The impact of mind-body therapies on the mental health of caregivers in palliative care settings: A systematic review

Abstract

Background:

Caregivers in palliative care settings often face significant emotional and psychological burdens, yet their mental health needs remain under-addressed. Mind-body therapies (MBTs) have shown promise in clinical populations, but their effects on caregivers are understudied and poorly synthesized.

Methods:

This systematic review followed PRISMA 2020 guidelines and was registered with PROSPERO (CRD42024571202). A comprehensive search of PubMed, EMBASE, Web of Science, and Cochrane Library identified randomized controlled trials (RCTs) and quasi-experimental studies (QES) assessing MBTs for caregivers in palliative care. Data were extracted using the PICOS framework. Study quality was appraised using JBI tools, risk of bias was assessed via ROB2 and ROBINS-I, and evidence certainty was rated with GRADE.

Results:

20 studies involving 1230 caregivers were included. Interventions encompassed 8 types of MBTs, with structured, therapist-guided formats demonstrating stronger effects than self-directed or digital approaches. Overall, MBTs showed short-term improvements in depression, anxiety, PTSD, distress, quality of life, and self-regulation. Informal and home hospice caregivers benefited most, while professional and spousal caregivers showed variable responsiveness. Methodological heterogeneity and limited follow-up constrained meta-analysis and generalizability.

Conclusion:

This review presents a comprehensive synthesis of existing evidence on MBTs targeting caregiver mental health in palliative care. It highlights the therapeutic potential of MBTs and identifies delivery modality and caregiver role as key moderators of intervention effectiveness. Furthermore, it contributes to the field by outlining practical implementation strategies for integrating MBTs into routine caregiver support, with an emphasis on contextual adaptation and policy-level facilitation. These findings support the development of scalable, person-centered interventions to meet the psychological needs of caregivers within palliative care systems.

1. Introduction

According to 2017 WHO statistics, approximately 20 million people globally require end-of-life palliative care services each year, with around 80% living in low- and middle-income countries,^[1] Among these, 67% are elderly (aged 60 years or older), and about 6% are children. With aging populations and rising chronic disease incidence, the demand for palliative care is increasing.^[2,3] Caregivers, as integral to palliative care, experience substantial physical, mental, and financial stress when caring for patients with advanced illnesses.^[4] Studies have shown that palliative care relies on caregivers who possess strong psychological, social, and emotional competencies to enable enhanced emotional regulation and empathy.^[5] Family members who assume long-term caregiving responsibilities frequently experience psychological issues such as depression and anxiety, along with significant physical health risks.^[6] The complexity of end-of-life caregiving imposes profound impacts on caregivers’ physical and mental well-being.^[7–9] During this period, caregivers often reduce participation in social and recreational activities, which further impairs their emotional regulation and stress management capacity.^[10,11] Therefore, maintaining caregivers’ mental health and supporting a balance between caregiving duties and personal needs are crucial to improving the overall quality of palliative care,^[12] and have emerged as urgent priorities in this field.^[13,14]

Mind-body therapies (MBTs) are conceptualized as therapeutic approaches emphasizing the dynamic interaction among the brain, mental processes, physical states, and behavior, with the primary aim of leveraging cognitive and emotional processes to influence physiological functioning and support overall health.^[15] As part of integrative multidisciplinary treatment, MBTs have been widely applied in cancer patients to alleviate anxiety, mood disorders, and chronic pain, thereby improving overall quality of life.^[16] In contrast, although caregivers play a critical role in palliative care, their mental health needs are often overlooked. Research on MBTs targeting caregivers remains limited, particularly regarding intervention modalities, outcome evaluations, and long-term effect.^[17] Preliminary evidence suggests that such interventions may offer psychological benefits for caregivers in palliative care settings. Lang study demonstrated that art therapy, as a mind-body intervention, significantly reduced anxiety in caregivers of cancer patients, helping to alleviate stress and promote positive emotion.^[18] Meanwhile, a RCT designed by Shani et al explored the application of Qigong in caregivers; preliminary findings showed that a Baduanjin-based intervention was feasible and showed potential in improving caregivers’ emotional distress, fatigue, sleep quality, and overall well-being.^[19] In addition, McConnell et al developed a best practice agenda for music therapy, highlighting its role as an art-based mind-body therapy that supports emotional regulation and psychological adaptation in informal caregivers throughout the end-of-life and bereavement proces.^[20] Shivarama, Mohana, and Sheelam noted that yoga and relaxation practices can enhance caregivers’ psychological resilience, significantly reduce perceived stress and burnout, and improve their sense of well-being.^[21]

However, current research on MBTs for caregivers presents 3 major limitations. First, intervention types are highly heterogeneous and study methodologies remain fragmented, making cross-study comparison and evidence synthesis difficult, thereby limiting clear conclusions on intervention efficacy. Easpaig et al noted substantial variation in intervention formats, settings, and outcome measures across studies, emphasized the absence of systematic classification systems and standardized evaluation frameworks.^[22] Second, caregivers are often not the sole or primary target population. MBTs were initially designed for patients and later extended to include caregivers, resulting in caregiver-specific outcomes being conflated and lacking clear definition or independent analysis. Third, most studies focus on short-term effects, with limited follow-up on the sustained impact of interventions. Consequently, the long-term effectiveness of MBTs in maintaining caregivers’ psychological well-being remains unclear,^[23] which hinders the optimization and broader implementation of such programs.

Therefore, this systematic review aims to address these limitations by evaluating the impact of MBTs on the mental health of caregivers in palliative care settings, analyzing variations in intervention outcomes, and exploring the potential implications of improved caregiver mental health for patient care quality. Through this study, we seek to provide a scientific basis for the development of targeted caregiver support strategies to enhance the quality of palliative care. This comprehensive analysis will not only clarify the current application of MBTs in palliative care but also offer important guidance for future research and practice.

2. Methods

This systematic review was meticulously designed in accordance with the updated guidelines outlined in the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (The PRISMA 2020 statement) checklist^[24] (Research Checklist), as well as the Cochrane Handbook for Systematic Reviews of Interventions.^[25] To ensure transparency and reproducibility, we registered our review protocol with the PROSPERO database under the registration number CRD42024571202. In developing the research question, the PICOS framework was adhered to, which stands for Population, Intervention, Comparison, Outcome, and Study design. This structured approach helps to clearly define the scope and focus of the review, ensuring comprehensive and systematic coverage of the topic at hand.

2.1. Study inclusion and exclusion criteria

This systematic review was designed to evaluate the impact of MBTs on the mental health of caregivers in palliative care settings. To ensure methodological rigor and the relevance of included evidence, we established predefined inclusion and exclusion criteria based on the PICOS framework (Population, Intervention, Comparison, Outcomes, Study design). Studies were included if they met the following conditions: the population comprised formal or informal caregivers providing care to patients in palliative care contexts; the intervention involved MBTs, such as mindfulness-based programs, meditation, acceptance and commitment therapy (ACT), yoga, music therapy, art therapy, or other integrated approaches incorporating both psychological and physical components; at least 1 caregiver-specific psychological outcome was reported, such as depression, anxiety, stress, emotional exhaustion, grief, or quality of life; the study design was a randomized controlled trial (RCT) or a quasi-experimental study (QES); the study was published in English; and it was a peer-reviewed full-text article.

Exclusion criteria were applied to remove studies that did not align with the scope or methodological standards of this review. Specifically, studies were excluded if they did not involve caregivers or reported outcomes exclusively for patients; did not employ MBTs as the intervention; failed to report caregiver-related psychological outcomes; were case reports, reviews, study protocols, non-English publications, or other forms of gray literature; or lacked sufficient methodological detail or outcome data. These criteria were rigorously and consistently applied throughout the study selection process to ensure the inclusion of only relevant, methodologically sound evidence.

Additionally, in this review, studies involving mixed populations (e.g., patients and caregivers) were only included if caregiver-specific psychological outcomes were reported separately. Studies that aggregated outcomes across groups without disaggregated caregiver data were excluded to ensure the specificity and validity of caregiver-focused analyses.

2.2. Electronic databases

This review searched for literature related to the impact of MBTs on the mental health of caregivers in palliative care settings. A comprehensive search was conducted across the following electronic databases: EMBASE, PubMed, Web of Science, and Cochrane, covering the time span from the inception of each database to July 2024 to ensure all relevant studies and data were included, thereby providing the most comprehensive informational support. In addition to the systematic searches of these specialized databases, a supplementary strategy was employed during the literature screening and retrieval process, utilizing a snowball method to further enhance the comprehensiveness and depth of the search.

2.3. Search strategy

The research team designed and implemented a detailed search strategy for relevant literature published since the inception of the databases (Table 1). This search strategy was independently executed by 3 researchers, who initially screened titles and abstracts. Key search terms included “mind-body therapies,” “palliative care settings,” and “caregivers,” along with related mental health issues such as depression, anxiety, emotional exhaustion, and post-traumatic stress disorder (PTSD). Researchers also utilized MeSH terms from each database to refine the search strings. To ensure comprehensiveness, the team reviewed the reference lists of included studies to identify additional potential research and contacted authors directly when necessary to obtain missing critical information. To prevent duplicates, all retrieved references were uploaded to the Zotero reference management tool.^[26]

Table 1.

Search strategies for English databases.

Number	Search terms
#1	Hospices [MeSH]
#2	Hospice care [MeSH]
#3	Hospice programs
#4	Hospice program
#5	Mental fatigue [MeSH]
#6	Hospice and palliative care nursing [MeSH]
#7	Hospice nusing
#8	Palliative nursing
#9	End-stage care
#10	#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9
#11	Caregivers [MeSH]
#12	Nurses [MeSH]
#13	Healthcare professionals
#14	Family members
#15	Companions
#16	#11 OR #12 OR #13 OR #14 OR #15
#17	Mental health [MeSH]
#18	Emotional states [MeSH]
#19	Anxiety [MeSH]
#20	Stress [MeSH]
#21	Fatigue [MeSH]
#22	Insomnia [MeSH]
#23	Psychological distress [MeSH]
#24	#17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23
#25	Mind-body therapy [MeSH]
#26	Mind-body therapies
#27	Mind-body medicine
#28	Art therapy [MeSH]
#29	Art therapies
#30	Drawing therapy
#31	Pottery therapy
#32	Meditation
#33	Mindfulness
#34	Drama therapy
#35	Music therapy
#36	Horticultural therapy
#37	#25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36
#38	#10 AND #16 AND #24 AND #37

MeSH = Medical subject headings.

2.4. Study selection

Three independent researchers participated in the screening, eligibility review, and selection process. J.F.S. was responsible for downloading and reviewing the screened articles, excluding irrelevant literature. Subsequently, after determining the eligibility of 20 documents, J.F.S. and L.Y.L. assessed and scored randomized controlled trials (RCTs) and QES using the JBI Clinical Assessment Tool.^[27] In cases of disagreement between the 2 reviewers, Researcher J.X. made the final decision. Eligible studies underwent double-checking by Researcher X.X.L. and Researcher G.H.H. This method was employed to systematically organize and evaluate the relevant information extracted from the literature, and the research process was reported using the PRISMA flow diagram (Fig. 1).

Figure 1.

PRISMA flow diagram of the literature selection process. PRISMA = preferred reporting items for systematic review and meta-analysis

2.5. Data extraction

This review utilized a systematic PICOS approach for the initial data extraction from the included studies. The extraction was independently completed by 3 researchers to ensure comprehensiveness and objectivity. To further guarantee data quality, approximately 20% of the extracted data were randomly selected for review by a fourth researcher to check for consistency and accuracy. This multi-researcher independent extraction with random verification significantly enhanced the reliability of data extraction. The research team designed a standardized Excel sheet based on the PICO(S) framework for systematic extraction of basic study information, including the first author, publication date, country of implementation, and journal.

Population characteristics (P): Basic information about caregivers, including gender, age, caregiving role (e.g., informal, spousal, or professional), and the context of care delivery (e.g., home-based, hospital-based, or hospice settings).

Intervention characteristics (I): Detailed records of the specific types of MBTs (e.g., mindfulness meditation, yoga, art therapy) and their implementation details (including content, frequency, duration per session, and total treatment duration).

Control group information (C): Description of the control group’s interventions.

Outcome indicators (O): Extraction of primary and secondary outcome indicators, including but not limited to depression, anxiety, stress levels, quality of life, caregiver burden, and social support.

Study design (S): Types of studies (RCTs or QES).

Additionally, during the data extraction process, if any information was unclear or missing, the research team attempted to contact the original authors for additional data to ensure the completeness of extracted information. All extracted data were systematically organized in a standardized Excel sheet for subsequent quality assessment, data synthesis, and analysis to ensure data reliability and consistency.

2.6. Statistical analysis

In RCTs, postintervention means, standard deviations (SDs), and sample sizes (n) for both intervention and control groups were extracted. Based on these values, standardized effect sizes (Hedges’ g) and their corresponding 95% confidence intervals (CIs) were calculated using the escalc function in Stata version 18 (StataCorp), following the procedures outlined by the Cochrane Handbook for Systematic Reviews of Interventions.^[28] When group-level means and SDs were unavailable but standardized effect sizes (e.g., Cohen d) were reported by study authors, those values were directly extracted for analysis.

In QES, including single-group pre-post designs, pre- and post-intervention means and SDs were extracted for the caregiver group. Within-group effect sizes were computed assuming a conservative pre-post correlation coefficient of R = 0.5, as recommended by Morris and DeShon (2002) and further applied in meta-analyses of psychological interventions.^[29,30]

2.7. Quality appraisal of the included studies

To ensure the quality of the literature included in this study, this systematic review employed a scoring system based on JBI guidelines to assess the quality of RCTs and QES.^[31] The quality assessment of RCTs and QES was independently conducted by J.F.S. and J.X., with Z.Z. participating in discussions to resolve any uncertainties. The quality assessment was scored according to JBI guidelines, with each “yes” receiving 1 point, while items that did not meet criteria or were unclear received 0 points. This scoring system facilitates horizontal comparison of study quality and ranking based on total scores. For RCTs, studies scoring 6 or below were categorized as low quality, scores from 7 to 9 were deemed medium quality, and those scoring 10 or above were considered high quality. Given the specificities of design and implementation for QES, the scoring criteria were adjusted accordingly: 4 points or below were classified as low quality, 5 to 7 points as medium quality, and 8 points or above as high quality.

2.8. Synthesis of the studies included

Due to the high heterogeneity of studies included in this systematic review, along with a lack of sufficient comparable quantitative data in existing research, a quantitative meta-analysis could not be performed. Consequently, the research team adopted a comprehensive qualitative analysis method based on the PICOS framework to address the diversity and complexity of the included studies.

2.9. Quality assessment of evidence and risk of bias

In the meticulous assessment of potential bias, this systematic review strictly adhered to the Cochrane Collaboration’s comprehensive guidelines. For RCTs, the risk of bias tool 2.0(ROB2), as delineated by Higgins et al^[25] and Sterne et al,^[32] was employed, allowing for evaluation across several domains, categorizing the risk as either low, high, or raising some concerns. Each trial was assigned an aggregate score based on these assessments, ensuring a nuanced appraisal of bias.

In addressing nonrandomized studies, the ROBIN-I tool, as described by Sterne et al,^[33] was utilized. This instrument facilitated a detailed examination of bias, rating each study on a gradient from low to critical risk or indicating no available information. The cumulative bias score for each study was derived following the prescribed methodological framework. This rigorous evaluative approach was pivotal in ensuring the reliability and validity of the findings, thereby contributing to the robustness of this systematic review.

Evidence quality was appraised using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) methodology, with evaluations carried out through the GRADEpro GDT platform (www.gradepro.org). Following GRADE criteria, each outcome was examined across 5 key aspects: study limitations, inconsistency of results, indirectness of evidence, imprecision of estimates, and potential publication bias. Based on the overall certainty in the estimated effects, evidence was graded into 4 levels: high, moderate, low, or very low.^[34] Two independent reviewers conducted the appraisal, and any discrepancies were resolved through discussion until consensus was achieved.

3. Results

3.1. Search results and studies description

To explore the effects of various MBTs on improving mental health issues among caregivers in palliative care settings, a comprehensive literature search and screening were conducted. This encompassed relevant studies published in PubMed, Web of Science, Cochrane Library and Embase. A total of 2282 records were initially retrieved through the database search. After removing duplicates and conducting title, abstract, and full-text screening, 20 studies met the inclusion criteria and were included for data extraction.

As detailed in Table 2, the compendium of synthesized evidence included 15 RCTs, which provided robust support for the effectiveness of MBTs. In addition, 5 QES, offered valuable insights by exploring the applicability of MBTs in palliative care settings and expanding the evidence base in the absence of randomized designs. Together, these investigations enhanced the understanding of the feasibility, acceptability, and contextual relevance of MBTs in diverse care settings. The included studies were published between 2006 and 2024 and collectively examined 8 distinct types of MBTs: accelerated resolution therapy (ART),^[35] existential behavioral therapy (EBT),^[36,37] meditation therapy,^[38–42] mindfulness therapy,^[43–46] ACT,^[47] music therapy,^[48–51] art therapy^[52,53]; and narrative interview^[54] (Fig. 2). These interventions demonstrated unique potential in enhancing emotional regulation, psychological resilience, and overall quality of life. Notably, they showed preliminary effectiveness in alleviating key psychological symptoms such as anxiety, depression, PTSD, and emotional distress. This resembles earlier discussions suggesting that MBTs may provide meaningful psychological benefits for caregivers in palliative care settings.^[55,56] Based on systematic data extraction and analysis of the included studies, several key findings have been identified. These not only confirm the short-term effectiveness of mind-body interventions in improving caregivers’ mental health, but also highlight the need for further exploration of their underlying mechanisms, delivery strategies, and long-term impact. With advantages such as high acceptability, low cost, and clearly defined therapeutic mechanisms, MBTs are emerging as practical and scalable approaches for caregiver support in end-of-life care. As noted by Gu et al, the systematic integration of such interventions into caregiver support policies may facilitate the development of a more person-centered long-term care system and help address structural gaps in existing mental health services.^[17]

Table 2.

Principal characteristics of all included studies in this review.

Authors	Ref	Study design	Nation/region	Clinical setting	Gender	Mean age	Follow-up	Sample sizes	Outcomes	Treatment group	Procedure times	Control group
Buck et al	[35]	RCT	USA	Outpatient department	M, F	68.7	Yes	32/22	ICG/PCL-5/CESD	ART	8 wk	WL
Kögler et al	[36]	RCT	Germany	Setting hospital	M, F	54.3	Yes	73/57	CAMS-R/BSI/SWLS/WHOQOL-BREF/SMiLE/FFM	EBT	6 wk	UC
Fegg et al	[37]	RCT	Germany	Setting hospital	M, F	54.5	Yes	81/79	BSI/SWLS/WHOQOL-BREF/QOL-NRS/PANAS	EBT	4 wk	UC
Tan et al	[38]	RCT	Malaysia	Setting hospital	M, F	45	No	20/20	SP/BIS	20-MMB	20 min	UC
Beng et al	[39]	RCT	Malaysia	Setting hospital	M, F	47–54	No	7/4	DT	5-MMB	5 min	LT
Kubo et al	[40]	RCT	USA	Setting hospital Home-care	M, F	62.65	Yes	22/17*	CQOLC/DT/HADS/FFMQ-SF	MI	6 wk	UC
Kubo et al	[41]	RCT	USA	Home-care	M, F	57.6	Yes	17/14*	DT/HADS/PROMIS/CQOLC/PTGI/ FFMQ-SF	MMI	8 wk	UC
Schellekens et al	[43]	RCT	Netherlands	Setting hospital	M, F	58.6	Yes	21/23*	HADS/SPPIC/IMS-S/FFMQ/SCS/RRS-BR/IES	MBSR	8 wk	UC
Milbury et al	[44]	RCT	USA	Home-care	M, F	63.9	Yes	26//24/25*	CES-D/IES/FACT-SP	CBM	7 wk	SE/UC
Türkoğlu and Kavuran	[45]	RCT	Turkey	Home-care	M, F	45.53	Yes	50/54	CSS/CQOLC	MBSR	12 wk	UC
Milbury et al	[46]	RCT	USA	Home-care	M, F	54.6	Yes	18/17	CES-D/MAAS/ SCS/PAIRI	CBM	12 wk	UC
Muscara et al	[47]	RCT	Australia	Setting hospital	M, F	37.17	Yes	37/44	PCL-5/DASS-21/PECI/FMM	ACT	8 wk	UC
Valero-Cantero et al	[48]	RCT	Spain	Home-care	M, F	62.71	No	41/41	QOL-FV/EuroQol-5D-5L/CSQ-8	PPM	7 d	UC
Valero-Cantero et al	[49]	RCT	Spain	Home-care	M, F	62.71	No	41/41	CSI	SCMI	7 d	BTER
Wlodarczyk	[51]	RCT	USA	Outpatient department	M, F	44	Yes	34/34	HCGI/CSF/WES	GMI	8 wk	UC
Karabuga Yakar et al	[42]	QES	Türkiye	Setting hospital	M, F	48.72	No	11	DT/BDI/BAI	MBMP	5 wk	N/A
Hilliard	[50]	QES	USA	Outpatient department	M, F	28–60	No	10/7*	CFS/TBQ	EMT/DMT	6 wk	N/A
Potash et al	[52]	QES	Hong Kong	Outpatient department	M, F	42	No	69/63*	MBI-GS/FFMQ/DAP-R	ATBG/SBS	6 wk	N/A
Salzano et al	[53]	QES	USA	Outpatient department	M, F	25~65	No	20	MBI-GS/SAWS	CAMI	1 mo	N/A
Kaimal et al	[54]	QES	USA	Home-care	M, F	69	No	7/7	PANAS/GSE/SWBS/PSS-10	CAT/NI	1 h	N/A

20-MMB = 20-min mindfulness breathing, 5-MMB = 5-min mindful breathing, ACT = acceptance and commitment therapy, ART = accelerated resolution therapy, ATBS = art-therapy-based supervision group, BAI = Beck Anxiety Inventory, BDI = Beck Depression Inventory, BIS = Bispectral Index Score, BSI = Brief Symptom Inventory, BTER = Basic Therapeutic Education Recording, CAMI = Collaborative Art-Making Intervention, CAMS-R = Cognitive and Affective Mindfulness Scale-Revised, CAT = Collage Art Therapy, CBM = Couple-Based Meditation, CESD = Center for Epidemiologic Studies Depression, CES-D = Center for Epidemiologic Studies Depression Scale, CFS = Compassion Fatigue Scale, CQOLC = Caregiver Quality of Life Index-Cancer, CSF = Compassion Satisfaction and Fatigue Test, CSI = Caregiver Strain Index, CSQ-8 = Client Satisfaction Questionnaire, CSS = Caregiver Stress Scale, DAP-R = Death Attitude Profile-Revised, DASS-21 = Depression, Anxiety, and Stress Scale-21, DMT = didactic music therapy approach, DT = distress thermometer, EBT = Existential Behavioural Therapy, EMT = ecological music therapy approach, EuroQol-5D-5L = European Quality of Life visual analogue scale, F = Female, FACT-G = Functional Assessment of Cancer Therapy General Scale, FACT-SP = Functional Assessment of Cancer Therapy - Spiritual Well-Being Scale, FFM = Five Facets of Mindfulness, FFMQ = Five Facet Mindfulness Questionnaire, FFMQ-SF = Five Facet Mindfulness Questionnaire–Short Form, FMM = Family Management Measure, GMI = Group Music Intervention, GSE = General Self-Efficacy Scale, HADS = Hospital Anxiety and Depression Scale, HCGI = Hospice Clinician Grief Inventory, ICG = Inventory of Complicated Grief, IES = Impact of Event Scale, IMS-S = Investment Model Scale-Satisfaction subscale, LT = Listening Therapy, M = Male, MBI-GS = Maslach Burnout Inventory-General Survey, MBMP = Meditation Based Mandala Programme, MBSR = Mindfulness-Based Stress Reduction, MBSR = Online Mindfulness-Based Stress Reduction, MI = mindfulness intervention, MMI = Mobile Mindfulness Intervention, N/A = not applicable, NI = narrative interview, PAIRI = Personal Assessment of Intimacy in Relationships Inventory, PANAS = Positive and Negative Affect Scale, PCL-5 = PTSD Checklist for DSM-5, PECI = Parent Experience of Child Illness Scale, PPM = personalized prerecorded music, PROMIS = Patient-Reported Outcomes Measurement Information System, PSS-10 = Perceived Stress Scale, PTGI = post-traumatic growth inventory, QOL-FV = Quality of Life Family Version, QOL-NRS = Quality of Life Numeric Rating Scale, RRS-BR = Ruminative Response Scale-Brooding subscale, SE = supportive-expressive therapy, SAWS = Support Appraisal for Work Stressors Inventory, SBS = skills-based supervision group, SCMI = Self-Chosen Music Intervention, SCS = Self-Compassion Scale, SMiLE = Schedule for Meaning in Life Evaluation, SP = Suffering Pictogram, SPPIC = Self-Perceived Pressure from Informal Care, SWLS = Satisfaction with Life Scale, TBQ = Team Building Questionnaire, UC = Usual Care, WES = Perception of Work Environment, WHOQOL-BREF = World Health Organization Quality of Life-BREF, WL = wait-list.

*ID-40: The study initially enrolled 39 caregivers (22 in the intervention group and 17 in the control group), of whom 27 completed the 12-wk follow-up assessments (14 intervention, 13 control).

ID-41: This RCT enrolled 44 caregivers, with 21 assigned to MBSR and 23 to usual care. Sample sizes declined over time due to withdrawal, illness progression, or loss to follow-up: n = 44 at baseline (T0), n = 35 post-intervention (T1), and n = 31 at 3-mo follow-up (T2).

ID-42: The study enrolled 31 caregivers, of whom 26 completed the intervention and post-intervention assessments (intervention group n = 13; control group n = 13).

ID-44: The experimental groups were divided into 2 parts: Couple-Based Meditation (CBM) group and Supportive-Expressive (SE) group. The control group was Usual Care (UC) group. A total of 75 patient-caregiver pairs completed the baseline self-report measures and were subsequently randomly assigned to the 3 groups. The specific allocation was: 26 pairs in the CBM group, 24 pairs in the SE group, and 25 pairs in the UC group.

ID-51: This quasi-experimental study included no control group. The Ecological Music Therapy group had 10 participants; the Didactic Music Therapy group had 7.

ID-53: This quasi-experimental study included 2 intervention groups without a control group: art-therapy-based supervision (n = 69) and skills-based supervision (n = 63).

ID-55: This quasi-experimental study did not include a control group. It comprised 2 intervention arms: the Narrative Interview Group (n = 7) and the Collage Art Therapy Group (n = 7), with participants self-selecting into groups based on preference.

Figure 2.

Classification of mind-body therapies in palliative care.

3.2. Methodological appraisal

According to the JBI critical appraisal tools, this systematic review included 15 RCTs out of 20 eligible studies. 3 studies were categorized as high quality,^[48,49,51] 11 studies as medium quality,^{[35–41,43–46]} and 1 study as low quality^[47] (Table 3). Overall, the included RCTs commonly exhibited insufficient blinding, inadequate allocation concealment, and unclear handling of attrition, resulting in the majority failing to meet high-quality standards. These methodological limitations compromised the internal validity and credibility of the findings, underscoring the need for more rigorous design in future research on mind-body interventions.

Table 3.

Quality of evidence in the included 20 reports based on JBI’s critical appraisal tools.

Study design	Author	Ref ID	Year	Journal	Evaluation for evidence reported
					1	2	3	4	5	6	7	8	9	10	11	12	13	Ranking
RCT	Buck et al	[35]	2020	American Journal of Hospice and Palliative Medicine®	Y	UN	Y	N	N	UN	Y	Y	Y	Y	Y	Y	Y	9
RCT	Kögler et al	[36]	2015	Palliative and Supportive Care	Y	UN	Y	N	N	UN	Y	UN	Y	Y	Y	Y	Y	8
RCT	Fegg et al	[37]	2013	Psycho-Oncology	Y	Y	Y	N	UN	N	Y	UN	Y	Y	Y	Y	Y	9
RCT	Tan et al	[38]	2020	American Journal of Hospice & Palliative Medicine®	Y	UN	Y	N	N	N	N	Y	Y	Y	Y	Y	Y	8
RCT	Beng et al	[39]	2016	American Journal of Hospice & Palliative Medicine®	Y	UN	Y	N	N	N	N	Y	Y	Y	Y	Y	Y	8
RCT	Kubo et al	[40]	2024	Psycho-Oncology	Y	UN	N	N	N	UN	Y	Y	Y	Y	Y	Y	Y	8
RCT	Kubo et al	[41]	2019	Integrative Cancer Therapies	Y	Y	N	N	N	UN	N	Y	Y	Y	Y	Y	Y	8
RCT	Schellekens et al	[43]	2017	Psycho-Oncology	Y	Y	N	N	N	UN	Y	N	Y	Y	Y	Y	Y	8
RCT	Milbury et al	[44]	2020	The Oncologist	Y	UN	Y	N	N	UN	Y	Y	Y	Y	Y	Y	Y	9
RCT	Türkoğlu & Kavuran	[45]	2020	BMC Paliative Care	Y	UN	Y	N	N	N	N	UN	Y	Y	Y	Y	Y	7
RCT	Milbury et al	[46]	2020	Journal of Pain and Symptom Management	Y	UN	Y	N	N	UN	Y	N	Y	Y	Y	Y	Y	8
RCT	Muscara et al	[47]	2020	JAMA Network Open	Y	Y	N	N	N	N	N	N	UN	Y	Y	Y	Y	6
RCT	Valero-Cantero et al	[48]	2023	The arts in psychotherapy	Y	UN	Y	Y	UN	UN	Y	Y	Y	Y	Y	Y	Y	11
RCT	Valero-Cantero et al	[49]	2022	Journal of the International Neuropsychological Society	Y	UN	Y	Y	UN	UN	Y	Y	Y	Y	Y	Y	Y	10
RCT	Wlodarczyk	[51]	2013	Progress in Palliative Care	Y	Y	Y	N	N	UN	Y	Y	Y	Y	Y	Y	Y	10
					100%	33%	73%	13%	0%	0%	67%	60%	93%	100%	100%	100%	100%
QES	Karabuga Yakar et al	[42]	2023	OMEGA – Journal of Death and Dying	Y	N	Y	Y	Y	N	Y	UC	Y	/	/	/	/	6
QES	Hilliard	[50]	2006	The Arts in Psychotherapy	Y	UC	Y	N	Y	N	Y	Y	Y	/	/	/	/	6
QES	Potash et al	[52]	2014	International Journal of Palliative Nursing	Y	N	Y	Y	Y	Y	Y	Y	Y	/	/	/	/	8
QES	Salzano et al	[53]	2013	The Arts in Psychotherapy	Y	Y	UC	Y	N	Y	Y	Y	Y	/	/	/	/	7
QES	Kaimal et al	[54]	2020	BMC Paliative Care	Y	N	Y	Y	Y	N	Y	UC	Y	/	/	/	/	6
					100%	17%	83%	83%	83%	67%	100%	67%	100%					100%

JBI critical appraisal checklist for randomized controlled trial: Was true randomization used for assignment of participants to treatment groups? Was allocation to treatment groups concealed? Were treatment groups similar at the baseline? Were participants blind to treatment assignment? Were those delivering treatment blind to treatment assignment? Were outcomes assessors blind to treatment assignment? Were treatment groups treated identically other than the intervention of interest? Was follow-up complete and if not, were differences between groups in terms of their follow-up adequately described and analyzed? Were participants analyzed in the groups to which they were randomized? Were outcomes measured in the same way for treatment groups? Were outcomes measured in a reliable way? Was appropriate statistical analysis used? Was the trial design appropriate, and any deviations from the standard RCT design (individual randomization, parallel groups) accounted for in the conduct and analysis of the trial?

JBI critical appraisal checklist for quasi-experimental studies: Y = yes; N = no; UC = unclear; NA = not applicable; JBI critical appraisal checklist for quasi-experimental studies: Q1 = Is it clear in the study what is the “cause” and what is the “effect” (i.e. there is no confusion about which variable comes first)? Q2 = Were the participants included in any comparisons similar? Q3 = Were the participants included in any comparisons receiving similar treatment/care, other than the exposure or intervention of interest? Q4 = Was there a control group? Q5 = Were there multiple measurements of the outcome both pre and post the intervention/exposure? Q6 = Was follow-up complete and if not, were differences between groups in terms of their follow-up adequately described and analyzed? Q7 = Were the outcomes of participants included in any comparisons measured in the same way? Q8 = Were outcomes measured in a reliable way? Q9 = Was appropriate statistical analysis used?

JBI = the Joanna Briggs Institute, N = no, QES = quasi-experimental studies, RCT = randomized controlled trial, UN = unclear, Y = yes.

In addition, this study included 5 QES, of which 1 study was classified as high quality^[52] and 4 studies as medium quality^{[42,50,53,54]} (Table 3). None of the QES were rated as low quality. The main methodological limitations were concentrated in 2 areas. First, most studies did not adequately report baseline comparability between groups, indicating a risk of selection bias. Second, incomplete follow-up data and insufficient analysis of attrition limited the interpretability and credibility of the findings.

3.3. Risk of bias assessment, certainty of evidence

This systematic review employed the cochrane risk of bias assessment tool to evaluate the methodological quality of the included studies (Fig. 3). Among the 15 RCTs, only 2 were judged to have a low overall risk of bias,^[41,45] 8 studies were rated as having “some concerns” in 1 or more domains,^{[35,37,38,40,43,44,48,49]} and 5 were considered to have a high overall risk of bias.^{[36,39,46,47,51]} Specifically, 13 studies were rated as having “some concerns” in the domain of selection of the reported result (D5), indicating a widespread issue with reporting transparency. Some studies also showed concerns related to randomization (D1) and missing outcome data (D3), although these issues were not predominant across the full sample.

Figure 3.

Risk of bias assessments for (A) randomized controlled trials (RoB2) and (B) nonrandomized studies (ROBINS-I).

The ROBINS-I assessment of the 5 nonrandomized studies showed that 1 study was rated as having a low overall risk of bias,^[50] while the remaining 4 were judged to have a moderate risk of bias.^[42,52–54] Overall, the studies were considered to present a moderate level of risk. While moderate concerns were identified in several domains – particularly confounding (D1) and participant selection (D2) – all studies were consistently rated as having low risk in both the classification of interventions (D3) and the selection of the reported result (D7).

The certainty of evidence for each psychological outcome was assessed using the GRADE framework, considering risk of bias, inconsistency, indirectness, and imprecision (Table 4). Overall, the evidence quality was rated as moderate for depression and low for all other outcomes. Specifically, depression (8 studies) demonstrated consistent moderate-to-large effects based on validated measures; although some CIs were wide, the consistent and substantial effect size warranted a downgrade for imprecision but an upgrade for magnitude of effect. In contrast, anxiety, stress, and psychological distress were rated as low-certainty outcomes due to inconsistent effect directions, heterogeneous measurement tools, and imprecision arising from small samples and CIs overlapping the null. Similarly, quality of life and PTSD were downgraded for indirectness, given the diversity of measurement instruments and intervention types, as well as for imprecision. Mindfulness and emotional exhaustion were also rated as low-certainty due to substantial variability in measurement approaches and effect estimates, with CIs frequently wide or crossing zero. Collectively, these limitations highlight the need for future trials to adopt more rigorous methodologies, including standardized outcome definitions, adequate sample sizes, and clearer reporting, to strengthen the evidence base for mind-body interventions in palliative caregiver populations.

Table 4.

GRADE evidence on mind-body therapies for caregiver mental health.

Outcomes	Number of studies	Total participants	Measurement tools	Certainty of evidence	Reason for downgrade
Depression	8	602	CESD/CES-D/HADS/DASS-21/BDI	⊕⊕⊕〇Moderate	Downgraded for imprecision due to some wide CIs; upgraded for large effects and consistent direction.
Anxiety	5	598	HADS/DASS-21/BAI	⊕⊕〇〇 Low	Downgraded for imprecision and inconsistency; small sample sizes and CI crossing zero in several studies.
Psychological distress	11	903	HADS/DASS-21/BSI/DT/ICG/PCL-5/IES/BAI/CES-D	⊕⊕〇〇 Low	Downgraded for inconsistency and imprecision due to wide CI and variability in results.
Stress	5	560	DASS-21/PSS-10/BSI/SPPIC/RRS-BR	⊕⊕〇〇 Low	Downgraded for inconsistency and imprecision due to heterogeneity in interventions and tools used.
Quality of life	7	560	CQOLC/WHOQOL-BREF/QOL-NRS/ QOL-FV/SWLS/PROMIS/EuroQol-5D-5	⊕⊕〇〇 Low	Downgraded for indirectness and imprecision; various tools and interventions applied inconsistently.
PTSD	4	486	PCL-5/ IES	⊕⊕〇〇 Low	Downgraded for imprecision due to wide CIs and limited number of studies; small total sample size and variation in effects.
Mindfulness	5	279	FFMQ/ FFMQ/ FFMQ-SF/ CAMS-R/ MAAS	⊕⊕〇〇 Low	Downgraded for imprecision and inconsistency; small to moderate effects with CI overlapping zero.
Emotional exhaustion	4	237	MBI-GS/ CFS/ CSF	⊕⊕〇〇 Low	Downgraded for inconsistency and imprecision; measurement variability and CI width.

BAI = Beck Anxiety Inventory, BDI = Beck Depression Inventory, BSI = Brief Symptom Inventory, CAMS-R = Cognitive and Affective Mindfulness Scale-Revised, CESD = Center for Epidemiologic Studies Depression, CES-D = Center for Epidemiologic Studies Depression Scale, CFS = Compassion Fatigue Scale, CQOLC = Caregiver Quality of Life Index-Cancer, CSF = Compassion Satisfaction and Fatigue Test, DASS-21 = Depression, Anxiety, and Stress Scale-21, DT = distress thermometer, EuroQol-5D-5L = European Quality of Life visual analogue scale, FFMQ = Five Facet Mindfulness Questionnaire, FFMQ-SF = Five Facet Mindfulness Questionnaire–Short Form, HADS = Hospital Anxiety and Depression Scale, ICG = Inventory of Complicated Grief, IES = Impact of Event Scale, MBI-GS = Maslach Burnout Inventory-General Survey, PCL-5 = PTSD Checklist for DSM-5, PROMIS = Patient-Reported Outcomes Measurement Information System, PSS-10 = Perceived Stress Scale, QOL-FV = Quality of Life Family Version, QOL-NRS = Quality of Life Numeric Rating Scale, RRS-BR = Ruminative Response Scale-Brooding subscale, SPPIC = Self-Perceived Pressure from Informal Care, SWLS = Satisfaction with Life Scale, WHOQOL-BREF = World Health Organization Quality of Life-BREF.

3.4. Description of the caregivers’ population

The included studies involved a total of 1230 caregivers across various caregiving roles (Table 2). These comprised informal caregivers, home hospice caregivers, spousal caregivers, and professional hospice caregivers, whose ages ranged from 25 to 69 years (Tables 2 and 5). Geographically, 9 studies were conducted in the United States, followed by 2 each in Germany, Malaysia, and Spain. Single studies were conducted in the Netherlands, Australia, Turkey, and Hong Kong (Table 2).

Table 5.

Effects of psychological outcomes of mind-body interventions in caregivers of palliative patients.

Author	Ref ID	Study participants	Interventions	Psychological indicators	P-value	Effect size	95% CI	Intervention effect	Results
Buck et al	[35]	Informal caregivers	Accelerated resolution therapy	Complicated grief	<.0001	−1.833 (g)	[−2.521, −1.144]	↑	Accelerated resolution therapy significantly alleviated complicated grief, PTSD, and depression among informal caregivers.
				PTSD	<.0001	−1.681 (g)	[−2.353, −1.009]	↑
				Depression	<.0001	−1.215 (g)	[−1.852, −0.578]	↑
Kögler et al	[36]	Informal caregivers	Existential behavioral therapy	Psychological distress	.02	NR	NR	△	Existential behavioral therapy significantly improved psychological distress, subjective quality of life, and meaning in life, especially in the short to mid term. Life satisfaction showed indirect improvement via mindfulness, while general quality of life did not significantly change.
				Subjective quality of life	.06	NR	NR	↑
				Quality of life	<.01	NR	NR	△
Fegg et al	[37]	Informal caregivers	Existential behavioral therapy	Somatisation	.79	0.025 (g)	[−0.321, 0.37]	×	The EBT significantly improved anxiety, depression, life satisfaction, and quality of life, while reducing negative affect in caregivers. Positive affect showed a nonsignificant trend, and somatization remained unchanged. Most indicators were valid and sensitive for evaluating EBT outcomes.
				Depression	.09	−0.132 (g)	[−0.478, 0.214]	△
				Anxiety	.006	−0.113 (g)	[−0.458, 0.233]	↑
				Life satisfaction	.009	0.159 (g)	[−0.187, 0.505]	↑
				Quality of life	.007	0.224 (g)	[−0.123, 0.57]	↑
				Subjective quality of life	<.001	0.426 (g)	[0.077, 0.776]	↑
				Positive affect	.06	0.457 (g)	[0.11, 0.804]	△
				Negative affect	.003	0.152 (g)	[−0.191, 0.495]	↑
Tan et al	[38]	Informal caregivers	Meditation therapy	Psychological suffering	.036	NR	NR	↑	The study showed that 20-min mindful breathing significantly reduced psychological suffering and stress, as measured by the Suffering Pictogram (P = .036) and BIS score (P < .0001). Both indicators proved valid and sensitive to the intervention.
				Stress	<.0001	NR	NR	↑
Beng et al	[39]	Family caregivers	Meditation therapy	Psychological distress	.001	NR	NR	↑	The study demonstrated that 5-min mindful breathing significantly reduced psychological distress among palliative caregivers, as measured by the Distress Thermometer. The indicator proved valid and sensitive, showing a greater reduction in distress compared to the control condition (P = .001).
Kubo et al	[40]	Family caregivers	Meditation therapy	Distress	.03	NR	NR	↑	The mindfulness-based intervention significantly improved distress and mindfulness among caregivers.
				Mindfulness	.03	NR	NR	↑
Kubo et al	[41]	Informal caregivers	Meditation therapy	Distress thermometer	.47	−0.44 (g)	[−1.22, 0.339]	×	The mHealth mindfulness intervention significantly improved caregivers’ mindfulness and showed borderline improvements in post-traumatic growth. No significant changes were observed in distress, anxiety, depression, sleep disturbance, fatigue, overall QoL, or other mindfulness dimensions.
				Depression	.29	−0.489 (g)	[−1.271, 0.293]	×
				Anxiety	.79	−0.084 (g)	[−0.853, 0.685]	×
				Mindfulness	.03	0.306 (g)	[−0.468, 1.08]	↑
				Sleep disturbance	.69	−0.222 (g)	[−0.994, 0.55]	×
				Fatigue	.23	−0.105 (g)	[−0.874, 0.665]	△
				Post-traumatic growth	.05	0.024 (g)	[−0.745, 0.793]	△
Schellekens et al	[43]	Home hospice caregivers	Mindfulness	Anxiety/depression	.661	0.061 (g)	[−0.604, 0.726]	×	The MBSR intervention was not effective in improving anxiety/depression, caregiver burden, mindfulness, rumination, or PTSD symptoms among caregivers in this study.
				Caregiver burden	.607	−0.208 (g)	[−0.875, 0.459]	×
				Mindfulness	.861	−0.076 (g)	[−0.741, 0.590]	×
				Rumination	.373	−0.21 (g)	[−0.878, 0.457]	×
				PTSD	.45	−0.074 (g)	[−0.739, 0.591]	×
Milbury et al	[44]	Spousal caregivers	Meditation therapy	Depression	.01	NR	NR	↑	The mindfulness-based intervention significantly reduced depression in caregivers and patients, with the strongest effects in caregivers. It showed marginal improvement for post-traumatic stress in patients and modest benefits for caregivers’ spiritual well-being. Depression proved the most responsive indicator.
				Post-traumatic stress	>.05	NR	NR	×
				Spiritual distress	.07	NR	NR	△
Türkoğlu & Kavuran	[45]	Home hospice caregivers	Mindfulness	Psychological stress	<.001	−1.256 (g)	[−1.678, −0.834]	↑	Both psychological stress and quality of life emerged as valid and sensitive indicators for evaluating the impact of the MBSR intervention, which demonstrated significant effectiveness in improving these outcomes among home hospice caregivers.
				Quality of life	<.001	0.664 (g)	[0.269, 1.06]	↑
Milbury et al	[46]	Spousal caregivers	Mindfulness therapy	Depression	>.05	NR	NR	×	The online couple-based meditation intervention did not yield significant improvements in spousal caregivers’ depression, mindfulness, or emotional resilience, despite reported subjective benefits, indicating limited psychological efficacy in this population.
				Mindfulness	>.05	NR	NR	×
				Emotional resilience	>.05	NR	NR	×
Muscara et al	[47]	Home hospice caregivers	Acceptance and commitment therapy	PTSD	.03	1.1 (d)	[0.61, 1.59]	↑	The ACT-based intervention significantly reduced PTSD symptoms and improved emotional resilience in caregivers. However, it did not show statistically significant effects on depression, anxiety, psychological stress, or overall mental distress. The intervention may be more effective for trauma-related symptoms than for general psychological symptoms.
				Depression	.41	0.51 (d)	[0.06, 0.96]	△
				Anxiety	.69	0.6 (d)	[0.15, 1.05]	△
				Psychological stress	.16	0.44 (d)	[0, 0.89]	×
				Emotional resilience	.002	0.95 (d)	[0.48, 1.42]	↑
				Mental distress	.03	0.98 (d)	[0.51, 1.44]	↑
Valero-Cantero et al	[48]	Home hospice caregivers	Music therapy	Psychological well-being	.004	0.649 (g)	[0.205, 1.094]	↑	Self-chosen music significantly improved overall and subjective quality of life in caregivers, while psychological well-being showed no significant change. Most indicators proved effective in evaluating the intervention.
				Quality of life	.008	0.423 (g)	[−0.015, 0.861]	↑
				Subjective quality of life	<.001	0.649 (g)	[0.204, 1.094]	↑
Valero-Cantero et al	[49]	Home hospice caregivers	Music therapy	Caregiver burden	.003	−0.411 (g)	[−0.849, 0.027]	↑	Self-chosen music effectively reduced caregiver burden, validating CSI as a sensitive indicator and highlighting music’s therapeutic potential.
Wlodarczyk	[51]	Professional hospice caregivers	Music therapy	Caregiver burden	.02	−0.148 (g)	[−0.624, 0.328]	↑	Music therapy significantly reduced caregiver burden, confirming its sensitivity as a psychological indicator. However, sadness and social isolation/anticipatory grief showed no significant change.
				Sadness	.93	−0.007 (g)	[−0.482, 0.468]	×
				Social isolation/anticipatory grief	.91	−0.01 (g)	[−0.485, 0.466]	×
Karabuga Yakar et al	[42]	Professional hospice caregivers	Meditation-Based Mandala Programme	Distress	.028	−0.834 (g)	[−1.713, 0.045]	↑	The mandala intervention effectively reduced distress, depression, and anxiety in caregivers, confirming the indicators’ validity and sensitivity.
				Depression	.013	−0.934 (g)	[−1.824, −0.045]	↑
				Anxiety	.045	−0.632 (g)	[−1.493, 0.228]	↑
Hilliard	[50]	Professional hospice caregivers	Music therapy	Perceived team support	.034	−1.699 (g)	[−2.983, −0.415]	↑	Music therapy did not significantly reduce compassion fatigue in this study. While the Compassion Fatigue Scale (CFS) was used, it lacked sensitivity to detect intervention effects, suggesting limited validity in this context.
				Compassion fatigue	.271	−0.293 (g)	[−1.348, 0.762]	×
Potash et al	[52]	Professional hospice caregivers	Art therapy	Emotional exhaustion	.011	−0.303 (g)	[−0.676, 0.07]	↑	This study found that art-therapy-based supervision significantly reduced emotional exhaustion, fear of death, and death avoidance, and improved emotional awareness and approach acceptance. Effects on cynicism and mindfulness subdomains were mixed and less pronounced. Overall, key indicators proved valid and sensitive in capturing the intervention’s psychological impact.
				Cynicism	.29	0.147 (g)	[−0.224, 0.518]	×
				Mindfulness	.002	0.391 (g)	[0.017, 0.765]	↑
				Emotional awareness	.19	0.157 (g)	[−0.214, 0.528]	×
				Attentional control	.013	−0.316 (g)	[−0.689, 0.057]	↑
				Emotional acceptance	.038	−0.2 (g)	[−0.571, 0.171]	↑
				Emotional regulation	.29	0.135 (g)	[−0.236, 0.506]	×
				Fear of death	.001	−0.344 (g)	[−0.717, 0.029]	↑
				Death avoidance	.037	−0.205 (g)	[−0.577, 0.166]	↑
				Existential acceptance	.64	0.069 (g)	[−0.302, 0.439]	×
Salzano et al	[53]	Professional hospice caregivers	Art therapy	Burnout	.01	NR	NR	↑	The collaborative art-making intervention effectively reduced burnout, exhaustion, and cynicism, while enhancing emotional support among hospice caregivers, but had no significant effect on professional efficacy.
				Emotional exhaustion	.01	NR	NR	↑
				Cynicism	.01	NR	NR	↑
				Personal accomplishment	.28	NR	NR	×
				Emotional social support	.01	NR	NR	↑
Kaimal et al	[54]	Home hospice caregivers	Narrative Interview	Positive emotional	.089	0.322 (g)	[−0.735, 1.379]	△	The narrative interview group showed significant improvement in spiritual well-being, with trends toward increased positive affect and reduced negative affect, while self-efficacy, perceived stress, and physiological indicators showed no notable changes.
				Negative emotional	.77	−0.333 (g)	[−1.391, 0.725]	△
				Spiritual well-being	.42	0.23 (g)	[−0.907, 1.367]	↑
				Self-efficacy	.67	−0.22 (g)	[−1.272, 0.832]	×
				Psychological stress	.916	0.016 (g)	[−1.032, 1.063]	×
				Physiological stress	.917	0 (g)	[−1.132, 1.132]	×
				Social-emotional bonding	.249	0.25 (g)	[−0.888, 1.388]	×

ID-51: Only data from the Didactic Music Therapy group were included due to the absence of a control group and the group’s greater improvement in key outcomes.

ID-53: Only data from the art-therapy-based supervision group were extracted for single-group pre-post analysis, due to the absence of a nonintervention control group in the study design.

ID-55:This study included 2 nonrandomized intervention groups. Data were extracted from the narrative interview group for single-group pre-post analysis.

↑ = significant improvement; △ = marginal improvement; × = no significant effect; MBSR = Mindfulness-Based Stress Reduction, NR = not reported, PTSD = post-traumatic stress disorder.

The caregiver population in the included studies primarily comprised family members and informal caregivers supporting patients in palliative care settings, who frequently experience high levels of psychological burden, including stress, anxiety, and depression. Across studies, considerable heterogeneity in the type and delivery of MBTs was observed, contributing to variations in intervention responsiveness across caregiver subgroups (Table 5). Notably, informal caregivers consistently demonstrated substantial and statistically significant improvements in psychological outcomes. For example, ART and meditation therapy produced large effect sizes in alleviating complicated grief (g = −1.833), PTSD (g = −1.681), and depression (g = −1.215).^[35] Similarly, EBT was associated with enhanced quality of life and life satisfaction.^[37] In contrast, spousal caregivers exhibited more variable responses. While 1 study reported a significant reduction in depressive symptoms following a mindfulness-based intervention,^[44] another found no significant effects on depression, emotional resilience, or mindfulness when the intervention was delivered online.^[46] These findings suggest that both delivery modality and emotional proximity to the patient may influence intervention receptivity in this group. Meanwhile, ACT and music therapy were particularly effective among home hospice caregivers, who generally demonstrated favorable responses to these interventions. For instance, ACT yielded robust effects on PTSD symptoms (d = 1.10) and emotional resilience (d = 0.95),^[47] while music therapy significantly reduced caregiver burden (g = −0.411) and improved subjective quality of life (g = 0.649).^[48,49] By contrast, findings related to professional hospice caregivers revealed a lack of consistent evidence regarding the psychological benefits of mind-body interventions. Although some creative interventions, such as art therapy and mandala-based meditation, showed promise in alleviating burnout and emotional exhaustion.^[42,53] other studies failed to detect significant improvements in key indicators, including compassion fatigue, emotional awareness, and personal accomplishment.^[50,52] These mixed results may reflect the distinct occupational stressors, role-related demands, and contextual appropriateness of specific interventions within formal care environments.

In summary, although mind-body interventions generally yielded positive psychological outcomes among caregivers, treatment responses varied by caregiver role and intervention type. Informal and home hospice caregivers showed greater responsiveness to structured therapies such as meditation, ACT, and music therapy. In contrast, effects among professional and spousal caregivers were more variable or limited. These findings underscore that while MBTs hold substantial potential for enhancing caregivers’ mental health in palliative care settings, their effectiveness is shaped by factors such as caregiver role, intervention modality, and delivery format. Tailoring interventions to the specific needs and contexts of caregiver subgroups is therefore essential to maximize therapeutic benefit.

3.5. Contents of mind-body therapie interventions

Eight distinct types of MBTs were identified across the included 20 studies: ART,^[35] EBT,^[36,37] meditation therapy,^[38–42] mindfulness therapy,^[43–46] ACT,^[47] music therapy,^[48–51] art therapy^[52,53]; and narrative interview^[54] (Table 5). Among these, meditation therapy and mindfulness-based interventions were commonly applied to support stress reduction and emotional regulation.^{[38–40,42–46]} music therapy and art therapy were frequently employed in professional or home hospice contexts, emphasizing emotional expression and social connectedness.^[48–53] In contrast, interventions such as ART, ACT, EBT, and narrative interview appeared less frequently but often targeted more specific psychological domains, including grief resolution, trauma-related symptoms, and existential distress.^{[35–37,47,54]}

Each therapy type displayed distinct functional emphases. ACT emphasized psychological flexibility and trauma processing,^[47] while EBT focused on reducing emotional distress and enhancing life satisfaction by helping caregivers process existential concerns such as suffering, loss, and meaning.^[36,37] Narrative interview aimed to reduce emotional burden and support caregivers’ psychological adjustment by helping them articulate and reframe their caregiving experiences.^[54] This functional diversity across intervention types reflects the multifaceted psychological needs of caregivers in palliative care settings and underscores the importance of selecting contextually appropriate therapeutic modalities.

In terms of implementation, most interventions were conducted over a period of 4 to 12 weeks, with 8-week protocols being the most frequently adopted format.^{[35,43,47,51]} Several studies employed brief interventions, including single-session mindfulness practices,^[38,39] 7-day music-based programs,^[48,49] and 1-time narrative or art-based activities.^[54] A limited number of studies employed extended interventions lasting up to 12 weeks, particularly within mindfulness-based and couple-focused formats.^[45,46] Beyond differences in duration, the studies also demonstrated notable variability in implementation settings and follow-up assessments. These interventions were implemented across a range of care environments, including hospital-based,^{[36–39,42,43,47]} home-based,^{[40,41,44–46,51]} and outpatient settings,^[35,50–53] reflecting the contextual adaptability of MBTs within palliative care. With regard to outcome evaluation, 11 studies incorporated follow-up assessments to examine the sustained effects of the interventions,^{[35–37,40,41,43–47,51]} whereas 9 studies assessed outcomes only at post-intervention.^{[38,39,42,48–50,52–54]}

Overall, the reviewed MBTs demonstrated considerable functional heterogeneity, targeting a broad spectrum of psychological domains including stress reduction, emotional regulation, trauma resolution, and existential adjustment. These interventions also demonstrate strong contextual adaptability, being applied across inpatient, home-based, and outpatient settings through both brief and extended delivery formats. The integration of therapeutic function with care context highlights the inherent flexibility of mind-body approaches in meeting the dynamic and multidimensional needs of caregivers in palliative care, positioning them as a critical mechanism for promoting sustained psychological well-being across diverse care environments.

3.6. Description of the control intervention

Across the included studies, control conditions primarily involved usual care (UC), referring to the routine medical, psychological, or supportive services provided to caregivers within palliative care settings (Table 2). UC was used as a standard-of-care comparator to evaluate whether MBTs provided additional psychological benefits beyond routine clinical support. UC was the most frequently employed comparator and was adopted in 13 of the 20 included studies, particularly in RCTs.^{[36–38,40,41,43–48,51]} In addition, some studies implemented alternative comparator strategies to account for nonspecific therapeutic effects. These included wait-list controls^[35] and listening therapy,^[39] both of which were designed to control for attention and expectancy effects. More active comparators were also employed in select trials, such as supportive-expressive therapy and basic therapeutic education recording,^[44,49] providing structured psychosocial engagement in the absence of the targeted intervention.

In contrast, most QES did not include a formal control group, relying instead on pre-post comparisons within a single intervention cohort.^{[42,50,52–54]} The absence of standardized control conditions across these studies introduces methodological variability that may affect the estimation of effect sizes and limit cross-study comparability.

3.7. Effectiveness and outcomes of interventions

3.7.1. Depression, anxiety, and PTSD

A total of 9 studies assessed emotional and trauma-related symptoms, specifically depression,^{[35,40–44,46,47]} anxiety,^{[37,40,42,44,47]} and PTSD^{[35,43,44,47]} (Table 5). Overall, most interventions yielded favorable results, particularly those utilizing trauma-focused techniques, structured mindfulness-based programs, and acceptance-oriented approaches. Among these, ART demonstrated the most robust effects. Buck et al reported large and statistically significant improvements in complicated grief (g = −1.833, P < .0001), PTSD (g = −1.681, P < .0001), and depression (g = −1.215, P < .0001) among informal caregivers.^[35] Similarly, ACT showed considerable benefits. Muscara et al observed large effects on PTSD (d = 1.10, P = .03) and overall mental distress (d = 0.98, P = .03), although only marginal effects were found for depression (d = 0.51, P = .41) and anxiety (d = 0.60, P = .69).^[47] EBT yielded more modest but consistent results. Fegg et al reported a significant reduction in anxiety (g = −0.113, P = .006) and a marginal reduction in depression (g = −0.132, P = .09), whereas somatization remained unaffected (g = 0.025, P = .79).^[37] Furthermore, a structured mandala-based meditation program demonstrated efficacy among professional caregivers. Karabuga Yakar et al found significant reductions in depression (g = −0.934, P = .013) and anxiety (g = −0.632, P = .045), along with improvements in overall distress (g = −0.834, P = .028).^[42] In contrast, low-intensity or digital mindfulness interventions tended to show limited effects. For instance, Kubo et al reported no significant improvements in depression (g = −0.489, P = .29) or anxiety (g = −0.084, P = .79).^[41] Similarly, Milbury et al found no significant changes in depression, mindfulness, or emotional resilience (P > .05) following an online couple-based meditation program.^[46] Schellekens et al also reported nonsignificant changes in anxiety and depression (g = 0.061, P = .661) among home hospice caregivers.^[43]

3.7.2. Psychological distress and perceived stress

Nine studies investigated outcomes related to psychological distress or perceived stress, with 6 studies reporting statistically significant improvements (Table 5). Effect sizes ranged from moderate to very large (g = −0.411 to − 1.256), predominantly in interventions involving facilitator-guided mindfulness, meditation, or acceptance-based approaches. For example, Türkoğlu and Kavuran observed a substantial reduction in psychological stress among home hospice caregivers following a mindfulness-based stress reduction (MBSR) program (g = −1.256, P < .001).^[45] Likewise, a mandala-based meditation intervention significantly alleviated distress in professional caregivers (g = −0.834, P = .028).^[42] Brief, low-dose interventions also demonstrated efficacy: Tan et al reported significant reductions in both psychological suffering (P = .036) and perceived stress (P < .0001) following a 20-minute mindful breathing exercise,^[38] while Beng et al found a significant decline in distress (P = .001) after just a 5-minute session.^[39] In another study, Kubo et al reported reduced distress (P = .03) among home hospice caregivers following a mindfulness-based intervention.^[40] EBT resulted in a significant reduction in psychological distress (P = .02), along with a marginal improvement in subjective quality of life (P = .06). However, effect sizes were not reported.^[36] While facilitator-guided interventions demonstrated significant psychological benefits, self-directed or digital formats yielded largely null or inconsistent outcomes. For instance, Kubo et al found no significant reduction in distress from a mobile mindfulness program (g = −0.44, P = .47)^[41]; Muscara et al reported a nonsignificant change following ACT (d = 0.44, P = .16)^[47]; and Kaimal et al observed no effect from narrative interviews (g = 0.016, P = .916).^[54]

3.7.3. Quality of life and emotional well-being

The 7 studies explored outcomes related to quality of life (QoL), life satisfaction, subjective well-being, and affective functioning (Table 5). Among them, 6 studies reported statistically significant or marginal improvements, particularly through music therapy, EBT, and acceptance- or mindfulness-based approaches. Music therapy produced the most consistent benefits; Valero-Cantero et al found significant improvements in both subjective QoL (g = 0.649, P < .001) and overall QoL (g = 0.423, P = .008) among home hospice caregivers,^[48] while a separate study by the same team showed a notable reduction in caregiver burden (g = –0.411, P = .003).^[49] In parallel, EBT (EBT) demonstrated multidimensional efficacy, with Fegg et al reporting significant gains in subjective QoL (g = 0.426, P < .001), overall QoL (g = 0.224, P = .007), and life satisfaction (g = 0.159, P = .009), along with reduced negative affect (g = 0.152, P = .003) and a marginal increase in positive affect (g = 0.457, P = .06).^[37] ACT also contributed to improved emotional outcomes; Muscara et al observed significant enhancements in emotional resilience (d = 0.95, P = .002) and reductions in mental distress (d = 0.98, P = .03).^[47] Additionally, art therapy–based clinical supervision supported emotional processing capacities, as Potash et al reported improvements in attentional control (g = –0.316, P = .013), emotional acceptance (g = –0.200, P = .038), and mindfulness (g = 0.391, P = .002), although effects on emotional exhaustion, awareness, and regulation were mixed or nonsignificant.^[52] Relative to the aforementioned interventions, narrative interview approaches exhibited comparatively modest effects. Kaimal et al observed no significant changes in spiritual well-being (g = 0.23, P = .42) or positive affect (g = 0.322, P = .089), and reported no discernible improvements in self-efficacy, perceived stress, or social bonding (P > .6 across subscales).^[54]

3.7.4. Mindfulness and self-regulation capacities

Six studies examined outcomes related to mindfulness, emotional regulation, attentional control, or resilience (Table 5). Among these, 4 studies reported statistically significant improvements, particularly when interventions were delivered in-person, facilitated by therapists, or incorporated expressive arts. For instance, Kubo et al found that an 6 weeks mindfulness-based intervention significantly enhanced mindfulness among home hospice caregivers (P = .03), although the effect size was not reported.^[40] In a mobile-delivered version of the same intervention, Kubo et al also observed a significant increase in mindfulness (g = 0.306, P = .03), despite null effects on distress, depression, and anxiety.^[41] Potash et al evaluated a clinical supervision program grounded in art therapy principles, which yielded multiple self-regulatory benefits, including significant improvements in mindfulness (g = 0.391, P = .002), attentional control (g = −0.316, P = .013), and emotional acceptance (g = −0.200, P = .038); however, gains in emotional awareness (g = 0.157, P = .19) and emotional regulation (g = 0.135, P = .29) did not reach statistical significance.^[52] Similarly, Muscara et al reported a large and significant effect of ACT on emotional resilience (d = 0.95, P = .002), further supporting its potential to enhance adaptive self-regulation among home hospice caregivers.^[47] Conversely, mindfulness-based interventions delivered through a couple-based online format^[46] or to professional hospice staff^[43] did not yield significant improvements in mindfulness or emotional resilience (P > .05).

Overall, this systematic review demonstrates that MBTs offer measurable psychological benefits for caregivers in palliative care, particularly across 4 domains: emotional and trauma-related symptoms (depression, anxiety, PTSD), psychological distress, quality of life, and self-regulation capacities. Across the included studies, interventions that were structured, therapist-guided, and contextually tailored consistently outperformed self-directed or digital formats in improving psychological outcomes.Trauma-focused therapies such as ART, acceptance-oriented interventions including ACT and EBT, as well as expressive arts-based approaches such as music therapy, art therapy, and mandala meditation, have demonstrated the greatest potential for producing positive outcomes. These findings highlight the critical role of delivering mind-body interventions in a personalized and guided format, preferably through in-person sessions facilitated by trained professionals, to optimize mental health improvements among caregivers in palliative care settings.

4. Discussion

This systematic review synthesizes evidence from 20 studies examining the impact of MBTs on caregivers’ mental health in palliative care settings, with particularly consistent benefits observed in reducing emotional distress and enhancing self-regulation capacities. The included studies demonstrated a wide range of psychological improvements, especially in domains related to depression, anxiety, PTSD, psychological distress, quality of life, and self-regulatory functioning. However, these therapeutic gains were not uniformly distributed across all intervention formats or caregiver subgroups. Structured, therapist-guided interventions generally produced more robust psychological outcomes than self-directed or digitally delivered formats. Interventions such as ART,^[35] ACT,^[47] and expressive arts-based approaches^{[42,48,49,52]} were consistently associated with improvements in both emotional and cognitive outcomes. In contrast, self-guided interventions like mobile mindfulness apps^[41] and online couple-based meditation programs^[46] showed limited effectiveness across key psychological indicators. These findings highlight the critical role of delivery modality in optimizing intervention impact: therapist-guided formats offer consistent engagement, emotional processing, and tailored feedback, while self-directed formats may lack the structure necessary to sustain psychological benefit. As Ahn et al noted, structured and guided interventions are particularly suited to meet the complex emotional needs of caregivers in high-burden settings by fostering cognitive-emotional integration and adaptive coping.^[57]

Importantly, the effectiveness of MBTs was also shaped by the caregiver’s role and relational context. Informal and home hospice caregivers showed the most consistent improvements, particularly when engaged in structured, therapist-led formats. These individuals, often lacking formal training and facing high emotional demands, appeared especially responsive to interventions targeting grief, trauma-related symptoms, and caregiver burden.^{[35–37,40,54]} In contrast, outcomes among spousal caregivers were more variable. While some in-person mindfulness-based programs demonstrated psychological benefit,^[44,45] digital self-guided formats yielded minimal effects,^[40,41] suggesting that emotional proximity to the patient and limited interpersonal engagement may constrain intervention responsiveness. Among professional hospice caregivers, the evidence was mixed. Creative modalities such as art therapy and mandala meditation offered some relief from burnout and emotional exhaustion,^[42,52,53] but other studies reported negligible improvements in broader indicators like compassion fatigue and emotional well-being.^[50,51] These disparities likely reflect role-specific demands, baseline resilience, and the degree to which interventions align with institutional care dynamics. Building on these subgroup-specific insights, it is worth noting that multiple national policy frameworks have increasingly recognized caregiver mental health as a critical component of long-term care systems. For instance, the UK Carers Action Plan,^[58] Australia’s National Carer Strategy,^[59] and the Organisation for Economic Co-operation and Development (OECD)^[60] have all emphasized the importance of enhancing caregivers’ emotional resilience and psychological support structures. However, these policies primarily focus on macro-level institutional guarantees and service subsidies, with insufficient attention to the heterogeneity of caregivers’ emotional needs, psychological burdens, and receptivity to interventions across subgroups.

This review reveals a critical role of caregiver type in moderating the effectiveness of MBTs. Compared to professional and spousal caregivers, informal and home hospice caregivers exhibited more stable and significant psychological improvements in response to structured, therapist-guided interventions. Such differences may stem not only from disparities in emotional stress and resource accessibility, but also from the degree of contextual alignment between intervention strategies and care environments. To support the clinical integration of MBTs into palliative care practice, several implementation strategies warrant consideration. First, interventions should be embedded into routine caregiver support services, particularly within hospice and home-based settings where informal caregivers face substantial emotional burdens. Modular, low-cost approaches such as MBSR, ACT, and expressive arts therapies can be adapted for delivery by trained facilitators or incorporated into community-based programs. Second, equipping palliative care teams with foundational training in mind-body techniques would enable trauma-informed support during caregiver interactions and enhance continuity of psychosocial care. Third, while digital MBTs show mixed efficacy, hybrid models combining online tools with intermittent therapist guidance may improve both accessibility and engagement, particularly in underserved contexts. Finally, institutional and policy-level support is essential, including the integration of MBTs into caregiver care pathways, routine psychological assessments, and the development of supportive policies to promote scalable and sustainable implementation. Tailoring intervention formats to specific caregiver roles and sociocultural contexts will be key to maximizing their relevance and therapeutic impact.

5. Limitations

This systematic review presents a comprehensive synthesis of MBTs for caregivers in palliative care settings, yet several limitations warrant consideration. First, the included studies exhibited substantial heterogeneity in intervention modalities, delivery formats, duration, and target populations. While this diversity reflects the adaptability of MBTs across settings, it also complicates direct comparisons and precludes meta-analytic synthesis. In particular, the lack of standardized intervention protocols, ranging from single-session breathing exercises to 12-week structured programs, makes it challenging to determine optimal intervention dosage and modality. Second, methodological limitations were evident in the primary studies. Many trials suffered from small sample sizes, inadequate blinding, and incomplete follow-up data, thereby affecting internal validity and limiting the capacity to draw firm conclusions about long-term effects. Although quality assessments were conducted using JBI and Cochrane tools, only a minority of randomized trials were rated as high quality, and several QES lacked control groups, increasing susceptibility to bias. Third, variations in outcome measurement tools introduced inconsistencies in evaluating psychological effects. Subjective outcomes such as depression, anxiety, and distress were assessed using diverse, and sometimes nonvalidated, instruments, which limits cross-study comparability. Additionally, only 11 of the 20 studies included follow-up assessments, and most evaluated outcomes immediately post intervention, thereby failing to capture the sustainability of therapeutic benefits. Furthermore, the exclusive inclusion of English-language publications may have introduced language bias and limited the scope of the evidence base.

Future research should prioritize standardized intervention protocols and validated outcome measures, adopt more rigorous experimental designs with sufficient statistical power, and examine long-term effects through extended follow-up. Additionally, stratified analyses by caregiver type, setting, and psychosocial context are needed to refine intervention targeting. Expanding research to include non-English studies and culturally diverse populations will enhance generalizability and support the development of tailored, equitable mental health interventions for caregivers in palliative care.

6. Conclusion

This systematic review synthesized findings from 20 studies assessing the effects of MBTs on caregivers’ mental health in palliative care. Interventions included mindfulness-based programs, ACT, ART, meditation, music, and art therapies, delivered through various formats and settings. Overall, MBTs yielded notable short-term improvements in depression, anxiety, PTSD, psychological distress, quality of life, and self-regulation. Therapist-guided interventions consistently outperformed self-directed or digital formats. Informal and home hospice caregivers benefited most, while effects among spousal and professional caregivers were mixed. However, methodological heterogeneity, inconsistent outcome measures, and limited follow-up constrained comparability and prevented meta-analysis. Future research should prioritize standardized protocols, validated assessment tools, and longer-term evaluations. Stratified analyses by caregiver type and culturally tailored delivery models are also needed to enhance intervention precision and applicability. In parallel, implementation strategies – including integration into routine care, caregiver-specific delivery models, and policy-level support – should be explored to translate MBT benefits into sustainable clinical practice. These findings highlight the potential of MBTs as scalable, person-centered strategies to support caregiver mental health within palliative care systems.

Acknowledgments

This study was supported by the Faculty of Humanities and Arts and the Faculty of Chinese Medicine at a university in Asia. The academic resources, guidance, and valuable advice provided by both faculties played a crucial role in the successful completion of the study. The research team is deeply grateful for the opportunity to collaborate with these faculties and appreciates their support and assistance throughout the research process.

The authors would like to express their sincere gratitude to Assistant Professor Jue Wang and Professor Guanghui Huang for their valuable guidance and support throughout the research and manuscript preparation.

Author contributions

Conceptualization: Jifeng Shen.

Data curation: Zhong Zheng, Linyang Li, Xuexing Luo.

Methodology: Xuexing Luo.

Supervision: Jue Wang, Guanghui Huang.

Validation: Zhong Zheng.

Writing – original draft: Jifeng Shen, Jing Xu, Zhong Zheng.

Writing – review & editing: Jifeng Shen, Jing Xu, Jue Wang, Guanghui Huang.