Hidden Glucose in Everyday Carbs Is Driving Fatty Liver and Diabetes

David Unwin argues that type 2 diabetes and poor metabolic health usually develop long before a person is told they are ill. The visible diagnosis is late. The earlier process is quieter: excess carbohydrate becomes glucose, insulin moves that glucose out of the bloodstream, fat accumulates around the abdomen and in the liver, insulin becomes less effective, and the pancreas is forced to produce more insulin until it eventually cannot keep up.

Unwin described this as the “long silent scream from the liver,” a phrase he connected to Professor Roy Taylor’s work at Newcastle University. The liver, he said, can become fatty for about 10 years without the person knowing. During that period, the person may only see a growing waist and assume it is “middle-aged spread.” Unwin made that mistake himself when he was a GP with a heavy biscuit habit, sitting for long hours in practice and taking in more carbohydrate than he needed to use.

His simple explanation of insulin was that it is protective. High blood glucose damages arteries, so insulin “pushes” glucose out of the bloodstream and into cells. Some glucose can be used for energy. But if a person takes in more carbohydrate than they need to “run around,” Unwin said, the body converts that sugar into fat because that is safer than leaving it in the blood. Over time, this produces abdominal fat and liver fat.

The problem then becomes self-reinforcing. Fatty liver interferes with insulin’s action, producing insulin resistance. If insulin is less powerful, the pancreas has to make more of it. Eventually, fat can be laid down in the pancreas itself, the gland that produces insulin. At that point, Unwin said, the ability to regulate blood sugar can collapse.

He stressed that this is not only an older-person problem. When he began general practice in 1986, in a community north of Liverpool, obesity was rare and his practice had “not a single case” of type 2 diabetes in anyone under 55. The condition was then commonly called maturity-onset diabetes. He said the name had to change because so many younger people now develop it. The day before the interview, he had seen two people under 25 with poorly controlled diabetes, one of whom was too heavy for him to weigh.

That change is why he described the present situation as a “pandemic of poor metabolic health.” Young people, in his view, are paying the greatest price because they have the longest futures to lose and the most years over which daily choices can compound.

The risk is not only that type 2 diabetes shortens life. Unwin said UK government figures show that each year of poorly controlled type 2 diabetes costs about 100 days of life. He also said perhaps a third of people globally with type 2 diabetes do not know they have it because they have not been tested.

Cardiovascular disease was the obvious answer when Steven Bartlett asked what kills people with type 2 diabetes. Unwin agreed that high blood glucose damages arteries over time, and he added that very high blood sugar can damage the glycocalyx, the “non-stick lining” of arteries, within six hours. But he also said a rising cause of mortality among people with diabetes is cancer, and that eight forms of cancer are strongly associated with diabetes.

The most actionable part of his argument was timing. In his own practice dataset, which he said he has tracked since 2013, people with prediabetes who go low-carbohydrate have, in his figures, a 93% chance of returning to completely normal blood sugar. If someone already has type 2 diabetes but is caught early, he said his chance of helping them achieve normal blood sugar is 73%. If they wait five years, that falls to about 50%.

His objection is that medicine often waits until the late stage. By then the visible disease has arrived, but the metabolic process has been running for years.

A major part of Unwin’s work is built around a communication problem: many people understand “sugar” to mean table sugar, sweets, chocolate, fizzy drinks, or something that tastes sweet. They do not necessarily understand that bread, rice, potatoes, pasta, breakfast cereal, and other starches break down into glucose.

His shorthand was that starchy carbohydrates are “glucose molecules holding hands.” Digestion breaks those bonds, and the starch becomes glucose in the bloodstream. He called this “schoolboy biology,” and one of the most important moments in his career came when a patient forced him to confront that he had not been making it clear.

The patient, whom he called “Mrs. Jones” rather than using her real name, had poorly controlled type 2 diabetes and had stopped taking Metformin. At the time, Unwin said, his practice monitored whether patients were taking Metformin partly because prescribing it was considered good practice and was linked to how the practice was paid. He wrote to her asking her to come in because she appeared not to be taking it.

She arrived angry. According to Unwin, she told him that when he checked her blood tests he would find her glucose was normal despite not taking Metformin. She asked whether he was qualified as a doctor because, over the previous decade, he had never once told her that bread, rice, or breakfast cereals were sugar. She had learned online that if she cut those foods, she no longer needed the drug. Unwin said he was frightened partly because complaints against GPs can last for years, but mainly because “every word she said was true.”

The blood test confirmed her claim. It was the first case of drug-free type 2 diabetes remission he had seen in 25 years of practice.

That incident changed his work. His wife, Jen Unwin, a clinical health psychologist, soon bought him John Briffa’s book Escape the Diet Trap, which described a low-carbohydrate approach to insulin resistance and type 2 diabetes. Unwin said the book gave a medical explanation for what his patient had done. Jen urged him to try one cheerful thing before retirement: do the diet himself and see whether patients would volunteer to do it with them.

His partners initially said no. Low carb was not respectable, he said, and they were concerned about using practice resources. Jen suggested that they work for free in the evening. They found 18 volunteers; with David and Jen, the group had 20 people. A nurse, Heather, also volunteered her time.

The early results startled him. Liver function improved first, often within weeks, sometimes by a third or 50%. Weight fell. Blood pressure improved. HbA1c, the average “sugariness” of blood over the previous three months, improved sharply. Patients also reported something he did not understand at first: they were not hungry. Some began asking whether breakfast was necessary. Unwin experienced the same thing himself.

This was the beginning of his low-carbohydrate work in 2013. Later, while discussing food addiction and relapse, Unwin said he had achieved drug-free type 2 diabetes remission 157 times.

The reason he developed his “teaspoon of sugar equivalent” system was not to claim that rice, bread, or potatoes literally contain that amount of table sugar. It was a clinical communication tool. Glycaemic index compares carbohydrates with pure glucose. Glycaemic load, which Unwin prefers, accounts for portion size and predicts the effect of that portion on blood glucose. But “grams of glucose” does not mean much to many patients. A teaspoon of sugar does.

So he converted glycaemic load into four-gram teaspoons of sugar. That is the unit he uses in clinic to explain what a portion of carbohydrate is likely to do to blood glucose.

In the demonstration with Bartlett, Unwin laid out cornflakes, a potato, 150 grams of boiled rice, a ripe banana, and a chocolate bar. Bartlett guessed how many teaspoon equivalents each contained, trying to answer as he might have before doing years of health interviews. He guessed one teaspoon for cereal, one for potato, one for rice, one for banana, and three for the chocolate bar.

Unwin’s answers, using his glycaemic-load-based system, were very different. The distinction matters because one on-screen summary briefly showed the potato as “0 grams,” meaning no literal sugar listed in that frame, while Unwin’s later answer treated the potato as a starch portion with a large blood-glucose effect.

A separate on-screen chart showed several common foods through the same lens.

Another on-screen chart showed why glycaemic load can differ from glycaemic index. Watermelon has a high glycaemic index, but because it is mostly water, a 120 gram serving was shown as 1.5 teaspoon equivalents, compared with banana at 5.8, black grapes at 4.0, and a Golden Delicious apple at 2.3.

Unwin said the white-rice fact has “astonished people all over the world.” His charts, produced through the Public Health Collaboration, are available in 35 languages and are not copyrighted; he said he wants people to “steal it, take it, use it.”

The deeper point is that people often think they have given up sugar because they no longer add it to tea or coffee, while still eating large amounts of starch. Patients often tell him they cannot understand why their blood sugar remains high after cutting obvious sugar. The answer, in his view, is that they have not counted bread, cereal, rice, pasta, potatoes, and similar foods as glucose-producing.

Unwin used a second demonstration to make the scale of blood glucose tangible. He asked Bartlett how much sugar would be present in all five litres of his blood if his blood sugar were normal. Bartlett guessed a cup.

Unwin held up one sugar cube.

That, he said, is all the glucose present in the entire bloodstream at normal levels. His point was that glucose is both vital and, at excess levels, toxic. The body controls it tightly. For someone with type 2 diabetes, he said, a banana can be too much; in his own case, a whole banana doubles his blood sugar because he cannot regulate glucose normally.

This is why he thinks continuous glucose monitors matter. A person can see, in real time, what a food does to their own blood sugar. He wore one during the interview and later showed his phone reading 5.5, with a flat graph. He said that was the blood sugar of someone with type 2 diabetes who had eaten in a way that did not spike it. Stress, he added, can also raise his blood glucose; he joked that because Bartlett had been kind and made him feel safe, there had been “no spiking.”

For Unwin, a continuous glucose monitor turns abstract advice into personal evidence. Bartlett said he had used one and learned that foods he thought contained no sugar had “loads” in their blood-glucose effect, including ketchup. Unwin replied that once a person has seen that on their phone, they cannot unsee it.

He described CGMs as “the cavalry coming over the hill” because, in his view, they make it harder for packaging and advertising to mislead people. Rather than relying on whether a food looks healthy, tastes sweet, or carries a wellness claim, the wearer can see the biological response.

That matters because Unwin repeatedly returned to packaging and food marketing as sources of confusion. Bartlett gave the example of a dried fruit snack business that had pitched on Dragons’ Den: dried pieces of exotic fruit, marketed as fruit, but 60% to 70% sugar on the back of the pack. Bartlett called it “basically candy.” Unwin agreed and said the word “fruit” creates a health halo that helps the product sell while obscuring the consequence.

Orange juice received the same treatment. Unwin said it contains a lot of sugar, and that juicing removes the fruit from the structure in which it was “meant to be,” creating a faster sugar hit. Blood sugar rises rapidly, insulin responds, blood sugar falls, and the person becomes hungry again. He connected that cycle to his own biscuit habit: biscuits raised his blood sugar; insulin then came in “heavy and slow,” leaving him with low blood sugar that felt like panic, so he ate more biscuits.

The problem, he argued, is not only the immediate glucose spike. It is the hunger that follows. Bartlett said Mars bars, rice, or similar foods can make him hungrier later, driving more sugar intake the same day and the next. Unwin said that was exactly what his patients noticed in 2013: the absence of hunger once they changed diet. He had been hungry all his life because he had been eating a carbohydrate-heavy diet, and had not realized “the more carbs you eat the hungrier you become.”

Highly processed carbohydrate can therefore produce a loop: spike, insulin response, hunger, repeat. In Unwin’s description of a typical day, many people eat cereal and orange juice at breakfast, snacks or crisps mid-morning, a muffin at school, sandwiches at lunch, cake or ice cream, then chips or pizza in the evening. “Sugar with your sugar, with your sugar,” he said. What is missing in that pattern is protein.

His practical advice for someone with insulin resistance or type 2 diabetes was to base meals on protein first: chicken, eggs, or another protein source; then green vegetables such as beans, salad, or frozen veg; then make those vegetables tasty with full-fat mayonnaise, butter, olive oil, or similar fats rather than sugary sauces. Bartlett noted that a bottle of barbecue sauce he had checked contained roughly 30 sugar cubes. Unwin said packet foods require vigilance and that “real food” not in packets is safer.

If people do eat from packets, Unwin said they should look not only at sugar, but at total carbohydrate. Bartlett said he usually looks at the “added sugar” line. Unwin said that is an error. The carbohydrate content matters because carbohydrate turns into sugar. He also noted that labels differ between the UK and the United States: in the UK, carbohydrate is listed separately from fibre, whereas in the US carbohydrate includes fibre, so the consumer needs to understand how much carbohydrate is actually absorbed.

He advised looking at the three macronutrients: protein, carbohydrate, and fat. Protein matters for muscle. Fats may matter for fat-soluble vitamins. Carbohydrate should be examined in light of the person’s insulin sensitivity, activity, and metabolic status. He was not arguing that every person must eat the same amount of carbohydrate. He repeatedly said “it depends.”

Bread was the clearest example. Bartlett cited research saying one slice of white bread contains about half a sugar cube of actual sugar and a loaf can contain up to 12 cubes. Unwin replied that this misses the wheat itself turning into glucose. On his teaspoon-equivalent system, he said, even a small slice of brown bread is about three teaspoons of sugar. An on-screen chart showed one 30 gram slice as 3.7 teaspoons for white bread, 3.3 for brown, 4.0 for 100% whole grain rye, and 2.6 for wholemeal stoneground flour.

Asked whether there is healthy bread, Unwin answered conditionally. If someone is young, exercises a lot, and has good insulin sensitivity, brown bread may be fine. For someone like him with insulin resistance, it would need to be low-carb bread; he would not normally eat bread otherwise. His recommendation was to experiment and measure. Try the bread and see, using a CGM or other feedback, what happens.

David Unwin did not present the ketogenic diet as a universal prescription. He framed carbohydrate reduction as a spectrum, with the right level depending on a person’s goal and response. The first question, he said, is what the person wants: weight loss, type 2 diabetes remission, mental clarity, psychiatric symptoms, or something else.

Steven Bartlett described using keto primarily for cognitive performance. His work requires long interviews, television appearances, meetings, and high verbal fluency. Some days he feels mentally sharp and articulate, while on other days he is embarrassed by his inability to “string a sentence together.” On keto, he said, he is consistently clearer. He also listed aesthetics, strength, and long healthspan as goals.

Unwin recognized that specificity as important. Bartlett had articulated a preferred future, and the more specific the goal, the more likely success becomes. That led into Jen Unwin’s behaviour-change model, which David Unwin demonstrated live: GRIN.

GRIN stands for Goals, Resources, Increments, and Noticing. The model, as Unwin presented it, is a compressed version of clinical behaviour-change work shaped by Jen Unwin’s psychology background.

First, define the goal in concrete terms. “Lose weight” is not specific enough. What difference would losing weight make? Bartlett’s answer became personal: he is afraid of having the same health profile as his father. He remembered walking down steep stairs in Bali to go white-water rafting with his fiancée and thinking that his father, who struggles with stairs, could not have taken part. His goal was not simply looking fit; it was being able to participate in life at 70 or 80 rather than losing mobility.

Second, identify resources. Unwin asked what Bartlett had done in the past that worked. Bartlett began by describing failed goals, including going to the gym every day and getting a six-pack for summer, but Unwin redirected him: what had worked? Bartlett said he had shifted from outcome goals to consistency. For four years, he had aimed to maintain a pattern rather than hit a fixed endpoint.

Third, identify small increments. Bartlett described a WhatsApp group with 10 friends where workouts are shared automatically. Each month the least consistent person is evicted and replaced. The group has medals, a league table, and a physical belt he once won. He also said shrinking the definition of success helped: if a day is difficult, an 18-minute workout still counts and preserves momentum.

Fourth, notice what changes when things go well. Bartlett said he feels better emotionally and energetically, and that consistent behaviour supports his identity as a healthy person who is in control. Unwin connected that to self-esteem.

The significance of GRIN, for Unwin, is that it moves the consultation away from blame and instruction. The doctor does not simply tell the patient what to do. Instead, the patient articulates the future they want, the strengths and supports they already have, the next small action, and the signs that it is working. Unwin said he uses this in nearly every surgery because it lets him discover useful information about the person and makes the work motivational.

That also explains his approach to keto. He asks where the person is now on the carbohydrate spectrum, what they are trying to achieve, what they can realistically reduce, and what happens after they try. Over time, he said, many of his patients go lower in carbohydrate because they like the mental clarity and appetite control.

Bartlett described a similar phenomenon: when he is keto, cravings for cinnamon rolls disappear. Unwin said he had experienced that with Christmas cake, which he called his “kryptonite.” At some point, he said, the brain stops treating those foods as food. It feels like a “superpower.”

The useful test is not whether a diet has a fashionable name. It is whether the person has a clear goal, a realistic next step, and feedback from their body or blood work. For Unwin, carbohydrate reduction is not a slogan; it is an experiment that should be measured.

David Unwin grounds his view of food addiction partly in personal experience. His first wife, he said, lived with severe multiple addictions for 12 years and later died. Living with addiction meant living with uncertainty: not knowing what he would come home to, not being able to fix the person he loved, and watching chaos enter ordinary life.

His current wife, Jen, later recognized her own problem as ultra-processed food addiction. Before that, both of them saw it as a weight problem. She would lose weight and regain it, repeatedly. She would bake tray bakes “for the children” and then eat them herself. He would try to solve the problem by throwing food away or confronting her, which led to defensiveness and arguments. Only later did they understand the behaviour as addiction: intelligent people repeatedly doing things that harm them.

Unwin said Jen would put the prevalence of some aspects of ultra-processed food addiction at about 14% of the population. A DOAC community note on screen referred to a British Medical Journal analysis drawing on 281 studies across 36 countries, estimating that 14% of adults met criteria for food addiction using the Yale Food Addiction Scale. The note added that this reflected addiction-like eating behaviours, such as intense cravings and withdrawal symptoms, rather than a formally recognised diagnosis for all foods.

Unwin’s most vivid clinical example was a 55-year-old business owner with type 2 diabetes, severe obesity, and destroyed knees. He needed major surgery, but his blood sugar control was so poor that the anaesthetist would not operate. Low carb helped for a while, but the weight returned repeatedly over four years. Unwin saw him monthly, trying to understand why.

The explanation came from the man’s wife. She told Unwin that her husband was getting up at four in the morning to eat bread from the fridge. She began putting leftover bread in the bin. He ate it from the bin. She poured detergent on the bread. He still ate it. Finally, she sprayed bleach on the bread and left the bleach can beside the bin so he would know not to touch it.

Unwin said that is addiction. The man was not unintelligent. He was a successful business owner. But bread had become a compulsive trigger.

What helped him, according to Unwin, was not a single intervention. It was low carb, a continuous glucose monitor, clinical support, and a low dose of a GLP-1 drug such as Ozempic. The medication, he said, reduced the “noise” and cravings in the man’s head. The CGM gave immediate feedback. Low carb and support supplied the framework. With all three, the man could abstain, though Unwin said he could not moderate. He eventually had the operation, but Unwin emphasized that maintenance remains lifelong.

That distinction between abstinence and moderation is central to his advice for people who cannot control certain foods. If a food has addictive potential for a person, “one biscuit” may not be a real option.

He said the first step is hard because people make excuses. In his own case, he rationalized biscuits as a reasonable response to work stress. It took him a year to stop. He weaned himself from chocolate ginger biscuits to plain digestives, then oat biscuits, then almonds. Jen believes in cold turkey and that he “should have” stopped all at once, but he did not.

The support step has a danger. If family or partners become heavy-handed, they may drive the person into secrecy. Bartlett described a relationship earlier in life where the other person’s intense health focus made him hide wrappers from foods he was eating. Unwin said that is exactly what happens. Jen did the same with him, hiding wrappers in the car. If policing forces deceit, the supporter should back off, because deceit damages self-esteem and can worsen the behaviour.

This was one of Unwin’s most practical distinctions: help is not surveillance. The person needs gentle support and tolerance, not judgmental enforcement.

He gave another diabetes example to show the stakes. One patient achieved drug-free remission, disappeared for a while, and returned with two dead toes that had to be amputated. Unwin said diabetes had damaged the blood supply, especially to the toes. The patient’s wounds were slow to heal because blood sugar was high. Unwin helped him back into remission, but only with renewed determination from him and his wife. He rejected the phrase “carb creep” for such cases. If a person loses part of a foot, he said, something more serious is going on.

David Unwin’s cancer point was less about treatment than prevention. Earlier, he had said cancer is a rising cause of mortality among people with diabetes and that eight forms of cancer are strongly associated with diabetes. When Steven Bartlett later raised the issue through a young friend with breast cancer, Unwin’s response was to ask why public discussion spends so much time on treating cancer and not enough on preventing it.

Bartlett then read a series of research claims prepared by his team. He said a French study associated 100 millilitres of sugary drinks per day with an almost 20% increased risk of overall cancer. He also read claims about diet drinks and early-onset colorectal cancer, sugar-sweetened beverages and estrogen-dependent endometrial cancer, sugary soda and telomere shortening, high sugar intake and chronic hyperinsulinemia, fructose metabolism, and inflammatory markers associated with tumour progression and metastasis.

Unwin did not walk through those mechanisms one by one as his own evidence review. He clarified the term chronic hyperinsulinemia as high insulin levels, connecting it to his earlier explanation of insulin. His broader response was that the prevention question deserved more attention. When he tells patients they have cancer, he said, “you feel it right here” because the diagnosis takes so much away. If diet is a major contributor after smoking, he asked, how serious does the evidence need to become before prevention is taken seriously?

Neither Bartlett nor Unwin claimed that all cancer is caused by diet. Bartlett made the limit explicit. His argument was about regret and avoidable contribution: he sometimes imagines receiving a devastating diagnosis and being told that lifestyle choices had contributed over the previous 5, 10, or 15 years. In that imagined moment, no sugary drink would seem worth it.

Unwin connected that mortality exercise to his own lifelong preoccupation with death. Death frightened him as a child, he said, but the fear gave him a drive not to waste time and to ask what the best use of today is.

That led into the distinction between lifespan and healthspan. A community note defined lifespan as total years lived and healthspan as years lived healthy, free from serious disease or disability. Unwin said healthspan is going down in the UK even as lifespan is “stuttering along.” Bartlett then read figures he attributed to 2024 to 2026 data from the Office for National Statistics and the Health Foundation: healthy life expectancy in the UK has fallen by roughly two years over the last decade; men and women can expect about 60 years in good health; and people may spend up to roughly 23 years at the end of life in poor health, chronic illness, or disability.

Unwin said that was exactly the point. The cost is not only personal suffering. He said government figures show every taxpayer in England pays an extra £7,000 per year for the consequences of ultra-processed food. He framed the larger cost as more than drugs: people not paying tax, not able to work, and not well enough to participate economically, including many young people.

Bartlett added a US comparison, saying the United States has the largest healthspan-to-lifespan gap on Earth and that Americans, as things stand, will be sicker for longer. Unwin replied that the UK is “trying to catch up,” darkly joking that it is doing its best.

The practical, low-cost test Unwin offered was waist-to-height ratio. Waist should be less than half height. His method is a piece of string as long as the person is tall, cut in half, then wrapped around the fattest part of the belly. If half the string fits around the waist, the person passes.

Bartlett, who said he is 6 foot 1, tried the test and just passed. Unwin emphasized that it was close and that belly fat matters more than fat on the legs or arms because abdominal fat is more associated with insulin resistance.

The larger significance of the string test is that it is cheap, visual, and actionable. It does not require a full medical screening. Unwin is interested in such measures because many of the people at risk will not access expensive testing, and because health inequality is worsening. He said he works only in the NHS, does not take private patients, and is troubled by the way health outcomes worsen as one moves north in the UK.

Bartlett described private health screening services he has used or invested in as attempts to make fuller testing cheaper and faster. Unwin’s caution was that screening should not simply frighten people. It should be connected to actions they can take. Otherwise it can leave people worried without improving outcomes.

David Unwin’s first principle on supplements was food first. His preferred diet is “real food,” lower or lowish carbohydrate, with plenty of protein and healthy fats. He said he is interested in farming and regenerative agriculture, and he asserted that the nutrient profile of crops grown today is not as good as it was 100 years ago because of soil problems, particularly for zinc and magnesium.

That claim remained his assertion, not a separately demonstrated point in the source. But it explained why he treated magnesium as the supplement exception. Although he would like people to get nutrients through diet, he said some things are now difficult to obtain in sufficient amounts. Magnesium, in his view, is one of them. He also said magnesium becomes harder to absorb with age, and some medications interfere with absorption, especially drugs for acidity.

He said magnesium supplementation was “magic” for his own muscle cramps and improved his sleep. But the type matters. For people who tend toward constipation, he suggested magnesium citrate because it has a more laxative effect and some is absorbed. If bowels are not a problem, and the goal is sleep or mood, he mentioned magnesium glycinate or threonate, adding that threonate crosses the blood-brain barrier but will not help constipation.

His memorable explanation came through a story about buying a cow. Jen once surprised him by taking him to a field in Lancashire and telling him she had prepaid for any cow he wanted, because he had always wanted one and they had a field. The farmer told him he could choose any cow, but only if he promised to buy magnesium supplements. The farmer had lost 15 cows that year to “the staggers,” which Unwin described as fitting caused by magnesium-poor grass.

At the same time, Unwin had a patient who kept having fits severe enough to require intensive care admissions, with no brain tumour or obvious explanation. He eventually connected the dots: magnesium deficiency related to the patient’s medication. That was the first time he took magnesium seriously.

Unwin also said blood magnesium is difficult to interpret because magnesium is mainly inside cells; serum magnesium may not reflect intracellular levels. His practical recommendation was therefore to try a magnesium supplement and see how one feels.

Steven Bartlett listed the supplements he takes: vitamin D, magnesium, creatine, a fibre supplement intended to reduce LDL cholesterol, omega-3 after testing showed deficiency, and a multivitamin. Unwin broadly approved but warned that many patients bring out “a carrier bag” of supplements and that oversupplementation is possible, including with vitamin D and other vitamins.

His general rule was conservative: start with food, use supplements for a reason, and avoid taking large stacks simply because they are available. Testing can help, but only if the result leads to an action rather than anxiety.

The underlying tension in David Unwin’s account was professional humility. For his first 25 years in practice, he tried to follow guidelines and be a good doctor. But the population he cared for was getting sicker, not healthier. He also felt uneasy about prescribing many drugs. If someone was taking six tablets a day, he asked, in what sense were they well?

His most painful admission was about weight-loss advice. He used to tell patients to eat less and move more. He sometimes invoked a harsh starvation analogy from Belsen, which he later acknowledged as a horrible thing to say because it blamed the patient. He advised two tablespoons of All-Bran, skimmed milk, multivitamins, and more skimmed milk. When it did not work, he blamed the patient.

That was part of his epiphany. The failure was not theirs; it was his. He had given advice that did not work and then blamed them for failing to follow it. “If you keep giving the same advice to people and it doesn’t work,” he asked, “shouldn’t I have questioned?”

He now sees lifestyle as central rather than an add-on to medication. He still presented Metformin fairly: at the time, the evidence and guidance supported it as good practice for type 2 diabetes, and not using it could be regarded as poor practice. But his patient’s remission without Metformin showed him that diet could change the disease process more profoundly than he had previously believed.

The credibility of his claims, in his own framing, rests on real-world data rather than selected ideal patients. He emphasized that he works in an NHS practice in the north of England and cannot cherry-pick whom he treats; patients are allocated by the state. His publications, he said, are based on baseline and follow-up data from ordinary health service patients near Liverpool. If similar results are replicated in Australia, New Zealand, North America, and elsewhere, he suggested, perhaps the model is true.

He ended by returning to children. Type 2 diabetes, he said, is now a new problem for pediatricians, doctors who treat children. A group of pediatricians had asked him to give a keynote because they had little training in the condition as a disease of children. For Unwin, that is the clearest sign of how far the metabolic problem has moved: a disease once associated with maturity is now appearing in childhood.