Why the hippocampus and posterior cingulate are the places to watch
Alzheimer's disease does not shrink the whole brain evenly. It tends to strike particular regions first, and two of the best-known are the hippocampus, a seahorse-shaped structure deep in the brain that is central to forming new memories, and the posterior cingulate cortex, a region toward the back-middle of the brain that is part of a network active when the mind is at rest. Because these areas change early, they are natural places to look when asking whether a lifestyle factor like sleep is nudging the brain toward Alzheimer's-type change.
The most direct test of this question followed cognitively normal older adults and used MRI to track gray matter volume while measuring sleep quality with the Pittsburgh Sleep Quality Index, a widely used questionnaire that scores how poor someone's sleep has been [1]. The finding was that poorer sleep went with faster volume loss in the hippocampus and posterior cingulate cortex specifically, not just the brain in general [1]. That regional specificity is what makes the result interesting: it lines up with where Alzheimer's pathology tends to begin.
It may not be sleep "quality" in general but particular sleep stages
Questionnaire-based sleep quality is a broad summary. Studies that instead record brain activity during sleep with polysomnography, which measures electrical activity across the night, point to specific stages. In 712 pre-symptomatic, healthy adults, a larger share of slow-wave sleep, the deep sleep stage thought to be important for clearing waste and consolidating memory, was associated with a lower Alzheimer's-related atrophy score and with thicker gray matter in the cortical regions that Alzheimer's typically affects [3]. That was a cross-sectional finding, meaning sleep and brain structure were measured at roughly the same time, so it shows an association rather than proving that sleep loss causes shrinkage.
The same study found no such relationship for REM sleep or for the lighter NREM stage 2 sleep, and it even found that more NREM stage 1 sleep went with thicker gray matter, mainly in prefrontal regions [3]. That mix of results is a useful caution: "poor sleep" is not one thing, and different stages may relate to the brain in different directions.
REM sleep shows up too, but the direction of cause is unclear
A more recent study used at-home polysomnography in 88 older adults, 68 of them cognitively normal and 20 with cognitive impairment, and looked at how much of the night was spent in REM sleep [2]. Lower REM percentage and longer REM latency, meaning it took longer to reach REM, were associated with lower scores on a cognitive test and with smaller hippocampal volume at the time of measurement [2]. Over time, lower REM percentage and longer REM latency were also associated with greater decline in cognition and hippocampal volume, along with rising amyloid and tau signals on PET scans [2].
The authors interpret this carefully: they suggest that the accumulation of Alzheimer's pathology over time may itself be affecting REM sleep patterns, rather than sleep loss simply causing the atrophy [2]. In other words, the arrow of cause could run partly backward. Because the cohort mixed cognitively normal and impaired people, these REM findings describe the group as a whole rather than cognitively normal adults alone.
What this means if you are a healthy older adult wondering about your sleep
The strongest support for the original question comes from the longitudinal questionnaire study, which found that poor sleep predicted faster shrinkage of the hippocampus and posterior cingulate in people who were cognitively normal at the start [1]. The polysomnography studies add that the story is probably about specific sleep stages, especially slow-wave and REM sleep, rather than a single global "sleep quality" number [2][3]. A study that included people with mild cognitive impairment and Alzheimer's disease also reported that sleep deprivation correlated with atrophy of the posterior cingulate cortex and hippocampus, but because it mixed diagnostic groups and was small, it mainly reinforces the pattern rather than settling it [4].
One further study looked at how sleep disturbance relates to the coordinated structure of large brain networks across the Alzheimer's continuum, and found that people with Alzheimer's and poor sleep showed different network patterns than those with normal sleep [5]. That work included people at many stages, not just cognitively normal adults, so it speaks to how sleep and brain organization interact as disease progresses rather than to the earliest phase alone. Taken together, the evidence is consistent with poor sleep being a marker or contributor to early Alzheimer's-related brain change in healthy older adults, but the studies differ in how they measure sleep and in whom they study, and at least one line of evidence suggests the relationship may run in both directions [2].
Sources used in this answer
Poor sleep accelerates hippocampal and posterior cingulate volume loss in cognitively normal healthy older adults
In cognitively normal older adults, poor sleep quality measured by questionnaire was associated with accelerated gray matter volume loss in the hippocampus and posterior cingulate cortex.
REM Sleep is Associated with Cognition and AD Biomarkers Longitudinally in Cognitively Normal and Impaired Older Adults
In a mixed group of cognitively normal and impaired older adults, lower REM sleep percentage and longer REM latency were associated with lower cognition and smaller hippocampal volume, and with greater decline in both over time, alongside rising amyloid and tau.
The relationship between Alzheimer's‐related brain atrophy patterns and sleep macro‐architecture
In 712 pre-symptomatic healthy adults, a higher proportion of slow-wave sleep was cross-sectionally associated with a lower Alzheimer's-related atrophy score and thicker gray matter in Alzheimer's-signature cortical regions, with no such effect for REM or NREM stage 2 sleep.
Effects of sleep deprivation on brain atrophy in individuals with mild cognitive impairment and Alzheimer's disease
In a small six-month study including people with mild cognitive impairment or Alzheimer's disease and healthy older adults, sleep deprivation correlated with posterior cingulate and hippocampal atrophy and lower total brain volume.
Impact of sleep disruptions on gray matter structural covariance networks across the Alzheimer's disease continuum
Across the Alzheimer's continuum, people with Alzheimer's and poor sleep showed distinct patterns of gray matter structural covariance network damage compared with normal sleepers.