📢 Obesity management medications are very effective anti-hypertensives. 📎 ncbi.nlm.nih.gov/pubmed/4290…

Despite being connected to a university library, I still don't have access to the full text of numerous journal articles. There must be a better way.

Using AI to protect us against AI doesn’t seem to get to the root cause of “alignment”

Semaglutide cut 🫀events by 20% in SELECT. What exactly mediates that effect is uncertain. The usual risk factors explain ~ a third of it. 📢 The prespecified SELECT mediation analysis (Colhoun et al., European Heart Journal 2026) put 👉weight, waist, hsCRP, HbA1c, blood pressure, LDL-C and kidney markers together at 31% of the benefit, with a 95% CI of −30% to 144%. 🟢 Waist (64%) and hsCRP (42%) carried the largest effect, and only hsCRP held up across all three analysis methods. ❌ Body weight alone: 20%, with confidence interval crossing zero 💭 Still lots of unknowns but can see how treating "obesity" is much more complicated than what happens on the scale! 🏁 doi.org/10.1093/eurheartj/eh…

The FLOW trial showed that semaglutide protects the kidneys in type 2 diabetes with chronic kidney disease. Here, REMODEL was designed to explain how. tldr; We still don't know! Secondary outcomes look interesting though! 👀 An interesting secondary finding: fat around the kidney ended 25% lower than placebo while body weight differed by only 5%. 📢 REMODEL (Tuttle, Bjornstad et al., Nature Medicine 2026) randomized 106 adults with type 2 diabetes and chronic kidney disease to semaglutide 1 mg weekly or placebo for a year, with kidney MRI in everyone and before-and-after kidney biopsies in 33. ❌ Primary otucomes: Kidney oxygen levels, blood flow and an MRI marker of inflammation did not differ from placebo. 📉 Resistance to blood flow in the kidney arteries fell 4% vs placebo (secondary outcome, statistically significant) 🟢 A second MRI measure read as scarring (water diffusion) held steady on semaglutide and worsened on placebo (secondary outcome), but biopsy scarring scores did not improve 📉 Fat around the kidney 25% lower than placebo, fat in the renal sinus 13% lower, body weight about 4.5 kg (5%) lower 📉 Urine albumin, a marker of kidney damage, fell 40% vs placebo (exploratory outcome) 📊 Kidney filtration (eGFR) fell 0.1 on semaglutide vs 3.4 mL/min/1.73 m² on placebo. 💭 I think this supports the notion that the location of fat loss & its associative dysfunction is much more important than the % of global fat loss. Fat around the heart (epicardial fat) has known direct and indirect effects on cardiac functon. A similar framework should hold for understanding perirenal fat and renal dysfunciton. 🏁 doi.org/10.1038/s41591-026-0…

Good riddance @calleymeans

A paper that should have never seen the light of day nor garnered any attention. Too much mental real estate was spent by intelligent people taking this down. Embarassing for those that should have known better: reviewers and editors at the JACC, the Lundquist Institute, and Dr. Budoff.

Retatrutide for sleep apnea — now in NEJM, two phase 3 trials, 80 weeks, with and without type 2 diabetes. 💥 It is an incredible time to be practicing medicine! 📉 AHI −34 events/h (9 mg) vs −10 on placebo without diabetes 📉 −25 vs −10 with type 2 diabetes 🫁 Hypoxic burden down ~70% 🎯 ~half without diabetes reached remission or mild OSA without sleepiness ⚖️ Weight −21 to −22% (no T2D), −18% (T2D) Similar AHI effect to tirzepatide in SURMOUNT-OSA. Smaller in diabetes. 🏁 doi.org/10.1056/NEJMoa261147…

If two people lose the same 10% of their body weight, does it matter whether they cut carbohydrate or fat? 📢 Petersen, Klein and colleagues (Cell Metabolism, 2026) supplied every meal to 55 adults with obesity, prediabetes and fatty liver, randomized to a ketogenic, Mediterranean or very-low-fat plant-forward diet. NIH funded; 42 finished, 14 per diet. 🔬 Weight loss matched at about 11.5 kg (10%) on every diet over about 5 months ✅ Muscle insulin sensitivity rose about 50% on all three, with no difference by diet 📈 Liver insulin sensitivity rose 2 to 3 times more on ketogenic than on the other diets 📉 Liver fat fell from 19% to 6% on ketogenic and from about 18% to 10% on the other two 🟢 HbA1c fell 0.6 points (7 mmol/mol) on ketogenic vs 0.1 and 0.4; prediabetes went into remission in 50% vs 29% and 7% ⚠️ LDL cholesterol fell 0.57 mmol/L (22 mg/dL) on Mediterranean and 0.26 mmol/L (10 mg/dL) on ketogenic; the trial was too small to call that difference either way ⚠️ The ketogenic diet also carried more protein (23% vs 15% of calories), and testing came after only 3 to 4 weeks of stable weight At the same weight loss, the ketogenic diet did more for the liver and blood sugar and the same for muscle. Whether that lasts, and how much of it comes from the extra protein, are still open ?'s. 🏁 doi.org/10.1016/j.cmet.2026.…

Should we offer continuous glucose monitors (CGM) to patients with type 2 diabetes who aren't on insulin? A recent NEJM published. tldr; YES. CONNECT (Oser et al., NEJM 2026) randomized 283 adults with HbA1c of 7.5% (58 mmol/mol) or higher at 22 US primary care practices. Both groups had six visits in 26 weeks, and a trial clinician reviewed glucose data and adjusted treatment at each one. 📊 HbA1c difference -0.9 points (95% CI -1.1 to -0.6), or -10 mmol/mol 📈 HbA1c below 7% (53 mmol/mol) in 46% with CGM vs 24% with usual care 📊 Glucose-lowering drugs were added in 31% vs 17%, mostly GLP-1 agonists ❌ Weight change did not differ (-2.5 vs -1.8 kg) 📊 Exploratory subgroups showed -2.1 points at HbA1c 10% (86 mmol/mol) or higher and -0.3 at 7.0 to 7.9%, with an interval crossing zero 🟢 No severe hypoglycemia, and 3 minor sensor-related bleeds The case is strongest for patients well above target who also get planned follow-up. 🏁 doi.org/10.1056/NEJMoa260857…