Magnesium is the second most-discussed PCOS/PMOS supplement after inositol, but almost everyone asking "is it working?" is looking at the wrong signal. The scale doesn't tell you. Your GP's serum magnesium test doesn't tell you — 85% of the body's magnesium is intracellular, which means serum levels can look perfectly normal while you're functionally deficient. The standard blood test only captures the magnesium floating free in your plasma. It misses the vast majority sitting inside your cells doing the actual work.
The only way to know whether magnesium is doing anything useful in a PCOS/PMOS body is to track what magnesium actually affects: sleep quality, cravings, insulin response, muscle tension, and cycle regularity. These are the signals that move first. If you're supplementing and not watching them, you're essentially running an experiment with your eyes closed.
Why PCOS/PMOS and Magnesium Are Deeply Connected
The connection between PCOS/PMOS and magnesium isn't incidental — it runs through four distinct biological mechanisms, each of which maps to a symptom cluster you're probably already experiencing.
Insulin sensitisation
Magnesium is a cofactor for over 300 enzymatic reactions, including those that govern GLUT4 transporter activity — the mechanism by which your cells take up glucose from the bloodstream. Low magnesium impairs glucose uptake independently of insulin. That matters enormously for PCOS/PMOS, because it means that even if metformin, lifestyle changes, or other interventions are working on the insulin signalling side of the equation, functional magnesium deficiency can blunt the effect by limiting what happens at the cellular end. Studies consistently show that women with PCOS/PMOS are twice as likely to be functionally magnesium-deficient compared to the general population — a figure that makes sense given how heavily the condition taxes the metabolic pathways magnesium supports.
If you've been tracking your insulin resistance markers and wondering why progress feels slower than expected, this is worth investigating. The complete insulin resistance tracking guide for PCOS/PMOS covers the full picture of what to measure and when — and magnesium status is one of the inputs the guide points back to repeatedly.
HPA axis and cortisol regulation
Magnesium acts as a natural brake on the HPA axis — the hypothalamic-pituitary-adrenal pathway that governs your cortisol response. When magnesium is low, that brake is less effective. The result is elevated baseline cortisol, which worsens insulin resistance, which worsens androgen production, which worsens PCOS/PMOS. This is the feedback loop that explains why stress doesn't just make you feel worse — it actively tanks your cycle.
The relationship between sleep, cortisol, and PCOS/PMOS runs through this exact pathway. Magnesium is one of the few nutrients that sits at the intersection of both the cortisol and sleep branches of that loop. Fixing one without addressing magnesium status often produces partial results that are frustrating to interpret.
Sleep architecture
Magnesium activates GABA receptors — the inhibitory neurotransmitter system that quiets the nervous system and enables the transition into deep sleep. Specifically, it promotes slow-wave sleep: the restorative deep sleep phase where cellular repair, hormone regulation, and memory consolidation all happen. Women with PCOS/PMOS disproportionately report poor sleep quality even when their total sleep duration is adequate. They go to bed on time, lie there for eight hours, and wake up exhausted. This disconnect between duration and quality is often magnesium-mediated — adequate hours of sleep, but not enough of the right kind of sleep.
This is also why sleep quality is one of the most sensitive early indicators of magnesium sufficiency. Changes to slow-wave architecture happen within the first few weeks of correcting a deficiency, well before other hormonal markers shift.
Progesterone synthesis
Magnesium is required for the conversion of cholesterol to pregnenolone — the precursor to progesterone. Low magnesium means blunted progesterone response, which tips the oestrogen-to-progesterone ratio toward oestrogen dominance, which worsens the full PCOS/PMOS symptom picture. This mechanism is particularly relevant for women with luteal phase defects, consistently short cycles, or mid-cycle spotting — all of which point toward insufficient progesterone in the second half of the cycle. If you've been told your progesterone is "a bit low" on a blood test, magnesium status is one of the upstream variables worth addressing before assuming the issue is purely hormonal.
Five Atlas Check-In Signals to Watch Over 12 Weeks
Rather than trying to measure magnesium directly — which, as noted, the standard serum test handles poorly — the approach that actually works is tracking the downstream signals that respond to magnesium status. There are five in Atlas that are worth logging consistently across a 12-week window.
Sleep quality rating, tracked as a 7-day rolling average, maps most directly to the GABA and slow-wave sleep mechanism. This is typically the first signal to move when magnesium levels improve — often within 2–3 weeks of consistent supplementation. The 7-day average is important because nightly sleep quality is noisy; the trend only becomes legible across a week or more.
Night-time cravings frequency sits at the intersection of the insulin sensitisation and cortisol interplay mechanisms. When blood glucose regulation improves and baseline cortisol drops, the urge to eat in the hours after dinner — the classic PCOS/PMOS craving window — typically decreases in frequency and intensity. This signal usually moves after sleep quality, in the four-to-six-week range.
Energy on waking is a distinct measure from sleep duration and maps to sleep architecture quality rather than simply how many hours you spent in bed. If you're sleeping eight hours but waking exhausted, and that pattern begins to improve, it's a meaningful signal — not just that you slept longer, but that the quality of those hours shifted.
Cycle length regularity connects to the progesterone synthesis mechanism. Cycles that have been consistently short (under 25 days) or long (over 35) beginning to drift toward a more regular pattern are a strong downstream signal. This one takes the longest to move — typically 8–12 weeks — and it's worth noting any change in direction even before a full normalisation appears.
Mood and anxiety score trend maps to the HPA axis braking effect. Baseline anxiety — the low-level background hum of it, distinct from acute stress responses — often improves as cortisol regulation improves. The connection between stress, cortisol, and PCOS/PMOS data explains in detail why this signal is worth tracking separately from energy and sleep, as it often moves on its own timeline.
Three 12-Week Tracking Experiments
The following three experiments are designed to be run sequentially or independently, depending on where you are in your supplementation journey.
Baseline first
Before you take a single magnesium supplement, track all five signals — sleep quality, night-time cravings, waking energy, cycle length, and mood/anxiety — for two full weeks. This gives you a genuine baseline to compare against, which most women never have. After two weeks, introduce magnesium glycinate at 310–360mg elemental magnesium at night. Glycinate is the form to choose: it's better absorbed than magnesium oxide and far less likely to cause the GI upset (loose stools, urgency) that oxide is notorious for. Per NHS guidance, 270mg/day is the standard recommendation for women, but therapeutic doses used in PCOS/PMOS studies range from 300–400mg elemental magnesium daily. Compare your five signals in 2-week blocks across 12 weeks. The data, not the feeling, is the answer.
Timing experiment
After a baseline period, take your full dose at night for six weeks, then split morning and evening for six weeks. Sleep quality rating is the key variable here. Many women find that concentrating the full dose at night has a measurably larger effect on sleep architecture than splitting it — the logic being that a higher single dose gives the GABA mechanism a bigger activation window during the sleep period itself. But this varies, and the only way to know how your body responds is to run both conditions and compare. Your Atlas sleep quality trend across the two 6-week blocks will tell you which timing works better for you specifically.
Diet-first experiment
Before or alongside supplementation, use the Atlas food log to track dietary magnesium for two weeks. The highest-magnesium foods are dark leafy greens (particularly spinach and Swiss chard), pumpkin seeds, dark chocolate at 70% or above, almonds, and black beans. If your dietary intake is genuinely low, improving it through food first is a useful data point — not least because it gives you something concrete to tell your GP. If your sleep quality improves meaningfully before you start supplementing, that's evidence of a dietary gap that a GP is more likely to take seriously than "I thought I might be deficient." It also gives you the satisfaction of a clear causal chain, which matters when you're making changes that are hard to sustain.
What to Tell Your GP
Most GPs will order a serum magnesium test if you ask. This is worth doing as a starting point, but it's important to understand what it can and can't show. Serum magnesium reflects the magnesium in your blood plasma, not inside your cells. A normal result doesn't rule out functional deficiency. What you're really looking for is an intracellular magnesium test — specifically an RBC (red blood cell) magnesium test, which measures magnesium inside the cells and gives a much more accurate picture of functional status. Not every GP will be familiar with this, but it's a legitimate test available in the UK and worth requesting by name.
When you go, bring your Atlas data. Your sleep quality trend over 8–12 weeks, your cycle length data, and your craving frequency log are exactly the kind of longitudinal pattern data that makes a GP's job easier. A single symptom conversation is easy to dismiss; a chart showing consistent improvement after introducing magnesium is harder to ignore.
It's also worth noting that magnesium supplementation at NHS-recommended doses is generally very safe, but it can interact with certain antibiotics (particularly tetracyclines and fluoroquinolones) and with some diabetes medications. If you're on any prescription medication, mention your magnesium supplementation to your GP or pharmacist before starting. The NHS recommends 270mg/day for women as a general intake target; therapeutic doses used in PCOS/PMOS research range from 300–400mg elemental magnesium per day. The gap between those numbers is not enormous, but it's worth being aware of when you're deciding on a dose.
For women who've also been tracking their inositol supplementation, the inositol tracking guide for PCOS/PMOS covers a similar structure — baseline, mechanism, signals — and the two protocols can be run in parallel without interfering with each other.
Start With Data, Not Hope
Magnesium is not a silver bullet. Nothing in PCOS/PMOS management is. But it is one of the few supplements with a plausible, well-evidenced mechanistic connection to PCOS/PMOS biology at multiple levels simultaneously — insulin sensitisation, HPA axis regulation, sleep architecture, and progesterone synthesis. That breadth of effect is genuinely unusual, and it's why the supplement consistently appears near the top of community discussions.
The reason most women can't tell if it's working is that they're not tracking the right signals. They look at the scale, or they wait for a GP test that measures the wrong thing, or they supplement for three weeks and conclude it didn't help because they don't feel dramatically different. Magnesium doesn't work dramatically. It works gradually, through improved sleep quality, slowly reducing baseline anxiety, fewer night-time cravings, and — eventually — more regular cycles.
With five specific Atlas metrics, a genuine two-week baseline, and 12 weeks of consistent data, you'll know whether it's working. Not guess. Not hope. Know.