What Causes Type 2 Diabetes?

Insulin resistance and high levels of insulin and lipids all precede the development of metabolic dysfunction. Which metabolic factor is to blame?

Barbara E. Corkey

FAT TRIANGLE: Fat cells (left) release lipids into the bloodstream. High lipid levels, in turn, can trigger the release of insulin from pancreatic β cells (middle). Insulin then travels to cells of the liver (right) and to the body’s periphery. Understanding the cause of metabolic dysfunction and diabetes will require a detailed understanding of how these different tissues and organs work together to regulate blood sugar.DANIELA MALIDE, NATIONAL HEART, LUNG, AND BLOOD INSTITUTE; NIH; © SCOTT CAMAZINE/SCIENCE SOURCE; © ISTOCK.COM/MAXPDIA

Type 2 diabetes is a multifactorial metabolic disease.1 Obesity, elevated levels of lipids and insulin in the blood, and insulin resistance all accompany the elevated blood glucose that defines diabetes. (Diabetes is defined as fasting blood glucose concentrations above 7 millimolar (mM), or above 11 mM two hours after ingestion of 75 grams of glucose.) But while researchers have made much progress in understanding these components of...

Lifestyle choices characterized by inactivity have been postulated as one possible cause.  Researchers have also pointed the finger at nutrition, postulating that poor food choices can contribute to metabolic disease. However, there is thus far weak support for these hypotheses. Changing to a healthy diet typically does not result in significant weight loss or the resolution of metabolic dysfunction, and it is rare to reverse obesity or diabetes through increased exercise. Furthermore, there does not appear to be a strong relationship between body-mass index (BMI) and activity level, though exercise clearly has many other health benefits.

With such macroscale factors unable to explain most cases of obesity and diabetes, scientists have looked to molecular mechanisms for answers. There are at least 40 genetic mutations known to be associated with type 2 diabetes. These genes tend to be involved in the function of pancreatic β cells, which secrete insulin in response to elevated levels of the three types of cellular fuel: sugar, fat, and protein. In healthy young adults, circulating glucose concentrations above about 5 millimolar (mM) trigger release of insulin from β cells. When fatty acids or amino acids are also elevated, the glucose-triggered insulin release is greater. Insulin facilitates the uptake of these molecules by the body’s tissues, leading to a decrease in their levels in the blood.

Excess fat deposits on the body can release triglycerides and free fatty acids into the blood, causing hyperlipidemia.
See full infographic: WEB | PDF
To date, most researchers have focused on insulin resistance, or the failure of insulin-sensitive cells in muscle, fat, liver, and other tissues to respond to the hormone, as the driver of dysfunction in this feedback cycle, assuming the other metabolic changes observed in type 2 diabetics to be symptoms of such flawed insulin signaling. However, some insulin-resistant people are capable of maintaining normal blood glucose levels, albeit by producing higher-than-normal levels of insulin. Moreover, if insulin resistance leads to metabolic dysfunction, then increasing levels of insulin should restore metabolic homeostasis. But treating insulin-resistant patients with drugs such as sulfonylureas or injected insulin is actually followed by greater metabolic imbalance.

As the lipids arrive at the pancreas, they are converted to free fatty acids and taken up by β cells, where they join with coenzyme A to form long-chain acyl-CoA. This can trigger insulin secretion, leading to hyperinsulinemia. As more insulin in secreted, β cells may begin to run low on their insulin reserves and insulin synthesizing capacity, such that they are not able to fully respond to the next surge in glucose.© 2016 ANDREA CHAREST & KARINA METCALFIn light of these findings, it’s time to begin considering what other metabolic correlates of diabetes might be driving factors. Elevated levels of lipids and insulin—both of which are seen prior to and at the onset of type 2 diabetes—are top candidates. Hyperlipidemia, the state of persistently high circulating levels of lipids called triglycerides, stimulates insulin secretion, leading to elevated insulin levels, or hyperinsulinemia. Inducing hyperinsulinemia in animal models can lead to insulin resistance and obesity.2

As insulin from the pancreas binds to receptors in the liver, those receptors get internalized and the insulin generates signals to stop gluconeogenesis. High levels of insulin can exhaust the receptors, such that the cells of the liver (and muscle and fat) become insulin resistant, unable to respond to spikes in insulin and continuing to make glucose even though blood sugar levels are inappropriately high.© 2016 ANDREA CHAREST & KARINA METCALFPrevention and treatment of type 2 diabetes depends on correctly determining the cause of metabolic failure. In fact, two available drugs (metformin and thiazolidinediones) developed to treat insulin resistance may actually work by lowering lipids. In combination with mounting evidence from our group and others that lowering circulating levels of insulin and lipids can reverse metabolic dysfunction in rodent models, researchers must now consider possible causes other than insulin resistance and try targeting these factors for new diabetes treatments.

Exploring hyperlipidemia and hyperinsulinemia

Obesity is accompanied by an uptick in circulating triglycerides and free fatty acids that come from increased adipose tissue mass. Triglycerides stored in fat cells (adipocytes) are broken down into fatty acids, and these lipids can enter β cells, where they generate signals to increase insulin secretion, resulting in elevated blood insulin levels. This signaling cascade is typically initiated inside the b cell by the attachment of coenzyme A (CoA) to the fatty acids, forming long-chain acyl-CoA. Acyl-CoA itself is a well-established and potent signaling molecule and is the precursor of other important signaling molecules such as diglycerides and monoglycerides. In pancreatic β cells, acyl-CoA has been shown to directly stimulate insulin exocytosis, change membrane ion channel activity, and influence Ca2+ handling.3 (See illustration below.)

To explore the importance of long-chain acyl-CoA in the development of diabetes, researchers studying rodent models have replaced dietary long-chain triglycerides with medium-chain triglycerides, which are rapidly oxidized in the mitochondria and thus do not generate cytoplasmic acyl-CoAs to stimulate insulin release. When a pair of researchers at McGill University in Quebec tried this treatment in mice, the animals had lower fasting insulin secretion and restored ability to respond to stimulatory glucose. They experienced no weight gain or increased fat mass, and they did not suffer from impaired glucose tolerance, which typically accompanies a high-fat diet. In addition, mice that consumed medium-chain triglycerides were satiated more quickly than animals fed long-chain fats, reducing their overall food consumption.4

Stimulating fat burning to decrease lipid stores can also reduce circulating lipids and cytoplasmic long-chain acyl-CoA. Activating the transcription factor PPARα increases expression of enzymes required for long-chain acyl CoA oxidation. In rodent models of high-fat diet–induced obesity, treatment with the PPARα agonist fenofibrate effectively stimulated fat burning. In obese mice, fenofibrate reduced circulating levels of fatty acids and triglycerides while reversing hyperinsulinemia and hyperglycemia; in rats, fenofibrate also increased metabolic rate.5

These findings point to hyperlipidemia as a driving force in the development of metabolic dysfunction: by reducing levels of circulating lipids, researchers have successfully stunted the development of diabetes in animal models. But there is also evidence that hyperinsulinemia is the initiating defect that leads to obesity, hyperlipidemia, and insulin resistance. In 2000, Christian Weyer, then at the Clinical Diabetes and Nutrition Section at the National Institutes of Health, and colleagues found that hyperinsulinemia precedes and predicts the development of diabetes in Pima Indians.6 Their findings fit with several previous studies of people in other ethnic groups prone to obesity, such as Mexican Americans and Nauruans, that also concluded hyperinsulinemia predicts diabetes.7,8,9,10 

By reducing levels of circulating lipids, researchers have successfully stunted the devel­opment of diabetes in animal models.

Other evidence for hyperinsulinemia as a cause of diabetes comes from gastric bypass surgery, an effective treatment for some patients with type 2 diabetes. In 2011, Walter Pories and G. Lynis Dohm at East Carolina University published a review documenting that, following surgery, patients experienced a decrease in fasting insulin levels, along with decreases in blood glucose and resolution of diabetes within a week—well before patients showed significant weight loss.11 Inhibiting insulin secretion in rats can prevent the development of the metabolic abnormalities induced by a high-sucrose diet;12 in vitro, curbing insulin secretion can keep human islets from deteriorating and becoming ineligible for islet transplantation.13

As for how hyperinsulinemia might cause metabolic problems, it has been known since the 1970s that insulin can downregulate its own receptor. When insulin binds its receptor, the cell internalizes the complex, digesting the attached insulin and recycling most of the receptors to the membrane surface. At chronically high insulin levels, such recycling can exhaust the receptors, such that there are few on the cell surface to respond to further increases in insulin. Moreover, my group recently showed that hypersecretion of insulin can deplete the insulin reserves of β cells in vitro. This leaves the cells unable to fully respond to a surge in glucose—a precursor to β-cell failure. In this case, we exposed the β cells to excess glucose and fat to cause basal insulin secretion, while inhibition of secretion preserved the cells’ insulin content.14 Partial inhibition of insulin secretion in obese mice with elevated fasting insulin did not result in increased glucose levels, nor did the animals fare worse on a glucose tolerance test, indicating that elevated insulin was not necessary to maintain normal levels of circulating glucose.15

Thus, evidence exists from in vitro studies and work in animal models that metabolic health and insulin secretory performance improve by preventing hyperinsulinemia or lowering the amount of ingested long-chain fatty acids. These approaches are now moving into clinical testing, with some early success.

Predicting and treating diabetes

New diabetes therapies are desperately needed. Although the current standard of care—daily administration of insulin or drugs such as various sulfonylureas that trigger increased insulin release from β cells—is sufficient to control metabolic disarray, there are many untoward near- and long-term side effects. I and others contend that some of the standard therapies adopted by medical practitioners may actually be causing metabolic dysfunction; further increasing or stimulating insulin secretion or insulin levels in the presence of hyperinsulinemia may accelerate β-cell deterioration.

Insulin secreted by the pancreas travels first to the liver, where it suppresses hepatic glucose production and is degraded. As a result, blood insulin levels entering the liver are three times higher than the concentrations that ultimately reach the periphery. This may cause hepatic insulin resistance to occur before muscle insulin resistance. To deliver adequate insulin supplies to the liver to overcome such insulin resistance, then, requires the administration of three times more insulin than is normally found in the periphery. Such excess insulin worsens muscle insulin resistance and promotes triglyceride synthesis, leading to increases in body fat and weight gain that further perpetuate metabolic dysfunction. High insulin levels may also promote cell growth and proliferation, increasing one’s risk of cancer.

Several small studies have assessed the effects of inhibiting insulin secretion to preserve β-cell insulin content in obese and prediabetic subjects. A study in healthy men showed a lower glucose level during an oral glucose tolerance test following a single dose of NN414, a small molecule that inhibits release of insulin from β cells;16 another study documented improved glucose-stimulated insulin secretion in diabetics after seven days of treatment with diazoxide, which has the same effect as NN414.17 Paired with exogenous insulin to maintain normal glucose levels, decreasing β-cell stimulation may also help preserve the remaining β-cell function in type 1 diabetics, whose metabolic dysfunction stems from an autoimmune attack on the pancreatic islets.18 (See “Taming Autoimmunity,” The Scientist, June 2016.)

Standard therapies adopted by medical practitioners may actually be causing metabolic dysfunction.

A few human studies have indicated that reversing hyperlipidemia can also stall the development of diabetes. As seen in animal models, consumption of medium-chain triglycerides (such as those found in palm kernel oil and coconut oil), instead of long-chain triglycerides (such as those in olive oil), increased energy expenditure, satiety, and fat loss in obese humans.4 More-extensive, longer trials are needed to identify patients who would benefit from such dietary intervention. Experimentally reducing carbohydrates in the diet of prediabetic and diabetic patients is also a promising strategy, as carbohydrates are necessary to form triglycerides, and most cells burn fat if glucose is not available.

PPARα agonists, particularly fenofibrate, have been inconsistent in their effects in human trials. But bezafibrate, a fibrate that interacts with all three PPAR isoforms (α, β, and γ), has consistently lowered triglycerides and improved glucose handling in diabetics.19,20,21 It should be noted, however, that most fibrate studies were designed to assess cardiac outcomes; the use of fibrates specifically to prevent deterioration of metabolic health before the onset of overt diabetes has not been studied in humans.

Unfortunately, the field has been slow to adopt hyperlipidemia and hyperinsulinemia as prime targets for diabetes therapy. It’s difficult to change scientific thinking, and most researchers are still stuck on insulin resistance as the ultimate molecular cause of metabolic dysfunction. Yet modern treatments often worsen prognosis. The time has come to focus on ways to protect the β cell, and research is now revealing just how to do that. Diverse stem cell therapies in development that are designed to stimulate the production of new β cells could also improve pancreatic function. We must follow the science as it leads us in new directions, and thoroughly test some of these novel approaches that have begun to show promise. 

Barbara E. Corkey is Zoltan Kohn Professor of Medicine, Director of the Obesity Research Center and Vice Chair for Research in the Department of Medicine at Boston University School of Medicine. She would like to express special appreciation to Dylan Thomas, Marie McDonnell, Richard Corkey, and Stanley Schwartz for their valuable suggestions on this article.


  1. S.S. Schwartz et al., “The time is right for a new classification system for diabetes: Rationale and implications of the β cell–centric classification schema,” Diabetes Care39:179-86, 2016.
  2. S. Del Prato et al.,  “Effect of sustained physiologic hyperinsulinaemia and hyperglycaemia on insulin secretion and insulin sensitivity in man,” Diabetologia37:1025-35, 1994.
  3. B.E. Corkey, J.T. Deeney, “Acyl CoA regulation of metabolism and signal transduction,” Prog Clin Biol Res, 321:217-32, 1990.
  4. M.P. St-Onge, P.J. Jones, “Physiological effects of medium-chain triglycerides: Potential agents in the prevention of obesity,” J Nutr, 132:329-32, 2002.
  5. F.P. Mancini et al., “Fenofibrate prevents and reduces body weight gain and adiposity in diet-induced obese rats,” FEBS Lett, 491:154-58, 2001.
  6. C. Weyer et al., “A high fasting plasma insulin concentration predicts type 2 diabetes independent of insulin resistance: Evidence for a pathogenic role of relative hyperinsulinemia,” Diabetes, 49:2094-101, 2000.
  7. G. Gulli et al., “The metabolic profile of NIDDM is fully established in glucose-tolerant offspring of two Mexican-American NIDDM parents,” Diabetes, 41:1575-86, 1992.
  8. S.M. Haffner et al., “Decreased insulin secretion and increased insulin resistance are independently related to the 7-year risk of NIDDM in Mexican-Americans,” Diabetes, 44:1386-91, 1995.
  9. S. Lillioja et al., “Insulin resistance and insulin secretory dysfunction as precursors of non–insulin-dependent diabetes mellitus. Prospective studies of Pima Indians,” N Engl J Med, 329:1988-92, 1993.
  10. R.A. Sicree et al., “Plasma insulin response among Nauruans: Prediction of deterioration in glucose tolerance over 6 yr,” Diabetes, 36:179-86, 1987.
  11. 1W.J. Pories, G.L. Dohm, “Diabetes: Have we got it all wrong?” Diabetes Care,  35:2438-42, 2012.
  12. R. Gutman et al., “Diazoxide prevents the development of hormonal and metabolic abnormalities present in rats fed a sucrose rich diet,” Horm Metab Res, 17:491-94, 1985.
  13. K. Maedler et al., “Glucose- and interleukin-1 β-induced β cell apoptosis requires Ca2+ influx and extracellular signal–regulated kinase (ERK) 1/2 activation and is prevented by a sulfonylurea receptor 1/inwardly rectifying K+ channel 6.2 (SUR/Kir6.2) selective potassium channel opener in human islets,” Diabetes, 53:1706-13, 2004.
  14. K.A. Erion et al., “Chronic exposure to excess nutrients left-shifts the concentration dependence of glucose-stimulated insulin secretion in pancreatic β cells,” J Biol Chem, 290;16191-201, 2015.
  15. R. Alemzadeh, K.M. Tushaus, “Modulation of adipoinsular axis in prediabetic zucker diabetic fatty rats by diazoxide,” Endocrinology, 145:5476-84, 2004.
  16. M. Zdravkovic et al., “The effects of NN414, a SUR1/Kir6.2 selective potassium channel opener, in healthy male subjects,” J Clin Pharmacol, 45:763-72, 2005.
  17. E. Qvigstad et al., “Nine weeks of bedtime diazoxide is well tolerated and improves beta-cell function in subjects with Type 2 diabetes,” Diabet Med, 21:73-76, 2004.
  18. M.A. Radtke et al., “Six months of diazoxide treatment at bedtime in newly diagnosed subjects with type 1 diabetes does not influence parameters of β cell function and autoimmunity but improves glycemic control,” Diabetes Care, 33:589-94, 2010.
  19. J.H. Flory et al., “Antidiabetic action of bezafibrate in a large observational database,” Diabetes Care, 32:547-51, 2009.
  20. H. Tenenbaum et al., “Long-term effect of bezafibrate on pancreatic beta cell function and insulin resistance in patients with diabetes,” Atherosclerosis, 194:265-71, 2007.
  21. T. Teramoto et al., “Effects of bezafibrate on lipid and glucose metabolism in dyslipidemic patients with diabetes: The J-BENEFIT study,” Cardiovasc Diabetol, 11:29, 2012.

Correction (August 10): A picture incorrectly described as depicting a fat cell filled with lipid droplets has been removed from the article. The Scientist regrets the error.

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?