Ten Healthy-Diet Scores and Epigenetic Aging: What a 6,470-Person Study Found

In 6,470 adults in Bonn, Germany, people who scored higher on healthy-eating patterns tended to have slower epigenetic aging readings in their blood. No single pattern was the clear winner, and only 18 people scored in the top quarter on all ten scores.^[1] The data are a snapshot, so they show a link, not a cause.

What the researchers did

The Rhineland Study is an ongoing study of adults aged 30 and older living in Bonn. The team used its baseline, one-time data from 6,470 participants (57% women, average age 56). Diet came from a food frequency questionnaire, and each person was scored on ten diet-quality indices: Mediterranean, DASH, MIND, AHEI-2010, Nordic, EAT-Lancet, a plant-based index, a healthful plant-based index, an unhealthful plant-based index and a dietary inflammatory index.^[1]

Epigenetic aging was read from DNA methylation in blood with three clocks: GrimAge and PhenoAge (both measure accumulated biological age) and DunedinPACE (the pace of aging). The team then repeated the analysis in 1,034 people from a second German cohort, EPIC-Potsdam.^[1]

What they found

• Overlap was small. Only 18 people were in the top 25% on all ten scores, and 77 on at least nine.^[1]
• Higher scores on the healthy patterns tended to go with lower epigenetic age readings. Scores for the two less healthy patterns (unhealthful plant-based and dietary inflammatory) tended to go with higher readings.^[1]
• All ten scores were associated with DunedinPACE after correction for multiple testing. DASH and Nordic were the two significantly associated with all three clocks.^[1]
• For a one standard deviation (SD) higher DASH score (4.4 points), DunedinPACE was 0.10 SD lower (95% confidence interval -0.12 to -0.08) and GrimAge acceleration 0.03 SD lower (-0.06 to -0.01). For Nordic (3.4 points), DunedinPACE was 0.07 SD lower (-0.09 to -0.05).^[1]
• The EAT-Lancet score showed no association with GrimAge or PhenoAge, but did with DunedinPACE.^[1]
• In the second cohort, associations were consistent in direction and magnitude for GrimAge and PhenoAge. The patterns left different methylation fingerprints, but those pointed to largely the same biological pathways.^[1]

What this study does not show

• It does not show that eating well slows aging. The analysis used one-time, cross-sectional data, so it cannot tell which came first or rule out other explanations. The authors note residual confounding cannot be ruled out.^[1]
• The effects are small. A 0.03 SD difference is tiny, and 0.10 SD is modest. This study did not translate them into years of life or disease risk.^[1]
• Diet came from a questionnaire, which the authors say is subject to measurement error.^[1]
• Both cohorts are mostly German adults of European ancestry, so the results may not carry over to other groups or diets.^[1]
• Epigenetic clocks are estimates of biological aging. Nobody in this study was followed to see who lived longer or stayed healthier.^[1]

What to do with it

The most useful read is a practical one: several different patterns lined up in the same direction, so there is little reason to argue over which brand of healthy eating is best.

1. Pick a pattern you will actually keep. The scores in this study measure how closely people followed a pattern, so sticking with it is what was compared. The study did not test which pattern is easiest to stick with, so choose by the food you will really eat.
2. If you already eat close to one of these, this paper is no reason to switch.
3. Check before big changes if you have a health condition. If you take medication or have a condition like kidney disease, talk to your clinician or a registered dietitian before changing how you eat.

More on this topic: foods for longevity and what the strongest research backs and healthy aging and what the research says matters most.

References

[1] Tavares JF, Liu D, Talevi V, Eichelmann F, Jannasch F, Schulze MB, Aziz NA, Nöthlings U, Breteler MMB. Associations between diet quality, epigenetic aging and epigenome in two population-based cohorts. Nature Communications, 2026;17(1):9232. Population: 6,470 adults aged 30-95 in the Rhineland Study (cross-sectional baseline data), replicated in 1,034 adults from EPIC-Potsdam. DOI: https://doi.org/10.1038/s41467-026-77064-4. Finding as used: minimal overlap in top-quartile adherence across ten diet quality scores (18 people on all ten, 77 on at least nine); healthy dietary patterns associated with lower GrimAge, PhenoAge and DunedinPACE acceleration, all ten scores associated with DunedinPACE after FDR correction; per 1 SD higher DASH score, DunedinPACE -0.10 SD (95% CI -0.12 to -0.08) and GrimAA -0.03 SD (-0.06 to -0.01); per 1 SD higher Nordic score, DunedinPACE -0.07 SD (-0.09 to -0.05); no EAT-Lancet association with GrimAA or PhenoAA; replication consistent for GrimAA and PhenoAA; authors note FFQ measurement error, residual confounding and limited generalizability beyond German adults of European ancestry. ↑︎

Kurt Greiner

Kurt is a digital strategist and IT professional blending emerging technology with practical application to help businesses and individuals streamline their digital presence. His current work focuses on the intersection of intentional living and technological resilience, exploring how individuals can leverage modern tools to navigate the second half of life with purpose.