Five-element music therapy based on emotion classification to improve sleep in patients with cancer-related depression: a multi-group randomized controlled trial

Wenjun Wang, Yong Sheng, Hongmei Xie, Jing‐Tao Zhao, Yang Zhao, Rui Zhou, Nan Zhao, Yu Wu

This study, published in Frontiers in Psychiatry, explores a personalized approach to a traditional intervention. Researchers conducted a multi-group randomized controlled trial to see if tailoring "Five-Element Music Therapy" (FEMT) to a patient’s specific emotional state could better alleviate sleep disorders and depression in cancer patients compared to standard music therapy.

TL;DR

By matching music to a patient's subjective emotional state using a novel classification software, researchers found that personalized music therapy improves sleep quality and mood in cancer patients significantly faster than traditional methods. This non-invasive, drug-free approach offers a promising adjuvant treatment for cancer-related depression.

What Problem Does This Paper Address?

Cancer-related depression (CRD) affects up to 69% of patients, often bringing severe sleep disturbances that cripple quality of life. While drugs like benzodiazepines are commonly prescribed, they carry risks of sedation, cognitive impairment, and addiction.

Traditional Chinese Medicine (TCM) has long used Five-Element Music Therapy (FEMT)—where different musical scales correspond to internal organs and emotional states—as a non-pharmacological alternative. However, standard FEMT is often "one-size-fits-all," ignoring the patient's immediate emotional needs. The authors sought to bridge this gap by creating an "emotion-based" FEMT protocol.

Methods

The researchers recruited 120 CRD patients and randomly assigned them to three groups:

  1. Emotional Classification Music Group (ECMG): Used custom software to select music from a library of 153 pieces based on their current emotional state and TCM syndrome. The protocol followed a specific "induction-then-calming" sequence.
  2. Traditional Music Group (TMG): Received standard FEMT without emotional matching or a specific playback order.
  3. No Music Group (NMG): Received routine oncological care (control).

The intervention lasted 4 weeks (30–40 minutes daily). Outcomes were measured using the Pittsburgh Sleep Quality Index (PSQI) for sleep, and the Hamilton scales (HAMD/HAMA) for depression and anxiety. They also tracked exploratory biological markers like serotonin (5-HT) and inflammatory cytokines.

Datasets

As a clinical trial, the primary "dataset" consists of the 120-patient cohort (reduced to 117 after minor dropouts).

  • The Music Library: A curated set of 153 music pieces screened and sorted by a panel of six experts in TCM, psychology, and musicology. This library served as the foundation for the personalized software intervention.

Key Insights and Ideas

  • Personalization Accelerates Recovery: The "aha moment" here is that while both music groups eventually improved, the personalized group (ECMG) saw significant sleep improvements by Day 14, whereas the traditional group (TMG) didn't reach similar levels until Day 28.
  • The "Induction-Calming" Strategy: The ECMG protocol wasn't just about "happy" music. It used a sequence of "emotional induction" (allowing the patient to express/process feelings) followed by "emotional calming." The authors suggest this might better regulate the autonomic nervous system than static listening.
  • Gender Sensitivity: Subgroup analysis suggested that female patients might be more sensitive to the therapeutic effects of FEMT, showing greater magnitude of change in sleep scores compared to male patients.

Key Results

  • Sleep Quality: Both ECMG and TMG significantly improved PSQI scores compared to the control. ECMG was notably faster, improving sleep disorder scores by Day 14 ($P=0.0016$).
  • Mood Regulation: Both music therapies were highly effective at reducing depression and anxiety. Secondary outcomes (HAMD/HAMA) showed high statistical power ($Pwr > 0.99$).
  • Biological Trends: In the ECMG group, researchers observed an upward trend in serotonin (5-HT) and IL-2 (an immune-regulating cytokine), while the other groups showed downward trends. However, these results were exploratory and not statistically definitive.

Limitations and Future Work

The authors are transparent about several hurdles:

  • Sample Size and Duration: The study was relatively small (120 participants) and short-term (4 weeks). It’s unclear if the benefits persist months after the therapy ends.
  • Blinding Challenges: While the outcome assessors were blinded, it is impossible to blind patients to whether they are listening to music, which may introduce a placebo effect.
  • Subjective Reporting: The study relied heavily on self-reported questionnaires. Future research should incorporate objective measures like polysomnography (sleep studies) or more robust biomarker tracking.
  • Mechanistic Depth: While the study hints at autonomic nervous system regulation, the exact neurological pathways through which personalized music speeds up recovery remain a "black box" for future investigation.

Summary updated

Abstract

https://www.chictr.org.cn/index.html, identifier ChiCTR2200062181.
BibTeX
@inproceedings{wang2025five,
  author = {Wenjun Wang and Yong Sheng and Hongmei Xie and Jing‐Tao Zhao and Yang Zhao and Rui Zhou and Nan Zhao and Yu Wu},
  title = {{Five-element music therapy based on emotion classification to improve sleep in patients with cancer-related depression: a multi-group randomized controlled trial}},
  booktitle = {Frontiers in Psychiatry},
  year = {2025},
  doi = {10.3389/fpsyt.2025.1633673},
}
Figure