SSRIs — a class of antidepressants that includes medications like sertraline, fluoxetine, and paroxetine — are one of the most commonly discussed treatments for PMDD and severe PMS. This large Cochrane review, updated in 2024, pulled together evidence from 34 randomized controlled trials involving over 4,500 women to assess how well these medications work and what side effects they carry.
The findings show that SSRIs probably reduce premenstrual symptoms, though they do not eliminate them entirely. The effect was moderate overall. When taken every day throughout the menstrual cycle (continuous dosing), SSRIs appeared to work somewhat better than when taken only during the two weeks before a period (luteal phase dosing). Both approaches showed benefits. All five types of SSRIs tested appeared to be helpful, and the treatment worked for both PMDD and PMS.
Side effects were common. About 1 in 8 women experienced nausea beyond what would be expected with a placebo, and about 1 in 9 experienced decreased energy. Other frequently reported side effects included sleepiness, dry mouth, sexual difficulties, insomnia, dizziness, and sweating. The review could not determine which side effects were temporary and which persisted over time, which is an important gap since many people take these medications for months or years.
Some caution is warranted when interpreting these results. More than two-thirds of the studies were funded by pharmaceutical companies, and there were signs that studies showing more favorable results may have been more likely to be published. Most studies lasted only a few months and were conducted in Western countries. The overall quality of evidence was rated as moderate, mainly because many studies did not clearly report their methods. Still, this remains one of the strongest bodies of evidence available on SSRI treatment for premenstrual conditions.
Key findings
- SSRIs probably reduce overall self-rated premenstrual symptoms compared to placebo (SMD −0.57, 95% CI −0.72 to −0.42; 12 studies, N=1742; moderate-certainty evidence)
- Continuous SSRI administration was probably more effective than luteal phase-only dosing (continuous SMD −0.69 vs luteal SMD −0.39; P=0.03 for subgroup difference)
- Nausea was the most common adverse effect with SSRI use (OR 3.30, 95% CI 2.58 to 4.21; NNTH 8), meaning approximately 1 in 8 women taking SSRIs experienced nausea beyond baseline risk
- SSRIs probably increased risk of all 12 assessed adverse events including sexual dysfunction (OR 2.32), insomnia (OR 1.99), asthenia/decreased energy (OR 3.28), and somnolence (OR 3.26)
- SSRIs were probably effective for both PMDD (SMD −0.57) and PMS (SMD −1.00), with no statistically significant subgroup difference (P=0.30)
- Response rates were significantly higher with SSRIs versus placebo (OR 2.45, 95% CI 2.04 to 2.94; 21 studies, N=3502), corresponding to 61% response in SSRI group versus 39% in placebo group
Methods, briefly
Cochrane systematic review and meta-analysis of 34 RCTs (N=4563 women with PMS or PMDD). Studies compared SSRIs (fluoxetine, paroxetine, sertraline, escitalopram, citalopram) to placebo. Treatment duration ranged from 2 to 6 menstrual cycles. Data pooled using random-effects model with standardised mean differences for continuous outcomes and odds ratios for dichotomous outcomes. GRADE assessment used. Searches conducted through November 2023 across CENTRAL, MEDLINE, Embase, PsycINFO and other databases. This is an update of a review last published in 2013.
Limitations to keep in mind
- 68% of included studies were funded by pharmaceutical companies, which may bias results in favor of SSRIs
- Poor reporting of study methodology — only 7 of 34 studies had adequate randomisation and allocation concealment
- Studies were predominantly conducted in Western settings (USA, Europe, Canada, Australia), limiting generalisability
- Treatment duration was short (2-6 cycles), providing no data on long-term efficacy or safety despite PMDD being a chronic condition
- Suspected publication bias was identified in the funnel plot for response rates
- Various definitions and diagnostic criteria for PMS and PMDD were used across studies
- Unable to distinguish between transient and persistent adverse effects from available data
The latest PMDD research, in your pocket
The app’s research feed is updated daily with newly published, peer-reviewed PMDD work - summarized in plain language.