The treatment for too much of a hormone turns out to be more of it

Picture the strangest possible natural experiment: a tumor on the pancreas that pumps out insulin nonstop, ignoring every signal the body sends telling it to slow down. Benjamin Bikman points to exactly this condition, called an insulinoma, as one of the clearest demonstrations of a principle that sounds almost backward: too much insulin causes insulin resistance. Patients with the most severe insulin overproduction from these tumors develop the most severe insulin resistance, while those with milder overproduction develop milder resistance. Every single one, eventually, develops some degree of it.

This isn't just a quirk of rare tumors. In a more controlled setting, researchers infused healthy, insulin-sensitive men with a steady drip of insulin, at doses the body would normally produce over a full day, and within just a few hours these men had become measurably insulin resistant. Bikman compares the underlying logic to how bacteria adapt to antibiotics or how a regular caffeine drinker needs more and more coffee for the same jolt: when a cell is bombarded constantly by a signal, it dials down its own response to survive the barrage.

Here's where the irony lands hardest. The standard treatment for advanced type 2 diabetes, once the pancreas can no longer keep up, is often to prescribe more insulin by injection. That insulin does bring blood sugar down, at least for a while, but because insulin itself drives resistance, it also raises the amount of insulin the body will need next time, creating a cycle of escalating doses. Even people with type 1 diabetes, who must take insulin to survive, can develop this same resistance, sometimes called double diabetes, if their diet forces them to inject unusually large amounts.

There is a hopeful thread inside this cycle, though. Bikman cites a study from the UK in which type 2 diabetes patients who cut carbohydrates for eight weeks saw a genuine recovery in how well their beta cells, the insulin-producing cells in the pancreas, functioned again. The assumption that those cells are simply gone once diabetes sets in, he argues, isn't always true. Sometimes they're not dead. They're just overworked, waiting for a chance to rest.

If you're on insulin therapy, start a conversation with your doctor about whether cutting carbohydrates could eventually lower the dose your body needs, since smaller blood sugar spikes generally demand less insulin to control. Pay attention if your usual dose starts causing lows, and treat that as good news worth adjusting for with medical guidance, rather than an inconvenience. Don't assume that damaged beta cells are gone forever; ask whether a supervised trial of carbohydrate restriction might let overworked cells recover some function, the way one UK study found happened within eight weeks. The cycle that escalating insulin doses create can run in reverse too, given the right conditions.