An intensive six-month program put type 2 diabetes into remission in 54% of younger adults, compared with 4% receiving usual care. The program combined a medically supervised very-low-calorie diet with aerobic and resistance exercise; it was not a diet-only test or a do-it-yourself plan.
What changed
The RESET for Remission trial randomly assigned 96 adults ages 18–45 with obesity and early-onset type 2 diabetes who were not using insulin. For 12 weeks, 50 participants followed an 800–900-calorie daily program with twice-weekly supervised exercise plus one independent session; they then moved to weight maintenance and independent exercise through week 24. The other 46 received usual care. Remission—HbA1c below 6.5% without glucose-lowering medicine for at least 12 weeks—occurred in 27 of 50 intervention participants versus 2 of 46 controls, a 50-percentage-point difference. The paper was published October 1, 2026; McGill reported the findings October 26.
How strong is the evidence?
This was a randomized, open-label human trial with blinded endpoint assessment. Ninety percent completed final assessments. No serious adverse events occurred, although temporary constipation, headache, dizziness, or fatigue were common early in the intensive program. Public disclosures included some author-industry relationships and donated or discounted study materials.
What remains uncertain
The trial was small, lasted only six months, and enrolled adults under 46. It cannot show whether remission lasts, whether results apply to older adults, or how much benefit came from exercise versus diet because there was no diet-only group. Remission is not a permanent cure.
Why it matters
The study shows that substantial remission is possible while preserving muscle in selected younger adults—but only under structured clinical supervision with medication review. Very-low-calorie diets can be unsafe when combined with diabetes or blood-pressure medicines without professional monitoring.
Related: Continuous glucose monitoring lowered HbA1c without insulin in a 283-person trial.
Sources
Original open-access trial in The Lancet Regional Health – Americas; McGill University Health Centre report.
Illustrative stock photograph by Sweet Life on Unsplash; it does not depict trial participants, meal replacements, supervised exercise, or results.
