Sleep changes in dementia follow a rough but predictable arc. In the early stages, most people notice difficulty falling asleep, frequent nighttime awakenings, and creeping daytime naps — with as many as 70% of patients in early-stage dementia experiencing some form of sleep disturbance. By the moderate stage, sundowning and fragmented sleep-wake cycles take hold, and the internal body clock begins to lose its grip on day and night. In severe dementia, the pattern can nearly reverse itself, with individuals spending about 40% of nighttime hours awake while sleeping much of the day. The progression is not identical for everyone — Lewy body dementia, for instance, causes far more severe sleep disruption than Alzheimer’s at every stage — but the general trajectory moves from mild insomnia toward a near-total breakdown of circadian rhythm.
Consider a woman in her early seventies who starts waking at 3 a.m. several nights a week, months before anyone suspects cognitive decline. Her husband chalks it up to aging. But those disrupted nights may be among the earliest measurable signs of neurodegeneration, and research now suggests they may even accelerate the disease process. This article walks through the specific sleep changes that occur at each stage of dementia, what recent science tells us about the relationship between sleep and brain decline, and what caregivers can realistically do when nighttime becomes a battleground.
Table of Contents
- What Sleep Changes Begin in Early-Stage Dementia?
- How Sundowning and Circadian Disruption Escalate in Moderate Dementia
- What Happens to Sleep in Severe and Late-Stage Dementia?
- What Can Caregivers Realistically Do About Dementia-Related Sleep Problems?
- The Bidirectional Link Between Poor Sleep and Dementia Risk
- What Does Brain Aging Research Tell Us About Sleep and Cognitive Decline?
- Where Is Dementia Sleep Research Headed?
- Conclusion
- Frequently Asked Questions
What Sleep Changes Begin in Early-Stage Dementia?
The earliest sleep changes tend to look unremarkable on their own. A person might take longer to fall asleep, wake up once or twice during the night, or start napping in the afternoon when they never used to. Brain wave studies show that measurable decreases in both dreaming (REM) and non-dreaming (non-REM) sleep stages are already underway in this phase. The prevalence of clinical insomnia among dementia patients ranges from 20% to 35%, but subclinical sleep disruption — the kind that doesn’t yet meet the threshold for a formal diagnosis — is far more common. What matters here is how dramatically the numbers differ depending on the type of dementia. Lewy body dementia shows significantly higher sleep disturbance prevalence at 49%, compared to 24% for Alzheimer’s disease, even at early stages. That gap is not a small footnote.
It means a person with early Lewy body dementia is roughly twice as likely to have noticeable sleep problems as someone with early Alzheimer’s. REM sleep behavior disorder — a condition in which people physically act out their dreams, sometimes violently — is particularly common in Lewy body dementia and can appear years before any cognitive symptoms. If a bed partner reports being kicked or punched during sleep, that is worth mentioning to a neurologist, not dismissing as a quirk. The danger in early-stage sleep changes is that they are easy to normalize. Plenty of older adults sleep poorly for reasons that have nothing to do with dementia. But when poor sleep accompanies even subtle memory complaints or personality shifts, it warrants attention. Sleep disruption at this stage is not just a symptom — emerging evidence suggests it may feed the disease itself.
How Sundowning and Circadian Disruption Escalate in Moderate Dementia
As dementia progresses into the moderate stage, sleep disturbances become harder to ignore. The Alzheimer’s Association notes that sleep problems affect up to 25% of people with mild to moderate dementia, rising to 50% in those with moderate to severe disease. The most visible change at this stage is often sundowning — a pattern of increased agitation, confusion, anxiety, wandering, and sometimes yelling that sets in during the late afternoon and evening hours. Sundowning affects up to 20% of dementia patients and is one of the primary reasons families begin considering residential care. The sleep-wake cycle becomes increasingly fragmented during this stage. Rather than one long sleep period at night and wakefulness during the day, sleep starts scattering across the 24-hour cycle, with longer daytime sleep bouts and more nighttime wakefulness. The circadian rhythm — the internal clock that tells the body when to be alert and when to rest — begins losing its signal strength.
Research published in December 2025 in the journal Neurology found that people with weaker circadian rhythms had nearly 2.5 times the risk of developing dementia, with a 54% increased risk per standard deviation decrease in circadian rhythm amplitude. That finding cuts both ways: circadian disruption may be both a consequence of neurodegeneration and a factor that accelerates it. However, not all agitation in the evening hours is true sundowning. Pain, medication side effects, overstimulation, and even hunger can produce similar behavior. Caregivers who assume every late-afternoon episode is sundowning may miss treatable causes. A urinary tract infection, for example, can cause sudden confusion and agitation in an older adult with dementia that looks identical to sundowning but resolves with antibiotics. The distinction matters because the interventions are completely different.
What Happens to Sleep in Severe and Late-Stage Dementia?
By the late stages of dementia, the relationship between day and night has often collapsed. Individuals may spend about 40% of nighttime hours awake while sleeping most of the day. Some patients sleep up to 14 to 15 hours per day, though the quality of that sleep is generally poor — fragmented, light, and punctuated by confusion upon waking. The circadian clock is severely degraded, and the body no longer receives clear signals about when to sleep and when to be alert. The type of dementia continues to shape the severity. Sleep problems affect approximately 40% of Alzheimer’s patients at this stage, but up to 90% of Lewy body dementia patients. That ninety percent figure is staggering and reflects the particular vulnerability of the brain regions that Lewy body disease attacks.
Common disorders at this stage include REM sleep behavior disorder, sleep-disordered breathing, and severe circadian rhythm disruption. For caregivers, this often means round-the-clock vigilance — a person who is awake and confused at 2 a.m. may attempt to leave the house, turn on the stove, or fall while trying to navigate in the dark. One pattern that catches many families off guard is the dramatic increase in total sleep time. A person who was restless and agitated at night during the moderate stage may begin sleeping 14 or more hours a day in late-stage dementia. Families sometimes interpret this as the person “getting better sleep,” but it usually reflects advancing brain deterioration rather than improved rest. The excessive sleep is often shallow and non-restorative, and the person may be difficult to fully rouse even during their waking hours.
What Can Caregivers Realistically Do About Dementia-Related Sleep Problems?
The honest answer is that no intervention fully restores normal sleep in someone with progressing dementia. But there is a meaningful difference between unmanaged sleep disruption and a household that has made reasonable adjustments. The first line of approach is non-pharmacological: maintaining consistent wake times, ensuring bright light exposure during the morning, limiting caffeine and long afternoon naps, keeping the bedroom dark and cool at night, and building a calming evening routine. These measures work best in the early and moderate stages, before circadian function deteriorates too badly. Pharmacological approaches carry tradeoffs that deserve frank conversation with a physician.
Melatonin is commonly tried and is generally safe, but evidence for its effectiveness in dementia patients is mixed at best. Sedative medications like benzodiazepines and antihistamines can increase fall risk, worsen confusion, and cause next-day grogginess — risks that are amplified in people with dementia. Newer options like suvorexant (an orexin receptor antagonist) have shown some promise in Alzheimer’s-related insomnia with a somewhat better side effect profile, but they are not universally effective and can still cause next-morning drowsiness. The comparison comes down to this: non-drug strategies have fewer downsides but require consistency and effort, while medications offer potential short-term relief but introduce real risks in a vulnerable population. For sundowning specifically, structured late-afternoon activities, reduced environmental stimulation in the evening, and light therapy have shown some benefit. But caregivers should also monitor for unmet needs — pain, thirst, boredom, overstimulation — that can masquerade as sundowning behavior.
The Bidirectional Link Between Poor Sleep and Dementia Risk
One of the most unsettling findings in recent dementia research is that sleep disruption is not merely a symptom of neurodegeneration — it appears to drive it forward. A September 2025 Mayo Clinic study found that people with chronic insomnia, defined as difficulty sleeping three or more days per week for three or more months, have a 40% higher risk of developing mild cognitive impairment or dementia. The researchers described this as equivalent to 3.5 additional years of brain aging. Separately, analysis suggests that roughly 12% of U.S. dementia cases could be tied to insomnia. The proposed mechanism involves the brain’s waste clearance system. During deep sleep, the glymphatic system flushes out metabolic waste products, including amyloid-beta, a protein that accumulates into the plaques characteristic of Alzheimer’s disease.
When sleep is disrupted, this clearance process is impaired, potentially allowing amyloid to build up faster. Circadian disruption may compound the problem by altering the timing and efficiency of this clearance. The result is a vicious cycle: neurodegeneration disrupts sleep, and disrupted sleep accelerates neurodegeneration. The limitation here is important to acknowledge. Most of this research is observational or based on animal models. We do not yet have definitive proof that treating insomnia in midlife prevents dementia, though several large clinical trials are underway. It would be premature — and irresponsible — to tell someone that fixing their sleep will prevent Alzheimer’s. But it is reasonable to say that chronic sleep problems deserve medical attention for multiple reasons, and potential brain health is one of them.
What Does Brain Aging Research Tell Us About Sleep and Cognitive Decline?
Recent large-scale studies have added quantitative heft to the sleep-dementia connection. Research examining over 25,000 participants found that suboptimal sleep predicts one to three years of MRI-derived brain age acceleration — meaning the brain looks physically older on imaging than it should for the person’s actual age. That is not an abstract statistic. It translates into measurable differences in brain volume and structure, particularly in regions involved in memory and executive function.
As a practical example, consider two 60-year-olds with similar genetics and lifestyle habits, except one has slept poorly for a decade. The imaging data suggest the poor sleeper’s brain may resemble that of a 62- or 63-year-old, while the good sleeper’s brain looks its age. Over decades, that gap compounds. This does not mean every person with insomnia will develop dementia, but it does mean chronic sleep deprivation is not a benign condition that the body simply adapts to.
Where Is Dementia Sleep Research Headed?
The December 2025 study in Neurology that linked weak circadian rhythms to dementia risk has shifted how researchers think about early detection. Circadian rhythm disruption is now being explored as a potential early warning sign — a biomarker that could flag elevated dementia risk years before cognitive symptoms appear. If validated in further studies, wearable devices that track rest-activity cycles might one day provide an inexpensive screening tool for primary care settings.
On the treatment side, clinical trials are testing whether improving sleep in at-risk populations can slow or delay cognitive decline. The stakes are significant: if even a fraction of the estimated 12% of dementia cases linked to insomnia can be prevented through better sleep management, the public health impact would be enormous. The science is not there yet, but the direction is clear — sleep is no longer a sideshow in dementia research. It is moving toward center stage.
Conclusion
Sleep changes in dementia are progressive, stage-dependent, and more varied than most people realize. Early-stage insomnia and fragmented sleep give way to sundowning and circadian disruption in the moderate stage, and eventually to a near-reversal of the day-night cycle in severe dementia. The type of dementia matters enormously — Lewy body dementia causes far more pervasive sleep disruption than Alzheimer’s at every stage.
And the relationship runs in both directions: poor sleep is not just a consequence of dementia but appears to accelerate the underlying brain degeneration. For caregivers, the practical takeaways are to address sleep changes early, start with non-drug strategies, monitor for treatable causes of nighttime agitation, and have honest conversations with physicians about the risks and limited benefits of sleep medications in this population. For anyone in midlife concerned about long-term brain health, chronic sleep problems are worth taking seriously — not with anxiety, but with the same practical attention you would give to blood pressure or cholesterol. The science connecting sleep to dementia risk is still maturing, but it has moved well past the point of speculation.
Frequently Asked Questions
Is it normal for someone with dementia to sleep all day?
Excessive daytime sleep is common in later stages of dementia, with some patients sleeping up to 14–15 hours per day. However, this is not healthy, restorative sleep. It reflects advancing neurodegeneration and a breakdown of the circadian rhythm. If a person with dementia begins sleeping dramatically more, it is worth discussing with their care team, as it sometimes signals a new medical issue like infection or medication side effects rather than disease progression alone.
What is sundowning, and when does it typically appear?
Sundowning refers to a pattern of increased confusion, agitation, anxiety, or wandering that occurs in the late afternoon and evening. It affects up to 20% of dementia patients and is most common during the moderate stage. The exact cause is not fully understood but appears related to circadian rhythm disruption, fatigue, and reduced ability to cope with stimulation as the day wears on.
Can poor sleep actually cause dementia?
The evidence increasingly suggests that chronic poor sleep raises dementia risk. A 2025 Mayo Clinic study found chronic insomnia is associated with a 40% higher risk of cognitive impairment, and an estimated 12% of U.S. dementia cases may be linked to insomnia. The proposed mechanism involves impaired clearance of amyloid proteins during disrupted sleep. However, this does not mean every person with insomnia will develop dementia — many factors contribute, and definitive proof of causation is still being established.
Why does Lewy body dementia cause worse sleep problems than Alzheimer’s?
Lewy body dementia attacks brain regions involved in sleep regulation more directly than Alzheimer’s does in its early stages. Sleep disturbance prevalence reaches 49% in early Lewy body dementia versus 24% in early Alzheimer’s, and up to 90% in late-stage Lewy body dementia. REM sleep behavior disorder — physically acting out dreams — is particularly characteristic of Lewy body dementia and can appear years before other symptoms.
Should I give melatonin to my family member with dementia?
Melatonin is generally considered safe and is commonly tried, but the evidence for its effectiveness in dementia-related sleep problems is mixed. It may provide modest benefit in some individuals, particularly in the earlier stages. It carries fewer risks than sedative medications, which can increase falls and worsen confusion in people with dementia. Any sleep medication, including melatonin, should be discussed with a physician who knows the patient’s full medical picture.
Are there wearable devices that can detect dementia-related sleep changes?
Consumer wearables that track rest-activity patterns are widely available, and researchers are actively studying whether circadian rhythm data from such devices could serve as an early warning sign for dementia risk. The December 2025 study linking weak circadian rhythms to elevated dementia risk has accelerated interest in this area. However, no wearable device is currently validated as a dementia screening tool, and abnormal readings should prompt a conversation with a physician rather than self-diagnosis.
