The finding suggests that one inexpensive way to improve the health of patients with type 2 diabetes might be to improve the duration and quality of their sleep.
“Sleep is modifiable,” said Kristen Knutson, research associate (assistant professor) in the department of health studies at the University of Chicago and first author of the paper. “We’ve known for some time that skimping on sleep can impair glucose tolerance even for healthy people. Now we have evidence connecting chronic partial sleep deprivation and reduced blood-sugar control in patients with diabetes.”
“Although we can’t be certain whether sleep loss makes diabetes worse or the diabetes interferes with sleep, it only makes sense for everyone, but especially patients with diabetes, to give themselves the opportunity to get enough sleep,” Knutson said.
The study focused on 161 African-American patients being treated at the University of Chicago Hospitals for type 2 diabetes. The researchers asked participants how much sleep they thought they needed at night and how much sleep they managed to get on weeknights and weekends. They also assessed the quality of their sleep using a standard 19-item questionnaire, the Pittsburgh Sleep Quality Index (PSQI).
To assess blood sugar control they measured glycosylated hemoglobin, a standard tool for management of patients with diabetes. Glycosylated hemoglobin, or HbA1c, reflects the average blood glucose level over the previous three months. A normal HbA1c result is between four and six percent. Higher levels represent poor glucose control. Patients with diabetes are considered to be under good control if they can keep their levels below seven percent.
The researchers found that, on average, the 161 diabetes patients got very little sleep and had poor glucose control. Mean sleep duration was six hours a night. Only six percent reported getting eight hours of sleep on weeknights and only 22 percent reported getting at least seven hours. Seventy-one percent had poor sleep quality. The median HbA1c score was 8.3 percent.
Many patients with diabetes have painful complications that can interfere with sleep. Even after the researchers excluded 39 patients who reported such pain, however, two out of three of the remaining 122 patients reported poor quality sleep. The average HbA1c among those patients was almost as high: 8.2 percent.
Insufficient or poor quality sleep was closely associated with higher HbA1c results. For patients with no complications of their diabetes, a three-hour “perceived sleep debt”—the difference between how much sleep they felt they needed and how much they think they got—was associated with a 1.1 percentage-point increase in HbA1c levels, for example from 7.5 percent up to 8.6 percent.
For patients with at least one complication of diabetes—such as nerve pain, kidney damage or coronary artery disease—decreased sleep quality appeared to be more important. An increase of five points (out of 21) on the PSQI was associated with a 1.9 percentage-point increase in HbA1c, for example from 8.7 percent up to 10.6 percent.
“The magnitude of these effects,” the authors note, “is comparable to those of widely used oral antidiabetic agents.”
A long series of laboratory and epidemiologic studies has suggested that cutting back on sleep has a harmful effect on glucose control, insulin secretion and metabolism in ways that might increase diabetes risk, said Eve Van Cauter, professor of medicine at the University of Chicago and senior author of the study. The current study asks the question: is glucose control in subjects who already have diabetes adversely affected by too little sleep or poor sleep?
"Our findings suggest, at least in this study population, that short or poor sleep is associated with decreased blood-sugar control in patients who already have diabetes," she said. "The growing tendency to burn the candle at both ends may be a significant contributor to the current epidemic of diabetes. One way to slow down this epidemic may be to avoid building a chronic sleep debt."
The MacArthur Foundation, the American Diabetes Association and the National Institutes of Health funded this study. Additional authors are Armand Ryden, of the University of Chicago, and Bryce Mander, now at Northwestern University.