Insulin Resistance(33 expert discussions analyzed)

Quick takeaways: Insulin resistance builds for years before blood sugar rises. Excess body fat, especially around the waist and in the liver, and inactivity are major drivers; sugary drinks are one of the easiest dietary drivers to remove. Resistance training improves insulin sensitivity independently of diet. Fiber supports insulin sensitivity through gut and hormonal pathways. Earlier intervention is far easier than reversing established disease. One important insight is that insulin resistance usually builds for years before fasting glucose or HbA1c reach the prediabetes range. Fasting insulin can reflect this, but it is not a recommended screening test, so in practice the early warning comes from reading blood sugar results alongside triglycerides, HDL, waist size and blood pressure. Beyond that point, the evidence becomes more nuanced. Nearly every expert here converges on cutting sugar and refined carbohydrates — and one argues that carbohydrate restriction misses the real problem entirely, which he says is nutrient deficiency. Experts also disagree on whether fructose is the primary driver or one of many, whether ketogenic eating is necessary or excessive for most people, and how much LDL alone tells us about cardiovascular risk. This topic maps where experts converge, where they differ, and what remains uncertain.

First synthesised Apr 19, 2026·Last reviewed Oct 18, 2026

3558 min of expert content · 29 min read|Summary:|

What matters

Primary Topic Intent

This page helps readers understand what insulin resistance actually is, what drives it, how to detect it early, and which interventions have the strongest evidence for reversing it — so they can have a more informed conversation with their doctor about their metabolic health.

Insulin resistance is a condition in which cells throughout the body stop responding normally to insulin — the hormone that signals cells to absorb glucose from the bloodstream. The pancreas compensates by producing more insulin, creating a state of chronically elevated insulin (hyperinsulinemia) that precedes elevated blood glucose by years or even decades. By the time glucose rises into the prediabetic range, insulin resistance may have been present for years. Some experts therefore argue for measuring fasting insulin; it is not a recommended routine test, because insulin assays are not standardized and there is no agreed cut-off. Andrew Koutnik, a metabolic researcher who manages his own Type 1 diabetes, frames this directly: 'Insulin moves before glucose. If you wait for the glucose flag, you've already missed years.'

The liver is the primary site where insulin resistance begins, according to Robert Lustig, a neuroendocrinologist at UCSF. His central mechanistic argument is that dietary fructose — particularly from added sugars in processed foods — is metabolized in the liver in a way that suppresses mitochondrial function, promotes fat accumulation in the liver, and drives systemic insulin resistance. He proposes that a fructose by-product called methylglyoxal impairs AMP kinase, an enzyme that signals cells to burn energy, and that uric acid from fructose processing limits fat burning; these mechanisms come mainly from laboratory and animal studies. His clinical claim, based on a small 9-day study in children without a control group, is that removing added sugar can improve metabolic syndrome markers within about ten days. David Unwin, a UK general practitioner with a published primary-care cohort, reinforces this with clinical evidence: many common 'healthy' carbohydrates (rice, bananas, wholemeal bread) act as hidden sugars in people who are already insulin-resistant, and in his practice audit, 51% of the 186 patients who chose a low-carbohydrate approach reached drug-free remission of type 2 diabetes (77% of those diagnosed within a year). This was an uncontrolled audit of self-selected patients. The DiRECT randomized trial reached remission by a different route, a low-calorie formula diet: 46% at one year, and at five years, 10 to 13% of those with follow-up data were still in remission (12 of 118 with any five-year data; 11 of 85 in the extension study), which is not a rate for everyone who started. Tim Noakes pushes this framing furthest, treating dietary carbohydrate as largely unnecessary — a position that goes beyond what the trial evidence supports and that sits outside mainstream nutrition science.

Beyond diet, the evidence points to a multi-lever picture. Muscle is the body's largest glucose sink — when you contract muscle, it absorbs glucose independently of insulin. This is why resistance training and post-meal walking are among the most consistently supported non-dietary interventions for insulin sensitivity. Mike Mutzel, a metabolic health educator, argues that exercise may reduce visceral fat — the metabolically active fat around organs — more effectively than dietary change alone. Will Bulsiewicz, a gastroenterologist, adds a distinct mechanism: fiber deficiency (affecting roughly 95% of Americans) promotes gut dysbiosis, which allows bacterial endotoxins called lipopolysaccharides to leak into the bloodstream, triggering chronic inflammation that directly impairs insulin signaling. The short-chain fatty acids produced when gut bacteria ferment fiber also stimulate gut hormones involved in appetite (GLP-1 and PYY), though at levels far below those reached with GLP-1 medicines such as semaglutide.

A minority position deserves separate treatment, because it inverts everything above. Peter Osborne argues that type 2 diabetes is substantially a nutrient deficiency disease: zinc and magnesium are required to manufacture insulin, chromium supports the insulin receptor, and deficiency in any of them may worsen glucose control. The underlying biology is real — these minerals genuinely are involved — and one of his specific points is independently well recognized: metformin depletes vitamin B12, and the resulting fatigue or nerve symptoms are easy to mistake for the diabetes itself. Where he departs from the rest of this evidence base is the claim that carbohydrate restriction misses the real issue. That contradicts the single strongest convergence in this topic. It is also relevant that he sells the nutrient testing and the supplements he recommends. The distinction that matters for a reader: these minerals are necessary for glucose metabolism, but being necessary in a pathway is not the same as being the limiting factor in a disease. Correcting a genuine deficiency can modestly shift numbers. There is no good evidence it changes the course of type 2 diabetes.

The cardiovascular implications of insulin resistance are significant and contested. Cardiologist Aseem Malhotra argues that standard LDL cholesterol is a weak independent predictor of heart disease risk, and that insulin resistance and chronic inflammation are the true upstream drivers of arterial damage. He and Lustig both point to the triglyceride-to-HDL ratio as a more useful marker. A high ratio does go with insulin resistance, but it is a rough guide: cut-offs differ by ethnicity and sex, depend on whether the lab reports in mg/dL or mmol/L, and are not part of clinical guidelines. Lustig's figures (above 2.5 as a concern, below 1.5 as healthy) are his rules of thumb, not validated thresholds, and the ratio does not replace LDL cholesterol or ApoB in assessing heart risk. The metabolic framing adds something real: insulin resistance is an independent cardiovascular risk factor worth recognizing, through routine measures such as blood sugar, triglycerides, HDL, waist size and blood pressure rather than insulin tests, and communicating absolute rather than relative risk is a genuine improvement in how statin decisions get made. But the claim that LDL is not causal runs against a large body of evidence — genetic studies and trials across several different LDL-lowering therapies consistently show that lowering LDL reduces cardiovascular events. The useful part of this argument and the overreaching part are separable, and worth separating.

The downstream consequences of unaddressed insulin resistance extend beyond cardiovascular disease. Benjamin Bikman, a metabolic researcher at Brigham Young University, frames insulin resistance as the common metabolic root under type 2 diabetes, heart disease, polycystic ovary syndrome, and parts of Alzheimer's risk. Dale Bredesen, a neurologist at UCLA, argues that brain insulin resistance — sometimes called 'type 3 diabetes' in research literature — is a central driver of Alzheimer's pathology, and that multi-domain lifestyle intervention addressing metabolic health, sleep, inflammation, and vascular function can reduce cognitive decline risk. This prevention framing is supported by the FINGER trial, a mainstream-backed randomized controlled trial of multi-domain lifestyle intervention. Bredesen's stronger claim — that established Alzheimer's can be reversed — is contested in mainstream neurology.

Best-supported action

Start by removing all liquid sugar — sodas, fruit juices, sweetened coffees, energy drinks, and sports drinks — for 30 days.

Eliminating liquid sugar is the most consistently supported starting point across every expert in this topic — but the right long-term strategy depends on what is actually driving your insulin resistance. Some people will see dramatic improvement from this single change because liquid sugar is their primary metabolic stressor. Others may need a more comprehensive approach because their insulin resistance is substantially driven by sleep deprivation, chronic stress, visceral fat accumulation, or gut dysbiosis — factors that dietary change alone won't fully address. Getting this distinction wrong can mean either stopping too soon after one change when multiple levers are needed, or over-engineering a complex protocol when a simple sugar reduction would have been sufficient.

Sources: 33 expert episodes · See sources

Understand where experts converge, where they differ, and what remains uncertain.

What we still don't know

Fasting insulin is not yet a standard routine lab in most primary care settings, and there is no universally agreed clinical threshold — Lustig suggests below 6 mIU/mL as a target for good metabolic health, but this is not a formal clinical guideline. The HOMA-IR calculation (which combines fasting glucose and insulin) is a research tool that has not been formally validated as a clinical diagnostic standard.

The optimal carbohydrate threshold for reversing insulin resistance is genuinely unknown and likely varies by individual. Koutnik and Unwin both note that most people respond to moderate carbohydrate reduction (100 to 150 grams per day) without needing strict ketogenic eating (under 50 grams per day), but there is no randomized controlled trial directly comparing these approaches at equivalent adherence levels.

Whether cutting carbohydrates reduces heart attacks independently of the weight loss it produces is not established. The carbohydrate-insulin model is often presented as though the cardiovascular benefit follows automatically from the metabolic improvement. The trials that exist mostly show changes in weight and glycaemic markers, not hard cardiovascular endpoints, and the long-term safety of very-low-carbohydrate or carnivore-style eating — particularly for heart and bone health — has not been established.

Whether correcting nutrient deficiencies changes the course of type 2 diabetes, rather than modestly shifting numbers, is not established. Magnesium, zinc and chromium are genuinely required for insulin production and glucose handling, but being necessary in a pathway is not the same as being the limiting factor in a disease. The trials that exist show modest biomarker movement, not altered disease trajectory.

Long-term safety data for sustained ketogenic eating in healthy adults is thin compared to Mediterranean-pattern data. D'Agostino acknowledges that 10 to 20 year outcomes are not well-established, and a subset of people develop elevated LDL and ApoB on ketogenic diets — this should not be dismissed because other markers improve: apoB-containing particles (measured as LDL cholesterol or ApoB) are an established cause of cardiovascular disease. What is uncertain is how much a keto-related rise changes long-term risk, because ketogenic diets lack long-term outcome studies.

The causal direction between mitochondrial dysfunction and insulin resistance is unresolved. Myhill argues mitochondrial damage is upstream; mainstream metabolic research more typically frames chronic insulin and substrate excess as damaging mitochondria over time. Both positions have mechanistic support and both converge on similar first-line interventions, but the ordering matters for downstream supplementation strategy.

Remission becomes less likely the longer type 2 diabetes has been present, even when insulin sensitivity improves with lifestyle change. In Unwin's audit, remission was most likely in people diagnosed recently (77% of those within a year of diagnosis, against 51% overall), and the DiRECT trial enrolled only people diagnosed within the previous six years. With long-standing disease, the insulin-producing beta cells may be too damaged for remission, even when insulin sensitivity improves. Earlier intervention is substantially more effective. Claims that type 2 diabetes is '100% reversible' overstate this.