Sustainability benefits of transitioning from current diets to plant-based alternatives or whole-food diets in Sweden

The study addresses how replacing animal-source foods (ASFs) with plant-based alternatives (PBAs) or whole foods (WFs) in Sweden affects environmental impacts, nutritional adequacy, and daily food expenditure. Motivated by the need for large-scale dietary changes to keep food systems within planetary boundaries while ensuring nutrition, the paper notes that many modeling studies show benefits of shifting toward plant-based diets, typically via whole foods like legumes. However, consumer barriers related to taste, convenience, and cultural values hinder transitions away from ASFs. PBAs—processed foods designed to mimic the sensory properties of meats, dairy, and seafood—may facilitate such transitions. Prior work has often been product-level or focused mainly on meat substitutes, with limited comparative assessment against WFs and little country-specific analysis. Sweden is a pertinent case given high ASF consumption with associated environmental and health burdens and a rapidly growing PBA market. The research aims to model BAU and plant-forward scenarios (vegan, vegetarian, flexitarian) using PBAs or WFs, under both mass- and energy-based functional units, to quantify environmental, nutritional, and economic implications.

Previous research indicates that plant-based dietary shifts can reduce environmental pressures and improve health, but most analyses replace ASFs with WFs rather than PBAs. Comparative LCA studies generally show PBAs have lower environmental impacts than their ASF counterparts, yet often higher than minimally processed WFs. There remains uncertainty about nutritional and long-term health effects of PBAs, with mixed evidence ranging from concerns about sodium and processing to findings that fortified PBAs can be nutritious alternatives to processed ASFs. The PBA market has expanded rapidly in Europe, suggesting growing relevance for diet-level assessment. Existing diet modeling typically focuses on subsets of PBAs (mainly meat alternatives) and compares only to ASFs, seldom including dairy and seafood alternatives or contrasting PBAs with WFs. There is also a need for country-specific studies reflecting local dietary patterns, environmental challenges, and cultural contexts. This study addresses these gaps by assessing a broader range of PBAs (meat, seafood, dairy, snacks) already available at retail, comparing them to WFs as replacements, and situating the analysis in Sweden with national data and nutrient recommendations.

Design: A modeling study constructing a current average Swedish diet (BAU) and six plant-forward scenarios replacing ASFs with PBAs or WFs, assessed under two functional units (mass and energy). System boundaries were cradle-to-consumer, including cooking at the consumer stage when necessary (e.g., for dried legumes). Diet construction: The BAU diet was derived from the Swedish Statistical Database representing household and institutional consumption; FAO food loss and waste factors were applied to reflect actual consumption. Six scenarios were developed: vegan with PBAs (VGNPBA), vegan with WFs (VGNWHOLE), vegetarian with PBAs (VGTPBA), vegetarian with WFs (VGTWHOLE), flexitarian replacing 50% of ASFs with PBAs (FLXPBA), and flexitarian replacing 50% of ASFs with WFs (FLXWHOLE). For WFs, upper reference values (e.g., legumes) from the EAT-Lancet Commission were used. Scenarios were constructed on both mass (grams) and nutritional energy (kcal) bases to explore functional unit uncertainty. Nutritional analysis: Each food item was paired with nutrient composition data (per kg or L) from the Swedish National Food Agency database covering 55 macro- and micronutrients. Nutrient adequacy was evaluated against Nordic Nutrition Recommendations for adults (most stringent where sex-specific), with essential amino acid requirements from FAO/WHO and EFSA (adult, 70 kg reference). Median nutrient values per item were used, with 25th/75th percentiles for uncertainty. Environmental impact analysis: Life-cycle environmental impacts of GHGe (kg CO2e), cropland use (m2), and consumptive freshwater use (L) were calculated. Primary Swedish food item LCA data were from Moberg et al.; for legumes, soy foods, and PBAs, additional Swedish consumer-guide sources were used. Where PBA data were missing (e.g., PB cheese, PB seafood, PB snacks), data were sourced from foodDB and harmonized into a single inventory. Some proxies were employed (e.g., fish fingers for PB seafood). Blue water consumption from agriculture was used where total water data were unavailable. Economic analysis: Retail price data (June 2022) were extracted from Willys (Axfood group), computing median prices for n=78 items aggregated into 15 food groups. Prices reflect market values and exclude cooking energy costs. Household waste at the purchase stage was not added, aligning with consumption-based modeling. Uncertainty analyses: (1) Functional unit comparison (mass vs energy) across nutrition, environmental, and cost outcomes. (2) Variability within food types using 25th/75th percentile nutrient values and environmental impacts, and price ranges from retailer data to reflect consumer choice for cheaper/more sustainable/more nutritious options.

- Nutritional adequacy: All six plant-forward scenarios improved multiple nutrients vs BAU: iron (+15–47%), fibre (+36–163%), folate (+13–160%), magnesium (+13–147%), and polyunsaturated fats (+26–92%); saturated fats decreased (−13–38%). Protein and zinc intakes remained above recommendations in all scenarios, though both decreased vs BAU (protein −6–22%; zinc −20–64%), lowest in VGNWHOLE and VGNPBA. Calcium met recommendations in all scenarios except VGNWHOLE and FLXWHOLE. Vitamin B12 fell below recommendations in VGNWHOLE and VGNPBA. Vitamin D decreased in all scenarios; only VGNPBA and FLXPBA remained adequate (due to fortified PBAs). Selenium was below recommendations in all diets, including BAU. Sodium exceeded recommendations in BAU and increased further in PBA scenarios (+11–24%) but decreased to recommended levels in WF scenarios. - Environmental impacts: Relative to BAU, vegan diets had the largest reductions. GHGe: −52% (VGNPBA) and −56% (VGNWHOLE). Land use: −44% (VGNPBA) and −32% (VGNWHOLE). Flexitarian reductions were moderate: GHGe −30% (FLXPBA), −20% (FLXWHOLE); LU −22% (FLXPBA), −27% (FLXWHOLE). Freshwater use decreased moderately (−14–26%) in most scenarios; remained similar in FLXWHOLE and VGNWHOLE due to higher fruit/vegetable intake. In BAU, ASFs provided 37% of energy but contributed 75% of GHGe (meat alone 43% GHGe) and were major contributors to LU (67%) and WU (38%). In VGNPBA, total GHGe dropped ~50%, with PBAs contributing 37% of remaining GHGe; PB snacks dominated (26%), largely driven by dark chocolate assumptions. - Daily food expenditure: BAU median cost was SEK 82/day (Q25–Q75: 66–105). Shifting to PBA diets slightly increased costs (+3–5%), except VGTPBA (similar to BAU). WF diets decreased costs (−4–17%), with VGTWHOLE yielding the largest reduction (−17%). In BAU, ASFs were 50% of expenses (meat 28%). PB meats and PB dairy had median prices similar to ASF equivalents; PB seafood was ~20% higher than seafood. - Uncertainty and functional units: Energy-based functional unit yielded slightly smaller nutrient differences vs BAU and higher estimated GHGe (+7–23%; 2.6–3.7 vs 2.1–3.4 kgCO2e), but scenarios still reduced GHGe 22–45% below BAU. WHOLE scenarios’ impacts rose slightly on an energy basis due to higher legumes/nuts shares; in PBA scenarios, PB dairy impact rose while PB meat fell. Costs slightly increased under energy basis except VGTPBA; vegetarian scenarios showed lowest costs. Within-item variation showed large price ranges (e.g., seafood +70% Q75 to +150% Q100), with BAU becoming most expensive if maximum prices were chosen.

The findings demonstrate that reducing ASF intake is the primary driver of environmental impact reductions, with both PBA- and WF-based replacements conferring substantial benefits; thus, choosing between PBAs and WFs is less consequential than the magnitude of ASF reduction. Nutritionally, all scenarios improved fibre, folate, PUFA, and reduced saturated fat, aligning with public health needs in Sweden. However, PBA scenarios increased sodium intake relative to BAU, reinforcing concerns over salt levels in processed PBAs; WF scenarios decreased sodium to recommended levels. Vitamin D declined across scenarios except those with fortified PBAs (VGNPBA, FLXPBA), consistent with Nordic contexts where supplementation is recommended. Selenium inadequacy persisted across all diets, reflecting broader Swedish concerns. While fortified PBAs helped maintain micronutrient profiles near BAU, questions remain regarding bioavailability and protein quality, warranting further research. Environmentally, vegan scenarios produced the greatest GHGe and LU reductions. The dominance of PB snacks (linked to dark chocolate) in VGNPBA impacts highlights the importance of product choice and the limitations of proxy data; oat-based or lower-impact confections could reduce this contribution. Economically, modeled diets were cost-competitive with BAU; PBAs slightly increased median costs while WFs decreased them, aligning with prior findings for high-income settings. Discrepancies with Swedish Consumer Agency estimates likely reflect use of a discount retailer, June 2022 price timing amid later inflation, and exclusion of household food waste and cooking energy in cost calculations. Methodologically, the study contributes by: (1) assessing a broad suite of retail-available PBAs (meat, dairy, seafood, snacks) in diet-level models and comparing them to WFs; (2) aligning nutrient assessment with Nordic Nutrition Recommendations; (3) evaluating both mass- and energy-based functional units; and (4) using country-specific data for contextual relevance. From a policy and behavior perspective, PBAs and WFs should be viewed as complementary strategies to reduce ASFs across diverse consumer groups. Including healthier PBAs in dietary guidance could help accommodate preferences, provided attention is given to sodium reduction, amino acid quality, and nutrient bioavailability. Affordability assessments should extend beyond food cost to broader socioeconomic factors affecting access and purchasing patterns.

Transitioning from current Swedish diets to more plant-based diets—via PBAs or WFs—can substantially lower greenhouse gas emissions, land use, and (often) freshwater use, while meeting most nutrient recommendations and remaining broadly cost-competitive. PBAs can facilitate shifts for consumers prioritizing convenience and sensory attributes, though current PBA products often have elevated sodium and may require continued fortification to cover nutrients of concern (e.g., vitamin D, B12). The results underscore prioritizing reductions in ASF consumption over the choice between PBAs and WFs, while encouraging diversification of WFs and reformulation of PBAs toward healthier profiles. Future research should expand environmental indicators, assess nutrient bioavailability and protein quality in PBA-rich diets, incorporate sociocultural dimensions of affordability and acceptability, and stratify analyses by demographic and socioeconomic groups. Continued inclusion of a wider array of retail-available PBAs (e.g., seafood, dairy, eggs, cheese, snacks) will refine understanding of their role in sustainable dietary transitions.

Key limitations include: (1) Consumption data gaps for PBAs, tofu, and legumes in national databases led to setting these to zero in BAU, likely underestimating current fibre and altering environmental/economic baselines. (2) The study did not evaluate bioavailability of fortified nutrients or processing effects; PBAs as processed foods can be reformulated (e.g., sodium reduction), but ultra-processing may carry health risks. (3) Environmental assessment was limited to GHGe, land use, and blue water use due to sparse PBA data; proxies were used for some PBAs (e.g., fish fingers for PB seafood), and total water use was not assessed. (4) Price data were conservative: derived from a discount retailer (Willys) in June 2022, not reflecting inflation thereafter; cooking energy costs were excluded; scenarios are consumption-based and omit household food waste at purchase. (5) Population-level adoption was assumed without stratification by gender, age, or socioeconomic status; acceptance and preferences for PBAs and WFs vary across groups, affecting real-world uptake.