Study helps explain why sleep becomes more fragmented with age

As people grow older, they often have difficulty falling asleep and staying asleep, and tend to awaken too early in the morning. In individuals with Alzheimer's disease, this common and troubling symptom of aging tends to be especially pronounced, often leading to nighttime confusion and wandering.

Now, a study led by researchers at Beth Israel Deaconess Medical Center (BIDMC) and the University of Toronto/Sunnybrook Health Sciences Center helps explain why sleep becomes more fragmented with age. Reported online today in the journal Brain, the new findings demonstrate for the first time that a group of inhibitory neurons, whose loss leads to sleep disruption in experimental animals, are substantially diminished among the elderly and individuals with Alzheimer's disease, and that this, in turn, is accompanied by sleep disruption.

"On average, a person in his 70s has about one hour less sleep per night than a person in his 20s," explains senior author Clifford B. Saper, MD, PhD, Chairman of Neurology at BIDMC and James Jackson Putnam Professor of Neurology at Harvard Medical School. "Sleep loss and sleep fragmentation is associated with a number of health issues, including cognitive dysfunction, increased blood pressure and vascular disease, and a tendency to develop type 2 diabetes. It now appears that loss of these neurons may be contributing to these various disorders as people age."

In 1996, the Saper lab first discovered that the ventrolateral preoptic nucleus, a key cell group of inhibitory neurons, was functioning as a "sleep switch" in rats, turning off the brain's arousal systems to enable animals to fall asleep. "Our experiments in animals showed that loss of these neurons produced profound insomnia, with animals sleeping only about 50 percent as much as normal and their remaining sleep being fragmented and disrupted," he explains.

A group of cells in the human brain, the intermediate nucleus, is located in a similar location and has the same inhibitory neurotransmitter, galanin, as the vetrolateral preoptic nucleus in rats. The authors hypothesized that if the intermediate nucleus was important for human sleep and was homologous to the animal's ventrolateral preoptic nucleus, then it may also similarly regulate humans' sleep-wake cycles.

In order to test this hypothesis, the investigators analyzed data from the Rush Memory and Aging Project, a community-based study of aging and dementia which began in 1997 and has been following a group of almost 1,000 subjects who entered the study as healthy 65-year-olds and are followed until their deaths, at which point their brains are donated for research.

"Since 2005, most of the subjects in the Memory and Aging Project have been undergoing actigraphic recording every two years. This consists of their wearing a small wristwatch-type device on their non-dominant arm for seven to 10 days," explains first author Andrew S. P. Lim, MD, of the University of Toronto and Sunnybrook Health Sciences Center and formerly a member of the Saper lab. The actigraphy device, which is waterproof, is worn 24 hours a day and thereby monitors all movements, large and small, divided into 15-second intervals. "Our previous work had determined that these actigraphic recordings are a good measure of the amount and quality of sleep," adds Lim.

The authors examined the brains of 45 study subjects (median age at death, 89.2), identifying ventrolateral preoptic neurons by staining the brains for the neurotransmitter galanin. They then correlated the actigraphic rest-activity behavior of the 45 individuals in the year prior to their deaths with the number of remaining ventrolateral preoptic neurons at autopsy.

"We found that in the older patients who did not have Alzheimer's disease, the number of ventrolateral preoptic neurons correlated inversely with the amount of sleep fragmentation," says Saper. "The fewer the neurons, the more fragmented the sleep became." The subjects with the largest amount of neurons (greater than 6,000) spent 50 percent or more of total rest time in the prolonged periods of non-movement most likely to represent sleep while subjects with the fewest ventrolateral preoptic neurons (less than 3,000) spent less than 40 percent of total rest time in extended periods of rest. The results further showed that among Alzheimer's patients, most sleep impairment seemed to be related to the number of ventrolateral preoptic neurons that had been lost.

"These findings provide the first evidence that the ventrolateral preoptic nucleus in humans probably plays a key role in causing sleep, and functions in a similar way to other species that have been studied," says Saper. "The loss of these neurons with aging and with Alzheimer's disease may be an important reason why older individuals often face sleep disruptions. These results may, therefore, lead to new methods to diminish sleep problems in the elderly and prevent sleep-deprivation-related cognitive decline in people with dementia."