One of the greatest advancements in diabetes treatment started out as an oddity. “I said to myself—what is this? This isn’t fitting with any of the known hemoglobins,” said Samuel Rahbar, MD, PhD, in an interview last July, recalling his first impressions of the hemoglobin A1C, a protein in human blood.
Rahbar died last November at the age of 83, but he will be remembered as the man who discovered that diabetes can raise blood levels of hemoglobin A1C, which is arguably one of the most important biological molecules in modern medicine—although it wasn’t immediately recognized as such a big deal. Since Rahbar’s A1C discovery in 1968, the protein has traveled a winding road to attain its current place in the center of diabetes medicine.
Here is a brief history of the A1C molecule, the star of the valuable blood test that, in a single simple percentage, offers a look at average blood glucose levels over the past two to three months.
Samuel Rahbar recognizes that hemoglobin A1C is elevated in people with diabetes.
Anthony Cerami and colleagues suggest that A1C is related to blood glucose control.
The American Diabetes Association first recommends A1C tests.
The Diabetes Control and Complications Trial establishes A1C as a valuable clinical marker in people with type 1 diabetes.
The United Kingdom Prospective Diabetes Study establishes A1C as a valuable clinical marker in people with type 2 diabetes.
The American Diabetes Association recommends using the A1C to diagnose diabetes and prediabetes.
In the 1960s, hemoglobin was all the rage in biological research. It “was the molecule du jour,” Rahbar recalled. The hemoglobin protein is a long chain of amino acids that folds upon itself to form a three-dimensional structure capable of performing a specific task. Hemoglobin’s job is to carry oxygen from the lungs through the blood to all the tissues in the body so cells can do their version of breathing. This oxygen mule makes up 97 percent of red blood cells’ dry weight, so even in the early days of molecular biology, when most cellular proteins were hard to come by, scientists could score large amounts of hemoglobin by culling the protein from blood samples.
Soon, researchers discovered that there is more than just one type of hemoglobin and that some hemoglobins are linked to disease. Sickle cell anemia, for example, is caused by a mutation in the hemoglobin gene that results in a malformed protein. The discovery of sickle cell hemoglobin, named hemoglobin S, set off a race to find additional versions of hemoglobin related to human disease.
Rahbar, working in Iran at the time, joined the race with gusto. “At 6 o’clock in the morning, someone would go with a motorcycle to pick up … small tubes of blood from the [Tehran University Hospitals]. I used to take their discarded blood samples,” he said. “I was screening 300 blood samples a day, and the lab was running like a factory.” Rahbar’s method of choice for seeking rare versions of hemoglobin was electrophoresis. That is a means of chemical separation that can tease out minor populations of hemoglobin from hemoglobin A, the most abundant type of hemoglobin in humans. “Now we know there are hundreds and hundreds of these variants,” says Anthony Cerami, MD, PhD, who is credited with developing the A1C blood test as a clinical tool. He is the founder and chairman of the board of two companies, Warren Pharmaceuticals and Araim Pharmaceuticals.
One day, Rahbar was screening a sample of blood and
detected a type of hemoglobin that he hadn’t seen before. Looking at
the medical records of the patient from whom the sample had been taken,
he noticed that the person had diabetes. And so Rahbar had his great
idea: Perhaps this mysterious hemoglobin was related to diabetes. To
test the hypothesis, Rahbar next studied the blood of 47 more people
with diabetes. “I will always remember: It was a weekend, Friday, and I
went [to the lab] and screened all of them. They all showed the same
hemoglobin,” said Rahbar. “I called it the diabetic component of
hemoglobin.” It would take additional research before the diabetic hemoglobin got its name.
What Is It?
At the same time Rahbar was doing his experiments, scientists around the world were studying hemoglobin using a different separation method called chromatography. They soon discovered several subtypes of hemoglobin A. “They could see a big peak of hemoglobin A, and all these little peaks,” says Cerami. In those little peaks were five new hemoglobins named A1a, A1b, A1c, A1d, and A1e, based on the order in which they emerged from the chromatograph.
As Rahbar scanned the scientific literature in search of answers, he decided to try chromatography on his blood samples. He soon realized that his diabetic hemoglobin appeared to have properties similar to hemoglobin A1C’s. What’s more, Rahbar found that the A1C molecule made up from 7.5 to 10.6 percent of the total hemoglobin in people with diabetes, while it constituted only 4 to 6 percent of hemoglobin in those without the disease, again establishing the link between the A1C molecule and diabetes.
Chemists began picking apart the hemoglobin A1C to figure out how, chemically, it differed from plain hemoglobin A. They discovered that the A1C molecule was essentially hemoglobin A—they have the same sequence of amino acids—but with one critical difference: a glucose molecule stuck to one end. Still, no one yet knew how the glucose got there and whether hemoglobin A1C tracked with blood glucose levels or simply was higher in people with diabetes. “For the first five to six years, no one believed this was something interesting,” said Rahbar, “but it turned out to be important.”
Very Important Protein
In the years that followed, Cerami ran a series of experiments in both animals and humans to help establish hemoglobin A1C’s role in diabetes. He eventually proved that the number of hemoglobin A proteins that become tethered to a glucose molecule, forming the A1C molecule, is proportional to the concentration of glucose in the blood. One study found that A1C levels mirrored urine glucose levels in people, suggesting that the molecule might offer a new opportunity to assess blood glucose control in people with diabetes.
“Clinicians at this time had no idea of what the control was of their patients,” says Cerami, because they based their assessment on measures like urine or blood glucose, which could change radically from day to day and hour to hour. The A1C was a game changer because it reported on the average blood glucose over two to three months (because the lifetime of a red blood cell is about two to three months). Before the A1C test, patients could eat well and exercise for two days before an exam and get a stellar blood glucose result, says David Sacks, MB ChB, FRCPath, chemist at the National Institutes of Health Clinical Center. Things are different now. “You can’t cheat on your hemoglobin A1C,” he says.
In the late 1970s, companies developed the first commercial tests of hemoglobin A1C, but clinicians were reluctant to use them because there was no proof yet that an A1C test could help improve patient health. “What really put hemoglobin A1C on the map was the DCCT,” says Sacks, referring to the Diabetes Control and Complications Trial, whose results were published in 1993. Researchers recorded the A1Cs, blood glucose measurements, and presence of diabetic complications in over 1,400 people with type 1 diabetes for up to 10 years. Participants who were assigned to receive intensive treatment to keep their blood glucose levels as close to normal as possible, as measured by the A1C, had fewer eye, nerve, and kidney diseases than people with higher levels. “A1C was much, much better than self-monitoring” at predicting who was at high risk for diabetes complications, says Sacks. That forever cemented the A1C as the gold-standard marker of long-term health in people with type 1 diabetes.
A few years later, in 1998, researchers published the results of the United Kingdom Prospective Diabetes Study, which was similar to the DCCT but included people with type 2 diabetes. Again, the A1C proved itself. The two studies led to the recommendation that people with diabetes maintain an A1C of 7 percent or less. In recent years, as more data on the costs and benefits of tight control have emerged, the emphasis has shifted from a hard A1C target to an individualized goal. Today, experts recommend that people with diabetes have an A1C blood test two to four times a year. Doctors and patients use the number to make decisions about medication selections and dosages and to assess the effects of lifestyle changes on blood glucose levels.
For a long time, scientists didn’t know for sure whether the A1C was a true measure of the average blood glucose level. There simply wasn’t a good way to test the hypothesis. Blood glucose varies a lot over the course of a day, and even testing with a meter several times a day doesn’t provide a complete picture. Then came continuous glucose monitors (CGMs), which test glucose levels every few minutes. Researchers finally had a way to see if A1C coincided with average blood glucose. In a 2008 study, 507 people, some with type 1 or type 2 diabetes and others without the disease, wore CGMs for three months. The researchers used the 2,700 glucose measurements they obtained from each participant during that time to calculate his or her average blood glucose, then compared the value to the measured A1Cs. The average glucose levels aligned with the A1Cs, showing the A1C test’s validity. Now, A1C test results can be translated in terms of an estimated average glucose (chart, below).
|The A1C is a measure of average blood glucose over the previous two to three months. The chart below gives the estimated average glucose for a given A1C, in the same units seen on a blood glucose meter. For an online calculator, go to diabetes.org/eag.|
|Estimated Average Glucose (mg/dl)|
Another recent development in the A1C test as a clinical tool is that experts now recommend using it to diagnose diabetes and prediabetes. An A1C of below 5.7 percent is considered normal, while A1C levels of 6.5 percent and above call for a diagnosis of diabetes. Levels between those two percentages signal prediabetes.
The A1C test used in the DCCT study has been the basis for almost all clinical measurements of the A1C since then. That’s a good thing, says Sacks, because “a sample run in Boise should be the same as in [Washington] D.C.” says Sacks. “Seven percent is 7 percent.”
Well, actually, 7 percent may not be 7 percent for long in some countries. The reason is that after the DCCT, some chemists developed a more accurate way to measure the hemoglobin A1C in the blood than the method used in the famous trial. “These are people who believe everything has to be as accurate as possible and exactly right,” says Sacks, who argues that such a change is clinically meaningless and potentially harmful to patients. The new test shows that A1Cs are actually from 1.5 to 2 percentage points lower than shown in the DCCT test. That could cause confusion among people who have long gotten A1C results from traditional testing. That is why the United States is sticking with the old percentage levels. Sacks, who was involved in the decision not to change the test, says: “I really only had one motive—that the patients shouldn’t suffer.”
The A1C has very likely prevented much suffering over the decades, as people with diabetes and their doctors have learned to use this tool to prevent complications and improve health. In June 2012, Rahbar accepted the Samuel Rahbar Outstanding Discovery Award from the American Diabetes Association. “I was very excited,” said Rahbar, mostly because he was honored in front of his wife, children, and grandchildren. Rahbar continued to work most days at the City of Hope National Medical Center in Duarte, Calif., waking at 5 a.m. to be in the laboratory by 5:30, until his death in November. The legacy of his work will remain a fixture in the lives of people with diabetes for the foreseeable future.