The Week’s Three Stories: Interactions, Dose Curves and the Fat the Scale Misses
Issue #17 · November 8, 2026
More Bleeding, Fewer Clots: What Two New Studies Say About GLP-1s and Blood Thinners
One November study found 18% more bleeding when blood-thinner patients started a GLP-1. An October study found 21% fewer clots in nearly the same population. Both numbers are real. Here is how they fit together.
Read →
Retatrutide Dosage: What the Trials Actually Used, Week by Week
There is no approved retatrutide dosage. Every number online traces back to a few trial protocols. Here is the real ladder - 2 to 12 mg - what each rung delivered, and where the benefit curve flattens.
Read →
Tesamorelin Cuts Visceral Fat and Adds Lean Mass. The Scale Barely Moves.
Two independent teams pooled the tesamorelin trial data in 2026. Visceral fat falls, liver fat falls, lean mass rises — and the scale barely notices. What the reviews really show, and what they don't.
Read →
The Week’s Three Stories: Interactions, Dose Curves and the Fat the Scale Misses
This week’s research asked three different questions about what a treatment changes—and what a headline leaves out. Two studies put GLP-1 medicines beside blood thinners and surfaced a small bleeding signal alongside evidence of fewer clots. Retatrutide’s trial record showed why dose escalation is part of the intervention, not a footnote. And two tesamorelin reviews found changes in visceral fat and lean mass that barely registered on the scale.
These are not three invitations to stack compounds or copy trial protocols. They are examples of why mechanism, population and outcome matter as much as the compound name. Observational associations are not randomized proof; trial doses are not personal instructions; and a finding in a narrowly defined population does not automatically travel to everyone. A relative-risk change can sound dramatic while the absolute difference remains small; a body-composition shift can be missed by a scale; and a dose-response curve can flatten before side effects do. Those distinctions are practical, not academic: they change what a patient should ask and what a prescriber should monitor. Across all three stories, the most responsible reading is often less tidy than the headline.
This week: semaglutide and tirzepatide in anticoagulated patients, investigational retatrutide’s dose-response curve, and tesamorelin’s body-composition evidence. The useful thread is precision: what was measured, in whom, and what the result cannot tell us.
GLP-1 medicines and blood thinners: two signals can coexist
Two observational studies in people with type 2 diabetes and atrial fibrillation reached findings that sound opposed. A Medicare target-trial emulation matched 930 older adults per group who started a GLP-1 medicine or a DPP-4 inhibitor while using a direct oral anticoagulant. Over roughly a year, the GLP-1 group had an 18% higher relative rate of recorded bleeding (hazard ratio 1.18; 95% CI 1.03–1.36). The estimate does not tell us that every event was serious, and its absolute importance depends on baseline risk.
A much larger TriNetX analysis matched 27,548 pairs and found fewer venous clots over one year among GLP-1 users: 2.07% versus 2.56%, hazard ratio 0.79. Pulmonary embolism and deep-vein thrombosis were also lower. These studies do not cancel each other out: they used different datasets and outcomes, and both remain vulnerable to residual confounding. Delayed gastric emptying could alter absorption of oral anticoagulants, while weight, blood pressure and inflammation changes could plausibly shift vascular risk in the other direction. The practical conclusion is not to stop anticoagulation; it is to have a prescriber or pharmacist review the combination and monitor appropriately.
Retatrutide: the ladder matters as much as the target
Retatrutide remains investigational and has no approved dosage. The published studies describe protocols, not a recommendation. In the phase 2 obesity trial, 338 adults received weekly doses from 1 to 12 mg for 48 weeks. Higher-dose groups began lower and escalated; that slower start mattered because gastrointestinal effects increased with dose. At week 48, average weight changes ranged from 8.7% at 1 mg to 24.2% at 12 mg, versus 2.1% with placebo. The result is substantial, but it is a group average under trial conditions—not a forecast for an individual. The phase 2 study also showed a dose-related rise in heart rate, peaking around week 24 before declining, a reminder that biological response changes over time. The headline percentage needs its denominator and timeframe attached: participants were followed for 48 weeks, and the placebo group also changed. Neither the largest arm nor the most striking average tells us who benefited, who discontinued, or what a future approved regimen might look like.
Phase 3 protocols used a measured staircase: start at 2 mg, then increase in four-week steps toward assigned targets. In TRIUMPH-1, the published 80-week estimates were 17.6%, 23.7% and 25.0% weight loss at 4, 9 and 12 mg, compared with 3.9% on placebo. The gain narrowed at the top: about six percentage points from 4 to 9 mg, then roughly one more from 9 to 12 mg. Nausea and vomiting rose with dose, while trial protocols included ways to reduce dose when tolerability or excessive loss became a concern. That balance—not a single maximum—is the information a protocol contains.
There is an important boundary here. The medicine’s roughly six-day half-life and carefully monitored, pharmaceutical-grade trial setting cannot be reproduced by copying a dosing chart or using an unregulated product. Differences in estimands also explain why headlines quote different weight-loss figures: an analysis of those who remain on treatment answers a different question from one that accounts for discontinuations. The strongest reading is methodological: titration, adherence and tolerability shape the result, and these trial numbers are not self-dosing guidance.
Tesamorelin: body composition can move while weight stands still
Two 2026 meta-analyses pooled a small, overlapping set of randomized trials of tesamorelin in people with HIV-associated abdominal fat accumulation. Across four trials and 909 participants, one analysis estimated a visceral-fat reduction, a 1.42 kg increase in lean body mass and a 1.61 cm reduction in waist circumference. A second review reported similar waist and lean-mass estimates, plus reductions in liver and trunk fat. BMI and subcutaneous fat did not significantly change. The striking point is not that the scale failed; it is that it measures total mass, not where tissue changes. The first analysis also found high heterogeneity for its visceral-fat outcome, which signals that the contributing trials did not all produce the same-sized effect. That uncertainty belongs alongside the pooled average, not beneath it. A pooled estimate is a useful map of the evidence, not a guarantee that each participant—or each clinical setting—will reproduce it.
The overlap between reviews matters: two analyses of much of the same trial base are not twice as much evidence. One visceral-fat estimate had substantial heterogeneity, and the results apply to a specific population and treatment context. Tesamorelin acts through the growth-hormone-releasing-hormone pathway, raising growth hormone and IGF-1; that mechanism is not a license to generalize findings to other growth-hormone-axis peptides or to use outside the approved indication. The pooled evidence also describes benefit while treatment continues, with long-term durability after stopping unresolved. A trial now testing function alongside exercise may help determine whether composition changes translate into meaningful physical outcomes.
What to watch next
The next useful evidence will come from studies that separate association from cause, track medication combinations and clinically important outcomes, and measure function rather than relying on weight alone. For retatrutide, further phase 3 results and regulatory decisions will matter more than online dose ladders. For tesamorelin, the question is whether changes in visceral fat and lean tissue improve function—and how durable those changes are. Until then, read populations and endpoints closely, and take treatment decisions to a qualified clinician. Watch especially for better estimates of absolute event rates in anticoagulated GLP-1 users, transparent reporting of who leaves long weight-loss trials, and tesamorelin follow-up beyond the treatment period. Those details could change how the evidence is used far more than another isolated headline percentage. For now, the careful answer is also the useful one: ask what changed, for whom, over what period, and at what cost.
This week’s three stories involve semaglutide and tirzepatide in the GLP-1 class, investigational retatrutide, and tesamorelin’s growth-hormone-releasing pathway. The BioStack Generator lets you map compounds against your goals and planned schedule, so potential overlaps and questions are visible before a clinician conversation. Use it to organize research—not to prescribe, validate an unapproved compound, or replace medical advice. Build your research stack at https://app.oriabiostack.com.