The global nutrition transition, which embraces major changes in how food is produced, distributed and consumed, is associated with rapid increases in the prevalence of obesity, but the implications for diabetes differ between populations. A simple conceptual model treats diabetes risk as the function of two interacting traits: ‘metabolic capacity,’ which promotes glucose homeostasis, and ‘metabolic load’, which challenges glucose homoeostasis. Population variability in diabetes prevalence is consistent with this conceptual model, indicating that the effect of obesity varies by ethnicity. Evolutionary life history theory can help explain why variability in metabolic capacity and metabolic load emerges. At the species level (hominin evolution), across human populations and within individual life courses, phenotypic variability emerges under selective pressure to maximise reproductive fitness rather than metabolic health. Those exposed to adverse environments may express or develop several metabolic traits that are individually beneficial for reproductive fitness, but which cumulatively increase diabetes risk. Public health interventions can help promote metabolic capacity, but there are limits to the benefits that can emerge within a single generation. This means that efforts to curb metabolic load (obesity, unhealthy lifestyles) must remain at the forefront of diabetes prevention. Such efforts should go beyond individuals and target the broader food system and socioeconomic factors, in order to maximise their efficacy.
