Radiofrequency device beat sham for postmenopausal sexual function in a 62-woman trial, but effect sizes look implausibly large
Journal: Menopause (New York, N.Y.) | Published: 2026-07-01 | Type: Randomized Controlled Trial | PMID: 41665605 Authors: Quezada-Bascuñán C, Ferri-Morales A, Martínez-Vizcaíno V, Gallego-Gómez C, Torres-Costoso A, Lirio-Romero C (Faculty of Physiotherapy and Nursing, University of Castilla-La Mancha, Toledo/Cuenca, Spain) Funding/COI: Funding source not listed. No conflicts of interest reported.
Sixty-two postmenopausal women with genitourinary syndrome of menopause (GSM) were randomized to six weekly sessions of capacitive-resistive monopolar radiofrequency or a sham procedure. The treatment group scored better than sham on both the Female Sexual Function Index (FSFI) and Vaginal Health Index (VHI) at the end of treatment and 12 weeks later. Vaginal cytology, a marker of estrogenic tissue change, didn't budge in either group, so whatever the device is doing, it isn't restoring estrogenic status.
This is a randomized, sham-controlled trial with a reasonable sample for a physical therapy intervention (32 vs 30), pre-specified outcome measures at three time points, and a 12-week follow-up to check durability. The use of a sham comparator rather than no treatment is a real methodological strength over a lot of pelvic floor device literature.
But the VHI effect size is a problem. A Cohen's d above 3.49 is not a "large effect" in any normal clinical sense, it's an effect size you almost never see outside of near-perfect group separation with minimal overlap in scores. That magnitude is more consistent with detection bias than a genuine physiological effect, especially since VHI is scored by a clinician examining the vagina, and radiofrequency sessions plausibly leave visible or palpable signs (warmth, tissue response) that could unblind the assessor. The paper is labeled "single-blind" but doesn't specify who was blinded, participant, assessor, or both, which matters a lot here since FSFI is entirely self-reported and susceptible to expectation effects if patients could tell which arm they were in.
The null finding on vaginal cytology cuts against the treatment's proposed mechanism (tissue remodeling toward a more estrogenized state) while the subjective/assessor-scored outcomes moved dramatically. That inconsistency deserves more scrutiny than the discussion appears to give it.
Worth a skeptical read, not a skip. The trial design (sham-controlled, multi-timepoint, multi-domain) is better than most device studies in this space, but the VHI effect size is so far outside normal clinical ranges that it raises more questions about blinding integrity than it answers about efficacy, and the disconnect between subjective improvement and unchanged vaginal cytology means the mechanism claimed in the conclusion isn't actually supported by this study's own data.