Review of plant-based milk analogue: its preparation, nutritional, physicochemical, and organoleptic properties

Abstract

In recent years, the market demand for plant-based milk analogues has been rising because of health concerns with bovine milk, like lactose intolerance and hypercholesteremia. Another reason is the lifestyle changes like adopting veganism. This review aims to offer a layout of the manufacturing process and discuss the different properties of plant-based milk analogues. The health benefits offered by the plant-based milk analogues and measures taken to eliminate the existing limitations are also discussed. Sensory profile and stability of plant-based milk analogues which add to the quality of the product were also taken into account and reviewed. The current review's objective is to present a comprehensive, scientifically comparable overview of the preparation procedures, nutritional content, and sensory characteristics of plant-based milk analogues. This is done while keeping in mind the potential of plant-based milk substitutes and associated challenges.

Introduction

Human or animal milk has been the first source of nutrition consumed worldwide by all humans. It continues to be a vital source of nutrients like protein, minerals like calcium and phosphorous, and vitamins like vitamins B12, A, and D (Aydar et al., 2020). However, consumption of certain constituents like cow protein and lactose can be problematic for people having cow protein allergy or intolerance to lactose. These health concerns lead to the demand for alternatives to animal milk, i.e., plant-based milk, which has increased by 61% since 2012 (Clay et al., 2022). Another factor contributing to the demand for plant-based milk is people's adoption of veganism for the welfare of animals and the environment. However, it is questionable whether we should consume plant-based milk regularly in the same way that we consume bovine milk. The US Food and Drug Administration currently defines “milk” and related milk products by the product source and the inherent nutrients provided by bovine milk. North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition Committee in response to a September 18, 2018 FDA Federal Register Request for Comment regarding the “Use of the Names of Dairy Foods in the Labelling of Plant-Based Products” (Docket No. FDA-2018-N-3522) commented that a child's nutritional status, growth, and development can be affected by milk substitution which does not provide a similar nutritional profile as that of cow milk (Merritt et al., 2020). They concluded in the letter “From a pediatric medical and nutritional standpoint, it is advisable that milk’ be: 1) milk products as currently defined by FDA, or 2) provide comparable nutritional value to standard ‘milk.’ Such labeling, and education regarding this labeling, may reduce adverse nutritional effects from consuming nutritionally non-equivalent plant-based products labeled as ‘milk.’ Thus, we can see that there are wide differences in the opinions of all the stakeholders such as consumers, producers, Regulatory bodies, etc. This review can help understand a broader perspective and can prove useful to harmonize the perspectives for the use of plant-based milk analogues.

Plant-based milks are essentially extracts obtained from seeds, nuts, and cereals. Although there isn't a defined classification of kinds of plant-based milk, based on the source one can classify plant-based milk into 5 kinds as shown in Fig. 1 (Romulo, 2022).

Fig. 1.

Classification of plant-based milk based on raw material used

Plant-based milk i.e., essentially the fluid extracted by breaking down the plant material (seeds, nuts, legumes, and cereals) in water followed by homogenization is carried out to imitate the appearance and viscosity of cow's milk. These extracts are further processed to enhance their nutritional quality, stability, and acceptability. Plant-based milks inherently have certain nutritional benefits like high fiber, healthy fats, and minerals like iron and copper, which are absent in animal milk (Vanga and Raghavan, 2018). They also contain bioactive compounds (such as isoflavones in soymilk; avenanthramides in oat-based milk; sesamolin found in sesame) that have health benefits. In plant-based milk, the bioavailability of proteins (as low in protein digestibility value) and minerals is a concern. The nutrition quality of plant-based milk can be further improved by eliminating or reducing the allergen and antinutritional factors present. Plant-based milk analogues are often fortified with minerals like calcium and phosphate. This review discusses the production methods of plant-based milks and their nutritional properties in detail.

The physicochemical and sensory characteristics of cow milk are often set as standard characteristics by consumers (Jeske et al., 2017). Efforts are put in to eliminate the off-flavor and increase consumer acceptability. Rheological properties of the plant-based milk and mouthfeel are also studied and improvised using techniques like Ultra High-Pressure Homogenization (UHPH) for improved mouthfeel. A lot of research is being done on understanding the techno-functional characteristics of plant-based milk and using them to create a plant-based product range. A few of these studies are included in this review to give a perspective on what more can be done with plant-based milk.

Preparation method for plant-based milk analogue

The flowchart for the preparation of plant-based milk analogue is presented in Fig. 2. Based on the process of milling, the preparation method has been classified as wet processing and dry processing. Owing to ease of handling, the most prevalent method of producing plant-based milk is extraction via wet processing (Aydar et al., 2020). In wet processing, water is added to the pre-treated raw material before or during size reduction. The ultimate concentration of plant-based milk may depend on several variables, including temperature, pH, water content, and extraction rate (Romulo, 2022). In dry processing, water is added after size reduction by milling. Colloidal mills are used for dry grinding. A comparative study on wet and dry milling of pistachio milk showed that the extraction of fat and protein was more efficient with dry milling (Shakerardekani, 2013). The next step is to filter the extract to remove residues. After this, the product is formulated, treated with heat, and homogenized to bring uniformity and stability to the final product. The final product is aseptically packaged and stored.

Fig. 2.

Flowcharts showing the preparation of plant-based milk analogues: (A) Wet processing; and (B) Dry processing (Aydar et al., 2020; Reyes-Jurado et al., 2021)

The significance of the important unit operations involved and the impact of operational parameters is further discussed.

Dehulling and peeling

Dehulling is the first step in the manufacturing of plant-based milk from shelled raw materials like walnuts. Dehulling and peeling of plant-based milk sources significantly enhance the nutritional profile and sensory qualities of the milk, as it reduces anti-nutrients and improves digestibility, ultimately leading to a more balanced and appealing milk alternative. For ease of dehulling, the raw material is soaked in hot water for a certain period. The dehulled raw material can be peeled after it is cooked and steeped in diluted acid and base. The ease of peeling will depend on raw material, pH and concentration of the solution, and time and temperature of soaking. For instance, tiger nuts took 24 h of soaking in 1% citric acid solution whereas walnuts took 2–3 min when soaked in 2% citric acid solution at 90 °C. Studies have also shown that walnuts need 18–20 h of soaking if soaked in water (Aydar et al., 2020; Kizzie-Hayford et al., 2016). After the treatment, washing is necessary to remove the residual acid or base. Dehulling and peeling contribute to improved nutrition properties through the removal of anti-nutrients and improved sensory properties by the removal of off-flavor-generating compounds (Aydar et al., 2020). For instance, Soybean soaked for 24 h and manually dehulled showed a significant reduction in trypsin inhibitor levels which is a major anti-nutrient present in soybean (Joshi and Varma, 2016).

Roasting

Roasting is a crucial step in the production of plant-based milk as it enhances the flavor profile and improves the nutritional quality by increasing antioxidant activity and reducing antinutritional compounds. Roasting conditions must be correctly optimized to avoid negative impact on the sensory and functional properties of the plant-based milk. Roasting conditions that have been studied and optimized in some cases have been enlisted in Table 1. Roasting of soybean helped in the reduction of off-flavor-producing volatile compounds like hexanal and hexanol, reduction in trypsin inhibitors, and heat treatment time. Along with the sensory and nutritional benefits, roasting improved the emulsion stability of the soymilk (Navicha et al., 2018). Further, the study suggested that roasting time and temperature combination may also negatively affect the yield, color, and sensory properties. Hence, optimum roasting conditions are to be set for lipoxygenase inactivation and protein extraction (Navicha et al., 2018). Conventional roasting can be replaced with microwave roasting for better product and nutritional quality (Le and Le, 2021). A study on the roasting of cashew milk concluded that roasting did not result in a significant change in sensory acceptance (Lima et al., 2020). Zaaboul et al. (2019) work on roasting peanuts enhanced protein solubility and emulsifying properties.

Table 1.

Roasting and soaking parameters, and extraction method used in plant-based milk preparation

Raw material	Conditions	References
Roasting
Soybean	110 °C for 20–100 min or 120 °C for 20 min	Navicha et al. (2018)
Red-brown rice	Microwave roasting (800 W; 1, 2 and 3 min)	Le and Le (2021)
Peanut	Short wave infrared roaster (220 °C for 25 min)	Zaaboul et al. (2019)
Soaking
Soybean	12 h before extraction in water	Niyibituronsa et al. (2019)
	16 h soaking in NaHCO3 solution followed by cooking
Soybean	55 and 60 °C for 2, 4 and 6 h	Nowshin et al. (2018)
Tiger nut	6 h	Asante et al. (2014)
Sesame	0, 0.5, and 1 g/100 mL NaHCO3 containing water 16 h at 25 °C	Ahmadian-Kouchaksaraei et al. (2014)

Raw material	Extraction method	References
Soybean	Microwave-assisted extraction Optimum parameters: 675 W, 80 °C, 160 rpm Increased yield and protein content	Varghese and Pare (2019)
Hemp seeds	Soaking in water at a temperature below 80 °C	Besir et al. (2022)
Rolled oats	Enzymatic extraction: α-amylase from Bacillus subtilis Optimum conditions: slurry concentration – 27.1% w/w; enzyme concentration – 2.1% w/w; liquefaction time – 49 min	Syed (2020)
Cottonseed milk	Enzymatic assisted aqueous extraction technique: Protease enzyme Optimized conditions: Enzyme concentration -0.50%; Temperature – 30°c; pH – 7.0; Hydrolysis time – 165.31 min Increase in extraction yield – 11.11%; Increase in protein content -21.35%	Subramani et al. (2022)

Soaking

To produce plant milk, it is vital to soak and heat-treat plant materials (Asante et al., 2014). This process removes antinutritional elements, improves the taste and digestibility of the milk, and softens plant materials to make grinding and extraction easier. Soaking soybeans in water resulted in high nutrient and isoflavone extraction (Niyibituronsa et al., 2019). Lipoxygenase activity was reduced in soaking. Increasing the soaking temperature had a positive effect on the solids extracted. Increasing soaking time did not show any impact on urease, phytate, and trypsin activity but it did reduce lipoxygenase activity (Nowshin et al., 2018). Protein solubility, ash content, stability, and lightness index of sesame milk increased on soaking. Table 1 covers the different soaking conditions that have been stated in recent literature.

Blanching

To increase the flavor and nutritional content of non-dairy milk alternatives, lipoxygenase, and trypsin inhibitors are inactivated by blanching in hot water. In plant-based milk made from oats, soybeans, almonds, etc., this treatment helps reduce the beany, rancid flavor. It also increases suspension stability and reduces chalkiness (Aydar et al., 2020). Raw materials are often blanched in hot water (85–100 °C for 2–5 min) to remove their skins and remove undesirable odors from non–dairy milk substitutes (Dhankhar and Kundu, 2021). Peanut milk when blanched under pressurized conditions (121 °C, 15 psi, 3 min) showed improved sensory and physicochemical characteristics (Jain et al., 2013).

Dry grinding

Dry raw materials are typically ground into powder form by using a dry mill to minimize the size of the particles. After that, the ground material is combined with water to create a paste. However, solids tend to settle out in the container during the production of plant-based milk made from pastes, resulting in an insufficient transfer of the content to the homogenizer and its loss as well. The particle stiffness and feed size must be taken into account to ensure that the dry grinding process operates effectively. Dry grinding produces a product with a higher quality of protein, carbs, fat, and minerals while using less energy and water, but it is less popular due to handling issues, such as dust and waste (Dhankhar and Kundu, 2021). Roasted almonds are dry-milled to create a proprietary almond milk substitute.

Extraction

The process of extraction is of utmost importance in the making of plant-based milk, as it has a substantial impact on the ultimate nutritional composition and sensory characteristics of the product (Silva et al., 2020). Various extraction methods have been reported and they significantly impact the composition of the final product (Table 1). By raising the pH using alkali or bicarbonate, raising the temperature, or employing enzymes, this step can be made more effective to enhance yield. To improve the protein extractability during extraction, an alkaline medium is necessary; however, a neutralization phase may also be necessary. Additionally, although protein denaturation reduces fat solubility, a greater extraction temperature increases the extractability of fat (Dhankhar and Kundu, 2021).

When compared to the steam infusion method of extraction, the use of microwave-aided extraction is an appealing alternative because it raises the extraction yield of soymilk as well as its crude protein, overall soluble solids, and protein solubility. When soy pulp was heated in a microwave, it showed more pores than when it was heated with steam. Therefore, polysaccharides and other trapped components are more easily soluble when heated using a microwave (Varghese and Pare, 2019). It has been reported that a different way of enhancing extraction yield involves the partial enzymatic hydrolysis of proteins and polysaccharides. For instance, neutral proteinase boosts soymilk extraction’s protein and solids yields by 40% and 24% respectively (Reyes-Jurado et al., 2021).

Wet milling

Wet milling is of utmost importance in the creation of plant-based milk due to its capacity to significantly improve the extraction of proteins, lipids, and vital nutrients from the ingredients. This process leads to enhanced sensory characteristics and increased nutritional value in the final product. The process entails using water to grind or mill the plant material to split up the external hull. Wet grinding is the process of creating a suspension by pulverizing fresh raw materials with water (Aydar et al., 2020). The wet grinding method is more frequently employed for production since it tends to create finer particle sizes of the ground material that results in higher stability of non-dairy milk alternatives. A colloid mill is typically used to reduce the size of raw materials in suspension. The raw material is first subjected to coarse grinding, which is then followed by fine grinding. Colloid milling performs a wide range of tasks in addition to disintegration, including mixing, blending, and homogenizing effects. Factors like the rotor speed of the mill, and the temperature of the milling will have an impact on the final product (Dhankhar and Kundu, 2021). The ratio of raw material and water important factor in wet milling due to the different compositions of raw materials. Ratios such as 1: 4, 1:2, 1:2, and 1:6 have been used for soybean, tiger nut, sesame, and peanut respectively with water (Nowshin et al., 2018; Asante et al., 2014; Ahmadian-Kouchaksaraei et al., 2014; Jain et al., 2013).

Filtration

Filtration plays a pivotal role as it significantly enhances the product's sensory attributes, product's stability, and particle size. Different methods have been used for filtration in the making of plant-based milk. Some of the filters used are ultrafiltration (membrane pressure 0.05–0.1 MPa, stirring rate 100–400 min⁻¹) in soy and oat milk; muslin cloth in oat milk; muslin cloth (pore size 180 μm) in almond milk; and filter paper in coconut milk (Eazhumalai et al., 2022; Hodúr et al., 2021; Manzoor et al., 2019; Sun et al., 2022). Eazhumalai et al. (2022), suggested that the size of the filter could be one of the reasons that impacted the composition of resultant oat milk. This highlights the added significance of filtration.

Product formulation

Product formulations are done by adding stabilizers, flavoring agents, preservatives, emulsifiers, sweeteners, coloring agents, etc. Other than those, alkalizing agents (like sodium bicarbonate) help prevent the settling of solid material affecting the milk stability. Some of the lists of commonly used additives and their roles are stabilizers (Xanthan gum, guar gum, carrageenan); preserving agents (sodium metabisulphite); antioxidants (citric acid); and alkalizing agents (sodium bicarbonate) (Dhankhar and Kundu, 2021). During this step, fortification of plant-based milk analogues can also be done. The bioavailability and stability of fortificants need to be checked.

Homogenization

By using shear forces, homogenization is used to reduce the dispersed phase components’ sizes in the range of 0.5 to 30 µm. When left to stand for a while, the particles of the dispersed phase, such as protein, fiber, starch, and other cellular materials, tend to settle to the bottom; however, the stabilization of suspension is accomplished during the production of plant-based milk because of the contribution of size reduction brought on by homogenization and the addition of emulsifying agents or hydrocolloids. To avoid phase separation, the technique helps break fat globules, which promotes the formation of creamier, more homogenized products (Dhankhar and Kundu, 2021). Valencia‐Flores et al. (2013) work on Ultra High-Pressure Homogenisation (UHPH) of Almond milk showed a reduction in the surface mean diameter from 1.4 to 0.29 µm at 200 MPa, 50 °C and 1.4 to 0.26 µm at 300 MPa, 50 °C. UHPH treatment on soymilk reduced the surface globule diameter from 0.37 to 0.17 µm, 0.15 µm and 0.14 µm at 70 MPa, 140 MPa and 210 MPa respectively (Mukherjee et al., 2017).

Heat treatment

The shelf life of plant-based milk is extended through the use of high-temperature processes like pasteurization, ultra-high-temperature (UHT) treatments, or sterilization. However, it has been noted that high-temperature treatments might cause the proteins to coagulate, which can make the plant-based milk unstable. This is because proteins expand at high temperatures, exposing nonpolar amino acid residues that engage in protein–protein interactions and subsequently display changes like aggregation, settling, or gel-forming. By destroying aggregates and lowering particle size dispersion, homogenization treatment after heat processing enhances suspension stability (Mäkinen et al., 2015). In addition to improving physical stability, these heat treatments also result in the simultaneous eradication of harmful bacteria in plant-based milk substitutes, increasing the plant-based milk shelf life (Aydar et al., 2020). Non-thermal heat treatments like Ultrasound, Pressure processing, and Pulsed Electric Field can be used for better nutrient retention. Manzoor et al. (2019), comparative study on heat treatment methods for almond milk showed that Pulse Electric Field treatment at an electric field (EF) strength of 28 kV/cm; pulse frequency of 1 kHz; pulse width of 40 µs and treatment time of 200 µs showed the same result as thermal treatment at 90 °C for 60 s. The Total Plate Count reduced from 4.2 log CFU/mL to ~ 0.5 log CFU/mL.

Nutritional and physicochemical properties of plant-based milk analogue

Nutritional profile of plant-based milk analogue

The nutritional profile of plant-based milk depends on the raw material used, its formulation, and the process of manufacturing (McClements et al., 2019). The nutritional profile of bovine milk and different kinds of plant-based milk analogues are compared in Table 2.

Table 2.

Comparison of nutritional values of cow’s milk and different plant-based milk analogues

	Cow milk	Hemp milk (seed)	Tiger nut milk (nut)	Black rice milk (cereal)	Quinoa milk (pseudocereal)	Pea milk (legume)
Raw material: water ratio	–	1:5	1:3	3:10	1:5	-
Energy (Cal)	63.9	19–53.09	6.32	33.31	–	120
Carbohydrate (%)	4.01 (lactose)	0.3–20 (sucrose and glucose)	3.17 (sucrose, starch without glucose, and fiber)	6.69 (maltose and glucose)	16.2 (glucose, sucrose, and starch)	12 (sucrose and glucose)
Protein (%)	4.09	0.83–4	6.95	0.13	0.68	8
Fat (%)	3.50	1.25 – 4.61	5.43	0.67	0.93	4.5
Moisture (%)	87.35	91.6	81.5	92.36	82.06	75.5
Fiber (%)	–	0.4	0.98	–	–	1
Ash (%)	1.05	0.47	1.95	0.15	0.13	–
References	Jeske et al. (2017) and Romulo (2022)	Besir et al. (2022) and Walther et al. (2022)	Gambo and Da’u (2014) and Opeyemi and Obuneme (2020)	Jeske et al. (2017) and Romulo. and Sadek (2022)	Kaur and Tanwar (2016) and Pineli et al. (2015)	Craig et al. (2021) and Kornet et al. (2020)

Protein, Carbohydrate, and Total fat content vary widely among the plant-based milks in itself. The carbohydrate content in plant-based milk is either comparable to or higher than bovine milk as they are sugar is added for consumer liking (Paul et al., 2020). Plant-based milk analogues have sucrose and glucose as their main sources of carbohydrates, in contrast to milk, which only contains lactose as a source of carbohydrates. The following carbohydrates—sucrose, fructose, glucose, lactose, and starch—were assessed in all milk analogues made from plants (Walther et al., 2022). Animal-based milk contains 0% fiber whereas, in plant-based milk, fiber content varies from 0% in coconut to 1.95% in tiger nut milk. These values depend on the process of manufacturing as some studies have shown that the fiber in coconut milk is as high as 6% (Silva and Sergiy, 2022). Fat content in plant-based milk is usually lower than bovine milk with a few exceptions like tiger nut and pea milk (Table 2). Plant-based milk usually contains less amount of protein except for soy milk and tiger nut milk which are comparable to that bovine milk (Chalupa-Krebzdak et al., 2018). However, qualitatively soy milk is inferior to bovine milk because of limiting essential amino acid quantities (Reyes-Jurado et al., 2021). The DIASS scores of bovine milk i.e., 109 and that of soy milk i.e., 90.7 attest to the difference in protein bioavailability (Chalupa-Krebzdak et al., 2018; van den Berg et al., 2022).

Physicochemical properties of plant-based milk analogues

The physicochemical properties of plant-based milk vary with the raw material used, formulation, and the treatments given in the course of production of the milk. It has been reported that most of the plant-based milk showed shear thinning fluids behavior with varying flow indexes (k = 165.4 and n = 0.671 for soymilk; k = 155.4 and n = 0.429 for oat milk; k = 8.5 and n = 0 for peanut milk; k = 83.9 and n = 0.449 for buckwheat milk) (Dąbrowski et al., 2022; De et al., 2022; Le et al., 2021; Salve et al, 2019; Sarangapany et al., 2022; Yao et al., 2022; Zabegalova et al., 2019). Jeske et al. (2019), study showed that lentil milks display Newtonian behavior. Similarly, other physicochemical properties should also be taken into account while designing the formulation and extraction method for plant-based milk. Particle size distribution, emulsifier type (thickness, charge, and hydrophobicity), and aqueous phase composition (ionic strength, pH, and polymer composition) are some of the key variables that food scientists control to modify the colloidal interactions in plant-based milk substitutes (McClements et al., 2019). The average droplet size of soymilk, oat milk, and peanut milk are 0.34 µm, 1.55 µm, and 0.29 µm. And pH ranged between 6.5 and 6.8. Peanut milk has a lower viscosity (5.52 mPa-s) than soy milk (17.09 mPa-s) because of its high-fat content, which may help proteins emulsify better and reduce the viscosity of the peanut. Milk made from buckwheat and oats appeared to be more viscous than other samples. Hence, both of these grains must be blended in ice water to prevent the formation of sticky slurries during high-intensity heat processing and to achieve the desired texture (Yao et al., 2022).

Peanut milk has a significantly high value of solid content (8.90 ± 0.61), of all the plant-based milks (Yao et al., 2022). The total solids content of milk made from beans was, higher than milk made from cereal. This might be because beans have higher fat and protein content than cereal, which helps produce emulsions and enhance the solid content. Plant-based milk had a pH value (ranging from 6.60 to 6.83), similar to bovine milk. Mäkinen et al. (2015), have studied the physicochemical properties of UHT-treated plant-based milk: soy, rice, oat, and quinoa. They found that plant-based milk analogues have varying rates of separation, with rice and oat milk being the most unstable products, with particle size ranging from 0.55 µm (soy) to 2.08 µm (quinoa), whereas bovine milk had an average size of 0.52 µm. Furthermore, when bovine, soy, and quinoa milk were acidified with glucono-δ-lactone, they produced structured gels with maximum storage moduli of 262, 187, and 105 Pa, respectively, whereas oat and rice milk produced no structured gel.

Although most plant-based milk also has a light, creamy appearance, it may differ noticeably from cow's milk in terms of color and lightness (McClements et al., 2019). For instance, due to the presence of natural pigments, milk made from oats or nuts may have a faintly brownish hue. It is helpful to know the type of pigments present in milk substitutes to either bleach or remove them because some consumers find their off-white tint to be unpleasant. Since soy milk and buckwheat milk contain a higher content of carotenoid and xanthophyll pigments than other prepared milk samples, they displayed significantly greater yellowness index (Yao et al., 2022). The methods that can be used to modify color indices are bleaching of the pigments, modifying the particle size, and identifying and removing any unwanted substances (McClements et al., 2019).

Benefits of plant-based milk analogue

A summary of the health benefits of plant-based milk analogue has been given in Fig. 3.

Fig. 3.

Health benefits of plant-based milk analogue (Akhlaghi et al., 2020; Chalupa-Krebzdak et al., 2018; Cheng et al., 2006; Ekanayaka et al., 2013; Giannetti et al., 2021; Koenig et al., 2011; Rasika et al., 2020; )

A solution to allergy and lactose intolerance related to mammalian milk

Cow milk protein can cause allergic responses in infants and toddlers. This response commonly known as Cow milk protein allergy is caused mainly by the soluble whey protein and supported by the insoluble casein protein present in cow milk (Giannetti et al., 2021). Symptoms could be observed in the gastrointestinal tract, respiratory tract, and skin. The anaphylactic shock which can lead to death is observed in 12% of cases. The only available treatment is having a diet free from animal milk. Hence, plant-based milk can be a solution to this problem (Silva et al., 2020). Lactose Intolerance is a non-immunological response to lactose and can be treated by external supplementation of lactase or by excluding lactose-containing dairy products (Silva et al., 2020).

Support the growth of probiotic

Probiotic fermentation of plant-based milk incorporates a significant number of active probiotics, improves nutritional value, flavor, texture, and stability, and ensures microbiological safety (Canaviri-Paz et al., 2021). The most prevalent probiotic strains utilized to create probiotic plant-based milk are those from the genus Lactobacillus and the genus Bifidobacterium (Rasika et al., 2020). Soy milk is known to support the growth of Bifidobacterium and Hazelnut milk for the growth of probiotic yeasts (Atalar, 2019; Mustafa et al., 2019). The probiotic viability of plant-based milk is strain-dependent and also depends on the packaging material of the product. Cui et al. (2021) comparative study proved the strain dependency of probiotic growth in soymilk yogurt. It was observed that B. animalis grew abundantly in soymilk yogurts compared to that of S. thermophilus, Lb. acidophilus and Lb. rhamnosus. Padma et al. (2019) study on rice milk beverages focused on the effect on the viability of different strains stored in Glass, High-Density Polyethylene (HDPE), and Low-Density Polyethylene (LDPE). The starter culture consisted of Lb. casei, Lb. bulgaricus, Lb. acidophilus, B. longum and S. thermophilus. It was found that except for St. thermophilus, all probiotic strains demonstrated greater viability in glass containers than they did in HDPE and LDPE containers. On the contrary, St. thermophilus displayed greater viability numbers in HDPE and LDPE than in glass containers, which is possible because these materials have higher oxygen permeability. Inclusion of Lb. plantarum in Soymilk significantly reduced renal function biomarkers and inflammatory adipokine Progranulin (PGRN) levels. This can help in renal function in nephropathy patients with type 2 diabetes (Miraghajani et al., 2019). Synbiotic approach with inulin as prebiotic and Lb. paracasei as a probiotic increased the antioxidant activity (Choudhary et al., 2019). de Carvalho et al. (2018), study on the effect of soymilk fermented with Enterococcus faecium and Bifidobacterium longum resulted in reduced weight gain and adipocytes and maintenance of fecal microbiota and immune profile of the animals suggested that it can be a good aid to obese patients.

Healthier fat source

Plant-based milks are free from cholesterol. This comes with preventive health benefits. In comparison to cow's milk, plant-based milk substitutes might contain healthier fats. α linoleic acid, an important omega-3 fatty acid, is present in hemp milk substitute at a concentration of 0.4 g per 100 mL which is 25% of the daily required consumption of 1.6 g. However, it is deficient in other crucial omega-3 fatty acids like Eicosapentaenoic Acid (EPA) and Docosahexaenoic Acid (DHA) (Chalupa-Krebzdak et al., 2018). Medium-chain triglycerides (MCT), which are present in coconut milk substitutes, have been shown to have a range of positive impacts on HDL and LDL cholesterol, body weight, waist-hip ratio, insulin levels, metabolic rate, and overall fat. Despite coconut milk's high saturated fat level, all of these health advantages are still present. The health advantages of the MCTs in coconut oil are also present in practical daily servings of coconut milk, according to a study that looked at the effects of consuming 200 mL of coconut milk, reconstituted from a market-bought coconut powder, for 2 months. Significant decreases in LDL were observed, along with increases in HDL (Ekanayaka et al., 2013).

Good source of fiber

Nuts, seeds, legumes, and cereals consist of dietary fiber (Aydar et al., 2020). In particular, dietary fiber is higher in cashew (1.15%), oat (0.8%), and soy milk (0.70%). β-glucans found in oat milk, a type of water-soluble dietary fiber, have frequently been shown to lower LDL cholesterol when ingested at or above 2.9 g per day (Hughes and Grafenauer, 2021). LDL cholesterol levels have been demonstrated to decrease with daily intake of 750 mL of an oat milk substitute with a 0.5 gm per 100 gm β-glucan content. As a result, oat milk substitutes would be a good substitute for cow's milk for people looking to reduce their LDL cholesterol.

Bioactive compounds with health benefits

Plant-based milk contains different bioactive compounds which have added health benefits. Isoflavones present in soymilk have been shown to reduce memory dysfunction in Alzheimer’s patients, increase bone mineral density, especially in the post-menopausal period, and reduction in blood pressure and insulin (Akhlaghi et al., 2020; Ko et al., 2018; Sathyapalan et al., 2018). Avenanthramides, a bioactive compound present in oat-based milk showed antiatherogenic and anti-inflammatory activity in rats (Koenig et al., 2011). The results of in vivo studies on sesamin and sesamolin, bioactive compounds found in sesame, suggested that sesamin and sesamolin exert effective neuroprotection against cerebral ischemia (Cheng et al., 2006). Cow milk is often poor in vitamin E, while almond milk has a vitamin E level of 6.33 mg per 100 g, which is 42% of the RDI of 15 mg. The antioxidant properties of vitamin E can provide several health benefits. (Chalupa-Krebzdak et al., 2018).

Limitations of plant-based milk analogue and measures taken for its elimination

Allergens and their removal

Food allergens are particular components that generate certain immunological responses leading to mild symptoms like swelling and abdominal discomfort to severe symptoms like anaphylaxis which may result in a fatal shock. Immunoglobulin E mediates a quick allergic response. Nuts and soy are the allergens of interest for plant-based milks. Allergen sources and compounds are enlisted in WHO/IUIS, Allergen Nomenclature, 2010. Studies have shown that soy, almond, and coconut milk produce the highest immunoglobulin E immune response (Vojdani et al., 2018).

The allergic activity of these allergens is dependent on their native structure. Reduction in the activity has been observed in case of loss in ordered structure for most of the allergens except the legumins in peanut and soybean. The allergic potential of gliadin and profilin reduced when they lost their 3D structure. The same was not observed for 2S albumin and Non‐specific lipid‐transfer proteins. Alkylation or reduction can reduce the allergic potential of gliadin and 2S albumin (Costa et al., 2022). Allergic potential of Cor a1 in hazelnut reduced after heat treatment. Thermal treatments like blanching, autoclaving, steaming, pasteurizing, and boiling are known to reduce the capacity of binding in immunoglobulin E (Costa et al., 2022). Changes in the secondary structure with reduction in immunoglobulin E binding capacity were observed in peanut extract treated at 138 °C for 30 min leading to loss in allergenicity (Cabanillas et al., 2012). Enzymatic treatment (protease) and cold plasma treatment (52 kHz/32 kV/118 s cycle) of peanuts have been reported to reduce Ara h1 and h2 by 95% and 65% respectively (Yu et al., 2011; Venkataratnam et al., 2020). A study by Briviba et al. (2016) showed the disappearance of protein antigen in almond milk under ultra-high-pressure homogenization (UHPH) (350 MPa at 85 ℃). Villa et al. (2020) showed that the reduction in allergic response was matrix-dependent. Hence, tests on plant-based milk followed by assays on patients with clinical allergies after the treatment of foods are necessary.

Limitation in micronutrient and protein quality; and fortification

Low protein quality in plant-based milk is evident from the lack of a complete amino acid profile. This can be resolved by using a combination of sources that complement each other's amino acid profile. Bonke et al. (2020), studied a combination of lentils, oats, and pulses pea to get a more balanced and complete amino acid profile. The final milk analogue contained significant amounts of Phenylalanine, Threonine, and Leucine, as well as moderate amounts of Isoleucine, Valine, and Methionine, as well as contributions from Histidine and Lysine.

Plant-based milk can be fortified to be a nutritionally equivalent analogue of cow's milk. PBMAs are usually fortified with calcium, vitamin D, and/or vitamin B12. Mineral salts like phosphates, citrates, and carbonates can be used for the fortification of potassium, calcium, and magnesium. According to Jeske et al. (2017), Zinc oxide, Folic Acid, Vitamin B12, Riboflavin, and Vitamin A palmitate have been used to fortify different plant-based milks. Incorporating mineral compounds does not necessarily mean an enhanced nutritional profile. The bioavailability of that compound will be the deciding factor. An example of the same would-be calcium carbonate is the preferred form of calcium fortification compared to tricalcium phosphate because of its greater bioavailability (Heaney et al., 2000; Kruger et al., 2005).

Antinutrients and their elimination

Antinutrients are compounds that interfere with the absorption of nutrients. Unprocessed seeds like cereals, legumes, and nuts contain antinutrients in high quantities (Reyes-Jurado et al., 2021). The most commonly found antinutrients are lectin present in cereals (0.5–7.3 mg per 100 g and nuts (37–144 μg per gram); trypsin inhibitor in legumes (6.7 mg/100 g); tannins in legumes (1.8–18 mg/g); phytic acids and phytates in legumes (386–714 mg/100 g), cereals (50–74 mg/100 g), and nuts (50–74 mg/100 g); saponins in legumes (106–170 mg/100 g); and oxalates in cereals and nuts—106–170 mg/100 g (Popova et al.,2019).

Germination activates genes and enzymes in the beans and causes physical and chemical changes which lead to the reduction of antinutrient factors. For instance, the phytase content increases on germination which reduces the phytic acid. The reduction % of antinutritional factors increased with an increase in germination time (Hu et al., 2022). The collapse of voids may have impacted lipoxygenase and upon the action of enzymes or physical changes, the disulfide bonds were altered thus, inactivating trypsin inhibitors (Hu et al., 2022). Trypsin inhibitor has a high thermal stability. Complete removal of trypsin inhibitor would require processing at ultra-high temperatures and long processing times which would negatively affect other nutritional and flavor properties. Hence, 20–25% residual of trypsin inhibitor in soy-based products is usually accepted (Kubo et al., 2021). During the inactivation of lipoxygenase and trypsin inhibitors, trypsin inhibitors can be the target compound because of their higher resistance (Kubo et al., 2021; Manassero et al., 2016). Complete inactivation of lipoxygenase was achieved by microwave treatment at 2450 MHz and 90 ℃ temperature for 9 min (Kubo et al., 2021). Soymilk treated with high hydrostatic pressure at 550 MPa and 60 °C with 15 mmol/L CaCl₂, 100% deactivation of trypsin inhibitor was achieved (Manassero et al., 2016)_. With radio frequency (27.12 MHz, 2.12 kW), 95.2% lipoxygenase was deactivated in 270 s. Also, urease activity was reduced to 14.1% of the original activity in 600 s and about 89.4% reduction in trypsin inhibitor activity in 300 s (Jiang et al., 2021).

Sensory characteristics

Sensory characteristic is a combination of visual appearances like color, and mouthfeel which is related to viscosity and flavor. Plant-based milk is formulated and given various treatments to imitate the homogenous, whitish appearance of bovine milk (Reyes-Jurado et al., 2021). A few of the compounds responsible for flavors in plant-based milk analogues are: benzaldehyde and nonanal in almond-based milk give a sweet taste; hexanal, 2,4-decadienal and 2-nonenal provide green off-flavor in soy-based milk; high levels of pyrazine, furans, alkanes in legume-based milk gives typical leguminous, beany and earthy notes; ketones and aldehydes in oat-based milk gives grainy, nutty, hay and grassy notes (Alasalvar et al., 2003; McGorrin, 2019; Nedele et al., 2021; Pérez-González, et al., 2015). The bitter, acidic, or astringent notes, and off-flavors generated are often resultant of such compounds. The presence of fibrous material in plant-based milk results in a gritty and sandy texture (Reyes-Jurado et al., 2021). Almond and cashew milk are known to be more liked than soy, oat, and rice because of their lack of coarse appearance and sweet taste (Reyes-Jurado et al., 2021). Almond and soy milk blend in the proportions of 50:50, 40:60, and 60:40 were tested and the 60:40 was most liked and close to bovine milk in terms of mouthfeel and taste (Kundu et al., 2018). Aldehydes, ketones, and alcohols make up the majority of the fragrance profile (nutty taste and odor) in almond-based milk analogues (Pointke et al., 2022). In addition to their nuttiness, almond-based beverages can also taste sweet, roast-like, soapy, and salty, and they may exhibit a thicker, lumpier texture (Vaikma et al., 2021). The coconut-based drinks, which had strong nutty flavors and a thicker texture, were the most comparable to almond-based beverages. Hazelnut beverages resemble both almond and coconut, however, cashew and brazil nuts possess more umami and saltiness. The majority of the constituents in oat and soy beverages belonged to the groupings of acids and alkanes (Pointke et al., 2022). Oat beverages usually had more alkane components than other plant-based beverages, and several compounds from this chemical category were linked with waxy characteristics. The Soy-based drinks are high in acid levels. A beany taste in soy-based products, a high level of bitterness, and poor textural quality, which can be brought on by a high starch content, are among the off flavors that are recognized. Additionally, volatile compounds like hexanal, hexanol, and pentanal produced during the oxidation of lipids can result in unpleasant tastes. Acids, alkanes, and organic compounds were the chemical classes that were most prevalent in soy beverages. Rice beverages seemed to have a sweet and salty flavor, but several of the items had an earthy, soapy flavor that may be described as off-flavor (Vaikma et al., 2021). Other rice and oat beverages likewise included a strong aftertaste, a sour taste, and a bitter taste. The darker color, nutty aroma, and astringent flavor of buckwheat and quinoa beverages seem to set them apart from oat and rice beverages.

Removal of unpleasant flavors

Pre-treatment of raw materials during the preparation of plant-based milk affects the sensory profile of the final milk. Table 3 summarizes the effects of pre-treatments and techniques applied to remove off-flavor on different raw materials. During storage, unpleasant flavors may develop due to the presence of unsaturated fatty acids and lipoxygenases. Soymilk often has a beany flavor which is associated with the presence of lipoxygenases and other compounds (Sethi et al., 2016). Retention of nutritional compounds and bioactive compounds could be a matter of concern after such treatments and should be checked. An Ultra-high pressure homogenization process (200 or 300 MPa at 75 °C) reduced tocopherol content in almond milk, but a similar treatment (200 MPa at 55 °C or 300 MPa at 75 °C) did not affect the amino acid profile (Ferragut et al., 2015; Toro-Funes, 2014).

Table 3.

Pretreatments and techniques applied to reduce the off flavor of plant-based milk analogues

Pre-treatments
Sr. No	Source	Pretreatment	Effect		References
1	Sesame	Roasting; Steaming (w & w/o soaking); Pressure cooker (w & w/o soaking)	Enhanced flavor Reduced bitterness and off flavor
2	Peanuts	Pressure blanching (121 °C, 15 psi, 3 min) before soaking and milling	Increased aroma score and taste acceptance		Jain et al. (2013)
3	Almonds	Hot water blanching (15 min)	High ratings for texture, flavor, color, and general acceptance

Techniques used to reduce off-flavor
Sr. No	Milk	Compounds	Description of off-flavor	Treatment/techniques	References
1	Soymilk	C-8 alcohols formed during soaking	Earthy and mushroom-like	Heat treatment under acidic conditions (pH 3–6) will transform octanols into volatile aldehydes and ketones	Feng and Hua (2022)
2	Soymilk	2,4-Decadienal, 2-pentyl furan, hexanol, hexanal, 1-octene-3-ol, 1-octene-3-one, 2-nonenal, 2,4-nonadienal along with aldehydes, ketones, furans and alcohol	Grassy, fatty, beany off flavor	High-Intensity Ultrasound-20 kHz, 400 W, 0-9 min. It was found that off flavor reduced till the treatment time was 7 min and increased after that	Mu et al. (2022)
3	Soymilk	2,4-Decadienal, 2-pentylfuran, hexanol, hexanal, 1-octene-3-ol, 1-octene-3-one, 2-nonenal, 2,4-nonadienal	Beany flavor	Radiofrequency- 27.12 MHz, 2.12 kW, 30–180 s. Hexanal and hexanol which are the main contributors to the beany flavor were significantly reduced. There wasn’t much effect on 1-octene-3-ol	Jiang et al. (2021)
4	Soymilk	Lipoxygenase, ethanol	Beany flavor	Pulsed light treatment with pretreatments like fully soaking beans, shaking, or spinning beans. Shaking beans reduced ethanol content and was most liked by the sensory panel	Alhendi et al. (2018)
5	Soymilk	Lipoxygenase	Beany Flavor	The lipoxygenase off flavor was masked by strong strawberry flavor by incorporation of strawberry pulp in different ratios (5, 10, 15, and 20%)	Verma et al. (2019)

Shelf life and stability improvement

Plant-based milk's stability depends on the size of dispersed particles like fat globules and protein, which may separate and form sediments during storage, the solubility of proteins, undisintegrated starch, and emulsion formation (Paul et al., 2020).

The type of mill used in making plant-based milk and the extraction temperature impact milk's shelf life and stability. In manufacturing pistachio milk, using a colloid mill enhanced the protein solubility and stability compared to mechanical grinders. High extraction temperatures will increase the fat extracted and denaturize the protein. This leads to a lowering of stability and shelf life. Therefore, optimum extraction temperatures should be chosen. Homogenization and thermal treatments often contribute to microbiological stability (Reyes-Jurado et al., 2021).

Enzymatic starch hydrolysis is a widely used method to increase stability in milk. For instance, the high starch content in rice milk results in poor emulsion stability, which can be resolved by hydrolyzing starch using α- and β-amylase or glucosidase (Mäkinen et al., 2015). The same method can be used to hydrolyze starch in oat milk which otherwise gelatinizes and forms highly viscous gels not liked by consumers. The use of Celluclast in soymilk has reduced its droplet size and enhanced physical stability. Adding emulsifying agents at 1% (w/w) to peanut milk resolved the problem of unstable emulsion caused due to high-fat content (Reyes-Jurado et al., 2021). The raw materials utilized, the manufacturing method, thermal and non-thermal treatments, packing, and storage temperature all affect how long plant-based milk will last on the shelf. Fresh pasteurized plant-based milk has a shelf life of 2–3 days at 4–5 °C, whether it is packaged or not. Pasteurized milk that has been aseptically packaged has a shelf life of 12–30 days at 4–5 °C. At room temperature, ultra-pasteurization and aseptic packing result in a 90–170-day shelf life.

For the preservation of Plant-based milk, heat treatment, which inactivates bacteria and enzymes, is frequently used. Emerging process methods enlisted in Table 4 have improved the stability of emulsions, eliminated enzymes that cause off flavors, decreased microbial loads, increased stability, and extended the shelf life of plant-based milk substitutes. Increased zeta potential indicates an increase in surface charges and repulsion which reduces protein aggregation and hence, it is an indicator of stability (Mukherjee et al., 2017). Ultrasound technology helps in homogenization as well as emulsification (Iorio et al., 2019).

Table 4.

Treatment techniques used to improve shelf life and stability of plant-based milk analogues

Sr. No	Milk	Treatment	Process parameters	Results	References
1	Hazelnut milk	Ultrasound	20 kHz, 100 W, 5 min and 10 min	Increased stability
2	Almond milk	Pulsed Electric Field	Electric field strength: 7, 14, 21, and 28 kV/cm for 200 µs at 1 kHz	Particle size reduction causes an increase in colloidal stability	Manzoor et al. (2019)
3	Peanut milk	Ultrasound	200, 300, and 400 W for 15 min	A TPC of 0.9 was achieved, and no phase separation after 48 h	Salve et al. (2019)
4	Peanut milk	Hydrodynamic cavitation	6, 8 and 10 bar for 15 min	At 10 bar pressure, a maximum log reduction of 1.2 was achieved, and no phase separation after 48 h	Salve et al. (2019)
5	Tiger nut milk	Ultra-High-Pressure Homogenization	200 and 300 MPa Ti = 40C	The shelf life of the milk was enhanced from 3 to 57 days at 300 MPa	Codina-Torrella et al. (2018)
6	Soymilk	Hydrocolloid addition – K carrageenan and Gum Arabic	0, 0.01,0.02,0.05% w/v; 5 weeks study	Reduction in protein aggregation, K carrageenan gave better results than Gum Arabic	Mukherjee et al. (2017)

Author Contributions

Conception and design of study: DD, SSA, KP. Acquisition of data: DD, KP, SSA. Drafting the manuscript: DD, KP. Revising the manuscript critically for important intellectual content: SSA.

Declarations

Conflict of interest

All authors have participated in the conception and design, analysis, interpretation of the data, drafting of the article or revising it critically for important intellectual content; and approval of the final version. This review article has not been submitted to, nor is under review at, another journal or other publishing venue. The authors have no affiliation with any organization with a direct or indirect financial interest in the subject matter discussed in the manuscript.