Fake meat or meat with benefits? How Dutch consumers perceive health and nutritional value of plant-based meat alternatives

P.J


Introduction
Animal agriculture has a substantial impact on the planet and population health. Current food production is responsible for more than 25% of the global greenhouse gas emissions and contributes to deforestation (Tilman & Clark, 2014). Meat production requires large amounts of water, land use, and other resources, which degrades and depletes the environment thus reducing earth's stability (Steinfeld et al., 2006). The unsustainable production of meat is evoked by a high demand for meat products. In many cultures, meat is an expensive but core dietary component. Meat consumption typically increases with increasing gross domestic product (GDP), except for various high-income countries where meat consumption already exceeds dietary recommendations (Gerhardt et al., 2020).
The high demand for meat jeopardizes health and well-being of both consumers and livestock animals. Excessive consumption of red and/or processed meat is a risk factor for non-communicable diseases, such as cardiovascular diseases (Cui et al., 2019), some forms of cancer (Larsson & Wolk, 2006), metabolic diseases, obesity (Fretts et al., 2015), and all-cause mortality (Larsson & Orsini, 2014). A high level of meat consumption also involves a high number of animals being slaughtered. It requires increasingly industrialized meat production, which promotes problematic animal rearing conditions (Robbins et al., 2016). Therefore, aiming to reduce the demand for animal-based products helps to create a more sustainable, equitable food system and to improve population health (de Boer & Aiking, 2022).
One method to reduce animal-based products consumption is by substituting meat products with plant-based meat alternatives. Meat alternatives often resemble meat with more corresponding sensory features than traditional protein rich plant-based alternatives such as tofu and legumes (Rubio, Xiang, & Kaplan, 2020). Meat alternatives are known by various synonyms including meat substitutes, meat analogues and plant-based meat. Defining a meat alternative can be challenging as different dimensions should be considered, such as production, sensory characteristics, nutritional value, sustainability, and consumer expectations (Ketelings et al., In Press). For the purpose of this study, we define meat alternatives as plant-based food products that aim to resemble meat in terms of sensory and nutritional aspects and are explicitly intended to replace meat as meal component.
Globally, the demand for meat is still very high, but the popularity of meat alternatives has increased markedly recently (Piernas et al., 2021). Despite the increasing demand for meat alternatives, research on consumers' perceptions of meat alternatives' healthiness and nutritional value is limited. There is a large body of evidence claiming nutritional benefits of plant-based diets, but this cannot be directly translated to meat alternatives (Tso et al., 2020). Many meat alternatives are classified as ultra-processed foods according to the NOVA classification (Bohrer, 2019;Monteiro et al., 2019), products that are often rich in salt and saturated fats. Therefore, some nutritionists argue or even warn that current meat alternatives in general are not healthier than meat products (Gibney et al., 2017). From a nutritional point of view, meat alternatives appear to be preferable in terms of nutrients like fibre, saturated fat and cholesterol (Bryant, 2022).
Meat is a prominent source of vitamin B12, zinc, and iron in an omnivorous diet (Godfray et al., 2018). Further, meat is typically protein dense, and this protein content is difficult to match on amino-acid level with plant-based meat alternatives since different sources of plant protein need to be combined. Even if it is matched, the bioavailability of plant protein is much less compared to its animal counterpart (Neacsu et al., 2017).
Manufacturers tend to highlight micronutrient and protein content on labels of meat alternatives as it allows them to signal that their product is 'meat-like' in terms of nutritional value (Lacy-Nichols et al., 2021). But such labels are not found on meat products and the potential risk of these labelled nutrition claims, then, is that it may lead consumers to overestimate the nutrition and health benefits of meat alternatives. Indeed, nutrition claims can shape the knowledge of consumers concerning the perceived healthiness of products, making food products with nutrition claims generally seem healthier (Oostenbach et al., 2019).
The current study was executed among Dutch consumers. We aimed to test the hypotheses that: (1) consumers perceive meat alternatives to be healthier than meat products; (2) consumers perceive meat alternatives to be higher in protein and fibre, and lower in salt and saturated fat compared to meat; (3) a nutrition claim (high in protein) promotes overestimation of a product's overall nutrient content and healthiness, particularly for meat alternatives; (4) perceived differences in nutritional value do not match with actual differences between meat products and meat alternatives as currently sold in supermarkets. Further, we explored consumers' need for health and nutrition information when buying food, in particular when buying meat alternatives.

Ethical considerations
All study procedures were approved by the Ethics Review Committee Inner City faculties (ERCIC) of Maastricht University (ERCIC_327_16_02_2022). Supplemental materials of the study can be found on OSF (https://osf.io/2cy5v/?view_only=5588b1e59ac64 502b11114418eee33de).

Study participants
A sample size calculation was performed using G*Power 3.1 to estimate the minimum number of participants needed in this study (Faul et al., 2007). Using an α rejection criterion of 0.05 and power of 0.90 (1-β) with an estimated medium effect size f = 0.3, a minimum sample size of 97 participants is required. People could participate if they are 18 years or older and speak the Dutch language. A total of 120 participants completed the study questionnaire via Prolific in April 2022. Participants who failed the attention checks in the online questionnaire or did not fill out the questionnaire completely were excluded from further analysis. Participants were informed about the study (i.e., the general aim and the exact procedure) and provided consent beforehand. Note, they were not informed about the specific aims or hypotheses of this study until after data collection was completed.

Study procedure
An online questionnaire was created in Dutch. The questionnaire consisted of four parts. Firstly, questions about grocery shopping behaviour and health status were addressed. Participants rated the selfreported use of sixteen different types of information typically found on food packaging (including price, shelf life, and health value) by answering the question: 'When buying food and drink products, how often do you look for the following information on the packaging?' on a 7-point Likert scale with 1 = 'Never' and 7 = 'Always'. The full list of characteristics can be found in the full translated questionnaire on OSF. This first part of the questionnaire also included questions about meat and meat alternative consumption frequency. Meat alternatives were defined to the participants as products that aim to resemble meat in terms of sensory aspects and are intended to replace meat as a meal component.
With the second part of the questionnaire, participants were randomly assigned to a 'claim' or 'no claim' group. Participants in the claim group were shown meat and meat alternative products with the nutrition claim 'high in protein'. In both groups, eight products with standardised packaging and labels were presented in a random order (four meat products and four meat alternative products) and participants had to indicate how much they agree with a statement about the respective product on a 7-point scale (1 = 'Completely disagree' to 7 = 'Completely agree'). The following statements were shown for each product: 'I would buy this product', 'I perceive this product as healthy', 'I perceive this product as sustainable', I perceive this product as natural', 'This product contains a lot of salt', 'This product contains a lot of protein', 'This product contains a lot of saturated fat', and 'This product contains a lot of fibre'.
Thirdly, the general health beliefs regarding meat and meat alternatives were assessed (with 7-point rating scales) on the following statements: 'Meat alternatives are healthy' and 'Meat is healthy'. Further, to examine what participants take into consideration when they judge meat and meat alternatives based on healthiness and sustainability, they had to rank order six different product characteristics (from most '1' to least important '6'): packaging design, taste and other sensory features, communicated information (labels and/or claims), ingredients, nutritional value, and organic origin.
Sociodemographic questions were asked in the fourth part of the questionnaire to collect information about age, gender, education level, and diet (whether consumers identified as vegan, vegetarian, flexitarian, pescatarian, or meat eater). After piloting the survey, it was uploaded to the online research platform Prolific. Prolific was designed for online research and maintains a large research participant pool allowing for participant screening prior to the study (Palan & Schitter, 2018). We estimated an overall duration of 10 minutes for participants to complete the questionnaire. Participants were remunerated (£1.24; £7.50/hour) via their Prolific account.
Nutritional values and nutrition claims on meat and meat alternatives were collected from three supermarket chains in the Netherlands (Jumbo, Albert Heijn, and EkoPlaza) between March and May 2022. Supermarket websites were searched for meat and meat alternative products. Next to the supermarkets' categorisation of products, an additional manual search was performed to ensure all the available products were included. Meat products included in the database were burgers, sausages, mince, and chicken. Products that are not specifically created to imitate meat products but are available in the dedicated meat alternative supermarket section, such as tempeh, tofu, and falafel, were not considered in this study.

Data analysis
To examine whether participants perceive meat alternatives to be healthier than meat and how that might be affected by the nutrition claim 'high in protein', we planned to conduct a split-plot, 2 (Product: meat versus alternative) × 2 (Claim: yes or no) ANOVA. The same twoway ANOVA was then repeated for willingness to buy and for each nutrition belief, that is, expected (1) salt content, (2) protein content, (3) saturated fat content, and (4) fibre content. Note that we did not include product sustainability and naturalness ratings in the data analysis plan as these outcome variables are not directly relevant for the current aim and hypotheses in this study. We included those variables to be able to explore a potential effect of a nutrition claim on the perception of sustainability and product 'naturalness', results of which may be reported elsewhere.
Further, we examined what product features participants are most likely to consider when assessing the health value of both meat and meat alternatives. Participants rank ordered characteristics from most to least important, as described above (in section 2.2). Friedman tests were conducted to determine for each product whether the ranking of product features deviated from random. In case of significant deviation, we fitted a Plackett-Luce model to further understand the relative importance of each product feature (Turner et al., 2020).

Results
Participant characteristics are displayed in Table 1. Importantly, no significant differences were found in age, gender, education level, and proportion of dietary identity between the group of participants who evaluated meat and meat alternatives with the nutrition claim 'high in protein' and the group who evaluated the same products in the absence of such a claim. These evaluation results are described in the sections below.

Willingness to buy and health beliefs
No significant effects were found when asking participants about how willing they were to buy (WTB) a meat or meat alternative product (largest F[1, 118] = 2.90). Also, the presence or absence of the nutrition claim 'high in protein' did not appear to influence the willingness to buy either product. A boxplot of these results is shown in Fig. 1, panel A. A significant main effect of Product (alternative versus meat) on health beliefs was found (F[1, 118] = 16.72, p < 0.001, η 2 G = 0.05). This implies that the participants generally believed meat alternatives to be healthier than meat. No effects of the nutrition claim 'high in protein' were found, (smallest p = 0.51). The results are shown in Fig. 1, panel B.

Nutrition beliefs
Salt. A significant main effect of Product (alternative versus meat) in expected salt content was observed (F[1, 118] = 8.02, p < 0.01, η 2 G = 0.03). This shows that the participants generally expected meat to contain less salt than a meat alternative. No claim effects or other effects were found (smallest p = 0.14). A boxplot of the results are shown in Fig. 2 was found F([1, 118] = 8.12, p < 0.01, η 2 G = 0.05. This implies that the participants generally expected products (both meat and meat alternative) to have a higher protein content when a packaging label claims that a product is "rich in protein". No other effects were found, with the smallest p = 0.09. A boxplot of the results are shown in Fig. 2, panel B.

Perceived versus actual nutritional value of meat alternatives
In Table 2, a comparison is presented of nutritional values between meat and meat alternative products available in Dutch supermarkets. When analysing protein, saturated fat, fibre, and salt content, several differences between product types were found. Meat alternatives contained significantly less protein and less saturated fat than meat. The levels of fibre and salt were significantly higher in meat alternatives compared to meat.

Consumers' assessment of a meat (alternative) product's health value
Need for health information. We first examined the need for health information in our panel of participating consumers by ordering the selfreported need for information on certain product aspects. Fig. 3 depicts the mean score of participants reporting to look for a specific item of product information when buying groceries, scoring each item on a Likert scale ranging from 1 (never) to 7 (always). The aspects considered most often were 'price', 'quantity', 'shelf life', and 'nutrients'. Health information was relatively infrequently considered when buying groceries.
Assessing healthiness of meat products. Participants rank ordered six product features they might consider when evaluating the health value of meat products from 1 (most likely to consider) to 6 (least likely to consider). Mean rank scores per product feature (+SD) are displayed in Table 3. A Friedman test indicated that the pattern of ranks significantly differed from random, Q(5 df) = 121.44, p < 0.001. Next, we fitted a Plackett-Luce model (PLM (Luce, 1959;Plackett, 1975;Turner et al., 2020); to the data with mean feature worth as reference to further examine the relative importance of the product features. Feature worth corresponds to the ranks provided by the participants such that features receiving a higher rank will have a higher worth value. Table 3 displays the worth estimates (+SE) for each product feature reflecting the importance of a feature relative to the average ranking across features.
Ingredients and nutrition information were both ranked significantly higher than average, whereas package design and organic origin were ranked significantly lower. The estimated probabilities that each feature received the highest rank are also displayed in Table 3, showing that information on ingredients was most likely to be ranked first. The ratio of the residual deviance (1429.9) and the degrees of freedom (1795) was 0.80, hence suggesting good fit of the model to the data (Agresti, 2013).
Assessing healthiness of meat alternatives. All participants ranked the features they consider when evaluating the health value of meat alternatives. Mean rank scores per product feature (+SD) are displayed in Table 4. The pattern of ranks significantly differed from random, Q(5 df) = 168.40, p < 0.001.
A PLM was fitted to the data with mean feature worth as reference. . Likert scores regarding meat alternatives and meat products presented either with (triangles) or without (circles) protein nutrition claim. Note: individual Likert scores were slightly jittered to limit overplotting.

Fig. 2. Expected degree of salt (A)
, protein (B), saturated fat (C), and fibre (D) content regarding meat alternatives and meat products presented either with (triangles) or without (circles) protein nutrition claim. Note: individual Likert scores were slightly jittered to limit overplotting.
General information and information on ingredients or nutritional value were all ranked significantly higher than average, whereas taste, package design, and organic origin were ranked significantly lower (see worth estimates with corresponding Z-and p-values in Table 4). Further, ingredients information was most likely to be ranked first. PLM goodness of fit was determined by calculating the ratio between residual deviance and the degrees of freedom. This ratio was 0.76, suggesting good fit.

General discussion
In this study, we found that Dutch consumers tend to view plantbased meat alternatives as being healthier than meat products. Vural, Ferriday, and Rogers (2021) showed similar findings with their study indicating that both meat-and non-meat eaters rate plant-based burgers and chicken nuggets as healthier compared to the conventional meat products (Vural et al., 2021). The current study showed that, despite these beliefs, consumers did not have a greater willingness to buy meat alternatives instead of meat. This may not be due to unfamiliarity with consuming meat alternatives, since the majority of the total sample of participating consumers (i.e., 90%) indicated having tried meat alternatives at least once. It is more likely that the willingness to buy meat alternatives is restrained by a higher price and lower palatability compared to meat. Indeed, most plant-based meat alternatives are more expensive than meat products and we found that price is a prominent  Fig. 3. Ordered bar plot showing mean self-reported frequency (x-axis, Likert score: 1 = 'never'; 7 = 'always') with which participants reported looking for specific items of product information (displayed on the y-axis) when buying groceries.

Table 3
Mean rank score per meat product feature when evaluating healthiness. For each feature, mean rank (+SD), worth estimate (+SE), Z-and p-values corresponding with the test that its worth departs from average, and the estimated probability that it is selected as most important feature are displayed.  Table 4 Mean rank score per meat alternative product feature when evaluating healthiness. For each feature, mean rank (+SD), worth estimate (+SE), Z-and p-values corresponding with the test that its worth departs from average, and the estimated probability that it is selected as most important feature are displayed. attribute consumers consider when buying food (Good Food Institute, 2022). Moreover, palatability of plant-based meat alternatives tends to be below that of meat products (Lee et al., 2020;Tso & Forde, 2021). Future research needs to provide further insights into how much price and palatability exactly affect willingness to buy meat alternatives. Our panel of consumers did not express a strong need for health information in general. However, when they do examine health aspects, nutrition information and ingredients are primarily considered. It remains unclear whether consumers completely understand these types of information. Previous work has indicated that this understanding may depend on several factors like country, age, and social economic status (Grunert et al., 2010). Many studies have focussed on the understanding and overall effect of front-of-pack health labelling. However, results about its effectiveness remain inconclusive (Egnell et al., 2018;Folkvord et al., 2021;Jürkenbeck et al., 2022).
We hypothesised that consumers perceive meat alternatives to be higher in protein and fibre, and lower in salt and saturated fat compared to meat. Notably, the current sample of consumers did seem to incorrectly estimate some of the nutritional values of plant-based meat alternatives when compared with the actual nutritional values based on our assessment of meat alternatives sold in three different supermarket chains in the Netherlands. Specifically, consumers appeared to overestimate the relative protein content of meat alternatives, generally expecting meat alternatives to contain as much protein as do meat products. These misunderstandings are unwarranted when juxtaposing the results from the questionnaire and supermarket product database.
Several researchers have investigated the nutritional composition of plant-based meat alternatives and compared those to their meat counterparts. Similar to our results, plant-based burgers were found to provide lower levels of saturated fat and protein, and higher levels of fibre and sodium (Cole et al., 2022). Another survey in the UK showed that meat alternatives generally have a more favourable nutrient profile compared to their meat counterparts, specifically for the total amount of fat, saturated fat and fibre content (Alessandrini et al., 2021). In line with our findings, Alessandrini et al. also found lower levels of protein and higher levels of salt in meat alternatives. Dissimilar to our findings, a nutritional comparison in Australia showed significantly lower levels of sodium in plant-based meat analogues (Melville et al., 2023). Salt content in meat alternatives compared to meat has been a point of discussion over the past years, where results remain inconclusive. Some studies indicate higher levels of salt for meat alternatives, other researchers report lower levels, and some researchers find no difference in salt content at all (Cole et al., 2022;Curtain & Grafenauer, 2019;Gastaldello et al., 2022;Melville et al., 2023). Nutritional values can vary substantially between individual meat alternative products. For example, we found that salt concentration in alternative meat products sold in three Dutch supermarkets ranged from as little as 0.15g/100g -2.60g/100g. Whilst our results show significantly higher levels of salt in meat alternatives, the absolute difference is small. With the current developments in processing technology and formulation of food products, meat alternatives have the potential to improve their nutritional profile by, for example, lowering salt content (Bryant, 2022).
There are not only differences in the nutritional values of the meat and meat alternative products, but also whether the product bear a health-or nutrition claim. Nutrition and health claims can be used by food business operators to highlight a beneficial effect of their products, in relation to health and nutrition, on the product label or in its advertising (Lacy-Nichols et al., 2021). We found that meat products are generally more protein rich than meat alternatives, but meat does not bear a 'high in protein' nutrition claim on their label. In other words, consumers might erroneously infer that meat alternatives with a 'high in protein' nutrition claim have an even higher protein content than meat products have. Since the nutrition claim 'high in protein' positively affected protein content expectancy and did so for both types of products to an equal degree. This effect did not spill over to bolstering a health belief or willingness to buy the product. This suggest that the nutrition claim of high protein content simply informed the consumer and nothing more.
As the demand for plant-based meat alternatives continues to grow, it is essential to understand its nutritional and health implications. Policymakers should be made aware of the potential nutritional misguidance about both meat and meat alternative products to consumers. Policies could be put into place for example to prevent this misguidance, since our study shows that nutrition claims will not rectify consumers' health and nutrition beliefs. Therefore, it is either important to raise these nutritional differences to the consumer or urge meat alternative producers to alter the nutritional content of their plant-based products to match the meat product it is intended to replace.

Strengths and limitations
A strength of the current study is its comprehensive approach. This is the first study to put consumer perceptions of meat and meat alternatives into contrast of products found in the supermarket. It also uses an extensive dataset of meat alternatives and meat products currently available in Dutch supermarkets. Another strength is that the panel also appears to consist of the main target consumer for meat alternative products, generating transferable results (Szenderák et al., 2022). However, the study is not free from limitations. Within the meat and meat alternative categories are differences with respect to level of processing, from a relatively unprocessed meat cut to an ultra-processed sausage. This could have an influence on the healthiness perception of the individual products, which was not taken into account in the present study. Furthermore, only the products containing a nutrition claim about micronutrients had information about its respective micronutrient content. All information of other products only contained the mandatory nutritional values. Therefore, no comparison between the products could be made regarding micronutrients.

Conclusion
The current study suggests that Dutch consumers consider plantbased meat alternatives as 'meat with benefits'. These benefits refer to the beliefs that meat alternatives are more environmentally friendly and healthier. That latter belief, however, warrants further investigation as although we found plant-based meat alternatives have higher levels of dietary fibre and less saturated fat, they have lower levels of protein and higher levels of salt compared to their meat counterparts. ". These results can thus assist policymakers and food business operators to create an environment that is fair, transparent and understandable for the conscious consumer with respect to the nutritional content of meat alternatives.

Ethical Review Committee Inner City Faculties -Maastricht University
After examination of the research study protocol entitled 'Consumers' attitudes and health perceptions towards meat substitutes' and relevant annexes, submitted by Dr. Alie de Boer, the Ethical Review Committee Inner City faculties (ERCIC) has concluded that there are no ethical objections to the execution of the research project.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
A link to Open Science Framework is included in the manuscript to give access to additional materials used for the study.