Skip to main content

Verified by Psychology Today


Cortisol: Harvard Study Finds "Moderate-response" May Be Key

Releasing "too much" or "too little" stress-induced cortisol may be detrimental.

Wikimedia/Life Sciences Database
Hippocampus in red.
Source: Wikimedia/Life Sciences Database

A new study reports that producing either “too much” or “too little” cortisol (colloquially known as “the stress hormone”) in response to a stressful experience is associated with someone feeling more negative emotions during times of distress than cohorts who sustained "moderate" levels of cortisol. Higher and lower cortisol responses during this experiment were also linked to less gray matter volume in the hippocampus.

This research was conducted by a team of researchers from Harvard Medical School in conjunction with McLean Hospital. The findings were published online July 24 in the Journal of Neuroscience.

Over the past few decades, researchers have made many advances that help us better understand the neurobiological mechanisms that drive stress responses. However, the role that various trajectories of cortisol play in how each individual responds emotionally to stress and the interplay between stress hormones and brain structure have remained enigmatic.

For this cortisol trajectory study, the researchers used a well-established laboratory procedure of acute stress induction that lasted for about an hour with a group of 79 healthy female participants. The goal was to measure cortisol levels via saliva sample in response to prolonged acute stress and then to have participants self-report the degree of negativity each person experienced subjectively during the experiment.

Additionally, the researchers used structural MRI brain scans to correlate various levels of cortisol response with gray matter volume in the hippocampus—which is a memory hub and brain structure with an abundance of cortisol receptors that plays a role in regulating stress responses.

The researchers were able to identify three distinct trajectories of cortisol response among the 79 participants: hyper-response (n=10), moderate-response (n=21), and mild-response (n=48). Although all three responder groups exhibited an uptick in cortisol release and a negative emotional response to the lab-induced stressful circumstance, both the hyper-response and mild-response groups reported significantly more negative affective responses in comparison to the moderate "n=21" cortisol group.

Also, MRI analysis showed that higher and lower levels of stress-induced cortisol release were associated with less hippocampal gray matter volume when compared to the moderate-response cortisol group. Based on these findings, one could speculate that having a moderate cortisol response to acute stress appears to represent a type of "sweet spot" that is better than having too much or too little cortisol.

Over the years, I've written extensively about cortisol in various Psychology Today blog posts such as, "Cortisol: Why 'The Stress Hormone' Is Public Enemy No. 1." I keep my antennae up for any new, pioneering research on this topic. Therefore, I was fascinated to read about the latest discovery of these three different trajectories of cortisol response to acute prolonged stress and correlations to hippocampal gray matter volume. But, I was also curious to learn more directly from the researchers about their somewhat counterintuitive findings. And to get some fresh insights about cortisol for Psychology Today readers.

Yesterday, I reached out to senior author Diego Pizzagalli, who is Director of Harvard Medical School's Center for Depression, Anxiety, and Stress Research at McLean Hospital, and lead author Roee Admon via email. Below are three questions from Christopher Bergland (in bold) and responses from Roee Admon in block quotes:

1. Many experts might make an educated guess that lower levels of cortisol would be a biomarker for someone feeling less stressed. But, your study shows that both higher and lower levels of cortisol were linked to people feeling more stressed out. Do you and your colleagues have a hypothesis as to why study participants with moderate-response levels of cortisol during acute prolonged stress self-reported less negative affect than those in the hyper-response and mild-response groups?

Unfortunately, the link between levels of cortisol and emotional response to stress is far from being trivial. Intuitively, it does make sense to think that low cortisol is associated with less negative emotional response. At the same time, extensive work has demonstrated that individuals diagnosed with stress related mental disorders actually exhibit less cortisol release in response to stress compared to healthy individuals. One potential explanation for this could be that repeated stress exposure may sensitize the endocrine system yielding a decrease in cortisol response as a result of overcompensation. Accordingly, we can speculate that “too little” and “too much” cortisol are both maladaptive. This is also in accordance with the well-established inverted U-shape relationship between cortisol levels and cognitive performance, with beneficial effects of moderately elevated cortisol levels on cognition.

2. Do you have any speculation as to why there is a correlation between less hippocampal gray matter volume and both higher or lower levels of stress-induced cortisol release in comparison to the moderate-response cortisol group's more robust hippocampal volumes?

Similar to previous reply, here as well, relationships between cortisol, emotional response, and hippocampal volume are not yet clear. If we follow the same line as in the previous response, repeated exposure to stress is also associated with morphological damage, particularly in the hippocampus (see the work of Bruce McEwen). Accordingly, the same individuals that exhibit low levels of cortisol release due to repeated stress exposure may also exhibit reduced hippocampal volume. If this is indeed the case one could further speculate that individuals with excessive cortisol release and negative response may eventually experience a decrease in their hippocampal volume, given sufficient exposure to stress.

3. From a clinical perspective, how do you see the findings of your latest research being applied to identify and treat individuals who are especially sensitive to the emotional toll of physiological stress responses?

Findings would need to be replicated before being applied to the clinical field. The most significant contribution of the current work is to highlight the importance of accounting for individual differences in response to stress, rather than overlooking them, when attempting to identify biological markers of stress vulnerability and resiliency in humans.

Diego Pizzagalli responded this morning in a separate email, "Roee's replies are spot on. With respect to the first point, as an example of what Roee mentioned, we recently found that completely remitted individuals with a history of major depression showed reduced cortisol release (but increased self-reported stress reactivity) in response to the same psychosocial stressor we used in Roee's study, "Acute stress impairs frontocingulate activation during error monitoring in remitted depression." (Published Jan. 2017 in the journal Psychoneuroendocrinology.)

Pizzagalli concluded, "These responses were associated with reduced activation in prefrontal brain regions important for stress and emotion regulation. Of note, reduced stress-induced activation in these brain regions predicted greater depressive symptoms 18 months later. Collectively, we hope that findings such as Roee's and the ones by Alexis Whitton will help us identify individuals at risk for first onset and recurrence of depression (several studies in this topic are currently ongoing in our center (CDASR))."

The latest findings by Roee Admon, Diego Pizzagalli, and colleagues offer exciting possibilities for advancing personalized diagnosis and treatment options for individuals suffering from depression, chronic stress, and/or anxiety. Please stay tuned for updates on this team's ongoing research and specific ways that future empirical evidence can be applied prescriptively in the clinical field.


“Distinct trajectories of cortisol response to prolonged acute stress are linked to affective responses and hippocampal gray matter volume in healthy females” by Roee Admon, Michael T. Treadway, Linda Valeri, Malavika Mehta, Samuel Douglas and Diego A. Pizzagalli in Journal of Neuroscience. Published online July 24, 2017 DOI: 10.1523/JNEUROSCI.1175-17.2017

Whitton, Alexis E., Ashlee Van’t Veer, Pragya Kakani, Daniel G. Dillon, Manon L. Ironside, Anja Haile, David J. Crowley, and Diego A. Pizzagalli. "Acute stress impairs frontocingulate activation during error monitoring in remitted depression." Psychoneuroendocrinology 75 (2017): 164-172. DOI: 10.1016/j.psyneuen.2016.10.007