Head Strong

Asprey's argument is that cognitive performance is fundamentally an energy problem. Neurons are among the most metabolically demanding cells in the body, mitochondrial output determines how much work they can do, and so fatigue, brain fog and poor focus should be treated as bioenergetic failures rather than character or willpower problems.

From that premise he builds a two-week programme with three moves: remove what he argues impairs mitochondrial function, supply what supports it, and add stressors that prompt adaptation. Removal covers refined sugar, certain oils, alcohol, poor-quality sleep and — a recurring personal thesis — mould and mycotoxin exposure in homes and food. Support covers a fat-forward, lower-carbohydrate diet, ketones, and a supplement list aimed at mitochondrial pathways. The adaptive stressors include brief intense exercise, cold, light exposure, and breathing practices, alongside devices such as red and near-infrared light and neurofeedback.

It is aimed at a reader who feels cognitively slower than they should and wants a structured two-week experiment, and it is readable for someone with no biology background.

On evidence, the foundation is genuine. Mitochondrial dysfunction is an active and serious area of research in neurodegeneration, fatigue and metabolic disease, and the basic supports Asprey lists — sleep, exercise, light, reduced alcohol, better food quality — are well supported for cognition. The leap is everything downstream of that. The claim that a specific 14-day protocol produces measurable, attributable cognitive gains is not established in controlled human trials, and the book offers self-report rather than testing. The mould and mycotoxin material is the clearest overreach: indoor mould exposure is a real clinical issue for some people, but the breadth of symptoms Asprey attributes to it, and the prevalence he implies, exceed what the literature supports. The framework is a reasonable way to think about cognitive fatigue; the specific protocol and its timeline should be treated as untested.