Differences in perspectives on sustainability attributes of dietary protein sources between reduced animal-based dieters and nondieters

Since the UN defined the Sustainable Development Goals in 2015, policies have increasingly targeted more sustainable production and consumption, including nutrition. Sustainable diets are those that promote health and wellbeing while exerting low environmental impact and remaining accessible, affordable, safe, equitable and culturally acceptable. Consumers are central to the transition, but evidence on how different dietary groups perceive the sustainability of protein sources is limited. Protein-rich foods strongly drive environmental footprints, making them a focal point for sustainable food choices. Front-of-pack nutritional scores are promoted across parts of Europe and may aid healthier choices, while emerging eco- or sustainability labels aim to steer consumers toward lower-impact foods, though skepticism exists. Prior choice-experiment studies have assessed label attributes affecting meat and meat-substitute choices, with some evidence that environmental cues (e.g., carbon footprint) are less influential than expected. This study addresses these gaps by comparing perceived sustainability (healthiness, environmental friendliness) and the relative importance of sustainability-related attributes (product category, nutritional score, ecological score, price) for eight common protein sources among two consumer groups in Hungary: nondieters (NDs) and reduced animal-based dieters (RABs). The novelty lies in focusing on RABs as a distinct, heterogeneous cohort (ranging from vegans to flexitarians), enabling an in-depth view of perceptions versus a mixed-diet group.

The paper situates its inquiry within literature on sustainable diets, front-of-pack labelling, and consumer choice experiments. Nutritional scoring systems (e.g., Nutri-Score-like approaches) are recommended for clearer nutrition communication and may improve perceived healthiness and purchase intentions. Environmental labelling, including eco-scores and carbon footprint indicators, has been tested with mixed effects: consumers often struggle to assess environmental impacts of foods and may undervalue environmental cues relative to other attributes. Studies on meat and substitutes show product category and brand often outweigh environmental metrics; carbon footprint can have low relative importance. Research calls for segment-specific analyses to tailor communication and labelling strategies, as consumer demographics and diet orientations (e.g., attitudes toward plant-based diets) shape perceptions and choices.

Design: Cross-sectional online survey conducted in Hungary via Jotform from April 7 to May 9, 2022. Recruitment used snowball sampling on Facebook (including paid ads) targeting demographics overrepresented among plant-based dieters. Sample: 639 respondents; after exclusions (<18 years; diet due to disease), N=541. Groups by self-reported diet: nondieters (NDs, n=247) and reduced animal-based dieters (RABs, n=294). RABs included vegans, vegetarians, flexitarians, and others reducing animal-based foods. Participation was anonymous and voluntary; ethical approval was obtained (protocol 102/2022). Survey components: (1) demographics; (2) perceived sustainability attributes (health and environmental friendliness) for eight protein sources on 5-point Likert scales; (3) food frequency questions for the same eight foods; (4) choice-based conjoint (CBC) tasks comparing product profiles varying by sustainability-related attributes. Stimuli/protein sources: poultry (chicken breast), pork (leg), cheese (Port-Salut), milk (2.8% fat), eggs (whole), fish (pangasius), almonds (plain), tofu (plain), selected per national dietary guidelines and population consumption patterns. Conjoint attributes and levels: product category (8 levels), nutritional score (present vs not), ecological score (present vs not), price (present vs not). Full factorial (8×2×2×2=64) was reduced using orthogonal design to 16 products (Set A) and paired with alternatives (Set B) to form binary choices plus a "no choice" option. Task prompt asked respondents to choose the more sustainable protein source given the information shown. Scoring systems and data sources:

• Portion correction: All metrics adjusted to typical portion sizes (from USDA FNDDS and validated by three MSc nutritionists).
• Nutritional score: SAIN/LIM-based algorithm (SAIN: potassium, protein, dietary fibre; LIM: energy, saturated fat, sodium, added sugar). RDIs from EFSA (corrected to 2000 kcal for added sugar and SFA); added sugar estimated via published methodology. Foods classified A–E by percentile cuts from a 25-food reference database (A healthiest, E least healthy).
• Ecological score: Integrated from carbon, ecological, and water footprints (Fondazione Barilla Double Pyramid). Each component normalized by deviation from the mean, averaged to an integrated ecoscore, and classified A–E (A most environmentally friendly, E least).
• Prices: Collected from Auchan Hungary online store; portion-adjusted. Analysis: Group comparisons of demographics via chi-square tests; food frequency via chi-square; perceived health and environmental friendliness via Mann–Whitney tests (non-normality by Shapiro–Wilk p<0.001 and unequal variances by Levene p<0.001). CBC analyzed with conditional fixed-effects regression (conditional logit) in R (survival package), estimating effects of product category, nutritional score, ecological score, and price; effects interpreted via exp(coef.) with significance at p<0.05.

Sample profile: Majority women (81%); RABs were 54% of sample. Significant differences between NDs and RABs in gender, age, family status, and food responsibility (all p<0.05). Younger age groups skewed toward RABs; NDs predominated in older groups. Most respondents lived in cities and had higher education. Sample not representative of the general population. Food consumption: Significant group differences for all eight categories (all p<0.001). Chicken: NDs mostly consumed 1–2 times/week, while most RABs never consumed. Tofu: most NDs never consumed; most RABs 1–2 times/week. Many RABs reported never consuming animal-based sources; among those who did, eggs and cheese were most common. Almonds were frequently consumed among RABs. Perceived healthiness (5-point Likert): All sources differed significantly between groups (p<0.001). Both groups: almonds rated healthiest; pork least healthy. RAB means (examples): almonds 4.8±0.5; tofu 4.3±0.8; fish 2.6±1.4; eggs 2.4±1.4; cheese 1.9±1.0; chicken 1.9±1.2; milk 1.6±0.9; pork 1.5±0.9. ND means: fish 4.2±1.0; eggs 4.1±1.0; chicken 3.7±1.0; cheese 3.3±0.9; milk 3.3±1.1; tofu 3.5±1.0; almonds 4.5±0.7; pork 3.0±1.0. Overall, NDs rated healthiness higher across products (ND all products mean ≈3.7±1.1 vs RABs ≈2.6±1.6). Perceived environmental friendliness: Except almonds (both 4.0±1.0), all differed significantly (p<0.001). Both groups: almonds most environmentally friendly, pork least (RABs 1.3±0.6; NDs 2.3±1.0). RABs: tofu 4.0±1.0 tied with almonds; eggs 2.0±1.0; fish 1.7±1.0; cheese 1.5±0.8; chicken 1.5±0.8; milk 1.4±0.7. NDs: eggs 3.3±1.0 and fish 3.0±1.5 higher than cheese/chicken/milk (≈2.6–2.7). Overall, RABs rated environmental friendliness lower (RABs ≈2.1±1.4; NDs ≈3.0±1.2). Conjoint (conditional logit; exp(coef.) reported):

• Nondieters (n events=3952): Positive significant effects: almonds 4.96; tofu 3.90; eggs 3.11; fish 2.22; milk 1.30; nutritional score 1.37. Non-significant: poultry 0.93; pork 1.04; price 0.91. Negative significant: cheese 0.34; ecological score 0.66. Likelihood ratio test=1246 on 11 df, p<0.001.
• RABs (n events=4704): Positive significant: tofu 11.90; almonds 10.85; nutritional score 1.19. Negative significant: eggs 0.76; fish 0.36; milk 0.27; cheese 0.16; pork 0.14; poultry 0.20; ecological score 0.75. Price 1.15, p=0.058 (ns). Likelihood ratio test=4526 on 11 df, p<0.001. Interpretation: Nutritional score presence increased choices in both groups; ecological score presence decreased choices. Plant-based categories (tofu, almonds) strongly increased choice; all animal-based categories reduced choice among RABs; among NDs, eggs and fish increased choice while cheese decreased it. Price had no significant effect in either group.

Findings show clear diet-type differences in perceived sustainability of protein sources. RABs rated animal-based products as less healthy and less environmentally friendly than NDs, while NDs tended to underestimate environmental impacts of protein sources. Both groups agreed almonds were healthiest and most environmentally friendly and pork the least, yet the environmental impact of nuts (notably freshwater footprint) may be underrated by consumers. Eggs and fish were viewed as the most favorable animal-based options in both groups, though RABs assigned substantially lower values. Milk and cheese were perceived poorly—especially by RABs—despite a relatively favorable nutritional score for milk, suggesting a negative bias and ongoing controversy about dairy health effects. Conjoint evidence demonstrates that front-of-pack nutritional information can nudge choices toward products perceived as more sustainable, aligning with prior work on nutrition labels. Conversely, the ecological score had a negative effect on choice—likely because many protein sources carry higher environmental burdens and an explicit eco classification (especially C–E) may deter selection. Product category remains a dominant driver of perceived sustainability. Price did not significantly influence sustainability choices in this design, potentially due to portion-based pricing and overshadowing by the score cues. Overall, the results indicate the need for tailored communication: RABs may benefit from balanced information on the health attributes of animal-based products, while NDs require better education on environmental impacts. Improving consumer understanding of multidimensional sustainability (health, environment, economic and sociocultural aspects) and the meaning of eco scores could help align perceptions with evidence.

This study compared nondieters and reduced animal-based dieters in Hungary regarding perceived sustainability of eight common protein sources and the influence of sustainability-related attributes in choice. Nutritional scores positively affected choices in both groups, while ecological scores reduced the likelihood of selecting products as more sustainable. Plant-based categories were the strongest positive drivers of choice for both groups, and all animal-based categories negatively influenced choices among RABs. Perceptions differed markedly by diet: RABs tended to underestimate the healthiness of animal-based products, and NDs underestimated the environmental impact of protein sources. These insights support targeted communication and labelling strategies—especially the combined use of clear nutrition labels and carefully framed environmental information—to promote healthier and lower-impact choices. Future work should examine how different consumer segments interpret and act on combined nutrition and environmental labels, refine eco-scoring communication to reduce misunderstanding, and consider cultural and dietary contexts in updating food-based dietary guidelines.

The sample was not representative of the general population (overrepresentation of women, younger adults, urban residents, and highly educated individuals due to snowball sampling via social media). Portion sizes used for score and price normalization lack international standardization, which may affect comparability. Price was presented in portion-corrected terms and may not have been salient, potentially attenuating price effects. The ecological and nutritional scoring systems were study-specific (though evidence-based) and derived from multiple data sources, which may introduce measurement and classification limitations. Cross-sectional, self-reported perceptions are subject to bias and cannot infer causality. The study context is Hungary, which may limit generalizability to other cultures and markets.