Can type 2 diabetes go into remission? What the DiRECT trial showed

Type 2 diabetes can go into remission — blood sugar back in the non-diabetic range without glucose-lowering medication — and the most reliable route to it is losing a large amount of weight and keeping it off. In one primary-care trial, 68 of 149 people (46%) offered an intensive weight-management programme were in remission a year later, against six of 149 (4%) given usual care [s2].

That is a real effect, and it changed how diabetes is talked about. But remission is not a cure, it is not available to everyone, and it decays: in the same trial the share in remission fell as the months passed and the weight crept back [s3].

What "remission" actually means

Until recently there was no agreed definition, which made the word easy to oversell. In 2021 an international expert group convened by the American Diabetes Association settled on one: remission is an HbA1c below 6.5% (48 mmol/mol) measured at least three months after stopping glucose-lowering drugs [s1]. The three-month gap matters — it is there to confirm the improvement is the body's own, not a lingering drug effect.

The group deliberately chose "remission" over "reversal" or "cure" [s1]. The distinction is not pedantic. Remission describes a state that is holding for now; it says nothing about whether it will last, and it does not mean the underlying tendency to high blood sugar has gone. The consensus notes that a return of normal glucose can happen spontaneously or after treatment, and may now be occurring more often as newer therapies drive larger weight loss [s1].

The trial that put a number on it

The clearest evidence comes from DiRECT, an open-label trial that randomised 49 general practices in Scotland and Tyneside to deliver either an intensive weight-management programme or best-practice care, with 149 patients per group analysed [s2]. Participants were aged 20 to 65, had been diagnosed with type 2 diabetes within the previous six years, had a body-mass index of 27–45, and were not on insulin [s2]. The programme withdrew antidiabetic and antihypertensive drugs at the start, replaced all food with an 825–853 kcal/day formula diet for three to five months, then reintroduced food and supported people to keep weight off [s2].

At 12 months, 36 intervention participants (24%) had lost at least 15 kg, against none in the control group; 68 (46%) were in remission, against six (4%) with usual care — an odds ratio of 19.7 (95% CI 7.8–49.8) [s2]. Mean weight fell by 10.0 kg in the intervention group [s2].

The single most important finding is that remission tracked weight loss almost mechanically. Across the whole trial population, none of the 76 people who gained weight reached remission; nor did more than six of the 89 (7%) who lost up to 5 kg [s2]. It rose to 34% of those who lost 5–10 kg, 57% of those who lost 10–15 kg, and 86% of the 36 people who lost 15 kg or more [s2]. Remission was, in effect, a dividend of sustained weight loss rather than a separate achievement.

What happened when the follow-up ran longer

The two-year results are the reason to be sober about it. At 24 months, remission had fallen to 53 intervention participants (36%) against five controls (3%), and the share who had kept off at least 15 kg dropped to 17 (11%) [s3]. Among everyone who managed to maintain a loss of at least 10 kg — 45 of 272 people with data — 29 (64%) were in remission; the problem is how few sustained that much [s3]. Weight regain, not a failure of the biology, is what pulls people back out.

There was better news on safety: serious adverse events were fewer in the intervention group than in the control group during the second year, nine against 22 [s3]. And it is worth noting who was studied — people diagnosed recently, still making their own insulin, and not yet on injections. The trial does not show that someone with 15 years of diabetes on multiple agents can expect the same, and it was not designed to.

How to read this

Three things follow. First, remission is genuine and can remove the need for medication in a meaningful minority of recently diagnosed people, but it is bought almost entirely with large, maintained weight loss — the harder half of the equation. Second, because it can fade, someone in remission still needs their blood sugar watched; the consensus explicitly calls for continued monitoring rather than discharge [s1]. Third, the closer to diagnosis and the more weight lost, the better the odds — which is an argument for acting early, not a promise.

None of this is a prescription. Whether an intensive diet, weight-loss medication such as a GLP-1 drug, or another approach is appropriate — and whether it is safe to withdraw any medication — depends on the person and belongs with the clinician managing their care. The weight-loss link runs through every route, and the dose–response between weight lost and metabolic benefit is the through-line of the whole field. What the evidence supports is a straightforward claim: type 2 diabetes is not always a one-way street, and how far it can be turned around depends heavily on how much weight is lost and held.