85
turbances were a symptom or a consequence of Alz
heimer’s. They assumed that as clumps of amyloid
proteins built up, then started to strangle and kill
nerve cells—particularly in the memory regions of
the brain—changes in sleep followed. Even older
people without Alzheimer’s can experience changes
in their sleep patterns, sleeping less and more lightly
as they age. So experts didn’t initially take these
shorter and more fragmented sleep cycles seriously.
But in the 1980s and 1990s, scientists began
studying whether there was any causal relationship
between sleep patterns and cognitive test perfor
mance among older people without Alzheimer’s by
studying them over longer periods of time. Those
studies suggested that people with poor sleep habits
tended to perform worse on cognitive tests over time.
“That got people thinking about the possibility that
sleep could be a risk factor in dementia,” says Spira.
Yaffe’s recent research, which focused on a group
of healthy older women, supported the idea that
what seemed to matter, in terms of dementia risk,
was the quality as opposed to the quantity of sleep.
Those who reported spending less time in bed actu
ally sleeping, and more time tossing and turning and
waking up throughout the night, were more likely to
develop any type of dementia five to 10 years later
than those who got betterquality sleep.
Others focused on explaining the biology behind
the sleepdementia connection. At this point,
Alzheimer’s researchers knew that a buildup in the
brain of amyloid and another protein called tau were
key features of the disease. At Washington Univer
sity School of Medicine in St. Louis, David Holtzman,
chair of the department of neurology, launched stud
ies to track exactly where in the brain this amyloid
originated. His search led him to nerve cells, which
release fragments of the protein as they go about
their normal business. Typically, these protein
byproducts (sometimes called amyloid beta) are re
leased into the circulatory system, where they float
around without causing problems, but in some cases
they remain in the brain, where they morph into a
sort of molecular Velcro, sticking together to form
amyloid plaques, which in turn damage neurons.
But what controls the production of amy
loid beta? In a 2009 study on mice, Holtzman
found that while the animals were awake, levels
of the protein fragments circulating in their brains
surged. When the mice slept, the levels dropped
dramatically— especially during the deeper stages
of nonREM sleep. And when he and his team de
prived the mice of nonREM sleep, more amyloid
built up in their brains over time than in mice who
got regular nightly rest. He saw similar changes when
he compared amyloid in the spinal fluid of people
who were well rested vs. sleepdeprived.