This systematic review and meta-analysis was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (registration no. CRD420251270027). The study was conducted according to a predefined protocol and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

A comprehensive literature search was conducted in CNKI, Wanfang Data, VIP, China Biology Medicine (CBM), PubMed, Web of Science, the Cochrane Library, and Embase from inception to November 2025. Search strategies were developed using a combination of controlled vocabulary terms (e.g., MeSH and Emtree) and free-text keywords. Chinese search terms included “音乐治疗 (music therapy)”, “音乐干预 (music intervention)”, “音乐疗法 (music treatment)”, “五音疗法 (five-tone therapy)”, “睡眠障碍 (sleep disorder)”, “失眠 (insomnia)”, “不寐 (sleeplessness)”, “睡眠质量 (sleep quality)”, “老年/老年人 (older adults/elderly)”, “老年疾病 (geriatric disease)”, and “养老 (elder care)”. English search terms included “music”, “music therapy”, “music intervention”, “sleep quality”, “sleep wake disorders”, “aged”, and “elderly”. Only studies published in Chinese or English were considered. The detailed search strategies for each database are provided in the supplementary materials.

Inclusion criteria: Studies were included if they met the following criteria: (1) Participants: Older adults aged ≥60 years with sleep problems. No restrictions were placed on diagnostic status, medication use, or other individual characteristics. (2) Intervention: Participants in the experimental groups received music-based interventions. (3) Comparators: Control groups received non-music-based interventions, including usual sleep care (e.g., sleep-related health guidance, optimization of the sleep environment, reduction of nocturnal disturbances, and daytime activity arrangement), sleep hygiene education (health education focused on sleep hygiene), routine oncology care (e.g., preoperative and postoperative nursing care), or pharmacological interventions. (4) the study design was a randomized controlled trial (RCT). (5) The study was published in Chinese or English.

Exclusion criteria: Studies were excluded if they met any of the following criteria: (1) duplicate publications, (2) study protocols, reviews, theoretical articles, case reports, or conference abstracts, or (3) studies in which music-based intervention was combined with other interventions.

Outcome measures: The primary outcome was sleep quality as assessed using the Pittsburgh Sleep Quality Index (PSQI). Secondary outcomes included depressive symptoms measured by using the Geriatric Depression Scale (GDS), anxiety symptoms as assessed by the Self-Rating Anxiety Scale (SAS), and cognitive function evaluated by using the Mini-Mental State Examination (MMSE).

Duplicates were identified and manually removed using EndNote 21. Based on the predefined inclusion and exclusion criteria, titles and abstracts were first screened to exclude obviously irrelevant studies, followed by full-text review to determine final eligibility. Three reviewers independently extracted key information, including authors, year of publication, participants’ age and sex, sample size, intervention and control conditions, outcome measures, and intervention frequency. Any discrepancies were resolved through discussion and, if necessary, adjudicated by a fourth reviewer to ensure the accuracy of study selection.

The methodological quality of the included studies was independently assessed by three reviewers using the Cochrane Risk of Bias Tool. The assessment covered seven domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, completeness of outcome data, selective reporting, and other potential sources of bias. Each domain was judged as “low risk,” “high risk,” or “unclear risk” of bias.

Meta-analyses were conducted using RevMan version 5.4 and Stata version 18.0. Continuous outcomes were pooled using mean differences (MDs) with corresponding 95% confidence intervals (95% CIs), applying the inverse-variance method. Statistical heterogeneity was assessed using χ² test and the I² statistic. A random-effects model was applied when substantial heterogeneity was detected (I² > 50% or P < 0.10), with between-study variance (τ²) estimated using the DerSimonian–Laird method.

Regarding publication bias, funnel plots were generated and Egger’s regression test was performed when at least 10 studies were included. When the number of included studies was limited or heterogeneity was high, the results of publication bias assessments were interpreted with caution.

A total of 14 studies published in Chinese and English were finally included, involving 1,730 participants, with 863 in the intervention groups and 867 in the control groups. The literature screening and selection process is presented in Figure 1, and the main characteristics of the included studies are summarized in Table 1.

The methodological quality of the included studies was assessed using the Cochrane Risk of Bias tool. Among the 14 included studies (Figures 2, 3), nine reported appropriate randomization methods, including random number tables or computer-generated allocation. Only two studies (19, 25) reported blinding of outcome assessors, while blinding was not reported in the remaining studies. One study (20) reported participant attrition, whereas outcome data were complete in the other studies. No other sources of bias were identified.

Subgroup analyses were conducted to explore potential sources of heterogeneity based on intervention duration, session length, intervention frequency, type of intervention, and target population (Table 2). Across all subgroups, music-based interventions were associated with improvements in sleep quality. Regarding intervention duration, interventions lasting 4–8 weeks showed the largest effect size (MD = −3.58, P < 0.00001). For session length, interventions lasting 30–45 min demonstrated the greatest improvement in sleep quality (MD = −3.60, P < 0.00001). In terms of intervention frequency, seven sessions per week were associated with the most pronounced effects (MD = −3.96, P < 0.00001). All intervention types showed significant effects on sleep quality. Improvements were more pronounced among community-dwelling older adults (MD = −2.99, P < 0.00001) compared with patients with Alzheimer’s disease.

Previous research has indicated that music-based interventions may influence sleep through multiple physiological and psychological pathways (13, 30). Listening to music characterized by slow tempo and gentle melodies before bedtime has been shown to reduce overall arousal levels and facilitate relaxation responses conducive to sleep initiation (31, 32). Objective polysomnographic studies further suggest that music interventions may decrease the proportion of light sleep (stage N1) and reduce high-frequency electroencephalographic activity, thereby promoting more stable sleep architecture (33). In addition, music can alleviate pre-sleep tension and rumination, redirecting attention away from stress-related cognitions and creating more favorable conditions for sleep onset (30).

On the basis of the mechanisms described above, the results of the present meta-analysis showed that music-based interventions significantly reduced the overall PSQI scores in older adults. The magnitude of improvement may be influenced by musical characteristics and intervention duration; specifically, sedative slow-tempo music and interventions lasting longer than 4 weeks appear to yield greater improvements in sleep quality. These findings are consistent with conclusions from previous systematic reviews and meta-analyses (34, 35). By including a larger number of randomized controlled trials and conducting subgroup analyses, the present study further confirmed the robustness of these findings and refined optimal intervention parameters.

The subgroup analyses revealed that music therapy improved sleep quality across different intervention durations, with the most pronounced effects observed in interventions lasting 4–8 weeks (MD = −3.58, P < 0.00001). This finding suggests that music therapy requires a sufficient intervention period to establish stable therapeutic effects. As a non-pharmacological approach, its benefits are typically achieved through repeated exposure, whereas excessively short interventions may be insufficient to induce sustained improvement, and overly prolonged interventions may reach a plateau due to habituation effects (11, 30, 32).

With regard to session length, interventions lasting 30–45 min demonstrated the greatest improvement in sleep quality (MD = −3.60, P < 0.00001), although beneficial effects were also observed with other durations. Moderate exposure time may help maintain engagement and relaxation, whereas longer sessions may induce fatigue or reduce adherence, thereby attenuating the intervention efficacy (10, 33).

In terms of intervention frequency, improvements in sleep quality were observed across studies employing regular music therapy, with the largest effect size associated with daily interventions (seven sessions per week; MD = −3.96, P < 0.00001). From a behavioral perspective, consistent and repeated musical exposure may strengthen the association between music listening and relaxation, facilitating the establishment of stable sleep–wake rhythms (11, 36).

The subgroup analyses by intervention modality indicated that multiple forms of music-based interventions significantly improved sleep quality in older adults (Table 2). This finding suggests that the therapeutic effects of music-based interventions may depend more on music-evoked emotional regulation and relaxation responses than on any specific musical form or theoretical framework (37). In addition, the subgroups in which music-based interventions were combined with usual care showed larger effect sizes in some comparisons, suggesting that multimodal approaches may provide synergistic benefits in sleep management for older adults.

Regarding population characteristics, beneficial effects were observed across different older adult groups, with relatively greater improvements among community-dwelling older adults (MD = −2.99, P < 0.00001). Previous studies suggest that baseline neurological status and disease burden may influence responsiveness to non-pharmacological interventions and that vulnerable populations may require longer or more individualized intervention strategies to achieve optimal outcomes (38, 39). Given the exploratory nature of the subgroup analyses, the limited number of studies in some subgroups, and the presence of heterogeneity, these findings should be interpreted with caution.

Music has been shown to activate brain regions involved in emotional processing, including the prefrontal cortex, limbic system, and reward-related networks, thereby modulating affective responses and enhancing positive emotional experiences (13). Music listening may also influence the autonomic nervous system and the hypothalamic–pituitary–adrenal axis, reducing stress-related physiological arousal and alleviating anxiety (14, 15). Moreover, music-induced relaxation and emotional resonance can reduce negative rumination and enhance subjective emotion regulation. In older adults, familiar and personally meaningful music appears particularly effective in promoting emotional stability and alleviating depressive and anxiety symptoms (10, 38).

The present meta-analysis demonstrated that music-based interventions significantly reduced scores on both the Geriatric Depression Scale and the Self-Rating Anxiety Scale, indicating stable beneficial effects on depressive and anxiety symptoms in older adults. These findings are consistent with previous evidence supporting the role of music interventions in emotional regulation (15, 40). By including more diverse older adult samples and simultaneously examining both depressive and anxiety outcomes, the present study further reinforces the potential of music-based interventions as a non-pharmacological psychological intervention. Taken together, the results suggest that improvements in sleep quality may be accompanied by parallel benefits in emotional well-being.

Music engagement activates widespread neural networks associated with cognitive processing and may support cognitive function through enhanced neuroplasticity (41). Neuroimaging studies have shown that music listening and music-based activities engage the prefrontal cortex, hippocampus, and default mode network, thereby supporting attention, memory, and executive functions (42). Regular musical engagement has also been proposed to help maintain functional connectivity within brain networks, which is particularly relevant in older adults and populations at risk of cognitive decline (43).

On this basis, the MMSE outcomes from the present meta-analysis suggest that music-based interventions may be associated with improved MMSE scores in older adults, consistent with findings from prior systematic reviews on music-based interventions and cognitive outcomes (41, 44). It should be noted, however, that the pooled MMSE analysis in the present study included two different intervention modalities—music listening and group receptive music therapy—resulting in a “mixed” estimate with limited clinical interpretability. Therefore, these findings should be interpreted with caution and are better considered as suggestive evidence. Given that improvements in sleep quality and reductions in emotional burden are closely linked to the maintenance of cognitive function (12, 45), future studies with consistent designs and clearly defined intervention modalities are needed to further clarify the independent contributions of different music-based interventions models to cognitive outcomes and to elucidate potential mechanisms.