Heavy periods: causes, anaemia and treatment

Older studies used a threshold of about 80 mL of blood lost per cycle, measured by a chemical method (alkaline hematin), but that is a research tool not available in the clinic. Clinically, the approach described in the Cochrane overview matters more: heavy menstrual bleeding is excessive blood loss that interferes with a woman's quality of life, regardless of the absolute amount. It means that what makes a period heavy is its effect on daily functioning, not merely the number of pads used.

A subjective estimate of the amount of blood is unreliable, which is why helper methods were developed. A systematic review of 71 studies on ways to measure menstrual blood loss found that every method has limitations, and that the best compromise between ease of use and accuracy comes from pictorial methods such as the pictorial blood assessment chart (PBAC). The sensitivity and specificity of the PBAC for identifying HMB fell in wide ranges (58-99% and 7.5-89% respectively), which shows that no score replaces a conversation about the effect of bleeding on life.

These data come from a specialist referral centre, so the proportions with anaemia and bleeding disorders are higher than they would be in the general population of women with heavy periods. They nonetheless illustrate two practical points. First, the great majority of women with HMB had low iron stores (ferritin below 30 µg/L in about three quarters), so the problem is far wider than anaemia alone. Second, merely prescribing iron does not solve the cause: if bleeding stays heavy, the iron stores will empty again.

The international federation FIGO organised the causes of abnormal uterine bleeding in the acronym PALM-COEIN. The 'PALM' part covers structural causes that can be seen on imaging or histology, and the 'COEIN' part covers non-structural causes. The classification is useful because it shows that in many women there is no 'visible' change, yet the bleeding is a real problem.

A meta-analysis of the aetiology of HMB, covering 45 studies and 41,541 patients, estimated that bleeding disorders were found on average in 30% (95% CI 14-46), von Willebrand disease in 8% (7-10), platelet function defects in 9% (7-12), abnormal thyroid function in 3% (0-6) and polycystic ovary syndrome in 8% (4-12). The wide confidence intervals reflect great diversity of the populations and methods studied, so we treat these figures as an order of magnitude and not an exact frequency.

The shared conclusions of three sources (the 2004 review, the 2025 meta-analysis and the 2026 French series) are consistent: von Willebrand disease and other bleeding disorders are clearly more common among women with HMB than in the general population, but most women do not have a bleeding disorder. In the 2025 meta-analysis bleeding disorders affected adolescents far more often (39%; 95% CI 18-60) than adults (16%; a wide confidence interval from -8 to 41), which calls for particularly careful questions about bleeding symptoms in young girls with heavy periods from their first period.

Iron deficiency develops in stages: first the stores are depleted (ferritin falls), then iron availability to tissues decreases, and only at the end does haemoglobin fall and anaemia is diagnosed. That is why with heavy periods the blood count alone may be normal even though stores are low. Symptoms of deficiency, such as fatigue, hair loss, poor concentration, restless legs or pallor, often appear before anaemia, but they are nonspecific, so their presence alone is not enough for a diagnosis.

The ferritin result and its thresholds are described more fully in our entries on ferritin and the blood count. We note that interpreting coagulation results is difficult: von Willebrand factor levels vary with blood group, stress, inflammation and cycle phase, so a single normal result does not always exclude the disease and is sometimes repeated.

This overview has important limitations: it rests on studies of low or moderate certainty, and effects on perception of improvement and satisfaction were uncertain (very low certainty) for all first-line methods. It is also worth remembering that an 'uncertain effect' for combined contraceptive pills does not mean they do not work: it means that the quality of the available studies does not permit a firm conclusion. In practice the choice of method depends on childbearing plans, the need for contraception, contraindications and the woman's preferences.

Procedures are a second-line option when conservative treatment fails, is poorly tolerated, or when a structural change needing removal has been found (a polyp, a submucosal fibroid). The Cochrane overview indicates that hysterectomy gave the largest reduction in blood loss and high satisfaction, and endometrial ablation (both resectoscopic and non-resectoscopic) also reduced bleeding, with moderate certainty of evidence for the non-resectoscopic methods. The differences lie in irreversibility and risk, however: hysterectomy is major surgery and rules out pregnancy, while after ablation pregnancy is inadvisable and may be dangerous, so it requires effective contraception.

The decision on a procedure should take into account childbearing plans, age, the cause of bleeding, previous attempts at treatment and the woman's preferences. If the cause is an unrecognised bleeding disorder, the risk of perioperative bleeding may be increased, which is why a history in this direction is important before any procedure.

Heavy periods are one of those health problems that are easy to dismiss because they are widespread, yet they respond well to treatment. We recommend treating them as a symptom that deserves a simple, orderly work-up: history and a bleeding diary, blood count and ferritin, TSH, ultrasound and, where indicated, coagulation tests. Treatment should be matched to the cause and to life plans, not the other way round. Data from the Cochrane overviews show that the levonorgestrel intrauterine system and tranexamic acid are the two best-documented conservative routes, and that surgery is a solution, not a failure.