Machine-learning MYTH diet linked to lower aging-related mortality in UK Biobank
In UK Biobank data, a 10-component dietary score built with machine learning was associated with lower aging-related mortality in internal and external validation cohorts; the abstract reports associations, not proven causes. [1]
Automatic summary · not editorially reviewed
The key findings
In 191,689 UK Biobank participants, 18 food groups were significantly associated with aging-related mortality, and a machine-learning model ranked their importance. [1]
Higher 10-component MYTH Diet scores were associated with lower aging-related mortality in internal and external validation cohorts. [1]
Higher MYTH scores were linked to slower biological aging in lungs, liver and pancreas and to lower risks for 15 aging-related diseases. [1]
The source at a glance
- Study type / format
- Observational cohort analysis with machine-learning dietary pattern development and validation (abstract).
- Population / context
- 191,689 UK Biobank participants, plus internal and external validation cohorts.
- Summary based on
- Study abstract
- Original publication
- August 7, 2026
MYTH score built from 18 food groups
The authors describe diet as a modifiable determinant of aging. Using data from 191,689 UK Biobank participants, they ran a food-wide association analysis and found 18 food groups significantly associated with aging-related mortality. A Light Gradient Boosting Machine model ranked food importance, and this led to a 10-component MYTH Diet score ranging from 0 to 10. The abstract does not list the 18 food groups or the 10 components.
Supporting source [1]
Lower mortality in two validation cohorts
Higher MYTH scores were consistently associated with reduced aging-related mortality in an internal validation cohort (highest versus lowest quartile: hazard ratio 0.79; 95% confidence interval 0.75–0.84) and in an external validation cohort (highest versus lowest quartile: hazard ratio 0.68; 95% confidence interval 0.58–0.80). A hazard ratio below 1 indicates lower hazard in the higher-score group in these comparisons. The abstract does not report follow-up duration or how the external cohort was defined.
Supporting source [1]
Multi-omics links and slower organ aging
Multi-omics analyses suggested that the diet's protective effects were partly mediated through proteomic, metabolic and inflammatory pathways. Named mediators include TNFRSF4, the proportion of polyunsaturated fatty acids (PUFA%) and lipid-related metabolites such as medium very-low-density lipoprotein phospholipids (M-VLDL-PL). Higher MYTH scores were also linked to slower biological aging in the lungs, liver and pancreas and to lower risks for 15 aging-related diseases. The authors propose the diet as a biologically informed, scalable framework for personalized nutrition; this is a proposal, not a validated product.
Supporting source [1]
How to interpret this
Abstract only; the full text was not reviewed. The abstract reports associations from observational UK Biobank data, not causal effects. It does not list the 18 food groups or 10 score components, describe follow-up, or define the external cohort. The multi-omics mediation and organ-aging results are reported without detail. The proposed framework is not a validated product.
Based on the study abstract; the full paper, tables and supplementary data were not reviewed.
Source & supporting passages
Miao Y, Li Z, Zhang X, Liu Z, Ma Y. · europepmc.org · August 7, 2026