The anxiety of early perimenopause, PMT and PMDD are on a continuum – and I treat all 3 with a little-known mineral with a bad reputation

This week in clinic, I have been supporting a 45-year-old woman in early perimenopause, struggling with her anxiety especially in the week before her period. Her periods were coming closer together, now 3 weeks apart, which meant that every 2 weeks she was barely able to contain the feelings of being severely irritable, snappy, intolerant at work, having crying spells and feeling overwhelmed. This situation had been escalating over the last few months.

These symptoms are not unusual in perimenopause and are a next-level version of premenstrual tension (PMT): the early stage of perimenopause sees oestrogen spiking wildly as it starts falling slowly overall, combined with erratic ovulation (the Progesterone Shift). One analysis found follicular oestrogen running 29% higher and premenstrual oestrogen 22% higher in perimenopausal women, with roughly a third of cycles showing an extra oestrogen surge during the luteal phase itself (1).

I like to think of oestrogen as fun-but-can-go-crazy when pushed too far, and that’s because it increases the sensitivity of the receptor that binds with glutamate, the excitatory neurotransmitter (2). At normal physiologic levels, this has the advantage of boosting learning and memory, and the ability to switch tasks and solve complex problems. But when oestrogen pushes glutamate goes too high, it can cause a hyper-excitatory state in the brain, driving specific, intense mood disturbances:

• Extreme irritability and rage: Unopposed oestrogen decreases the brain’s calming GABA transmission while ramping up glutamate. This shifts the neurological baseline away from relaxation, resulting in unprovoked agitation, a “short fuse,” and episodes of sudden rage.
• Hyper-reactive anxiety and panic: High glutamate speeds up neural firing in emotional centres like the amygdala. Without ALLO to act as a brake, the nervous system remains in a constant state of high alert, triggering acute anxiety, restlessness, and new-onset panic attacks.
• Intense mood swings: Perimenopause is defined by rapid hormonal ups and downs rather than a steady decline. While surges of unopposed oestrogen cause hyper-excitability (anxiety, rage, panic), the sudden drops in oestrogen during perimenopause cause glutamate levels in the brain to plummet, which heavily drives perimenopausal depression, fatigue, and severe brain fog. It is this constant swinging between hyper-glutamatergic excitement and hypo-glutamatergic depletion that can make early perimenopause feel like a roller-coaster of extreme emotional highs and lows (3).
• Insomnia and wired restlessness: Glutamate is highly wake-promoting. An elevated glutamate-to-GABA ratio leaves the brain feeling “wired but tired,” making it difficult to quiet the mind, fall asleep, or stay asleep.
• Sensory overwhelm: Glutamate amplifies how the brain processes incoming information. When overactivated, everyday sensory inputs – such as loud noises, bright lights, or crowded environments – can feel physically distressing and emotionally overwhelming.

If ovulation was happening, progesterone would be made and it would balance out oestrogen.

Progesterone is oestrogen’s calm, relaxed sister (but has a tendency to lethargy and feeling low if pushed too far). Progesterone converts to allopregnenolone – known as ALLO. ALLO ALLO! Here comes some peace and quiet: coming up to menstruation, ALLO travels to the brain, binds with the GABA receptor, and creates a calm, relaxed state of mind that GABA, the calming neurotransmitter, is known for.

But in early perimenopause, not only is oestrogen spiking, but progesterone often fails to be made in sufficient amounts because of fewer and fewer ovulatory cycles – and less progesterone means less raw material to convert into ALLO.

At the same time, even if ovulation does occur to make progesterone to balance oestrogen, the amount may not be enough to prevent the mood swings because of a conversion issue: a study found that women with PMS had significantly lower ALLO on day 26 despite having normal progesterone levels compared to controls – meaning they converted less of the same progesterone pool into ALLO (5).

In some women, there can also be another problem. To use ALLO and get the calming effect, the brain relies on the specialised landing docks, the GABA receptors. These receptors are built from several different moving parts, but the most important one for emotional stability is a component called the delta subunit. Think of the delta subunit as a docking station. These docking stations don’t just sit there passively waiting for ALLO to arrive – they actively expand to meet it. The brain builds more docking capacity in real time so that by the time ALLO peaks in the days before your period, there’s enough dock space to receive it cleanly. Then ALLO docks in the way it’s meant to, acting as a gentle, natural sedative, and the rising hormone produces a stable, relaxed luteal phase.

In women with PMS, the docking capacity grows, but not quite enough to keep pace with the hormone rise. So as progesterone and ALLO climb, symptoms climb with them in real time – irritability, anxiety, sugar cravings, and low mood building steadily as ALLO rises but doesn’t dock, cresting in the final few premenstrual days, and clearing within days of your period starting.

This mechanism is heightened in women with premenstrual dysphoric disorder (PMDD), the most severe and disabling end of the spectrum of premenstrual disorders. Here the docking stations are barely made at all. Direct measurement in women with PMDD shows lower subunit expression during the luteal phase and tracks with heightened amygdala reactivity – the brain’s threat-detection centre going into overdrive (6). Without that capacity, ALLO keeps arriving at a system with nowhere to go, and past a certain point, this shrunken docking system doesn’t just fail to create calm – it converts the ALLO signal into rage, extreme anxiety, and depression. So PMS and PMDD are the same story, told at different volumes.

In essence, 4 things may be going on:

1. Oestrogen spikes → glutamate spikes. Oestrogen surges in early perimenopause directly overstimulate glutamate signalling, regardless of what’s going on with progesterone or ALLO.

2. Progesterone is present, but has poor conversion to ALLO. Even in a cycle where ovulation does happen and progesterone rises normally, the conversion to ALLO can be too low.

3. ALLO present, but the docking stations aren’t responding The receptor is not receiving the calm signal.

4. No ovulation → no progesterone/ALLO at all. No ovulation, no corpus luteum, no ALLO whatsoever.

You may be thinking that 4 different ways things can go wrong mean there need to be 4 solutions, but in this case, that’s not true.

In my practice, I work in 2 lanes: a fast lane to relieve immediate symptoms as quickly as possible while also working in the slower lane, on root causes.

The root cause of these mood swings is, of course, the excess oestrogen. This means that clearing oestrogen will reduce symptoms, and obviate the need to support progesterone. I am addressing this with my client by increasing the clearance of oestrogen via the liver, optimising gut function so that the cleared oestrogen doesn’t recycle back into the body, and keeping blood sugar stable to avoid insulin spiking oestrogen. This approach can take 8-12 weeks to bed in, but in the meantime, nobody wants to wait that long to stop feeling like their life is unbearable.

My go-to approach for quick resolution of hormonal mood swings in all of the 4 cases above is a mineral with a bad reputation: lithium. The lithium most people have heard of – lithium carbonate – is a prescription-only psychiatric drug used at high doses (typically 900-1800mg a day) to treat bipolar disorder, and it requires regular blood tests because the effective dose sits close to a toxic one. But there is another version of lithium which is a completely different animal. Lithium orotate is an over-the-counter mineral supplement providing a tiny fraction of that dose – often only a few milligrams of actual lithium, nowhere near prescription strength – and it doesn’t require blood monitoring. In my practice, I have found that low-dose lithium orotate has a consistently strong, positive effect on the mood swings of early perimenopause, PMS, and PMDD – and it doesn’t seem to matter which end of the spectrum a woman sits on, or what her particular underlying picture looks like.

In just 5 days on lithium orotate, my client experienced a profound change in her mood – she had gone from feeling chronically irritable and tearful to feeling much calmer and centred; her mood volatility had already reduced by 50%. This, of course, makes it much easier to continue working in the slow lane, modulating excess oestrogen.

Given that I have seen lithium orotate have similar positive effects on mood swings arising from these different underlying pictures, it is interesting to consider how it works.

Lithium strengthens brain circuitry via four separate, well-established actions:

1. It boosts GABA, the brain’s main “off switch”. GABA is the brain’s principal calming chemical – think of it as the brake pedal for an overactive nervous system, damping down excitability rather than firing it up. The delta docking station is one route GABA uses to apply that brake, but it isn’t the only one. Lithium increases how much GABA gets released from nerve cells in the first place (7). More GABA released means more of the brake signal getting through via the routes that are working normally, even while the ALLO docking station remains under-built.
2. It turns down the brain’s excitatory “accelerator,” glutamate. If GABA is the brake pedal, glutamate is the accelerator – the brain’s main excitatory chemical, responsible for revving up neural activity. Too much unchecked glutamate signalling is linked to anxiety, agitation, and overstimulation. Lithium has one of its most repeatedly confirmed effects right here: it reduces how strongly nerve cells respond to glutamate at a specific docking point called the NMDA receptor, largely by blocking a chemical modification that would otherwise make that receptor more reactive (8, 9). This doesn’t necessarily reduce how much glutamate is floating around – the evidence on that is mixed – but it reliably reduces how loudly the receptor answers when glutamate arrives, which turns down the net excitatory effect regardless of what’s driving it upstream.
3. It boosts serotonin, the brain’s own mood-stabilising chemical. Lithium reliably increases serotonin signalling in the brain, including measurable increases in serotonin receptor activity on brain imaging (10, 11). This is the same serotonergic action underlying lithium’s long-established reputation as an antidepressant – and, notably, one of the few agents ever shown to specifically reduce suicide risk (12). It gives the system a steadier serotonin floor to fall back on, independent of what’s happening with ALLO and the delta subunit receptor at all.
4. It physically strengthens the amygdala circuit under strain. Lithium reliably increases the volume of, and improves connectivity within, the amygdala and prefrontal cortex (13, 14) – the amygdala being the brain’s threat-detection and emotional-reactivity centre, and the prefrontal cortex being the region that normally keeps it in check. Lithium builds a sturdier emotional-regulation circuit with more capacity to absorb the emotional disruptions, whatever specific form that disruption takes in a given woman. In my practice, symptom relief usually starts within days, as with my client above, but the fuller, more durable stabilisation – where mood swings settle down for good rather than just easing – tends to build over 4-6 weeks of consistent use.

None of these four actions treats the upstream causes directly – the oestrogen spike, the poor ALLO conversion, the under-built docking station, or the skipped ovulation itself. Instead, they act one step downstream, on the shared brain circuitry that all four of these problems ultimately end up disrupting. This also matches something lithium is independently, robustly known for outside of any hormonal context at all: reducing irritability, impulsive aggression, and emotional reactivity across a huge range of unrelated conditions, and even in people with no psychiatric diagnosis whatsoever, as shown in controlled trials in prison populations decades ago (15, 16).

The hormonal mood issues relate to downstream amygdala-driven irritability and emotional reactivity. Lithium’s best-documented effect is dampening exactly that, regardless of what’s driving it upstream – which is exactly why it doesn’t seem to matter which version of the underlying cause a given client has. But to be clear, as good as it is, lithium orotate is only working on the downstream effects of the root-cause issues – the root causes still need to be addressed for long-term health and vitality.

REFERENCES

1. Prior, J.C. Perimenopause hormone review – https://storage.imrpress.com/imr/journal/FBS/article/495448/1752777891274.pdf
2. Oestrogen effects on GABA/glutamate – https://pmc.ncbi.nlm.nih.gov/articles/PMC6737530/
3. Sex hormones and neurotransmitters review (dopamine) – https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2015.00037/full
4. Glutamate and sensory hyper-reactivity (autism) – https://pmc.ncbi.nlm.nih.gov/articles/PMC8322079/
5. Progesterone metabolite allopregnanolone in PMS – https://pubmed.ncbi.nlm.nih.gov/9351749/
6. Timby et al., 2025, delta subunit/amygdala in PMDD – https://pmc.ncbi.nlm.nih.gov/articles/PMC12287299/
7. Lithium and presynaptic GABA release – https://pmc.ncbi.nlm.nih.gov/articles/PMC8069239/
8. Hashimoto et al. 2002, NMDA receptor mechanism – https://pubmed.ncbi.nlm.nih.gov/11841566/
9. Nonaka et al. 1998, PNAS glutamate/NMDA – https://www.pnas.org/doi/full/10.1073/pnas.95.5.2642
10. Lithium mechanisms review (bipolar disorder) – https://pmc.ncbi.nlm.nih.gov/articles/PMC5125816/
11. PET imaging of 5-HT1A with mood stabilisers – https://www.nature.com/articles/s41398-026-03868-z
12. Cipriani et al., lithium and suicide prevention – https://pmc.ncbi.nlm.nih.gov/articles/PMC12530009/
13. Lithium and amygdala/hippocampal volume – https://pubmed.ncbi.nlm.nih.gov/19573050/
14. Brain imaging review, lithium structural/connectivity – https://pmc.ncbi.nlm.nih.gov/articles/PMC9537229/
15. Sheard et al., impulsive aggression trial – https://pubmed.ncbi.nlm.nih.gov/984241/
16. Lithium as an antiaggressive agent independent of bipolar diagnosis – https://pubmed.ncbi.nlm.nih.gov/1639003/