Personalised Nutrition — What the Science Can and Can't Yet Deliver
The idea that the optimal diet varies between individuals is almost certainly correct. The evidence that we can currently identify what that optimal diet looks like for a specific person, using the tools commercially available, is considerably more mixed. Personalised nutrition sits at an intersection of genuine scientific advance and significant commercial hype, and telling those two things apart is the most useful thing this page can do.
The science behind individual variation in nutritional response is real and fascinating. The landmark Weizmann Institute study — 800 people, continuous glucose monitors, identical meals — found that two people could eat the same food and produce blood glucose responses so different that what constituted a high-glycaemic food for one was a low-glycaemic food for the other. The responses were predicted not by standard nutritional tables but by the individual's gut microbiome composition, genetics, sleep, and physical activity. That finding, published in Cell in 2015, shifted how nutritional researchers think about population-level dietary guidelines. It did not, however, immediately translate into clinical tools that can reliably predict what any given individual should eat — a distinction the wellness industry has been slow to observe.
What individual variation actually means in practice
The individual variation finding is important but easily misread. It does not mean that the broad foundations of a healthy diet — adequate protein, dietary fibre, whole plant foods, limited ultra-processed food, omega-3 provision — vary meaningfully between people. Those foundations are robust across populations and the evidence for them is drawn from studies large enough that individual variation is accounted for statistically. What varies is the fine-tuning: which specific carbohydrate sources work best for a given person's blood glucose management, which dietary pattern achieves satiety most effectively, which micronutrient gaps are most relevant for a specific individual's profile.
The practical consequence is that personalisation is most valuable at the margins — once the foundations are in place — rather than as a substitute for them. Someone spending money on nutrigenomics testing while still eating a predominantly ultra-processed diet is optimising the wrong layer. The foundations matter far more than the fine-tuning, and for most people, getting the foundations right is sufficient to produce meaningful health outcomes without any sophisticated testing.
Continuous glucose monitoring — the tool with the most immediate practical value
Continuous glucose monitors (CGMs) — small sensors worn on the upper arm that measure interstitial glucose every few minutes — represent the personalised nutrition tool with the strongest evidence base and the most immediately actionable output. Historically restricted to people with diabetes, CGMs are now available to the general public through services like Levels, Zoe, and others, typically at a cost of £150–400 for a two-week monitoring period.
What CGM data reveals is genuinely informative: how your blood glucose responds to specific foods, meal combinations, sleep quality, stress, and exercise. The Weizmann Institute findings suggest that postprandial glucose response — how high blood glucose rises after a meal and how quickly it returns to baseline — is one of the more important modifiable predictors of metabolic health, and that this response is sufficiently individual to make general glycaemic index tables a poor guide for personal dietary decisions. Two weeks of CGM data can identify your specific high-response foods (which are not always the obvious ones), the meals and meal combinations that produce the flattest glucose profiles, and the lifestyle factors — exercise timing, sleep quality, stress — that most affect your response.
The honest caveat is that CGM data requires interpretation. A high postprandial glucose spike is not inherently harmful in the context of an otherwise healthy metabolic profile, and the clinical significance of the glucose variability metrics marketed by consumer CGM services is not yet fully established for non-diabetic individuals. The data is informative rather than diagnostic, and it is most useful when used to develop intuitions about your own metabolic responses rather than as a source of anxiety about every meal.
Nutrigenomics — real science, premature commercialisation
Nutrigenomics — the study of how genetic variants affect nutritional requirements and responses — is a legitimate and growing field of research. Specific genetic variants do affect metabolically relevant processes: MTHFR variants affect folate metabolism and methylation capacity; APOE4 significantly elevates Alzheimer's risk and modifies the response to dietary fat; FADS1/FADS2 variants affect the efficiency of omega-3 and omega-6 fatty acid conversion; FTO variants influence appetite regulation and obesity risk. These are real genetic influences on nutritional metabolism with meaningful implications.
The gap between that science and the commercially available nutrigenomics tests marketed for dietary personalisation is substantial. Most consumer genetic tests for nutrition report on dozens or hundreds of variants, the majority of which have small effect sizes that are not yet replicated consistently across populations, and generate dietary recommendations that in practice diverge very little from general healthy eating advice. A 2019 review in the British Medical Journal found no consistent evidence that nutrigenomics-based dietary advice produced better health outcomes than standard dietary guidance. The science is advancing rapidly — this picture may look different in five years — but the current commercial offerings should be approached with realistic expectations about what they can deliver.
APOE4 is the exception worth knowing specifically. Carried by roughly 25% of the population, the APOE4 allele significantly elevates Alzheimer's risk and appears to modify the response to dietary saturated fat — APOE4 carriers show larger LDL increases in response to saturated fat than non-carriers. If you have a family history of Alzheimer's disease and haven't been tested for APOE4 status, it is a conversation worth having with your GP, primarily because it changes the risk calculus around cardiovascular risk management and may influence the urgency of lifestyle interventions.
Gut microbiome testing — where the science and the products diverge most
The gut microbiome is genuinely central to nutrition and health — that case has been made compellingly in the research over the past decade. The microbiome affects nutrient absorption, immune function, inflammatory tone, metabolic responses to food, and even mood and cognition through the gut-brain axis. Individual microbiome composition varies enormously between people and is shaped by diet, medication history, birth mode, early life exposures, and genetics. The Weizmann Institute's finding that microbiome composition predicted individual glycaemic responses better than standard nutritional tables was a landmark result.
What does not yet follow from that science is the clinical utility of commercial microbiome tests for dietary guidance. The field lacks consensus on what a healthy microbiome looks like — there is no established reference range in the way there is for blood glucose or cholesterol. The relationship between any specific microbiome composition and health outcomes is correlational rather than causal in most human data. And the dietary interventions recommended by commercial microbiome tests — typically involving specific probiotic supplements and prebiotic foods — have not been validated in clinical trials against the tests' own outputs. The tests are scientifically interesting. The dietary recommendations they generate should be treated with appropriate scepticism.
What is well-evidenced is that a diverse, fibre-rich diet — the kind described throughout this nutrition section — consistently promotes microbiome diversity, and that microbiome diversity is associated with better metabolic and immune outcomes. This conclusion does not require a test. It is supported by the general dietary evidence and does not change based on an individual's specific microbiome profile.
Blood testing — the most clinically grounded form of nutritional personalisation
The most evidence-based and immediately actionable form of nutritional personalisation is not genetic testing or microbiome analysis — it is standard and extended blood testing, which can identify specific deficiencies and risk factors that should directly modify dietary and supplementation choices. Vitamin D deficiency is extremely common in the UK, particularly in adults who spend limited time outdoors, and supplementation in deficient individuals produces well-established health benefits. Ferritin (iron stores) and B12 are commonly low in people eating plant-heavy diets and in older adults with reduced gastric acid production. Thyroid function affects metabolic rate and therefore macronutrient requirements in ways that are meaningless to personalise around without knowing thyroid status.
An NHS Health Check provides total cholesterol, blood glucose, and blood pressure — a useful starting point. A more comprehensive private blood panel — typically £150–250 from services like Medichecks or Thriva — adds vitamin D, ferritin, B12, folate, full thyroid panel, full lipid panel including Lp(a), HbA1c, and inflammatory markers. This is the most cost-effective form of nutritional personalisation available, because it identifies specific deficiencies that individual dietary adjustment can address, rather than generating recommendations based on population-level genetic associations with small effect sizes.
'The science of individual variation in nutritional response is real and fascinating. The commercial products claiming to translate that science into personalised dietary plans are considerably ahead of what the evidence currently supports. Knowing the difference is worth more than any test.'
Building your own personalised picture — the practical hierarchy
The most useful framework is to think about personalisation in layers, ordered by evidence strength. The foundation — adequate protein, dietary fibre, whole plant foods, omega-3 provision, limited ultra-processed food — is universal and does not require personalisation. The second layer is blood-test-informed supplementation: knowing your vitamin D, B12, ferritin, and thyroid status, and adjusting diet and supplements accordingly. The third layer, for those interested in going further, is CGM-informed understanding of personal glucose responses — genuinely informative and practically useful for metabolic health management. Genetic and microbiome testing sit at the fourth layer — scientifically interesting, commercially oversold, most useful for specific high-impact variants like APOE4 rather than as comprehensive dietary guides.
The reader who works through this hierarchy systematically — foundations first, blood testing second, CGM if motivated, genetic context where specifically relevant — will have a more genuinely personalised nutritional picture than one who invests in a commercial DNA diet kit and leaves the foundations unaddressed. Personalisation is real. The tools to achieve it are unevenly mature. Starting with the ones that are clinically established is the rational approach.
Start Slowing the Clock
Expert tips and insights on living younger for longer — straight to your inbox, every week.
No spam, ever. Unsubscribe any time.
Start Slowing the Clock
Expert tips and insights on living younger for longer — straight to your inbox, every week.
No spam, ever. Unsubscribe any time.