The science

If section 1 made the case that nutrition sits at the centre of long-term health, this section is for readers who want to see the workings. We've structured it around six topics — the levers that most affect how well you age, ranked roughly by the strength of recent evidence. For each, we walk through the mechanism (what food is actually doing in the body), the evidence (what the research shows), and where the picture is still being worked out.

Four underlying mechanisms recur throughout: blood sugar regulation, the gut microbiome, muscle maintenance, and chronic low-grade inflammation. Almost everything nutrition does to long-term health travels through one or more of these pathways. Keep them in mind as you read — they're the connective tissue between the topics.

1. Ultra-processed foods

The mechanism

Ultra-processed foods (UPFs) are industrial formulations of cheap ingredients — refined oils, isolated proteins, added sugars, starches, emulsifiers, flavourings — engineered to be shelf-stable, cheap, and highly palatable. The NOVA classification, developed by researchers at the University of São Paulo, distinguishes them from foods that are merely processed. Bread, cheese, tinned tomatoes, yoghurt — all processed, all fine. UPFs are what you get when industrial reformulation goes further: ingredients you wouldn't find in a home kitchen, structures engineered for a specific mouthfeel, combinations of fat, sugar and salt calibrated to override satiety.

The distinction matters because the harm from UPFs doesn't seem to operate mainly through nutrients. If you matched calories, fat, sugar, sodium and fibre between a UPF diet and an unprocessed one, older nutrition science would predict similar outcomes. That prediction appears to be wrong. Something about how UPFs are designed — their energy density, the speed at which people eat them, the disruption of the feedback signals that tell the brain you've had enough — drives overconsumption directly.

The evidence

The clearest causal evidence comes from a controlled-feeding trial at the US National Institutes of Health. Twenty adults were admitted to a metabolic research ward and put on matched ultra-processed and unprocessed diets for two weeks each, in a randomised crossover design. The diets were matched for calories, sugar, fat, sodium and fibre. Participants were told to eat as much as they wanted. On the UPF arm, they consumed around 500 extra calories a day — and gained weight, measurably, over just two weeks. [1] The effect size is large enough that, extrapolated across a lifetime of modern eating, it could plausibly account for much of the rise in obesity and metabolic disease.

The epidemiology matches. A 2024 umbrella review published in the BMJ — the most comprehensive synthesis of UPF research to date — covered 45 separate meta-analyses and data on 9.9 million people. Higher UPF consumption was associated with 32 different adverse health outcomes: a roughly 50 per cent higher risk of cardiovascular-related death, a 48 to 53 per cent higher risk of anxiety and common mental disorders, a 12 per cent higher risk of type 2 diabetes, and a 21 per cent higher risk of all-cause mortality. [2] In November 2025, the Lancet published a three-part series written by 43 international experts — including the originators of the NOVA classification — that reviewed over 100 studies and concluded the evidence is now strong enough to justify immediate public health action. [3]

The scale of the shift in UK diets is significant. Over half the calories UK adults now consume come from UPFs; for teenagers it's closer to two-thirds. That is the largest single change in how people eat over the last fifty years.

What's contested

The main critique of the evidence base is that the NOVA category is heterogeneous. It groups sugar-sweetened drinks with whole-grain wraps, frozen ready meals with plant-based milks. Critics argue this makes the research harder to interpret, and the advice harder to act on. The mainstream response, accepted in the Lancet series, is that the category still tracks real health outcomes consistently across populations — and that the pattern matters more than the edges. The boundary cases are likely to get sharper over the next few years; the overall signal is clear enough now to act on. Reducing UPFs in a diet moves in the right direction, regardless of where the exact lines end up being drawn.

2. Protein and muscle

The mechanism

Muscle is the largest metabolic organ in the body. It's not just what moves you — it's a major site of glucose disposal, a reservoir of amino acids, and one of the strongest predictors of healthspan in adults over 60. From around the mid-thirties onwards, most people start to lose muscle mass at roughly one per cent per year. The rate accelerates after sixty. By eighty, if nothing intervenes, many people have lost thirty per cent of the muscle they had at thirty. This is sarcopenia, and it underpins much of what we associate with ageing: weakness, falls, loss of independence, metabolic deterioration.

Protein is the main dietary lever. Each meal containing adequate protein triggers a pulse of muscle protein synthesis — a signal that tells the body to build or maintain muscle tissue. Below a threshold, the signal doesn't fire strongly enough to offset the continuous background rate of muscle breakdown. Over years, those missed signals accumulate.

The evidence

The PROT-AGE international expert group, convened specifically to reassess protein needs in older adults, recommends healthy people over 65 consume 1.0 to 1.2 grams of protein per kilogram of bodyweight per day. [4] That is noticeably above the long-standing Recommended Dietary Allowance of 0.8g/kg, which was set in an era when the research focused on preventing outright deficiency rather than maintaining healthspan. For someone weighing 70kg, the difference is about 15 grams a day — roughly the protein in a small chicken breast.

Distribution across meals matters as much as total intake. Research suggests that muscle protein synthesis responds in a threshold fashion: roughly 25 to 30 grams of high-quality protein per meal appears to be the level at which the signal fires fully. Below that, the meal contributes less to muscle maintenance than the numbers might suggest. Most UK adults load their protein at dinner — eggs or toast for breakfast, a sandwich for lunch, a substantial dinner — which is the opposite of the optimal distribution. Breakfast is where the shortfall is most visible.

Strength training amplifies the effect significantly. Resistance exercise without adequate protein achieves a fraction of what it could; protein without resistance exercise maintains less muscle than the combination. The two work together, which is why this topic connects to the Movement pillar as much as to nutrition.

What's contested

There is some debate about whether the 1.0–1.2g/kg recommendation extends to very active older adults — where some research suggests higher intakes (up to 1.6g/kg) may be beneficial — and about the extent to which plant-based proteins can fully substitute for animal sources without careful combining. These are real questions, but they operate at the margins. The headline finding — that most adults over 40 are under-eating protein relative to their actual needs — is not seriously contested.

3. Fibre and the microbiome

The mechanism

The community of bacteria living in the large intestine — collectively, the gut microbiome — has turned out to be far more consequential than researchers expected twenty years ago. It affects immunity, inflammation, mood, cognition, and even the pace of biological ageing, largely through the short-chain fatty acids its residents produce when they ferment dietary fibre, and through what's called the gut-brain axis: a bidirectional communication system between the gut and the central nervous system.

The microbiome feeds almost exclusively on what you eat, and in particular on fibre from plants and the live cultures in fermented foods. A fibre-poor, plant-poor diet starves the beneficial species; over time, diversity declines. Low microbiome diversity correlates with almost every chronic condition that has been studied in this area — inflammatory bowel disease, depression, metabolic syndrome, autoimmune conditions, and accelerated cognitive ageing.

The evidence

The UK data is stark. Average fibre intake is around 19 grams per day, against a target of 30g set by the Scientific Advisory Committee on Nutrition. Only around nine per cent of UK adults meet the recommendation. [5] This is not a marginal shortfall — it's a gap large enough to matter at the scale of national disease burden, and it's been flagged as a contributor to the stalling of UK life expectancy gains.

The most striking recent intervention evidence came from a randomised trial at Stanford in 2021. Researchers split 36 healthy adults into two arms over ten weeks: one group progressively increased fibre intake, the other progressively increased fermented foods (kefir, live yoghurt, kimchi, kombucha, unpasteurised sauerkraut). The fermented-food group showed measurably increased microbiome diversity and decreases in 19 separate inflammatory proteins. The fibre group did not show the same diversity effect in the same timeframe. [6] This was surprising — the consensus view had been that fibre was the main diversity-driver. The likely explanation is that people who already had low-diversity microbiomes lacked the bacteria needed to ferment the increased fibre; the fermented foods effectively seeded new species directly.

Two practical implications follow. First, variety matters as much as quantity — thirty different plants per week appears to be a useful target, drawing from fruits, vegetables, legumes, nuts, seeds, whole grains, and herbs and spices. Second, fermented foods appear to do something specific that fibre alone does not. Both are worth including, not either-or.

What's contested

The microbiome field moves fast, and individual responses vary more than most nutrition science — what works for one person's gut may do less for another's. Personalised microbiome testing and recommendations are a growing commercial space; the science underneath them is real, but the precision of the current tests is debated. The general principles — more fibre, more plant diversity, regular fermented foods — are robust enough to act on without testing. The personalised recommendations are still catching up to their marketing.

4. Saturated fat and cardiometabolic health

The mechanism

Saturated fat affects cardiometabolic health through several overlapping pathways. It raises LDL cholesterol, which contributes to the build-up of atherosclerotic plaque in arteries. It appears to drive low-grade inflammation — particularly when consumed in excess and in the context of refined carbohydrates. [7] And higher sustained intakes are associated with reduced insulin sensitivity over time, contributing to the development of metabolic syndrome and type 2 diabetes. The three pathways are connected: inflammation and insulin resistance amplify each other, and both accelerate cardiovascular disease.

What matters isn't just saturated fat in isolation, but what else is in the diet and what saturated fat replaces (or is replaced by). Saturated fat consumed as part of an ultra-processed food — say, palm oil in a packaged biscuit — appears to behave differently from saturated fat consumed as part of a whole-food pattern, like cheese in a traditional Mediterranean diet. The food matrix matters.

The evidence

The strongest independent evidence for reducing saturated fat comes from a 2020 Cochrane review of randomised controlled trials. Pooling data from 12 trials covering 53,300 participants who sustained reduced saturated fat intake for at least two years, the review found a 17 per cent reduction in cardiovascular events. [8] The effect was strongest when saturated fat was replaced with polyunsaturated fats — particularly from plant sources like seeds, nuts, and oily fish.

The American Heart Association's 2017 presidential advisory synthesised the RCT evidence and concluded that replacing saturated fat with polyunsaturated fat reduces cardiovascular events by approximately 30 per cent — comparable in magnitude to the effect of statin treatment. [9] What doesn't work, according to the same body of evidence, is replacing saturated fat with refined carbohydrates. That kind of substitution produces little benefit, and sometimes worse outcomes, because refined carbohydrates drive their own problems with blood sugar and triglycerides.

The observational picture is more complicated. A 2015 systematic review published in the BMJ found that saturated fat intake, assessed through dietary recall studies rather than controlled trials, was not clearly associated with all-cause mortality, coronary heart disease mortality, or type 2 diabetes. [10] Trans fats, by contrast, showed clear associations with all three. This observational-vs-trial discrepancy is one of the reasons the saturated fat debate has been so persistent: the RCT evidence supports reduction; the observational evidence is messier.

What's contested

A serious scientific debate exists about whether the traditional "limit saturated fat" guidance has been too simplistic. A 2020 reassessment published in the Journal of the American College of Cardiology, authored by twelve senior researchers including some who had previously supported the standard guidance, argued that food-based recommendations (avoid processed meat, limit refined carbohydrates, eat more fish and plants) serve people better than nutrient-based ones. Their central point was that saturated fat from different whole-food sources — the cheese in a Mediterranean meal, the fat in a piece of oily fish, the cream in unsweetened yoghurt — doesn't seem to behave uniformly, and that blanket nutrient restrictions may miss more than they capture. [11]

A 2023 overview of systematic reviews — covering the full evidence base — concluded that the certainty of evidence for saturated fat causing mortality and major cardiovascular events is low to moderate, not high. [12] This is not the same as saying saturated fat is harmless. It is saying that the confidence bands around the headline numbers are wider than some guidelines imply.

The weight of independent evidence still supports reducing saturated fat, particularly by replacing it with unsaturated fats from whole-food sources — olive oil, nuts, seeds, oily fish. The "all saturated fat is bad" framing of earlier decades has not aged well. What the evidence actually supports is a plant-forward pattern with unsaturated fats as the default dietary fat, and saturated fat from whole foods — not from ultra-processed foods — fitting within the wider diet in moderation.

The signal on saturated fat is real, but the direction of travel — plant-forward, replace rather than just cut, whole foods over engineered ones — matters more than any single number.

5. Omega-3 fatty acids

The mechanism

Omega-3 fatty acids — particularly EPA and DHA, found mostly in oily fish — are structural components of every cell membrane in the body and precursors to signalling molecules involved in resolving inflammation. Unlike most fats, they appear to shift the body's inflammatory balance toward the anti-inflammatory side. They also lower triglycerides, modestly affect heart rate variability, and appear to protect against the electrical instability in heart tissue that underlies some arrhythmias.

A separate form of omega-3, alpha-linolenic acid (ALA), is found in plants — walnuts, flaxseed, chia, rapeseed oil. The body can convert ALA into EPA and DHA, but the conversion rate is inefficient (typically under 10 per cent). For people who eat oily fish regularly, this doesn't matter. For strict vegetarians and vegans, it means algae-derived supplements are the most reliable way to meet EPA and DHA needs.

The evidence

A 2024 meta-analysis published in the European Journal of Preventive Cardiology pooled 18 randomised trials covering 134,144 participants across primary and secondary cardiovascular prevention. Omega-3 supplementation reduced myocardial infarction, cardiovascular death, and coronary revascularisation — with benefits maintained in patients already on statin therapy. [13] The effect sizes are modest rather than dramatic (cardiovascular death reduced by around 8 per cent; myocardial infarction by around 11 per cent), but they are consistent and clinically meaningful.

The UK recommendation is two portions of oily fish per week, of which at least one should be oily (salmon, mackerel, sardines, herring, trout). Most UK adults eat less than one portion. For those who don't eat fish, a modest algae-based EPA/DHA supplement is the cleanest equivalent.

What's contested

Higher-dose supplementation — particularly purified EPA at pharmacological doses — shows stronger effects in some trials but has also been associated with small increases in the risk of atrial fibrillation. [13] For the general dietary recommendation (two portions of oily fish a week), this isn't a concern. For pharmacological supplementation, it's an ongoing question that your GP is better placed to weigh than we are.

6. Blood sugar and meal composition

The mechanism

When you eat — particularly refined carbohydrates and sugars — glucose rises in the bloodstream, and the pancreas releases insulin to move that glucose into cells. Do this repeatedly, across decades, and the system begins to lose sensitivity. Cells that used to take up glucose efficiently require more insulin to do the same job. This is insulin resistance, and it sits upstream of type 2 diabetes, metabolic syndrome, non-alcoholic fatty liver disease, and much of what's called "metabolic ageing."

The key insight from recent research is that it's not just average blood sugar that matters — it's variability. Large, fast glucose spikes appear to be more problematic than the same total glucose delivered more gradually. Two people with the same HbA1c can have very different metabolic trajectories depending on the shape of their glucose response throughout the day.

The evidence

What blunts the glucose response is partly what you eat, and partly the order in which you eat it. Protein, fibre and fat before or alongside carbohydrates slow gastric emptying and reduce the post-meal glucose spike — sometimes substantially. Controlled studies have shown that eating the protein and vegetables on a plate before the carbohydrates can reduce the post-meal glucose spike by 30 per cent or more, with the same foods and the same total calories. Whole foods produce smaller spikes than refined ones (steel-cut oats vs. instant; brown rice vs. white; whole fruit vs. juice). Adding acid — vinegar in dressings, lemon juice — also modestly reduces the glucose response to carbohydrates.

Continuous glucose monitors (CGMs) have made personal glucose variability visible in a way it wasn't before. For people with diabetes, this is transformative. For people without, the evidence for routine CGM use is still developing — the devices show variability that may or may not be clinically meaningful, and the interpretation requires some care. The meal-composition principle is robust and worth acting on with or without a CGM; the technology is interesting but, for otherwise healthy adults, not yet a first-line recommendation.

What's contested

There is genuine disagreement about how much everyday glucose variability matters for people without diabetes or prediabetes. The mechanistic case is strong; the hard-outcome evidence in otherwise healthy adults is still being built. What's not contested: the dietary pattern that produces favourable glucose responses — whole foods, adequate protein and fibre, moderate portions of refined carbohydrates, meals rather than grazing — is the same pattern supported by almost every other strand of nutrition evidence. Whether or not glucose variability itself is the mechanism, the practical recommendations overlap with everything else in this section.

7. Micronutrients, plant bioactives, and inflammation

The mechanism

Modern diets are calorie-abundant but often nutrient-poor. A diet can meet energy needs, fit within macronutrient guidelines, and still leave the body short on vitamins and minerals it needs for long-term health. This is sometimes called "hidden hunger" — not the acute deficiency that causes scurvy or rickets, but the chronic sub-optimal intakes that degrade function gradually over decades. Low-level inadequacies in vitamins D and B12, folate, iron, magnesium, iodine, selenium and others — each with specific roles in energy metabolism, immune function, bone health, cognition and thyroid function — accumulate into the kind of health debt that makes ageing harder than it needs to be.

Alongside these classical micronutrients, plants deliver a second category of compounds that nutrition science has only recently started to take seriously: polyphenols. These include flavonoids (found in berries, tea, cocoa, onions, citrus), phenolic acids (in coffee, whole grains), stilbenes (grapes, peanuts) and lignans (seeds, whole grains). They aren't technically essential nutrients, but they do meaningful work in the body. Part of that work is as antioxidants — neutralising reactive oxygen species (free radicals), which are unstable molecules produced during normal metabolism that can damage cells, DNA and tissues over time. Part of it is anti-inflammatory: polyphenols modulate the signalling pathways that drive chronic low-grade inflammation, the shared pathway underneath cardiovascular disease, type 2 diabetes, some cancers and neurodegeneration. And part of it is indirect — many polyphenols are metabolised by gut bacteria into compounds that circulate in the bloodstream and affect tissues far from the gut.

The anti-inflammatory dimension deserves particular attention because inflammation connects to almost every other topic in this section. Diets high in ultra-processed foods, refined sugars and industrial trans fats push inflammatory markers up; diets rich in oily fish, colourful plants, extra virgin olive oil, nuts, legumes, herbs and spices push them down. Specific bioactive compounds — omega-3 fatty acids, polyphenols, fibre-derived short-chain fatty acids from the microbiome — each contribute through different mechanisms to the same overall effect.

The evidence

On the micronutrient side, the UK picture is stark. The most recent National Diet and Nutrition Survey shows that around 25 per cent of UK women aged 19-64 have iron intakes below the Lower Reference Nutrient Intake — the floor level below which deficiency becomes likely. 16 per cent of adults have vitamin D blood levels indicating deficiency risk, rising further in winter. Selenium intakes average around half of the recommended level across the UK population — a national-scale shortfall driven partly by UK soil chemistry. Magnesium, iodine, folate and B12 all show significant subpopulations below the floor, particularly in women and in younger adults. [5] The pattern is consistent: people who eat mostly ultra-processed foods, skip plants, or follow restrictive diets without planning tend to be short on several nutrients at once.

On polyphenols, the evidence has strengthened significantly over the last decade. A 2019 prospective study published in Nature Communications followed 56,048 Danish adults for 23 years and found that moderate flavonoid intake was inversely associated with all-cause, cardiovascular, and cancer mortality — with the protective effect plateauing around 500mg per day. [14] A 2020 meta-analysis pooling 16 cohort studies and over 462,000 participants confirmed that people in the highest category of flavonoid intake had roughly 13 per cent lower all-cause mortality and 15 per cent lower cardiovascular mortality than those in the lowest category. [15] The effect is particularly pronounced in smokers and heavy drinkers, suggesting polyphenols may help buffer the oxidative stress those behaviours drive.

On anti-inflammatory action, the most consistent evidence points to dietary patterns rather than individual foods. A 2025 meta-analysis of randomised controlled trials published in Nutrition Reviews found that the Mediterranean dietary pattern significantly reduced high-sensitivity C-reactive protein (a key inflammatory marker), interleukin-6, and interleukin-17 — three of the most-studied markers of systemic inflammation. [16] The effect is driven by the combination of components: monounsaturated fats from olive oil, omega-3s from oily fish, polyphenols and antioxidants from colourful plants, fibre feeding beneficial gut bacteria, and the relative absence of ultra-processed foods and refined sugars. Individual "anti-inflammatory foods" make for good headlines, but the evidence is strongest for the overall pattern.

Practically, the nutrients, the polyphenols and the anti-inflammatory compounds tend to come from the same foods. Colourful fruits and vegetables, oily fish, nuts and seeds, whole grains, legumes, extra virgin olive oil, herbs and spices, and tea and coffee between them supply most of what the research shows to matter. Variety is doing real work — different plants contain different polyphenol profiles, and eating across the spectrum of colour and type delivers a broader set of compounds than eating the same few items repeatedly. The old advice to "eat the rainbow" has a sharper evidence base now than when it was first coined.

Where supplements fit

Food first, but supplements have a legitimate role in filling identified gaps. Getting the full spectrum of micronutrients from food alone is genuinely difficult — particularly in the UK, where soil selenium levels are low, vitamin D is hard to obtain from diet and sunlight for much of the year, and busy modern eating patterns frequently fall short of recommended intakes for magnesium, iron, iodine and B12. For these nutrients, supplementation is not a compromise on a whole-foods approach; it is a sensible way to meet needs that diet alone may not reliably cover.

The evidence supports several specific uses. Vitamin D through the autumn and winter months is recommended by Public Health England for the entire UK population. Omega-3 supplementation (fish oil or algae-derived) is supported by cardiovascular trial evidence for those who don't regularly eat oily fish. Magnesium supplementation has good evidence for sleep, muscle function and glucose regulation, particularly given how difficult it is to reach the RNI through modern food sources alone. B12 is essential for anyone following a vegan diet and worth considering for older adults whose absorption declines with age. A quality multivitamin can help close gaps for people with restricted diets, high training loads, or periods of stress. Probiotics have an emerging evidence base for specific gut and immune contexts.

Many people seriously interested in longevity take a thoughtful combination of these — not because they think supplements replace good food, but because closing nutrient gaps through diet alone is often impractical, and getting the balance right is genuinely hard. The honest framing is that whole foods should provide the foundation, and supplements can be a targeted, evidence-based way to fill gaps in that foundation. What the evidence does not support is taking supplements without purpose, or using high-dose isolated extracts (like standalone polyphenol capsules) to try to replicate the effects of whole plants — the compounds in real food work together in ways that isolated pills haven't reproduced in trials.

Seven topics, hundreds of pages of underlying research, and a handful of recurring patterns. Almost all of it travels through blood sugar regulation, the gut microbiome, muscle maintenance, or chronic inflammation — the four mechanisms flagged at the start. Almost all of it points toward the same dietary pattern: whole foods over processed ones, plant-forward without being dogmatic about it, adequate protein distributed across the day, plenty of fibre and plant diversity, oily fish a couple of times a week, unsaturated fats as the default, saturated fat from whole-food sources in moderation, and a sensible relationship with refined carbohydrates and added sugars. The plants carry most of the polyphenols; the variety carries the micronutrients; the whole-foods framing handles almost everything else.

This is not a diet. It's a pattern — one that looks a lot like the Mediterranean way of eating, but that doesn't need to be labelled to work. Members who follow most of this most of the time are pulling on almost every lever that nutrition research has shown to matter for long-term health. Section 4 makes this actionable: where to start, depending on where you are.