Eight Nutrition and Dietetics Developments from September 2026

Note: This article is for scientific information and is not personal diagnosis, treatment, or nutrition advice. Relative risk is not the same as absolute risk. Each finding is interpreted within its study design and uncertainties.

1. Ultra-processed foods: what does a meta-analysis of 8.8 million people show?

A dose-response meta-analysis published in Family Medicine and Community Health on 22 September pooled 8,819,894 adults from 51 prospective cohorts. Comparing the highest with the lowest consumption categories, ultra-processed food intake was associated with higher risks of cardiovascular events, cancer, overweight or obesity, metabolic syndrome or diabetes, depression or anxiety, digestive disease, and all-cause mortality. Each additional 100 g per day was associated with 14% higher relative risk of cardiovascular events, 4% higher cancer risk, and 3% higher all-cause mortality.

A synthesis of this size reveals a consistent pattern across diseases, but 8.8 million participants do not remove the problem of causality. All included studies were observational; intake was commonly measured with error-prone food-frequency questionnaires or 24-hour recalls, and application of NOVA could vary across studies. The authors rated the certainty of evidence as low or moderate. The result does not mean that everyone who consumes an extra 100 g experiences the same risk increase; it is large-scale association evidence supporting the replacement of ultra-processed-food-heavy patterns with higher-quality alternatives.

FigureThis month’s headlines do not generate the same kind of evidence

Synthesis

Prospective cohort meta-analysis

It reveals large, consistent patterns; residual confounding and measurement error may remain.

Intervention

Randomized sub-study

It compares assigned groups; small samples and attrition can reduce certainty.

Institution

Risk assessment and field report

It supports regulatory and public-health decisions; it is not a clinical efficacy trial.

Study size matters, but causal inference, generalizability, and translation into practice depend on design.

2. Weekend shifts in meal timing may relate to cardiovascular health

A NutriNet-Santé study published in Communications Medicine on 1 September examined weekday–weekend differences in meal timing—‘eating jetlag’—among 104,806 adults. Meal-timing exposure was derived from repeated 24-hour records collected during the first two years, with follow-up from 2009 to 2023. In men, each additional hour of weekday–weekend discrepancy was associated with approximately 14% higher cardiovascular disease risk; the same association was not statistically demonstrated in women.

The finding should not be translated as ‘delaying weekend breakfast by one hour causes heart disease.’ Meal timing changes alongside sleep, shift work, chronotype, physical activity, alcohol use, and social schedules. Although the analysis adjusted for many factors, unmeasured confounding may remain. The study nevertheless shows that chrononutrition is not only about clock time; the stability of eating rhythms across the week may also matter.

3. A new index for diet and biological ageing: EDISEA

A Nature Communications study published on 25 September developed EDISEA—the Empirical Dietary Index for Slower Epigenetic Aging—to capture dietary features associated with slower biological ageing. The food-based score was derived to predict acceleration in GrimAge2, a DNA-methylation measure. Higher EDISEA scores were associated with lower epigenetic age acceleration in an independent sample and with lower mortality and selected age-related outcomes across other cohorts.

Data-driven development followed by testing in independent samples is an important strength. EDISEA is not yet a clinical tool that reveals someone’s ‘true age,’ nor does it prove that a particular diet makes people younger. Epigenetic clocks are indirect markers of biological ageing, while dietary records and health outcomes remain observational. Its most useful role is not to identify isolated ‘anti-ageing’ foods, but to study how long-term dietary patterns intersect with ageing biology.

4. EFSA: salt of aspartame–acesulfame is safe at current exposure levels

On 10 September, EFSA announced the re-evaluation of salt of aspartame–acesulfame, coded E962. After ingestion, the sweetener dissociates into aspartame and acesulfame. The panel found no safety concern for E950, E951, or E962 at currently reported uses and use levels. The assessment relied on the reconfirmed acceptable daily intake of 40 mg/kg/day for aspartame and the recently revised 15 mg/kg/day for acesulfame K; estimated exposures in all age groups remained below these values.

Here, ‘safe’ is a conditional risk-assessment term: it applies to current uses and exposure levels. The conclusion does not mean sweeteners guarantee weight loss, are independent of overall diet quality, or can be consumed without limit. EFSA also recommended updated production specifications for E962 and a review of arsenic and lead limits in aspartame specifications. The practical message is simple: toxicological safety is assessed by dose, while a product’s place in the diet must be considered in the context of the total dietary pattern.

5. Why do diet–microbiome relationships differ across countries?

A Nature Communications study published on 21 September compared gut microbiome and dietary data from 1,976 participants across five countries using harmonized methods. Microbiome composition and variation within microbial genomes were strongly related to geography. Sparse canonical correlation and machine-learning analyses suggested that although some diet–microbiome links were shared, many were population-specific and generalized poorly across countries.

The finding is an important warning against universal prescriptions such as ‘eat this food to increase that bacterium.’ The same food may show different relationships depending on local dietary patterns, preparation, lifestyle, medication use, environment, and the existing microbiome. The study does not make personalized nutrition impossible; it shows that models need independent, geographically diverse validation before they are exported beyond the populations in which they were developed.

6. Metabolic effects of sweeteners: the SWEET sub-study

A SWEET sub-study published in Nutrients on 2 September evaluated the acute and longer-term metabolic effects of products containing sweeteners and sweetness enhancers instead of sugar. Twenty-six adults with overweight or obesity from the one-year randomized SWEET trial entered the baseline assessment, but only 16 completed the final test day. Outcomes included fat and carbohydrate oxidation, energy expenditure, glucose, and insulin responses.

No significant between-group differences were detected in these metabolic outcomes. The controlled design is useful because it reduces some user differences that complicate observational studies; however, the small sample and substantial attrition limit interpretation, especially of a ‘no difference’ result. Failure to detect significance does not prove that the approaches are equivalent in every setting or that all sweeteners have no long-term effects. This is one small piece of evidence, not a complete answer to safety or health-effect questions.

7. Conflicts of interest in nutrition policy: balancing expertise and independence

A BMJ investigative report dated 9 September stated that six of 19 members of the UK Scientific Advisory Committee on Nutrition had food-industry links classified as ‘significant’ under the committee’s own rules. This was not a new clinical study; it was a governance investigation into membership, disclosure, and conflict-of-interest management in a body that informs public nutrition guidance.

Past industry collaboration does not by itself invalidate scientific judgment or prove that a decision was influenced. In nutrition, knowledge of products, production, and implementation can add valuable expertise. But the nature, recency, and role of such relationships should be disclosed clearly, with recusal, non-voting, and balanced independent membership used where appropriate. Trust depends not only on reaching a sound conclusion, but also on making the route to that conclusion visible.

8. Child malnutrition: environmental exposures and the humanitarian crisis in Haiti

A UNICEF brief published in September 2026 emphasized that child undernutrition and micronutrient deficiencies cannot be explained by food quantity alone. Nutrition during the first 1,000 days shapes survival, growth, development, and even intergenerational outcomes. Water, sanitation, and hygiene have long been linked to about 10% of undernutrition cases, while the nutritional effects of heat, climate, chemical pollutants, toxicants, and biological hazards remain under-recognized.

In the same month, UNICEF, FAO, and WFP reported that 253,725 children aged 6–59 months in Haiti were expected to experience acute malnutrition between June 2026 and May 2027. This included 93,873 projected severe and 159,852 moderate cases, while 25,750 pregnant and breastfeeding women were expected to need nutrition support. These figures are a 12-month estimated caseload, not the number of children simultaneously affected on the announcement date. Amid violence, displacement, and disrupted access to food and care, an adequate response requires more than therapeutic products: safe water, healthcare, livelihoods, local food production, and early screening must be strengthened together.

Take-home message

September’s shared lesson: read the design before the number, and the context before the headline

These eight developments show the different layers of nutrition science. Meta-analyses involving millions reveal patterns; small randomized studies test defined mechanisms; risk assessments establish acceptable exposure; and humanitarian reports reveal the social conditions of nutrition. None is sufficient alone. Scientifically honest interpretation begins by not presenting relative risk as an absolute outcome, separating association from causation, explaining that ‘safe’ depends on dose and context, and refusing to reduce nutrition problems to individual willpower.

Eight Nutrition and Dietetics Developments That Mattered in September 2026