A low-calorie diet and supervised exercise put type 2 diabetes into remission in 27 of 50 adults aged 18 to 45

Twenty-seven of 50 adults aged 18 to 45 with type 2 diabetes had blood sugar below the diabetic range, without any diabetes medication, after 24 weeks of a low-calorie diet combined with supervised exercise. Among 46 similar people who carried on with their usual care, two did. The trial behind those numbers was published on October 1 in The Lancet Regional Health Americas and presented at the annual meeting of the European Association for the Study of Diabetes in Milan.

It was led by Kaberi Dasgupta of McGill University in Montreal and Tom Yates of the University of Leicester, with colleagues at the University of Alberta. The people they recruited had what doctors call early-onset type 2 diabetes, diagnosed before about 45. The researchers describe it as a distinct form of the disease with rapid deterioration in blood sugar control, early onset of complications, and lower life expectancy than type 2 diabetes that begins later.

What type 2 diabetes remission means

Remission has a formal definition. In 2021 an international expert group convened by the American Diabetes Association proposed a single test for it: an HbA1c below 6.5%, measured at least 3 months after cessation of glucose-lowering pharmacotherapy, that is, after diabetes drugs have been stopped. HbA1c is the blood test on which a diabetes diagnosis usually rests, and 6.5% is the line.

The group picked the word with care. Terms such as reversal and cure had all been used for the same outcome, and after discussing cure, the group agreed that the term should be avoided in the context of type 2 diabetes. Their reasoning was that the word implies nothing more needs doing. Remission, by contrast, fits the view that a person may require ongoing support to forestall relapse, and regular monitoring to allow intervention should hyperglycemia recur. Hyperglycemia means high blood sugar.

That definition is the one the new trial used. Its main outcome was remission at 24 weeks, with participants off glucose-lowering medication for at least 12 of them.

How the diet and exercise trial worked

Very low-calorie diets, which researchers call low energy diets, are an established route to remission. According to the trial report, they have not been studied in combination with structured supervised exercise. The plan the team published in advance sets out why the pairing might matter. Low energy diets incorporating meal replacements can achieve remission through weight loss, but may reduce lean mass, the protocol says, using the technical term for the body’s non-fat tissue, much of it muscle. Training might protect against that loss, because exercise can preserve lean mass and directly lower insulin resistance, the reduced response to insulin that underlies type 2 diabetes.

The trial, called RESET for Remission, ran at three sites in Leicester, Montreal and Edmonton. It enrolled adults aged 18 to 45 with type 2 diabetes and obesity who were not on insulin therapy. Ninety-six people were randomly assigned, 50 to the combined plan and 46 to usual care, between September 2021 and June 2025. Their mean age was 38, and their average body mass index was 35.2, well into the obese range. Thirty-seven were of European background, 27 South Asian and 32 of other ethnicities.

Those in the combined group stopped their glucose-lowering and blood pressure medication at the outset and began a diet of 800 to 900 calories a day. The first two weeks comprised total meal replacement, using commercial formula products. From week 3 to week 12 the calorie level stayed the same with more variety, allowing one food-based meal per day.

Exercise began alongside the diet. Twice a week people trained under supervision, and once more on their own. The protocol describes brisk treadmill walking for the aerobic part, and resistance training with machines or free weights, including bench press, seated row and leg press. After 12 weeks the intensive phase ended, followed by a weight-maintenance diet with independent exercise until week 24.

The comparison group received usual diabetes care. Under the protocol, people in that group were to be offered a 12-week low energy diet intervention once their final measurements were complete.

The trial was open-label with a blinded endpoint. Participants and their trainers knew who was dieting, as they must in a trial like this, but the outcome was a laboratory value assessed without knowledge of group.

Remission in 54% on diet and exercise against 4% on usual care

Remission at 24 weeks occurred in 27 of 50 people, or 54%, in the combined group and 2 of 46, or 4%, in the usual-care group. The difference was 50 percentage points, and the range compatible with the data ran from 33 to 64. After adjustment, the odds of remission were about 21 times higher in the combined group. That estimate is imprecise, with an interval running from about 6 to about 115, because so few people in the comparison group reached remission.

The analysis was conservative in one respect. It counted everyone as randomized, assuming no remission when data were unavailable. Ten of the 96 participants did not attend the final assessment, and each was treated as still having diabetes.

Weight fell sharply. The combined group had a mean weight of 100.7 kg at the start, about 222 pounds. Weight loss between weeks 0 and 24 in the intervention group was 8.4 kg, or 18.5 lb, against 1.2 kg in the usual-care group. After adjustment, people on the combined plan lost 7.5 kg more than control participants by week 24, which included 6.6 more kg of fat mass lost. Roughly nine-tenths of the extra weight lost, in other words, was fat.

Preserving muscle is what the exercise was meant to achieve, though without a diet-only group the trial cannot show that the training was responsible. “Our RESET for Remission study targeted muscle during the intensive weight loss phase through resistance training, and our findings indicate fat-free mass and muscle preservation is achievable with rapid weight loss,” the authors said in a statement released in Leicester.

The plan was not comfortable. Most intervention group participants experienced at least one adverse event, 43 of 50 compared with 6 of 46 on usual care. These were mainly early effects of the diet, such as constipation, headache, dizziness and fatigue, and they resolved. Remission was achieved without serious adverse events.

How the result compares with earlier diabetes remission trials

The authors place their result high among comparable studies. “Our remission rates are at the upper end of those previously reported in the literature for trials of similar duration,” they said.

Study	Who took part	Follow-up	In remission
England and Canada trial (RESET for Remission)	96 adults aged 18 to 45	24 weeks	54% against 4% on usual care
Scotland and England trial (DiRECT)	306 adults aged 20 to 65	12 months	46% against 4%
Qatar trial (DIADEM-I)	158 adults aged 18 to 50	12 months	Over 60%
England, NHS remission service	710 adults with two blood tests on record	1 year	27%
United States trial (Look AHEAD)	Adults in a less intensive diet and exercise plan	1 year	11.5%

The benchmark is DiRECT, a diet-only trial run at 49 primary care practices in Scotland and the Tyneside region of England. Its participants replaced all meals with a formula diet of about 850 calories a day for three to five months, then reintroduced food in steps. At 12 months, the DiRECT team reported, almost half of participants achieved remission to a non-diabetic state and off antidiabetic drugs. The RESET protocol gives the figure exactly: in DiRECT, remission was 46% at 1 year, against 4% in its comparison group.

A trial closer in spirit to the new one took place in Qatar. DIADEM-I, done in primary care and community settings in Qatar, tested meal replacement with physical activity support in adults aged 18 to 50 who had had diabetes for three years or less. Its exercise component was advice and self-monitoring instead of supervised sessions. At 12 months the program was associated with diabetes remission in over 60% of participants.

Less demanding plans have done less. In the American Look AHEAD trial, which paired a diet of 1,200 to 1,800 calories with an exercise target, remission, a secondary outcome, was 11.5% at 1 year.

Outside trials the numbers drop too. England’s National Health Service launched a meal-replacement service in September 2020. Among 710 participants with two blood tests on record, 190, or 27%, were in remission at a year. The evaluators concluded that remission of type 2 diabetes is possible outside of research settings, through at-scale service delivery, but that the rate of remission achieved is lower and the ascertainment of data is more limited with implementation in the real world.

These rows are not a ranking. The studies differ in the age of participants, how long they had been diabetic, and when remission was measured, and 24 weeks is a shorter test than 12 months.

How long diabetes remission lasts after the diet ends

The harder question is durability, and the new trial stops before it can answer. DiRECT kept going. After two years its program had produced a mean weight loss of 7.6 kg, with 36% of participants in remission of type 2 diabetes. Continued low-intensity dietary support was then offered up to 5 years from baseline, and among the 85 people in that extension group at five years, 11, or 13%, were in remission. Counted another way, of those in remission at year 2, 26% remained in remission at 5 years.

The pattern follows weight. The RESET authors acknowledge that weight-loss efforts of every kind, low energy diets included, are followed by a period of either partial or full weight regain for most people. The consensus group reached a similar position on remission itself: long-term remissions are possible, but not assured. It listed durability among the open research questions, writing that the expected duration of a remission induced by various interventions is still not well defined.

Time spent in remission may still count for something. In the DiRECT follow-up, serious adverse events in the original intervention group were under half those in the control group over the five years. The DiRECT team worded its conclusion with care, saying the extended program suggested improved health status over 5 years.

The RESET team argues that muscle is where its approach could differ over the long run. “Preserving muscle during remission-oriented weight loss may have important implications for long-term metabolic health, physical function, and disease trajectory. This requires examination in a longer-term study,” the authors said.

What the trial cannot show about diet and exercise

The design compared the whole package with usual care. There was no group that dieted without training, so the trial cannot say how much of the remission came from the gym. The protocol states this plainly: RESET will not distinguish the impacts of the individual dietary and exercise components. A remission rate of 54% at 24 weeks is in the range that diet-only trials have reported, and the case for adding exercise rests on fat loss and muscle more than on remission itself.

Size is a second limit: with 96 participants the estimate of the difference between groups is imprecise. Length is a third. Follow-up ended 12 weeks after the intensive diet did, too soon to observe the weight regain that the authors themselves expect for most people.

Delivery is a fourth. Participants had an exercise physiologist, formula products and regular medical review, at three university-linked sites. The NHS experience suggests results thin out when a program leaves the research setting.

Finally, remission is a blood test result. The trial did not run long enough to count heart attacks, kidney disease or eye damage. The consensus report cautions that at present, there is no long-term evidence indicating that any of the usually recommended assessments for complications can safely be discontinued.

What changes for people with early type 2 diabetes

Nothing changes in practice on the strength of one 24-week trial, and the authors do not claim otherwise. Of their approach they wrote: “It warrants further evaluation in longer term studies in real-world settings.”

Participants in RESET came off their medication at the start under the supervision of the study’s physicians, who reviewed them during the diet. Do not stop or change a prescribed medicine without talking to your doctor. Whether an intensive diet is suitable for a particular person depends on their medication and other conditions, which is assessed by their doctor.

Just over half of the adults who dieted and trained met the blood-test definition of remission at 24 weeks, and how many will remain so at five years is a question this trial was not built to answer.