There are books that stay with you for years.
I first read Haruki Murakami’s Norwegian Wood as a teenager, and Midori has stayed with me ever since: bright, charming, uninhibited and bursting with life, she is one of those female characters who leaps off the page. Of all the scenes in the book, one small exchange has remained particularly vivid. Midori jokes that, during her period, she should wear a red hat as a warning to everyone else. Perhaps that scene stayed with me because it taps into such a familiar stereotype: the idea that menstruating women become irritable, unreasonable or somehow harder to be around.
Looking back now, as a woman, a doctor and a researcher, I understand that there is much more behind these symptoms than a familiar stereotype. For some people, these cyclical changes are severe, disabling and part of a recognised condition: premenstrual dysphoric disorder (PMDD).
For people with PMDD, severe mood symptoms (e.g., irritability, anxiety, low mood, emotional sensitivity) typically emerge during the luteal phase (after ovulation and before menstruation) and usually ease shortly after menstruation begins. PMDD is much more than simply having a “bad period”: symptoms can seriously disrupt relationships, work and everyday life. (Roux C, 2025; Aspey 2025, Hantsoo L and Payne JL, 2023).
PMDD is not simply caused by abnormal hormone levels. Instead, some people seem to be particularly sensitive to the normal hormonal changes that occur across the menstrual cycle (Hantsoo L and Payne JL, 2023). But where might this sensitivity show up in the brain?
In their recent study, Stiernman and colleagues (2026) aimed to answer this question.
Methods
Stiernman and colleagues used functional magnetic resonance imaging (fMRI) to compare 29 women with premenstrual dysphoric disorder (PMDD) to 27 controls across the menstrual cycle. Participants completed two emotional tasks designed to examine emotion generation and emotion regulation, allowing the researchers to distinguish responses to negative emotional information from deliberate attempts to regulate those responses.
The researchers compared brain activity and connectivity between the groups and across menstrual-cycle phases. They focused on large-scale networks involved in emotion, including the salience network (i.e., helps us detect and prioritise information that seems important) and default mode network (i.e., involved in internally focused processes like thinking about ourselves), and examined whether group differences were associated with premenstrual symptom severity or serum levels of progesterone-derived neuroactive steroids.
Results
The brain’s “importance detector” responded more strongly in PMDD.
As expected, women with premenstrual dysphoric disorder (PMDD) reported higher levels of negative emotions during the luteal phase compared with the follicular phase, particularly when participants simply looked at negative images (F(1,54)=10.83, p=.002, η²=.025), when they tried to distance themselves from negative images (F(1,54)=5.14, p=.027, η²=.011) and from neutral images (F(1,54)=4.55, p=.038, η²=.020). This difference between phases was not found in controls.
Moreover, women with PMDD showed greater activity in the bilateral amygdala, a key region of the salience network, during the luteal phase (when PMDD symptoms typically occur) compared with the follicular phase. This pattern was not seen in women without PMDD (p<0.05, TFCE corrected), and it survived correction for multiple comparisons. This suggests that negative emotional information may generate a stronger neural response in women with PMDD. There was also weaker, subthreshold (corrected p values ranging between 0.07 and 0.08) whole-brain evidence of greater activity in regions associated with the default mode network during the luteal phase compared with the follicular phase.
Differences in the absence of symptoms
When compared with controls, women with premenstrual dysphoric disorder (PMDD) showed greater activity in the dorsal anterior cingulate cortex (a key part of the salience network) and in the cerebellum, in both the follicular (i.e., when symptoms are typically absent or reduced) and the luteal phase (corrected p<0.05). Moreover, greater follicular-phase salience-network activity was related to the severity of their premenstrual symptoms.
Finally, in women with PMDD, greater dorsal anterior cingulate cortex activity during the follicular phase was moderately associated with more severe premenstrual depressive symptoms (r=.40, p=.03).
This raises an interesting possibility. Rather than the brain changing only when PMDD symptoms appear (luteal phase), some women with PMDD may have a more reactive emotional-processing system even when relatively symptom-free (follicular phase). Hormonal changes later in the cycle might interact with this underlying vulnerability, increasing susceptibility to severe mood symptoms.
But there is an important “might” in that sentence. The study shows an association. It cannot tell us whether heightened salience-network activity causes later symptoms, is a consequence of living with recurring PMDD symptoms, or reflects something else entirely.
Consciously controlling emotions?
Both women with premenstrual dysphoric disorder (PMDD) and controls showed similar network connectivity patterns and similar activity in brain regions associated with the conscious control of emotion.
These findings do not strongly support the theory that intense emotions in PMDD result from weaker emotional control. Instead, the more convincing differences appeared during emotion generation: the initial response to emotionally negative information. These data do not suggest that people with PMDD have weaker “emotional brakes”, but that the emotional signal itself may be more intense.
What about hormones?
One influential theory proposes that symptoms arise not from abnormal hormone concentrations themselves, but from an altered brain response to normal hormonal fluctuations (Stumper et al, 2026). This study found no relationship between the observed differences in brain activity and serum concentrations of progesterone-derived neuroactive steroids.
This does not mean that neuroactive steroids are irrelevant to PMDD, just that the specific relationship examined here was not demonstrated.
Conclusions
- Stiernman and colleagues found evidence of increased responsiveness in the brain’s salience network in women with premenstrual dysphoric disorder (PMDD), including during the relatively symptom-free follicular phase.
- Greater salience-network activity during this phase was also associated with more severe premenstrual symptoms.
- In contrast, the study found little evidence that PMDD involves abnormalities in brain regions involved in consciously regulating emotion.
- Perhaps the question is not simply “Why are emotions harder to control?” but “Why might the emotional signal be stronger in the first place?”.
Strengths and limitations
There is quite a lot to like about this study, alongside some reasons to remain cautious about its findings.
A major strength of this study is its repeated-measures design. Premenstrual dysphoric disorder (PMDD) is defined by symptoms that change across the menstrual cycle, so assessing the same participants at different cycle phases is more informative than relying on a single phase. The inclusion of a control group (women without PMDD), alongside symptom ratings, hormone measures and fMRI, also allowed the researchers to test several possible explanations for the observed differences.
The distinction between emotion generation and emotion regulation was particularly useful. Rather than assuming that intense emotions in PMDD reflect poorer emotional control, the researchers were able to examine whether differences emerged earlier, when emotional responses were first generated.
However, the study was small, including just 29 women with PMDD and 27 controls. This limits the precision of the findings and is particularly important in neuroimaging research, where many brain regions and connections can be tested at once. Moreover, the study cannot determine whether heightened salience-network activity causes PMDD symptoms or results from them.
The authors controlled for several confounding factors, including psychiatric illness and psychotropic medications. However, recent stress or stressful life events do not appear to have been specifically measured or accounted for and could potentially have influenced both emotional responses and brain activity.
The authors also note that participants were relatively young and mostly nulliparous (i.e., have never given birth to a live child), which may limit the generalisability of the findings. In addition, there is limited information on the ethnic or broader sociodemographic diversity of the sample, making it difficult to judge how representative it was of the wider PMDD population.
Implications for practice
So, should clinicians change their day-to-day practice after reading this study?
This is a question I often struggle with when it comes to neuroimaging research. The beauty of these studies is not in their ability to immediately change clinical care, but in how they help us understand what may be happening underneath the symptoms. Brain imaging can give us clues about the processes that connect hormones, brain function and what people actually experience. That may not change practice tomorrow, but it can help us ask better questions today.
For now, these findings do not make fMRI a diagnostic test for premenstrual dysphoric disorder (PMDD), nor do they point to an immediate change in treatment. Their clinical value is more indirect: they reinforce the importance of taking cyclical emotional symptoms seriously. For clinicians, this means asking carefully how symptoms change across the menstrual cycle. Indeed, prospective daily symptom tracking across at least two cycles remains important when assessing possible PMDD (Reilly TJ et al, 2024).
The findings may also help shape treatment research. If PMDD involves altered responses to hormonal changes and heightened salience processing, future studies could ask whether different approaches, including hormonal, psychological or other biological treatments, act on different parts of this pathway, and whether successful treatment changes the brain responses observed here. For example:
- Is heightened salience-network activity specific to PMDD?
- Does it precede symptoms, or develop after years of recurrent episodes?
- Does it change with effective treatment?
Perhaps the most useful takeaway is that PMDD deserves to be understood as a complex, cyclical condition, not dismissed as an inability to control emotions.
In this light, Midori’s red hat feels less like a warning sign and more like a reminder of how easily menstrual symptoms can be reduced to a stereotype.
Rather than asking why menstruating women can be so “difficult”, this study asks a more interesting question: why might some people’s brains respond so strongly to otherwise normal hormonal changes? That does not give us the full answer to PMDD, but it takes us a little further from stereotype and a little closer to mechanism.
Statement of interests
Maria (Etta) Nettis declares no conflicts of interest. AI was used to generate the blog’s first image and to assist with language refinement.
Editor
Edited by Éimear Foley. ChatGPT assisted with language refinement and formatting during the editorial phase.
Links
Primary paper
Stiernman Louise, Dubol Manon, Comasco Erika, Johansson Maja, Stiernman Lars, Bixo Marie. (2026) Emotion Generation and Regulation in Premenstrual Dysphoric Disorder: Dysregulation of Large-Scale Brain Networks Across the Menstrual Cycle. Biological Psychiatry 99(3): 227-236. https://doi.org/10.1016/j.biopsych.2025.05.025
Other references
Murakami H. (2000) Norwegian Wood. Translated by Jay Rubin. London: Harvill
Press. Roux C. (2025) When help feels out of reach: mental health and the menstrual cycle. The Mental Elf, 10 June 2025.
Aspey C. (2025) From menstruation to menopause: how sex-steroids shape women’s mental health across the life course. The Mental Elf, 1 October 2025.
Hantsoo L, Payne JL. (2023) Towards understanding the biology of premenstrual dysphoric disorder: From genes to GABA. Neuroscience & Biobehavioral Reviews 149:105168. https://doi.org/10.1016/j.neubiorev.2023.105168
Stumper A, Schmalenberger KM, Eisenlohr-Moul TA, Peters JR. (2026) Affective Sensitivity to Ovarian Steroid Hormone Flux Across the Menstrual Cycle: Manifestations and Biopsychosocial Risk Factors. Annual Review of Clinical Psychology, 22:77–104. doi:10.1146/annurev-clinpsy-061724-083756
Reilly TJ, Patel S, Unachukwu IC, Knox C-L, Wilson CA, Craig MC, Schmalenberger KM, Eisenlohr-Moul TA, Cullen AE. (2024) The prevalence of premenstrual dysphoric disorder: systematic review and meta-analysis. Journal of Affective Disorders 349:534-540. https://doi.org/10.1016/j.jad.2024.01.066