Simultaneous optimisation of the nutritional quality
and environmental sustainability of diets
The purpose of this thesis was to explore how the nutritional quality and environmental sustainability of diets can be optimised simultaneously. This work operationalised and quantified the concept of sustainable diets, as defined by the Food and Agricultural Organization (FAO), by using examples from the Netherlands. Given the need to redefine dietary guidelines to include health, food security and sustainability issues, quantitative ways of combining and optimising nutritional quality and environmental sustainability were developed. Such approaches increase the concreteness of dietary advice, thus enabling consumers to make dietary choices that satisfy both nutritional and sustainability requirements.
Indicators
Achieving simultaneous optimisation necessitates methods of measuring the nutritional quality and the environmental sustainability. Greenhouse gas emissions (GHGE) and land use (LU) appeared to be applicable and representative indicators of the environmental sustainability of diets. In evaluating nutritional quality, the current focus is usually on negative health effects, especially those caused by sodium, saturated fats and added sugars. This thesis expanded this approach by demonstrating how positive effects can also be incorporated into measurements. Accordingly, plant proteins, essential fatty acids and dietary fibre were selected as indicators of high nutritional quality.
Application
The indicators were applied to evaluate and compare the current average Dutch diet with existing predefined dietary regimens, namely, adherence to dietary guidelines and flexitarian, vegetarian, vegan and Mediterranean diets. The composition of these dietary approaches were confirmed as ensuring high nutritional quality whilst simultaneously resulting in high environmental sustainability.
Optimisation of diets through linear programming
Subsequently, the thesis demonstrated that linear programming is an effective tool for simultaneously optimising the nutritional quality and environmental sustainability of diets. Linear programming was applied to two sample diets, after which an explanation was provided as to how elements of the FAO definition of sustainable diets can be quantified. Such quantification not only covers the optimisation of natural and human resources but also the guarantee of nutritional adequacy, thus resulting in diets with only half of the environmental impacts exerted by current diets. On the basis of the FAO definition, additional constraints were added to the quantification, thereby enabling the design of low-impact diets that are culturally acceptable and affordable. The linear programming approach generated three main findings:
• An optimised, traditional Low Lands diet is as healthy as a traditional Mediterranean diet, but is more sustainable.
• A sustainable diet is not necessarily more expensive. A healthy, sustainable weekly shopping basket for a two-person household can be filled for less than € 40 per week.
• Linear programming enables the modelling of domains of nutritional quality, economic effects (price) and environmental sustainability across disciplines, thus paving the way for the creation of dietary programmes that are more sustainable, cheaper and more culturally acceptable than predefined dietary regimens.
Synergy at product level
The in-depth analysis of the predefined dietary regimens and optimised diet models revealed a synergy between nutritional quality and environmental impact at the product level. Environmental quality (especially impacts on climate change and land use) could be connected to nutritional quality at diet and product levels. This synergy is described by an index that was developed in this thesis on the basis of seven nutritional quality indicators. The index is called the Sustainable Nutrient-Rich Foods (SNRF) index. With the SNRF index as a reference, consumers and practitioners can categorise food into the following four groups, which are identifiable by colour:
1. Products rich in sodium, saturated fats and/or added sugars (red; SNRF < –1)
2. Moderately nutrient-dense animal products (white; SNFR = –1 to 0)
3. Energy-dense, moderately nutrient-dense plant products (brown; SNFR = 0 to 1)
4. Low energy-dense, high nutrient-dense plant products (green; SNFR > 1)
The SNRF index can also assist in the rating of food products. For consumers, such rating helps in adhering to diets that are both healthy and sustainable; for public health organisations, SNRF-assisted rating can help them formulate well-founded recommendations for healthy and sustainable diets. In general, healthy and sustainable diets (i.e. Low Lands, Mediterranean, New Nordic and pesco-vegetarian diets) contain few products from the red group, moderate quantities of products from the white group and a high proportion of products from the brown and green groups in equal proportions (average SNRF = 1, i.e. all essential macronutrients that cover energy requirements are included in the diets).
New functional unit describing nutrient density
This thesis also developed a new functional unit called the Nutrient Density Unit (NDU). A functional unit is a measure of the function of a studied system —the life cycle analysis of a product or diet— providing a reference to which environmental impacts can be related. From the perspective of a healthy diet, comparing the environmental impacts of products by using metabolic energy content (kilocalorie), or even more accurately, nutrient density (defined as NDU) as a functional unit is a more practicable approach than using weight (kilogram). For example, cereals and other carbohydrate-rich products from the brown group exert the lowest environmental impact per kilocalorie, whereas vegetables, fruits, pulses and mushrooms from the green group pose the lowest environmental impact per NDU. In the end, the function, and there with the choice of the functional unit, dictates the ranking of the environmental impact of products.
Opportunities for improving diet practices
The final step was to test the practical applicability of the preceding results through a consumer questionnaire. The aim was to identify changes in actual food group consumption that may improve both Health and Sustainability Scores of different Dutch population subgroups (described by gender, age, education, work, income, socio-economic status, and social milieus). Opportunities for improving both scores of different population subgroups lie primarily in the reduction of meat consumption and total metabolic energy consumption. With the quantitative analysis of the diets followed by the different subgroups as a reference, four strategies were formulated:
I) Replacing snacks with fruit, especially between meals
II) Replacing cheese with vegetables at lunch and dinner
III) Partly replacing meat with fish
IV) Decreasing the consumption of alcoholic drinks and increasing the consumption of water, which reduces energy intake.
These strategies can support different population subgroups in efforts to eat both healthy and sustainable diets. If deriving the highest environmental gain is the purpose of dietary adjustment, a fifth strategy is applicable:
V) Halving the portions of consumed meat.
Actual unsustainable diet practices of specific subgroups require dedicated transition strategies.
Consumer target groups
If institutions would promote more healthy and simultaneously more sustainable diets, effective strategies will be necessarily context bound and depend on current dietary patterns of specific target groups. A quintile (20%) of the Dutch population already consumes a healthy and sustainable diet. This percentage predominantly consists of older, unemployed or retired women, including groups of post-materialists and traditionals. The opposite quintile (having a relative unhealthy and unsustainable diet) comprise mainly (young) employed men, who have been determined in this thesis as the appropriate main target group for improvements to overall diet. Applying the optimised diet models and strategies to high-impact groups (high greenhouse gas emissions) can elevate nutritional quality at the population level whilst simultaneously reducing the environmental impacts of the Dutch diet.
Next steps
The insights derived in this research can help individual consumers make informed food choices and, at the same time, contribute to the development of baseline consumer education and future dietary guidelines that consider both nutritional quality and environmental sustainability. The first step is to recommend the consumption of more plant-based foods and less meat and the selection of products that pose the least environmental within each product group. Halving the environmental impacts of food consumption can be accomplished primarily by a combination of four actions: reducing metabolic energy intake, transitioning from an animal-based diet to a more plant-based one and replacing snacks and drinks produced with high GHGEs with healthier alternatives.