GLP-1s Require Personalized Dosing for Effectiveness

GLP-1s broke the standard dose model. P4 medicine is how you fix it.

You cannot run a fixed titration schedule on a molecule that behaves this differently across people and call that good medicine.

Predictive: before anyone starts a GLP-1, you need to know what you are working with. Fasting insulin, HbA1c, CRP, DEXA for body composition, lipids, thyroid, sex hormones, liver and kidney function. Casey Means makes a point in Good Energy that stuck with me: only 6.8% of US adults have all five basic metabolic markers in range without medication. That means the starting point varies enormously. Someone who is metabolically far from baseline will likely need a different dose than someone who is not. If you don't measure first, you are guessing.

Preventive: the goal is not to prescribe the drug and see what happens. Labs can move before weight does. A wearable picks up that resting heart rate is creeping up, or HRV is dropping, or sleep is getting worse, before anyone says they feel terrible. That is the signal a dose is too high. Catching it early is prevention. Ignoring it is how you get dropouts, and the STEP 1 extension data is clear on what dropout costs: about two thirds of the weight lost came back within a year of stopping.

Personalised: the dose where appetite is controlled, not suppressed, is different for everyone. Past the right threshold people under-eat, lose muscle with fat, feel bad, quit. Below it nothing changes. A fifth to a tenth of the standard starting dose is where some clinicians begin, not because it is a pharmacology standard, but because finding the individual threshold requires starting low and measuring the response.

Participatory: none of this works if the person on the drug is not in the loop. They are the ones reporting mood, logging food, wearing the CGM, sending back the data between blood draws.