Personal Fat Threshold Theory Explained

In 2004, in an operating room in St. Louis, surgeons spent an afternoon removing about ten kilograms of fat from the abdomens of fifteen women (00:19). Eight had normal blood sugar, seven had type 2 diabetes. The researcher running the study, Dr. Samuel Klein, wanted to answer a question that had divided obesity science for decades: does body fat itself cause the diseases that travel with it? If so, removing it should help. Twelve weeks later, and again four years later, Klein checked every marker that mattered: insulin sensitivity, blood pressure, cholesterol, inflammation. None of it moved (00:58). Ten kilograms of fat came out, and the disease stayed exactly where it started.

"10 kilos of fat were taken out of the human, and the disease didn't notice." (Dr. Jordan Feigenbaum, [01:16])

That result should have settled something. Instead it produced a more specific theory: maybe it wasn't the fat under the skin, called subcutaneous fat, that mattered, but the fat packed around the organs, called visceral fat, which drains directly into the liver through a vein most people never think about. That idea organized the field for thirty years. Two follow-up experiments, led by researcher Fabrini in 2010, tested it directly. Patients getting gastric bypass surgery either had a structure called the omentum, an apron of visceral fat over the intestines, removed or left alone. A separate group had only the omentum removed with no other treatment. In both experiments, removing the visceral fat changed nothing (16:07, 16:26). Visceral fat is still real trouble: it makes proteins that thicken blood and raise blood pressure, and too little of a protective hormone called adiponectin (02:03). But cutting it out does not fix the underlying disease. It is a downstream symptom, not the original cause.

A Garage With a Fixed Number of Spaces

Podcast host Dr. Jordan Feigenbaum builds his alternative around a model that already has a name in the research literature: the personal fat threshold (10:20). It starts with insulin, the hormone the pancreas releases after eating, which travels through the blood telling tissues to absorb incoming fuel and, for fat cells specifically, to hold onto what they already have (06:33). Fat tissue works like a garage, storing boxes of energy that would otherwise cause trouble sitting in the blood. Garage size differs from person to person and is largely fixed by adulthood. The number of fat cells someone has is mostly set by the end of adolescence, something researchers proved using leftover radioactive carbon from nuclear testing in the 1950s and 1960s, which lets scientists date the birth of a cell the way tree rings date a trunk. About one in ten fat cells gets replaced every year, but the total count holds steady for life (14:00).

"The garage never gets more parking spaces. You can't really expand it. Whatever you were given is what you're working with." (Dr. Jordan Feigenbaum, [14:21])

As long as the boxes fit, plenty of body fat can sit there without causing disease. Trouble starts once the garage is full and fuel keeps arriving. Full fat cells stop responding to insulin's signal to hold on, so they leak fatty acids into the blood all day (13:02). A 2005 tracer study found that only about 15 percent of the fat inside a fatty liver comes directly from food. Close to 60 percent leaked out of the person's own overflowing fat tissue (12:39). That fat lands in the liver, the pancreas, the kidneys, and the muscle around the heart, tissues never built to store fuel. Doctors call this ectopic fat, meaning fat sitting in the wrong place (11:20). It is what actually triggers insulin resistance: a cell already packed with more fuel than it can burn has no reason to let more in, so it stops answering insulin's knock (11:39).

Two natural experiments make the case that location, not total fat mass, drives the disease. People born with congenital generalized lipodystrophy have almost no fat tissue anywhere on their bodies, yet by adolescence most develop severe insulin resistance and damaged, fatty livers (23:02). In 2002, a Yale and NIH research group gave three of these patients injections of leptin, the hormone fat tissue normally produces to tell the brain how much storage exists. Over three months, with no change in body weight, their liver fat fell more than 80 percent, their muscle fat dropped by about a third, and all three came off diabetes medication (26:46). The opposite experiment ran in reverse. A drug called pioglitazone tells fat cells under the skin to build more storage. Patients with type 2 diabetes who took it gained about three kilograms of fat, and their livers handled insulin nearly twice as well as before, while fat around their organs actually went down (28:21). Weight went up, and the disease improved anyway.

None of this proves the theory by itself. Feigenbaum is careful to note that every model gets patched to survive bad results, and the trick is telling a good patch from a bad one. He borrows an example from nineteenth-century astronomy: the French astronomer who correctly predicted the existence of Neptune from irregularities in Uranus's orbit later invented a fictional planet, Vulcan, to explain a similar irregularity in Mercury's orbit. Same method, one triumph and one dead end (31:22).

"A theory that's working right makes new predictions that then come true. A theory that's dying only explains away the things that already went wrong." (Dr. Jordan Feigenbaum, [32:03])

By that standard, the storage model has some support. It predicted that normal-weight people could put diabetes into remission after losing a small share of body weight, which has been observed. It predicted that genes controlling storage capacity should affect disease risk, and a genetic study of roughly 30,000 people found that gene variants linked to higher liver fat causally raise the risk of type 2 diabetes by about a quarter (33:29).

Why the Usual Numbers Miss the Point

This matters because the tools doctors use to screen for this disease were not built to see it. Body mass index, or BMI, weight divided by height squared, correctly identifies excess fat about 95 percent of the time when it comes back above 30, but that same cutoff misses roughly half of people who genuinely carry too much fat (04:06). Waist circumference does a little better at catching people and a little worse at being right when it flags them. Across 31 studies covering 300,000 adults, adding a waist measurement to BMI improved accuracy by only three to five percentage points (04:48). In 2025 the Lancet Commission on Clinical Obesity, with 58 commissioners signing on, restricted BMI to population-level screening and required that any individual diagnosis be confirmed with a direct fat measurement or waist ratio (37:35).

The gap shows up starkly in lean patients. Feigenbaum describes a 41-year-old man with a normal BMI of 24 who lifts three times a week and has a blood sugar marker called A1c of 7.1, consistent with type 2 diabetes (35:07). A previous doctor told him his liver enzymes were normal and that he could not have a fatty liver because he wasn't overweight. Both claims were wrong. In a study of 100 people with type 2 diabetes and normal liver enzymes, three-quarters turned out to have fatty liver on a scan, including about two-thirds of those who were not obese (35:48). A separate 20-year Swedish study of 646 biopsy-confirmed fatty liver patients found the risk of severe liver disease ran from about twice normal, in people with no scarring yet, up to roughly 100 times normal in those with cirrhosis. Inflammation markers added no extra predictive value once scarring was already known (37:12).

Feigenbaum's practical suggestions follow from this. He recommends checking a waist-to-height ratio, aiming for a number between 0.4 and 0.5 (41:02), and calculating the TYG index, a number built from a routine fasting lipid panel and glucose test that costs nothing extra and outperforms the more common triglyceride-to-HDL ratio, especially in Black patients, who often develop insulin resistance without the elevated triglycerides that ratio depends on (42:21). For lean patients with a new diabetes diagnosis, he recommends ruling out type 1 diabetes and a rare inherited form called MODY, since one study found that one in six lean patients labeled type 2 actually had something else (43:48). And when fatty liver is suspected, he recommends the FIB4 score, a formula built from age, two liver enzymes, and platelet count, followed by elastography, an ultrasound that estimates liver stiffness, rather than a biopsy (45:07).

None of these tools, old or new, can measure the one thing that actually determines risk: how much room a particular person's fat tissue had to begin with. Two people can carry the same thirty kilograms of fat, and one has spare capacity while the other is already spilling fat into the liver and the pancreas. That distinction reframes obesity not as a number on a scale but as a mismatch between fuel arriving and room to store it.

"The overflow is the disease we call obesity. It's the point where somebody's storage runs out and starts spilling into organs that were never built to hold it." (Dr. Jordan Feigenbaum, [46:06])

The practical upshot is almost anticlimactic: anything that keeps fuel from piling up over a long stretch helps, whether that means losing fat or, as Feigenbaum previews for a future episode, exercising in a way that pulls fat out of the liver without changing the number on the scale at all (46:53). The scale, it turns out, was never really the story.