2026-07-08 · type 2 diabetes, remission, DiRECT, DIADEM-I, STAMPEDE, clinical trials, weight loss · 16 min read
Written by Nora Kim
Nora Kim covers medical and surgical weight loss options, GLP-1 therapies, and evidence-based supplements. She focuses on explaining clinical research, safety considerations, and practical next steps so readers can discuss treatment choices with their care teams.
Type 2 Diabetes Remission via Weight Loss: What DiRECT, DIADEM-I, and STAMPEDE Actually Show About the 15 kg Threshold
Type 2 diabetes remission — an HbA1c below 6.5% sustained off all glucose-lowering medication — is real, replicated, and, in early disease, reachable with a defined amount of weight loss. The clearest signal comes from the DiRECT trial (Lean 2018, Lancet), a UK primary-care randomized study in which 46% of the very-low-calorie diet (VLCD) arm reached remission at 1 year versus 4% of the control arm. At two years, 36% were still in remission (Lean 2019, Lancet Diabetes and Endocrinology); at five years, about 11% overall — but 64% of those who kept off at least 10 kg (Lean 2024, Lancet Diabetes and Endocrinology). DIADEM-I (Taheri 2020, Lancet Diabetes and Endocrinology) replicated the finding in a MENA population with 61% remission at 1 year in early diabetes.
The mechanism is Roy Taylor’s twin-cycle hypothesis: ectopic fat in the liver and pancreas drives hepatic insulin resistance and beta-cell de-differentiation, and weight loss below an individual’s personal fat threshold clears it (Taylor 2018, Cell Metabolism). This guide covers the ADA/EASD 2021 remission definition, the four-driver mechanism, what DiRECT, DIADEM-I, Look AHEAD, and STAMPEDE actually show, a five-step DiRECT-style protocol, and how lifestyle, bariatric surgery, and GLP-1 medications compare on 1- and 5-year remission rates.
Who this is for — and who it is not for
This guide is for adults with type 2 diabetes — especially within about six years of diagnosis and not yet insulin-dependent — who want to understand whether remission is realistically achievable through weight loss, and what the evidence-based protocol looks like.
It is not for:
- Type 1 diabetes — an autoimmune loss of beta-cell mass; the twin-cycle mechanism does not apply. See our type 1 diabetes and weight loss guide.
- Monogenic (MODY) diabetes — remission genetics are different; specialist input is required.
- Long-standing insulin-requiring type 2 diabetes with markedly reduced beta-cell reserve — remission is much less likely, and deprescribing is higher risk. HbA1c reduction and cardiovascular protection remain worthwhile targets — see cardiovascular disease and weight loss for the SELECT-era MACE evidence that applies here — but the DiRECT model was not designed for this group.
- Type 2 diabetes with active DKA, hyperosmolar hyperglycemic state, or a recent cardiovascular event — stabilize first, then discuss weight-loss protocols with the treating clinician.
If you have prediabetes rather than diabetes, the intervention conversation is prevention rather than remission — see our prediabetes and weight loss and diabetes prevention program guides. If your broader picture is central adiposity plus insulin resistance without an A1c in the diabetic range, see metabolic syndrome and weight loss.
The remission definition — 5-row primer
The 2021 ADA/EASD consensus (Riddle 2021) supersedes the older 2009 Buse definition and is the framework every current trial and guideline uses.
| Term | Threshold | Duration | Medication status | Notes |
|---|---|---|---|---|
| Remission (2021 ADA/EASD) | HbA1c < 6.5% (48 mmol/mol) | ≥ 3 months | Off all glucose-lowering medication, including GLP-1 receptor agonists and SGLT2 inhibitors | Current standard; the definition used by DiRECT 2 y and 5 y and by remission registries |
| Partial remission (2009, older) | HbA1c 5.7–6.4% | ≥ 1 year | Off medication | Superseded — still appears in older papers |
| Complete / prolonged remission (2009, older) | HbA1c < 5.7% | ≥ 5 years | Off medication | Rare in real-world data; mostly a research concept |
| Confounders that can invalidate an A1c | — | — | — | SGLT2 washout, GLP-1 washout, honeymoon phase after diagnosis, hemoglobinopathy, iron-deficiency anemia, recent transfusion — an HbA1c under these conditions may not reflect true glycemic status |
| What remission is NOT | — | — | — | Cure; permanent; guaranteed to hold if weight is regained; a signal that microvascular surveillance can stop |
The four-driver mechanism — Taylor’s twin cycle
Roy Taylor’s twin-cycle hypothesis (Taylor 2018, Cell Metabolism) explains why type 2 diabetes is mechanically reversible in many people and why weight loss below an individual threshold — not below a population BMI — is the trigger.
Driver 1 — hepatic fat and fasting hyperglycemia. Excess energy intake drives triglyceride storage in the liver. A fatty liver becomes insulin-resistant and over-produces glucose overnight, elevating fasting glucose. The pancreas compensates by secreting more insulin, encouraging more fat storage — the cycle self-reinforces.
Driver 2 — pancreatic fat and beta-cell de-differentiation. Fat eventually accumulates inside the pancreas, and lipotoxicity drives beta cells to dedifferentiate — losing insulin-secretory identity rather than dying outright (Steven 2016). Post-meal glucose rises, and the clinical diagnosis emerges.
Driver 3 — weight loss depletes ectopic fat in a predictable order. In the Counterbalance study (Taylor 2013, Diabetologia), an 8-week 600–800 kcal/day VLCD cleared liver fat within days (fasting glucose normalized in a week) and pancreatic fat over about 8 weeks (beta-cell function partially recovered). Weight loss below the personal fat threshold is what allows this depletion.
Driver 4 — the personal fat threshold (Taylor 2015, 2020). Each person has an individual capacity for ectopic fat storage before the twin cycle activates — and this capacity is not tightly linked to BMI. South Asian and East Asian populations often cross the threshold at a BMI of 25 or lower, which is why DIADEM-I found high remission rates in adults who by Western BMI cutoffs would not appear obese. Remission occurs when you cross below your own threshold, not below any universal weight.
For related mechanisms of hepatic fat and metabolic dysfunction, see our insulin resistance and weight loss and resmetirom (Rezdiffra) for MASH guides — DiRECT-style loss reverses steatohepatitis on the same weight trajectory that reverses diabetes.
What the evidence actually shows
DiRECT (Lean 2018, 2019, 2024). In a UK primary-care randomized trial of 306 adults with type 2 diabetes of less than six years’ duration and BMI 27–45, an intensive lifestyle arm (12–20 weeks of total diet replacement at ~825 kcal/day, stepped food reintroduction, and long-term weight-loss maintenance support) produced 46% remission at 1 year vs 4% in the control arm (Lean 2018, Lancet). At 2 years, remission was 36% vs 3% (Lean 2019, Lancet Diab Endo). Remission was tightly linked to sustained weight loss: 86% of those who kept off ≥ 15 kg were in remission at 2 years. At 5-year follow-up, overall remission had fallen to about 11%, but participants who kept off at least 10 kg had remission of roughly 64% (Lean 2024, Lancet Diab Endo). The takeaway is durable at the group level and clear at the individual level: hold the weight, hold the remission.
DIADEM-I (Taheri 2020, Lancet Diab Endo). A primary-care VLCD trial in Qatar, Kuwait, and the UAE recruited 158 adults with type 2 diabetes diagnosed within the previous three years. 61% of the intensive arm were in remission at 1 year vs 12% of usual-care controls. DIADEM-I was decisive on two points: DiRECT-style remission generalizes outside a Northern European population, and duration of diabetes at enrollment matters — earlier is better.
Look AHEAD (Gregg 2012, JAMA). In the multi-center Look AHEAD lifestyle-intervention trial for established type 2 diabetes, an intensive lifestyle intervention reduced weight by ~8.6% at 1 year and produced partial remission in 11.5% of participants at year 1 vs 2% in usual care. Look AHEAD ran a less aggressive protocol than DiRECT (moderate calorie deficit, portion-controlled meal plan) and enrolled adults with longer-duration diabetes — both features predict lower remission rates. The signal was real but weaker.
STAMPEDE (Schauer 2017, N Engl J Med). The Cleveland Clinic STAMPEDE trial randomized 150 adults with type 2 diabetes and BMI 27–43 to intensive medical therapy alone, medical therapy plus Roux-en-Y gastric bypass, or medical therapy plus sleeve gastrectomy. At 5 years, remission (HbA1c ≤ 6.0% off medication) reached 29% with bypass, 23% with sleeve, and 5% with medical therapy alone. STAMPEDE anchors the bariatric benchmark and shows the surgical remission signal is durable.
SURMOUNT-2 and SURPASS (Jastreboff 2023; Frías 2021; Rosenstock 2021). The tirzepatide and semaglutide trial programs in type 2 diabetes produce weight loss in DiRECT territory (12–22%) and HbA1c reductions of 2.0–2.5 points from baseline. But these results include the direct drug effect on insulin secretion and gastric emptying — they do not meet the ADA/EASD remission definition, which requires 3 months off the medication.
Time course — what happens over hours, weeks, and years
The DiRECT and Counterbalance timelines show a striking pattern: liver fat clears in days, pancreas fat clears in weeks, and the remission-durability curve is set in years.
| Timepoint | What is happening | Source |
|---|---|---|
| Day 1 of VLCD | Glycogen and water losses account for the first 1–3 kg; insulin levels drop | Lim 2011 |
| Day 7 | Fasting glucose approaches normal as intrahepatic triglyceride depletes | Lim 2011; Taylor 2013 |
| Week 8 | Pancreatic fat approaches non-diabetic levels; first-phase insulin response begins to recover | Taylor 2013 (Counterbalance) |
| Month 6 | Peak weight loss on the DiRECT protocol; the majority of remission events are captured by this point | Lean 2018 |
| Year 2 | Remission durability begins to correlate strongly with maintained weight loss | Lean 2019 |
| Year 5 | Overall remission ~11%; ~64% among those who maintained ≥ 10 kg off | Lean 2024 |
Five-step “how to actually attempt remission” protocol
This is not medical advice — it is a description of the DiRECT protocol so that readers know what a physician-supervised remission attempt looks like.
Step 1 — Screen for candidacy. Diabetes duration under about six years, BMI ≥ 27, not on insulin, no active DKA history, no severe cardiovascular disease within the last six months, and a treating clinician willing to co-manage medication withdrawal.
Step 2 — Coordinate medication withdrawal. With a physician, withdraw insulin and sulfonylureas first (before beginning the VLCD, to avoid hypoglycemia). SGLT2 inhibitors are usually paused with attention to euglycemic DKA risk. GLP-1 receptor agonists are typically discontinued as weight loss reduces the need. Metformin is usually last — it is the safest, and by the point it comes off, HbA1c has been non-diabetic for weeks.
Step 3 — Total diet replacement at ~825 kcal/day for 12 weeks. In DiRECT this used a formulaic meal-replacement (Optifast or similar) providing complete micronutrients. VLCDs at this intensity are for medical use and require monitoring for hypoglycemia, orthostatic symptoms, gout flares, and gallstone risk.
Step 4 — Structured food reintroduction over 4–8 weeks. Guided by a dietitian, gradually re-introduce whole foods while tracking weight, fasting glucose, and appetite. This step is often where adherence starts to slip; structured coaching contact matters.
Step 5 — Long-term weight-loss maintenance protocol. The single biggest failure point in DiRECT was regain during years 2–5 — see weight-loss maintenance for the evidence-based approach (weekly weigh-ins, high protein, resistance training, cognitive-behavioral relapse-prevention). Remission durability is essentially weight-maintenance durability.
Treatment comparison — 1-year and 5-year remission rates
| Option | Typical weight loss | 1-year remission | 5-year remission | Trial anchor |
|---|---|---|---|---|
| DiRECT intensive VLCD lifestyle | ~13 kg at 12 mo; ~9 kg at 24 mo | ~46% | ~11% overall (64% if ≥10 kg kept off) | Lean 2018 / 2019 / 2024 |
| Look AHEAD moderate lifestyle | ~8.6% at 12 mo | ~11.5% partial remission | Not designed to sustain remission | Gregg 2012 |
| Bariatric surgery (RYGB / sleeve) | 25–30% of body weight | ~40–60% | ~23–29% | Schauer 2017 (STAMPEDE) |
| GLP-1 / GIP-GLP-1 (semaglutide, tirzepatide) | 12–22% at 12–18 mo | HbA1c normalization common but does not meet ADA/EASD remission (on-drug) | Not established off-drug | Jastreboff 2023; Frías 2021 |
| Metformin alone | ~2 kg | Not designed to induce remission | ~0% | ADA 2024 Standards of Care |
| Dietary weight-neutral A1c-lowering (e.g. Mediterranean without deficit) | Minimal | Rare | Not established | PREDIMED-Plus is the closest analog |
The comparison chart in bariatric surgery vs GLP-1 medications covers the surgical-vs-pharmacologic decision in more detail; the same trade-offs shape a lifestyle-vs-surgery remission choice.
Special situations
Newly diagnosed vs long-standing diabetes
DiRECT restricted enrollment to under six years since diagnosis, and even within that window shorter duration predicted higher remission. DIADEM-I limited to under three years. Beyond about ten years, beta-cell reserve is usually depleted enough that full remission is uncommon, though HbA1c improvement and medication reduction remain worthwhile targets.
Insulin-dependent type 2 diabetes
Remission is technically possible but the medication-withdrawal side of the protocol is materially higher-risk. Insulin dosing typically has to drop dramatically within the first days of a VLCD; without careful monitoring, hypoglycemia risk is high. This is a specialty-endocrinology protocol, not a primary-care one.
Chronic kidney disease (CKD)
SGLT2 inhibitors and GLP-1 receptor agonists have independent kidney and cardiovascular benefits that a remission protocol removes. In CKD stages 3–5, coordinate with nephrology before withdrawing either class — see our chronic kidney disease and weight loss guide. VLCDs in CKD require specialist monitoring.
MASH / fatty liver disease
DiRECT-style loss reverses steatosis and often reverses steatohepatitis on the same weight-loss timeline that reverses diabetes. See resmetirom (Rezdiffra) for MASH.
Ozempic-face and rapid loss
Rapid weight loss of 15–25% often produces cosmetic changes covered in Ozempic face and weight loss; a DiRECT-style VLCD produces the same effect on a shorter timeline.
Post-bariatric conversion
Some adults enter remission after bariatric surgery, regain weight, and lose remission; the surgical-to-medical-therapy conversion pathway is covered in bariatric surgery vs GLP-1 medications.
South Asian and East Asian populations
Personal fat thresholds tend to be lower — an adult with a BMI of 25 may already have crossed threshold. DIADEM-I is the strongest evidence for effective remission at lower BMIs, and international guidelines now recommend earlier screening and lower BMI thresholds for these populations.
Older adults
In adults over 70, deprescribing risk (hypoglycemia, falls, sarcopenia) usually dominates. Full remission is less commonly the primary goal — HbA1c targets are individualized and moderate loss with medication simplification often provides most of the benefit. See our weight loss for older adults guide.
Six myths and red flags — what to correct
- “Type 2 diabetes is always progressive and never reversible.” DiRECT, DIADEM-I, and STAMPEDE all show it is often reversible in early disease. It is not always reversible — beta-cell reserve matters — but the “always progressive” framing is out of date.
- “You need bariatric surgery to reverse type 2 diabetes.” DiRECT achieved 46% remission at 1 year in primary care with no surgery. Bariatric surgery does outperform lifestyle at 5 years, but it is not required for remission.
- “Any weight loss will produce remission.” The signal clusters around ~15 kg or roughly 15% of body weight for most adults. A 3–5% loss lowers HbA1c meaningfully but rarely produces remission.
- “Remission means cure — you can stop worrying about diabetes forever.” Remission is not cure. Beta-cell function and metabolic vulnerability remain. Continue microvascular surveillance and monitor HbA1c on the same schedule.
- “Keto is the only diet that reverses diabetes.” DiRECT used a formula VLCD, not keto. Mediterranean, low-carb, and total-diet-replacement patterns can all produce the needed weight loss; the active ingredient is the sustained energy deficit and the ectopic-fat depletion it drives.
- “Metformin prevents remission.” Metformin is compatible with remission attempts and is usually the last drug withdrawn. It does not block the twin-cycle-reversing effect of weight loss.
Emergency lines: call 911 for suspected DKA (fruity breath, deep rapid breathing, vomiting, confusion) or HHS during any low-calorie protocol, and call your clinician for a fasting glucose below 70 mg/dL while off medication. Call or text 988 for behavior-change or eating-disorder crisis.
Bottom line
Type 2 diabetes remission is real, replicated across DiRECT, DIADEM-I, Look AHEAD, and the bariatric literature, and defined precisely enough (HbA1c < 6.5% off all glucose-lowering medication for ≥ 3 months, ADA/EASD 2021) that clinicians and patients can hold themselves to the same standard as the trials. The evidence-based path is early diagnosis, ~15 kg sustained weight loss, physician-supervised medication withdrawal, and structured long-term maintenance. Remission is not cure — durability tracks weight-maintenance durability — but for adults within about six years of diagnosis, the DiRECT protocol offers a genuine, well-replicated chance at putting the disease behind them.
Sources
- Lean MEJ, Leslie WS, Barnes AC, Brosnahan N, Thom G, McCombie L, et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial. Lancet (2018).
- Lean MEJ, Leslie WS, Barnes AC, Brosnahan N, Thom G, McCombie L, et al. Durability of a primary care-led weight-management intervention for remission of type 2 diabetes: 2-year results of the DiRECT open-label, cluster-randomised trial. Lancet Diabetes and Endocrinology (2019).
- Lean MEJ, Leslie WS, Barnes AC, Brosnahan N, Thom G, McCombie L, et al. 5-year follow-up of the randomised Diabetes Remission Clinical Trial (DiRECT) of continued support for weight loss maintenance. Lancet Diabetes and Endocrinology (2024).
- Taylor R. Type 2 diabetes and remission: practical management guided by pathophysiology. Cell Metabolism (2018).
- Taylor R, Holman RR. Normal weight individuals who develop type 2 diabetes: the personal fat threshold. Diabetologia (2015).
- Steven S, Hollingsworth KG, Al-Mrabeh A, Avery L, Aribisala B, Caslake M, et al. Very-low-calorie diet and 6 months of weight stability in type 2 diabetes: pathophysiological changes in responders and nonresponders (Counterbalance). Diabetes Care / Diabetologia (2013–2016).
- Lim EL, Hollingsworth KG, Aribisala BS, Chen MJ, Mathers JC, Taylor R. Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia (2011).
- Taheri S, Zaghloul H, Chagoury O, Elhadad S, Ahmed SH, El Khatib N, et al. Effect of intensive lifestyle intervention on bodyweight and glycaemia in early type 2 diabetes (DIADEM-I): an open-label, parallel-group, randomised controlled trial. Lancet Diabetes and Endocrinology (2020).
- Gregg EW, Chen H, Wagenknecht LE, Clark JM, Delahanty LM, Bantle J, et al. Association of an intensive lifestyle intervention with remission of type 2 diabetes (Look AHEAD). JAMA (2012).
- Schauer PR, Bhatt DL, Kirwan JP, Wolski K, Aminian A, Brethauer SA, et al. Bariatric surgery versus intensive medical therapy for diabetes — 5-year outcomes (STAMPEDE). New England Journal of Medicine (2017).
- Mingrone G, Panunzi S, De Gaetano A, Guidone C, Iaconelli A, Nanni G, et al. Bariatric–metabolic surgery versus conventional medical treatment in obese patients with type 2 diabetes: 5-year follow-up. Lancet (2015).
- Aminian A, Brethauer SA, Andalib A, Punchai S, Mackey J, Rodriguez J, et al. Individualized metabolic surgery score: procedure selection based on diabetes severity. Annals of Surgery (2020).
- Riddle MC, Cefalu WT, Evans PH, Gerstein HC, Nauck MA, Oh WK, et al. Consensus report: definition and interpretation of remission in type 2 diabetes. Diabetes Care (2021).
- Buse JB, Caprio S, Cefalu WT, Ceriello A, Del Prato S, Inzucchi SE, et al. How do we define cure of diabetes? Diabetes Care (2009).
- Jastreboff AM, le Roux CW, Stefanski A, Aronne LJ, Halpern B, Wharton S, et al. Tirzepatide once weekly for the treatment of obesity in people with type 2 diabetes (SURMOUNT-2). New England Journal of Medicine (2023).
- Frías JP, Davies MJ, Rosenstock J, Pérez Manghi FC, Fernández Landó L, Bergman BK, et al. Tirzepatide versus insulin glargine in type 2 diabetes (SURPASS-4). New England Journal of Medicine (2021).
- Rosenstock J, Wysham C, Frías JP, Kaneko S, Lee CJ, Fernández Landó L, et al. Efficacy and safety of a novel dual GIP and GLP-1 receptor agonist tirzepatide in patients with type 2 diabetes (SURPASS-1): a double-blind, randomised, phase 3 trial. Diabetes Care (2021).
- American Diabetes Association. Standards of Care in Diabetes — 2024. Sections 3 (Prevention or Delay of Type 2 Diabetes) and 9 (Pharmacologic Approaches). Diabetes Care (2024).