Researchers at Uppsala University tracked 90 women with PMDD and 44 women without the disorder across two consecutive menstrual cycles, collecting daily self-reported symptoms. Computer models revealed three distinct subtypes among those with PMDD, suggesting the condition, estimated to affect 3 to 8 percent of people who regularly menstruate, is more heterogeneous than it is often treated as being.
The first subtype, labeled 'severe full luteal', involves symptoms that begin shortly after ovulation and gradually become more severe in the weeks leading up to menstruation, typically ceasing once bleeding starts. The authors found that irritability commonly follows this pattern.
A second subtype, 'severe late luteal', sees symptoms start later in the luteal phase and ramp up quickly to become severe, and they can linger for four to five days after bleeding begins. Anxiety commonly follows this trajectory.
The third subtype, 'moderate late luteal', involves symptoms that appear just before menstruation and are generally less severe, disappearing soon after bleeding starts. Depression, anxiety, and mood swings commonly follow this timeline. Several physical symptoms, including joint pain and headaches, also follow it, though physical symptoms tend to be less severe than in other patterns.
The study found that 98 percent of participants with PMDD experienced more than one symptom subtype during the two cycles tracked. The researchers noted that to date, PMDD symptomatology and its temporal onset and offset have generally been treated as homogeneous across and stable within individuals, with little attention given to inter- and intra-individual heterogeneity.
A previous study in 2019 identified three very similar symptom trajectories among 74 individuals prospectively diagnosed with PMDD, but it tracked only one menstrual cycle.
The study is small and ovulation was not directly measured, but the authors argue their findings support the idea that PMDD is a complex, cyclical disorder rooted in psychoneuroendocrine dynamics. They argue that by integrating dynamic symptom modeling with clinical practice, precision in mental health care for people who menstruate could advance and reduce the burden of PMDD.
The researchers noted that a DSM-5 diagnosis currently depends on PMDD worsening in the week before menstruation and improving shortly after bleeding begins, but symptoms do not always present this way.
Awareness of PMDD is poor, the article says. A 2026 editorial in the British Journal of General Practice argued that while there is overlap between PMDD and PMS, and current guidelines largely recommend similar management approaches, the severity and cyclical nature of symptoms may not be fully recognized. The editorial's researchers added that patients with PMDD may be managed in the same way as those with milder PMS, which may lead to under-recognition of the impact on functioning, delayed escalation of treatment, and a lack of appropriate follow-up.
Even when properly diagnosed, treatment options are limited. Psychological support such as cognitive behavioral therapy may be recommended, and pharmacological treatments such as oral contraceptive pills or selective serotonin reuptake inhibitors could be advised, but a significant percentage of patients do not respond to available therapies. A 2021 meta-analysis in the Journal of Women's Health found that PMDD was associated with nearly sevenfold higher odds of suicide attempts.
The researchers argue that acknowledging differences in symptom timelines could help explain why some patients respond to treatment and others do not.
