The Four-Body Hypothesis Take four people, all 47. One thin. One normal weight. One mildly overweight. One obese. They look nothing alike. But suppose several of them share a hidden condition: skeletal muscle that is fat-infiltrated, insulin-resistant, mitochondrially weak, and metabolically noisy — muscle that has stopped signaling properly. A real metabolic phenotype sits underneath this thought experiment. It has a name. Metabolically obese, normal weight. MONW. Thin outside, fat inside. Normal BMI, visceral fat, intramuscular lipid, insulin resistance. These people can be invisible to BMI, because the scale says they are fine. Body composition, waist measures, insulin markers, and imaging can find them. We just rarely look. So the thin man may not be the control. He may be the same patient in a smaller package. Now put all four on a GLP-1. They will not lose the same weight. They will not lose the same ratio of fat to lean mass. The obese man may shed enormous fat. The thin man may barely move on the scale. The overweight man may lose a disproportionate share of lean tissue. But what if the weight loss is not the output that matters? What if lower insulin, lower energy intake, reduced lipid overflow, and altered nutrient signaling force each body to remodel whatever tissue is most metabolically inefficient — not because the drug recognizes bad muscle, but because bad muscle is what stops being affordable? Different bodies. Different curves. Possibly the same endpoint: less metabolic static, better muscle signaling, restored NK-cell function. The scale would say these are four different experiments. The immune system might say they are one. That is the speculation. And there may already be a hint sitting on a hard drive in Ireland. In the 2023 Dublin study, six months of semaglutide improved NK-cell cytotoxicity, IFN-γ and granzyme B. Only 9 of 20 participants lost more than 5% of body weight — and the authors reported that the immune improvements appeared largely independent of weight loss. So the first question is already answerable, with data that exists: Within that cohort, does NK improvement correlate with baseline BMI, with baseline metabolic dysfunction, or with weight lost? If it tracks weight loss, the convergence hypothesis weakens and this looks more like a fat-loss story. If it doesn't, convergence stays alive — though with n=20, a null correlation could just be low power. And every one of those twenty had obesity, so that cohort can never test four body types. It can only tell us whether the question is worth asking properly. The real test needs a new cohort: thin, normal weight, mildly overweight, obese — matched by age, and stratified by actual muscle quality rather than BMI. One afternoon with the existing spreadsheet can tell us whether the next study is worth running. It cannot replace the next study. And to be clear: none of this says a thin person should take these drugs. It says their muscle may be broken in a way the scale cannot see. The four-body hypothesis was never an argument for the drug. It is an argument against using BMI as a proxy for metabolic or muscle health.

@PatternNotes Two people can weigh the same and one is secretly metabolically wrecked on the inside. The scale is the worst health diagnostic we ever made mainstream.