Diabetes

What new research says about type 2 diabetes

By Lena Filatova · 26 June 2026 · 8 min read

🇺🇦 Читати українською

The science of type 2 diabetes is moving faster than the headlines suggest — and some of the most important findings are ones you're unlikely to hear in a ten-minute GP appointment. Here's what recent research actually says, and what's genuinely worth trying.

The story told about type 2 diabetes for the past few decades has been fairly consistent: it develops gradually, usually from lifestyle factors, it is progressive, and once you have it, you manage it forever. The conversation tends to centre on what you did to get there, and what medication you will need over time.

That story is being substantially rewritten.

Researchers are now questioning whether type 2 diabetes is even a single condition. Remission data from rigorous clinical trials is challenging the idea that it is permanent. And a growing body of evidence is pointing to tools — some surprisingly simple — that make a meaningful difference to blood glucose, independently of medication.

None of this is a miracle narrative. But it is worth knowing.

Remission is real — and it can last

The DiRECT trial ↗, which ran in GP practices across the UK and Scotland, set out to test whether type 2 diabetes could go into remission through a structured dietary intervention. The results, now at five years, are striking.

Almost half of participants — 46% — achieved remission within the first year. By year two, that figure was 36%. And critically, 26% of those who had reached remission at two years were still in remission at five years, with no diabetes medication.

There was also a halving of serious medical events requiring hospitalisation in the diet intervention group compared to those receiving standard care.

The intervention itself was demanding: an 830-calorie formula diet for twelve weeks to produce significant weight loss, followed by supported reintroduction of ordinary food and ongoing help with weight maintenance. For those who lost more than 10 kg, three-quarters achieved remission at both one and two years.

This does not mean remission is achievable or appropriate for everyone. But it firmly dismantles the assumption that type 2 diabetes is irreversible — and that assumption has, for a long time, quietly removed hope from the conversation.

It may not be one disease

One of the more significant shifts happening in diabetes research right now is the move away from treating type 2 diabetes as a single condition.

A 2025 paper in The Lancet Diabetes & Endocrinology ↗ put it plainly: for decades, type 2 diabetes has been managed as a monolithic condition treated via uniform, stepwise algorithms — an approach that increasingly fails to account for its biological heterogeneity.

Four broad subtypes have now been identified: mild age-related diabetes (more common in older adults, with relatively low complication risk); mild obesity-related diabetes (linked to excess weight, with a generally milder metabolic profile); severe insulin-resistant diabetes (a higher risk of kidney disease); and severe insulin-deficient diabetes (closer in some ways to type 1, with a higher risk of eye disease).

Why does this matter practically? Because it means what works well for one person with a type 2 diagnosis may do very little for another — and it makes the habit of attributing the condition to personal choices even less defensible than it already was.

The direction of travel in research is towards precision medicine: matching treatment and lifestyle strategies to individual biology, rather than everyone following the same protocol.

When you eat may matter as much as what you eat

Two findings here are worth knowing about.

Meal order. Research from Weill Cornell Medical College and others has consistently found that eating vegetables and protein before carbohydrates significantly reduces the blood sugar spike after a meal. In people with type 2 diabetes, starting with vegetables and protein before carbohydrates reduced post-meal blood sugar peaks by around 54% compared with eating carbohydrates first. One researcher described the magnitude of effect as comparable to what you might expect from a blood glucose medication.

A longer-term retrospective study following patients for five years ↗ found that those using a food-order approach — vegetables and protein first, carbohydrates last — saw their HbA1c improve from 8.5% to 7.6%, while the control group saw no meaningful change.

This costs nothing, requires no prescription, and can be applied at any meal.

Time-restricted eating. A 2025 meta-analysis of eight randomised controlled trials ↗ found that time-restricted eating — eating within a defined window each day, typically eight to ten hours — significantly reduced fasting glucose and HbA1c, and increased time in range for blood glucose. The improvements were modest but consistent across studies. A separate 2024 randomised controlled trial found HbA1c improved comparably in participants following time-restricted eating and those receiving individualised dietetic guidance.

Neither approach is a replacement for medication, and anyone adjusting their eating pattern significantly should speak with their healthcare team first. But both have meaningful evidence behind them and are worth discussing.

Sleep and blood sugar — especially for women

This is an area that receives far less attention than it deserves.

Research from the National Heart, Lung, and Blood Institute ↗ found that chronic insufficient sleep significantly increases insulin resistance in otherwise healthy women — with more pronounced effects in postmenopausal women. The effect was independent of body weight, meaning poor sleep acts on blood sugar through its own pathways, not simply by making people less active or more likely to eat poorly.

Insulin resistance is central to type 2 diabetes risk. And yet sleep quality is rarely addressed in a standard diabetes review.

For women in their forties and fifties — where perimenopause is already disrupting sleep, and oestrogen is affecting insulin sensitivity — this compounds quickly. It is one more reason to treat sleep as a non-negotiable pillar of metabolic health, not something to sort out later.

What the research says

UK resources worth knowing about

What to take from this

Type 2 diabetes research in 2024 and 2025 points in a consistent direction: the condition is more varied, more responsive to change, and less fixed than we were led to believe.

Remission is not guaranteed — but it is possible for some people, and that possibility is grounded in rigorous trial data, not wishful thinking. The biology is more complex than a single narrative about lifestyle, and the tools available — including some very simple ones like meal order — are broader than most people realise.

If you have a type 2 diagnosis, or are managing elevated blood sugar, the research increasingly supports bringing sleep, meal timing and eating sequence into the conversation alongside medication and diet — not as replacements, but as pieces of the same picture.

Every decision about your management should be made with your healthcare team. The goal here is to make sure you walk into those appointments knowing what to ask about.

The information on this website is educational and is not medical advice. Please consult your doctor if you have any doubts or further questions.