Dr. Michael Snyder on Huberman Lab: Why One Diet Fails All

Snyder’s lab and others show people have distinct “glucotypes”: some spike strongly on potatoes, others on grapes, bananas, pasta, white vs brown bread, etc. Glycemic index tables are poor predictors at the individual level. A continuous glucose monitor (CGM) worn for ~14 days lets you see exactly which foods cause high and/or prolonged excursions (>140 mg/dL for non-diabetics, >180 mg/dL for diabetics) and which do not, so you can keep time-in-range high and tailor your carbohydrate sources.

Brisk 15–20 minute walks after high‑carb meals (e.g., white rice) noticeably flatten glucose curves by letting working muscles act as glucose sponges. Snyder notes that companies like January AI now actively coach users to walk after meals when CGM data shows a spike. Even simple “exercise snacks” (squats, pacing, calf raises) throughout the day help counteract the harms of prolonged sitting and improve metabolic control.

Beyond the classic type 1 vs type 2 split, Snyder highlights subphenotypes: muscle insulin resistance, beta‑cell secretion defects (his own case), hepatic (liver) insulin resistance, and incretin (GLP‑1) defects, plus combinations. These subtypes respond differently to foods, exercise timing, and drugs. For example, increasing muscle mass did little for Snyder because his main issue is beta‑cell insulin release, whereas muscle insulin–resistant individuals benefit more from morning activity and build muscle to improve glucose disposal.

“Fiber” is not one thing; there are many classes (arabinoxylan, inulin, beta‑glucan, resistant starch, etc.), each interacting differently with your microbiome and immune system. In Snyder’s crossover trial, arabinoxylan generally lowered LDL cholesterol ~25% but had no effect in some individuals, while inulin helped those same non‑responders. Other work (Sonnenburg, Gardner) shows extra fiber can lower inflammation in some but increase it in others. Practical implication: if one fiber supplement (e.g., Metamucil/arabinoxylan) doesn’t move your LDL or makes you feel worse, try a different fiber type instead of assuming “fiber doesn’t work.”

GLP‑1 agonists (Ozempic, Mounjaro, Farxiga-class) dramatically lowered Snyder’s HbA1c from 8.4 to 5.7 and stripped visceral fat and fatty liver on his serial whole‑body MRIs, but also caused significant weight loss (144→128 lb) he found cosmetically and functionally undesirable. He counters muscle loss risk by daily resistance training (alternating heavy and light days). GLP‑1s may also improve cognition and reduce cravings, but likely act via supraphysiological GLP‑1 elevations (~1000‑fold). For non‑exercisers, these drugs risk sarcopenia and frailty; for motivated patients, they can be a bridge to better cardiometabolic health.