Low-carbon diets across diverse dietary patterns: Addressing population heterogeneity under constrained optimization

Area of Science:

• Nutritional Science
• Environmental Science
• Computational Science

Background:

• Mathematical optimization models diets for health and sustainability.
• Existing models overlook population-level dietary diversity.
• This study addresses this gap by incorporating diverse dietary patterns.

Purpose of the Study:

• To develop a method for optimizing low-carbon diets considering population dietary diversity.
• To investigate trade-offs between greenhouse gas emission (GHGE) reduction and consumer inconvenience.
• To analyze differences in optimized diets and required changes across population subgroups.

Main Methods:

• K-means clustering applied to dietary intake data from Denmark to identify distinct dietary patterns (clusters).
• Quadratic programming used to minimize a consumer inconvenience index (total dietary changes) while meeting nutritional and GHGE constraints.
• GHGE constraints were incrementally tightened to assess impact on inconvenience and diet composition.

Main Results:

• A significant increase in consumer inconvenience was observed below approximately 3 kg CO2e/10 MJ GHGE.
• Optimized diets generally increased cereals, eggs, and fish, while decreasing beef, lamb, cheese, and alcohol.
• Dietary changes required varied across clusters, indicating differential impacts on population subgroups.

Conclusions:

• Population-specific optimization is crucial for effective dietary transitions towards sustainability.
• The proposed method can guide targeted interventions for specific sub-populations.
• The methodology allows for future integration of additional sustainability metrics and consumer preferences.