Issue: February 2014
February 01, 2014
15 min read

As insulin nears 10th decade of discovery, experts reflect

Issue: February 2014
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Before the discovery of insulin in the dawn of the 20th century, patients with diabetes were given poor prognoses. Since its inception in 1921, and the first patient treated in 1922, researchers have used insulin as a means for controlling the most troublesome cases of diabetes.

Now, as it nears its 10th decade in use, researchers continue to learn more about the underlying pathophysiology of diabetes to develop more targeted approaches using insulin therapy.

Paresh Dandona, MD, PhD, State University of New York distinguished professor and chief of endocrinology of the University at Buffalo School of Medicine and Biosciences. 

Paresh Dandona, MD, PhD, State University of New York distinguished professor and chief of endocrinology of the University at Buffalo School of Medicine and Biosciences, identified insulin as an anti-inflammatory and antiatherosclerotic hormone. Photo courtesy of Paresh Dandona, MD, PhD.

Brief history of insulin

In 1869, a German medical student discovered what would one day become known as insulin-producing beta cells. Thirty years later, Oskar Minkowski, MD, a German physiologist, and physician Josef von Mering, MD, found that if a pancreas gland was removed from a dog, the animal developed symptoms of diabetes and died soon thereafter. Minkowski and von Mering were the first to propose that the pancreas played a vital role in sugar metabolism.

By 1920, Frederick Banting, MB, a Toronto-based surgeon stumbled upon a publication that spiked his interest in duct ligation. He and J.J. R. Macleod, MB,J. Bertram Collip, MA, PhD, and C.H. Best would go on to develop the life-sustaining therapy known as insulin for the treatment of diabetes.

The discovery of insulin led to the development of protamine zinc insulin in the 1930s and neutral protamine Hagedorn (NPH) insulin in 1950; synthetic insulin became “human insulin” by the 1980s and analogue insulin lispro in the late 1990s.

Jesse Roth, MD, FACP, investigator and head of the laboratory for diabetes and diabetes-related disorders at the Feinstein Institute for Medical Research in New York, has actively worked in the field for approximately 50 years.

“The question we asked back in the 1960s was how does the cell know that insulin is there?” Roth told Endocrine Today. “We now know that insulin, like every hormone, in order to do its work, needs to bind to a receptor — its own receptor.”

Roth and his colleagues were the first to introduce the direct study of cell surface receptors, including the insulin receptor. Studies from this group identified the role that the insulin receptor plays in genetic and acquired disorders of glucose metabolism.

Clinical ramifications

When looking at the history of insulin and how it was discovered, the initial insulins were very impure, according to Etie Moghissi, MD, FACP, FACE, an associate clinical professor of medicine at the UCLA David Geffen School of Medicine and a consultant in diabetes, endocrinology and metabolism in Marina del Rey, Calif.

Etie Moghissi

Etie Moghissi

“When insulin analogues came about, we got much closer to normal physiology, and that is why most of the time now, we use long-acting insulin as the basal and rapid-acting insulins; and we tend to back away from using regular insulin as much, especially in the outpatient setting,” Moghissi told Endocrine Today.

Rapid-acting insulins are viewed with a much greater anticipation because the risk for hypoglycemia and weight are associated with older therapies, she said.

“Overall, as a clinician who has been in practice for almost 30 years, I have seen that the role of insulin has evolved, the type of insulin that we have available are much better than before, and the devices and pens have made it much easier for me to teach patients to go on insulin and for them to accept insulin therapy as an option,” Moghissi said.


Overall, Moghissi said insulin’s past and present has been positive, and advancements in the development of insulin pumps have made a great difference in the delivery of the life-saving therapy.

“We have a come a long way. It has been so positive these past nearly 10 decades that we are making huge progress with the type of insulin and the technology to deliver insulin; and patients are benefiting from it,” she said.

However, it is important to make the distinction between type 1 diabetes and type 2 diabetes, Moghissi said. For example, patients with type 1 diabetes must be administered insulin. Metformin is the gold standard for patients with type 2 diabetes, and other agents can be considered. Still, many patients depend on insulin.

Concerns over timed intervals

It used to be thought that the injection itself was the limiting feature for patients’ acceptance of insulin, according to Alan J. Garber, MD, PhD, chief medical editor of Endocrine Today.

Alan J. Garber

Alan J. Garber

“It turns out that insulin injection, perse, is not a primary concern of the majority patients, and we see patients readily accepting injectable glucagon-like peptide-1 incretin agonists for type 2 diabetes. The concern seems to be insulin itself; and that’s what has dominated the evolution of insulin,” he said.

Irl B. Hirsch, MD, of the University of Washington Medical Center, told Endocrine Today that he may have been the first to coin the term “lag time” with respect to lengthening the intervals between the dose of insulin and time of meals.

“It’s important because what happens to the blood glucose is not just the dose of the insulin, but it has to do with getting the insulin into the bloodstream to the point that the glucose is, ideally, already starting to go down; so when a carbohydrate is eaten, you don’t have as high as a spike,” Hirsch said. “What we knew, but didn’t understand as well as we do now, is that before we had the rapid-acting analogues, we knew that, ideally, patients should wait at least 30 minutes before eating.”

Hirsch recalls working with pregnant patients in the 1980s, when those spikes in blood glucose were thought to be bad for the fetus.

“The obstetricians were having the patients wait 60 to 90 minutes, especially if their blood sugar was not perfect before they ate because their regular insulin was so slow,” Hirsch said.

Irl B. Hirsch

However, this lag time could potentially lead to hypoglycemia after the meal was absorbed, he said. Therefore, this method did not appear successful.

When insulin lispro was introduced in 1996, Hirsch and colleagues soon discovered that the lag time was still necessary. But in the past 5 years or so, Hirsch said the scientific community has become much more sophisticated with the use of continuous glucose monitoring.

“We’ve learned that if the glucose is trending upward, patients should wait longer than if it is a flat trend or a downward trend,” Hirsch said.

During the Juvenile Diabetes Research Foundation’s Continious Glucose Sensor Study, Hirsch and colleagues developed specific algorithms in the protocol, indicating that patients should wait a minimum of 10 minutes if their blood glucose was in the 100s; 20 minutes if in the 200s; and 30 minutes if in the 300s. They found that HbA1c levels decreased by 0.52%.

When they asked patients what they did differently, the most common response was that they waited between administering their insulin and eating meals.

“It’s the timing,” Hirsch said. “What most people don’t appreciate is that, for patients, especially those who are extremely insulin deficient, the timing of the insulin is a critical factor in terms of success or lack thereof; and it’s amazing how few people really appreciate that.”


Insulin’s anti-inflammatory, cardioprotective effects

More recently, Paresh Dandona, MD, PhD, State University of New York distinguished professor and chief of endocrinology of the University at Buffalo School of Medicine and Biosciences, identified insulin as an anti-inflammatory and antiatherosclerotic hormone.

Recent studies have demonstrated that insulin exerts an anti-inflammatory effect in addition to inhibiting platelet aggregation and the expression of other prothrombotic factors, according to Dandona.

“We made our first publication on this subject in 2001. And earlier, in 2000, we published a study showing an in vitro anti-inflammatory effect of insulin using human and aortic endothelial cells and cultures,” Dandona told Endocrine Today. “In those cultures, we demonstrated that insulin suppresses a proinflammatory transcription factor.”

When Dandona and colleagues used low-dose insulin at 2 units per hour, along with 5% dextrose, they found that within 2 hours of beginning the infusion, there was suppression of free radical generation, and with it, there also was suppression of transcription factor, he said.

“Clearly, now we were moving into an area where we wanted to ask whether a similar regimen of insulin infusion in acute myocardial infarction would lead to protection of the heart,” Dandona said.

In 2004, Dandona and other researchers published a paper in Circulation showing that a low-dose insulin infusion reduced the size of the infarct as measured by the increase in the enzyme creatine kinase, according to data.

“We also showed in that study that insulin suppressed C-reactive protein, and C-reactive protein increased the size of infarct,” he said.

Hyperinsulinemia, or the high insulins that predict MI and stroke, are essentially the product of underlying inflammation and, therefore, must be looked at that way, according to Dandona.

He is currently working on studies that he hopes will establish that insulin is definitely cardioprotective, he said.

Antiatherogenic effects of insulin

The relationship between insulin’s effect on the development of atherosclerosis has been debated for more than 40 years, according to George L. King, MD, director of research and head of the section on vascular cell biology at Joslin Diabetes Center and professor of medicine at Harvard Medical School.

George L. King,

George L. King,

“This is due to the fact that there’s a higher rate of cardiovascular disease in both type 1 diabetes and type 2 diabetes, especially in type 2 diabetes; and there’s a correlation of increased risk for CVD inpatients with insulin resistance,” King told Endocrine Today.

Tight glycemic control has been shown to improve some outcomes of CVD in type 1 diabetes, according to King.

“In addition, it has been suggested that insulin may contribute to the risk of CVD due to hyperinsulinemia, which can occur when insulin is administered exogenously,” he said.

The risk for CVD increases before the diagnosis of diabetes because patients with type 2 diabetes already have insulin resistance and hyperinsulinemia at the prediabetes stage, King said.

“Several studies testing glycemic control alone did not show too much improvement on CVD outcomes. However, it has been difficult to cleanly separate the contribution of insulin resistances from hyperinsulinemia on the vessel wall for the development of atherosclerosis,” he said.

The concern is that the peripheral hyperinsulinemia due to the higher insulin injections also may have a role in accelerating atherosclerosis, according to King.

“Since the 1980s, it has been shown that endothelial cells respond to insulin in a specific manner by increasing the production of nitric oxide, which has many anti-inflammatory properties and basal dilation. Therefore, it seems to have an antiatherogenic action,” he said.

This debate of whether insulin has an antiatherogenic or proatherogenic action and whether insulin can do both led to a proposal by King and colleagues in the 1990s that insulin could have both actions at different levels of insulin and throughdifferent mechanisms.


“It’s known that insulin’s metabolic action activates a glucose transport in lipid metabolism going through the AKT pathway [rs1 and rs2], whereas its monogenic action or long-term actions go through the mitogen-activated protein kinase (MAPK) pathway,” King said.

The literature has begun to prove this idea, he said. In mouse models, it has been shown that if insulin receptors or insulin action on the endothelial cells are eliminated, and the mice are exposed to high lipid diets, atherosclerosis will develop. However, if the insulin receptor only is removed in the endothelial cells, the insulin effect is lost. Therefore, these mice would develop atherosclerosis two to three times faster than mice with insulin receptors in the blood vessels.

“That suggests that insulin’s action on the blood vessel is actually antiatherogenic,” King said.

Higher doses for hyperinsulinemia

In a recently published paper in Endocrinology, Willa Hsueh, MD, and colleagues wrote that they sought to determine the cardiac responses to a high-fat feeding-induced hyperinsulinemia diet.

Willa Hsueh

Willa Hsueh

Hsueh, director of the Methodist Diabetes and Metabolism Institute in Houston and its head section chief in the division of diabetes, obesity and lipids, told Endocrine Today that her research with young and middle-aged male mice translates into clinical practice.

“Obviously, having insulin is crucial. The question is when patients with type 2 diabetes become insulin resistant, what is the role of hyperinsulinemia?” she said.

For example, there is some histological data indicating that islet cells become full of fat under certain circumstances, Hsueh said.

“Eventually, that fat causes lipotoxicity, which we think causes islets cell defects and leads to the so-called lipotoxicity and not enough insulin to overcome the insulin resistance, thus resulting in type 2 diabetes,” she said.

The study data indicated that high-fat diet-fed mice demonstrated increased cardiac glucose uptake. This was most recognized among the middle-aged mice in the absence of cardiac contractile dysfunction or hypertrophy.

In addition, researchers found an enhanced mitochondrial oxidation of palmitoylcarnitine, glutamate and succinate, and greater basal insulin in the hearts of high-fat diet-fed mice. This finding suggests that cardiac insulin sensitivity was maintained despite the subjects’ insulin resistance, according to data.

“This was the first evidence that, at least in some situations of obesity, having that extra insulin tic could be important to protect the heart in mitochondrial function,” Hsueh said. “The question is whether that is also true for the blood vessels because insulin has a growth effect and, in some situations, hypertrophy of muscle cells, and those kinds of changes … could potentially lead to more vascular injury.”

Hsueh said her institution currently gives insulin during heart failure, but it is unclear whether giving high doses of insulin would be clinically beneficial.

“Maybe now we could do more constant high levels of insulin, whereas in the past we didn’t have all of these different choices, and we couldn’t regulate the amount of time the insulin was on board,” she said.

An unaffordable future

Despite its many innovations, insulin has become unaffordable for many, Hirsch wrote in a recent editorial published in Diabetes Technology & Therapeutics.

“In the younger type 1 population, we don’t see nearly the amount of proliferative retinopathy causing blindness and end stage renal disease that we used to see. Frequent neurologic complications such as disabling gastroparesis, orthostasis, diarrhea and gustatory sweating are seen less often. The horrible things we used to see when I was a medical student are now very rare, which is wonderful,” Hirsch said. “But all of this has come at a cost; a cost that only the rich or very well insured can afford.”

According to Hirsch, the price of insulin has “skyrocketed in the past few years.” He wrote that while each insurer will require a separate co-payment, the co-payments are rising to match the retail price.

In his own analysis of 10 pharmacies within 3 miles from the University of Washington Medical Center, Hirsch reported that the price for one vial of insulin glargine (Lantus, Sanofi-Aventis) ranged from $185 to $231, with an average of $206. Pen insulin was more expensive, and there was a similar average for insulin detemir (Levemir, Novo Nordisk) at $192 per vial, $182 for one vial of insulin lispro (Humalog, Lilly), and $187 for one vial of insulin aspart (NovoLog, Novo Nordisk).

“What makes this even more amazing is the fact that in April 2012 (just 19 months ago), I was complaining that these same insulins cost about $130 per vial, an increase of about 120% for the 7 years prior to that. Now for the past 19 months we have seen an additional increase of more than 50%,” Hirsch wrote.

He told Endocrine Today that he fears the day when there is a cure for type 1 diabetes but patients are unable to afford the cost of the medication.

Although it is difficult to pinpoint what the next breakthrough in insulin therapy could be, researchers and clinicians agree that devices and ultra-rapid-acting insulin will surely play a role. – by Samantha Costa

Aljada A. J Clin Endocrinol. 2001;86:450-453.
Chaudhuri A. Circulation. 2004;109:849-854.
Dandona P. Clin Cornerstone. 2003;4:S13-20.
Dandona P. JAMA Intern Med. 2013;173:1905-1906.
Dandona P. MedGenMed. 2004;29:6.
Gupte AA. Endocrinology. 2013;154:2650-2562.
Hirsch IB. Diabetes Technol Ther. 2013;15:212-213.
Roth J. Diabetes Metab Res Rev. 2012;28:293-304.
Stout RW. Diabetes. 1996;45:S45-S46.
Paresh Dandona, MD, PhD, can be reached at the University at Buffalo School of Medicine and Biosciences, Dent Tower, 3980 Sheridan Drive, Amherst, NY 14226; email:
Alan J. Garber, MD, PhD, is the chief medical editor of Endocrine Today.
Irl B. Hirsch, MD, can be reached at the University of Washington Medical Center, 4225 Roosevelt Way NE, Suite 101, Seattle, WA 98105; email:
Willa Hsueh, MD, can be reached at the Methodist Hospital Research Institute, 6670 Bertner Ave., Houston, TX 77030; email:
George L. King, MD, can be reached at Joslin Diabetes Center, Harvard Medical School One Joslin Place, Room 4504, Boston, MA 02120; email:
Etie Moghissi, MD, FACP, FACE, can be reached at UCLA David Geffen School of Medicine, University of California, 4644 Lincoln Blvd., Suite 409, Marina del Rey, Los Angeles, CA 90292; email:
Jesse Roth, MD, FACP, can be reached at the Feinstein Institute for Medical Research, 350 Community Drive, Manhasset, NY 11030; email:


One hundred years from now, will there be a need for insulin, or will bariatric surgery and/or pharmacotherapies make it obsolete?


There will always be a need for insulin.

The basic defect of diabetes is either a complete destruction of the beta cells or defect in insulin secretion. We are either going to have perfect cell replacement therapies or small molecules that will totally mimic the action of insulin. Therefore, there will always be a need for insulin.

Gary Lewis

Gary Lewis

It’s clear that bariatric surgery plays a short-term role in that it’s our most effective weight-loss treatment. People with type 2 diabetes can completely abolish type 2 diabetes. Hopefully, in the long run, it will play less and less of a role as we make more discoveries on how to treat obesity with medications and lifestyle modifications. My prediction is that, in the future, we’ll look back at bariatric surgery as something archaic, and it will have no role.

I don’t think pharmacotherapies will completely replace insulin, but I believe as they get better and [there are] more choices, we will have less use for insulin and we’ll be able to keep people off insulin with other therapies for longer time periods.

We’re standing at an exciting time in terms of new discoveries and potential discoveries. I think cell-based, for example, stem cell-based insulin-producing cells is an exciting possibility for the future. Mechanical devices like insulin pumps are getting better and better.

It’s an exciting time and I’m sure there will be genetic therapy within the next 100 years but we can’t even predict all the possible advances even beyond the next 5 or 10 years. So, I believe that it’s going to be an exciting time in diabetes research discovery and, ultimately, it will have huge advantages for the patient.

Gary Lewis, MD, FRCPC, is Director of the Banting and Best Diabetes Center at the University of Toronto. He can be reached at the MaRS Centre at the Toronto Medical Discovery Tower, 10th Floor, Room 203, 101 College St., Toronto, Ontario, Canada M5G 1L7; email: He reports no relevant financial disclosures.


There will always be a need for insulin.

I think it will very much be needed. The question is how will it be delivered in another 10 decades or a century from now? I sense that the science of device development will be much more sophisticated, and I have no ability to declare myself as a prophet, but I sense that in 100 years the insulin-requiring patient with diabetes will be insulin delivered through some mechanical means by which the body senses glucose and its flux, meaning up and down. The current injection of insulin will be history.

Robert H. Eckel

Robert H. Eckel

There’s no question bariatric surgery can take a lot of people with diabetes and restore them to normal levels over a period of time. Patients with more longstanding diabetes who are being insulin treated today do not in fact often have total resolution of diabetes.

There may be some periods of improvement in their glucose control in the absence of insulin during active weight loss, but people who have more advanced diabetes or longer duration to diabetes are not insulin-free after bariatric surgery very often. And those are the people who fail; the people who have longstanding diabetes who are on an insulin treatment. So, cut to the chase, in 100 years, insulin will still be needed but will be delivered very differently from how we deliver it today. It will pretty much be automated.

Robert H. Eckel, MD, is the Charles A. Boettcher II endowed chair in atherosclerosis, professor of medicine with appointments in the division of endocrinology, metabolism and diabetes and the division of cardiology, and professor of physiology and biophysics at the University of Colorado Denver. He can be reached at the University of Colorado Anschutz Medical Center, Research Complex 1 South, 12801 E.17th Ave., Room 7107-8106, Aurora, CO 80045; email: He reports no relevant financial disclosures.


Weight-directed therapy won’t make it obsolete

Weight-directed therapy won’t make it obsolete, but type 2 diabetes has two distinct processes: one is that patients are resistant to insulin most often because they’re overweight but the other is they make less insulin. Limiting obesity would only take care of that part of the equation.

Our approach to obesity now with bariatric surgery is a good first step, but there’s so much about bariatric surgery that we don’t understand. When you look at the hormonal effects of bariatric surgery, it’s possible that in the future, we may find that bariatric surgery was just a means to an end that we can get to in a different way.

We may also discover that gut hormones have huge effects on appetite. Perhaps we could regulate appetite 100 years from now in a way that we can’t even imagine today. It may be possible to make everybody’s set point right where we want it to be.

I think our treatments for overweight and obesity 100 years from now are going to be much more based on the drivers of hunger and satiety. We’ll make much more progress on that front which may just be the trick to reducing the global epidemic.

Robert W. Lash, MD, is professor of internal medicine and chief of staff at the University of Michigan Health System. He can be reached at the University of Michigan Endocrinology & Metabolism, Domino's Farms –Lobby C, 24 Frank Lloyd Wright Dr., Ann Arbor, MI 48105; email: He reports no relevant financial disclosures.