Results of the SNi DELTA Model have this week been published in the Journal of Nutrition. The paper details the construction of the model, the scope of its use, and some key results that challenge widely repeated ideas in the food system.
As detailed in the paper, the DELTA Model is an online tool that allows users to design global food system scenarios (in terms of production, waste and use) and see the outcomes for human nutrition. For example, a user can design a scenario with increased food production and decreased food waste, which the model will then use to calculate the nutrients available to the global population. This value is then compared to the nutrients that the population requires, calculated using demographic data coupled with age and gender specific nutrient requirements.
How does the model work?
DELTA was developed using publicly available datasets from international organisations, complemented with key information from the scientific literature. It takes global food production totals and runs them through a calculation pipeline:
- Allocation: food items are allocated to their uses. This includes use as animal feed, processing into other food or non-food commodities, seed for following growing seasons, and non-food use (such as sugar crops for biofuel production). The amount wasted along the supply chain is also deducted.
- Consumer use: a substantial amount of food matter is not consumed, either because it is considered non-edible (such as animal bones and vegetable peel), or it is simply thrown away uneaten.
- Conversion to nutrients: food composition data are used to convert the total amount of food consumed into a total amount of nutrients consumed. 29 essential nutrients are included.
- Bioavailability scaling: specific nutrients are scaled for bioavailability, i.e. the ability of the body to utilise these nutrients when consumed in certain foods.
The total amount of bioavailable nutrients is then compared to the requirements of the global population. This can either be today’s population, or a forecast population in the future, and DELTA considers the demographic makeup of these populations when calculating nutrient requirements, because not all individuals have the same nutrient needs.
What are the results of the model?
In this paper, the DELTA Model was used to address a number of different questions.
Where are the gaps in our current food production system?
Using 2018 data, it was found that there was sufficient availability of most nutrients to feed the global population. The only exceptions were calcium and Vitamin E. There was even sufficient macronutrients (e.g. protein, energy, fat) to nourish the 2030 population. This demonstrates that the problems of undernutrition present in the world are partly due to the inequitable distribution of food.
What about food waste?
The DELTA Model found that even completely removing food waste doesn’t solve all our nutritional problems. There still wouldn’t be enough calcium or Vitamin E for the global population. This is because we waste different amounts of different nutrients, so reducing waste has a varied impact on nutrient availability.
What foods should we be producing in the future?
The DELTA Model wasn’t designed to be prescriptive, so no optimal food production system is given. Instead, a few example future scenarios are discussed, each of which fails to meet nutritional needs in some way.
Scaling up food supply with the population doesn’t resolve nutrient gaps, it just keeps them from growing any larger.
Removing meat and seafood production to achieve a globally vegetarian diet increases the food available to people due to reduced animal feed demand, but leaves gaps for key nutrients for which these foods are major contributors, such as iron, zinc and Vitamin B-12.
Increases in plant food production can help to feed a growing population, but using this technique alone to meet nutrient requirements may come with an excess intake of energy.
Finally, halving waste by 2050 is an admirable goal, but not the whole answer. Doing so would keep macronutrients above requirement, but would leave many micronutrient deficiencies.
What’s novel about DELTA?
There exist several other models for global nutrition which perform a similar role to DELTA (e.g. GENuS, The Global Nutrient Database, Beal et al.). The key points of difference and novelty in DELTA are:
- Accessibility: few of the alternative models are openly available for the general public to use to gain a better understanding of global nutrition. Moreover, many of the alternative models use data that is not publicly accessible, making interpretation more challenging.
- Bioavailability of nutrients: the DELTA Model takes into account the fact that 100g of a nutrient from one food is not the same as 100g of that nutrient from another food. For example, DELTA scales the availability of protein and the essential amino acids based on protein quality: the ability of the body to obtain and use these nutrients from different sources. The model also scales the iron and zinc requirements of the global population based on the foods available to meet these requirements. Consideration of bioavailability allows for a more realistic comparison between availability and requirement than is possible from considering content alone.
- Inclusion of upper and lower safe intake levels for nutrients: we all know about recommended daily intakes for nutrients. However, many nutrients have additional reference values, such as for safe lower and upper intake levels. The DELTA Model displays these in addition to a target intake, to give the user more information about the adequacy of nutrient supply. For example, in several scenarios detailed in the paper, there is sufficient availability of most required nutrients. However, these scenarios also feature energy availability above the safe upper intake level, implying that obtaining the target levels of other nutrients in these scenarios may necessitate excess energy intake.
The future of the DELTA Model
The current DELTA Model has a number of limitations for the SNi team to address in the future. For example, the current version considers 29 nutrients, but this does not include all that are essential to good nutrition. Inclusion of essential fatty acids will be a key next step.
Bioavailability is currently only included for some of the nutrients in the model. This will be developed in the future, but is limited by the availability of data for all nutrients from all foods.
Future versions of the model will also need to calculate the environmental impacts of food system scenarios. At present, it is the user’s responsibility to decide whether their scenario is possible from an environmental perspective. We are currently working on land use sub-models, which will allow the user to see the required land necessary for food production in their scenario. Other environmental impacts will follow.
Under justified scrutiny from an environmental sustainability perspective, the global food system needs to change. However, it is essential that the nutritional implications of any change are not forgotten. The DELTA Model is a tool that allows us to investigate future food system scenarios to see what is possible from a nutrition perspective, to be considered alongside the other aspects of sustainability.
All content relates to or was reproduced from Smith et al. (2021) Journal of Nutrition