Mind Over Molecules

A New Study Reveals that Your Perception of Stress, More than Your Cortisol Levels, Predicts Changes in Gene Expression

September 23, 2025
stock photo with DNA

A new study from Whole Communities–Whole Health researchers shows that how stressed you feel may be just as important as how stressed your body is, at least when it comes to changes in your DNA.

Led by psychology professor Frances Champagne (College of Liberal Arts), the study tracked biological responses in 55 undergraduate participants who were put through a semi-virtual version of the Trier Social Stress Test, a gold-standard protocol that includes an impromptu speech and a surprise math test.  

The twist this time? Participants were judged via Zoom.  

"Even though the judges weren’t physically in the room, the experience still triggered a pretty robust stress response,” Champagne said. “We were quite happy to see the typical stress response curves, both in terms of perceived stress and hormone levels."

"If I say I’m really stressed, it’s not just in my head. It’s something you can see at the level of controlling how your genes operate."

— Frances Champagne, College of Liberal Arts

This new study builds on earlier work by Champagne and colleagues, who previously demonstrated that the semi-virtual Trier test could reliably trigger psychological and hormonal stress responses in lab participants. That research helped validate the adapted stress-testing protocol and set the stage for a deeper investigation: Could a single stressful experience cause immediate changes in gene function? 

Enter DNA methylation (DNAm), a chemical process that helps regulate which genes are turned on or off. Methylation is one of the most studied forms of epigenetic change: modifications that don’t alter your DNA sequence but can affect how your genes function. 

Using saliva samples collected before and after the test, Champagne’s team found that DNAm changes happened rapidly — within just 90 minutes. More importantly, they found those changes didn’t track with cortisol, the body’s primary stress hormone. Rather, they tracked with subjective stress — in other words, how stressed participants said they felt. "If I say I’m really stressed, it’s not just in my head," Champagne said. "It’s something you can see at the level of controlling how your genes operate."

Spit Takes

To make sense of the findings, it helps to understand that DNAm is like a volume knob for genes: it doesn’t rewrite your genetic code, but it can influence how loudly a gene “speaks.” And while methylation has long been seen as relatively stable, new research like this suggests it’s far more dynamic and responsive to the environment than once thought.

Champagne’s team also examined changes in what's known as epigenetic age, a measure of how quickly or slowly a person is biologically aging based on patterns of DNAm. They found that while perceived stress best predicted real-time methylation changes, hormone levels were more closely tied to epigenetic age. "It’s this idea that we can accelerate or decelerate our aging process," Champagne said. "Luckily, it goes both ways. If someone’s aging really fast, there are interventions — stress management, diet, exercise — that can slow the clock down."

"Paying attention to how people feel, as opposed to some objective measure, might actually be more useful in predicting molecular changes. Subjective stress really does have a biology to it."

— Frances Champagne, lead author

The findings are now informing Whole Communities–Whole Health’s flagship longitudinal study, which tracks multiple health markers, including stress and biological aging, in nearly 500 parents and children throughout southeast Austin. Participants contribute annual saliva samples, from which researchers calculate their epigenetic age and offer suggestions for managing stress, building social support and improving well-being.

While the Trier test won’t be used on community participants, the lab-based results offer clues. "We could look at the genes that are most stress-sensitive in the social stress test and see if those same genes show up in people who’ve experienced high levels of stress in real life," Champagne said.

Ultimately, the goal is to develop less invasive, more personalized ways to identify who might be most vulnerable to stress, and then to intervene before the health consequences take root. "Paying attention to how people feel, as opposed to some objective measure, might actually be more useful in predicting molecular changes," Champagne said. "Subjective stress really does have a biology to it."