Weight & Metabolic

Weight and metabolic disease are the cleanest pharmacology story in adult medicine right now. The 2026 GLP-1 class — semaglutide, tirzepatide, and the next-gen triple agonists like retatrutide — has rewritten what's possible. Tirzepatide hit a mean 20.9% body-weight loss at 72 weeks in SURMOUNT-1. Retatrutide hit roughly 24% at 48 weeks in Phase 2. These are numbers that historically belonged only to bariatric surgery. The trade-off is also real: lean-mass loss runs 20–40% of total weight loss in most trials, GI side effects are common in the titration phase, and the discontinuation rate after a year is high.

Berberine is the most over-marketed substitute for this class. The data show it lowers fasting glucose and HbA1c comparably to a low-dose metformin in head-to-head trials. It does not produce 20%-range body-weight loss and was never going to. The "nature's Ozempic" framing collapses the moment you put the actual numbers next to each other. Where berberine is defensible is in mild insulin resistance, prediabetes, and as a foundation while waiting on a GLP-1 referral.

Body composition is the variable the trial endpoints often ignore. Two interventions can produce the same scale weight loss and very different physiologic outcomes — the lean-mass-sparing protocols (resistance training during the cut, protein at 1.6–2.2 g/kg, electrolyte management) move the long-term metabolic curve more than the drug picks. GLP-1 fragments like Fragment 176-191 and AOD-9604 keep coming up in the comparative literature for precisely this reason: they target fat oxidation without the lean-mass cost, though their human data is thin.

Metformin remains a quiet workhorse. It is cheap, well-tolerated by most, and the only widely-prescribed insulin sensitizer with decades of safety data. The TAME trial will tell us whether it earns a place outside diabetes care. The CGM data we've published shows that in non-diabetics, glycemic-variability reduction from metformin is modest and behavioral interventions (timing of carbs, post-meal walks, sleep) compete with it for the same endpoint.

Intermittent fasting and metabolic-syndrome cluster work remain useful but are no longer the headline. The fasting work that survived 2024–2025 replication was the time-restricted-eating data showing the effect is mostly calorie compression, not a magic metabolic switch. That is not a defeat — it is just a more honest understanding of what the lever actually is.