Insulin Resistance — Medicine’s Systemic Blindness

There is a silent epidemic in much of the industrialized world. As our lives become modernized, our nutrition and hygiene improve, and our knowledge of medicine enables the early detection and treatment of an increasing number of pathologies, a novel illness also creeps in. Completely asymptomatic for years, even decades, by the time it manifests, it has morphed into a monstrous entity, well beyond the reach of even the most advanced therapies modern science can offer.

I have seen this monster everywhere in the hospital — most prominently as type 2 diabetes, but also as heart attacks and strokes, as fatty liver, as menstrual irregularities (polycystic ovarian disease), as sleep apnea, as gout, as endometrial, hepatocellular, colorectal, and pancreatic cancers, and as Alzheimer’s disease. The condition is given a rather abstract name: insulin resistance.

Throughout the medical curriculum, insulin resistance appears repeatedly as a contributing factor, and yet I never received a comprehensive course on what it is or how it develops. Nearly all doctors agree it is lifestyle-related and rapidly rising. Yet few truly understand what it is. This ignorance is not due to intellectual laziness or bad faith — insulin resistance escapes clinical awareness precisely because it does not fit the fundamental logic of modern medical thinking. The harder you study to become a doctor, the more you learn what to look for, and the more likely you are to create a blindspot for insulin resistance. Let me explain.

The food we eat is delivered to the cells through the circulation, but most of the cells in the body cannot just senselessly uptake nutrients without any concern for the rest of the body. If cells were continuously uptaking glucose from the blood, then blood glucose would be rapidly eliminated, and either we would have to keep replenishing it by eating, or the liver would need to constantly expend energy to make glucose. Instead, most cells in the body will only uptake glucose when they are “allowed” to do so by a hormonal signal known as insulin. Insulin rises in response to feeding. When there’s a bulk delivery of glucose, it is secreted into the blood to instruct bodily cells to uptake and store or use it. Once the glucose level returns to a certain low level, insulin release is stopped, and uptake by the cells is halted.

Insulin resistance is predominantly a cellular process — the cells’ machinery to respond to insulin is dampened, or the cell becomes less sensitive to insulin, and it takes a stronger signal (more insulin) to achieve the same cellular uptake of glucose. So is insulin resistance a disease where the insulin-responding machinery of the cell is broken by some insult? Not so. Rather, cells themselves have ways to turn it off — there is an “insulin sensitivity” tuner programmed into every cell, and it can modulate how much to respond to insulin as it sees fit. This was a major surprise for me — why would we have a preprogrammed pathway that could make us sick in so many ways? The big mystery is in how “it sees fit” — what makes cells reduce responsiveness to insulin, and what is the point of it?

Insulin is a very crude signal to distribute nutrients to the body. There are certain contexts where glucose is released into the circulation for a specific organ, but any increase in glucose would trigger insulin release. In these circumstances, insulin shouldn’t trigger all body cells to uptake glucose, but only certain groups of cells. Two organ systems can induce insulin resistance to direct glucose their way. First is the brain through the stress response. Stressful scenarios demand nutrient mobilization, and the brain must work quickly and accurately to compute what to do next. Stress hormones like cortisol prevent peripheral organs from using the glucose, so that more is available for the brain. Another organ system is the immune system. Mounting an immune response is highly energetically costly, and activated immune cells require more glucose than other types of nutrients. Inflammatory cytokines therefore prevent other organs from uptaking the released glucose so it will be allocated to them. Therefore, acute insulin resistance is a common physiological phenomenon critical for survival — not a disease process on its own.

The problem is that these states are never supposed to last too long. In evolution, both stressful events and infection are fleeting, with the expectation that a period of lower glucose demand would soon resume. Modern lifestyle, however, can keep this “emergency” protocol activated chronically. Chronic stress can be psychological, or due to circadian disruption and sleep deprivation, and chronic inflammation can be due to obesity, lack of exercise, or potentially from poor diet leading to a leaky gut. 

Another major reason for chronic insulin resistance is the chronic elevation of insulin itself — insulin turns off its own signaling (apparently insulin is “supposed” to be an acute signal). But when a person eats a lot of insulin-demanding food without long periods of fasting, insulin levels are chronically high, which can also reduce responsiveness to insulin. 

Fructose, a component of table sugar, could also increase insulin resistance. Fructose is not just a caloric source, but apparently an environmental signal for upcoming winter — in nature, fructose is not supposed to be available all year round, but is rather seasonal. Fruits become enriched in fructose after a whole summer of photosynthesis, prior to nutrient-scarce winter. Eating fructose may signal the body to prepare for the upcoming fasting — if you are not likely to eat a lot of food for months (or nothing at all during hibernation), then you must pack tons of fat and use glucose wisely. The body can only store so much glucose, and making glucose from other parts is costly, so insulin resistance is induced to prioritize the brain that needs glucose more. However, in modern life, fructose availability is not restricted to the fall. When you drink coke, you might be misleading your body to prepare for metabolic “winter” that never actually arrives — you only get fat and insulin resistant. These are some of the ways how a modern lifestyle leads to a persistent insulin resistance state.

Insulin resistance itself has absolutely no symptoms, and the reduced responsiveness to insulin is easily compensated by the pancreatic beta cells making more insulin. When insulin resistance becomes prolonged, pancreatic beta cells also expand in mass to produce chronically elevated insulin levels. This compensation can work out until the end of a person’s life, if the person is lucky and has good genetics.

However, pancreatic beta cells, much like any other endocrine glands, also have a design vulnerability. Endocrine cells regulate the amount of hormonal production by the number of endocrine cells. To dynamically meet the body’s ever-changing situations, they are constantly dividing and dying to adjust the overall amount of insulin secretion. If the glucose is too high, more proliferate and fewer die. If too low, fewer proliferate and more die. Beta cells therefore have a mechanism to detect how much glucose is available, and this detection system is tied to glucose metabolism within the cells. But metabolism is a complex multi-step process, each step prone to mutations. If the beta cells produce more metabolic signals despite glucose being normal, the cell could falsely assume glucose is high and proliferate more. The daughter cells would also have the faulty mutations that would wrongly assume high glucose. These mutants will now produce more insulin than normal, leading to low glucose. The “normal” beta cells would now die more, to try to bring the glucose back up. Soon, the whole pancreas will be dominated by mutants — guarding blood glucose will be impossible. To avoid this mutant takeover, beta cells have a mechanism known as glucotoxicity. When there are a lot of metabolic signals within the beta cells, they intentionally have weak protective mechanisms against damaging metabolites produced from excessive metabolism (they have low antioxidant enzymes) so that they will be removed. Such “death under excessively high glucose” didn’t become a major problem in evolution, as excessive feeding and prolonged high glucose levels were rare. However, if insulin resistance constantly demands the beta cells to create more insulin, gradually this “glucotoxicity” will start to kill off beta cells. Some people simply have stronger beta cells due to their genetics, but others will begin to deplete their beta cell population. This is, however, a very slow process over years to decades. Beta cells have a surprisingly large reserve capacity, and only when the majority of the mass is lost would the blood sugar start to increase. But when it does, the disease is in the terminal stage — most of the beta cells have already died, and it is very difficult to reverse it.

If blood sugar rises above a certain level, we call it diabetes. Glucose is an aldehyde that can react with amino groups of proteins. This reactivity is kept at bay when the concentration is low, but if its levels chronically double, the reaction is unleashed. Modifying proteins this way will damage their structure and function. One of the reasons why glucose level in blood, but not other nutrient levels, is obsessively controlled is because it can be a potent poison even at double concentration. Particularly the inner linings of blood vessels, peripheral nerves, kidneys, and retinal cells, that cannot self-regulate the amount of glucose flowing into the cells, are the first to be damaged by elevated glucose. Nonetheless, this is not like acute poison killing the cells in minutes — the timescale is again years or even more than a decade. With damage so slowly accumulating over time, once it is fully established, there is no way of reversing it.

The clinical problem of insulin resistance, however, does not necessarily require beta cell failure. Even with completely normal glycemia, insulin resistance in different organs causes various diseases. This is because insulin resistance does not mean complete insensitivity to insulin, but rather that some pathways remain sensitive to insulin, whose chronic activation cause disease. Fatty liver due to hepatic insulin resistance, now the most common liver disease, is due to lipid metabolism downstream of insulin stimulation being preserved (FYI: hepatic insulin resistance mostly impairs suppression of gluconeogenesis). So increased insulin still triggers lipid synthesis inside the liver. This floods the liver with too much fat to clear, and since lipid clearance takes a lot of oxygen, the liver becomes hypoxic. To protect itself from dying of suffocation, the liver turns off lipid clearance to avoid requiring more oxygen. This completely tips the balance and packs the liver with fat, impairing function and ultimately increasing the risk of liver cancer.

Excessive insulin and insulin resistance also impair ovarian function, leading to polycystic ovarian syndrome (PCOS), now the most common endocrine disorder in women. Producing offspring is tremendously energy-costly, and child-rearing would cost even more. Therefore, the maternal body is extra-sensitive to energy availability signals in order to initiate a menstrual cycle that could potentially lead to pregnancy. If insulin is chronically elevated, this confusing nutrient signal disrupts the normal cascade of brain-ovary communication, leading to menstrual problems (for those curious: usually menstrual cycle is regulated by sequential rise of 2 hormones, FSH and LH, regulated by GnRH. PCOS leads to irregular/rapid pulse GnRH that leads to mostly LH secretion, not FSH. LH stimulates theca androgen synthesis, which is preserved, but low FSH prevents its conversion to estradiol, leading to hyperandrogenism and anovulation).

Blood vessels, particularly their inner lining (endothelial cells), can also develop insulin resistance, but again, resistance impacts a particular branch of signaling (PI3K/Akt/eNOS) without affecting other branches (MAPK/ERK). Combined with insulin resistance-induced elevation in blood lipids (by failing to suppress their release from adipose tissue or the liver), this can lead to atherosclerosis, which can precipitate heart attack or stroke. I once asked a cardiothoracic surgeon how much overlap there is between diabetics/prediabetics and patients requiring cardiac bypass due to atherosclerosis in the coronary arteries. The doctor told me it is surely over 80%. Considering these are people already showing signs of abnormal blood glucose regulation, which happens only after beta cell failure, my guess is that almost everyone with clinical atherosclerosis has insulin resistance.

We have delved deep into the pathophysiology of insulin resistance. This was necessary to understand why the current clinical care is deeply inadequate, not necessarily due to physician negligence but rather due to how modern clinical medicine approaches disease. The first feature of insulin resistance and its associated diseases is that it is a very chronic disorder, and for the majority of its existence it is completely asymptomatic. Only at terminal stages does the pathology precipitate symptoms. In the asymptomatic phase, there is essentially no easily measurable biomarker or diagnostic test that could reliably screen for insulin resistance. As we’ve discussed above, prior to beta cell failure, blood glucose is completely normal. One way to probe insulin resistance is to check fasting blood insulin. Glucose, being a redox-active aldehyde, is easy and cheap to measure by doing a redox reaction in a small device called a glucometer — all you need is just a single drop of blood. Insulin, however, is a protein with a specific structure, whose detection is not simple. The first reliable quantification with radioimmunoassay got a Nobel prize. Today, we can use ELISA, but both capture insulin by specialized antibodies that can uniquely bind to insulin protein. Antibodies are expensive, ruling it out as a routine method of diagnosis.

Clinical medicine begins either with a symptom or, in a minority of cases, with a biomarker — typically a blood parameter. Insulin resistance provides neither, and is therefore ignored most of the time. Only when a symptom emerges, or beta cell failure is so advanced that blood glucose starts to rise, would clinical care begin. Some studies have reported that persistent insulin resistance could start as early as the 20s, suggesting decades of disease progression before symptom emergence or glycemic disruption. In many cases, clinical care does not start until years of elevated glucose level itself causing further problems, such as diabetic nephropathy, retinopathy, and neuropathies. These are unanimously untreatable. Why would we allow for such a duration of disease progression, and start addressing it only at advanced stages? Obviously it would be difficult to reverse pathologies that were established over such a long time, as demonstrated by the difficulty of therapy against type 2 diabetes.

I remember in medical school, during a lecture by a diabetologist, I asked him how he addressed the problem of all these years of insulin resistance that actually provides the best window for reversing the disease. I asked if a patient hasn’t changed glucose levels, how would he approach treating it. He appeared puzzled, and told me, if there’s no change in glycemia, there is nothing to treat. He cannot tell patients to preemptively “live healthier” and reverse insulin resistance. In fact, the whole specialty of diabetology is just about controlling blood glucose — without it, they wouldn’t even think it’s their domain to do something about it. In this sense, it is probably more accurate to say the current paradigm in diabetes-care is “disease-care” rather than “health-care.”

In fact, the clinical obsession with glucose micromanagement, undoubtedly fueled by its easy point-of-care detection, has led to aggressive insulin therapies that have proven mostly useless and even harmful. Insulin itself can drive much of the pathologies of insulin resistance due to partially preserved signaling, so just treating blood glucose aggressively with insulin — which does address direct glucose toxicity — nonetheless fails to address insulin-driven pathologies.

Another systemic challenge revealed by a condition like insulin resistance is the current division of labor in clinical care. Insulin resistance is truly a systemic disease, with manifestations in many different organ systems. If we view PCOS as purely a gynecological problem, or fatty liver as a concern for hepatologists, and avoid going beyond one’s specialty for the fear of “overstepping”, we miss the greater metabolic derailment that requires our attention. Management of these conditions should be united to provide a general improvement of metabolic health, not just direct peripheral treatment of the issue at hand, as it is usually addressed in clinics today. We need a concrete evidence-based plan of weight management, dietary intervention, exercise, sleep tracking, and general lifestyle guidance to be treated with similar seriousness as we treat pharmacotherapy. Alas, despite the millions that go into (often futile) drug hunting, similar investment is not made in prevention — meaning, finding a better way to help people live a better lifestyle.

Addressing these systemic problems requires fundamental restructuring of clinical care itself. The way we approach disease today is that we only check in with doctors when we have complaints or occasionally for checkups to look for pathological deviations. Meanwhile, insulin resistance requires routine monitoring and assessment of health states — we need to probe biology, not just statistically associated surrogate markers, like glucose, as we do now in the clinics (see also Clinical Medicine is Bayesian). Remember insulin resistance itself is not a disease, but rather the lack of flexibility to switch it off over a long time is the disease. Therefore, it requires a kind of omnipresent health monitoring beyond the current mentality of “check back in when something’s wrong.” At the risk of sounding like sci-fi, I often envision a future where medicine is a more routine presence in a person’s life, where we are able to probe biology itself daily, rather than starting from symptoms. 

Modern medicine has doubtlessly saved many lives through its innovations and theoretical frameworks. But we must be aware that most of the philosophies and approaches of clinical medicine were not the ultimate logical conclusion but rather a temporary solution to help organize information. Once exceptions and anomalies accumulate to a certain quantity, there is no ultimate reason to hold on to the current paradigm of how we approach diseases and treatment. Considering a new therapeutic philosophy does not mean that we must instantly abandon the current clinical framework — we could simply extend it to address major gaps. But this process must be completely conscious, as the current approach of medical information gathering is so ingrained in both the public and the clinicians’ imagination that it feels almost intuitive. In order to fuel a paradigm change, we must invest extra attention to current “misfits”, like insulin resistance. A new generation of medicine is not created by ever deeper characterization of biomarkers and subtle protocol sequences, as much of clinical research does. If we really want to address the root cause of diseases, then we cannot rationalize away how personal lifestyles are beyond the domain of clinicians — as much creative effort should go into trouble-shooting this preventive aspect as we currently spend on drug development.

All images are from Wikipedia