Pocket metabolic coaches
New consumer AI tools are turning raw glucose and bloodwork into actionable reports — Blood Sugar Journal (iOS) uses GPT‑4o‑mini to synthesize glucose/insulin data with iCloud sync, while a Claude‑built metabolic calculator computes HOMA‑IR and TG/HDL to flag early insulin resistance. Folks are even pitching AI eye‑scan concepts — yellow tint for liver issues, redness for fatigue, blurred patterns for glucose irregularities — as low‑cost preventive screens. (x.com) (x.com) (x.com)
Blood Sugar Journal is listed on the App Store under developer Dmitry Mashkin and the listing describes a free app with in‑app purchases and a Pro tier that adds AI‑generated summaries, plus local export and Apple Health integration. (apps.apple.com) The app’s Product Hunt page highlights an iOS‑26‑optimised build and promotes fast, design‑forward reporting and iCloud sync for cross‑device data, according to its product listing. (producthunt.com) Independent coverage and reviews note the app currently relies on manual entry for many metrics, that extended AI summaries are gated behind a paid subscription, and that device integrations and automated CGM imports are not universal. (toolradar.com) Multiple public metabolic calculators already compute HOMA‑IR and TG/HDL as screening metrics; one aggregator that offers both calculators and interpretation guidance is Metabolicum, which recommends combining markers because waist‑to‑height ratio misses insulin resistance in about 40% of lean people. (metabolicum.org) Clinical references use HOMA‑IR thresholds in population screening — a HOMA‑IR cutoff near ≥2.5 is commonly cited (NHANES contexts), while TG/HDL guidance varies from “optimal” below ~1.5 up to warnings above ~3.0 in different practitioner resources. (mdcalc.com) (reversing‑insulin‑resistance.com) Eye‑based screening ideas have precedent: conjunctival (scleral) icterus becomes visible once bilirubin reaches roughly 2–3 mg/dL and is a recognized early sign of liver dysfunction, while AI retina‑imaging has recently shown >99% accuracy for diabetic retinopathy in a study presented at ENDO 2025 — both findings are cited by ophthalmology and endocrine sources. (msdmanuals.com) (my.clevelandclinic.org) (endocrine.org)
