MOST trial: modified sound therapy, 4-arm double-blind RCT (DFCRS arm)

Record details

• Authors: Tang D, Gu D, Gong J, et al.
• Journal / source: eClinicalMedicine (2025)
• Design: 4-arm multicentre double-blind RCT (participants + clinicians blinded); daily 2-h listening for 9 months; NO no-treatment or sham arm — unmodified music (UM) is the reference
• Sample: 440 randomized (UM 111 / UM+NBN 110 / HFEM 108 / DFCRS 111), all ITT
• Primary endpoint: Two-part: complete remission (THI=0 at any visit within 9 months), else THI change baseline→9 months
• Randomized: Yes · Sham/placebo-controlled: No · Independent of the manufacturer: Yes

What this research found

The largest double-blind sound-therapy RCT. Only 1 of 440 participants reached complete remission. All arms improved (overall THI median 50→35, p<0.0001 — within-group). Between arms (GEE vs unmodified music): the customized DFCRS arm was better by −4.37 THI points [95% CI −6.25, −2.48], p<0.0001 — statistically significant but SMALL; fixed high-frequency enhancement (HFEM) was WORSE than unmodified music (+2.11, p=0.039); narrowband-noise mixing added nothing (p=0.75).

VAS-loudness fell in ALL arms (DFCRS greater in the first 3 months); the design cannot separate sound-specific benefit from placebo/natural fluctuation — the authors say so explicitly. No multiplicity adjustment. No adverse events in any group. ChiCTR2000039007; Chinese public funding; no competing interests declared. Registered on ChiCTR — invisible to CTgov-only discovery (audit finding).

Limitations and conflicts of interest

No no-treatment/sham arm; placebo effects cannot be excluded; no multiplicity adjustment; adherence not objectively monitored; 3-month posttreatment follow-up; single-region (Shanghai).

Funding / conflicts: Independent (public funding; no competing interests declared).

Primary sources

• PubMed record (41497514)
• Publisher (DOI 10.1016/j.eclinm.2025.103671)
• Source link

Used as evidence for

• Sound therapy & masking