Type 2 Diabetes Remission: What Doctors Actually Mean by 'Remission'

Can type 2 diabetes go away?

You'll hear wildly different answers.

One person says:

"Diabetes is permanent."

Another says:

"I reversed mine completely."

Someone else says:

"My A1C is normal now, so I'm cured."

The medical terminology is more precise.

Type 2 diabetes can enter remission in some people.

And there is now an internationally agreed definition for what that means.

The basic definition is surprisingly simple

In 2021, an international expert group convened by the American Diabetes Association — with participation from the European Association for the Study of Diabetes, Endocrine Society, Diabetes UK, and other organizations — proposed this definition:

A1C below 6.5% for at least 3 months without glucose-lowering medication.

That's the usual criterion for:

type 2 diabetes remission.

There are nuances.

But that one sentence clears up a lot of confusion.

Remission does NOT require an A1C below 5.7%

This surprises many people.

An A1C of:

6.4%

is still in what is normally called the prediabetes range.

But if someone previously had established type 2 diabetes and now maintains an A1C below:

6.5%

for at least three months without glucose-lowering medication...

they can meet the consensus definition of:

remission.

The definition does not require:

A1C 5.6%.

It does not require:

A1C 5.0%.

It requires glucose to remain below the diagnostic threshold for diabetes without medication maintaining that result.

That is an important distinction between:

"normal blood sugar"

and:

"formal diabetes remission."

They aren't exactly the same term.

Here's an example

Imagine two people who previously had type 2 diabetes.

Person A

A1C:

6.2%

No glucose-lowering medication for four months.

Under the consensus definition:

This can qualify as remission.

Now consider:

Person B

A1C:

5.5%

but still taking metformin specifically to control diabetes.

Their glucose control is excellent.

But formally:

this would not be classified as drug-independent remission.

That's not because Person B is somehow doing worse.

It's because the definition is trying to answer a particular scientific question:

Can glucose remain below the diabetes threshold without a glucose-lowering medication producing that effect?

Why does medication matter to the definition?

Suppose your A1C is:

5.8%

while taking medication.

That's excellent glucose control.

But we can't know from that number alone what your A1C would be:

without the medication.

That's why the international consensus requires a period without glucose-lowering pharmacotherapy before formally diagnosing remission.

The usual requirement is:

at least three months.

That gives enough time for the medication's effect to disappear and for the new glucose environment to become reflected in A1C.

But this creates a modern complication

Some diabetes medications now do much more than lower glucose.

For example:

GLP-1 receptor agonists

may also be prescribed for obesity or cardiovascular benefit.

SGLT2 inhibitors

may be used for:

• heart failure
• chronic kidney disease
• cardiovascular protection

even when glucose lowering is not the primary reason.

Metformin

may sometimes be continued for reasons other than immediate glycemic control.

The consensus acknowledges this problem.

If someone continues a drug that also lowers glucose, researchers generally cannot establish drug-independent remission under the formal definition.

But that does not mean the medication should automatically be stopped.

The label "remission" is not more important than appropriate treatment.

If a medication is providing kidney, cardiovascular, weight-management, or other benefits, stopping it simply to prove you qualify for the word remission may make no clinical sense.

That decision belongs between the individual and their healthcare professional.

So remission and excellent diabetes control aren't the same thing

This gives us three different situations.

Active diabetes with elevated glucose

For example:

A1C 8.2%.

Well-controlled type 2 diabetes

For example:

A1C 5.9% while taking glucose-lowering medication.

Type 2 diabetes in remission

For example:

A1C 5.9% maintained for at least three months without glucose-lowering medication.

All three describe something different.

And importantly:

remission is not a moral achievement.

Someone who needs medication to maintain excellent glucose control has not "failed."

Type 2 diabetes varies substantially between individuals.

The goal is the best long-term health outcome — not winning a terminology contest.

Why doctors say "remission" instead of "cure"

If someone's A1C has been normal for a year without diabetes medication, it can be tempting to say:

"The diabetes is gone."

The expert consensus intentionally advises against the word:

cure.

Why?

Because the underlying susceptibility can remain.

Glucose can rise again.

Several things may contribute to recurrence:

• weight regain
• declining beta-cell function
• illness
• medications that raise glucose
• aging
• return of insulin resistance

So remission means:

The diabetes is currently below its diagnostic threshold without glucose-lowering treatment.

It does not mean:

The person is biologically incapable of developing diabetes again.

This becomes obvious when we look at long-term studies

One of the most important remission trials is the:

DiRECT trial.

The original study included adults with type 2 diabetes diagnosed within the previous six years.

Participants were not using insulin.

Researchers tested an intensive primary-care weight-management program.

At one year:

46%

of participants in the intervention group were in remission.

Only:

4%

of the usual-care group achieved remission.

That's an impressive result.

But follow-up tells us why the word remission is so appropriate.

At two years, remission fell from 46% to 36%

At the two-year follow-up:

36%

of the intervention group remained in remission.

The probability of remission was strongly related to maintaining weight loss.

Among participants who maintained at least:

10 kg — about 22 pounds — of weight loss

at two years:

64% were in remission.

So achieving remission was one challenge.

Maintaining it was another.

The five-year results make the point even clearer

Participants were subsequently followed for five years.

Among the group that continued receiving extension support, approximately:

13%

were in remission at year five.

Among people who had been in remission at year two and were followed to year five:

26% remained in remission.

Those who remained in remission tended to have maintained greater weight loss.

This doesn't mean the intervention "stopped working."

It reveals something fundamental about type 2 diabetes:

remission can be durable — but relapse is common.

That's exactly why clinicians prefer the word remission over cure.

Weight loss strongly predicts remission — but not everybody responds the same way

The original DiRECT results showed a striking relationship.

Among all participants:

Lost less than 5 kg

About:

7% achieved remission.

Lost 5–10 kg

About:

34% achieved remission.

Lost 10–15 kg

About:

57% achieved remission.

Lost at least 15 kg

Approximately:

86% achieved remission.

That's one of the clearest demonstrations that substantial weight loss can fundamentally alter glucose regulation in some people with relatively recent type 2 diabetes.

But notice the word:

some.

Even among people who lost substantial weight, remission was not guaranteed.

Why doesn't everyone respond the same way?

Type 2 diabetes isn't simply:

too much body weight = high glucose.

Its biology includes:

• insulin resistance
• excess liver fat
• excess pancreatic fat
• beta-cell dysfunction
• genetics
• duration of diabetes
• individual fat-storage capacity

Two people may lose the same amount of weight and have very different glucose outcomes.

One major factor appears to be:

how much recoverable beta-cell function remains.

The pancreas has to produce enough insulin to control glucose once insulin resistance improves.

If type 2 diabetes has been present for a long time and beta-cell function has declined substantially, remission may be more difficult.

That's one reason remission studies often find better results in people with:

shorter diabetes duration.

Remission isn't only achieved through diet

There isn't one official "remission diet."

Different interventions can produce remission.

These include:

Substantial lifestyle-driven weight loss

The DiRECT approach is one example.

Metabolic surgery

Procedures such as gastric bypass and sleeve gastrectomy can produce substantial weight loss and frequently lead to major improvements in glucose control.

Current ADA guidelines state that metabolic surgery can result in diabetes remission for some people with type 2 diabetes.

Medical weight management

Modern obesity medications can produce much larger weight losses than older medications.

This raises a newer terminology problem because several of these therapies also directly lower glucose.

Someone may have:

A1C 5.5%

and major weight loss while using a GLP-1–based therapy.

Metabolically, that may represent an enormous improvement.

But under the strict consensus definition, ongoing use of a glucose-lowering drug prevents formal classification of drug-independent remission.

Again:

the health outcome matters more than the label.

Remission doesn't mean you can stop paying attention

This is probably the biggest misconception.

Someone gets an A1C of:

5.8%

without medication.

They celebrate.

Then think:

"Great. I don't have diabetes anymore. I never need another test."

That's not what medical organizations recommend.

The remission consensus says glucose should continue to be checked:

at least annually.

Why?

Because hyperglycemia can return.

And because someone who previously had diabetes may still have accumulated cardiovascular or microvascular risk from the years before remission.

Get the research, without the noise.

Practical, sourced writing on blood sugar and metabolic health. One email a week.

Email addressSubscribe

One email a week. Unsubscribe any time. See our privacy policy.

Diabetes complication screening still matters

The consensus specifically recommends continuing the routine monitoring that would normally be appropriate for diabetes complications.

That can include assessment of:

• eyes
• kidneys
• feet
• blood pressure
• cardiovascular risk

depending on the person's clinical circumstances.

There is an important biological reason for this caution.

Past exposure to high glucose may have long-term effects even after glucose improves.

Researchers sometimes refer to this phenomenon as:

metabolic memory

or:

the legacy effect.

Remission greatly improves the metabolic situation.

It does not erase someone's medical history.

But remission may still have meaningful long-term benefits

This is the other side of the story.

Remission isn't merely a nicer laboratory label.

Longer-term observational analyses suggest people who achieve remission may have lower risks of important diabetes outcomes.

The five-year DiRECT follow-up found fewer serious adverse events in the original intervention group than in controls.

Other analyses of diabetes-remission cohorts have reported associations between remission and lower rates of:

• cardiovascular disease
• chronic kidney disease
• diabetes-related complications

But long-term remission research is still developing.

The 2021 international consensus explicitly noted that we need better evidence about exactly how remission changes long-term risks.

So we shouldn't claim:

"Once you reach remission, your complication risk becomes the same as someone who never had diabetes."

That hasn't been established.

What happens if A1C rises above 6.5% again?

Suppose someone maintains:

A1C 5.8%

without medication for two years.

Then gradually regains weight.

Their next A1C is:

6.7%.

They no longer meet the remission definition.

Their type 2 diabetes has:

relapsed.

That doesn't erase the previous period of remission.

Nor does it mean the effort was wasted.

They spent two years with substantially lower glucose exposure.

The practical response is the same as with any other change in diabetes:

figure out what changed and determine the appropriate treatment.

Remission should not become another perfection standard

There is also a psychological trap hidden inside this conversation.

Imagine someone starts at:

A1C 10.2%.

Six months later:

6.6%.

They have lost 25 pounds.

Stopped insulin under medical supervision.

Need less medication.

Feel substantially better.

But technically:

they aren't in remission.

Calling that a failure would be absurd.

Likewise, someone whose A1C reaches:

5.9% while taking medication

may have achieved extraordinarily good diabetes control even though the formal remission definition isn't met.

Remission is useful because it gives researchers and clinicians a consistent definition.

It should not turn diabetes care into:

remission or failure.

There is an enormous amount of valuable territory between uncontrolled diabetes and formal remission.

A simple way to think about it

Here's the distinction.

Controlled

Your glucose is being kept below a desired threshold, potentially with medication.

Remission

Your A1C remains below the diabetes diagnostic threshold for at least three months without glucose-lowering medication.

Cure

Would imply that diabetes is permanently gone and will not return.

Current expert consensus says:

don't use "cure."

Because we cannot reliably make that promise.

The Health Facts takeaway

Type 2 diabetes remission is real.

But the medical definition is more specific — and more modest — than many internet claims suggest.

Under the international consensus definition:

A1C <6.5%

for:

at least 3 months

without:

glucose-lowering medication

can qualify as remission.

That means remission does not necessarily require an A1C below 5.7%.

It does not mean insulin resistance has vanished forever.

And it does not mean glucose can never rise again.

The larger picture looks like this:

type 2 diabetes

↓

metabolic conditions improve

↓

glucose falls below the diabetes threshold

↓

glucose-lowering medication is no longer needed

↓

A1C remains <6.5%

↓

remission

And after that:

maintenance still matters.

The important lesson isn't that diabetes can be magically erased.

It's that type 2 diabetes can be far more modifiable than many people once assumed.

For some people — particularly when meaningful metabolic changes occur relatively early in the disease — glucose regulation can improve enough that diabetes medication is no longer required to remain below the diagnostic threshold.

That's a meaningful biological outcome.

But it isn't a lifetime guarantee.

So perhaps the best way to think about remission is:

The disease is quiet enough that it no longer meets its diagnostic threshold without glucose-lowering treatment — but the conditions that allowed that to happen still need to be protected.

That's why the word is:

remission.

Not cure.