Type 2 diabetes was for a long time described as chronic and progressive — manageable, requiring escalating treatment over time, and not reversible.
Trial evidence over the past decade has changed that picture substantially for a subset of people, and the change is significant enough to be worth understanding properly.
What the trials showed
A landmark trial examined an intensive weight management programme delivered in primary care, involving a period of total diet replacement using a low-calorie formula, followed by structured food reintroduction and long-term maintenance support.
A substantial proportion of participants achieved remission — defined as glycaemic measures below the diabetic threshold without glucose-lowering medication — at one year.
Remission rates were strongly related to weight loss achieved. Among those losing the most weight, remission rates were high.
Follow-up at two years found that a meaningful proportion sustained remission, with rates declining as weight was regained.
Related research has proposed a mechanism involving reduction of fat in the liver and pancreas, with restoration of beta cell function as that fat is removed.
What remission means
An important definitional point. Remission is not cure.
The underlying susceptibility remains. Blood glucose returns to non-diabetic levels while the conditions producing remission are maintained, and diabetes can return if weight is regained.
Consensus definitions generally specify glycaemic measures below the diagnostic threshold, sustained for a defined period, in the absence of glucose-lowering medication.
People in remission are generally advised to continue regular monitoring, including for diabetes-related complications, because the risk history doesn't disappear.
Who it's most likely to work for
The evidence points fairly clearly at some predictors.
Shorter duration since diagnosis. Remission is substantially more likely in people diagnosed within the past few years than in those with long-standing diabetes. Beta cell function declines over time and the potential for recovery diminishes.
Greater weight loss achieved. The relationship is strong and dose-dependent.
Not requiring insulin. People on insulin, particularly long term, have lower remission rates.
Higher starting weight. There's more to lose and the relationship between fat accumulation in liver and pancreas and glucose control is more likely to be the operative mechanism.
This means remission is not a realistic goal for everybody with the diagnosis, and framing it as universally achievable risks making people feel responsible for an outcome that wasn't available to them.
The routes
Intensive dietary weight management, as in the trials described. Requires structured support and is delivered as a programme rather than as advice.
Bariatric surgery produces high rates of remission, with substantial evidence including randomised comparisons against medical management. It carries surgical risk and requires long-term follow-up and nutritional monitoring.
Substantial weight loss by other means, including through newer weight-loss medications, with evidence accumulating.
Notably, all routes work through weight loss, which supports the underlying mechanism.
What this doesn't mean
Some important qualifications.
It doesn't mean people who don't achieve remission have failed. The predictors above are largely not under individual control, and someone with long-standing diabetes doing everything correctly may not achieve it.
It doesn't mean medication is unnecessary. Glucose-lowering treatment substantially reduces complications, and stopping medication should only happen under clinical supervision.
And it doesn't apply to type 1 diabetes, which is an autoimmune condition with a completely different mechanism. Any suggestion that type 1 can be reversed through diet is dangerous and wrong.
The practical position
If you have been diagnosed relatively recently, are carrying excess weight, and are not on insulin, remission is a realistic possibility worth discussing with your clinician.
The programmes with evidence behind them involve structured support rather than general advice, and availability varies by health system. Asking specifically about remission programmes is worth doing, because they may not be offered proactively.
If remission isn't achievable, weight loss still improves glycaemic control, blood pressure and lipids, and reduces complication risk. Benefit doesn't require reaching a threshold.
And for anybody attempting significant dietary change while on glucose-lowering medication, this must be supervised. Medication doses frequently need adjustment as glucose control improves, and continuing unchanged doses can cause dangerously low blood glucose.
That last point is the most important practical one in this article. Do not undertake substantial dietary change while on diabetes medication without speaking to your prescriber first.
General information only. This is not medical advice. Diabetes management and any changes to treatment must be discussed with a qualified healthcare professional.
What maintenance actually requires
The part that determines whether remission lasts, and it receives less attention than the initial weight loss.
The trials that produced durable results included structured ongoing support — regular contact, monitoring, and a plan for responding to weight regain early rather than after it had accumulated.
That reflects the physiology described in the wider weight literature. The adaptations opposing weight loss persist, which means maintenance is an active process rather than a resting state.
Practically, the people who sustain remission tend to weigh themselves regularly, have a defined threshold at which they act, and have some form of continued support. Framing remission as an achievement to be maintained, rather than a finish line, appears to be a large part of what distinguishes the durable cases.