Charlene Laino

June 18, 2003

June 18, 2003 (New Orleans) — Drinking four or more cups of coffee a day appears to be associated with a lower risk of type 2 diabetes, Harvard researchers report.

It's not the first time that a study has suggested that the popular brew may protect against the disorder that affects at least 17 million Americans. Last year, scientists at Vrije University in Amsterdam reported that heavy coffee drinkers are half as likely to develop diabetes as people who consume two cups or less a day.

But that work set off a major controversy, motivating the Harvard researchers to try to replicate the findings in the 100,000-plus men and women whose health they have been following for about two decades in the Health Professionals Follow-up Study and the Nurses' Health Study.

Frank Hu, MD, PhD, associate professor of nutrition and epidemiology at the Harvard School of Public Health, talked about the research here Tuesday at the American Diabetes Association 63rd Scientific Sessions.

"If it was really true, maybe we could just forget about exercise and sit in front of the TV and sip coffee all day," he told colleagues in jest.

In all seriousness, there was every reason for disbelief, Dr. Hu said. Studies suggest that acute administration of caffeine decreases insulin sensitivity and impairs glucose tolerance.

On the other hand, caffeine stimulates thermogenesis and increases energy expenditure, he said. "This could decrease obesity rates and thus diabetes."

One problem, he noted, is that tolerance can develop after several days of caffeine use, making it very difficult to extrapolate from short-term to long-term studies.

Also, caffeine might be coffee's best known ingredient, but it's not the only one, Dr. Hu said. There are substantial amounts of magnesium, niacin, potassium, and even such antioxidants as tocopherol, he said. To try to tease out the effects of caffeine, as opposed to some of these other substances, Dr. Hu's team looked at rates of diabetes in not only coffee drinkers, but also in tea and decaffeinated coffee drinkers. The Dutch researchers didn't do that.

"Decaf has the same amount of these other substances, but less caffeine," Dr. Hu said. "Tea has other substances and is relatively low in caffeine."

The analysis was based on data collected from 42,888 male healthcare professionals from 1986 to 1998 and 85,056 female nurses from 1980 to 1998. All of the men and women were free of diabetes, cancer, and cardiovascular disease at baseline.

Every two to four years, the participants filled out validated dietary questionnaires that included items on coffee consumption.

Over the course of the study, 1,333 men and 4,085 women developed type 2 diabetes, Dr. Hu said. Using a Cox proportional hazard model that adjusted for age, body mass index, smoking, and other diabetes risk factors as well as for fiber and fat intake, the researchers found:

  • Men who drink six or more cups of coffee a day were less than half as likely to develop diabetes compared with nondrinkers. Drinking four to six cups helps too: Those men saw their risk cut by 29%.

  • Women who consume four cups or more a day also reduced their risk of developing the disease by about 30%. But in their case, six cups did not seem to be any more protective than four cups, Dr. Hu said. Tea, on the other hand, had no impact on diabetes risk, the study showed.

  • Decaf was associated with a "modest reduction" in risk for those who drank four cups or more a day, he said. But those results were not controlled for the amount of regular coffee that was consumed.

The researchers then looked at total caffeine intake from coffee, colas, and other foods, and found that it too appeared to protect against the development of diabetes. Men and women in the highest quintile of caffeine intake were 22% and 30% less likely to develop diabetes, respectively, compared with those in the lowest quintile.

"The strengths of the study include large cohorts, long follow-ups and repeated and validated measures of diet and lifestyle," Dr. Hu said. "Limitations include self-reporting, but our validation studies suggest the data is good."

And by eliminating individuals with any type of major disease at baseline, the study addresses the potential confounding effect of pre-existing conditions on diabetes risk, he added.

The new findings have made a believer out of at least one former skeptic, but a key one — Terry E. Graham, PhD, who presented a study showing that a dose of caffeine equivalent to that contained in two strong cups of coffee reduced insulin sensitivity and increased blood glucose levels in 12 men with type 2 diabetes.

"The findings completely contradict everything we have shown in acute studies of caffeine," he said. "When the Dutch study came out, I was shocked. But now with this second study that shows the same thing, you start to believe it," said Dr. Graham, chair of human biology and nutritional sciences at the University of Guelph in Ontario, Canada.

The Harvard study is more thorough, he added, with more accurate data about the participants' coffee habits over 10 to 15 years. "Plus, they evaluated tea and decaf, which the first study didn't."

Some clues into why the short-term and long-term studies arrive at such different results might come soon, he said. Vanderbilt University researchers are studying compounds produced when coffee beans are roasted that appear to counter some of caffeine's effects.

In the meantime, Dr. Hu said that he is not yet ready to recommend coffee for diabetes prevention. "We still need more study," he said.

But for the diabetic patients who ask if coffee is detrimental, Dr Hu said, "I don't think they have to worry."

ADA 63rd Scientific Sessions: Abstracts 307, 308. Presented June 17, 2003.

Reviewed by Gary D. Vogin, MD


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: