PCOS Period Pain Relief: What the Trials Actually Show

Why does it matter that these trials weren’t done in PCOS?

Almost every trial below was run in “primary dysmenorrhoea” — cyclical cramping that follows ovulation — not in women selected for a PCOS diagnosis, and that gap matters more here than usual. The article on why PCOS period pain happens covers the mechanism in full: many PCOS cycles are anovulatory, so there is often less of the prostaglandin surge that drives textbook cramping. If your pain is mild, comes and goes, and doesn’t reliably track a bleed, the options below are treating a pain type your cycle may not generate as strongly. If your cramping is the familiar day-one, worsens-then-eases kind, the evidence below transfers more directly.

PCOS was renamed polyendocrine metabolic ovarian syndrome (PMOS) in May 2026 by a global consensus of more than 50 medical organisations. Same condition, same mechanism — this article uses PCOS because that is still what most readers search.

How do NSAIDs actually work on period pain, and how well?

NSAIDs block the enzyme that manufactures prostaglandins, the compounds that drive the uterine contractions behind classic cramping — a mechanistic reason they work, not just an empirical one. Elevated prostaglandin levels are the documented driver of primary dysmenorrhoea, and cyclooxygenase inhibitors — the drug class NSAIDs belong to — reduce that release directly, which is why they’re the established first-line option for this kind of pain.

A 2015 Cochrane review pooling 80 randomised trials in 5,820 women with primary dysmenorrhoea found NSAIDs significantly more effective than placebo (odds ratio 4.37, 95% CI 3.76 to 5.09) — if 18% on placebo get moderate-to-excellent relief, roughly 45–53% do on an NSAID. NSAIDs also outperformed paracetamol (OR 1.89, three smaller trials), and no individual NSAID stood out as clearly safer or more effective than another.

That effectiveness carries a real cost — the same review found NSAID users had more adverse effects overall (OR 1.29), more gastrointestinal side effects (OR 1.58), and more than twice the rate of neurological side effects such as headache or drowsiness (OR 2.74) than placebo. “Very effective but not free” is what that looks like in numbers.

Does heat work as well as a pill?

Heat matched oral ibuprofen for pain relief in the one well-designed trial that tested them head-to-head. In an 84-person randomised, placebo-and-active-controlled trial, a continuous low-level heat patch worn roughly 12 hours a day produced relief (mean score 3.27) statistically indistinguishable from oral ibuprofen alone (mean score 3.07) — both significantly better than an unheated patch with a placebo pill (mean score 1.95, p < 0.001 for both). Combining heat with ibuprofen didn’t improve relief over ibuprofen alone (p = 0.096), but it worked faster: median time to noticeable relief was 1.5 hours for the combination versus 2.79 hours for ibuprofen alone (p = 0.01).

What about exercise — how much, and how reliably?

Exercise produced a large pooled reduction in period pain across dozens of small trials, though the exact number is a direction, not a promise. A 2025 systematic review of 29 randomised trials in people with primary dysmenorrhoea found exercise reduced pain scores significantly versus usual care or no intervention (weighted mean difference −2.62, p < 0.001). Sessions longer than 30 minutes, more than three times a week, programmes of eight weeks or longer, and a weekly total of at least 90 minutes each showed a larger effect than shorter or less frequent routines, with strength training the strongest category.

Honest caveat: most of the 29 trials were small and heterogeneous, so “exercise reliably helps” is fair; “exercise cuts pain by exactly 2.62 points” is not.

Is TENS worth trying?

TENS reduced period pain compared with placebo in a current Cochrane review, through a mechanism that doesn’t touch prostaglandins at all. The 2024 Cochrane review of 20 trials in 585 women found high-frequency TENS reduced pain versus placebo or no treatment (mean difference −1.39, 95% CI −2.51 to −0.28; low-certainty evidence) and low-frequency TENS did too (mean difference −2.04, 95% CI −2.95 to −1.14; low-certainty evidence). TENS interferes with pain-signal transmission at the spinal cord and prompts release of the body’s own opioid-like pain relievers — worth layering alongside the others rather than instead of them. Certainty was rated low throughout, due to risk-of-bias in the underlying trials, and the review couldn’t say whether TENS beats acupressure, paracetamol, or NSAIDs specifically — only that it beats nothing.

Does magnesium help period pain?

Magnesium’s period-pain evidence is old, small, and hasn’t been repeated in a modern placebo trial — “popular supplement, thin proof” is the honest summary, not “proven to work.” A 1992 Italian trial gave 30 women with primary dysmenorrhoea magnesium pidolate across six cycles, using each woman’s own untreated first cycle as her baseline rather than a separate placebo group. Pain on the first day of bleeding dropped significantly by the sixth treated cycle (p < 0.05); days two and three trended the same way without reaching significance. A 30-woman, non-placebo-controlled design from over three decades ago doesn’t support a confident recommendation, and no larger trial has replaced it since. The proposed mechanism — uterine muscle relaxation and reduced local prostaglandin F2-alpha — is biologically plausible; plausible isn’t the same as demonstrated at scale.

That’s a different question from magnesium’s role in PCOS more broadly: the evidence on magnesium for PCOS-specific insulin and hormone markers is its own thin picture, covering different outcomes than period pain.

Where do hormonal options fit?

Hormonal contraception works through a mechanism the options above don’t touch — suppressing ovulation removes the progesterone-withdrawal signal behind prostaglandin release, rather than blocking prostaglandins once they exist. A 2023 Cochrane review of 21 randomised trials in 3,723 women with primary dysmenorrhoea found combined oral contraceptives produced a moderate reduction in pain versus placebo (standardised mean difference −0.58, 95% CI −0.74 to −0.41; 6 trials, 588 women; high-quality evidence), and improved the odds of reporting relief overall (risk ratio 1.65; 6 trials, 717 women; lower-quality evidence). Continuous dosing was probably more effective than the standard cyclic regimen, though long-term safety data on continuous use is limited, and the review couldn’t say with confidence whether NSAIDs or contraceptives work better head-to-head — only one small trial has ever compared them directly.

The 2023 international PCOS guideline centres on diagnosis, metabolic risk, and fertility; period-pain treatment isn’t something it addresses in detail, so the grading here comes from dysmenorrhoea trials and Cochrane reviews in the general population, not a PCOS-specific recommendation. Combined contraception itself is not right for everyone — it is not recommended with a personal history of blood clots or VTE, migraine with aura, or smoking past age 35, when a progestin-only option is the usual alternative. Which option makes sense depends on that history, whether pregnancy is a current goal, and — for pain alongside heavy or unpredictable bleeding — the fuller table of contraceptive and progestin options already covering that pattern. This is a decision for a clinician who knows your history, not a table alone.

Who this relief plan will not work for

None of this helps if the mechanism driving your pain isn’t the one these trials targeted, worth naming plainly rather than implying one plan fits every PCOS period. If your pain comes from the anovulatory, no-bleed pattern — cramping without a period, driven by unopposed oestrogen or a stalled follicle rather than a prostaglandin surge — NSAIDs have a less direct target, though heat, exercise and TENS may still help by acting on pain generally. And if pain is progressively worsening cycle over cycle, occurs with sex or bowel movements, or simply isn’t responding to anything here, another remedy from this list won’t sort that out — that pattern warrants its own evaluation, and which condition explains it is a question for a clinician and the right tests, not something a symptom article can settle.

Your next step

Track which relief method you try alongside the pain-pattern log from the main period-pain article — same side, same timing, how it responds. That record shows which option works for your pattern, and gives a clinician what they need if it changes. For the wider set of PCOS symptoms worth tracking, see the symptoms hub.