A dominant narrative that gets the problem backwards
Insulin resistance is most often told the same way: too many carbohydrates, too much insulin, cells that can no longer use fat. The answer that follows seems obvious: cut carbohydrates and push the body to burn more fat. Mathys Chaslin, biologist and nutritionist specialising in metabolic health, does not dispute that a calorie excess damages metabolism. He disputes the direction of causality.
His thesis, set out at the Simplycure webinar on 1 October 2026: in people with type 2 diabetes, fats are not poorly oxidised; if anything, they are oxidised too much. Glucose, on the other hand, is poorly oxidised, and it is the excess of fatty acids that slows its use. To support this, he draws on clamp and metabolic chamber studies, and acknowledges that several points are still debated.
Two pathways for glucose
Once inside the cell, glucose goes through glycolysis and splits into two pyruvates. Two paths then open up:
- Complete oxidation: pyruvate enters the mitochondrion, goes through the Krebs cycle and then the electron transport chain (oxidative phosphorylation). The yield is about 32 ATP per glucose molecule, and this pathway depends on oxygen.
- Non-oxidative glycolysis: pyruvate is converted into lactate, outside the mitochondrion. The pathway is fast but produces only 2 ATP.
For Mathys Chaslin, it is the split between these two pathways that lies at the heart of insulin resistance, far more than the entry of glucose into the cell alone.
What the Del Prato study shows
In the fasting state
The Del Prato study compares healthy subjects with diabetic patients. In the fasting state, the diabetic patients unsurprisingly have higher insulin, blood glucose and free fatty acid levels. They also have higher lactate, a sign of non-oxidative glycolysis that is already increased.
Under clamp
The researchers then put both groups under clamp, in three conditions. At equal blood glucose and moderate insulin, the diabetic patients take up less glucose, synthesise less glycogen and oxidise less glucose. On the other hand, their non-oxidative glycolysis is increased, and their lipid oxidation is higher than that of healthy subjects. The exact opposite of the idea that people with diabetes "can no longer burn fat".
In hyperglycaemia, glucose uptake normalises without more insulin, simply through a gradient effect. Glucose oxidation nevertheless remains below normal.
With a lot of insulin
Under massive hyperinsulinaemia, uptake, glycogen synthesis and lipid oxidation normalise. Glucose oxidation increases but never reaches that of healthy subjects, and non-oxidative glycolysis reaches its highest level. Of the insulin-dependent share of the glucose flux, healthy subjects oxidise about 85% and direct 15% towards lactate. In the diabetic patients, in the first condition, the split is reversed, at around 25% versus 75%, and it only rises to about 50/50 in the other two.
The authors attribute part of the improvement under insulin to an indirect effect: insulin strongly curbs the release of fatty acids and their oxidation, which lifts part of the inhibition of pyruvate dehydrogenase and reopens the pathway to glucose. For Mathys Chaslin, the conclusion is clear: if the problem were only an impaired response to insulin, massively increasing the doses should correct everything. That is not the case.
In muscle, an enzyme imbalance
Skeletal muscle accounts for most of the insulin-dependent glucose flux. Another study presented by Mathys Chaslin measured the activity of enzymes of glycolysis, the Krebs cycle and the electron transport chain there, in lean subjects, obese subjects who were still glucose tolerant and people with type 2 diabetes.
The result: the more glycolytic capacity dominates oxidative capacity, the lower insulin sensitivity is. The best predictor found is the hexokinase / citrate synthase ratio, highest in people with diabetes, intermediate in obese subjects and lowest in lean subjects.
Free fatty acids, the common causal factor
Why is glucose poorly oxidised? Mathys Chaslin presents two models. The first draws on the Randle cycle, described as early as 1963: the influx of lipid metabolites increases the acetyl-CoA/CoA and NADH/NAD+ ratios, slows pyruvate dehydrogenase, causes citrate to accumulate, and slows phosphofructokinase and then hexokinase. Glucose builds up in the cell, its entry is blocked, and hyperglycaemia sets in upstream. The second model, put forward by Roden, places the brake instead at the level of glucose transport (GLUT4) and its phosphorylation.
The exact site of the block is still debated. For Mathys Chaslin, the key point lies elsewhere: both models share the same cause, excess fatty acids. In a proof-of-concept study in healthy subjects, at identical blood glucose and insulin, a simple infusion of free fatty acids reduces glucose uptake by 54% and its oxidation by 40%. This is lipid-induced insulin resistance.
Low-carb and ketogenic diets: a physiological insulin resistance
Sharply reducing carbohydrates, whether through low-carb, ketogenic eating or fasting, increases circulating free fatty acids and therefore induces insulin resistance. Mathys Chaslin specifies that it is physiological and reversible: when the diet stops, sensitivity returns within a few weeks, depending on the individual. According to him, it does not cause type 2 diabetes. It does show, however, that these strategies do not restore the cell's capacity to oxidise glucose.
The central pillar: adipose tissue saturation
While insulin resistance has several causes, the pillar he considers central is adipose tissue saturation: subcutaneous, but above all visceral, ectopic and intramuscular. Certain lipid intermediates, in particular diacylglycerols and ceramides, disrupt insulin signalling. Hypertrophied adipocytes release more fatty acids and, when they rupture, inflammatory mediators that keep the problem going. Stress hormones also play a part.
The difference with low-carb lies in duration: on a ketogenic diet, free fatty acids rise for as long as the diet lasts; when adipose tissue is saturated, they are there permanently, until it is desaturated.
He also cautions about BMI. An obese person can compensate for a long time through beta-cell hypertrophy and high insulin secretion, with blood glucose still normal. Conversely, a person with a normal BMI but a large amount of visceral and ectopic fat can tip into prediabetes more quickly. Body composition tells you more than weight.
Carbohydrates, hyperinsulinaemia and weight gain
Calorie surplus first
A surplus of 1,000 kcal a day causes weight gain, worsens body composition and promotes insulin resistance, whether it comes from carbohydrates or fats. That is not where the debate lies. The real question, according to him, is whether a high share of carbohydrates in energy intake, at constant calories, causes hyperinsulinaemia and insulin resistance.
On weight loss, he cites Kevin Hall's study (2015), carried out in a metabolic chamber with fully controlled food intake: at equal calories, carbohydrate restriction brings no advantage for fat mass loss, and the study even shows a slight advantage for the highest-carbohydrate diet. The differences observed in real life are mainly due to confounding factors: fewer foods available, an effect on satiety in some people, and above all patient adherence.
Insulin, a brake on metabolic stress
Mathys Chaslin also invites us to stop seeing insulin as a harmful hormone. Beyond glucose entry, it curbs lipolysis, proteolysis, glycogenolysis, gluconeogenesis and ketogenesis, and counters glucagon, adrenaline and cortisol. By lowering circulating fatty acids, it helps the body use glucose. For him, its rise is a response to the poor use of glucose linked to lipotoxicity, not the root cause.
He adds that the ketogenic diet increases stress hormones. Circulating cortisol returns to normal after a few weeks, but the tissue activity of 11β-HSD1, the enzyme that regenerates cortisol, remains increased: according to him, the body does not do without these hormones, it becomes more efficient at using them.
What very high-carbohydrate diets show
Kempner's Rice Diet
Mathys Chaslin presents this study as a proof of concept, not as a recommendation. People with type 2 diabetes and vascular disease followed a diet of about 565 g of carbohydrates, 25 g of protein and 5 g of fat a day, based on white rice, fruit, white sugar and fruit juice. Of 100 patients, followed for 3 months to 11 years (22 months on average), 63 improved, 22 remained stable and 15 deteriorated. Mean fasting blood glucose fell from more than 2 g/L to 1.5 g/L, insulin requirements from 25 to 17 units a day, and glycosuria fell markedly.
He himself highlights its limitations: an average weight loss of 6 kg, a diet that was not controlled and very hard to follow, and unknown adherence.
Anderson's study (1979)
To rule out weight loss, he cites a metabolic ward study in 20 people with type 2 diabetes on insulin, with a high-carbohydrate, high-fibre, isocaloric diet at stable weight. In 16 days, insulin doses fell on average from 26 to 11 units a day and insulin therapy could be stopped in more than half of the patients, with improved fasting and postprandial blood glucose.
Brunzell's study
Finally, to rule out the effect of fibre, he cites a metabolic chamber study: after 10 days on a very high-carbohydrate diet, mainly dextrose, people with type 2 diabetes have lower fasting blood glucose and insulin, and an improved area under the curve on the oral glucose tolerance test, with no increase in the insulin response. For Mathys Chaslin, if a relative excess of carbohydrates created hyperinsulinaemia and insulin resistance, these results would be impossible.
He also mentions populations in which more than 70%, or even more than 90%, of energy intake comes from fibre-rich carbohydrates, such as tubers and sweet potato, without type 2 diabetes or obesity.
These studies were carried out under medical supervision. Any change to insulin therapy or antidiabetic treatment is a matter for the prescribing doctor.
Reading a blood glucose curve: the descent matters
Mathys Chaslin draws attention to how curves are read. People often focus on the peaks, whereas the speed of return to normal weighs more heavily in the area under the curve, the parameter that reflects insulin sensitivity. Rapid excursions to 1.6 or 1.8 g/L can be seen in very insulin-sensitive people without this being a problem. A high peak followed by a rapid descent can also point to a defect in the first phase of insulin secretion, a situation different from insulin resistance.
Acting early, while beta cells are still there
The message he considers most important: at the time type 2 diabetes is diagnosed, about 50% of pancreatic beta cells are already lost, and this loss is irreversible. Once past this stage, even improved insulin sensitivity can leave blood glucose disturbed.
Hence the value of spotting a rise in fasting insulin or HOMA early, while beta cells are still fully functional and the scope for action is greatest.
In practice: Mathys Chaslin's strategy
During the question session, he described how he works with patients who are receptive to this approach:
- Protein at around 1.6 to 2 g per kilo of body weight, never as low as in the Rice Diet.
- Fat at around 20% of energy intake at the start, then 25 to 30%, temporarily lower in some cases, rarely below 15%. In a deficit, he would rather reduce fat than carbohydrates, to preserve satiety.
- Carbohydrates relatively high, preferably fibre-rich, adjusted with the patient.
- The calorie deficit, when needed, serves to desaturate visceral and ectopic adipose tissue: it is the underlying lever. Intermittent fasting does not in itself improve insulin sensitivity, but can help some patients sustain this deficit, and make others give up.
- Physical activity supports mitochondrial oxidative capacity. He points out that per unit of oxygen and per unit of time, glucose is a more advantageous energy substrate than fatty acids.
For very advanced type 2 diabetes, with substantial beta-cell loss, a long-term ketogenic diet can be defended if the patient wishes. He points out that another option exists, to be discussed with the doctor: acting on the causes, desaturating adipose tissue, and adding insulin therapy if needed. The choice belongs to the patient, and adherence is part of the strategy.
Finally, he notes that falling oestrogen, from the two years before menopause, promotes visceral storage in a surplus situation, as does inflammation, which sustains a cycle that is hard to break.
What practitioners can take away
For Mathys Chaslin, insulin resistance is first and foremost a lipid overload: lipotoxicity, saturation of visceral and ectopic adipose tissue, then glucotoxicity. Hyperinsulinaemia is a response to it, not the cause, and a relative excess of carbohydrates, at constant calories, does not trigger it. The priority is to act on modifiable factors, early, with a strategy the patient can sustain over time. A reading that runs counter to part of current thinking, and that every practitioner can test against their own practice.
Watch the webinar
The full replay (in French) is available below, with chapters. The last part is devoted to practitioners' questions: intermittent fasting, very active populations, protein intake in the Rice Diet, insulin-dependent tissues and the role of physical activity.
Find the whole micronutrition ecosystem on Simplycure
Simplycure brings together more than 4,500 products from 250 brands on a single platform: built-in comparison tool, intake questionnaires, expert protocols and a personalised shop to share with your patients. More than 15,000 practitioners use it every day. Create your free practitioner account and build multi-brand recommendations in a single order for the patient.
