145: Why GLP-1 Drugs Don’t Fix High Insulin

Bikman builds this protocol from his review of the evidence. He repeatedly shows that it is insulin resistance, not obesity per se, that blunts GLP-1; the twin study proved that obese yet insulin-sensitive individuals have normal GLP-1. Therefore, targeting insulin rather than just weight is key. He advises practical interventions: reducing consumption of refined starches and sugars, which are the primary drivers of insulin spikes and chronic hyperinsulinemia; losing weight to shrink fat cells that contribute to insulin resistance; and improving stress and sleep (both of which raise cortisol and insulin). He notes that while this is the most fundamental approach, it is speculative whether L-cells can fully recover, but the logic is ‘sound’ and consistent with the gradient evidence. He contrasts this with GLP-1 drugs, which bypass but do not fix the L-cell defect.

Bikman acknowledges that GLP-1 drugs work precisely because they override the dysfunctional L-cell, but they don’t repair it. However, he speculates that if a person uses the drug to get relief from intense carb cravings and successfully transitions to a diet that keeps insulin low, then perhaps after weaning, the L-cells’ responsiveness will have returned. He frames this as an open question that would be worth studying. His overall message is that while drugs are powerful, they treat a downstream consequence; the real goal should be to restore the endogenous system by lowering insulin. This protocol is the pragmatic marriage of the two approaches: use the drug to buy time and willpower while attacking the root cause with diet.

After establishing that GLP-1’s main physiological job is to apply a ‘gastric brake’ that keeps postprandial glucose and insulin modest, Bikman walks through a series of human studies showing that the worse the insulin resistance, the more profoundly GLP-1 secretion after a meal is blunted. A 2001 JCEM paper in 54 type 2 diabetics found a clear gradient: the diabetic group had the lowest GLP-1, the prediabetic group was intermediate, and healthy controls were highest. Crucially, the insulin area-under-the-curve was the strongest negative predictor of GLP-1 output, independent of BMI. A 2008 study across BMIs from 20 to 61 showed that both obesity and glucose intolerance independently lowered GLP-1. The definitive evidence came from a 2009 University of Toronto paper in Endocrinology: L-cells (the GLP-1 factories) express insulin receptors. When L-cells are exposed to insulin acutely, GLP-1 secretion increases, cooperating in a healthy meal. But when they were chronically bathed in high insulin — the state of hyperinsulinemia — the L-cells became insulin resistant: insulin receptor expression dropped, downstream signaling was blunted, and the cells could no longer mount a normal GLP-1 secretory response, even to other stimuli. This was confirmed in insulin-resistant mice. The conclusion: chronic hyperinsulinemia doesn’t just cause classic insulin resistance in muscle and liver; it also silences the gut’s L-cells, creating a GLP-1 deficiency precisely when it is most needed.

Bikman dismantles the conventional wisdom by citing two key experiments. First, Flint et al. (2001) infused GLP-1 into obese men before a fixed breakfast. GLP-1 slowed gastric emptying so powerfully that nutrient absorption dropped by about a third and blood glucose barely rose above fasting. The insulin peak in the GLP-1 group was less than half of the placebo group. GLP-1 did not stimulate insulin; it reduced it by removing the glucose stimulus. Second, Hirota et al. (2019) gave healthy volunteers exenatide (a short-acting GLP-1 drug) before an oral glucose tolerance test and found both glucose and insulin peaks were ‘remarkably delayed and reduced.’ The authors described this explicitly as an ‘insulin-sparing effect.’ Bikman summarizes that GLP-1 under realistic conditions is a brake on postprandial insulin, not an accelerant. This reframe matters because it explains the consequences of losing GLP-1: without the brake, nutrients flood in, glucose spikes, and insulin is forced to rise far higher, feeding the vicious cycle.

Bikman highlights several studies: the 2001 JCEM paper that found insulin AUC was the independent negative predictor of GLP-1, not BMI; a 2002 (or 2001) study using the hyperinsulinemic-euglycemic clamp that divided non-diabetic men into insulin-sensitivity tertiles and found that at just 15 minutes post-meal, the most insulin-resistant men had GLP-1 levels only half that of the most insulin-sensitive; and the 2014 twin study in Diabetes Care, where in monozygotic and dizygotic twin pairs discordant for weight, whenever the heavier twin had higher insulin resistance, GLP-1 was blunted — but if the heavier twin had preserved insulin sensitivity, GLP-1 was normal. This pattern demonstrates that insulin resistance, not obesity per se, is the key driver. Bikman also notes the 1996 Gut study where obese women had a blunt GLP-1 response only to a high-carbohydrate meal (which drives a sharp insulin spike) and not to a high-fat meal (which does not), further implicating insulin. All together, these data suggest that GLP-1 deficiency is an acquired consequence of chronic hyperinsulinemia and could, in theory, be restored by lowering insulin.

Bikman walks through why this matters. Carbohydrate drives a sharp insulin spike; fat does not. The obese women’s L-cells were likely insulin-resistant due to chronic hyperinsulinemia, so the carb-induced insulin surge further silenced them, resulting in no GLP-1. In contrast, the high-fat meal did not trigger the same insulin spike so the L-cells could still mount a small response. He then translates this to satiety: the lean group got a big GLP-1-driven fullness signal and felt satisfied after the high-carb meal, while the obese group got no such signal and would likely go back for seconds or thirds. This directly links hyperinsulinemia, GLP-1 deficiency, and the behavioral drive to overeat carbohydrates, creating a feedback loop that perpetuates insulin resistance.