A new observational study spanning two human cohorts has found a link between theobromine, a phytochemical abundant in cocoa, and slower epigenetic aging [1].
A yummy geroprotector?
Popular wisdom has it that not many things in life are both healthy and delicious, but according to research, one of them is chocolate. Studies have repeatedly linked chocolate and cocoa consumption with improved health outcomes, including beneficial effects on cholesterol, inflammation, and cellular senescence [2].
Cocoa and coffee share a family of alkaloids called methylxanthines, which includes caffeine, theophylline, paraxanthine, theobromine, and 7-methylxanthine. Coffee is caffeine-heavy, while cocoa is theobromine-heavy.
Theobromine has been reported to extend lifespan in certain strains of C. elegans nematode worms [3] and has been linked to cardiovascular and other health benefits in observational human data [4]. A new study conducted by scientists from King’s College London and several German institutions, published in Aging, looks at theobromine in the context of human aging by analyzing the correlation between the chemical’s blood levels and biological age acceleration.
Two clocks confirm slower aging
The team used two metrics: GrimAge, one of the most robust methylation clocks and trained to predict mortality, and DNAmTL, a DNA methylation-based estimator of telomere length. The latter predicts telomere length in leukocytes from methylation at a specific set of CpG sites.
The researchers pulled together two population cohorts in which people had both blood metabolomics and DNA methylation data. TwinsUK, which includes 509 women with a median age of around 60, served as the discovery cohort. The results were then replicated in KORA, a cohort of 1,160 German adults of both sexes.
They started by measuring the association between the difference between the GrimAge reading and the person’s chronological age (clock acceleration) and six coffee/cocoa-related metabolites: five methylxanthines and the amino acid theanine.
Theobromine immediately stood out, showing the strongest negative correlation. In the researchers’ main model, higher theobromine levels were linked to roughly 1.6 years less GrimAge acceleration per standard step up in theobromine. Theobromine levels were also associated with higher DNAmTL readings, suggesting slower telomere attrition.
Testing the results
The researchers then extensively stress-tested these results. To see if this was just a generic “coffee drinker” signal, they added caffeine and its breakdown products to the model and asked whether the theobromine effect would disappear. It did not; the association with slower GrimAge acceleration stayed, becoming only slightly weaker.
Next, they used penalized regression models (LASSO and elastic net), which automatically shrink or drop less informative variables. Even under these harsher conditions, theobromine consistently remained one of the key predictors of GrimAge acceleration.
In TwinsUK, metabolomics and methylation measurements could be taken up to 5 years apart, so the researchers re-ran the analyses within narrower time windows to see how the association would change. The shorter the distance between the two readings (latency), the stronger the effect size became, strengthening their confidence that this was not just an artifact of long gaps between measurements.
In the KORA replication cohort, higher serum theobromine again tracked with younger epigenetic profiles. After adjustment for age, BMI, blood cell composition, technical factors, and the other methylxanthines, each standard step up in theobromine was linked to about one year of reduced GrimAge acceleration and to slightly longer telomeres.
Because the discovery cohort consisted entirely of women, the authors checked whether the pattern held in women from the KORA study and found a similar, slightly weaker association there, reinforcing their original discovery. In the full KORA cohort, which includes both men and women, the overall effect of theobromine on GrimAge was actually stronger than in women alone. This suggests that men also show a negative association between theobromine and epigenetic age that is at least as strong, if not stronger, than in women.
Dr. Ramy Saad, lead researcher at King’s College London, who is also a researcher at University College London and holds a doctorate in clinical genetics, said: “This is a very exciting finding, and the next important questions are what is behind this association and how can we explore the interactions between dietary metabolites and our epigenome further? This approach could lead us to important discoveries towards ageing, and beyond, in common and rare diseases.”
Dr. Ricardo Costeira, a Postdoctoral Research Associate from King’s College London, added: “This study identifies another molecular mechanism through which naturally occurring compounds in cocoa may support health. While more research is needed, the findings from this study highlight the value of population-level analyses in aging and genetics.”
Caveats and limitations
Being observational and cross-sectional, this study cannot prove a causal relationship between theobromine and aging or even the methylation-based aging metrics that the researchers used. While they adjusted for age, BMI, blood cell counts and several related metabolites, residual confounding by factors like overall diet, lifestyle, socioeconomic status, or other cocoa components, especially flavanols, remains a real possibility.
Both cohorts are European and middle-aged/older, and the discovery sample is composed entirely of female twins, which limits generalization to men, younger people, and other ancestries. Finally, epigenetic clocks and DNAm-based telomere estimates are useful but still imperfect proxies for biological aging. Nevertheless, it’s more encouraging news for chocolate lovers.
Literature
[1] Bell, J., Saad, R., Costeira, R., Garcia, P. M., Villicaña, S., Gieger, C., … & Waldenberger, M. (2025). Theobromine is Associated with Slower Epigenetic Ageing. Aging.
[2] Arranz, S., Valderas‐Martinez, P., Chiva‐Blanch, G., Casas, R., Urpi‐Sarda, M., Lamuela‐Raventos, R. M., & Estruch, R. (2013). Cardioprotective effects of cocoa: Clinical evidence from randomized clinical intervention trials in humans. Molecular Nutrition & Food Research, 57(6), 936-947.
[3] Li, H., Roxo, M., Cheng, X., Zhang, S., Cheng, H., & Wink, M. (2019). Pro-oxidant and lifespan extension effects of caffeine and related methylxanthines in Caenorhabditis elegans. Food Chemistry: X, 1, 100005.
[4] Sharifi‐Zahabi, E., Hajizadeh‐Sharafabad, F., Nachvak, S. M., Mirzaian, S., Darbandi, S., & Shidfar, F. (2023). A comprehensive insight into the molecular effect of theobromine on cardiovascular‐related risk factors: A systematic review of in vitro and in vivo studies. Phytotherapy Research, 37(9), 3765-3779.
;){kind=link}
;){kind=link}