Researchers at Center for BrainHealth at The University of Texas at Dallas have identified a specific feature of sleep that helps older adults regulate anxiety overnight — and shown that, as it deteriorates with age, emotional wellbeing deteriorates along with it.

The feature is slow-wave activity: the large, rolling brain oscillations of deep non-REM sleep. In a study recently published in Nature Communications Psychology, the team found that older adults who generated fewer slow waves during the night woke up more anxious the next morning. Those who maintained stronger slow-wave sleep did not show the same pattern, regardless of their chronological age.

The finding reframes how scientists think about late-life anxiety. For years, the prevailing assumption has been that anxiety in older adults is a downstream consequence of brain aging — a natural, if unwelcome, companion to structural changes that accumulate over decades. The new study suggests the picture is more specific, and potentially more tractable.

“Deep sleep acts as a kind of nightly recalibration for the anxious brain. When that recalibration is impaired, anxiety doesn’t fully resolve overnight,” said lead author Eti Ben Simon, PhD, associate director of the Sleep Innovation Laboratories at Center for BrainHealth and research assistant professor at UT Dallas. “The encouraging part is that sleep is modifiable in ways that brain structure is not, offering a powerful lever to help the aging brain continue to adapt and thrive.”

Why Anxiety in Aging Deserves More Attention

Anxiety disorders are the most prevalent mental health condition among adults over 60, yet they receive considerably less clinical attention than conditions such as dementia or depression. The consequences are well-documented: Late-life anxiety accelerates cognitive decline, increases dementia risk, and is associated with earlier loss of independence.

“The stakes here are real,” said senior author Matthew Walker, PhD, director of the Sleep Innovation Laboratories, Laurie and Todd Platt Chair in BrainHealth and professor of neuroscience and biomedical engineering at UT Dallas. “As a society, our brainspans are not keeping up with our extended lifespans, but research continues to reveal new avenues that we can influence to proactively improve and optimize our brain health – including sleep.”

The Study

The research team studied 61 cognitively healthy older adults from the UC Berkeley Aging Cohort, with a mean age of 74.6 years. Each participant spent a night in the sleep laboratory, where brain activity was recorded via polysomnography. Participants reported anxiety levels before sleep and again the following morning. Each participant also underwent a high-resolution MRI scan. A subset of 24 participants returned approximately four years later for repeat assessment.

A consistent pattern emerged: participants who generated less slow-wave activity during deep sleep woke with higher anxiety. The relationship held after controlling for age, sex, total sleep duration, and baseline anxiety traits. Notably, REM sleep — which is more commonly associated with emotional processing — did not show the same predictive relationship. The effect was specific to the slow oscillations of non-REM sleep.

Brain imaging pointed toward a mechanism. Participants with greater shrinkage in key emotional regulatory regions — the amygdala, insula, and cingulate cortex — generated fewer slow waves. Mediation analysis showed that impaired slow-wave activity fully accounted for the relationship between brain atrophy and next-day anxiety.

“The brain atrophy is not directly causing the anxiety,” Walker explained. “It impairs the brain’s capacity to generate deep sleep, and it is that sleep impairment that drives the anxiety. This distinction matters because it points to a target that is more amenable to intervention than structural brain change. The atrophy is upstream — it’s the power cut, not the darkness itself.”

“What surprised us was how specific the effect turned out to be,” Ben Simon added. “We had expected REM sleep to be the primary driver — that is where most emotional processing research has focused. The slow-wave finding was more targeted than we anticipated, and that specificity is what makes it so interesting from a clinical standpoint.”

What the Longitudinal Data Showed

Participants who returned four years later showed a measurable decline in slow-wave activity — roughly 20 percent fewer slow waves on average — and their self-reported anxiety had increased correspondingly over the same period.

Importantly, there was also meaningful individual variation, suggesting that chronological age alone was not a reliable predictor of emotional outcomes.

“Two individuals who are both 75 years old can have quite different anxiety trajectories depending on the quality of their deep sleep,” Ben Simon noted. “Age matters, but it is not the whole story.”

“What the four-year follow-up gives us is a sense of direction,” Ben Simon added. “This is not a static vulnerability — it progresses. And the rate at which it progresses varies considerably between individuals. That variability is where we think targeted intervention could have the most impact.”

IMAGE CREDIT: NASA.