In mice made diabetic with a high-fat diet plus streptozotocin, low- and high-dose D-psicose (a rare sugar) improved serum lipids, lowered ALT and AST, reduced circulating inflammatory cytokines and cut hepatic fat deposition. Treated animals also showed better intestinal morphology, higher ZO-1 and Occludin, lower circulating lipopolysaccharide, and more Akkermansia and Lactobacillus at the high dose. Serum tryptophan and indole-3-propionic acid rose, kynurenine and the Kyn/Trp ratio fell, hepatic TDO2 dropped, AhR rose and NF-κB p65 phosphorylation declined.
All of this evidence comes from mice, and the authors rightly stress that it is associative. They ran no microbiota-transfer or AhR-blocking experiments, so the gut-tryptophan-AhR story is a hypothesis, not a demonstrated mechanism. The single HFD/STZ model, which mimics insulin deficiency plus diet-induced obesity, also limits how far the results travel. Human data on D-psicose are thin: small trials suggest modest effects on post-meal glucose, but nothing yet shows benefit for fatty liver or long-term outcomes. The microbiome angle (Akkermansia, indole metabolites, tighter gut barrier) fits a crowded literature in which many fibres, polyphenols and sweeteners produce similar signatures, so the novelty is incremental. Readers should not treat psicose as a proven liver therapy; the realistic takeaway is that it is a plausible candidate for human trials in people with diabetes and fatty liver, while weight loss, exercise and diet quality remain the interventions with established benefit.