The Neurobiology of PMDD: What Brain Imaging Actually Shows Us
For a long time, PMDD sat in an awkward position in medicine — taken seriously by the women living with it, but often treated by the wider system as something closer to a personality trait than a biological condition. That's changed substantially over the last three decades, mostly because of neuroimaging. We can now watch, in real time, how the brains of women with PMDD respond differently to the same hormones that every woman produces. This post pulls together what that research actually shows — not the simplified version, but the real findings.
It starts with a basic question: do hormones change the brain at all?
Before we could understand PMDD specifically, researchers first needed to establish that gonadal hormones affect brain function in the first place. One of the clearest demonstrations of this came from a PET scanning study that measured blood flow in the brains of women during a demanding cognitive task (a card-sorting test known to reliably activate the prefrontal cortex) under three carefully controlled hormonal conditions: ovarian suppression, suppression plus oestrogen replacement, and suppression plus progesterone replacement.
The result was striking. With gonadal hormones essentially removed from the picture, the women showed almost no activation in the dorsolateral prefrontal cortex during the task — even though they performed it just as accurately. Once either oestrogen or progesterone was added back, normal prefrontal activation returned. In other words, the prefrontal cortex doesn't just tolerate the presence of gonadal hormones — it appears to need them to engage normally during demanding cognitive work.
That single finding reframed the whole conversation. If the frontal lobes require hormonal input to function typically, then it starts to make sense that fluctuating hormones across a cycle might have real, measurable effects on mood, cognition, and emotional regulation — not as some vague "hormonal" hand-wave, but as a specific neurophysiological mechanism.
The emotional brain across the cycle
Once that groundwork was laid, research turned to the parts of the brain most relevant to mood and emotional reactivity — particularly the amygdala, a structure central to processing threat and negative emotion.
One imaging study scanned naturally cycling women twice: once in the early follicular phase, when both oestrogen and progesterone are low, and once in the mid-luteal phase, when progesterone is at its peak. While viewing negatively arousing images, both the amygdala and hippocampus showed significantly greater activity during the luteal phase. Because progesterone — not oestrogen — was the hormone that differed significantly between the two sessions, this pointed to rising progesterone as the likely driver of heightened emotional reactivity.
This fits with a broader pattern across the wider research literature. A large review examining methodologically sound menstrual cycle studies found that while cognitive differences across the cycle (things like mental rotation or verbal memory) are small and inconsistent, emotion-related changes are much more reliably observed — and they track more closely with progesterone than with oestrogen. Higher progesterone levels were consistently linked to increased amygdala reactivity and stronger emotional memory formation.
A related imaging study looked specifically at the orbitofrontal cortex (OFC), a region tightly connected to the amygdala and involved in regulating emotional responses. It found a genuine dissociation across the cycle: the medial OFC became more active in response to negative words during the premenstrual (luteal) phase, while the lateral OFC became less active during that same window. This was interpreted as evidence of increased top-down regulatory demand on the brain premenstrually — essentially, the emotional brain becomes more reactive, and the regulatory regions have to work harder to manage that reactivity, even in women with no premenstrual mood symptoms at all.
It isn't only about function — the brain's structure shifts too
A structural MRI study looked at grey matter volume across the cycle, rather than just activity. It found that grey matter in the right anterior hippocampus was relatively increased in the postmenstrual (late follicular) phase compared to the premenstrual phase, with the opposite pattern in part of the basal ganglia. Verbal declarative memory performance also improved postmenstrually. This is a genuinely important finding: it tells us the hormonal cycle isn't just changing how active certain brain regions are moment to moment, it's associated with detectable structural differences in brain tissue across a single month.
Widening the lens: stress and reward circuitry
Two further studies extend this picture beyond pure emotion processing.
One study compared men to women scanned in different cycle phases while viewing stressful images, looking specifically at the brain's stress-response circuitry — the amygdala, hypothalamus, hippocampus, and prefrontal regulatory regions. It found that women in the late follicular/midcycle phase (when oestrogen is high, unopposed by progesterone) showed markedly attenuated activation in this circuitry compared to men, particularly in the medial and ventromedial prefrontal cortex. The implication is that oestrogen, when not opposed by progesterone, appears to have a genuinely protective, dampening effect on the brain's stress response — a natural hormonal capacity for stress regulation that shifts across the cycle.
Another study took a different angle again, looking at the brain's reward circuitry — the amygdala, orbitofrontal cortex, and striatum — during a monetary reward task. It found these regions were more reactive during the follicular phase than the luteal phase, both during the anticipation of a reward and at the moment of receiving one. This is a useful reminder that hormone-brain interactions aren't confined to negative emotion; they touch reward processing and motivation too, which may have relevance well beyond PMDD, including to conditions involving impulsivity or substance use.
So what's actually different in PMDD?
Everything above describes normal hormonal modulation of the brain in women without significant premenstrual symptoms. The genuinely important question is what happens differently in women who do have PMDD — and here, the research becomes more specific and, frankly, more compelling.
One study directly compared women with PMDD to asymptomatic controls during an emotional word task, scanned in both the premenstrual and postmenstrual phases. In the premenstrual phase specifically, women with PMDD showed an increased amygdala response to negative words compared to controls, alongside a decreased ventral striatum response to positive words. This is a meaningfully different pattern from simple cycle-related fluctuation — it suggests that in PMDD, the emotional brain isn't just more reactive premenstrually in the way every woman's is to some degree; it's reactive in a way that's measurably abnormal compared to women without the condition.
A further multimodal study, using both PET and fMRI in the same patients, went further still. Women with PMDD were scanned during a working memory task under three controlled hormonal conditions (ovarian suppression, oestrogen add-back, progesterone add-back) — deliberately designed so that any differences found couldn't simply be attributed to different circulating hormone levels between groups, since those levels were tightly controlled. Across both imaging methods, PMDD patients showed abnormally increased activation in the dorsolateral prefrontal cortex compared to controls — and critically, the degree of that overactivation correlated with real clinical measures: functional disability scores, age at symptom onset, and how severely mood symptoms had changed between the premenstrual and postmenstrual weeks before the study even began. Earlier onset and greater disability were both associated with more pronounced prefrontal overactivation. This is strong evidence that the abnormal brain response in PMDD isn't a side effect of mood state on the day of scanning — it looks like a stable, trait-level marker of vulnerability to the condition.
A further study adds another layer, showing that women with PMDD display altered prefrontal reactivity specifically during the anticipation of emotional stimuli — not just in response to them. That matters clinically, because so much of the lived experience of PMDD involves dread and anticipatory anxiety in the days before symptoms peak, not just reaction to what's already happening.
Beyond the amygdala and prefrontal cortex: what a wider review adds
A more recent systematic review pulled together findings from eighteen separate neuroimaging studies of PMDD, covering just over 360 women in total, and it both confirms and extends the picture above.
On confirmation: pooling across this much broader set of studies, the same core pattern holds up — women with PMDD reliably show enhanced amygdala reactivity alongside reduced frontal-cortical reactivity in response to emotional stimuli, compared to women without the condition. That's genuinely reassuring from a scientific standpoint. A single study finding something is interesting; eighteen independent studies converging on the same corticolimbic pattern is much harder to dismiss as coincidence.
But the review also surfaces two things that hadn't shown up in the studies discussed so far. First, several structural imaging studies found greater grey matter volume and altered metabolism in the cerebellum in women with PMDD — a brain region traditionally associated with movement and coordination, but increasingly understood to also play a role in emotional processing. This wasn't something I expected going into this research, and it's a reminder that PMDD's neural signature may extend beyond the "usual suspects" of amygdala and prefrontal cortex. Second, several studies identified altered serotonergic and GABAergic neurotransmission in the PMDD brain — meaning the chemical signalling systems themselves, not just which regions activate, appear to function differently. That's a particularly interesting finding given that both of these neurotransmitter systems are the direct targets of the two most established PMDD treatments: SSRIs act on serotonin, and allopregnanolone (progesterone's neuroactive metabolite) acts directly on the GABA system.
The review also found that differences between women with PMDD and controls weren't constant across the whole cycle — they were often specific to particular menstrual cycle phases, showing up in the amygdala, insula, prefrontal, occipital and cerebellar regions at some points in the cycle but not others. That fits precisely with the clinical picture: PMDD isn't a fixed trait that's always present, it's a brain that responds differently specifically when exposed to particular hormonal transitions.
Bringing it together
A comprehensive systematic review pulling together the imaging literature on oestrogen and progesterone effects on emotional and cognitive brain function found a consistent overall picture: cognitive effects of the cycle are modest and inconsistent, but emotional processing is reliably and substantially modulated by ovarian hormones, with progesterone and the luteal phase featuring especially heavily in relation to amygdala reactivity.
What all of this adds up to, I think, is a genuinely coherent biological story. Gonadal hormones don't simply exist somewhere in the bloodstream and separately affect mood as an afterthought — they act directly and specifically on the brain circuits responsible for emotional reactivity, regulation, memory, and stress response, and they do so differently across the menstrual cycle in every woman. In PMDD, that same system responds abnormally: not just more strongly in the way any sensitive woman's might, but in patterns of over- and under-activation that are measurably different from women without the condition, and that correlate directly with how severe and disabling someone's symptoms are.
It's also worth pausing on why the serotonin and GABA findings matter beyond the lab. They're not just an interesting biological detail — they line up directly with why our two most established PMDD treatments actually work. SSRIs act on serotonin; strategies that stabilise or dampen allopregnanolone's effect on the GABA system (including some newer approaches) target that pathway directly. The neuroimaging isn't just describing PMDD, it's pointing at exactly the systems we're already intervening on clinically, and helping explain why they help.
This is really the scientific foundation for something I keep coming back to across this whole podcast series: PMDD is not a psychological overreaction to normal hormones. It's a specific, identifiable difference in how certain brain circuits — the prefrontal cortex, the amygdala, the orbitofrontal cortex, the hippocampus — respond to the same hormonal signals every woman experiences. The hormones aren't the problem. The brain's response to them is.