Nutrition research, explained and sourced

Population

31 randomised trials, 23,684 adults with prediabetes

What the study found

A 2025 systematic review and meta-analysis of 31 randomised trials involving 23,684 adults with prediabetes found that lifestyle interventions reduced type 2 diabetes incidence overall and increased reversion to normoglycemia. Effects differed across prediabetes phenotypes, which is an important reminder that prediabetes is not a single uniform state.

Why it matters

Prediabetes is common, and this review gives a high-level picture of how lifestyle programmes performed in trials. It also shows that results were not identical across groups, so it is background evidence rather than a forecast for any one person.

Keep in mind

The review found limited evidence of benefit in some phenotype subgroups. This does not mean lifestyle support is ineffective for every individual in those groups, and it does not establish outcomes for Foodi-AI.

Population

27 studies of lifestyle interventions in people with different prediabetes phenotypes

What the study found

A systematic review of 27 studies found that responses to lifestyle intervention differed between prediabetes phenotypes. In the included evidence, reductions in diabetes incidence were clearer for impaired glucose tolerance, with or without impaired fasting glucose, than for isolated impaired fasting glucose.

Why it matters

It helps explain why prediabetes is described using different glucose patterns, and why those differences may matter when thinking about more personalised support.

Keep in mind

Phenotype-specific evidence is still developing. The findings support careful personalisation, not a conclusion that one group cannot benefit from lifestyle change.

Population

77 randomised studies, 22,629 adults with prediabetes

What the study found

A 2026 meta-analysis of 77 randomised studies examined which characteristics of lifestyle programmes were linked with stronger outcomes in adults with prediabetes. The findings reinforce that intervention structure and the changes participants achieve can matter, not simply whether lifestyle advice is offered.

Why it matters

It moves the question from whether lifestyle advice is offered to how programmes are designed, which is a useful lens when reading any lifestyle research.

Keep in mind

Meta-regression can identify relationships across studies but cannot prove that every programme component directly caused better outcomes. Avoid turning subgroup findings into universal prescriptions.

Population

Primary care teams in Catalonia, Spain, and their patients, including those at high risk of diabetes

What the study found

Catalonia's DP-TRANSFERS programme trained primary care teams to deliver structured lifestyle coaching. After implementation, more people with prediabetes were identified and more high-risk patients were referred into nutrition services.

Why it matters

Primary care is often where prediabetes is first identified. This project shows how care teams were prepared to spot risk and connect people with nutrition support.

Keep in mind

This was an implementation project in one region, and changes seen after a programme starts cannot on their own show that the programme caused them. Health systems differ between countries.

Population

Participants who completed a period of energy restriction (see the source for details)

What the study found

Investigators observed that ghrelin, GLP-1 and peptide YY responses after a period of energy restriction were correlated with how well weight and glycaemic improvements were sustained afterwards.

Why it matters

It points to biology as one possible reason why keeping improvements going can be harder than achieving them in the first place.

Keep in mind

A correlation does not show that these hormones cause weight regain or the loss of glycaemic improvement. This is a single study, so check the source for details about the participants.

Population

People taking part in 29 produce prescription programmes

What the study found

A systematic review of 29 produce prescription programmes reported higher fruit and vegetable intake across the included programmes, alongside reductions in HbA1c and blood pressure. Effect sizes varied widely between programmes.

Why it matters

Produce prescriptions are a real-world way of making fruit and vegetables easier to access. The review gives an overview of how such programmes have been evaluated and what was reported.

Keep in mind

Most of the included studies were not randomised, so the review cannot show that the programmes themselves caused the changes. Programmes and participants differed from one another.

Population

Adults with prediabetes

What the study found

Researchers linked high MEDLIFE scores combined with low visceral adipose tissue to lower odds of type 2 diabetes and microvascular complications among adults with prediabetes. The association was seen for the lifestyle pattern as a whole, not for diet alone.

Why it matters

It looks at a lifestyle pattern as a whole rather than a single food or nutrient, a useful reminder that eating habits sit alongside other factors.

Keep in mind

This reports an association, so it cannot show that the lifestyle pattern or lower visceral fat caused the lower odds. It describes adults with prediabetes in a single study.

Population

People with prediabetes in a cardiac rehabilitation programme

What the study found

The DIABEPIC-1 cardiac rehabilitation study added a 10-hour eating window and an ultra-processed food intake below 10% of calories to standard care. The intervention group had higher rates of normoglycaemia and larger blood pressure reductions than the control group.

Why it matters

It looks at two food-related changes together, the timing of eating and the degree of food processing, in a rehabilitation setting. It shows what was observed for the combined approach.

Keep in mind

The intervention combined several changes, so the study cannot separate the contribution of the eating window from that of lower ultra-processed intake. It is a single study in a cardiac rehabilitation setting.

Population

Participants in a crossover trial of an early 10-hour eating window (see the source for details)

What the study found

In a crossover trial, an early 10-hour eating window was associated with more time in range on continuous glucose monitoring and lower glucose variability, without weight change. This suggests that the timing of eating may matter independently of how much is eaten.

Why it matters

It explores when people eat as a question separate from how much they eat. The glucose measures come from continuous monitoring during the trial.

Keep in mind

The outcomes are glucose measures taken during the trial, not long-term health outcomes, and this is a single trial. Results may differ for other eating schedules and other groups of people.

Population

25 randomised trials (36 interventions), 851 adults

What the study found

A meta-analysis of 25 randomised trials found that exercise training improved insulin-stimulated glucose disposal in adults. Weight loss strengthened the effect, but the evidence reinforces that exercise can influence metabolic health through more than body-weight change alone.

Why it matters

Body weight is only one way to look at metabolic health. This review points to insulin action as another outcome on which exercise programmes can be assessed.

Keep in mind

The trials used a specialised clamp method to assess insulin action and included varied adult populations. The findings should not be interpreted as a guaranteed individual response.

Population

Adults: 53 studies identified, 39 included in the meta-analysis

What the study found

A 2026 systematic review found that interrupting prolonged sitting with brief activity breaks reduced post-meal glucose and insulin responses in adults. Walking breaks showed the strongest effects among the activity modes analysed, while standing alone produced smaller changes.

Why it matters

It shows how a small, low-effort change around meals was examined in controlled studies. It describes short-term responses, so it is context for the idea of moving around meals rather than evidence of long-term results.

Keep in mind

Most included studies were short laboratory experiments, often with small samples. The results describe acute post-meal responses rather than long-term diabetes outcomes.

Population

33 studies, 14,398 adults with prediabetes

What the study found

A systematic review and meta-analysis of 33 studies involving 14,398 adults with prediabetes found that digital health lifestyle interventions produced greater weight loss than comparison conditions. Improvements were also reported for several secondary outcomes, although the strength of evidence varied.

Why it matters

Digital tools are widely used for lifestyle support, so it is useful to see how they have performed in research. This review looks at digital programmes as a group, in adults with prediabetes.

Keep in mind

The interventions ranged from partly digital programmes to fully digital approaches, so the results should not be treated as evidence for any specific app or feature, including Foodi-AI.

Population

Adults with prediabetes

What the study found

In a randomised trial in adults with prediabetes, dietitian consultations paired with real-time continuous glucose monitoring were followed by greater gains in time in range and lower fasting glucose than dietary education alone.

Why it matters

It asks whether seeing glucose data in real time, alongside professional guidance, adds anything beyond standard education. It is evidence about individualised feedback delivered with dietitian support.

Keep in mind

Glucose monitoring was combined with dietitian consultations, so the trial cannot show what monitoring alone would do. The outcomes are short-term glucose measures, and the findings do not apply to apps or tools that do not use glucose monitoring.

Population

31 studies, 23,684 people with prediabetes

What the study found

A 2025 systematic review compared face-to-face, blended and digital lifestyle interventions for adults with prediabetes. Across the included trials, lifestyle intervention reduced progression to type 2 diabetes and increased reversion to normoglycemia, while effects varied by delivery approach.

Why it matters

It compares how support was delivered, which gives context for conversations about digital and in-person approaches. It suggests that the delivery format is one factor among several.

Keep in mind

The categories included different programmes, intensities and populations. Do not use the review to claim that digital delivery is always superior to face-to-face care.

Population

Community participants in Hunan, China

What the study found

A community cohort in Hunan, China found that replacing table salt with potassium-enriched blends was associated with lower urinary sodium-to-potassium ratios and lower systolic blood pressure in ordinary daily use.

Why it matters

Changing the type of salt used at home is a small, practical dietary change, and this study looked at it in ordinary daily life.

Keep in mind

This was a cohort in one region of China, so results may differ elsewhere. Potassium-enriched substitutes are not suitable for some people, notably those with kidney disease or taking certain medication, so check with a doctor first.

Population

Participants in a population imaging study, compared by prediabetes status

What the study found

A population imaging study linked prediabetes to thicker carotid intima-media layers and stiffer arteries.

Why it matters

It adds context for why glucose levels in the prediabetic range may be discussed alongside heart and blood vessel health, not only as a glucose issue.

Keep in mind

This reports an association, so it cannot show that prediabetes caused the vessel changes. Subclinical means the changes were found before symptoms or a diagnosed condition.