A peanut-derived plasma lipidomic signature (PeLS), constructed by combining lipid changes observed in a 6-month randomized trial (ARISTOTLE, n=63), was associated with an 11% lower hazard of incident type-2 diabetes (HR=0.89, 95%CI: 0.80–0.99) in the EPIC-Potsdam case-cohort study (774 T2D cases). The PeLS was dominated by very long-chain saturated fatty acid (VLCSFA)-containing sphingolipids and showed no significant link to cardiovascular disease. Strikingly, BMI explained an estimated 89% of the association's mediating pathway.
The study's two-stage architecture — using a controlled intervention to identify biologically responsive lipids, then testing those signatures in a large prospective cohort — is methodologically elegant and more causally informative than typical dietary recall analyses. That said, the ARISTOTLE trial was tiny (63 participants, healthy young adults), limiting the precision of the lipidomic signature itself. The EPIC-Potsdam association, while statistically significant, is modest, and the mediation estimate's confidence interval spans 57%–100%, meaning the BMI pathway interpretation carries real uncertainty.
Practically, 25g of peanuts daily — roughly a small handful — is achievable for most adults. The finding that benefit may flow primarily through modest BMI effects rather than direct lipid-mediated mechanisms is both clarifying and humbling: it suggests peanuts are likely useful as a satiating, nutrient-dense snack within a calorie-conscious diet. Confirmatory trials in older or at-risk populations are needed before clinical recommendations evolve.