Can type 2 diabetes be reversed? What the evidence says

For decades type 2 diabetes was taught as a progressive, irreversible disease: it got worse and medication only slowed it down. We now know that is not accurate. And the change did not come from an internet guru but from published clinical trials and an international consensus.

First, the right word: remission

Professionals do not say "cure" or "reversal", and it is not excessive caution. We say remission, and it is precisely defined.

In 2021 an international panel — including the American Diabetes Association, the European Association for the Study of Diabetes, the Endocrine Society and Diabetes UK — published the consensus definition:

And now my own bar, which is higher than that. 6.5% is the diabetes threshold, but below it there is still the prediabetes band. So whenever it is possible, I prefer not to talk about remission until the numbers are fully normal:

I say whenever it is possible on purpose: it is not always possible, and it is not always the goal. For some people getting below 6.5% is already a huge result, and staying there, on less medication and in better health, is exactly where they should be. What I do not do is call it remission when it is not quite there yet.

The difference from "cure" matters, and it is honest: the predisposition is still there. If you return to the previous situation, the glucose returns. Remission means that right now, without drugs, your values are normal. It is an enormous achievement, but it is not a free pass.

Why it is possible: the explanation that changed everything

Professor Roy Taylor, of Newcastle University, proposed an explanation that today underpins much of what we do. Here it is without jargon.

Each person has a personal fat threshold: an amount of fat they can safely store under the skin. That threshold varies enormously from one person to another — which is why, in my practice, I have seen people develop diabetes at 78 kg and others not develop it at 110. It is not the norm, but it happens.

When you exceed it, the surplus fat has to go somewhere, and it goes where it should not: first to the liver, and from there to the pancreas. A liver full of fat stops responding properly to insulin and releases glucose when it should not. A pancreas full of fat stops producing insulin correctly. The two cycles feed each other.

The practical consequence is hopeful: if that fat is removed, the mechanism can start working again. And not in years. In the Counterpoint study, the liver's handling of glucose normalised within seven days, and pancreatic cell function recovered substantially over eight weeks.

The proof: the DiRECT trial

DiRECT is the study that took this from theory into primary care. It was published in The Lancet in 2018.

306 people with type 2 diabetes of less than six years' duration took part, with overweight or obesity (BMI 27–45) and not on insulin. The intervention was delivered in primary care, not in a specialist hospital.

Almost half, in remission at one year. No surgery and no new drugs.

One nuance: in DiRECT remission was measured after at least two months off medication, not the three of the 2021 definition.

How much weight you need to lose

This is the most useful part of the study, because the relationship turned out to be almost linear:

Read it slowly. This is not "people who look after themselves improve a bit". It is that the amount of weight lost predicts the outcome. And it also says the opposite, which has to be said: losing 3 kg is usually not enough for remission, even if it improves other things.

What the weight you lose is made of

What predicted remission was how much weight came off. But what that weight is made of is a different question, and it is the one that decides whether you can hold on to it.

When you lose weight you do not only lose fat: you lose muscle too. And muscle is precisely the store where most of the sugar you eat ends up — around 80% of what insulin clears. Losing weight at the expense of muscle takes room away from glucose just when you need it most.

And looking five years out, which is where the real problem is: remission held mostly in those who kept the weight off. Keeping weight off is far easier with muscle than without it.

Which exercise, in what order and how much

Who it is most likely in

Let us be honest, because this is where a lot of people sell false hope. Remission is clearly more likely if:

• You were diagnosed recently. Under six years' duration the odds are considerably better. The longer it goes on, the more pancreatic function is lost in ways that are hard to recover.
• You do not use insulin. DiRECT participants were not on insulin. That does not mean people who use it cannot improve — they can — it means full remission is less frequent.
• You have weight to lose. When excess fat is the engine of the problem, removing it is what switches it off.
• You can sustain the change. At five years, remission held mainly in those who kept the weight off, but it is not for ever: of those in remission at two years, one in four was still in remission at five.

What you must NOT do

It is one of the reasons doing this with support matters: not only for the plan, but because somebody has to be watching what happens while you change.

So what can you expect?

If you were diagnosed recently, do not use insulin and have weight to lose: remission is a realistic goal and worth pursuing seriously.

If you have had it for fifteen years, use insulin and your weight is normal: full remission is unlikely, and anybody promising otherwise is lying to you. But — and this is also true — you can lower your HbA1c, reduce doses, gain energy, improve your body composition and reduce your cardiovascular risk. That is not remission, but it changes your life and your outlook.

Tell me your case and I will tell you honestly which of the two you are in. I would rather lose a client than sell them a false expectation.