All About Scientific Research Studies
From hypothesis to trial, here’s how research happens—and how it affects your health
When David Nathan, MD, first started working as an endocrinologist in the 1970s, insulin had been in use for more than 50 years. Type 1 diabetes had been transformed from a lethal disease to a treatable condition.
Yet diabetes management was still an imprecise affair. Most people relied on urine tests to check their glucose levels and injected insulin once or twice a day. “People took insulin to make sure they weren’t spilling glucose into the urine,” recalls Nathan, now the director of the Diabetes Center and Clinical Research Center at Massachusetts General Hospital in Boston. “Measurement devices, let alone insulin pumps, weren’t available.”
As people with type 1 aged, doctors started seeing serious complications in people who had lived with the disease for 15 or 20 years. Amputations, kidney disease, and heart attacks were all frightening facts of life. Retinopathy was so common that the waiting room of the Boston clinic where Nathan worked had an area for seeing-eye dogs.
Even though any endocrinologist could see that people with type 1 had more complications than people without the disease, “there was no clear understanding as to how you should treat diabetes and what were appropriate glucose goals,” Nathan says. “There was a hypothesis that if you controlled glucose you’d prevent complications, but no one knew for sure.”
To find out, the National Institutes of Health (NIH)—the federal government’s medical research agency—funded one of the most ambitious, expensive studies it had ever taken on: the Diabetes Control and Complications Trial (DCCT). It was designed to test the idea that intensive glucose management (keeping levels as close to nondiabetic levels as possible) would reduce the risk of complications.
Researchers divided over a thousand patients into two groups: One got what was considered standard care at the time. The other was asked to keep their blood glucose tightly managed—no easy task at a time when checking blood glucose required clumsy equipment and a large drop of blood. The two groups were monitored for more than six years to see if those with tightly controlled blood glucose had fewer complications than the standard-care “control” group.
When the results of the DCCT were announced in 1993, the world changed for people with diabetes. It was conclusively proven that intensive therapy reduced long-term complications in people with type 1 by up to 76 percent. “We proved the hypothesis and changed the standard of care. Complications have demonstrably fallen,” Nathan says. “The study was extremely expensive, but it was inexpensive in that the frequency of complications, which cause human suffering and are extremely costly, began to decrease. It was a grand slam for the NIH. And we rarely have seeing-eye dogs in the office.”
The Path to Discovery
If the DCCT was a grand slam, most studies are base hits—or strikeouts. Research into new drugs, devices, and methods to care for diabetes and other diseases is a long, slow process, sometimes taking decades and drawing on the work of hundreds or thousands
of researchers around the world. Most studies begin in a petri dish, or in mice, and never get much further than that.
News headlines can give a different impression, bombarding people with a steady stream of dramatic claims and conflicting recommendations (memorably, coffee has been both blamed for increasing heart disease risk and bad cholesterol and credited with reducing strokes and preventing type 2 diabetes). Making sense of it all can be confusing for patients, who are constantly reading about the results of studies that promise to change the course of their disease—someday. The fact is, there’s a long path from the laboratory to the patient.
Along the way, there’s a need for different kinds of studies. Most people think immediately of intervention trials like the DCCT, where the effects of a treatment or drug are measured in a large group of people. But to get there can take decades of different approaches: mice and other animal models, broad analyses of health data that reveal differences in disease risk, and smaller observational studies (see “Scientific Terms to Know,” link below) that pave the way for large-scale efforts. “Observational studies are extremely important,” says Barbara Linder, MD, PhD, program director in the Division of Diabetes, Endocrinology, and Metabolic Diseases at the National Institute of Diabetes and Digestive and Kidney Diseases. “They help us determine the rates of disease and the distribution of illness in populations.”
But they have their limitations. For instance, imagine a study that links health records to diet surveys and shows that people who eat blueberries have better heart health. So-called observational studies can’t prove blueberries were the cause. “They can only tell us about association, not cause and effect,” Linder says. Perhaps blueberry eaters are healthier overall, or some other aspect of their lifestyles, other than blueberry-eating, led to the lower heart risk. (When reading about new research, terms such as “linked to” and “associated with” are clues that cause and effect wasn’t proven.)
To prove cause and effect, a study needs to show that an effect (better heart health) is directly caused by something (blueberries). That could mean giving one group blueberries for years, and then comparing their heart disease rates to a group that didn’t eat any blueberries, all while making sure other behaviors that might affect heart health (such as exercise and smoking) were the same between the two groups.
Small pilot studies might show a relationship, but small numbers make their conclusions less convincing: Without a lot of data, it’s hard to tell if a result is a fluke or not. “When you get something on a very small scale, before you change practice you want it to be repeated,” Linder says.
The DCCT is an example of a successful intervention study that confirmed a theory suggested by earlier observational studies: that people with intensive diabetes management aimed at lower blood glucose targets have fewer complications than those not on intensive therapy. Prior to the DCCT, researchers “had looked at a large population and seen that people with higher blood sugar did worse,” Nathan says. “That’s an association, not a causation. To prove an intervention will work, you have to do the experiment.”
Expect the Unexpected
Sometimes such large-scale experiments come up with unexpected results. Take, for example, vitamin E. Or rather, Linder says, don’t. After observational studies suggested that men who took vitamin E supplements had a lower risk of prostate cancer, the NIH spent over $100 million on a clinical trial to see if taking vitamin E regularly prevented prostate cancer.
Begun in 2001, the Selenium and Vitamin E Cancer Prevention Trial (SELECT) was one of the largest cancer prevention trials ever conducted. After looking at over 35,000 men assigned to take vitamin E, selenium, or a placebo, researchers concluded it didn’t matter, at least as far as cancer was concerned. “The trials didn’t show any decrease in cancer risk,” Linder says.
The best research has to account for the fact that medicine is more than statistics. Everything from genes to behavior can vary from person to person, potentially impacting their health. Consider the Diabetes Prevention Program (DPP). This landmark type 2 diabetes study separated participants into three groups: One got normal care (they were given general information about a healthy diet and exercise program with no other intervention), another was given metformin, and a third participated in intensive lifestyle modification that encouraged them to lose weight by changing their diet and exercising more.
The DPP showed that weight loss and lifestyle changes made a massive difference in reducing type 2 diabetes risk by 58 percent. It was a turning point in diabetes care, prompting governments, health insurance companies, and hospitals to change the way they treated people at high risk for the disease.
But 40 percent of people in the study still developed type 2 diabetes. “Whatever the science is, we always believe patients need individualized treatment,” Linder says. “Patients have to remember human beings are complicated. Diseases are not all the same. Diabetes is not the same in every patient. A given patient may respond differently to medications than another patient.”
That’s why large clinical trials remain the gold standard for policy decisions. “To establish cause, you have to go to trials,” says Linder. “Randomly assigning people to different treatments helps ensure that any difference in outcome is the result of the treatment.”
That very question was the origin of the DPP trial, the one that showed losing weight via lifestyle changes was so effective at reducing diabetes risk. Nathan was involved in that study, too, this time as its head. “We showed, across all these populations, that lifestyle interventions worked,” he says. “Programs that imitated the DPP lifestyle intervention popped up all over the place.”
And when it comes to your health, a clear outcome is the best outcome researchers could hope for.
How to Join a Study
From tiny pilot studies to large government-funded trials, medical research depends on volunteers. “The lifeblood of our studies is our participants,” says David Nathan, MD, director of the Diabetes Center and Clinical Research Center at Massachusetts General Hospital in Boston. “This is a team sport, and the most important players are the participants. Without them, studies like [the Diabetes Control and Complications Trial] can’t happen.”
Many researchers draw participants from clinics or organizations in their area. If you’re interested in participating, ask your health care providers if they know of anyone who’s recruiting. For information on larger studies, check out clinicaltrials.gov.
Before signing up, make sure you understand how the trial will work, the goal of the study, and the potential benefits and risks compared with the treatment you’re already getting. Be sure you can handle the time commitment. Head here for more information.
Diabetes Clinical Trial Hall of Fame
Much of your doctors’ advice comes from a handful of landmark clinical trials. Here are three of the biggest.
Diabetes Control and Complications Trial
Outcome: Intensive blood glucose management reduced the likelihood of eye, kidney, and nerve complications for people with type 1 diabetes by up to 76 percent.
Lasting Impact: People with type 1 know to keep their blood glucose as close to normal as safely possible, beginning soon after diagnosis.
United Kingdom Prospective Diabetes Study
Outcome: Intensive glucose management reduced complications in people with type 2 diabetes.
Lasting Impact: People with type 2 can benefit from tight glucose management.
Diabetes Prevention Program
Outcome: By losing weight via diet and exercise, people at high risk for type 2 diabetes lowered their risk of developing the disease by 58 percent over about three years. Metformin reduced the risk by 31 percent.
Lasting Impact: Weight loss and the consideration of metformin are now standard recommendations for preventing or delaying the onset of type 2 diabetes. The study helped inspire the National Diabetes Prevention Program, which helps people at risk for type 2 diabetes make lifestyle changes to reduce their risk for the disease and improve their health.