*Review* **Natural Sweeteners: The Relevance of Food Naturalness for Consumers, Food Security Aspects, Sustainability and Health Impacts**

#### **Ariana Saraiva <sup>1</sup> , Conrado Carrascosa <sup>1</sup> , Dele Raheem <sup>2</sup> , Fernando Ramos 3,4 and António Raposo 5,6,\***


Received: 28 July 2020; Accepted: 27 August 2020; Published: 28 August 2020

**Abstract:** At a moment when the population is increasingly aware and involved in what it eats, both consumers and the food sector are showing more interest in natural foods. This review work discusses, addresses and provides details of the most important aspects of consumer's perceptions of and attitudes to natural foods and in-depth research into natural sweeteners. It also includes issues about their use and development as regards health impacts, food security and sustainability. In line with our main research outcome, we can assume that consumers are very keen on choosing foods with clean labelling, natural ingredients, preferably with other functional properties, without the loss of taste. In response to such a phenomenon, the food industry offers consumers alternative natural sweeteners with the advantage of added health benefits. It is noteworthy that Nature is a superb source of desirable substances, and many have a sweet taste, and many still need to be studied. Finally, we must stress that being natural does not necessarily guarantee market success.

**Keywords:** consumer's perceptions and attitudes; food industry; food security; health impacts; natural food products; natural sweeteners; sustainability

#### **1. Introduction**

In the 20th century, developed countries resolved the lack of food security with a major contribution from agri-food industrialisation [1–3]. Food processing has played a vital role in prolonging food products' shelf life, mitigating food losses and reducing waste and in enhancing the production of nutrients and their availability [4,5]. Yet day-to-day consumer perceptions rely on other factors apart from these achievements. In modern societies, more globalised markets and more manufacturing efforts made in the food chain have led to knowledge gaps and a perceived separation between local manufacturers and citizens (e.g., how foods are produced, where they are produced, etc.) [6,7]. Consumers are gradually becoming more aware to natural ingredients, while the growing importance of naturalness among consumers has meant key implications for the food industry [8]. This could well have implications for not only developing and selling food, but also for the increase in emerging food

technologies. It is possible that those food products not perceived as natural are not accepted by lots of consumers in the majority of countries.

The demand for zero-calorie and naturally derived sweeteners has dramatically grown in the last decade because consumers are more mindful of their health [9]. For decades sweeteners have been used to make food more flavourful and to attract consumers. They were first adopted because of high-calorie sugar-to-diet ratio, and this favoured obesity in the general population, which became widespread in infants and children [10]. Thus, a low-calorie sweetener, saccharine, became available in the 1980s. As this sweetener was so popular, others followed, including cyclamates, aspartame and acesulfame K, which are the most widespread. Sweeteners have long-since been the object of controversies and conflicts over the years, which have included allegations of liver and bladder toxicity, carcinogenicity, foetus malformations, along with other dangers [11]. Whereas all these allegations have been investigated, sweeteners were considered safe [12], although some loss of consumer trust remains as some are not permitted in the USA, while others are allowed in the EU (e.g., cyclamate and cyclamic acid), but are not permitted in the USA (under E 952). Hence the need for natural substitutes is crucial [13]. Natural sweeteners and synthetic sweeteners have the same purpose: to act as a sweet flavour while fewer or no calories at all to diet. Natural sweeteners can be classified as two categories: high-potency sweeteners and bulk sweeteners. The former's potency is greater than the sweetness of one sucrose molecule. The latter's potency is the equivalent to one sucrose molecule, or less, with sucrose being the international standard for sweetness.

The main aims of this review were to provide details of and understand consumers perceptions and attitudes to natural food products, and to study in-depth natural sweeteners, mostly aspects related to their use and production in health impacts, food security and sustainability terms. Special attention was given to sweeteners, which are unanimously considered in the literature to have a good taste, high solubility and high stability, be safe with an acceptable cost-on-use, namely, erythritol glycyrrhizin, tagatose, steviol glycosides and thaumatin.

#### **2. Consumer Perceptions of and Attitudes to Natural Food Products**

Humans are inherently connected to natural objects [14], so it should come as no surprise that most humans have clearly preferred natural foods in the last few decades [8,15]. The findings of the Nielsen Global Health and Wellness Survey [16], which was conducted in 60 countries and involved 30,000 consumers, revealed that the most essential food characteristics are naturalness, freshness and minimal processing. The findings of the Kampffmeyer Food Innovation Study [17], conducted with over 4000 consumers from eight European countries, indicate that food naturalness is a "decisive buying incentive," and that approximately three quarters of respondents perceived a close "natural" + "health" link. The market research outcomes generally suggest that many consumers in developed countries usually eat natural foods. From the natural science point of view, naturalness definitely does not imply that a food product is healthier, less dangerous or tastier, although this is not how most people perceive naturalness [18,19].

Consumers perceive naturalness as a beneficial characteristic of food items. However, the relative importance of food naturalness varies across lands, cultures and throughout history [20]. Human beings have conventionally sought to monitor and reduce environmental threats. The arrival of increasingly processed foods in developed countries in the 1950s provided longer food shelf life, and better food and nutrition security [4,5]. It was at that time when consumers started showing a strong preference for processed foods. Conversely, consumers' everyday experience depends on other things apart from these accomplishments. Today, highly globalised markets and intensified food chain production in industrialized economies have added to a knowledge gap and perceived distance between food manufacturers and consumers [6,7].

Globalisation and industrialisation go hand in hand with a more man-made and higher risk that enhance citizens' perception of modernity risks [21]. In the last few decades, food safety incidents have impacted Europe, such as dioxin and bovine spongiform encephalopathy [22,23]. Consumers are concerned about excess pesticide use by traditional industrial agricultural methods [24], the employment of artificial ingredients, colourants or additives like E133 [25], and questionable food innovations, like genetically modified organisms [26], being introduced. This has made consumers suspicious or sceptical about the negative health consequences that this food system entails [3]. Growing public concern about what the food system does to climate change and its general negative impact on sustainability [27] mean that consumers now question the social and environmental consequences of food production [28,29].

Consumers generally prefer food to be nutrient-fed and satiated, while price and taste are other important factors [30,31]. However, it is often suggested that food consumption in industrialized societies is presently affected by three particular major trends: convenience, health concerns, sustainability [32]. Health concerns are driven by consumers' affluence, but are also explained by not only the rising number of food- and lifestyle-related diseases (i.e., obesity, diabetes, etc.) [7,33], but also increasing intolerances and allergies to certain components and specific food products like gluten. These aspects also motivate consumers to pay more attention to healthy food items that promote healthier lifestyles at older ages and that lower the incidence of some diseases. Sustainability concerns come over according to increasing knowledge of emissions released by traditional agricultural activities [1]. This has led to an ever-growing expansion of organic farming and markets, and can also explain why some consumers seek products, such as 'local food' products (food miles) and why they are willing to pay more for water-saving products [34]. Convenience food refers to the number of meals not eaten at home or home-delivered meals as opposed to homemade meals. This figure has significantly risen in past decades [35], which suggests that consumers are involved in additional capabilities of food items to save time (e.g., frozen foods, ready meals, microwavable food, etc.).

By analysing the factors that impact individual differences in the perceived importance of naturalness, although high mean values of the importance of naturalness in foods (INF) were found in most studies, individual differences appeared in how important naturalness was perceived [8]. For psychological factors, several works have indicated the importance of consumers' values for explaining INF. Idealism [36], tradition and universalism [37] were positively associated with INF, whereas hedonism and power correlated negatively with INF [37]. Interest shown in health correlated positively with INF [38,39], and attitudes to novel technologies, chemicals and functional foods correlated negatively [38,40–45]. Some experiments revealed a certain conceptual similarity between predictors and how INF was calculated. Attitudes to traditional and organic food, along with food involvement and neophobia were positively related to INF in a number of research works [46–51]. The research by Olbrich et al. [47], which included over 10,000 German consumers, revealed that attitudes to organic food were related to INF. Similar findings are reported in Taiwan by Hsu et al. [48]. These two experiments indicated that no difference appeared between INF and items assessing organic food preferences. Very few experiments examined the relation linking INF with personality characteristics. Steptoe et al. [39] reported a positive association for INF and the control locus, while Huotilainen et al. [52] found that the perceived INF value was not related to consumer willingness to accept food innovations.

As for consumer attitudes to food naturalness affecting their behaviour and intentions, some studies report INF measurement items overlapping measuring intentions or behaviours in relation to eating organic food [49,51,53,54]. INF influences intention to eat in a more enviro-friendly way [55] with fresh [56], local/traditional [52,57] natural [58] and low-calorie food items [59]. The results cross-country analysis obtained by Hemmerling et al. [57] were inconclusive; INF had a strong impact on local/traditional foods in Italy and Germany, but was negligible in The Netherlands, Poland, France and Switzerland. Only two experiments have investigated how INF affects consumer decisions about eating functional food, but the results were inconclusive. Kraus [60] reported substantial effects in Poland, but the research by Urala and Lähteenmäki in Finland [61] proved unsubstantial. Lähteenmäki et al. [62] also reported that INF adversely affected purchase intention to buy genetically modified cheese in Finland, Denmark, Sweden and Norway. In their works, Lusk et al. [63] saw

that INF increased the probability of selecting non-clone or non-hormone milk as opposed to the clone or hormone variety. Many research works undertaken in different countries have revealed that INF significantly affects eating healthy [64–68], organic [69] and traditional foods [46,70], and eating unhealthy [66,67] and convenience foods [71]. Roininen and Tuorila [72] and Zandstra et al. [73] found that INF did not affect unhealthy or healthy eating. Small sample sizes (*n* = 144 and *n* = 132) could potentially explain such insignificant results. With their survey of 197 Spanish consumers, Carrillo et al. [74] found that perceived naturalness of functional foods increased their sales.

Regarding natural food ingredients, which have attracted far more attention from public and food manufacturers, for a few decades now it has been worth emphasising that consumers mainly choose food without additives but, if they are not available, the same consumer chooses food containing natural additives rather than synthetic ones [11,13]. Consumer research has revealed that consumers have increasingly become more knowledgeable of food additives and more frequently prefer natural additives to their artificial analogues [40,75,76]. Unlike artificial sweeteners, which are all capable of structural modification in the hope that better tasting analogues will be discovered, natural sweeteners must be used 'as are' simply because any structural change made to a natural sweetener to improve its taste profile will automatically destroy the 'natural' proposition and position. So, although the consumer interest level is high, identifying a natural sweetener with the requisite sensory quality is no trivial undertaking [77].

#### **3. Natural Sweeteners**

Preference for sweet taste is not only innate, but universal [78]. Food products related to a sweet taste characteristically contain simple carbohydrates in the forms of fructose, glucose and sucrose, which are metabolised to produce energy rapidly, as wee as complex carbohydrates in the form of starch for long sustained energy and storage. However, the sweet taste can be induced by the presence of peptides, D-amino acids, glycosides, proteins, coumarins, ureas, substituted aromatic compounds, dihydrochalcones and other nitrogenous substances [79]. Yet all sweet-tasting compounds interact and activate a single receptor, which is expressed on the surface of taste buds, the TAS1R2-TAS1R3 heterodimer, and contains multiple binding sites to explain the range of compounds that induce perceived sweetness [80].

Honey used to be the main sweetener in human diet. However, in the 18th century, the process of extracting sucrose from sugar beet and sugar cane grew exponentially and clearly assumed preponderance. Nowadays, sucrose, or common table sugar, remains the most traditionally used sweetener, and is available in a variety of refined forms [81]. In 2018 and 2019, global sucrose consumption came to 174 million metric tons [82]. In the last few years, sugar overconsumption has become pandemic, with serious consequences in public health terms. There is clear evidence for an association between eating too much sugar and being at higher risk for dental caries, type II diabetes obesity and cardiovascular diseases, among other non-communicable diseases [83]. Given this scenario, sweeteners in food products have spread, and this product is major a target of much interest for the industrial and scientific communities.

Many synthetic sweeteners have been developed, but today demand undoubtedly lies in natural sweeteners, preferably the high-intensity kind; i.e., with low-calorie contributions. This trend stems from not only growing consumer concern about the harmful effects of a diet that includes too much sugar, but also the problems that arise from employing artificial food additives. Although many low-calorie sweeteners are readily available, only a few can be used by the food industry, mainly because of safety concerns and technological problems [81]. It is worth noting that, apart from sweetening, these compounds can influence a product's colour, flavour, texture and shelf life [84].

The most important aspects when selecting a sweetener have to do with its physico-chemical properties, such as thermal stability and solubility in water, but also production cost and safety [85,86]. Its sweetness potency is extremely relevant; that is to say, sweeteners can be classified in line with their intrinsic characteristics (nutritive value, sweetening power) and origin (synthetic, semisynthetic and

natural) [87]. Depending on their sweetness level in relation to sucrose, considered an international reference (sweetness potency = 1), sweeteners are grouped into two classes: bulk and intense. Bulk sweeteners have a similar or less sweetening potency vs. sucrose, and are used to typically confer low-calorie food products preservative action, bulk and texture [88]. They can be employed in baked food products, breakfast cereals, preserved foods, desserts, cakes, jams, ice cream and sauces [89]. To the sweeteners in this category we indicate sugar alcohols like maltitol, sorbitol, lactitol, xylitol, erythritol, mannitol, isomalt, hydrogenated starch hydrolysates and hydrogenated glucose syrups [86]. Trehalose and tagatose are two new compounds with a similar suitability to sugar alcohols [86,89]. Bulk sweeteners are frequently utilised in the food industry for the benefits they offer over sucrose in both functional aspects (e.g., lowering the freezing point of a frozen desert or of the Maillard reaction) and nutritional terms (e.g., slower assimilation). However, these sweeteners do not substantially lower the calorie value of a food [89].

The sweetness that intense sweeteners provide is much more than sucrose and in different potencies [88]. Given their high sweetness potency, very small amounts are required to accomplish the desired sweetening effect. Hence their contribution to a product's energy value is minimum, which is most advantageous [87]. Despite them often being called "artificial sweeteners", such compounds can be synthetic (e.g., saccharin, aspartame, sucralose, acesulfame-potassium, cyclamate, alitame, neotame, dulcin), semisynthetic (e.g., neohesperidine dihydrochalcone) or natural (e.g., rebaudioside and stevioside) [88]. Intense sweeteners are of widespread use in processed foods, especially carbonated and non-carbonated drinks, canned food, baked food, sweets, jellies and puddings [89].

Although an origin-based sweetener classification is not considered by authorities like the European Food Safety Authority (EFSA) or the US Food and Drug Administration (FDA), we adopt this classification because, given their growing popularity, we pay attention to natural sweeteners.

Natural sweeteners encompass wide-ranging compounds like sugars, sugar alcohols, amino acids, proteins, terpenoid glycosides and some polyphenols [84,85]. Having said that, only those that possess relevant characteristics, e.g., safety, good taste, high stability, good solubility and reasonable cost, are found on the market as widely used sweeteners [11,90]. The present review focuses on the natural sweeteners that comply with these principles.

#### *3.1. Sugars*

The relative sweetness potency of carbohydrates is consistently lower than that of sucrose (a reference compound), except for fructose, which is the sweetest natural sugar (relative sweetness = 1.43) which is abundance in fruit, agave nectar, honey and some vegetables [85]. Fructose and glucose and are the two most widely adopted monosaccharides as natural sweeteners by the food industry [85]. Fructose replaces sucrose in a range of food products thanks to its lower glycaemic index, sweetening strength and low cost, and its ability to improve overall end product quality characteristics; flavour, colour texture, and shelf-life stability [84]. High fructose syrups are widely employed, mostly those produced with corn starch by an elaborate technological process, and given their interesting texturising ability, flavour profile [85]. When ingested as large amounts, malabsorption and consequent gastrointestinal disturbances may occur, and excessive intake may lead to metabolic changes; e.g., insulin resistance, high plasma triglycerides, etc. [84,91].

In this context it is worth highlighting two new compounds: trehalose and tagatose. They have been relatively recently approved as novel food ingredients or novel foods and appear on the market as sucrose alternatives [86]. The trehalose disaccharide comprises two glucose units linked by an α-1,1-glucosidic bond and occurs naturally in plants, fungi, insects, algae bacteria and yeasts [86]. The commercial product is acquired from starch by following an enzymatic process [92] and it has a relative sweetness power of 0.43 [85]. Trehalose is well appeciated because it induces a low glycaemic response and helps to maintain dehydrated and frozen food products fresh via the stabilisation of colour, texture and flavour [86]. It reduces starch retrodegradation and does not participate in Maillard reactions. It is an ingredient frequently found in sports drinks and health bars [93]. Tagatose is a

fructose isomer found naturally-occurring in certain fruit and dairy products [90]. It is considered a prebiotic and a flavour enhancer [11]. In industrial terms, it is produced from lactose following a multistep enzymatic process, plus fractionation and purification [14]. Its sweetness potency comes close to that of sucrose, 0.92, with the advantage of being differently metabolised by contributing to fewer calories, evoking a weaker glycaemic response [86,87,90] and does not favour dental caries. As it is only partly digested, tagatose can bring about diarrhoea, abdominal discomfort and flatulence when ingested as high doses [86]. Tagatose is used to prepare energy bars, breakfast cereals, chocolate gums, caramel, ice cream, soft drinks and yogurt [11,90].

#### *3.2. Sugar Alcohols*

Sugar alcohols, or polyols, are low-digestible carbohydrates that occur naturally in fruit, vegetables, mushrooms and algae [94], and have been employed as an alternative type of sweetener in recent years. The sugar alcohols allowed by the food industry to be used as nutritive or bulk sweeteners include maltitol, mannitol, sorbitol, xylitol, erythritol, isomalt and lactitol. Some other relevant compounds that enter this category are arabitol and hydrogenated starch hydrolysates, despite not being permitted in the EU [87]. They are obtained generally by catalytic hydrogenation from the corresponding aldose sugars [84]. For certain sugar alcohols, like erythritol, methods based on fermentation or enzymatic conversion with osmophilic yeasts or fungi have been followed [11]. By the way, mannitol, sorbitol and maltitol are easily extracted from brown algae (i.e., *Laminaria* species) [95]. Sugar alcohols are frequently employed for food product reformulation purposes where, other than sweetness, texture and the bulk of sugar play a key role in sugar-free cookies, cakes, sweets, chocolate and gums [86,96]. They are applied to pharmaceutical products like throat lozenges [86]. Polyols offer two major advantages over sugar as food ingredients: (1) do not favour tooth decay because they are not fermentable by oral bacteria; (2) lower calorie content and glycaemic index, which are most interesting for diabetics [86]. Sugar alcohols also present prebiotic properties and, like fibres, contribute to healthy intestinal microbiota [97]. Sugar alcohols are universally considered safe and have no established acceptable daily intake (ADI), but should be used in line with good manufacturing practices (GMP) [84,94].

Sugar alcohols are frequently used together with intense sweeteners to mask off-flavours of the latter, while conferring the bulkiness that low-calorie sweeteners cannot [84,98]. Unlike sugars, they are not subject to Maillard reactions and leave a cooling effect in the mouth, which could be desirable for certain products, but particularly in many others (e.g., baked goods) [94]. Overall, of all the allowed sugar alcohols, the properties of xylitol, erythritol and maltitol come the closest to those of sucrose, with a relative sweetness of 0.63, 0.87 and 0.97, respectively. This is why they are the most widely used [85,99]. Unlike sucrose and glucose, sugar alcohols are not totally digested, which is why excess ingestion can lead to gastrointestinal symptoms even in healthy people, like laxative side effects, so consumers with inflammatory bowel disease should be very careful with them [97]. Notwithstanding, frequent intake seems to result in better tolerance [99]. Besides the above-cited side effect, they display no other health-related problems in association with high-potency artificial sweeteners [10]. As with all sweeteners, the safety of sugar alcohols is currently being reassessed by EFSA and new data are expected to become available at the end of 2020 [100].

#### *3.3. Terpenoid Glycosides*

Steviol glycosides are a group of sweet compounds which are extracted from the leaves of a plant native to South America called *Stevia rebaudiana* Bertoni (Asteraceae), which is currently grown in some countries in Europe and Asia. Typically, the above-mentioned glycosides represent up to 15% of the dry matter in plant leaves. Ten main ent-kaurane diterpenoid glycosides exist, and they all have the same steviol core structure. Stevioside, followed by rebaudioside A, are the two most abundant and commercially relevant ones [90]. They leave a very sweet taste in the mouth that is hundreds of times superior to sucrose, which makes them very interesting sweeteners [85]. Rebaudioside A, whose relative sweetness is 250–450, is the most appealing steviol glycoside, and offers a taste like sucrose with no off-tastes, while stevioside has a slight bitter side effect [85,101]. Seeing as the leaves of the plant cannot be utilised directly in the USA and EU, steviol glycosides are extracted with water before being redissolved and recrystallixed from a hydro-alcoholic solution [87]. Steviol glycosides are hydrolysed to steviol by colonic microbiota. Most steviol is absorbed by the intestine before reaching the liver, where it goes through a process of conjugation with glucuronic acid to produce steviol glucoronides, which are finally excreted mostly in urine [102]. Consuming steviol glycosides is safe provided it lies within the 4 mg/kg of body weight/day limit [103]. Both their calorie contribution is non-significant, so they are suitable for diabetic patients. They have also been attributed anti-inflammatory and immunomodulatory diuretic and anti-hypertensive properties [104]. As for their physico-chemical characteristics, they remain moderately stable at high temperature and may be used within a pH range from 2 to 10 [87]. Steviol glycosides have been widely used to produce confectionery, chocolates, baked goods, yoghurts, ice cream, gums, sauces, jam dairy products and drinks [11,105].

Another interesting sweetening compound is glycyrrhizin, or glycyrrhizic acid, which is isolated from *Glycyrrhiza glabra* L. roots (Fabaceae), which is a liquorice plant [90]. This molecule offers a relative sweetness of 90 [85]. Its use as a sweetener is permitted in Japan and other countries, but not in the USA and EU [85], where glycyrrhizin is approved only as a surfactant and flavouring agent, and also in the form of ammoniated glycyrrhizin, which is considered Generally Recognised as Safe (GRAS) [106]. Glycyrrhizin intake should never exceed 100 mg/day, considering all its sources in the diet, given the risk of toxic effects: hypertension and hypokalaemia-induced secondary disorders [107,108]. Some authors indicate that glycyrrhizin could have beneficial effects on intestinal microbiota [97]. The applications of glycyrrhizin as foaming agent and flavour enhancer include baked goods, ice cream, confectionery, gums and beverages [99].

#### *3.4. Proteins*

Sweet-tasting proteins are naturally-occurring in some exotic plants, and their sweetness is hundreds to thousands of times superior to sucrose [85]. Thaumatin comprises a mixture of six closely interrelated proteins, thaumatin I, II, III, a, b and c, extracted from *Thaumatococcus daniellii* Benth fruit (Marantaceae), which is native to western Africa. Thaumatin I and II are the main forms, despite all isoforms being sweet-tasting [90]. No unanimous value for its sweetening potency exists, but it is estimated to be about 1600–3000-fold higher than sucrose [11,85]. Extraction is performed by water and mechanical methods [87]. Current thaumatin production does not meet demand, and alternative methods to produce it through microorganisms and transgenic plants are growing [109]. Thaumatin is permitted in both the EU and the USA, where it is GRAS [97]. Owing to lack of toxicity, its ADI is still to be established [110]. There is, however, a risk of allergic reactions [111]. The metabolism of thaumatin is similar to that of any other protein in human diet. Its energy input is 4 kcal/g, which is negligible as minor amounts are used in practice [101]. The main problems with its use are late onset of action and a slight liquorice off-taste, which may interfere with consumer acceptance. Hence its use in large amounts is not recommended. Nevertheless, it works extremely well when employed in conjunction with other sweeteners to diminish bitterness and confer foods an umami flavour [90]. As regards physico-chemical properties, it is highly soluble in water, and well resists high temperature and acidic pH [87]. Thaumatin is frequently employed in processed vegetables, sauces, soups, poultry, products deriving from egg, gums and fruit juice [87].

Several other sweet proteins are known, with the most promising ones being brazzein, mabinlin, monelin, miraculin, pentadine, curculin (neoculin) and lysozyme, but more studies are necessary to ensure their safety and applicability [90].

#### **4. Production of Safe Enviro-Friendly Natural Sweeteners**

Natural sweetener production must remain safe with no adverse environmental consequences. It is presently necessary to guarantee that our food system does not pose health problems to consumers and our planet, as reflected in the recent 'EU green deal' [112]. This section summarises the production

of the following sweeteners, which are highlighted in preceding sections given their relevance for industrial food processing.

#### *4.1. Erythritol*

Industrial erythritol production has gained prominence with the rapid development of the electrochemical process. During this process, erythrose and erythritol are produced by the electrolytic decarboxylation of arabinoic or ribonic acid. The substrates for the reaction are obtained by the decarboxylation of C-6 sugars [113]. However, a more natural method involves the biotechnological process, which results in higher yields from fermenting a sugar source. Erythritol derives from fermentation processes, conducted mostly by fungi or synthesized by lactic acid bacteria. In order to produce erythritol, the common pathways among east-like fungi genera include: *Trigonopsis*, *Candida*, *Pichia*, *Moniliella*, *Yarrowia*, *Pseudozyma*, *Trichosporonoides*, *Aureobasidium*, *Trichoderma* [114]. For industrial production purposes, *Yarrowia lipolytica*, *Moniliella pollinis* and *Trichosporonoides megachiliensis* are reported as effective [113]. One main part of the production process involves separation and purification steps because they are crucial when erythritol is taken as a food additive. A patent describes that to recover erythritol from the culture medium, separation from fermenting microorganisms is required, followed by ion exchange chromatography and crystallization. Moreover, a chromatographic separation step was subject to activated-carbon treatment in order to recover the erythritol fraction [115].

#### *4.2. Tagatose*

As a natural sweetener, biotechnological tagatose production by enzymatic isomerisation is a preferred alternative to chemical processes. For biological D-tagatose manufacturing, several biocatalyst sources can be resorted to; e.g., L-arabinose isomerase (l-AI) EC 5.3.1.4, which can catalyse the conversion of D-galactose into D-tagatose, and also for converting L-arabinose into L-ribulose, due to the similar configurations of substrates [116,117] yet biological D-tagatose production is limited given the less bioconversion efficiency of l-AI, a metal ion requirement, and the poor thermostability and low affinity of the enzyme for D-galactose. It has, thus, been suggested that applying protein engineering and genomic tools may enhance the bioconversion efficiency for D-tagatose production by amending the functional properties of l-AI [118]. Applying high-throughput screening or a selection method helps to evaluate individual protein variants and, hence, increase the possibility of screening specific mutants with greater catalytic activity. During D-tagatose production, the safety problem caused by enzyme or cells of not GRAS hosts can be overcome by transferring the gene of L-arabinose isomerase to GRAS hosts like *C. Glutamicum*, *Corynebacterium ammonagenes* and *Bacillus megaterium* [119]. Ultimately, more research needs to be conducted to explore new sources of biocatalysts from GRAS microorganisms, apart from enzyme secretion and expression in a food-grade microbial host.

#### *4.3. Steviol Glycosides*

The raw materials employed in the manufacturing process of Steviol glycosides preparations are crushed leaves from the perennial shrub *Stevia rebaudiana* (Bertoni) Bertoni of the family Asteraceae (Compositae). The literature indicates several alcohols and ion exchange resins used during the manufacturing process [120]. Extracting glycosides from stevia leaves involves thermal extraction and maceration. Both the quality and yield of the extracted products can increase by following techniques like supercritical fluids, ultrasonic waves and microwaving [121]. Besides, a multistage membrane process, which has been developed to concentrate glycoside sweeteners, is also highlighted in the report, with bitter-tasting components from the sweetener concentrate being washed out during the nanofiltration process.

The conventional extraction processes described in the literature often follow a similar methodology, whereby stevia leaves are extracted with hot water or alcohols. In certain cases, leaves are pre-treated with non-polar solvents (e.g., hexane or chloroform hexane) to eliminate lipids, essential oils, chlorophyll and other non-polar substances. With this pre-treatment, extracts are clarified by

precipitation with either salt or alkaline solutions, and then finally concentrated and redissolved in methanol for the crystallisation of glycosides [122]. Other extraction procedure steps involved are described in [123], where stevia leaves were soaked in warm water to dissolve glycosides before the precipitation and filtration of the resultant solution, followed by concentration by evaporation, ion exchange purification, spray drying and crystallisation to produce a white powder and crystals. Rao et al. [124] applied ultra- and nano-filtration membranes to develop a simple eco-friendly and low-cost process to isolate steviol glycosides, which resulted in the final product's improved taste profile.

#### *4.4. Glycyrrhizin*

The methods followed to prepare glycyrrhizic acid (GA) from liquorice roots have been investigated by several researchers. The literature reports a number of procedures as regards solvent extraction by various organic solvents, purification by ion exchange and polymeric resins, chromatographic separation, adsorption, foam separation, supercritical fluid extraction, microwave-assisted extraction (MAE) and multistage counter-current extraction (MCE) to extract GA [125]. Most existing processes to extract and purify the sweet ingredients from liquorice roots involve several steps and large quantities of solvents and chemicals. Extracting GA from roots includes extraction with hot water at ambient pressure in the presence of a number of additives, such as alkalis, as well as mineral acids, ethyl alcohol like aqueous ammonia, methanol and ethanol, which are the most well-accepted technologies. The primary aqueous extract from liquorice roots contains GA and many other water-soluble substances, which are then subjected to further process more purified products. Pure GA is also prepared from liquorice roots using alcohol as the extraction solvent in an ultrasonic device, followed by purification [126]. The conventional solvent extraction technique followed to extract GA from liquorice offers several disadvantages, namely considerable solvent requirements, longer extraction time, lower yields and higher extraction temperature. All this requires developing an effective economical extraction method [126]. The purification procedure involves the acidification of the extract by adding acids like H2SO<sup>4</sup> or HCl acids to form the solid product of GA salt (at pH 1–2). Ultrasound assisted extraction has shown that the extraction rate rises due to cavitation because the developed cavity grows in size and then abruptly collapses with the release of energy at an enormous rate, which thus increases the local temperature and pressure [127]. Therefore, greater solvent penetration in cellular materials takes place, which improves the cell content release in the bulk medium [127].

#### *4.5. Thaumatin*

The thaumatin production process can be strongly affected depending on the quality and availability of source materials [128]. In order to achieve stabler protein production that meets its demand, a series of studies were conducted, which involved thaumatin production with genetically engineered microorganisms and transgenic plants (see the studies by [129,130]).

Although using a plant system offers some advantages over microbial systems in terms of its scalability, safety and economy, they still lack some benefits that can be obtained from microbial hosts, such as the possibility of controling growth conditions and product consistencies from batch to batch [128]. Biochemical production methods have been considered because the natural production of these proteins is normally too expensive. Recombinant DNA technology is applied to produce sweet proteins in a host organism. The most promising host known is the methylotrophic yeast *Pichia pastoris*. This yeast has a tight regulated methanol-induced promoter that well controls recombinant protein production [128]. Despite thaumatin having been studied by several researchers in the last 30 years, there is still much to be done to improve its production by biochemical routes. As the literature evidences, biological products are emerging as a promising applicant in the food industry, hence the huge potential for future research to centre on using advanced computational techniques to optimise thaumatin bioproduction.

Other natural sweeteners with enviro-friendly production methods that are becoming popular food ingredients for health-conscious consumers are briefly described below:


Consumers are eager to purchase products with natural ingredients and clean labels, preferably with further functional properties, but which do not compromise taste. In order to achieve this trend, food industries are now willing to reformulate their food products to include alternative natural sweeteners to sugar.

#### **5. Health Impacts**

For natural sweeteners are deemed suitable to be extensively used and marketed, they must be safe, offer good flavour with a high degree of solubility and a good level of stability, and offer reasonable cost-effective applications [135]. This paper only investigates the natural sweeteners that meet all these criteria [11] in relation to their health impacts.

The two major compounds of bulk sweeteners are erythritol and tagatose. Erythritol is allowed in both the USA and the EU, but there are restrictions on use in drinks with the latter. As a bulk sweetener, it has approximately 65% of sucrose sweetness, but it does not lead to tooth decay and is neither toxic nor carcinogenic for the amounts added to food. The main products for which erythritol is employed are baked goods, frostings, coatings, chocolate, fermented milk, low-calorie beverages, chewing gums, sweets, among others [129,135]. In 2014, a scientific panel, mandated by EFSA, ruled out its laxative properties and declared it safe for use without defining its acceptable daily intake (ADI) [136]. Based on acute toxicity investigations, and following oral administration, erythritol is graded as being essentially non-toxic. Subchronic research further enhances erythritol's safety. Chronic research (up to 2 years) has demonstrated that erythritol has no effect on either survival or carcinogenicity [137,138]. At high doses (up to 16 g/kg body weight), erythritol affects neither the reproductive capacity nor fertility of parental rats. No adverse effects have been observed on developing foetuses [137–140]. Erythritol has no mutagenic potential, as observed in the Ames and chromosomal aberration tests [137,138,141,142]. Animal toxicity tests and human clinical trials have reliably shown that erythritol is safe. Erythritol has never been predicted to have adverse effects when applied for its intended use in food [137,138].

Erythritol has been found to reduce the risk of caries in several trials [143–147]. As erythritol does not affect insulin levels or glucose, it is an appropriate sugar substitute in diabetes patients, and also for individuals who wish or need to regulate their blood sugar levels because of prediabetes or compromised carbohydrate metabolism [148,149]. Diabetes patients can benefit from the vascular effects of erythritol, as mentioned above. It is assumed that endothelium is not compromised by erythritol in non-diabetic subjects, but in diabetic subjects where endothelium is under diabetic stress, erythritol can transfer a range of damage and dysfunction parameters to a safer side, as *in vitro*, *ex vivo* and *in vivo* studies report [148,150,151]. Erythritol can also be regarded as a substance with a beneficial impact on the endothelium under high-glucose conditions by contributing to avoid or delay the onset of diabetic complications [152]. The erythritol attribute has minor effects on several targets and can also prove beneficial. A compound with a strong biological effect is not as appropriate for chronic supplementation as required in diabetes. The option would be to use a substance like erythritol with moderate protective effects. Erythritol is not only valuable, but should be considered a recommended sugar replacement for the rapidly increasing numbers of people with diabetes or prediabetes to reduce their chances of developing diabetic complications [152,153].

Tagatose comes in very small amounts in fruit and heat-treated dairy products. Its potency vs. sucrose is 92 %, which means that it comes close in taste, but only adds 1.5 kcal/g, which makes it safe for diabetics to use without harming teeth. Tagatose is approved in the US as a GRAS compound, is permitted in the EU as a food ingredient and in many other countries with practically no toxicity associated with its use. Tagatose uses in the food industry include yoghurts, frostings, cereals, beverages, chewing gum, fudge, caramel, fondant, chocolate and ice cream [129,135,154].

Tagatose's safety and toxicity dimensions have been explored in animal and human subjects [118]. As tagatose intake increases above 10%, adverse reactions (increased liver weight and hypertrophy) have been reported in rats [155]. Consequently, the 5% tagatose level is a known safe dose that has no side effects. at reproductive performance is not impaired, even when tagatose intake is as high as 20 g/kg body weight/day [156]. Human clinical experiments to study D-tagatose use have been based mainly on its gastrointestinal and urecaemic consequences. High plasma uric acid levels are associated with purine metabolism disorder and gout development. A significant rise in the plasma uric acid concentration occurs in both the healthy and non-insulin-dependent diabetes mellitus populations after a single oral 75 g dose of D-tagatose [157]. A lower D-tagatose dose (45 g/day; 15 g, 3 times/day [TID]) is considered safe for healthy human subjects because it has no adverse effects on glycogen levels, plasma uric acid, and liver function [158]. An intake of 45 g D-tagatose/day (15 g TID) for 1 year does not induce any adverse effects on plasma uric acid levels in patients with non-insulin-dependent diabetes mellitus [159]. The above D-tagatose dose also tends to reduce postprandial plasma glucose levels. However, very few records suggest any gastrointestinal problems (nauseas, mild to severe flatulence and diarrhoea) following the intake of 30 g of D-tagatose as a single dose [160]. Given the above considerations, the "No Observed Adverse Effect Level" (NOAEL) for tagatose is set at 45 g/day or 0.75 g/kg body weight/day [161].

Regarding high-potency sweeteners, steviol glycosides (E 960) [162] are a good example of natural compounds disseminated widely worldwide. Steviosides have been used in large quantities in Japan for more than 20 years and have no documented side effects. Stevia safety is also responsible mostly for the low-absorption steviol glycosides in both humans and rats in stomach and upper intestine [121].

The use and safety of steviol glycosides has been reviewed and evaluated worldwide by a range of scientific bodies and regulatory organisation. High-purity extracts of stevia leaves have been approved for use in food and beverages by over 150 countries and regions [163]. During its 69th meeting, the Joint FAO/WHO Expert Committee on Food Additives (JECFA) set an ADI of 4 mg/kg bw/day for steviol glycosides in 2008, expressed as steviol equivalents. JECFA reaffirmed this ADI during its 82nd meeting in 2016 [164]. The Food Standards Australia New Zealand [165] and EFSA [166] have defined an ADI of 4 mg/kg bw/day for steviol glycosides (expressed as steviol equivalents). Stevia mutagenicity has been studied in many trials, although they gave contradictory results. For example, two studies concluded that, in certain assays, stevia demonstrated a dose-dependent mutagenic effect, but the same studies also concluded that stevioside is non-mutagenic [167,168]. Several other findings indicate that

the plant lacks mutagenic effects [169,170]. Despite reports not being harmonious, the FDA continues to monitor this herb as a sugar replacement, while other findings reveal that steviol and stevioside do not interfere with DNA and have no genotoxicity [171]. Mizushina et al. [172] suggested that stevioside is not involved in bladder carcinogenesis. Up to 2500 mg/kg body weight/day has been safely used in rats and enabled their normal growth and reproduction [173]. After 14 consecutive days of administering steviosides as part of acute toxicity trials, no histopathologicity, no lethality and no morphological modifications were recorded in rodents [174]. In another study, the oral administration of an aqueous extract taken from stevia leaves (up to 10%) revealed no adverse effects on female rat fertility and no teratogenic effects [175]. It has also been shown that both stevia and stevioside are safe when used as sweeteners. This is appropriate for diabetic and phenylketonuric patients, and also for obese individuals who wish to lose weight and to remove sucrose from their diet. After intake, no allergic reactions or toxicity were reported [176]. In the long term, randomised, double-blind, placebo-controlled trials indicate using steviol glycosides as a sweetener with no toxic effects for humans [177]. Stevia's safety has also been confirmed by recent studies, which demonstrated that steviol glycosides are not mutagenic, carcinogenic or teratogenic, they and do not cause toxicity [178,179]. Recently, following oral administration, a toxicological stevia leaf ethanolic extract evaluation revealed no harmful effects on subchronic oral toxicity and genotoxicity. The authors proposed that stevia leaves have the potential to be considered functional food and a nutritional supplement, rather than sweetener [180].

Another high-potency sweetener is glycyrrhizin (E 958) [181]. This compound, which is also known as glycyrrhizic acid, can act as a sweetener with a potency 50-fold sweeter than sucrose, but is also employed as a foaming agent and a flavour enhancer. This substance is legally used in both the US and EU as mono-ammonium glycyrrhizinate and ammoniated glycyrrhizin. Glycyrrhizin has antiviral, anticancer antioxidant, anti-inflammatory and hepatoprotective effects [97], but also has potential hypertensive effects and an intense aftertaste [182]. In the gut, glycyrrhizin is de-glycosylated to glycyrrhetic acid (a major product) by *Eubacterium* spp. *Bacteroides* J-37 and to 18β glycyrrhetic acid 3-*O*-monoglucuronide (the minor product) by *Bacteroides* J-37 and *Streptococcus* LJ-22. *Eubacterium* spp. can be used to convert 18β-glycyrrhetic acid 3-*O*-monoglucuronide into glycyrrhetic acid [182–184]. These glycyrrhizin metabolites (particularly 18β-glycyrrhetinic acid) are significant anti-tumour cytotoxic agents with potent inhibitory effects on anti-platelet aggregation activity and rotavirus infection [185]. Some results indicate that the glycyrrhizin/intestinal microbiota interaction has beneficial effects on hosts [183,184,186].

Thaumatin (E 957), a mixture of five proteins (thaumatin I, I, III, a, b), is also employed as a sweetener in many countries. If we consider its health effects, thaumatin does not induce tooth decay and is suitable for diabetics, as opposed to artificial sweeteners [187]. The metabolism of this sweetener is the equivalent to other dietary proteins. The research work by Hsu et al. [188] demonstrates that thaumatin is digested more quickly than egg albumin. Moreover, several studies addressing thaumatin safety aspects indicate that this sweetener induces neither toxicity nor allergenicity [128]. Some studies have evaluated thaumatin toxicity; e.g., the Joint FAO/WHO Expert Committee on Food Additives, Food and Agriculture Organization of the United Nations and World Health Organization [189] study reveals that protein is void of toxic, genotoxic or teratogenic effects. Several studies offer compelling evidence that thaumatin is not an allergen to either oral mucosa or other treatment-associated allergic effects [128]. Higginbotham et al. [190] also state that thaumatin has no harmful impact when employed as a flavour additive or a partial sweetener at a particular intake level. This protein's safety has been evaluated by the Scientific Committee for Food of the European Commission (SCF) and JECFA, which concluded that it should be listed as an acceptable ingredient [110]. This sweet protein has been approved in the European Union since 1984 (E957) according to Annex II of Regulation (EC) No. 1333/2008 [110] and maintains its GRAS status in the USA. It was licensed for use in pharmaceuticals and food in the UK in 1983, except for baby food. It is an approved high-intensity sweetener and flavour enhancer in most other countries [191]. The Panel on Additives and Products or Substances

used in Animal Feed (FEEDAP) [192] also indicates the safety of this protein in animals and its use is permitted as an additive from 1 to 5 mg/kg. Thaumatin is also employed as a sweetener in some foodstuffs like ice cream and sweets at the permitted 50 mg/kg dose. In dairy products and soft drinks, it is primarily utilised as a flavour enhancer within the range from 0.5 mg/L and 5 mg/kg [89].

There are a few other natural sweeteners that can be used in the future, but are not actually found in foodstuffs. Some examples of these substances are brazzein and monatin, which is attributable to their rarity and low yield when isolated from plant matrices.

Table 1 lists the main attributes of natural sweeteners for their use by taking into account health impacts.


**Table 1.** The main attributes of natural sweeteners for their use by taking into account health impacts.

#### **6. Conclusions**

Society is becoming increasingly aware of the utmost importance of eating a balanced diet to maintain and promote health. Excess sugar consumption is now a cross-cutting concern, but this habit is not an easy one to break, so sugar-free or low-sugar foods and drinks are in great demand and the sweetening agents that make them feasible are high-value ingredients. Today the food industry applies bulk and intense sweeteners, which are mainly synthetic in origin, to substitute sugar (sucrose). Consumers are all the more eager to eat products with natural ingredients and clean labels, preferably with other functional properties, and that do not compromise taste. To achieve this trend, the food industry now has alternative natural sweeteners at its disposal, like high-fructose corn syrup, sugar alcohols (polyols) and, quite recently, steviol glycosides tagatose and thaumatin, which offer consumers the advantage of additional health benefits. Nature is an incredible source of valuable compounds, including those with a sweet taste, of which many have not yet been explored. Nevertheless, it must be emphasised that being natural does not ensure their success on the market. It should also be noted that a long traditional use in some restricted societies and areas around the globe, and this despite providing some reassurance, cannot rule out the need to conduct detailed scientific studies to prove the safety of the natural compounds to be used as food additives and, for example, as sweeteners. The food industry needs to face the challenge of developing new products with natural functional sweeteners to continue innovating and satisfying consumers. Finally, although compounds like glycyrrhizin, an approved flavour enhancer, are not used as a sweetener, can play a relevant role in improving product characteristics, such as flavour, and need to be considered by industry.

**Author Contributions:** Conceptualization, A.S., C.C., D.R., F.R. and A.R.; methodology, A.S., C.C., D.R., F.R. and A.R.; software, A.S., C.C., D.R., F.R. and A.R.; validation, A.S., C.C., D.R., F.R. and A.R.; formal analysis, A.S., C.C., D.R., F.R. and A.R.; investigation, A.S., C.C., D.R., F.R. and A.R.; resources, A.S., C.C., D.R., F.R. and A.R.; data curation, A.S., C.C., D.R., F.R. and A.R.; writing—original draft preparation, A.S., C.C., D.R., F.R. and A.R.; writing—review and editing, A.S., C.C., D.R., F.R. and A.R.; visualization, A.S., C.C., D.R., F.R. and A.R.; supervision, A.S., C.C., D.R., F.R. and A.R.; project administration, A.S., C.C., D.R., F.R. and A.R.; funding acquisition, A.S., C.C., D.R., F.R. and A.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors are very grateful to their families and friends for all the support they provided. **Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Review* **Microbial Biofilms in the Food Industry—A Comprehensive Review**

**Conrado Carrascosa 1,\*, Dele Raheem <sup>2</sup> , Fernando Ramos 3,4 , Ariana Saraiva <sup>1</sup> and António Raposo 5,\***


**Abstract:** Biofilms, present as microorganisms and surviving on surfaces, can increase food crosscontamination, leading to changes in the food industry's cleaning and disinfection dynamics. Biofilm is an association of microorganisms that is irreversibly linked with a surface, contained in an extracellular polymeric substance matrix, which poses a formidable challenge for food industries. To avoid biofilms from forming, and to eliminate them from reversible attachment and irreversible stages, where attached microorganisms improve surface adhesion, a strong disinfectant is required to eliminate bacterial attachments. This review paper tackles biofilm problems from all perspectives, including biofilm-forming pathogens in the food industry, disinfectant resistance of biofilm, and identification methods. As biofilms are largely responsible for food spoilage and outbreaks, they are also considered responsible for damage to food processing equipment. Hence the need to gain good knowledge about all of the factors favouring their development or growth, such as the attachment surface, food matrix components, environmental conditions, the bacterial cells involved, and electrostatic charging of surfaces. Overall, this review study shows the real threat of biofilms in the food industry due to the resistance of disinfectants and the mechanisms developed for their survival, including the intercellular signalling system, the cyclic nucleotide second messenger, and biofilm-associated proteins.

**Keywords:** biofilms; food industry; food microbiology; food safety

#### **1. Introduction**

Typically, bacteria bind to surfaces and form spatially structured communities inside a self-produced matrix, which consist of extracellular polymeric substances (EPS) known as biofilms [1,2]. Biofilms imply major challenges for the food industry because they allow bacteria to bind to a range of surfaces, including rubber, polypropylene, plastic, glass, stainless steel, and even food products, within just a few minutes, which is followed by mature biofilms developing within a few days (or even hours) [3].

Since ancient times, this sessile life form has been followed as an excellent survival technique for microorganisms, given the protective barrier generated and physiological changes made by the biofilm matrix, while it fights against the adverse environmental circumstances faced typically by bacteria in man-made and natural settings, even in foodprocessing facilities [4,5]. Hence, biofilms are believed responsible for damaged equipment, more expensive energy costs, outbreaks, and food spoilage [5–8]. Biofilms have become more robust to disinfections in many wide-ranging food industries, such as processing

**Citation:** Carrascosa, C.; Raheem, D.; Ramos, F.; Saraiva, A.; Raposo, A. Microbial Biofilms in the Food Industry—A Comprehensive Review. *Int. J. Environ. Res. Public Health* **2021**, *18*, 2014. https://doi.org/10.3390/ ijerph18042014

Academic Editor: Paul B. Tchounwou

Received: 26 December 2020 Accepted: 7 February 2021 Published: 19 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

seafood, brewing, dairy processing, and meat and poultry processing [9] There is compelling evidence for biofilm lifestyle making them more resilient to antimicrobial agents, particularly compared to planktonic cells (Figures 1 and 2). This entails having to remove them from surfaces of food processing plants, which poses a massive task [10–12]. as to whether residual microbial species are found at a suitable level, and if harmful microorganisms are removed. The obtained results will allow criteria to be set, such as how to clean surfaces and food product quality [13,14].

ment, more expensive energy costs, outbreaks, and food spoilage [5–8]. Biofilms have become more robust to disinfections in many wide-ranging food industries, such as processing seafood, brewing, dairy processing, and meat and poultry processing [9] There is compelling evidence for biofilm lifestyle making them more resilient to antimicrobial agents, particularly compared to planktonic cells (Figures 1 and 2). This entails having to

Microbiological surface management is relevant for assessing and making decisions

*Int. J. Environ. Res. Public Health* **2021**, *18*, x FOR PEER REVIEW 2 of 32

**Figure 1.** Biofilm formation and development stages. **Figure 1.** Biofilm formation and development stages.

Sensory tests that involve visually inspecting surfaces with good lighting, smelling unpleasant odours, and feeling encrusted or greasy surfaces are run as a process regulation to instantly overcome visible sanitation defects, while microbiological evaluations are Microbiological surface management is relevant for assessing and making decisions as to whether residual microbial species are found at a suitable level, and if harmful microorganisms are removed. The obtained results will allow criteria to be set, such as how to clean surfaces and food product quality [13,14].

often made to guarantee consistency with microbial standards and to make improvements to sanitation procedures [15]. The fact that visual inspection cannot coincide with bacterial counts has been well-documented [16]. The hygienic conditions of food-contact surfaces must be properly examined for all of the above-cited purposes. Lack of convergence between the various approaches followed to detect and quantify biofilms does, however, make it more difficult for the food industry to locate the most effective ones [17]. The Hazard Analysis and Critical Control Points (HACCP) system and good manufacturing practices have been developed to regulate food safety and quality. Bacterial biofilms are not directly mentioned in the HACCP system employed on food processing facilities. Hence, an updated HACCP system that contemplates evaluating biofilms in food environments, and establishes an apt sanitation plan, is expected to provide much clearer contamination information, and to facilitate production in the food industry's biofilm-free processing systems [18]. The importance and impact of biofilms on the food industry have become clear in several works where the cross-contamination is common among these food products, with a wide range of pathogens, including *Listeria monocytogenes*, *Yersinia enterocolitica*, *Campylobacter jejuni, Salmonella* spp., *Staphylococcus* spp., *Bacillus cereus*, and *Echerichia coli* O157:H7 [19]. Sensory tests that involve visually inspecting surfaces with good lighting, smelling unpleasant odours, and feeling encrusted or greasy surfaces are run as a process regulation to instantly overcome visible sanitation defects, while microbiological evaluations are often made to guarantee consistency with microbial standards and to make improvements to sanitation procedures [15]. The fact that visual inspection cannot coincide with bacterial counts has been well-documented [16]. The hygienic conditions of food-contact surfaces must be properly examined for all of the above-cited purposes. Lack of convergence between the various approaches followed to detect and quantify biofilms does, however, make it more difficult for the food industry to locate the most effective ones [17]. The Hazard Analysis and Critical Control Points (HACCP) system and good manufacturing practices have been developed to regulate food safety and quality. Bacterial biofilms are not directly mentioned in the HACCP system employed on food processing facilities. Hence, an updated HACCP system that contemplates evaluating biofilms in food environments, and establishes an apt sanitation plan, is expected to provide much clearer contamination information, and to facilitate production in the food industry's biofilm-free processing systems [18]. The importance and impact of biofilms on the food industry have become clear in several works where the cross-contamination is common among these food products, with a wide range of pathogens, including *Listeria monocytogenes*, *Yersinia enterocolitica*, *Campylobacter jejuni, Salmonella* spp., *Staphylococcus* spp., *Bacillus cereus*, and *Echerichia coli* O157:H7 [19].

**Figure 2.** Biofilm structure in the growth and maturation stages [20]. **Figure 2.** Biofilm structure in the growth and maturation stages [20].

The main objectives of this review were to identify the most important biofilm examples in the food industry and to present methods to visualise in situ biofilm production, how to avoid this production, and methods to remove biofilms. This study focused on the microbial biofilms that affect the food industry and provides an overview of their importance in cross-contamination when food comes into contact with surfaces. Although going into detail in each discipline, specific to microbiology for biofilm isolation and identification, is not the object of this work, it contributes new knowledge about techniques to The main objectives of this review were to identify the most important biofilm examples in the food industry and to present methods to visualise in situ biofilm production, how to avoid this production, and methods to remove biofilms. This study focused on the microbial biofilms that affect the food industry and provides an overview of their importance in cross-contamination when food comes into contact with surfaces. Although going into detail in each discipline, specific to microbiology for biofilm isolation and identification, is not the object of this work, it contributes new knowledge about techniques to control and eradicate biofilms in the food industry from food safety and quality perspectives.

#### control and eradicate biofilms in the food industry from food safety and quality perspec-**2. Biofilm Development in Food Processing Environments**

tives. **2. Biofilm Development in Food Processing Environments**  Modern food processing lines are a suitable environment for biofilms to form on food contact surfaces, primarily due to manufacturing plants' complexity, long production periods, mass product generation, and large biofilm growth areas [21]. Many food-borne bacteria may, therefore, bind to the contact surfaces present in these areas, which could Modern food processing lines are a suitable environment for biofilms to form on food contact surfaces, primarily due to manufacturing plants' complexity, long production periods, mass product generation, and large biofilm growth areas [21]. Many food-borne bacteria may, therefore, bind to the contact surfaces present in these areas, which could contribute to increase the risk of bacterial food-borne diseases. By way of example, 80% of bacterial infections in the USA are believed to be related specifically to food-borne pathogens in biofilms [9].

contribute to increase the risk of bacterial food-borne diseases. By way of example, 80% of bacterial infections in the USA are believed to be related specifically to food-borne patho-Mixed-species biofilm production is extremely dynamic and depends on the attachment surface's characteristics [22], food matrix components [23], environmental conditions [24], and involved bacterial cells [8,25].

gens in biofilms [9]. Mixed-species biofilm production is extremely dynamic and depends on the attachment surface's characteristics [22], food matrix components [23], environmental conditions [24], and involved bacterial cells [8,25]. Attachment surface properties, such as hydrophobicity, electrostatic charging, interface roughness, and topography impact biofilm formation and, thus, affect the overall hygiene status of the surface [22,26]. Nevertheless, the precise consequence of some parameters vastly varies under specific laboratory conditions. Some experiments have revealed that bacterial attachment is more likely to happen on rougher surfaces [22,27], while others have found no association between roughness and bacterial attachment [28]. Hydrophobic surfaces tend to attract more bacteria, but studies that have tested the hydrophobicity effect present opposing results [29,30], and other experiments indicate that hydrophilic surfaces enable more bacterial adherence than hydrophobic equivalents [27,28]. The fact that clear results are lacking might lie in the various methods and bacterial strains Attachment surface properties, such as hydrophobicity, electrostatic charging, interface roughness, and topography impact biofilm formation and, thus, affect the overall hygiene status of the surface [22,26]. Nevertheless, the precise consequence of some parameters vastly varies under specific laboratory conditions. Some experiments have revealed that bacterial attachment is more likely to happen on rougher surfaces [22,27], while others have found no association between roughness and bacterial attachment [28]. Hydrophobic surfaces tend to attract more bacteria, but studies that have tested the hydrophobicity effect present opposing results [29,30], and other experiments indicate that hydrophilic surfaces enable more bacterial adherence than hydrophobic equivalents [27,28]. The fact that clear results are lacking might lie in the various methods and bacterial strains employed, and in overall attachment likely being established for several reasons. The most popular food contact material in the food industry is stainless steel type 304 because it is chemically inert, easy to clean, and extremely corrosion-resistant at a range of processing temperatures. Given its continuous usage, this material's topography typically displays crevices and cracks that protect bacteria from sanitising treatments and mechanical cleaning methods.

employed, and in overall attachment likely being established for several reasons. The most

temperatures. Given its continuous usage, this material's topography typically displays crevices and cracks that protect bacteria from sanitising treatments and mechanical clean-

ing methods.

The food matrix components in food processing environments also influence bacterial attachment [31]; e.g., food waste, such as milk and meat exudates enriched in fats, proteins, and carbohydrates, facilitate microorganism growth and multiplication, and favour dual-species biofilm development by *E. Coli* and *Staphylococcus aureus* [32]. Milk lactose improves biofilm production by both *Bacillus subtilis,* by activating the LuxS-mediated quorum-sensing system [33], and *S. aureus* through intercellular polysaccharide adhesion development [34]. Improved biofilm production by *Geobacillus* spp. in milk results in high concentrations of free Ca2+ and Mg2+ [35].

Microbial cell properties, especially hydrophobicity, cellular membrane components (e.g., protein and lipopolysaccharide), appendages (e.g., pili, flagella, fimbriae) and bacteriasecreted EPS, also play a key role in stimulating biofilm production [22]. Fluctuations in biofilm-forming capability among species or strains of different genotypes and serotypes have been identified, which reveals the evolution of enhanced biofilm formation from various genetic backgrounds [8,36]. Similar species can also impact one another in a mixed microbial community, which culminates in the co-colonisation of certain species.

#### **3. Examples of the Most Relevant Biofilms in the Food Industry**

In the food industry, biofilm-forming species appear in factory environments and can be pathogenic to humans because they develop biofilm structures. The processing environments of the food industry, e.g., wood, glass, stainless steel, polyethylene, rubber, polypropylene, etc., act as artificial substrates for these pathogens [37,38]. The characteristics of the bacterial growth form on food in a processing environment involve different behaviours when considering cleaning and disinfection processes. Controlling biofilm formations in the food industry can prove difficult when having to decide the right strategy.

Examples of these relevant biofilm-forming pathogens for the food industry are briefly described in Table 1.

#### *3.1. Bacillus Cereus*

*Bacillus cereus* is a Gram-positive anaerobic or facultative anaerobic spore-forming bacterium that can grow in various environments at wide-ranging temperatures (4 ◦C–50 ◦C). It is resistant to chemicals, heat treatment, and radiation [39]. *B. cereus* is a frequently isolated soil inhabitant from food and food products, such as rice, dairy products, vegetables and meat. It secretes toxins that can cause sickness and diarrhoea symptoms in humans.

*B. cereus* is responsible for biofilm formation on food contact surfaces, such as stainless steel pipes, conveyor belts and storage tanks. It can also form floating or immersed biofilms, which can secrete a vast array of bacteriocins, metabolites, surfactants, as well as enzymes, such as proteases and lipases, in biofilms, which can affect food sensorial qualities [40]. Motility by bacterial flagella confers access to suitable biofilm formation surfaces, and is required for biofilms to spread on non-colonised surfaces. However, *B. cereus* flagella have not been found to be directly involved in adhesion to glass surfaces, but can play a key role in biofilm formation via their motility [55].

#### *3.2. Campylobacter Jejuni*

*Campylobacter* spp., mainly *C. jejuni*, are Gram-negative spiral, rod-shaped, or curved thermophilic and bipolar flagellated motile bacteria [41]. *C. jejuni*, also known as an anaerobic bacterium, can develop biofilms under both microaerophilic (5% O<sup>2</sup> and 10% CO2) and aerobic (20% O2) conditions [56]. Despite it being a fastidious organism, *C. jejuni* can survive outside the avian intestinal tract before it reaches a human host. A range of environmental elements initiates the formation of biofilms, which are then affected by a set of intrinsic factors [57]. The European Union One Health 2018 Zoonoses Report classifies *C. jejuni* as an opportunistic pathogen that is believed to be the causative agent of most bacterial gastroenteritis cases, and has been regarded as a common commensal of food animals and poultry, with turkeys and hens in particular [42]. When the preparation and processing areas of food products or water become contaminated, such as unpasteurised

milk, *C. jejuni* reaches the human host by infecting and colonising the gastrointestinal tract to cause disease [43].


**Table 1.** Biofilm-forming pathogens in the food industry.

#### *3.3. Enterohaemorrhagic Escherichia coli (EHEC)*

*Escherichia coli* is a Gram-negative and rod-shaped bacterium. Most *E. coli* strains form part of human intestinal microbiota and pose no health problem. However, the virulence types of *E. coli* include enterotoxigenic (ETEC), enteroinvasive (EIEC), enteropathogenic (EPEC), and Vero cytotoxigenic (VTEC). O157:H7 EHEC is the most frequent serotype associated with EHEC infections in humans in the USA [58]. Widespread *E. coli* dissemination

in natural environments is, to a great extent, due to its ability to grow as a biofilm. It is worth considering that several *E. coli* strains may cause disease in humans, and that Enterohaemorrhagic *E. coli* (EHEC) strains are the most relevant for the food industry. EHEC serotype O157:H7 is the human pathogen responsible for bloody diarrhoea outbreaks and haemolytic uremic syndrome (HUS) worldwide. They can be transmitted by raw milk, drinking water or fresh meat, fruit, and vegetables; e.g., melons, tomatoes, parsley, coriander, spinach, lettuce, etc. [44].

*E. coli* can employ pili, flagella and membrane proteins to initiate attachment to inanimate surfaces when flagella are lost after attachment and bacteria start producing an extracellular polymeric substance (EPS) that helps to confer bacteria better resistance to disinfectants [59]. There are reports indicating that although EHEC can form biofilms on different food industry surfaces, neither an effective means to prevent EHEC biofilm formation nor an effective treatment for its infections exists because antibiotic treatment tends to increase the risk of haemolytic-uremic syndrome and kidney failure [60].

#### *3.4. Listeria Monocytogenes*

*Listeria monocytogenes* is a Gram-positive bacterium and a ubiquitous food-borne pathogen that can appear in soil, food, and water. Its ingestion can result in abortions in pregnant women, and other serious complications in the elderly and children. The pathogen can be transmitted to several food types, such as dairy products, seafood, meat, fruit, ready-to-eat meals, ice cream, soft cheeses, unpasteurised milk, frozen vegetables, candied apples, and poultry [45,46], but it is not known to be resistant to pasteurisation treatments [61]. The pathogen proliferates at low temperature, and is able to form pure culture biofilms or grow in multispecies biofilms [62]. *L. monocytogenes* can survive under acidic conditions for lengthy periods and can form biofilms that grow without oxygen. Its numbers are likely to rise or lower in biofilms depending on the competing microbes present [63].

Given the presence of pili, flagella and membrane proteins, prevalent *L. monocytogenes* strains possess good adhesion ability in food processing environments [64].

#### *3.5. Salmonella Enterica*

*Salmonella enterica* is a Gram-negative, rod-shaped, flagellate and facultative aerobic bacterium, and a species of the genus *Salmonella* [65].

It can cause gastroenteritis or septicaemia (in some serovars) [66]. *Salmonella* spp. express proteinaceous extracellular fibres known as curli, which are involved in surface and cell-cell contacts, and in promoting community behaviour and host colonisation [67]. Besides curli, different fimbrial adhesins have been identified with biofilm formation implications that are serotype-dependent [40]. *S. enterica serovar Enteritidis* is the most frequent serotype to cause fever, vomiting, nausea diarrhoea, and abdominal pain as main symptoms [47]. Poultry meat is a frequent reservoir for these bacteria in processed food, whose importance as a food pathogen has been demonstrated by the fact that *S. enterica* biofilm formation on food surfaces was the first reported case in 1966 to possess complex multicellular structures [48].

When contaminating a food pipeline biofilm, *S. enterica* may cause massive outbreaks, and even death in infants and the elderly. It can grow on stainless steel surfaces to form a three-dimensional (3D) structure with several call layers of different morphologies depending on available nutrients, such as the reticular shaped ones generated when cultured on tryptic soy broth (TSB) medium [68].

#### *3.6. Staphylococcus Aureus*

*Staphylococcus aureus* is a Gram-positive, non-spore-forming, non-motile, facultative anaerobic bacterium capable of producing enterotoxins from 10–46 ◦C. *S. aureus* can multiply on the skin and mucous membranes of food handlers, and can become a major issue in food factories [49]. These enterotoxins are heat-stable and can be secreted during *S.*

*aureus* growth in foods contaminated by food handlers. The bacterium grows well in high salt- or sugar-content foods with little water activity. The foods frequently implicated in Staphylococcal food-borne disease are meat and meat products, poultry and egg products, milk and dairy products, bakery products, salads, and particularly cream-filled cakes and pastries and sandwich fillings [50]. *S. aureus* is known for its numerous enteric toxins. These enterotoxins bind to class II MHC (major histocompatibility complex) in T-cells, which results in their activation that can lead to acute toxic shock with sickness and diarrhoea [69].

#### *3.7. Pseudomonas spp.*

*Pseudomonas* is a heterotrophic, motile, Gram-negative rod-shaped bacterium. Pseudomonads are generally ubiquitous psychrotrophic spoilage organisms that are often found in food processing environments, including floors and drains, and also on fruit, vegetables, and meat surfaces, and in low-acid dairy products [17,62]. The extracellular filamentous appendages produced by motile microorganisms result in both the attachment process and the interaction with surfaces in different ways. Flagella and pili have been thoroughly studied [70].

When biofilms develop and their regulation by quorum sensing is considered, *Pseudomonas aeruginosa* can be taken as a model organism [71], which is about 1–5 µm long and 0.5–1.0 µm wide. A facultative aerobe grows via aerobic and anaerobic respiration with nitrate as the terminal electron acceptor [71].

*Pseudomonas* spp. produce huge amounts of EPS and are known to attach and form biofilms on stainless steel surfaces. They can co-exist with other pathogens in biofilms to form multispecies biofilms, which make them more resistant and stable [62]. These biofilms can be accompanied by a distinct blue discolouration (pyocyanin) on fresh cheese produced by *P. fluorescens* [72].

#### *3.8. Geobacillus stearothermophilus*

*Geobacillus stearothermophilus* is a Gram-positive, thermophilic, aerobic, or facultative anaerobic bacterium [73]. Thermophiles, such as *G. Stearothermophilus,* formerly known as *Bacillus stearothermophilus,* can attach to stainless steel surfaces on processing lines in evaporators and plate heat exchangers, which allows them to grow and produce biofilms, which implies the potential release of single cells or aggregates of cells into the final dry product [74]. *B. stearothermophilus* are able to form biofilms on clean stainless steel surfaces and to release bacteria into milk during dairy industry processing [75]. The above-cited authors observed that the conditions for a biofilm in a laminar flow milk system were more adequate for the growth of spore-forming bacteria, which are thermophilic. Their growth as a culture medium in milk is quite difficult [75].

#### *3.9. Anoxybacillus flavithermus*

*Anoxybacillus flavithermus* is another Gram-positive, thermophilic, and spore-forming organism that is facultatively anaerobic and non-pathogenic [76]. *A. flavithermus* is a potential contaminant of dairy products, and poses a problem for the milk powder processing industry, as high levels will reduce milk powder acceptability for both local and international markets [77]. *A. flavithermus* spores are very heat-resistant and their vegetative cells can grow at temperatures up to 65 ◦C with a significant increase in bacterial adhesion on stainless steel surfaces in the presence of skimmed milk. This indicates that milk positively influences these species' biofilm formation [78]. In the dairy industry, the commonest biofilm-forming isolates are thermophilic genera [79]. In many parts of the world, *A. flavithermus* and *G. stearothermophilus* are regarded as the most dominant thermophilic microbial contaminants of milk powders [78].

#### *3.10. Pectinatus spp.*

*Pectinatus* is Gram-negative, non-spore-forming, and anaerobic bacteria that have been linked with a high concentration of biofilms in breweries due to sanitation problems [80]. Spoilage bacteria were first isolated from a brewery in the USA in unpasteurised beer stored at 30 ◦C [81]. *P. cerevisiiphilus* have also been isolated from many breweries in Germany, Spain, Norway, Japan, the Netherlands, Sweden, and France [80].

#### *3.11. Synergistic Pathogens*

A combination of several pathogens can synergistically interact to form biofilms in the food industry. In food-processing environments, bacteria are able to exist as multispecies biofilms, from where both spoilage and pathogenic bacteria can contaminate food [82]. For instance in the fishing industry, fresh fish products can suffer from biofilm formation by mixed pathogenic species (*Aeromonas hydrophila*, *L. monocytogenes*, *S. enterica,* or *Vibrio* spp.), which can imply significant health and economic issues [83]. Synergistic interactions have been observed in a fresh-cut produce processing plant, where *E. coli* interacted with *Burkholderia caryophylli* and *Ralstonia insidiosa* to form mixed biofilms. Acylhomoserine lactones (AHLs) can control biofilm formation in synergistic interactions among mixed species. Interference of AHLs is manifested by AHL lactonases and acylases, both of which are present in Gram-positive and Gram-negative bacteria [60].

Bacteria use quorum sensing to coordinate biofilm production and dispersion, when bacteria attach to a biotic or abiotic surface, and cell-to-cell attachment engages in communication via a quorum sensing (QS)-based extracellular cell signalling system [84]. The importance of cell signalling for bacterial biofilm formation has been further confirmed by the control of exopolysaccharide synthesis by quorum-sensing signals, as in *Vibrio cholera* [85].

Synergistic pathogens are found in several works, where biofilm levels of the fourspecies consortia have been further examined and compared to the biofilm production levels of each isolate under monospecies conditions. They have revealed that *P. aeruginosa* and *A. junii* isolated from different samples to contribute as best biofilm producers, including poor or non-biofilm-producing isolates, which increases the overall biofilm formation in the included consortia [86]. Several authors [87] have found positive synergistics in other studies by investigating mixed species of biofilms, such as *Candida albicans.*

In food industries, biofilm-related effects (pathogenicity, corrosion of metal surfaces, and alteration to organoleptic properties due to the secretion of proteases or lipases) are critically important. For example, in the dairy industry several processes and structures (pipelines, raw milk tanks, butter centrifuges, pasteurisers, cheese tanks, packing tools) can act as surface substrates for biofilm formation at different temperatures and involve several mixed colonising species. Thus, it is essential that accurate methods to visualise biofilms in situ be set up to avoid contamination and to ensure food safety in the food industry.

#### **4. Biofilm Control and Elimination**

It is well-known that biofilm bacteria present a distinct phenotype with a genotype as regards gene transcription and growth rates under very particular conditions that differ from planktonic conditions [88]. Biofilms are capable of adhering to a very wide diversity of surfaces with distinct biotic and abiotic compositions, including human tissue and medical devices. Once biofilms form, they are a major threat because they cause infectious diseases and economic loss. In the 1940s, several authors produced further research works into biofilm evolution and surface relations for marine microorganisms [89] and seawater [90]. Nevertheless, marked progress has been made given the incorporation of the electron microscope, which allows high-resolution photomicroscopy at much higher magnifications than light microscopy [85]. Indeed, the most revealing discovery of the relation with biofilm elimination was a description of its structure, the surrounding matrix material, and the cells enclosed in these biofilms were polysaccharides, as by special stains revealed [85]. Doubtlessly, disinfectants have proven more efficient in fighting against biofilms since 1973, while Characklis (1973) [91] showed marked persistence and resistance to disinfectants, such as chlorine.

#### *4.1. Biofilm Elimination in the Food Industry*

Biofilm formation has been investigated in food industry and hospital environments. Perhaps, in the research conducted, hospitals, to eliminate biofilms, have been more successful, thanks to the easier applications and special surface compositions (antibiofilm activity) in medical surroundings, such as implants, prostheses, tools, and surfaces for operating theatres.

To date, many efforts have been made to reduce biofilm formation on food industry surfaces, but those works were based mainly on new disinfectants with different efficacies. These results have improved in line with specific mechanisms for initial surface attachment, developing a group structure and ecosystem, and detachments [75] with dissimilar results.

Nowadays, disinfectants are the best ally to eliminate biofilms. However, other research fields, such as the composition of surfaces for materials to prevent bacterial adhesion and developing phages to combat biofilm-forming bacteria, have obtained favourable results. Doubtlessly, most research works have focused on the bacteriological biofilm, without discussing the hypothesis of filamentous fungi being responsible for biofilm formation. Several authors [92,93] support this theory, where the presence of Aspergillus fumigatus has been presented as biofilm-responsible. In this case, the marked similitude of bacteriological and filamentous fungi biofilms is based on morphological changes, the presence of an extracellular polymeric matrix, differential gene expression, and distinct sensitivity to antifungal drugs compared to diffuse or loosely associated (planktonic) colonies [94,95].

Before we go on to explain several factors that could influence bacterial adhesion and biofilm formation, we should bear the food industry's hygiene design in mind. In order to prevent microorganisms entering food production, factories, and the employed hygienic equipment should be designed to limit microorganisms from accessing. Aseptic equipment must be isolated from microorganisms and foreign particulates. To prevent microorganism growth, equipment should be designed to prevent any areas where microorganisms can harbour and grow, along with gaps, crevices, and dead areas. This is also important during production, when microorganisms can grow very quickly under favourable conditions [96].

According to such premises, food companies have the capacity to apply innovation to design the industry and its equipment. In both the USA and the European Union (EU), the trend in regulations in this field is not so much command and control by government regulators, but lies more in self-determination by the food industry. In particular, hazard analysis and critical control points (HACCP) systems provide the skills to replace detailed regulatory requirements with the overall goals to be fulfilled [97]. The Food and Drug Administration (FDA) and the United States Department of Agriculture (USDA) Food Safety Inspection Service (FSIS) share primary responsibility for regulating food safety in the USA. One example is the recommendation for equipment and process controls: "Seams on food-contact surfaces shall be smoothly bonded or maintained to reduce accumulation of food particles, dirt, and organic matter and thus minimise the opportunity for growth of microorganisms" [85].

European Union (EU) statutory instruments include EC regulation no. 852/2004 (Hygiene of Foodstuffs) and EC regulation no. 853/2004 (Specific Rules Food of Animal Origin), which expect food manufacturers to control food safety risks by HACCP systems.

In short, the performance of a cleaning and disinfection programme (CDP) to avoid biofilm formation should start by aptly designing the hygiene of equipment, surfaces, and devices. Today the CDP is a proven effective measure in fighting against biofilms.

#### *4.2. Factors Associated with Bacteria*

Several factors associated with surfaces or bacteria can influence adhesion from the planktonic phase and biofilm evolution, such as:


Bacterial cellular surface components:

Hydrophobic interactions tend to increase with the enhanced non-polar nature of one involved surface or more, and most bacteria are negatively charged [85]. This means that the cell surface's hydrophobicity is a relevant factor during adhesion.

• Fimbria, pili, and flagella: fimbriae, non-flagellar appendages other than those implicated in the transfer of viral or bacterial nucleic acids (known as pili) are responsible for cell surface hydrophobicity. Most investigated fimbriae contain a high proportion of hydrophobic amino acid residues [100]. Fimbriae include adhesins that attach to some sort of substratum so that bacteria can withstand shear forces and obtain nutrients. Therefore, fimbriae play a role in cell surface hydrophobicity and attachment, presumably by overcoming the initial electrostatic repulsion barrier between the cell and the substratum [101].

Studies' responses surface methods are used to not only develop and optimise models for food processing systems and operations, but also to better elucidate bacterial adhesion and biofilm formation processes. The response surface method provides valuable information to help decision making about disinfection and cleaning procedures for the utensils, equipment and containers employed in the food industry [102]. Hence, surfaces and materials of equipment, plus floors and walls, also impact biofilms, along with dead spaces, crevices, porous and rough material surfaces, which must be eliminated to avoid biofilm formation [12].

The most adopted strategies for controlling biofilms are sanitation procedures that combine detergents and disinfectants. Alkaline detergent eliminates organic and inorganic acid detergent waste from surfaces, while disinfectants reduce spoilage microorganisms, and diminish or eliminate pathogens, to safe levels [12]. Enzymatic detergents have replaced traditional alkaline and acid detergents, because enzymes (proteases, lipases, amylases) can remove biofilms in the food industry [102] as enzymes reduce the physical integrity of PS by weakening the structural bonds of the lipids, proteins, and carbohydrates that form its structure [103]. Other advantages over detergents include low-toxicity and biodegradability, but application costs and requirements (temperature, time) are higher than detergents. This is why several detergent manufacturers have marketed a synergetic combination of enzymes, chelating agents, surfactants, and solvents.

#### *4.3. Disinfectants and Biofilm Resistance*

The most widely used disinfectants in the food industry's disinfection programmes are quaternary ammonium compounds (QAC), amphoteric compounds, hypochlorites, peroxides (peracetic acid and hydrogen peroxide) [15], aldehydes (formaldehyde, glutaraldehyde, paraformaldehyde), and phenolics. This product list remains unchanged after 18 years. Today, alkyl amines, chlorine dioxide and quaternary ammonium blends are incorporated into disinfection programmes. Besides these, alcohols, phenolic compounds, aldehydes, and chlorhexidine are also resorted to, but mostly in health services. In the food processing industry, disinfectants can remain on surfaces for longer due to microorganisms' prolonged exposure to the employed disinfectant, which improves their efficacy [104].

#### 1. Sodium hypochlorite (NaOCI):

It is one of the most widespread disinfectants in the food industry despite its disadvantages and the growing use of new products on the market. Its desirable reaction produces

both hypochlorous acid (HOCl) and the hypochlorite ion (-OCl), which are strong oxidising agents that eliminate cells given their ability to cross the cell membrane, and to oxidise the sulfhydryl groups of certain enzymes participating on the glycolytic pathway [97]. It has been described to react with wide-ranging biological molecules, such as proteins [105], amino acids [106], lipids [107], peptides [108], and DNA [109] under physiological pH conditions [110].

However, sodium hypochlorite may be affected by organic matter because free chlorine might react with natural organic matter and be converted into inorganic chloramines to generate trihalomethanes, by reducing antimicrobial activity against biofilms [111], and it is known to be less reactive than free chlorine. Peracetic acid is reported as being the most effective sanitiser against biofilms because it is a strong oxidising agent that does not interact with organic matter waste [112,113].

#### 2. Quaternary ammonium:

QACs, such as benzalkonium chloride, cetrimide, didecyldimethylammonium chloride, and cetylpyridinium chloride, are cationic detergents (surfactants or surface-active agents). They reduce surface tension and form micelles to lead to dispersion in a liquid. This property is resourceful for removing microorganisms. They are membrane-active agents that interact with not only the cytoplasmic membrane of bacteria, but with the plasma membrane of yeast. Their hydrophobic activity also makes them effective against lipid-containing viruses. QACs interact with intracellular targets and bind to DNA [114]. However, their efficacy is still questioned given the appearance of relatively high resistance to *Listeria monocytogenes* (10%), *Staphylococcus* spp. (13%), and *Pseudomonas* spp. (30%), and lower resistance to lactic acid the bacteria (1.5%) and coliforms (1%) isolated from food and the food processing industry [115].

#### 3. Peracetic acid:

In the last decade, peracetic acid (PAA) has been widely used by the food industry in water and wastewater treatment, even in paper machines [116], to control biofilms. Its antimicrobial effect is probably due to the oxidation of thiol groups in proteins, disruption of membranes [117], or damage to bases in DNA [118]. Its use has been shown to increase the sensitivity of bacterial spores to heat [119]. The efficacy and environmental safety of peracetic acid make it an attractive disinfecting agent for industrial use.

Bacterial regrowth after oxidant treatment (peracetic acid and free chlorine) depends on the absence or presence of organic matter. The oxidation-reduction (redox) potential of PAA (1.385 V vs. standard hydrogen electrode or standard hydrogen electrode (SHE) for the redox couple of CH3COOOH(aq)/(CH3COO− (aq) + H2O(l))) comes close to that of free chlorine (1.288 V vs. SHE for the redox couple of HOCl(aq)/Cl<sup>−</sup> (aq)) under biochemical standard state conditions (pH 7.0, 25 ◦C, 101.325 Pa) [120]. For this reason, acid peracetic and free chlorine may have similar efficiencies in preventing planktonic bacteria regrowth in the absence of organic matter. As PAA reacts with organic matter more slowly than free chlorine, its self-decomposition is slower [121].

#### Resistance to Disinfectants

Bacterial resistance to disinfectants in the planktonic phase can hardly be compared to biofilm resistance. Yet several studies have shown contrary results to widespread belief, such as the existence of wide interspecific variability of resistance to disinfectants. Grampositive strains generally appear to better resist than Gram-negative strains. This resistance is also variable among strains of the same species [122].

Given the growing interest in knowing biofilm resistance to chlorine, quaternary ammonium and peracetic acid, many studies have been conducted [123]. The bacteria in mature biofilms are 10- to 1000-fold more resistant to antibiotics than the bacteria in the planktonic phase [124], and this resistance appears against biocides. However, this natural resistance is still unknown, and probably depends on many factors, mainly of structural biofilm barriers and genetic factors for adaptation. To explain this resistance,

several authors [125] have suggested three possible causes by three hypotheses: the first is based on the slow or incomplete diffusion of antibiotics into inner biofilm layers. The second lies in the changes taking place in the biofilm microenvironment as some biofilm bacteria fall into a slow growth state due to lack of nutrients or given the accumulation of harmful metabolites and, therefore, survive [126]. Finally, the third hypothesis indicates a subpopulation of cells in the biofilm whose differentiation resembles the spore formation process. They have a unique and highly resistant phenotype to protect them from the effects of antibiotics, and are a biologically programmed response to the sessile life form of bacteria [127].

In addition, aquatic fauna is less affected by PAA than by chlorine [128]. Therefore, PAA is considered a green alternative to chlorine for disinfection purposes, and its disinfection performance is currently being investigated [120].

As previously mentioned, there are two very different situations for action against biofilm formation: the food industry and the health field. There is still a lot to do in both fields, but it is true that CDP is practically the only implementation in the food industry. The health field has witnessed much more progress: phages [129], aerosolisation [130], sonication brush [131], and metal ion solutions (silver, copper, platinum, gold, and palladium) [132].

In short, some authors [133] establish two strategies for fighting biofilms in the food industry: structural modification of surfaces or application of antibacterial or antibiofilm coatings [134]. Thus, several alternative products to classic disinfectants (chlorine, quaternary ammonium, etc.), such as, plant-derived antimicrobials (essential oils: orange-sorrel, lemon, lavender, chamomile, peppermint, oregano), with thymol and carvacrol being the compounds that display more significant antimicrobial action in shorter action times.

#### *4.4. Alternative Methods to Eliminate Biofilms*

Phages: bacteriophages are specific "viruses" of microbial cells that are specific to the different serotypes or strains of microbial species, and are obligate parasites with a genetic parasitism [135]. Bacteriophages inject their DNA and force the cell to produce the bacteriophage genome and structures (e.g., capsid and tail). When phages are complete, they lyse cells, which means that bacteriophage infection can destroy the entire colony [136,137]. In the last few years, the FDA/USA approved preparing bacteriophages (LISTEX P100) to combat the direct presence of *L. monocytogenes* in foods [138,139].

Aerosolisation: a disinfection method with different disinfectants applied to working areas by pulverisation. Several authors [138] have shown its efficacy as a biofilm control method in the food industry and hospitals by using hydrogen peroxide [140], sodium hypochlorite, and peracetic acid [141].

Knowledge about the resistance mechanisms associated with biofilm evolution could be primordial to develop new actions or strategies by biocides and antibiotics [142], such as modified wound dressings with phyto-nanostructured coatings to prevent staphylococcal and pseudomonal biofilm development [143].

Involving bacterial adherence: in recent years, new biochemistry methods have been studied to prevent biofilm formation. The most efficient strategy would interfere with bacterial adherence, as this first step is paramount in biofilm formation, performed by the direct blockage of surface receptors [144] or by a non-specific strategy, which normally involves compounds with anti-adherence properties [145]. Another biofilm inhibition form would be to impede communication processes between bacteria to enter the biofilm by employing different natural or artificially synthesised compounds [146]. One example of this is *P. aeruginosa,* which uses quorum sensing for modular biofilm evolution, and proposes that agents are capable of blocking quorum sensing (QS), and could be useful for avoiding biofilm formation [144].

The role of the QS in biofilms includes controlling the cell-to-cell communicating system in response to small diffusible signal molecules, such as N-acyl-homoserine lactones (AHLs), produced by Gram-negative bacteria [147]. It has been shown to play crucial roles

in biofilm formation by activating the transcription of related genes [148]. Specifically, different AHLs have been detected in biofilm reactors and several bacterial species have been identified to possess the capacity to produce AHLs [149]. Thus, AHLs-based QS has been widely reported to regulate many microbial behaviours, including EPS expression, nitrogen transformation, organic pollutant degradation and microbial community construction [150].

Current advances in nanoscience nanotechnology and nanosensors [151] have emerged for applications to detect microorganisms and biofilms [152] with high sensitivity and good spatial resolution on nanoscale scopes [153]. Recent works have been published on imaging approaches of biofilm microprocesses [154], in-situ surface-enhanced Raman Scattering (SERS) analyses, biofilm visualisation [155] and biosensors of bacteria in foods [156]. The enormous advantages and great potential observed in bioassays based on multifunctional optical nanosensors are promising to continue with a view to ensure and promote food safety and quality. From the detection targets perspective, QS detection might become a new biofilm research trend based on evidence that biofilm formation can be inhibited by blocking QS [151].

#### **5. Biofilm Identification Techniques and Methods to Visualise Biofilms In Situ**

#### *5.1. General Aspects of Biofilm Study Techniques*

Doubtlessly, biofilms can pose a major challenge in both clinical microbiology and hygiene food areas. In the latter area, several authors consider them a real threat. Currently, methods aim to analyse biofilm formation and development, which have not yet been standardised. Different methods have been followed to qualitatively and quantitatively evaluate biofilms, and each one is useful for estimating one peculiar biofilm lifestyle aspect [157]. Nevertheless, research to identify and acquire knowledge of biofilms has allowed distinct techniques to be developed and adapted from microbiology or cell histology. It is essential to evaluate biofilm formation for a sensitive, specific and reproducible methodology for biofilm quantification to become available.

Different approaches classify the methods followed to detection biofilms on very distinct surfaces: (i) the simplest classification of methods is direct or indirect [17]. (ii) Rapid tests of hygienic control, and methods for microscopic, biomolecular, extracellular polymeric, physical, or chemical substances (EPS) are another possible classification [158]. (iii) A recent publication [159] only refers to the technology of the referred methods being classified as physics, physico-chemistry or chemistry, and recommends three effective approaches for testing biofilms: (a) observations by various microscopic methods with different view fields at the same point; (b) in-depth data analyses during microscopic image processing; (c) a combination study using atomic force microscopy (AFM) and chemical analysis. Perhaps this is the most advanced and appropriate methodology for biofilm analyses, where the detailed image of the surface will help to build a relation among the biofilm matrix, interactions and other factors like pH, surfaces [159]. However, there is another case in which bacterial species can help to reduce biofilms, where *Bacillus licheniformis* can express hydrolytic enzymes capable of reducing detrimental biofilms [160].

Therefore, depending on the set objectives, that is, what we wish to achieve with the biofilm, we should choose a technique according to our study. Not all techniques are suitable for a certain purpose, but might be compatible. Thus, some methods are suited for quantifying the biofilm matrix, while others are able to evaluate both living and dead cells, or exclusively quantify viable cells in biofilms.

By considering the complexity and heterogeneity of the biofilm structure, the exact research objective should be set. The amount of EPS, the total number of bacterial cells embedded in biofilms, or the actual number of living bacteria in biofilms must be considered to be different targets that require distinct experimental approaches [157]. We should bear in mind that the biofilm volume is constituted mainly by an extracellular matrix (95–65% range), which is composed mostly of proteins (>2%), and other constituents, such as polysaccharides (1–2%), DNA molecules (<1%), RNA (<1%), ions (bound and free), and finally 97% water [161]. Thus, the biofilm research methodology should address the identification of bacteria and other matrix constituents.

In order to obtain a fundamental understanding of the formation and presence of bacterial biofilms, our analysis should include the detection of bacteria and the matrix. The most frequently followed methods to assess biofilm heterogeneity are direct microscopic imaging of the local biofilm morphology or microscopic measurements of local biofilm thickness [162]. For many applications, time-lapse microscopy with Confocal Laser Scanning Microscopy (CLSM) is an ideal tool for monitoring at a spatial resolution in the order of micrometres, and it allows the non-destructive study of biofilms by examining all layers at different depths. In this way, it is possible to reconstruct a three-dimensional structure [163]. Matrix detection can be achieved by a double-staining technique combined with CLSM, which allows the simultaneous imaging of bacterial cells and glycocalyx in biofilms [164].

#### *5.2. Colorimetric Methods*

#### 5.2.1. Evaluating the Biofilm Matrix

Staining the biofilms grown in microtiter plates wells is widely utilised by researchers to screen and compare biofilm formation by different bacteria or under various conditions [165]. Of the methods described in the literature, crystal violet (CAS number 931418 92 7) [166] is the most widespread for biofilm biomass quantification [167,168]. This basic dye binds negatively charged molecules and, thus, stains are able to dye both bacteria and the surrounding biofilm matrix. Acetic acid can be used as the extraction solvent and be measured by absorbance at 700–600 nm. Safranin staining can also be employed for biofilm biomass quantification [165,169], but results in lower optical densities than crystal violet staining and, therefore, may not be as sensitive to detect small amounts of biofilm [165].

Crystal violet staining tests the concentration of the dye incorporated into bacterial cell walls, and depends on cells' integrity, but not on viability. However, other methods like ATP bioluminescence report the cell's metabolic status and drops to undetectable limits within minutes after cell death. Resorting to both methods can provide supplementary information on the cell exposed to disinfectant. The results can indicate that, despite the drastic drop in viable cell numbers in the biofilm after disinfectant treatment, a significant number of intact cells, or cellular debris, may still be capable of retaining the dye. This observation leads to the question about the reliability of crystal violet staining as a method to monitor biofilm disinfection [170].

Another colorimetric method for living cells is fluorescein diacetate (CAS number 596 09 8), which employs a useful live-cell fluorescent stain that is hydrolysed to fluorescent fluorescein in live cells. The signal can be spectrophotometrically measured. This is suitable for cell viability assays with intact membranes as dead cells are unable to metabolise fluorescein diacetate. Thus, there is no fluorescent signal [157].

#### 5.2.2. Cell Staining

Visualising a cell with fluorescent compounds provides a wide variety of information to analyse cell functions. Various activities and cell structures can be targeted for staining with fluorescent compounds [171]. These cell components are mostly cell membranes, nucleotides, and proteins. The stain can pass to cells depending on the molecule charge, hydrophobicity, or reactivity. Thus, small neutral and positively charged fluorescent compounds can normally reach mitochondria for dyeing. Negatively charged molecules cannot pass through viable cell membranes. Ester is a suitable functional group for staining viable cells because it can pass through viable cell membranes, where it is hydrolysed by cellular esterases into a negatively charged compound [171].

Other complementary techniques can be run to examine the performances of advanced microscopic techniques employed to study microbial biofilms (i.e., confocal laser scanning microscopy, mass spectrometry, electron microscopy, Raman spectroscopy) [157].

Spectrofluorometric assays for the quantification of biofilms of gram-negative and gram-positive bacteria is a method that utilises the specific binding of the wheat germ agglutinin-Alexa Fluor 488 conjugate (WGA) to N-acetylglucosamine in biofilms [172]. This lectin conjugate also binds to N-acetylneuraminic acid on the peptidoglycan layer of gram-positive bacteria. WGA specifically binds to polysaccharide adhesin (poly Nacetylglucosamine), which is involved in biofilm formation by both gram-positive and gram-negative bacteria. Burton et al. [172] compared the colorimetric assay with the spectrofluorometric assay, whose results revealed that WGA staining may be a more specific means of *E. coli* and *Staphylococcus epidermidis* biofilm detection and quantification.

#### 5.2.3. LIVE/DEAD

This method is based on employing two different nucleic acid binding stains. The first dye is green fluorescent (Syto9, λex 486 nm and λem 501 nm), which is able to cross all bacterial membranes and bind to the DNA of both Gram-positive and Gram-negative bacteria. The second dye is red-fluorescent (propidium-iodide (PI), CAS number 25535-16-4, λex 530 nm and λem 620 nm), which crosses only damaged bacterial membranes. Stained samples are observed under a fluorescent optical microscopy to evaluate live and dead bacterial populations (see Figures 3–5). In fact, live bacteria fluoresce in green and dead bacteria fluoresce in orange/red [173]. The efficiency of both stains is conditioned by some factors, such as the reagent's binding affinity to cells [169], physiological cell state [174], reagent concentration [175], and temperature and incubation time [176].

Both stains are suitable for use in fluorescence microscopy, confocal laser scanning microscopy, fluorometry flow and cytometry, and can be employed as a nuclear counterstain. LIVE/DEAD staining cannot be performed for the direct staining of biofilms on surfaces because of interference between the stain and polysaccharides of the biofilm matrix and slime [177].

This method's main downside involves having to observe a statistically relevant portion of the sample, which is representative of the whole population. Overall, the method provides only semiquantitative results because the total count of bacterial cells is not possible [178]. Nevertheless, this inconvenience can be prevented by employing imaging software, such as cellSens®, which can count and measure cells depending on the staining cell. *Int. J. Environ. Res. Public Health* **2021**, *18*, x FOR PEER REVIEW 16 of 32

**Figure 3.** Planktonic state of *Pseudomonas fluorescens* stained with LIVE/DEAD (SYTO 9 and propidium iodine) after treatment with sodium hypochlorite (500 ppm) for 15 minutes. Viable bacteria (green) and damaged bacteria (red). Magnification ×100. **Figure 3.** Planktonic state of *Pseudomonas fluorescens* stained with LIVE/DEAD (SYTO 9 and propidium iodine) after treatment with sodium hypochlorite (500 ppm) for 15 min. Viable bacteria (green) and damaged bacteria (red). Magnification ×100.

**Figure 4.** Planktonic state of *Pseudomonas fluorescens* stained with LIVE/DEAD (SYTO 9 and propidium iodine) after treatment with peracetic acid (250 ppm) for 15 min. Viable bacteria (green) and

damaged bacteria (red). Magnification ×100.

(green) and damaged bacteria (red). Magnification ×100.

**Figure 4.** Planktonic state of *Pseudomonas fluorescens* stained with LIVE/DEAD (SYTO 9 and propidium iodine) after treatment with peracetic acid (250 ppm) for 15 min. Viable bacteria (green) and damaged bacteria (red). Magnification ×100. **Figure 4.** Planktonic state of *Pseudomonas fluorescens* stained with LIVE/DEAD (SYTO 9 and propidium iodine) after treatment with peracetic acid (250 ppm) for 15 min. Viable bacteria (green) and damaged bacteria (red). Magnification ×100. *Int. J. Environ. Res. Public Health* **2021**, *18*, x FOR PEER REVIEW 17 of 32

**Figure 5.** Planktonic state of *Pseudomonas fluorescens* stained with LIVE/DEAD (SYTO 9 and propidium iodine) after treatment with Sodium hypochlorite (350 ppm) for 15 min. Viable bacteria (green) and damaged bacteria (red). Magnification ×100. **Figure 5.** Planktonic state of *Pseudomonas fluorescens* stained with LIVE/DEAD (SYTO 9 and propidium iodine) after treatment with Sodium hypochlorite (350 ppm) for 15 min. Viable bacteria (green) and damaged bacteria (red). Magnification ×100.

#### 5.2.4. Different Fluorescents Stainings

5.2.4. Different Fluorescents Stainings

tinguishing between viable and dead cells [182].

Both stains are suitable for use in fluorescence microscopy, confocal laser scanning microscopy, fluorometry flow and cytometry, and can be employed as a nuclear counterstain. LIVE/DEAD staining cannot be performed for the direct staining of biofilms on surfaces because of interference between the stain and polysaccharides of the biofilm matrix and slime [177]. The application of fluorescent stains to cells and food soil can be useful for the quantitative analysis of surface cleanability. Thus, the stain combination and working concentration are essential for assessing the hygienic conditions of surfaces [179] or testing disinfectant efficacy against bacteria. Different methods can be followed to visualise and differentiate cells and organic matter. The staining techniques to measure surface coverage by the two

tion of the sample, which is representative of the whole population. Overall, the method provides only semiquantitative results because the total count of bacterial cells is not possible [178]. Nevertheless, this inconvenience can be prevented by employing imaging software, such as cellSens®, which can count and measure cells depending on the staining cell.

The application of fluorescent stains to cells and food soil can be useful for the quantitative analysis of surface cleanability. Thus, the stain combination and working concentration are essential for assessing the hygienic conditions of surfaces [179] or testing disinfectant efficacy against bacteria. Different methods can be followed to visualise and differentiate cells and organic matter. The staining techniques to measure surface coverage by the two stains by image analysis are highlighted [180] using DAPI and Rhodamine B, DAPI and Fluorescein, or non-specific stains, such as acridine orange, and are also available and specific for particular organic matter [180], and/or for microorganisms [181].

The use of different staining types can be explained by the results obtained in each study after checking the best biofilm and cells staining. These results depend on bacterial species, residual organic matter on surfaces, pH, disinfectant, etc. Whitehead et al. [182] conducted a large study with different dyeing, and concluded that the best combination was DAPI (CAS no. 28718-90-3, λex 340 nm, λem 488 nm, blue) and Rhodamine B (CAS no. 81-88-9, λex 553 nm, λem 627 nm, red), as it allowed the quantitative determination of *L. monocytogenes* and whey on a surface with fluorescent staining under epifluorescence microscopy. It is also useful for demonstrating the hygienic status of surfaces (Figures 7 and 8). The other tested staining procedures were unsatisfactory, or only slightly so, for dis-

DAPI staining is suitable for studying cell viability in planktonic situations (initial attachment) and biofilms attached to surfaces (proliferation and growth–maturation) (Figure 6). Nevertheless, coculture biofilm studies need to spatially discriminate between stains by image analysis are highlighted [180] using DAPI and Rhodamine B, DAPI and Fluorescein, or non-specific stains, such as acridine orange, and are also available and specific for particular organic matter [180], and/or for microorganisms [181].

The use of different staining types can be explained by the results obtained in each study after checking the best biofilm and cells staining. These results depend on bacterial species, residual organic matter on surfaces, pH, disinfectant, etc. Whitehead et al. [182] conducted a large study with different dyeing, and concluded that the best combination was DAPI (CAS no. 28718-90-3, λex 340 nm, λem 488 nm, blue) and Rhodamine B (CAS no. 81-88-9, λex 553 nm, λem 627 nm, red), as it allowed the quantitative determination of *L. monocytogenes* and whey on a surface with fluorescent staining under epifluorescence microscopy. It is also useful for demonstrating the hygienic status of surfaces (Figures 7 and 8). The other tested staining procedures were unsatisfactory, or only slightly so, for distinguishing between viable and dead cells [182]. *Int. J. Environ. Res. Public Health* **2021**, *18*, x FOR PEER REVIEW 18 of 32

> DAPI staining is suitable for studying cell viability in planktonic situations (initial attachment) and biofilms attached to surfaces (proliferation and growth–maturation) (Figure 6). Nevertheless, coculture biofilm studies need to spatially discriminate between species, and classic methods, such as crystal violet (CV), SYTO9/propidium iodide, and DAPI staining are insufficient given their non-specific nature [183], and selectively bind to each species. This burden can be overcome by applications, such as mutants expressing green fluorescent protein (GFP) [182,184], fluorescently labelled antibodies [185], and fluorescence in situ hybridisation (FISH). species, and classic methods, such as crystal violet (CV), SYTO9/propidium iodide, and DAPI staining are insufficient given their non-specific nature [183], and selectively bind to each species. This burden can be overcome by applications, such as mutants expressing green fluorescent protein (GFP) [182,184], fluorescently labelled antibodies [185], and fluorescence in situ hybridisation (FISH). The DAPI/Rhodamine B combination in biofilms offers the best resolution and quan-

> The DAPI/Rhodamine B combination in biofilms offers the best resolution and quantification power between cells and organic matter (Figures 6–8). Several authors, such as Almeida et al. [183], have applied peptide nucleic acid fluorescence in situ hybridization (PNA FISH) combined with DAPI as a steady method to evaluate, validate, quantify, and characterise the initial adhesion and biofilm formation of three microorganisms: *Salmonella enterica*, *Listeria monocytogenes* and *Escherichia coli.* tification power between cells and organic matter (Figures 6–8). Several authors, such as Almeida et al. [183], have applied peptide nucleic acid fluorescence in situ hybridization (PNA FISH) combined with DAPI as a steady method to evaluate, validate, quantify, and characterise the initial adhesion and biofilm formation of three microorganisms: *Salmonella enterica*, *Listeria monocytogenes* and *Escherichia coli.*

**Figure 6.** *Pseudomonas aeruginosa* biofilm on the stainless steel coupons. Stained with DAPI (0.1 mg/mL; 10 µL) (Magnification ×100). **Figure 6.** *Pseudomonas aeruginosa* biofilm on the stainless steel coupons. Stained with DAPI (0.1 mg/mL; 10 µL) (Magnification ×100).

**Figure 7.** *Pseudomonas fluorescens* biofilm (24 h) on the stainless steel. Coupons stained with DAPI

(0.1 mg/mL; 10 µL) and Rhodamine B (0.1mg/mL; 10 µL) (Magnification × 100).

species, and classic methods, such as crystal violet (CV), SYTO9/propidium iodide, and DAPI staining are insufficient given their non-specific nature [183], and selectively bind to each species. This burden can be overcome by applications, such as mutants expressing green fluorescent protein (GFP) [182,184], fluorescently labelled antibodies [185], and flu-

The DAPI/Rhodamine B combination in biofilms offers the best resolution and quantification power between cells and organic matter (Figures 6–8). Several authors, such as Almeida et al. [183], have applied peptide nucleic acid fluorescence in situ hybridization (PNA FISH) combined with DAPI as a steady method to evaluate, validate, quantify, and characterise the initial adhesion and biofilm formation of three microorganisms: *Salmo-*

orescence in situ hybridisation (FISH).

mg/mL; 10 µL) (Magnification ×100).

*nella enterica*, *Listeria monocytogenes* and *Escherichia coli.* 

**Figure 7.** *Pseudomonas fluorescens* biofilm (24 h) on the stainless steel. Coupons stained with DAPI (0.1 mg/mL; 10 µL) and Rhodamine B (0.1mg/mL; 10 µL) (Magnification × 100). **Figure 7.** *Pseudomonas fluorescens* biofilm (24 h) on the stainless steel. Coupons stained with DAPI (0.1 mg/mL; 10 µL) and Rhodamine B (0.1 mg/mL; 10 µL) (Magnification ×100).

**Figure 8.** *Pseudomonas fluorescens* biofilm (7 days) on the stainless steel. Coupons stained with DAPI (0.1 mg/mL; 10 µL) and Rhodamine B (0.1mg/mL; 10 µL) (Magnification × 100). **Figure 8.** *Pseudomonas fluorescens* biofilm (7 days) on the stainless steel. Coupons stained with DAPI (0.1 mg/mL; 10 µL) and Rhodamine B (0.1 mg/mL; 10 µL) (Magnification ×100).

5.2.5. Confocal Laser Scanning Microscopy (CLSM)

5.2.5. Confocal Laser Scanning Microscopy (CLSM) Confocal laser scanning microscopy (CLSM) is an optical microscope equipped with a laser beam that is particularly useful for examining thick samples like microbial biofilms. Confocal laser scanning microscopy (CLSM) is an optical microscope equipped with a laser beam that is particularly useful for examining thick samples like microbial biofilms. Samples are stained with specific fluorescent dye insofar as the fluorescent light from

illuminated spot is collected on the objective and transformed by a photodiode into an electrical signal to be computer-processed [160] given the complexity of the microbial biofilm's extracellular matrix formed by heterogeneous compounds: polysaccharide, lipids,

whole biofilm matrix owing to its different compositions, which depend on each bacterium and environmental condition, which means that each matrix component must be individually stained. Unfortunately, however, a general stain for polysaccharides does not exist because the chemical structure of matrix polysaccharides differs between distinct

However, no fluorescence labelling method is currently available for visualising the

Extracellular DNA has been related to bacterial attachment and early biofilm formation stages in many species across the phylogenetic tree. These findings were discovered by employing combined stains, such as PicoGreen® and SYTOX®, PI, 1,3-dichloro-7 hydroxy-9,9-dimethyl-2(9H)-acridinone (DDAO), TOTO®-1, TO-PRO® 3. Most reports employed DDAO for staining eDNA in biofilms after the first publications by Allesen-Holm et al. [187] and Conover et al. [188]. Excellent efficacy has been reported for TOTO®- 1, SYTOX® Green, while PI provides the most reliable results. TO-PRO®-3 and DDAO are

With biofilm proteins, which may sometimes be more important than polysaccha-

rides, this occurs in cell wall-anchored proteins in *Staphylococcus aureus* and *S. epidermidis,* and contributes to aggregation by homophilic interactions [190], or interacts with matrix components that originate from the host, such as fibronectin, collagen, or fibrin [191]. These biofilm proteins can be visualised with strains FilmTracer™ SyPro® [192]. Several proteins also play a key role in the *P. aeruginosa* biofilm matrix, such as CdrA and others, perform functions that range from nutrient acquisition to protection from oxidative stress [193]. Moreover, serine-protease inhibitor ecotin has been identified as a matrix protein

enzymes, extracellular DNA, and proteins [186].

bacteria: Gram + and Gram− [186].

not completely cell-impermeant [189].

that binds to Psl [194].

the illuminated spot is collected on the objective and transformed by a photodiode into an electrical signal to be computer-processed [160] given the complexity of the microbial biofilm's extracellular matrix formed by heterogeneous compounds: polysaccharide, lipids, enzymes, extracellular DNA, and proteins [186].

However, no fluorescence labelling method is currently available for visualising the whole biofilm matrix owing to its different compositions, which depend on each bacterium and environmental condition, which means that each matrix component must be individually stained. Unfortunately, however, a general stain for polysaccharides does not exist because the chemical structure of matrix polysaccharides differs between distinct bacteria: Gram + and Gram− [186].

Extracellular DNA has been related to bacterial attachment and early biofilm formation stages in many species across the phylogenetic tree. These findings were discovered by employing combined stains, such as PicoGreen® and SYTOX®, PI, 1,3-dichloro-7-hydroxy-9,9-dimethyl-2(9H)-acridinone (DDAO), TOTO®-1, TO-PRO® 3. Most reports employed DDAO for staining eDNA in biofilms after the first publications by Allesen-Holm et al. [187] and Conover et al. [188]. Excellent efficacy has been reported for TOTO®-1, SYTOX® Green, while PI provides the most reliable results. TO-PRO®-3 and DDAO are not completely cell-impermeant [189].

With biofilm proteins, which may sometimes be more important than polysaccharides, this occurs in cell wall-anchored proteins in *Staphylococcus aureus* and *S. epidermidis,* and contributes to aggregation by homophilic interactions [190], or interacts with matrix components that originate from the host, such as fibronectin, collagen, or fibrin [191]. These biofilm proteins can be visualised with strains FilmTracer™ SyPro® [192]. Several proteins also play a key role in the *P. aeruginosa* biofilm matrix, such as CdrA and others, perform functions that range from nutrient acquisition to protection from oxidative stress [193]. Moreover, serine-protease inhibitor ecotin has been identified as a matrix protein that binds to Psl [194].

Nowadays, confocal microscopy is a relevant tool for studying the structure of biofilms thanks to its excellent real-time visualisation capability of fully hydrated living samples. The limitation of light microscopy's spatial resolution is improved by a fluorescence technique and by coupling CLSM with other imaging techniques [157]. The PNA FISH and CLSM combination allows the spatial organisation of and changes in specific members of complex microbial populations to be studied without disturbing the biofilm structure [195,196].

#### *5.3. Raman Microscopy (RM)*

This non-destructive analytical technique provides fingerprint spectra with the spatial resolution of an optical microscope [197]. This original technique permits the quantitative, label-free, non-invasive, and rapid monitoring of biochemical changes in complex biofilm matrices with high sensitivity and specificity [198]. Raman spectra studies are characterised by high specificity, and by usually revealing sharper clearer bands than IR spectra, and a small water background. Compared to IR microscopy, excitation with visible light can be employed in Raman spectroscopy, which allows standard optics to be utilised. Other advantages include its application to characterise and identify different biological systems (fungi, bacteria, yeasts) because all biologically associated molecules (e.g., nucleic acids, proteins, lipids, carbohydrates) exhibit distinct spectral features [197]. Therefore, Ivleva et al. [197] analysed seven different specific microorganisms by RM to characterise microorganisms in biofilms.

Another author evaluated the antibiotic effect on biofilms [198], and the oxidation of graphene as antibacterial activity against the *Pseudomonas putida* biofilm with variable ages [199].

#### *5.4. Scanning Electron Microscopy (SEM)*

Scanning electron microscopy (SEM) provides useful information about size, shape, and localisation in the biofilms of single bacteria, and in biofilm formation process steps about bacterial interactions and EPS production [200]. Surface topography has been widely discussed as a parameter that influences microbial adhesion. In line with this, the experiments by Kouider et al. [201], which employed SEM to establish the effect of stainless steel surface roughness on *Staphylococcus aureus* adhesion, revealed that the adhesion level largely depends on substrate roughness with a maximum at Ra = 0.025 µm and a minimum at Ra = 0.8 µm. [202]. Mallouki et al. studied the anti-adhesive effect of fucans by SEM and a MATLAB programme to determine the number and characteristics of adhered cells [203].

SEM has been extensively used to qualitatively observe biofilm disruption owing to its high resolution, and is usually applied in combination with biological assays of biofilm removal efficiency [204,205]. With SEM images, simple thresholding cannot often be implemented because biofilm normal surface intensity values are similar due to the same effective contrast seen by SEM. Rough (textured) biomaterial surfaces complicate image analyses, and advanced segmentation methods, such as semi-supervised machinelearning techniques, are usually needed [206]. The biofilm might be segmented from the surface using the Trainable Weka Segmentation plugin, which utilises a collection of machine-learning algorithms for segmentation purposes [207].

As with other previously mentioned techniques, SEM is a widely used resource for confirming the presence of bacteria and the exopolysaccharide matrix when studying biofilms (Figures 9–11). These studies usually obtain SEM results and are supplemented with the results of other techniques like confocal [208,209], surface-enhanced Raman scattering (SERS) spectroscopy [210], epifluorescence microscopy (DAPI/Rhodamine B), and contact plates [211]. *Int. J. Environ. Res. Public Health* **2021**, *18*, x FOR PEER REVIEW 21 of 32

**Figure 9.** Scanning electronic microscopy (SEM) images of stainless steel of 3-day biofilms formed by *Pseudomonas fluorescens*. **Figure 9.** Scanning electronic microscopy (SEM) images of stainless steel of 3-day biofilms formed by *Pseudomonas fluorescens*.

**Figure 10.** Scanning electronic microscopy (SEM) images of stainless steel of 7-day biofilms formed

by *Pseudomonas fluorescens.*

by *Pseudomonas fluorescens*.

**Figure 10.** Scanning electronic microscopy (SEM) images of stainless steel of 7-day biofilms formed by *Pseudomonas fluorescens.* **Figure 10.** Scanning electronic microscopy (SEM) images of stainless steel of 7-day biofilms formed by *Pseudomonas fluorescens. Int. J. Environ. Res. Public Health* **2021**, *18*, x FOR PEER REVIEW 22 of 32

**Figure 11.** Scanning electronic microscopy (SEM) images of stainless steel of 7-day biofilms formed by *Pseudomonas fluorescens* after treatment with peracetic acid (250 ppm) for 15 min. **Figure 11.** Scanning electronic microscopy (SEM) images of stainless steel of 7-day biofilms formed by *Pseudomonas fluorescens* after treatment with peracetic acid (250 ppm) for 15 min.

#### *5.5. Microbiological Methods*

VBNC.

against several bacterial species.

*5.5. Microbiological Methods*  The estimation of the total number of organisms (total viable count) is the most widely used technique to estimate biofilm viable cells. This count is done on agar media The estimation of the total number of organisms (total viable count) is the most widely used technique to estimate biofilm viable cells. This count is done on agar media and its result is colony-forming units (CFU). Based on the serial dilution series approach followed

followed to quantify microorganisms, this technique is easy and requires no special equipment [158]. Surface samples (stainless steel, plastic, rubber coupons) with biofilms are analysed by swab or sonication, and transferred to agar plates. This culture medium can be specific for either the studied species or non-specific species (plate count agar media).

Several authors like [212] discovered that some bacterial species can enter a distinct state called the viable, but non-culturable (VBNC) state. These living cells have lost the ability to grow on plate agar media. However, this method has serious drawbacks and limitations [213]: (i) the fraction of detached live cells may not be representative of the initial biofilm population; (ii) a subpopulation of biofilm cells can be viable, but non-culturable (VBNC), and cannot be detected by the CFU approach for the CFU estimation of the recovery and quantification of viable biofilm cells. Several authors, such as Cerca et al. [214] and Olivera et al. [215], have proposed applying flow cytometry coupled with a few possible fluorophores as an alternative to the total viable count from biofilms because flow cytometry solves both CFU counting limitations by distinguishing total, dead, and

The total viable count technique is fundamental for the evolution of biofilm studies, as are studies about the efficacy of industrial disinfectants and increased resistance to the application of different disinfectants. Table 2 shows some results of disinfectant efficacy to quantify microorganisms, this technique is easy and requires no special equipment [158]. Surface samples (stainless steel, plastic, rubber coupons) with biofilms are analysed by swab or sonication, and transferred to agar plates. This culture medium can be specific for either the studied species or non-specific species (plate count agar media).

Several authors like [212] discovered that some bacterial species can enter a distinct state called the viable, but non-culturable (VBNC) state. These living cells have lost the ability to grow on plate agar media. However, this method has serious drawbacks and limitations [213]: (i) the fraction of detached live cells may not be representative of the initial biofilm population; (ii) a subpopulation of biofilm cells can be viable, but non-culturable (VBNC), and cannot be detected by the CFU approach for the CFU estimation of the recovery and quantification of viable biofilm cells. Several authors, such as Cerca et al. [214] and Olivera et al. [215], have proposed applying flow cytometry coupled with a few possible fluorophores as an alternative to the total viable count from biofilms because flow cytometry solves both CFU counting limitations by distinguishing total, dead, and VBNC.

The total viable count technique is fundamental for the evolution of biofilm studies, as are studies about the efficacy of industrial disinfectants and increased resistance to the application of different disinfectants. Table 2 shows some results of disinfectant efficacy against several bacterial species.


**Table 2.** Resistance of several bacterial species to disinfectant on different material surfaces.

Colony-forming units (CFU)**;** cleaning-in-place (CIP)**;** polypropylene (PP)**;** quaternary ammonium compounds (QAC).

#### **6. Conclusions**

Biofilms have become a major environmental microbiology concern in the food industry over the last 30 years. This topic is prominent due to the potential for contamination of food from biofilms; they are responsible for more than 20% of food poisoning cases and for being up to 1000-fold more tolerant to antibiotics than their planktonic counterparts [219].

Many bacterial species have the ability to form biofilms, such as microbial subsistences (when faced with hostilities from the environment), antibiotics, and disinfectants. For these reasons, cleaning and disinfecting in the food industry must bring about changes that favour eliminating biofilms, because once they form, the resulting costs and risks will be very high. As previously discovered in many publications, the ability of bacteria to form biofilms is greater than the discoveries. Thus, they must be eliminated. The advancement of new, non-destructive technologies (e.g., laser dissection) to study biofilms and their results should be applied to biofilm diagnoses in the food industry, to better understand the physiological anatomy of microbes and biofilms, and future applications in the food industry.

**Author Contributions:** Conceptualization, A.S., C.C., D.R., F.R., and A.R.; methodology, A.S., C.C., D.R., F.R., and A.R.; software, A.S., C.C., D.R., F.R., and A.R.; validation, A.S., C.C., D.R., F.R., and A.R.; formal analysis, A.S., C.C., D.R., F.R., and A.R.; investigation, A.S., C.C., D.R., F.R., and A.R.; resources, A.S., C.C., D.R., F.R., and A.R.; data curation, A.S., C.C., D.R., F.R., and A.R.; writing original draft preparation, A.S., C.C., D.R., F.R., and A.R.; writing—review and editing, A.S., C.C., D.R., F.R., and A.R.; visualization, A.S., C.C., D.R., F.R., and A.R.; supervision, A.S., C.C., D.R., F.R., and A.R.; project administration, A.S., C.C., D.R., F.R., and A.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors are very grateful to their families and friends for all of the support they provided.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Review* **Texture-Modified Food for Dysphagic Patients: A Comprehensive Review**

**Dele Raheem <sup>1</sup> , Conrado Carrascosa 2,\*, Fernando Ramos 3,4 , Ariana Saraiva <sup>2</sup> and António Raposo 5,\***


**Abstract:** Food texture is a major food quality parameter. The physicochemical properties of food changes when processed in households or industries, resulting in modified textures. A better understanding of these properties is important for the sensory and textural characteristics of foods that target consumers of all ages, from children to the elderly, especially when food product development is considered for dysphagia. Texture modifications in foods suitable for dysphagic patients will grow as the numbers of elderly citizens increase. Dysphagia management should ensure that texture-modified (TM) food is nutritious and easy to swallow. This review addresses how texture and rheology can be assessed in the food industry by placing particular emphasis on dysphagia. It also discusses how the structure of TM food depends not only on food ingredients, such as hydrocolloids, emulsifiers, and thickening and gelling agents, but also on the applied processing methods, including microencapsulation, microgels as delivery systems, and 3D printing. In addition, we address how to modify texture for individuals with dysphagia in all age groups, and highlight different strategies to develop appropriate food products for dysphagic patients.

**Keywords:** dysphagia; the elderly; food industry; food products; nutrition; processing; rheology; texture

#### **1. Introduction**

Food colloids are multi-component, multi-phase systems, involving a complex mixture of water, proteins, polysaccharides, lipids, and many minor constituents that contribute to food textures [1]. While eating and swallowing food, sensory tasks require the tongue's motor behavior to explore, squeeze, or move a bolus to ascertain its flow properties [2]. However, eating and swallowing food can pose problems that result in dysphagia; those with this condition are dysphagic patients.

Dysphagia refers to difficulty in swallowing, or sometimes the impossibility of swallow liquid or semisolid/solid food [3]. This condition affects almost 580 million people worldwide, especially infants and the elderly, and it leads to nutritional deficiencies [4,5]. As populations in many developed countries age, the number of dysphagic patients is likely to rise. Approximately 2 billion people will be aged 60 and over by 2050, in many countries, (e.g., Japan, Germany and Korea); around 15% of their populations will be over 80 years old [6]. The older population is the global population's fastest growing segment. Average life expectancy at birth is expected to rise from the present 70 years to 77 years by 2045, with more than 400 million individuals older than 80 years by 2050 [7]. Hence, urgent attention must be paid to the food and nutrition requirements of the elderly, particularly

**Citation:** Raheem, D.; Carrascosa, C.; Ramos, F.; Saraiva, A.; Raposo, A. Texture-Modified Food for Dysphagic Patients: A Comprehensive Review. *Int. J. Environ. Res. Public Health* **2021**, *18*, 5125. https://doi.org/10.3390/ ijerph18105125

Academic Editor: Alberto Mantovani

Received: 5 March 2021 Accepted: 8 May 2021 Published: 12 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

those who are very old and frail. This creates an excellent opportunity for food scientists to respond by formulating food products that meet this demand [8]. Apart from the elderly, and infants whose muscle mass and strength—related to swallowing foods—are weak, those with other medical conditions, such as trauma, cancer, surgery, cerebral palsy, stroke, and other neurological conditions in any age group, may also suffer dysphagia.

Modifying food texture and liquid thickness, without compromising nutritional quality, will play a key role in dysphagia management to ensure that patients can meet their nutritional requirements [9]. For example, studies on bolus rheology by Ishihara et al. [10] suggest that bolus viscoelasticity balance is important to ease swallowing. Other researchers recommended that food texture for dysphagia diets be soft, smooth, moist, elastic, and easy to swallow [11,12]. Handling viscous food components will involve more studies on their rheological parameters. However, a general understanding of the parameters defining texture-modified (TM) food for dysphagia patients worldwide is generally lacking. One extremely important matter for dysphagia management and treatment is to implement the same terminology for it to be universally accepted. A classification system for food viscosity and texture based on sound empirical evidence to help with dysphagia management is necessary. There is a gap in communicating and collaborating among experts in food services and clinical staff. To bridge this gap, in 2012, the International Dysphagia Diet Standardization Initiative (IDDSI) was founded to provide a globally standardized terminology and definitions for TM food and liquids that are applicable to dysphagia individuals of all ages, in all care settings, and for all cultures [13]. We need to conduct more studies to maintain a valid and quantitatively defined scale for different food/fluid textures that can be tested under clinical conditions. Likewise, developing standard recipes for TM food and fluid is also important. For example, in order to provide foods with suitable textures to dysphagic patients, healthcare personnel will have to communicate what this texture is to food producers.

Texture is a sensory multiparameter attribute. It includes all the attributes of the rheological and structural properties of a food product, perceptible by mechanical, visual, auditory and tactile preceptors [14,15]. The roots of the multiparameter attributes of texture lies in its molecular, microscopic, or macroscopic structure. Moreover, certain texture aspects can be seen by the naked eye (e.g., coarse or fine cake texture) or heard by ears (e.g., sounds made when biting on a crunchy celery stalk or a crisp piece of toast) [16]. Dysphagic patients need nutritious foods; such foods need to be of the right texture to improve their consumption and deliver the required nutrients. The need for better intervention strategies is addressed in previous works that target elderly hospitalized patients; this is important because it has the potential to improve patient treatments and outcomes [17]. There are concerns that some TM strategies, such as the IDDSI, do not address the nutritional aspects of foods [9].

Food industries are concerned about variations in taste that come about with changes in viscosity and flow behavior. For instance, evidence suggests that increasing solution viscosity in regular syrup substantially lowers taste intensity, while an increased non-Newtonian flow property observed in light syrup diminishes taste intensity [18]. A better understanding of rheological properties would allow the systemic development of food products to be designed for desired texture and taste interactions. Texture, for many food materials, is a key quality factor. Knowledge gained from the rheological and mechanical properties of various food systems will be relevant for designing flow processes to ensure quality, and to predict storage and stability measurements. Rheological behavior is directly associated with sensory qualities, which significantly influence taste, mouth feel, and stable shelf life. Hence, there is a need for caterers and food scientists to formulate suitable food products for aging populations, which requires a classification system based on rheological properties, consistency, and texture for dysphagia management. These products can be developed for dysphagic patients by blending food ingredients according to personalized recipes for TM food and fluids [19]. To supply dysphagic patients with appropriately

textured food, healthcare personnel have to communicate this texture to food service providers by utilizing the same terminology in dysphagia management and treatment [20].

In order to overcome this dilemma, a guide for TM food was developed, in collaboration with dieticians, speech and language pathologists, and a food company specialized in TM diets (Table 1). The purpose of this guide was to develop different food texture definitions based on several Swedish documents. This guide was influenced by the guidelines developed by the British Dietetic Association in collaboration with the Royal College of Speech and Language Therapists [21,22]. It was a pioneering work in the early years (2000–2002), which objectively defined and quantified categories of texture-modified food by conducting rheological measurements and sensory analyses [23]. However, there are limitations in the work, as analyses were conducted on 15 representative TM sample food items. Moreover, individual medical research will be needed to provide diet recommendations to dysphagic patients.

**Table 1.** Descriptions of the consistencies in the texture guide [23].


Today, the literature on the impacts of TM food, developed by food scientists, on food swallowing, remains scarce. Food processing industries are adopting various treatments including thermal and non-thermal treatments—to modify texture. Future trends will likely include a combination of three-dimensional (3D) printing and drying to design foods, and to enhance textural and sensory characteristics for dysphagic patients [24]. A good starting point to develop these new food products is to gain a better understanding on sensory and rheological characteristics (see Table 1), which will be useful for modifying food texture. The objectives of this review article are to raise awareness about the importance of texture modification in the foods provided to dysphagic patients, describe methods to assess viscosity and texture properties in TM food for dysphagia, and compile those aspects that are related to the nutritional quality of foods for dysphagia. Section 2 describes various textural properties by highlighting methods to assess texture in general, particularly referring to dysphagia. Section 3 describes the varying effects of ingredients and processing methods on food texture. Section 4 discusses texture modification for dysphagic patients. Section 5 offers some food examples developed for dysphagic patients. Finally, Section 6 concludes this review.

#### **2. Methods to Assess Texture in the Food Industry**

The International Organization of Standardization (ISO) recognizes texture as both a sensory quality attribute and a multiparameter attribute. The commonly accepted ISO defines texture as all rheological and structural (geometric and surface) attributes of a food product perceptible by means of mechanical, tactile and, wherever appropriate, visual and auditory preceptors [14].

Texture and rheology are important parameters that need to be assessed when developing food products. One of the physical properties in food technological and sensory analyses that agrees with the ISO definition related to food texture is given by Szczesniak [25] as: "the sensory and functional manifestation of the structural, mechanical and surface properties of foods detected through the senses of vision, hearing, touch and kinesthetics". Texture is one of the key food attributes that is used to define product quality and acceptability [26], and even shelf life. This characteristic is present in all food, can affect its handling and processing, and can even be decisive for both shelf life and consumer acceptance. It will depend on the analyzed food type. Thus raw material food, handling and processing conditions, such as storage temperature, can have a significant influence on, for example, meat textural properties [27]. To understand this physical food property, we should understand the role of rheology in food. It was defined by Steffe [28] as "a branch of physics that studies the deformation and flow of matter". This means that it is the condition under which materials respond to an applied force or deformation, despite the fact that many authors relate rheology to liquid or semiliquid food sensory properties rather than to solids. In cases where swallowing food is difficult, hydrocolloids, which exhibit many functionalities in foods, including thickening, gelling, water-holding, dispersing, stabilizing, film-forming, and foaming agents are useful [29]. They have been used as a texture modifiers in almost all kinds of processed food products [29]. All materials have rheological properties that can be employed to assess raw materials and process characteristics, as well as their behavior and stability, throughout storage time until they are eaten to determine their customer acceptance [30]. This means that rheological analyses are necessary to identify the most suitable foods in accordance with final consumer requirements and to ensure the uniformity of different batches over time [30].

Food rheology has been defined as "the science of the deformation and flow of matter" [31]. Therefore, food texture characterization is no new science. Founded in 1929, the American Society of Rheology has already considered experimental foodstuff rheology [30] and consumers previously employed the food rheology/texture as food quality parameters. Conscientious interest has always been shown in its analysis, which has led it to be characterized in the past with senses by means of sensorial techniques. Nevertheless, this analysis is completely subjective and, thus, it is apt and necessary to perform instrumental analyses, which can relate their results to sensorial tests [32].

Matter starts deforming or flowing only when it is acted upon by forces that may be applied deliberately or accidentally; moreover, there is the all-pervading force of gravity that causes "soft" bodies to flow and lose shape. Rheology is, thus, mainly concerned with forces, deformations, and time [31]. Time matters in many ways, but it is often introduced into measuring rates of changes of deformations and forces. The passage of time does not actually bring about changes in materials. Chemical changes in foodstuff often occur with time, but can be studied by rheological methods. Temperature is also important and frequently appears in rheological equations [31].

In 1958, Blair [31] classified the frequently used instrumental techniques to measure texture into three main groups:


A widely used imitative test today in the food technology field is the so-called Texture Profile Analysis (TPA). The TPA is not only widespread, but also convenient for rapid food texture evaluations [33], although texture can be measured by expert people with

sensorial analyses. This test involves double compression to determine food textural properties. Any food texture identity is rarely a simple matter of understanding a singular characteristic, such as toughness or cohesion. The texture of each food is versatile and related to consumers' sensory expectations. It is not enough to deliver food with the target hardness and elasticity if consumers do not like it and it does not meet their expectations for that food type [34].

Food oral processing is described as a complex and dynamic pathway that involves mechano/chemoreceptors, mixing with saliva, temperature, friction, etc. When thickening formulas for dysphagia are considered, imitation by means of instrumental techniques is difficult as the physico-chemical features of each specific hydrocolloid or food involved in diet will be differently perceived in the mouth [35]. Viscosity is a fundamental property that is obtained from rheological measurements, and is used as the most important criterion in developing thickeners for dysphagia patients. The American Dietetic Association reached an agreement, which was published in the National Dysphagia Diet T [36], and categorizes foods according to their viscosity (at 50 s−<sup>1</sup> ) shear rate range values. The categories are: (1) nectar-like (51–350 cP); (2) honey-like (351–1750 cP); (3) spoon thick (>1750 cP)) to ensure safe swallowing and to facilitate palliative care procedures for different types of patient needs, although the categorization does not consider very relevant sensory aspects. Although viscosity values are obtained at 50 s−<sup>1</sup> , no consensus has been reached by the scientific community about the shear rate value of the swallowing process [36]. A study that considered rheological and tribological responses of biopolymer-based thickening solutions incorporated into different food matrices for dysphagic patients observed that an increase in the biopolymer concentration significantly affected rheological properties as xanthan gum showed the highest viscosity, pseudoplasticity, and viscoelasticity, followed by flaxseed gum [37].

ISO 11036:2020 [38] sensory profiling methods can be used for these attributes. This ISO document specifies a method for developing a texture profile of food products (solids, semisolids, liquids). This method is one approach to the sensory texture profile analysis. Chemical composition determines the basic physical structure of foods which, in turn, influences their texture. An understanding of textural properties will, therefore, require studying the physical structure of foods. Other methods based on physical structure that can offer a description of a product's textural attributes include light and/or electron microscopy, and an X-ray diffraction analysis that provides information about crystalline structure. Differential scanning calorimetry provides information about melting, solidification, and other phase or state transitions, while a particle size analysis and sedimentation methods offer information on particle size distribution and particle shape [39]. Conventional profiling via QDA®, flash profiling and projective mapping performed by panels were used by Albert et al. [40] to describe foods with complex textures. The application of QDA®, flash profiling and projective mapping using panels with different degrees of training helps to overcome issues in the sensory description of served hot food with a complex texture [40].

However, a qualitative empirical method on test conditions that can better measure viscosity is lacking, as the literature on dysphagia indicates [41]. Researchers have used a quick empirical test, the line spread test (LST), to compare relative viscosities of several similar products. It measures the consistency of a liquid using the distance that a standard amount of liquid spreads over a horizontal surface when released from a confined chamber [42].

Dr. Szczesniak developed and improved sensory descriptions for the texture of specific food while searching for more universal descriptors to be applied to a broader array of food. One of the goals was to develop a common lexicon and a set of procedures to allow objective and repeatable sensory texture evaluation tests to be run in different laboratories, with several operators, and for many distinct food types [34]. These experiments described and introduced food sensory analyses as five basic independent mechanical parameters: hardness, cohesiveness, adhesiveness, viscosity, elasticity, and into three more dependent

parameters (brittleness, chewiness, gumminess) [34]. These mechanical parameters [34] can be read from the curve and compared to the observed sensory characteristics. A high correlation between the measurements taken by this technique and sensory evaluations has been shown.

Figure 1 shows a typical TPA graph for food, which is a popular double-compression test run to establish textural food material properties [25] and to quantify mechanical parameters from recorded force-deformation curves.

**Figure 1.** Generalized instrumental texture profile curve modified from Szczesniak [25]. A<sup>1</sup> : positive force area during first compression; A<sup>2</sup> : positive force area during second compression; A<sup>3</sup> : negative force area during first compression; A<sup>4</sup> : negative force area during second compression; Z<sup>1</sup> : height of the maximum force during first compression; Z<sup>2</sup> : height of the maximum force during second compression; Y<sup>1</sup> : time of maxime force during first compression; Y<sup>2</sup> : time of maxime force during second compression; x: time between the first and second compression.

Generally, the parameters observed in the texture profile analysis, i.e., hardness, adhesiveness, and cohesiveness, are used to compare the sensory attributes and rheological properties of various foods. They are employed to examine the material properties of commercial oral moisturizers and denture adhesives, which are relevant to dysphagia [43].

In the curves generated in the two TPA cycles, when foods are chewed over time, as shown in Figure 1**,** hardness (N/m<sup>2</sup> ): the peak force in the first compression cycle (Z1); adhesiveness: negative force area A<sup>3</sup> for the first bite; cohesiveness (J/m<sup>3</sup> ): the ratio between the positive force area during the second compression and that during the first compression (A2/A1); springiness: Y2/Y1.

It is important to emphasize that TPA has not been broadly used in texture measurements for dysphagia as it does not assess some of the core attributes that are relevant to their foods, which are important; they include slipperiness, humidity, and mouth coating. However, other new tests were developed to help complete the knowledge about the physicochemical properties of food in different analytical fields, such as microscopic, submicroscopic, and molecular [44]. This technical progress was assisted by computer science. Indeed, without computer aid, modern spectroscopy, calorimetry, microscopy, and rheological equipment would not have been able to help texture analyses [44].

Nowadays, new texture analyses include a range of food texture-related parameters: firmness, hardness, consistency, fibrosity, tenderness, elasticity, resistance, gel strength, stickiness, adhesiveness, spreadability, bloom force, extensibility, cohesiveness, chewiness extrudability, texture profile analyses, rubberiness, and resilience. Touch characteristics can be classified as mechanical, which measure chewing effort, geometric, related to shape, and others, such as moisture and fat content. Therefore, most of these characteristics are perceived in the mouth if we bear in mind that texture includes all the steps from the first

bite to swallowing [45]. Food mastication covers different processes, including deformation and flow (rheology), size reduction (comminution), and mixing and hydration with saliva. Other physical behaviors that can also be relevant for texture are changes in temperature and surface roughness (rugosity). Food researchers should run rheological tests to describe only a portion of the physical properties sensed in our mouth while chewing [46].

An assessment of rheological properties, particularly in relation to the dysphagia field, includes tests on the flowability or consistency of food. For these tests, a Bostwick Consistometer can be applied to assess the slump of sauces and condiments using a volume of 75 mL, which is released to flow along a channel. The distance traveled by the liquid over 30 s is used to classify consistency [47]. An adaptation of slumping with a reference to dysphagia drinks is called the line-spread test. The IDDSI flow test can be applied using a standard 10 mL Luer slip tip syringe as the "funnel". This test classifies consistency based on the volume of the residual liquid in the syringe after a period of a 10 s flow. The resulting levels are then defined as level 0 thin (0–1 mL liquid remaining), level 1 slightly-thick (1–4 mL), level 2 mildly thick (4–8 mL) and level 3, moderately thick (8–10 mL) [48]. The new International Dysphagia Diet Standardization Initiative (IDDSI) classification system considers practical measurements for liquids that could be used in kitchens, bedsides and in laboratories. In addition, devices capable of modeling human swallowing will provide more accurate measurement information on shear rates during swallowing in dysphagic patients. Clinicians can employ either manometry or video-fluoroscopy for this purpose. With a manometry, a probe is inserted into the patient's pharynx, which obstructs the bolus flow and causes discomfort [49]. During video-fluoroscopic analyses, swallowing of fluids is monitored by X-ray imaging and the entire swallowing process is recorded, which, therefore, enables the examiner to follow the swallowing sequence frame by frame [50].

Studies about food texture rheological properties have been systematically conducted since the early 1950s, while the rheological properties of several food types have been studied, and are summarized in many publications, e.g., for roast turkey breast muscle [51]; Japanese sweets [52], the rheology of food dispersion [53]; and food rheology [54]. Many variables can influence rheological properties, including ripeness, processing methods, temperature, composition, time, instrumental techniques, and analytical assumptions and methods), and modify the results obtained by one test [54]. However, not all tests focus on solving the swallowing problem. Suebsaen et al. [55] prepared banana gels from hydrocolloids for the elderly with dysphagia, modified texture and hardness, to obtain a dessert. This product had different characteristics in instrumental rheology, texture properties, and sensory attributes terms. To improve the swallowing ability of foodstuff, different thickeners are added to normal food and drinks, which may be gum- or starchbased [56].

Food technologists are interested in the mastication process, rheological changes, and other textural properties that occur during this process [46]. For dysphagic patients, sensory tasks that require motor behaviors of the tongue to explore, squeeze, or move a bolus to ascertain its flow properties are challenging tasks related to eating and swallowing foods. In addition to taste receptors in the mouth, trigeminal nerve receptors in the mouth and tongue are capable of detecting both static and dynamic characteristics of items placed in the mouth, such as shape, size, volume, mass, location, temperature, two-point discrimination, and flow or movement [57]. It is, therefore, interesting to understand the sensory function of the tongue for tasks that may be relevant for detecting differences in the flow characteristics of swallowing.

Most of the available information on rheological properties of ready-to-eat dysphagiaoriented products only focuses on viscosity [10]. However, new tests on hardness will be necessary to reveal the effect of elastic modulus on the swallowing ability of solid foods for dysphagia [29].

Several aspects could be considered in the foods for dysphagic patients: the positive effect of dysphagia-oriented products on the quality of life of dysphagic patients; improve their nutritional status and prevent more weight loss. Designing standardized diets for each type of dysphagia is proposed as a desirable approach in rheological studies that are related to the management of dysphagia [29].

#### *2.1. Gels: Rheological Characterization*

Some of the most popular foods, such as gelatin desserts, cooked egg whites, frankfurters, surimi-based seafood analogs, and fruit jellies, can be considered gels. In short, they are solid-in-liquid, and the solid phase immobilizes to the liquid phase [54].

Rheological properties can be measured by (a) puncture test, which is one of the simplest methods to obtain a stress strain curve, and is widely used in both solid and semisolid foods; (b) the torsion test, a method that applies shear stress to samples in a twisting fashion; (c) the folding test, which can be used to measure the binding structure of gels, especially surimi gels, and can be interpreted in cohesiveness terms; (d) the oscillatory test, dynamic rheological testing that evaluates the properties of gel systems, which are suitable for testing the characteristics of gels, gelation, and melting; (e) the stress relaxation test, namely rapid deformation applied to food samples [54]. It can be done while under compression, extension, or shear; (f) yield stress, used for predicting how products respond to processing and/or how they endure performance; (g) rheological characterization of time-dependent fluids, it analyzes the flow behavior (or viscosity) of liquid and semi-liquid food. It is an intrinsic parameter and a measure of fluids' resistance to flow when shearing stress is applied [54].

#### *2.2. Emulsions: Rheology of Food Emulsions*

Emulsions are dispersions of one liquid phase in the form of fine droplets in another immiscible liquid phase. The immiscible phases are usually oil and water, so emulsions can be broadly classified as oil in water or water in oil emulsions, depending on the dispersed phase (mild cream, ice cream, butter, margarine, salad dressing, and meat emulsions) [58], although the rheology of food emulsions is mainly dependent on the strength of inter-droplet interactions and dilute emulsions (that is, milk) have a lowviscosity Newtonian behavior. Nevertheless, concentrated food emulsions show gel-like rheological characteristics [59].

#### *2.3. Rheological Measurements: Equipment*

The rheometer, or viscometer, measures resistance to flow when a known force or stress is produced by a known amount of flow, and is crucial equipment in food rheological studies. Such equipment can be capillary viscometers, falling-ball viscometers, and rotational and oscillatory rheometers, which are used to take rheological measurements [54]. Tests must be carried out under certain conditions for samples, such as steady flow, laminar flow, and uniform temperature [60]. Figure 2 shows the different tests used in rheology.

**Figure 2.** Rheological tests used in food characterization [54].

These rheological measurements use experimentation with and observation of sampled food to compare data, whose main goals are to analyze materials' mechanical properties and identify molecular interactions and foodstuff composition [61]. Nevertheless, these results should be tested with people. Currently, there are two method groups for rheological studies [29], and both are based on measuring force and deformation according to time:


Lack of oral cavity control, poor bolus preparation, or a delayed swallowing response are some reasons for using thicker food and drink for dysphagic patients because thickened foods change the speed at which they are transported through the throat, which is related to delayed swallowing response and, therefore, reduces the aspiration risk [62]. Thickeners, which are typically gum- or starch-based, are added to food to slow down the flow of the bolus. Thickened liquids are highly recommended for dysphagic patients as slowing down the flow rate can provide the time required to close airways [63]. However, excessively thickened food may require much more force on the tongue and pharynx during swallowing [63].

Moreover, we should take into account the texture profile panel, which is a valuable tool for describing and quantifying textural characteristics of food products when the panel is carefully selected, trained, and maintained [64]. Nor should we forget the application from trained panelists or consumer panels in these tests. Thus Saldaña et al. [65] obtained suitable results when sensory hardness correlated positively with instrumental springiness in light mortadella analyses. Other authors, such as Yates [66], performed a descriptive analysis of Gouda cheese texture by a sensory panel and Barden et al. [67] did so on cheddar cheese.

#### **3. The Effects of Processing Methods and Ingredients on Food Texture**

The structure of modified foods depends very much on the ingredients making it up, and also on the processes involved in their development [68]. Based on these premises, it should be noted that the main building blocks for developing most TM food items are carbohydrates, lipids, and proteins [69]

When heated, globular proteins unfold and denature, which increases liquid viscosity (e.g., in protein drinks). They can self-assemble as nano-sized aggregates and fibrils upon additional heating, and eventually become the network chains of gels [70]. Proteins are appreciated not only for these structural applications, but also for certain essential amino acids (e.g., leucine), whose high hydrolysate content tends to facilitate muscle protein synthesis during aging [71].

Polysaccharides have been used as gelling agents to thicken aqueous food dispersions, and to stabilize emulsions and foams [72]. Nishinari et al. [73] offer an exceptional overview of the rheological properties of polysaccharide solutions and gels associated with tasting and swallowing of TM foods. Dextrins include viscous clear solutions that are often employed as thickening agents and encapsulating matrices for nutrients, colorants, flavors, enzymes, and antioxidants, along with starch and gum [74]. Dietary fiber, such as cellulose derivatives (e.g., microcrystalline cellulose), or resistant starch, which may alleviate constipation, can be added directly to food [75]. Although typically used to thicken liquids, starch has been underexploited as a texture-modifier in the pastes and gels utilized as TM food [76]. As starch granules accumulate large quantities of water during gelatinization, they can be preloaded during this process with water-soluble micronutrients and bioactives [77]. Starch can also be partially gelatinized so that various glycemic responses are elicited [78].

Given their amphiphilic nature, phospholipids and monoglycerides can be used as emulsifiers in interfaces or self-associated with a plurality of nano-sized structures (e.g., vesicles and micelles) as bioactive and nutrient delivery vehicles. Triacylglycerol molecules crystallize from a molten state and cluster to form aggregates and, ultimately, a fat network to occlude parts of liquid fat, which ends in a plastic matrix [79]. Food nano- and microemulsion, a topic reviewed recently by McClements [80], can be used to encapsulate and deliver hydrophobic components, such as nutraceuticals, vitamins, and flavors. Oleogels, formed by a liquid lipid process trapped inside a stable gel network, are involved in drug delivery applications as carriers of unsaturated fats and increase food texture [81].

In order to develop TM foods, with a view to retain the overall flavor and appearance of whole pieces while softening their structure, several known technologies achieve this texture-softening effect: freeze-thawing (with/without enzyme infusion), enzyme impregnation, high-pressure processing, pulsed electric fields, and sonication [6]. The regular supervision of process variables maintains the color and flavor of food products, while adjusting their soft texture to various degrees.

Many technologies contribute to small particles and may have applications in TM foods. The food industry has long since been aware of the aggregation and microparticulation of proteins and products employed as thickening agents and fat replacers for beverages and semisolid food [82]. Technologies focus primarily on globular proteins' capacity to undergo denaturation and aggregation in solution, which results in several morphologies (e.g., spherical particles, fibrils, and flexible strands), whose main dimensions range from approximately 10 nm to a few microns [83].

The aim of another group of techniques is developing fibers and soft particles from biopolymer solutions. Microgels are small soft and stable particles (e.g., sizes from <1–100 mm) and come with a wide variety of shapes, sizes, and textural properties that can be tuned to structures [84]. Microgel formation is often performed by direct gelling, often under shear, in a particle or fiber shape, or by reducing bulk gel size by mechanical means. Microgel suspension is typically free-flowing as opposed to bulk gels with prevailing viscoelastic behavior [85].

Apart from their function as texture modifiers, microgels have been proposed as delivery vehicles for non-polar compounds, such as vitamins, flavors antimicrobials and antioxidants, which can be spread in tiny micelles or more functional liposomes (20 nm and a few hundred mm) in the aqueous phase [86]. Given their soft texture and flowability, micron-sized hydrocolloid gel particles with their high water content (e.g., >95%) are very appealing to be employed as structuring agents to consolidate dispersed phases, and also as soup and sauce thickeners. These hydrocolloid microparticles are generally formed by shear gelling or preformed droplet gelation [87]. A recent study showed that a combination of 0.5% Alcalase, and two-step heating at 37 ◦C and 90 ◦C was useful for improving the physico-functional properties of a novel surimi gel for people with dysphagia [88].

Some innovative micro-technology techniques have recently emerged and may lead to revolutionary TM food design and manufacturing applications [89–99]. In channels with cross-sections of a few hundred microns, microfluidic systems handle minute quantities of fluids. Systems have been developed to produce foams and emulsions of identical size and different shapes with a monodispersed discontinuous phase and gel microspheres [92]. 3D printing is a rapid prototyping technique based on digitally-controlled material depositing and layer-by-layer stacking. From "printable" mixtures of carbohydrates it is possible to obtain lipids and proteins, and complex food structures based on liquid deposition or powder binding. According to Kouzani et al. [100], 3D printing reduced design, and fabrication time, improved the consistency and repeatability of 3D printed tuna fish (consisting of tuna, puréed pumpkin, and puréed beetroot), and optimized sensory characteristics of this puréed food for dysphagic patients. Electrospinning employs a high-voltage electrical field to create biopolymer solution electrically-charged jets that become nanofibers upon solvent evaporation. During the encapsulation of bioactives and probiotics, electrospun protein fibers (e.g., <1 µm in diameter) are used as dietary supplements, and also confer food mouthfeel and texture. The nanofibers employed to encapsulate bioactives or as entangled mats to simulate meat are suggested electrospinning technology applications to manufacture TM food [96–98]. Electro-spraying is another electrohydrodynamic manufacturing technique whereby near-spherical droplets are produced from a jet flowing through a nozzle submitted to an external electrical field that yields micro- or nanoparticles upon solvent evaporation. Microencapsulation matrices used to protect biologically active compounds is a suggested use for electro-spraying technology in TM food manufacturing [99].

The relevant role thickeners play in TM food while swallowing, slowing down the flow of liquids and stopping them from being aspired via the airway is highlighted [4]. Currently, starch and gums are the most popular commercial choices. Thus increasing the availability of thickeners to be employed in TM food and extending their properties can be challenging [29].

Gel microparticles are excellent alternatives to tailor food rheological properties thanks to their small tunable size, soft texture and free-flowing state [101]. To be able to change their texture perception and flow behavior, they can be blended into thin liquids or incorporated into purées. They evoke a stronger aroma during mouth breakdown if filled with flavors and supplied with a thin delicate texture [102]. Artificial caviars introduced by molecular cuisine proved to be the most innovative use of soft gel particles. Tiny spheres with a soft core and a tough outer layer were formed by dipping droplets in a calcium bath of colored and flavored alginate solutions [103]. Artificial caviars are now often featured in main dishes, desserts, drinks, etc., and are offered in contemporary restaurants [104,105]. Using tiny "gelatinous" beads and other light-molecular cooking creations (i.e., foam or "air") has been proposed to inspire the elderly to produce attractive TM foods [106].

Recently, the extensive literature on gel microparticles essentially endorses employing them as encapsulating agents and delivery systems rather than applying them to alter texture or to act as major nutritional functions [107,108]. For example, by adding protein microparticles, the texture control of matrices can be achieved [109] and elderly people are likely to try protein-enriched foods if they need a higher protein intake [110]. Conversely, by introducing a dispersed gas phase into bubble form, softness and density can be adjusted [111], which provides the added beneficial effect of a higher perceived intensity of tastants in the gel phase [112]. Insoluble fiber can be filled with gelled microparticles to increase fecal bulk and to prevent constipation, while partially masking the insipid fiber flavor and its rough texture [75,113]. Lastly, emulsion gels are food items in which lipid droplets are enclosed inside a soft biopolymer matrix (e.g., sauces, yogurt, frankfurters, etc.). Gelled emulsion microparticles are small biphasic structures in which a lipid phase offers many opportunities [114]. The incorporation of whey protein isolate (WPI)-based gelled microspheres loaded with lipids into food bars, soups, and other food systems has been suggested by Egan et al. [115]. These microparticles can also be employed as delivery systems for bioactive lipophilic ingredients (fatty acid ω-3, phytosterols, carotenoids, etc.), tastants, and fat-soluble aromas [116]. WPI microgels can lower the plasma insulin peak and postpone the postprandial amino acid profile in relation to protein powder in the interface with drugs [117].

#### **4. Modifying Texture for Dysphagic Patients**

Food contains several phases and hierarchical structures that vary from nanoscopic to microscopic length scales [118]. The configurations offer some features like texture control and nutritional value, or support for processing and shelf-life stability [101]. Texture control and alteration are common ways to control dysphagia. Modified diets are believed to minimize the risk of choking and the need for chewing or oral food processing [119]. Eating thickened fluids is indicated to help safe swallowing as the act of swallowing is delayed and the transit time of food with an altered consistency in modified foods is typically longer than for non-modified foods. This gives the glottis more time to close and avoids food or fluid aspiration to the lungs of dysphagic patients [120].

Food texture can be modulated and altered to meet consumers' nutritional demands. Texture modification and thickening of fluids are normal features of dysphagia evaluation and therapy [121]. TM foods can be defined based on many variables, such as viscosity, density, and fluid flow rate. However, using viscosity to describe thickened beverages for dysphagia management has been questioned as no viscosity measurements are available for most clinicians and caregivers [122].

When designing healthy foods for the elderly, significant factors need to be addressed. TM foods prescribed for seniors' dysphagia management and dietary intake should be soft, moist, smooth, elastic, and simple to swallow [5]. One important key for designing texture and bolus rheology is understanding dynamic food structure changes during oral processing. This rheological state should allow the more cohesively mass flow of bolus throughout the pharyngeal phase to help to improve easy swallowing in dysphagic patients [10].

The IDDSI framework provides standardized terms and descriptions to classify TM food and thickened liquids for dysphagia patients [122]. The IDDSI framework consists in a continuum of eight levels (0–7), as shown in Figure 3. In addition, the syringe flow test classifies IDDSI levels from 0 to 3 based on the flow rate, while a fork pressure test is best used to assess the foods of levels 4 to 7 [5].

**Figure 3.** The IDDSI Framework for the TM food and thickened liquids used with dysphagic individuals from all age groups, in all healthcare facilities and of all cultures [122]. Note. 0: thin; 1: slightly thick; 2: slightly thick; 3: liquidized/moderately thick; 4: puréed/extremely thick; 5: minced and moist; 6: soft and bite-sized; 7: easy to chew/regular.

It is worth noting that foods classified as levels 4 to 7 are texture-modified foods for dysphagic patients [5]. Sungsinchai et al. [5] described the various levels as follows: puréed foods at level 4 do not require chewing, and include products like potato purée, carrot purée, and avocado purée; level 5 (minced and moist) represents soft and moist food with no separate thin liquid; small lumps (of 2–4 mm in size) may be visible in food and minimal chewing is required. Level 5 foodstuff includes items like minced meat and fish, mashed fruit, fully softened cereal, and rice (not sticky or glutinous); level 6 (soft and bite-sized) food that can be mashed and broken down by applying pressure with forks, spoons, or chopsticks that are soft, tender, and moist throughout, but with no separate thin liquid. Chewing is required for this food class, which include cooked tender meat, cooked fish, and steamed or boiled vegetables. Level 7 is regular food with various textures (that can be hard, crunchy and naturally soft).

An example of the TM foods defined in the IDDSI framework is puréed food at the fourth level of the IDDSI framework. Puréed foods are typically ground and/or mixed in a form that involves less chewing and oral manipulation. A cohesive swallowable mass, referred to as 'bolus,' is formed that is easy to push with the tongue to the pharynx [123], which can make swallowing simpler and avoid bolus regurgitation, which causes dysphagia aspiration. Other examples of dysphagia-specific standardized scales when considering TM foods for dysphagia include the Penetration-Aspiration Scale [124] SWAL-QOL and SWAL-CARE [125], the Dysphagia Outcome Severity Scale [126], and the Functional Oral Intake Scale [127].

Using thickeners to improve bolus viscosity in post-stroke oral dysphagia has been proposed as a countervailing clinical technique against aspiration. Nonetheless, this strategy has been questioned because the number of experiments is limited and methodologies vary [128]. One experiment has indicated improved safe swallowing when patients

received altered starch and xanthan gum thickeners with 'spoon-thick' viscosity. The therapeutic effect of these thickeners was due to a counterbalancing process that brought about no major change in swallow reaction timing [129]. Another research study revealed that enhanced bolus viscosity promotes safe swallowing and lowers mid-term pneumonia in patients with oral dysphagia [130]. Some studies have demonstrated that elevated viscosity impairs swallowing effectiveness in oral dysphagia by increasing oropharyngeal residue. Other studies argue that the effect of thickeners on swallow reaction physiology is still not fully understood [129]. Analyzing the effect of augmented bolus viscosity on swallowing safety in patients with dysphagia poses a research challenge. However, novel naturally sourced thickeners from food biopolymers are drawing significant attention and enable on-demand dysphagia management, where fluidal food must be adequately thickened for patients. A recent study investigated the rheological behaviors of a novel thickener with a carboxymethylated curdlan potential for dysphagia, which was a traditional food thickener of konjac glucomannan and its mixtures in both water and model nutrition emulsions. It reported both the efficacy and applicability of these thickened fluids and compared them to those of xanthan gum, taken as the reference. It showed that carboxymethylated curdlan, which is similar to xanthan, displayed a unique viscosity-enhancing ability in both water and emulsions, and proved promising feasible as a novel dysphagia-oriented thickener [131]. Furthermore, the modification of viscosity with thickeners was used as a strategy to circumvent oropharyngeal dysphagia patients' swallowing problems. Generally, the formulations of commercial food products with thickening properties often contain xanthan and starch. However, flaxseed gum was found to improve rheological behavior in liquid foods and can be considered a potential thickener with additional health benefits [132]. These results offer the opportunity to tailor the rheological characteristics of food systems by adding and combining natural ingredients to improve technological and nutritional properties.

Hydrocolloids are often used by the food industry to enhance consistency and cohesiveness, and to reduce TM foods syneresis. Enhanced food consistency and cohesiveness make it safe to swallow [133]. Although all hydrocolloids can be used as thickeners, they are not all capable of forming a cross-linked gel network to be employed to confer modified food solidity. Food mixture thickening that involves hydrocolloids is mainly the result of polymer chains that entangle while their concentration rises. Entanglements in dilute systems are less common, and polymer chains are free to move and viscosity is minimum. After eating thickened food mixtures, saliva dilutes and breaks, which leads to substantially reduced viscosity. Lower viscosity is a problem, particularly when starch-based thickeners are used because saliva contains α-amylase that breaks down amylopectin and amylose [134]. Non-starch gums can be employed to reduce this, even though they do not completely remove undesirable viscosity declination. When non-starch biopolymer gums are resorted to as thickeners, non-specific entanglement can come into play which, above a given concentration, can increase stickiness, which impairs the ability to swallow. Hydrocolloids in TM food have been reported to have an effect on particle breakdown, microstructure, deformation force during mastication, mouth coating, and bolus lubrication [133]. These properties have an implication for oral processing and sensory food perception. Thickened liquids have also been reported to be considerably less palatable than their non-thick counterparts [135]. It is also necessary to produce new thickening agents that are well-defined in terms of sensory properties, and can be employed to enhance swallowing while preserving palatability. This will include a plan to control dysphagia in order to avoid the detrimental effects of decreased palatability and increasing residual viscosity when complying with therapy. In order to increase palatability, TM foods need to be homogeneous in appearance, and particular attention must be paid to their flavor and odor. Adapting the sensory characteristics of dishes to dysphagia in association with cerebral palsy was possible by the check all-that-apply (CATA) method [136]. CATA is faster, more economical, and does not require trained judges. It is sufficiently robust to obtain the profile of a wide range of food products to be developed.

When modifying texture for dysphagic patients, the influence of two natural different hydrocolloids (apple and citrus pectin) on physical, rheological, and textural parameters, bioactive compounds, and antioxidant activity of courgette (*Cucurbita pepo*) purée was studied. Pectin was added within the 0.1–0.3% range to courgette purée and ohmically heated at 20 V/cm for 3 min. Ohmic heating was utilized to improve and preserve the main properties of purées. Antioxidant activity has also increased with ohmic heating, up to 58% compared to the control sample [137]. The study shows the potential of this treatment for ready-to-eat courgettes as food that can be developed for dysphagic patients.

It should be noted that although several hydrocolloids can be used, they have different physicochemical properties, and even different behavior when preparation variables, such as temperature, shearing, and pressure, are applied.

#### **5. Developed Food Products for Dysphagic Patients**

People usually eat raw or cooked foods, but swallowing is the key issue. It starts during the mastication process in the mouth, and passes from oropharyngeal safe food transfer to the esophagus to reach the stomach, where the gastro intestinal digestion process starts to allow nutritional food use [138].

Processing food to increase ease of swallowing requires modifications to texture, and also to physicochemical and rheological properties. For instance, plasma processing is effective in improving the cooking properties of brown rice. Swelling of starch granules due to water uptake not only cuts cooking time, but also softens the cooked rice texture and makes it easier to chew. Bran layer fissure significantly improves water absorption and reduces cooking time [5].

Calorie and nutrient requirements diets for dysphagic patients are similar to those presented by persons of the same age and sex, unless co-existent diseases are present [139]. Just like all people, dysphagic patients require suitable food. As previously mentioned in the introduction, this suitability is not only related to food texture, which needs to be appropriate, but should also offer nutritional value and adequate palatability/acceptability and, if at all possible, it must be visually appealing. Combining all of these characteristics is truly challenging [140].

Thus, if we consider not only hospitalized patients, but also the number of older people [141] living in institutionalized settings, and those with dysphagic problems, the major role of food and pharmaceutical industries in developing TM foods is a welcoming development [142,143] The main food texture characteristics that affect dysphagia management can be classified as [138]: adhesiveness (effort made to overcome food adhering to the palate), cohesiveness (if food is deformed or sheared when compressed), firmness (force needed to compress semisolid food), "fracturability" (force required to break solid food), hardness (force required to compress food to attain a certain deformation), springiness (rate or degree that food goes back to its original shape after being compressed), viscosity (rate of flow per unit of force) and yield stress (minimum shear stress applied before flow begins) [134]. The recommended food texture of dysphagic diets must be, at least, smooth, moist, soft and elastic if we contemplate that these attributes should combine TM food rheological properties and patients' difficulty swallowing [29,119].

The IDDSI framework shown in Figure 3 is a useful guide when a dysphagic patient's diet is considered. Variations based on individual circumstances may exist. The Functional Diet Scale, in addition to this framework (IDDSI–FDS), permits levels 2–5 as being suitable food and drinks for dysphagic patients. In a Canadian study with adults living in long-term care institutions, IDDSI Functional Diet Scale scores were derived based on diet orders and were compared between residents with and without dysphagia. The IDDSI–FDS for residents with no dysphagia risk ranged from 4 to 8, which reflects the lack of severe diet texture restrictions, while the probability of having an IDDSI–FDS score of <5 was significantly higher in individuals at dysphagia risk [142]. When foods are prepared or formulated for dysphagic patients, it is important for the bolus to be swallowed safely if it is not chewed. Thus particle size and moisture content of food are key criteria, especially for minced and moist foods at level 5. Simple and inexpensive tests at home and in residential care or nursing settings recommended by the IDDSI are: the spoon tilt test to ensure that food is not too dry or sticky; the fork drip test to guarantee that the food is not too runny [9]. In many German nursing homes, minced and moist texture diets are available, which are easy to produce because only a blender and no special knowledge are needed. In these settings, the puréed texture is the most elaborate because they should be lump-free and require special equipment (e.g., a bowl cutter) for several food types (e.g., meat) because of natural fiber content [144].

In one intervention study conducted in a long-term care facility in Canada, the presentation to dysphagic patients of developed foods based on texture and shape resulted in increased body weight, and higher energy and nutrient intake, after 12 weeks in residents receiving reshaped TM diets compared to a control group on unshaped TM diets. Another study in the USA showed a 15% higher food intake after changing to the 3D preparation of puréed foods [145]. This shows that reshaping food components enhances the visual appeal of meals significantly and they are more likely to be eaten. Therefore, it is essential that either liquid or solid food is modified for them to offer appropriate nutritional properties to make swallowing easy for dysphagic patients. Different strategies have been applied to achieve this goal. The main ones can be summarized by following thermal processing and non-thermal technologies, and employing thickeners [139,146].

The simplest form of thermal processing is using hot water, which is known to be effective in transforming hard food into soft food. It is also known that some nutrients are especially heat-sensitive and using thermal processing leads to marked vitamin loss, especially in food rich in these essential nutrients like fruit and vegetables [147]. It is noteworthy that this thermal processing type is often used at home and in industry. To solve this problem, food and pharmaceutical industries usually apply two strategies: first, addition of the micronutrients lost from processing; second, using non-thermal technologies. The most widespread thermal technologies applied to obtain TM food are pulsed electric field, high-pressure processing, high hydrodynamic pressure, ultrasound, and gamma-irradiation [5].

The above-mentioned non-thermal technologies can be applied to meat [148], fish, or its by-products [11], rice [63], starch, and carbohydrate-based products [149], or fruit [150] and vegetables [151]. The use of non-thermal technologies helps to maintain bioactive compounds (especially heat-labile compounds) in food and, thus, promotes health benefits for dysphagic patients. As dysphagic patients benefit from soft food that is safe for swallowing, the characteristics and gel properties of starch play an important role in the desired final product quality.

Irradiation can increase gelatinization temperature, water solubility, water absorption capacity, and oil absorption capacity, but can lower peak, trough, final breakdown, and setback viscosities in starch-based foods. Irradiation has been shown to induce the depolymerization and destruction of the crystalline structure of chickpea flour, which resulted in gamma-irradiated flour being cooked more easily with less retrogradation [5].

The texture of solid food can be generally classified into four grades, as shown in Figure 3. Nevertheless, if regular/unmodified everyday food is not considered, then only three categories are useful for dysphagic patients: (i) "soft"—food is naturally soft (e.g., ripe banana) or cooked or cut to alter food texture; (ii) "minced and moist"—food easily forms into bolus using only the tongue; (iii) "smooth puréed"—food is cohesive enough to maintain its shape on a spoon, similar to the consistency of commercial puddings [152].

No international harmonized terminology is available for thickened liquids, although four or five categories have been defined according to respective viscosity values [4]. However, if the so-called water-like viscosity (<50 cP) is excluded, in a very simplistic form, and according to the "fork test" [63], it is possible to classify thickeners into three texture grades: (i) "nectar"—can be drunk in a cup or with the help of a straw (51-350 cP); (ii) "honey"—can be drunk in a cup, but not with a straw (351–1750 cP); (iii) "pudding" should be eaten with a spoon (>1750 cP). [4,139].

Commercial TM food was developed to improve nutritional intervention in dysphagic patients. To minimize the risk of aspiration and dehydration, ready-to-serve commercially packaged pre-thickened (CPPT) and instant food thickeners (IFT) are used to modify beverage consistency in dysphagia management [20]. The test of masticating and swallowing solids (TOMASS), an international study that performs quantitative solid bolus ingestion assessments [153], and in vitro testing, such as that performed by Mathieu and co-workers [154] or by Qazi et al. [155], are relevant tools that contribute to assess that these food types do not require further preparation by patients' families and/or caregivers.

Thus, in commercial terms, food and pharmaceutical companies have made different product types available on the market, which can be summarized as:


The variety and supply of TM foods targeted at elderly consumers in Asian countries is more promising. The market in Japan is steadily expanding; in South Korea, the market value of the "senior-friendly" food industry in 2010 was around USD 4 million and growing at a rate of 11% per year [6]. The guidelines for TM foods in Japan have been issued by several initiatives, such as food for special dietary uses (FOSDU), the dysphagia diet 2013, and "Smile-Care" foods [6]. Several companies have a special product line that consists mainly in thickened beverages and purées for individuals with swallowing disorders. TM foods offer food companies the opportunity to tailor-make products with soft textures because the products for this market segment have been slow to appear in Europe [162].

Finally, diets should be as varied as possible, and ought to supply sufficient energy and protein. In addition, dishes ought to be pleasantly presented to encourage whetting people's appetite. Servings should be small and frequent rather than a few copious meals a day. For such purposes, molecular gastronomy [163] and 3D printing technology [100] have been used to produce food from various raw material sources with a variety of textures to enhance diet and to make them more palatable and esthetically appealing. In order to apply 3D printing to food, it is necessary for the food material to possess suitable rheological characteristics to allow its extrusion and for it to be cohesive enough to maintain its shape. However, further research into the application of selected non-thermal technologies as a means to modify food texture for subsequent 3D printing will be worthwhile [5]. In the future, it is envisaged that the food industry will advance toward convergence technology by the utilization of digital solutions, such as machine learning to food systems, as shown in the results of a recent study. This suggests a pioneering framework to identify the rheological levels of foods for the elderly by combining experimental results with machine learning technology in the food application domain [164].

#### **6. Conclusions**

Food texture modifications are essential to suit the nutritional diets of dysphagic patients. It is necessary to gain a better understanding of the complex factors that influence the colloidal food matrix from a multidisciplinary perspective. Individual and household food service operators, including nursing homes, need to acquire better knowledge about food texture, nutrition, and sensory properties. The elderly and dysphagic patients require a sourcing of special foods that are not only soft and easy and safe to swallow, but are also nutritious and tasty. This is vital for them to achieve their nutritional needs. Current food product development initiatives on the TM foods industrial scale also need to employ novel technologies to ensure dysphagic patients' access to appropriate TM food products. Additional quality criteria and clinical guidelines that target dysphagic patients, and based on the rheological parameters discussed in this review, need to be introduced by the food industry, and healthcare and catering services. In the very near future, it is hoped that more food processors will engage in the commercialization of cost-effective TM foods to put innovative technologies in this field to the best possible use.

**Author Contributions:** Conceptualization, A.S., C.C., D.R., F.R. and A.R.; methodology, A.S., C.C., D.R., F.R. and A.R.; software, A.S., C.C., D.R., F.R. and A.R.; validation, A.S., C.C., D.R., F.R. and A.R.; formal analysis, A.S., C.C., D.R., F.R. and A.R.; investigation, A.S., C.C., D.R., F.R. and A.R.; resources, A.S., C.C., D.R., F.R. and A.R.; data curation, A.S., C.C., D.R., F.R. and A.R.; writing—original draft preparation, A.S., C.C., D.R., F.R. and A.R.; writing—review and editing, A.S., C.C., D.R., F.R. and A.R.; visualization, A.S., C.C., D.R., F.R. and A.R.; supervision, A.S., C.C., D.R., F.R., and A.R.; project administration, A.S., C.C., D.R., F.R. and A.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors are very grateful to their families and friends for all of the support they provided.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


International Journal of *Environmental Research and Public Health*
