Complete Food and Home Cooking

Complete foods are powders formulated to be a whole diet: mixed with water they supply calories, protein, fat, fibre and the full set of vitamins and minerals. Huel, Jimmy Joy, Mana and Soylent are the best-known brands, and most of the European ones are vegan. Part of their appeal is environmental, since nothing spoils, portions are exact and nothing is cooked. The paper compares them with a regional, seasonal, whole-food vegan diet cooked at home in Germany. Both sides are vegan on purpose: dropping animal products is already the largest lever on a diet’s footprint, so holding it fixed leaves the narrower question of whether the form the plants arrive in changes anything.

The short answer is that it changes little. Both options land somewhere between 1 and 3 kg CO₂e per 2000 kcal, roughly one day of food for one adult. A meat-heavy diet measured the same way comes to about 10 kg (Scarborough et al. 2023). Choosing to eat plants at all is a factor of four; choosing between a shake and a stew is a factor of three at most, and much of that is measurement uncertainty rather than a real difference.

Why the comparison is harder than it looks

Food footprints come from life-cycle assessments, and two choices decide what such an assessment says. The first is the system boundary, where the accounting stops: at the factory gate, at the supermarket shelf, or only after cooking and waste at home. The second is the unit the footprint is expressed per. A complete food is a dry powder at about 417 kcal per 100 g, while fresh vegetables are mostly water. Compared per kilogram, the powder looks about ten times better purely because of that water. The paper therefore expresses everything per 2000 kcal and adds household waste, cooking energy and the trip to the shop as separate lines, so that a comparison can be re-run at a different boundary.

There is also very little to compare against. No independent, peer-reviewed assessment of any complete-food product exists. Huel publishes an absolute figure and has its emissions inventory independently assured, but it does not publish the method behind the figure. Jimmy Joy reports an amount of offsetting, which is not a measurement of a footprint. The other brands publish nothing usable. Every number about these products is either a manufacturer claim or a reconstruction from ingredient data.

Refining is where the footprint comes from

The paper reconstructs one product, Jimmy Joy’s Plenny Shake, from its declared ingredient order and nutrition panel, and applies verified ingredient emission factors from CarbonCloud’s ClimateHub database. The result is about 1.0 kg CO₂e per 2000 kcal, with a plausible range of roughly 0.9 to 1.3. The step people picture as “industrial food”, mixing the powders in a factory, is about 0.2 % of the total. The footprint is already in the ingredients when they arrive.

What makes an ingredient expensive is refining it. A whole soybean carries 1.22 kg CO₂e per kilogram; turned into protein isolate it carries 7.23, about six times as much, and 84 % of that comes from processing energy rather than farming (CarbonCloud, soy protein isolate). The same pattern holds for hydrolysed oat flour and for oil dried into a powder. The consequence is that two products in the same category can differ from each other more than either differs from home cooking: a formulation built from flours should sit near the bottom of the range, one built from protein isolates near the top. The vitamin and mineral premix is also not the rounding error it is assumed to be. At 15 kg CO₂e per kilogram it is more carbon-intensive than beef per kilogram, and even at about 2 % of the product’s weight it contributes roughly a sixth of the ingredient footprint (CarbonCloud, premix).

Huel’s own figure works out at about 3 kg CO₂e per 2000 kcal, three times the reconstruction. A different recipe built around isolates explains perhaps half of the gap. The rest most plausibly comes from a company-level figure divided by meals sold, which includes offices, travel and marketing; that would make the manufacturer’s number conservative rather than flattering. Without a published method this stays a hypothesis.

The home-cooked side

Two model days of 2000 kcal each were built from the French Agribalyse 3.1 database: one optimised around German seasonal and stored winter produce, one reflecting an ordinary vegan shop with imported and greenhouse items. Including waste, cooking and shopping they come to about 2.0 and 2.6 kg CO₂e. Three results matter more than the totals.

Seasonality is a large effect, and it comes from heating rather than distance. Agribalyse puts an out-of-season tomato at 3.4 times the footprint of a seasonal one, and almost all of the difference is in the agricultural stage, the heated greenhouse, while transport barely moves. A single portion of winter tomato is the largest line in the ordinary day. Transport, on the other hand, is a large share of a small number for plant foods, which is why food miles mislead: the share is high only because growing plants is cheap. Cooking energy, finally, is the largest line outside the food itself, about half the footprint of the food in the optimised day, and most published comparisons leave it out. It is also, together with waste, where a powder has a genuine structural advantage.

Health

On health the evidence is missing rather than balanced. No study has followed healthy adults living mainly on a complete food for longer than four weeks. The usual objection, that these products are ultra-processed, does not transfer cleanly: in a large European cohort the higher disease risk attached to ultra-processed food concentrated in animal-based products and sweetened drinks, while plant-based alternatives showed no detectable association (Cordova et al. 2023). Against that stands the one randomised trial that isolated processing itself, in which people ate about 500 kcal a day more on an ultra-processed diet without being asked to (Hall et al. 2019). Nobody has run that experiment with a complete food.

The home-cooked vegan diet has a documented weakness of its own. In the EPIC-Oxford cohort, vegans had 2.31 times the hip-fracture rate of meat eaters (Tong et al. 2020), which points at nutrient adequacy rather than at plant food as such. A fortified product is designed to close exactly that gap, but whether it does has never been tested.

What follows

Environmental impact is not a good reason to choose either way of eating. The levers that actually move the number cut across both: avoiding heated-greenhouse produce out of season, not wasting food, not driving to the shop, and how the cooking is powered. Where the motivation is convenience, cost or reliable micronutrient coverage, the environmental evidence neither supports nor blocks a complete food, and partial substitution captures most of the waste and cooking-energy advantage while keeping the variety of a mixed diet.

The full paper, with the ingredient tables, the reconstruction, both model days and every source, is the PDF on this page.