveyed about what effect they view meat alternatives as having on the environment. Frey reported that approximately 40 percent of all shoppers surveyed perceived meat alternatives as having a positive environmental impact, whereas 20 percent perceived them as having no such impact; more than 20 percent indicated they were not sure, while 8 percent reported a perceived negative impact.

Frey also reported the results of research using a digital ethnography method, which entailed examining consumer conversations to capture underlying trends that might not be revealed in direct consumer research. That research showed, she said, that sustainable farming (i.e., farming methods that do not rely on industrial fertilizers and monocropping) is important to consumers and that sustainability concerns contribute to their desire to avoid animal products. However, she added, consumers also indicated concerns that a plant-based diet could be associated with unsustainable plant production methods.

According to Frey, the food and beverage industry appears to be responding to consumer environmental concerns through an increase in the environmental claims made about products, including sustainability production and packaging practices, as well as claims about social responsibility and animal welfare. She presented global data indicating that animal welfare claims are among the top-five drivers of purchase interest among consumers across the food and beverage, personal care, and household care categories. These claims include “animal welfare certified,” “cruelty free,” “free of animal by-products,” and “certified human raised and handled.” Frey added that the increasing popularity of a flexitarian diet (a diet that intersperses meat consumption with meat-free meals) is playing a role in the growth of the meat-alternative market. Only 1 percent of households buy only meat alternatives, compared with 21 percent of households that buy both traditional meat and meat alternatives. Thus, Frey stated, the market of people who buy both traditional and alternative meat presents an opportunity for the industry. She also referenced research indicating some trends among heavy buyers of meat alternatives that included larger proportions of Asian people, larger households, people with high incomes, and individuals ages 45–54.

Slowing Sales of Meat Alternatives

Frey went on to discuss a recent decline in sales of plant-based meat alternatives that has caused concern about whether the market for these products is slowing. In her view, the numbers of new repeat shoppers being brought incrementally into the market should be factored in with sales numbers in

assessing the alternative proteins market. Over the past 2 years, she reported, a quarter of all meat-alternative buyers have been new to these products; 37 percent of these consumers have returned to make a repeat purchase, and more than half of them have made additional repeat purchases. She added that among alternative ground meat, meat patties, chicken tender and nugget analogs, and tofu, the category of meat patties accounts for the highest percentages of first-time, first-repeat, and additional-repeat purchasers. Frey noted that the convenience aspect of patties contributes to their sales.

Frey remarked that at the time of this workshop, McDonald’s had recently announced it would phase out the McPlant alternative patty in the United States, and Cracker Barrel faced backlash after announcing the addition of a plant-based sausage patty to its menu. According to Frey, these events can be attributed in part to some people associating plant-based products with political leanings. She stressed, however, that although the meat-alternative market has seen some decline since 2021, it is nearly double what it was in 2019.

Frey presented the results of a survey designed to shed light on how many consumers are trying meat-alternative products and repeatedly purchasing them, as well as the reasons why some opt not to repurchase. The percentage of meat-alternative buyers who repurchase these products has seen an overall decline, she reported. Among those who purchased meat-alternative products once or twice, the top reasons given for not making additional purchases were taste, not meeting expectations, and cost. Frey noted that the reasons given for not continuing to purchase alternative-meat products were consistent regardless of whether the products were made with beans, vegetables, and grains or with isolated proteins and starches.

Frey also observed that consumers are paying attention to the health attributes and processing of meat alternatives. Digital ethnography, she said, indicated preferences for natural products free of artificial ingredients and containing minimal additives. Approximately 94 percent of alternative-meat products contain artificial ingredients, she noted; 89 percent have more than 200 calories per serving; 86 percent exceed the saturated fat limit; and more than 80 percent are made with more than 21 ingredients. Frey pointed out that regardless of these nutritional issues, health is one of the main reasons consumers cite for consuming alternative proteins.

Projecting Future Protein Trends

Frey next reviewed data collected to inform projections about future trends in alternative proteins. Consumers were asked whether they expected their consumption of plant-based foods and beverages to increase or decrease in the future. Frey reported that almost half of all shoppers said they expected their consumption of these products to increase, and this

proportion was particularly high for consumers of animal-product alternatives (62 percent). Only 7 percent of all shoppers and 5 percent of alternative-product consumers anticipated that their consumption of plant-based foods would decrease.

Frey also presented global data collected to track consumer perspectives on food innovations. Consumers were surveyed about their willingness to try new foods derived from such innovations as regenerative agriculture, vertical farming, and cultured or laboratory-grown meat. Approximately 29 percent of those surveyed were unaware of or unwilling to try products created via regenerative agriculture and vertical farming, said Frey, whereas more than half were unaware of or unwilling to try cultured or laboratory-grown meat (30 percent and 23 percent of those surveyed, respectively). About 54 percent of consumers were willing to try regenerative agriculture or vertical farming products, compared with 37 percent of consumers willing to try cultured or laboratory-grown meat. Frey added that awareness of and willingness to try cultured or laboratory-grown meat varied geographically. For instance, consumers in India and China reported higher awareness of these products, as well as greater willingness to try them, compared with consumers in other regions. Frey noted that these data were collected to help gauge consumer interest in alternatives and to assess the evolution of alternative products.

Frey concluded her presentation by stating that consumer interest in plant-based and alternative proteins is strong and is expected to grow, but consumers indicate confusion and some dissatisfaction with these products. She stressed that different brands have varying levels of repeat customers, customer loyalty, and customer satisfaction, adding that taste, health, and cost are major factors in how consumers meet their protein needs.

SAFETY PERSPECTIVES ON ALTERNATIVE PROTEIN PRODUCTS

Paul Hanlon, Abbott Nutrition, provided his perspective as an industry toxicologist responsible for evaluating the safety of alternative protein products. He outlined the components of food safety assessments, described the process of precision fermentation, and highlighted challenges in conducting hypothesis-driven safety assessments for alternative proteins.

Food Ingredient Safety Assessments

Hanlon opened with an overview of the process through which novel alternative proteins—like all new food ingredients—are evaluated for safety before they can be used in manufactured products. He explained that a new or updated safety assessment of a food ingredient is conducted when an ingredient (1) has never been consumed as a food, (2) is produced via a new

manufacturing process, (3) is being added to a new food product, or (4) is being added to an existing food product in higher amounts.

Hanlon explained that alternative protein ingredients fall into two broad categories: protein isolated from traditional foods and those manufactured using the process of precision fermentation. Proteins extracted from traditional foods, he elaborated, are derived from agricultural commodities such as legumes, oats, corn, soy, dairy milk, nuts, and hemp, whereas precision fermentation is a process in which a microorganism is used to produce a protein (e.g., fungal-produced beta-lactoglobulin, yeast-produced beta-lactoglobulin, yeast-produced ovomucoid). Hanlon noted that even if a protein produced by precision fermentation already exists in the food supply, a safety assessment is required because fermentation is a new method of manufacturing that protein.

Hanlon went on to describe the three key components of a safety assessment for any food substance, including alternative proteins. The first is to define the ingredient by evaluating its manufacturing process and creating an appropriate specification for it. The second step is to assess how much of the ingredient people would be consuming, using information about the foods in which it will be used and estimating the quantity of those foods that people eat. This process informs the third step—determination of whether the amounts of the ingredient expected to be consumed will be safe. Hanlon explained that this step involves evaluating any existing data regarding the inherent safety of the ingredient and, in cases of data gaps, conducting additional safety studies of the ingredient. He noted that safety assessments of food ingredients have been conducted successfully for several decades. Thus, he said, although alternative proteins are relatively new to the food supply, the safety of these novel ingredients is assessed with models that have long been used to evaluate new ingredients.

Precision Fermentation in Food Production

Before describing the three components of the food safety assessment process in more detail, Hanlon provided an overview of how the process of precision fermentation is used to create novel, alternative proteins. He remarked that because the process of precision fermentation uses microorganisms to manufacture these proteins, it can seem futuristic and abstract compared with the process of extracting protein from traditionally consumed foods. To illustrate this point, he compared the concept of consuming a corn protein with that of consuming a milk protein that was never in contact with a cow but instead was produced by a fungus. However, he stressed, microorganisms have been used throughout human history to make such foods as bread, beer, yogurt, cheese, and wine. Hanlon described precision fermentation as an evolution of this long-standing practice that has led to

a more effective, precise method of using microorganisms to create a highly specific food ingredient.

Hanlon elaborated on the point that although precision fermentation is an iteration of an ancient method, the process has been in use for decades. Riboflavin (vitamin B2), for example, a common vitamin added to food products, was originally produced using chemical synthesis beginning in the 1930s (Liu et al., 2020; Revuelta et al., 2017; Schwechheimer et al., 2016). Microorganisms were subsequently identified as a potential mechanism for producing B vitamins, and in the 1960s, nonmodified microorganisms were used to generate riboflavin. Hanlon explained that the yields of that process were not high in comparison with chemical synthesis, and so this original microorganism-driven process never became commercially viable. Then in the 1990s, researchers created a cost-effective method of producing riboflavin using a genetically modified microorganism. According to Hanlon, the new process proved to be so effective that by 2015, it was being used to produce virtually all riboflavin.

Ingredient Specifications

Hanlon returned to his overview of the food safety assessment process by elaborating on the critical first component: establishing a specification for an ingredient. He stated that various criteria are used to define the elements that should and should not be present in an ingredient. Desirable elements, he added, constitute an ingredient’s purity. For an alternative protein, he explained, specifications around purity could relate to nutritive content (e.g., protein, fat, and carbohydrate content) or other desirable content, including amino acid profile, fiber, and minerals such as zinc. The relevant impurities that should be addressed by the specification, Hanlon continued, are determined by evaluating which substances could be present in the finished ingredient at relevant concentrations. Impurities could also include substances inherent in the environment—such as heavy metals and microbiological contaminants—that are incorporated into agricultural commodities and transferred into finished products; substances present in the source of the ingredient, such as alkaloid glycosides in fava beans or glycoalkaloids in protein; or substances introduced during the manufacturing process. Hanlon explained that once an ingredient’s purities and impurities have been identified, minimum and maximum amounts are defined to establish a specification for that ingredient. He added that specifications are similar for proteins isolated from traditional foods and those derived via precision fermentation.

___________________

¹ More information about the GRAS Notice Inventory is available at https://www.fda.gov/food/generally-recognized-safe-gras/gras-notice-inventory (accessed October 12, 2022).

are present that require additional safety studies. Hanlon observed that numerous factors are considered in evaluating whether a traditional and alternative protein are equivalent with respect to safety. For example, the digestibility of traditional and alternative proteins is assessed using both in vitro and in vivo methods.

The safety assessment also explores potential risks introduced by the production process, said Hanlon. For example, a safety assessment of a protein produced with precision fermentation using a fungus will include examining the microorganism to ensure that it is not pathogenic or toxigenic. If a microorganism has been genetically modified, Hanlon added, the genes inserted into the organism will be assessed for safety concerns. Existing literature is reviewed for data regarding any identified potential risks, he noted, and additional safety studies can be required when data are insufficient to support the conclusion that differences between the alternative and traditional proteins are safe. He also described common safety studies used for food ingredients to assess general toxicity (the ability of a substance to produce systemic toxicity) or genotoxicity (the ability of a substance to damage genetic information within cells). General toxicity studies typically involve conducting a 90-day feeding study in rats, he elaborated, whereas a panel of assays is typically used to assess genotoxicity.

Hanlon reflected on some of the challenges that arise in assessing the safety of alternative proteins using these methods. For instance, in the 90-day feeding study in rats, food ingredients are typically administered at an amount far greater than a human would be expected to consume. Hanlon explained that because approximately 20 percent of a standard rat diet is composed of protein, it is not possible to increase a rat’s protein intake by 10- or 100-fold without disrupting the nutritional balance of the animal’s diet. Hanlon also provided the example of safety studies conducted on food enzymes, proteins often produced via precision fermentation. He noted that a review of more than 200 genotoxicity and general toxicity studies of enzymes found an absence of any adverse effects (Ladics and Sewalt, 2018). He added that proteins, including enzymes, are readily digested into amino acids by mammals after consumption. Thus, he maintained, animal studies are unlikely to have any value for evaluating the safety of most alternative proteins, and other methods of demonstrating the safety of these proteins, including nonanimal methods, may be more helpful. This matter is gaining worldwide traction, Hanlon observed, noting that in November 2022 the World Health Organization (WHO) and FAO would be holding a joint expert meeting as a first step toward developing global guidance on the safety assessment of alternative proteins.

Fasano elaborated, the FDA maintains a voluntary premarket program under which the agency evaluates a firm’s conclusion of GRAS status for its intended use of an ingredient. The submissions and the agency’s responses are posted to the GRAS Notice Inventory on the FDA website.

Fasano emphasized that the standard for safety and evidence required for a GRAS conclusion is the same as that for an approved food additive. He explained that for the GRAS provision to apply to an ingredient, safety data must not only demonstrate reasonable certainty of no harm from the intended use but also be (1) generally available (i.e., not confidential and published in peer-reviewed publications) and (2) generally accepted (i.e., through consensus among qualified scientific experts regarding safety). Fasano remarked that in a sense, it can sometimes be more challenging to satisfy the criteria for GRAS status than those for a food additive because the latter requires approval only by the FDA, whereas meeting the GRAS criteria involves convincing the relevant scientific community as a whole that a particular use of an ingredient is safe.

Fasano explained that the safety standard for food additives outlined in the FD&C Act requires reasonable certainty that a substance will not cause harm, not absolute proof of safety. The FDA’s approach to assessing the safety of substances added to food, he summarized, includes identifying the ingredient, determining the exposure that will result from its intended use, considering its relevant properties, and reviewing appropriate data. As described earlier by Hanlon, the basic elements of the safety assessment are (1) defining and characterizing an ingredient, its manufacturing process, and specifications; (2) identifying its intended uses; (3) estimating maximum levels of consumption for the intended use; and (4) determining whether it will be safe at the estimated exposure levels. Fasano noted that this process applies to any substance being added to food.

Safety Assessment of Precision Fermentation Products

Fasano reiterated Hanlon’s explanation that the method of precision fermentation, or recombinant fermentation, uses microorganisms to produce food ingredients (Figure 3-1), and that bacteria and fungi have a long history in traditional food production to make yogurt, cheese, pickles, wine, bread, and other foods. Fasano noted that fermentation has also been used as a source of enzymes to treat or otherwise modify food. Originally, he said, these enzymes were those already produced by the organism. In 1990, however, the FDA evaluated a genetically engineered microorganism used in the production of an ingredient for use in conventional food for the first time (recombinant chymosin). In the decades since, Fasano observed, the FDA has routinely evaluated recombinant ingredients derived from microbial fermentation through its premarket programs.

for contaminants specific to the new process. Likewise, he added, any differential posttranslational modification made in producing a protein from a particular source would be assessed to determine whether this change would affect digestibility or other factors. Any potential differences resulting from a process are identified to evaluate their relevance from a safety perspective, said Fasano.

Safety Assessment of Animal Cell Culture in Food and Feed Production

Fasano explained that animal cell culture is a process that is widely used in a therapeutic and research context and is now being explored for use in food production. In a partnership established in 2019, the FDA and USDA’s Food Safety Inspection Service (FSIS) jointly oversee products that feature cultured livestock or poultry cells to be used in human food.² The FDA oversees cell collection and cell culturing and conducts premarket consultations on production processes. Both the FDA and FSIS provide oversight over the harvesting of live cellular material in the field. FSIS oversees conventional processing, packaging, and labeling of products containing harvested cellular material. The FDA oversees the entire process, Fasano noted, from collection through market, for human foods incorporating cultured fish and seafood cells, with the exception of Siluriformes fish.

The FDA’s agreement with USDA includes a commitment to multiple activities related to the use of this new food production technology, said Fasano. He detailed these activities as (1) conducting premarket consultations; (2) overseeing cell collection, banking, and culture; (3) coordinating with FSIS to oversee the harvesting of cellular material from livestock, poultry, and Siluriformes fish; (4) enforcing FDA requirements applicable during production; (5) conducting inspections as necessary; and (6) overseeing food products incorporating cultured fish and seafood cells. Fasano noted that during premarket consultation, numerous factors are evaluated, such as the production materials used as inputs in the process, the process for initial tissue collection to establish cell lines, the development and maintenance of cell lines and banks, and the properties of the material at collection.

In summary, Fasano stated that the FDA has methods for evaluating substances or additives to ensure that they will not adulterate the foods to which they are added. Production processes are considered with regard to any effects they may have on the safety or properties of food. Fasano

___________________

² More information about the formal agreement between the FDA and USDA regarding oversight of human food containing cultured animal cells is available at https://www.fda.gov/food/domestic-interagency-agreements-food/formal-agreement-between-fda-and-usda-regarding-oversight-human-food-produced-using-animal-cell (accessed October 13, 2022).

remarked that the FDA’s goal is not to identify differences between foods with and without the substances or additives but, more specifically, to identify differences that may be relevant from a food safety perspective. Thus, as production processes change, he concluded, so too may the data required to establish safety.

REQUIREMENTS FOR LABELING OF ALTERNATIVE PROTEIN FOODS

Douglas Balentine, FDA CFSAN, outlined the required components of food labels. He opened by explaining that existing labeling requirements for all foods are applied to all alternative protein foods to ensure that labels are meeting a consumer’s expectations and needs. The FDA may consider issuing guidance around how to apply the labeling guidelines appropriately for these new products as it deems necessary, he added, as it is doing for labeling of plant-based alternative foods. Balentine clarified that his presentation would focus on labels overseen by the FDA, noting that USDA has labeling authority over many food products containing meat, poultry, and certain types of fish. As mentioned by Fasano, the FDA and USDA also have a shared responsibility for labeling food products derived from cell culture technologies. As these products evolve and enter the marketplace, said Balentine, the two agencies may take somewhat different approaches, but generally the methods they use with respect to the labeling of food products are aligned. He highlighted as one notable difference the use of premarket approval for labels by USDA and not by the FDA, which conducts postmarket label reviews.

For most consumers, Balentine remarked, the food label is an important source of information for guiding food choices and understanding the nutritional composition and benefits of foods they consume. He observed that in general, the label outlines a food’s nutritional composition and may contain claims and other messages. More specifically, mandatory components of the label include (1) a statement of identity (i.e., what type of food the package contains); (2) the net quantity of contents within the package; (3) the name and location of the company that manufactures the product; (4) an ingredient statement; (5) nutrition labeling; (6) allergen labeling if applicable; and (7) material facts about the food, such as safe handling information or preparation instructions. He noted that existing labeling regulations apply to all foods, including emerging alternative protein products.

Statement of Identity

A product’s statement of identity is the name of the food, said Balentine, adding that the name may be required by law or legislation (e.g., butter,

cheddar cheese), whereas in other cases, a common name is used for a food (e.g., tomato soup, chocolate chip cookie). Such terms have emerged over time, he noted, and are commonly understood to represent specific foods. Balentine added that marketers often use a fanciful, branded, or registered trademark name for a particular food in an effort to make the food stand out. If the name a marketer has selected for a food is not sufficiently clear, he said, an appropriately descriptive term or set of terms must be included to provide clarity to consumers. According to Balentine, descriptive terms are often needed for alternative protein foods to ensure that consumers understand what they are purchasing. For example, the statement of identity for a plant-based yogurt could be “soy-based yogurt alternative,” making clear that the product is an alternative to traditional yogurt. Thus, stated Balentine, statements of identity help consumers understand the nature of the food they are buying and how it is to be used. He emphasized that this can be particularly important for foods that are substituting for or replacing animal-based products.

Ingredients and Allergens

An ingredient statement listing all food ingredients in descending order of predominance by weight is a requirement for all packaged foods, said Balentine. The common names of ingredients must be used, and sub-ingredients must be included. Names of novel ingredients, such as those featured in some alternative protein products, can include a comment or bullet to add further clarity about the ingredient’s identity. Balentine pointed out that using similar names for an ingredient across product categories helps consumers recognize the ingredient from one product to another. Any ingredients that constitute less than 2 percent of the food’s weight may be grouped together.

Balentine stated that allergen labeling is required for any major food allergens, as defined by the FD&C Act. He stressed that this feature is important for consumers with food sensitivities or allergies and noted that alternative proteins made from fermentation processes are potentially allergenic; for example, dairy proteins produced from yeast or other fungi are likely to have the same allergenic potential as dairy proteins produced by animals. Thus, Balentine stressed, appropriate disclosure is critically important to consumers with milk allergies. Likewise, he pointed out that insect proteins may have the potential to cause cross-reactions for people with crustacean allergies, depending on how the proteins are purified and isolated and the nature of the ingredient itself. Thus, he said, as new ingredients come to the marketplace, it is important to monitor for the emergence of any cross-reactions or new allergies to appropriately address any allergen-related risks.

___________________

³ More information about nutrient content claims is available at https://www.fda.gov/regulatory-information/search-fda-guidance-documents/guidance-industry-food-labeling-guide (accessed October 14, 2022).

⁴ More information about the Codex is available at https://www.fao.org/fao-who-codexalimentarius/en/ (accessed October 31, 2022).

standards regarding nutritional composition and labeling requirements to move toward a common set of international guidelines.

DISCUSSION

The discussion following the presentations summarized above focused on regenerative agriculture, regulatory processes, the role of mimicry in plant-based alternatives, repurposing agricultural land, consumer perceptions of processing and ingredient count, and defining alternative protein.

Regenerative Agriculture

Williamson asked whether regenerative agricultural practices could serve as an intermediary step toward eliminating animal agriculture, allowing some level of meat production while decreasing the harmful effects of traditional animal agriculture. Brown remarked that “regenerative agriculture” is not a well-defined term, and that it is often applied to a certain grazing approach for which little evidence indicates that it improves carbon levels and some contrary evidence exists. He argued that replacing millions of acres of the Amazon rainforest with cows regeneratively grazing and ostensibly putting carbon into the soil will never approach the climate benefit of removing 22 years’ worth of fossil fuel emissions and restoring native ecosystems. Any effort to decrease harmful effects on the climate is a step in the right direction, he acknowledged, but the levels of effectiveness of regenerative agriculture and elimination of animal agriculture are vastly different.

Regulatory Processes

Williamson posed a question from an audience member about the safety and regulatory processes involved when a new substance is added to food. Hanlon responded that an assessment is performed on any new substance—whether a protein is extracted from an agricultural source or created via precision fermentation—to ensure that its safety is equivalent to that of similar substances already available to consumers. This process, he explained, involves examining the ingredient’s composition and consumption patterns. To illustrate, he used the example of a milk protein produced by a fungus for which it could be determined that there were no compositional differences from a milk protein produced by a cow, so that assurance of the safety of the fungal-produced protein could rely on data demonstrating the safety of its counterpart produced by a cow.

In response to a question about whether companies can volunteer additional safety testing data during or after the GRAS approval process,

Fasano clarified that the FDA does not issue GRAS approvals; rather, if a substance meets GRAS criteria, it is exempt from the requirement to obtain premarket approval. Thus, he said, GRAS status is an exemption from an approval process, not an approval in and of itself. For a substance to meet GRAS criteria, he added, evidence of safety must be publicly available and supported by broad consensus. Fasano noted that given the level of evidence required to meet GRAS criteria, a company making a robust case for a GRAS exemption would not need to submit additional data. Therefore, further data would be relevant only in specific circumstances, such as a modification being made in the production process.

Role of Mimicry in Plant-Based Alternatives

Williamson asked about an approach for decreasing consumption of animal-based proteins by creating new plant-based products that do not attempt to mimic animal products. For example, falafel nuggets could provide a plant-based alternative to animal protein without having to match the taste, texture, and appeal of an animal-based counterpart. Frey reiterated that consumer responses indicate the importance of similarity to a dairy or meat counterpart as a factor in plant-based purchases. This preference, she asserted, also is reflected in the popularity of plant-based patties and nuggets, which are both convenient for and familiar to consumers. At the same time, however, Frey pointed out that consumer interest in plant-based products across categories indicates an opportunity for innovation beyond efforts to mimic analogs. Messina referenced several studies indicating that plant-based products that do not attempt to replicate the experience of consuming animal-based products are more readily accepted by people already on a plant-based diet than by those on an omnivorous diet. He cited as an example black bean burgers, remarking that they do not mimic beef and, despite having been available for decades, have neither achieved mass appeal nor shifted the animal-to-plant protein intake ratio in any country. He added that people who consume animal-based products are more likely to eat products similar in taste and texture to the products with which they are familiar than unfamiliar plant-based products.

Repurposing Agricultural Land

A question was raised about the potential for land recovered through reduced animal farming to be used for nonsustainable purposes. Brown responded that this topic is being actively considered at Impossible Foods. He stressed that animal agriculture is the most land-intensive human activity by a wide margin, with all cities and suburbs in the world occupying about 3 percent of the amount of land used for animal agriculture. Therefore, he

argued, strong economic drivers for the repurposing of agricultural land for commercial construction are absent. Brown commented further that, according to several studies, abandoned agricultural land can readily recover native ecosystems provided it has not been severely damaged. He observed that forests are flourishing in what used to be New England farmland, and forest is even found at the Chernobyl site just 35 years after the nuclear accident there occurred. Thus, he maintained, land will default to recovery of biomass unless economic incentives drive its use for other purposes. Brown also highlighted movement toward creating economic incentives for carbon capture. Although he acknowledged that the future of this market is unpredictable, he pointed out that the current demand for carbon capture in the voluntary carbon market exceeds the supply of credible carbon capture products. Thus, he asserted, there could be significant earning potential in this area for farmers, given that U.S. cattle ranchers average single-digit earnings per acre, whereas restoring native ecosystems and selling carbon credits at current voluntary market carbon prices offers earnings several-fold greater. He argued further that economic incentives for restoring ecosystems and capturing carbon are more likely than incentives for other land uses.

Consumer Perceptions of Processing and Ingredient Count

Hanlon remarked that as a food toxicologist, he does not relate his food choices to the number of additives or processing techniques used to produce a particular food. Pointing to the heavy emphasis of classification systems such as NOVA on the degree of processing and number of additives for a food, rather than on its nutrient content or energy density, he asked about the effect of such classification systems on consumer choices. Frey replied that in many countries around the world, consumers are increasingly reporting attention to the level of processing used for foods. This conversation is starting to gain attention in the United States as well, she noted, along with the idea that consumers should avoid products containing hard-to-pronounce ingredients. Over the past few years, she pointed out, enormous growth has occurred in “clean labeling,” which features claims that products do not contain artificial ingredients, flavors, or colors. Beyond foods, she added, preferences for natural products are increasingly applied to home and body products. In recent months, she said, ingredient count has emerged as another consumer focus. She stated that this is not a positive or negative development but a lens through which to understand how consumers view products with higher or lower ingredient counts. In summary, Frey observed that consumers are examining labels, reading ingredients, and indicating a preference for natural products with clean labels.

Messina contended that the food industry has done a poor job of educating the public around the role of food processing and ingredient count.

The number of ingredients is completely unrelated to a product’s nutrient quality, he remarked. To illustrate he pointed to such products as the Impossible Burger that have long ingredient lists because they are fortified with nutrients. Furthermore, he maintained, the complexity of ingredient names on labels does not necessarily reflect the safety of the ingredients. And he noted that “ascorbic acid” is more difficult to pronounce than “vitamin C,” yet these are two names for the same nutrient.

Williamson asked Frey whether consumers have indicated that they pay attention to nutrient density. Frey replied that consumers pay attention to nutrients listed on nutrition labels but that the topic of nutrient density is absent from consumer responses. She added that researchers are discussing different metrics for measuring the nutrient content of products. Williamson commented that products sometimes feature claims about nutrient content, such as the claim that alternative protein sources have higher fiber content than that of their animal counterparts.

Defining Alternative Protein

Naomi Fukagawa, USDA, asked the panel to define “alternative protein.” Balentine remarked that in his work with Codex, the discussion features a distinction between long-standing plant-based proteins, such as tempeh, tofu, and black bean burgers, and products developed to resemble and replace animal-based counterparts, such as beef, chicken, or salmon. In this sense, he suggested, the term “alternative protein” does not apply to all sources of plant-based dietary protein. He pointed out, for example, that legumes are a source of protein but would not be considered an alternative protein food, as they are not intended to directly replace a specific animal-based product. Fasano stated that many forums refer to alternative proteins as a conceptual group, but from the safety perspective, these foods are produced with varied technologies that often have little relationship to one another. Thus, he maintained, the functional or consumer perspective on alternative proteins is quite different from the technical one. Hanlon said he applies the term “alternative protein” to a food that is new, featuring either novel composition or an existing protein produced via a new manufacturing method. Frey observed that consumers wishing to increase consumption of plant-based foods can eat either inherently plant-based foods, such as fruits and vegetables, or products that mimic conventional products. In the data she shared in her presentation, “alternative protein” refers to counterparts to conventional proteins that offer consumers an alternative to traditional foods. Her company is tracking protein trends in terms of whether they are proven, growing, or developing because this information helps researchers understand the evolution of the plant-based protein perspective.