Caregiver burden, quality of life, and psychosocial outcomes among informal caregivers of patients with glioma: A systematic review and meta-analysis Authors: Haimeng Wei, Shuo Zhang, Shuai Tao, Jing Li, Xiaonan Liu Published: 2026 ABSTRACT AbstractObjectiveInformal caregivers of people with glioma frequently provide complex, long-term support in the context of neurological, cognitive and behavioural impairment, which may adversely affect caregiver burden, quality of life (QoL), and broader psychosocial wellbeing, including anxiety, depression, and distress. We synthesised evidence on caregiver burden and psychosocial outcomes among caregivers of patients with glioma, and compared outcomes with caregivers of other cancers where data permitted.MethodsWe conducted a systematic review and meta-analysis in accordance with PRISMA 2020. PubMed, Embase, Scopus, PsycINFO and the Cochrane Library were searched from inception to the final search date. Eligible studies reported caregiver burden, QoL, anxiety, depression and/or distress among caregivers of patients with glioma. For single-group analyses, pooled prevalences of burden, anxiety, depression and distress were estimated using random-effects meta-analysis of proportions (Freeman–Tukey transformation). For comparative analyses, standardized mean differences (SMDs) were pooled using random-effects inverse-variance models (DerSimonian–Laird) comparing caregivers of patients with glioma with caregivers of other cancers.ResultsTen studies were included. Clinically significant caregiver burden was common (pooled prevalence = 60%). Pooled prevalences of depression and anxiety were 35% and 36%, respectively, while distress was high (68%). Comparative meta-analysis (3 studies; 710 participants) showed no statistically significant difference in caregiver QoL between glioma and other-cancer caregiver groups (SMD = −0.21). Heterogeneity was substantial across outcomes.ConclusionsInformal caregivers of patients with glioma experience substantial psychosocial morbidity, including high distress, frequent burden, and common anxiety and depressive symptoms. Routine caregiver assessment and integrated caregiver-focused supportive care should be embedded in neuro-oncology pathways.Keywords: Caregivers, Glioma, Meta-analysis, Quality of life Introduction Gliomas are the most common malignant primary tumours of the central nervous system and are frequently associated with substantial morbidity across neurological, cognitive, and behavioural domains.1 High-grade glioma particularly glioblastoma, typically follows an aggressive clinical course requiring multimodal treatment, repeated monitoring, and rapid adaptation to functional decline.2 Population-level registry data continue to show that malignant brain tumours carry a major health burden, with glioblastoma remaining a predominant entity among adult malignant brain tumours and contributing disproportionately to mortality and disability.3 Most day-to-day care services are delivered outside the hospital setting by informal caregivers, such as spouses, partners, family members, or close friends. Glioma-related neurological, cognitive, and behavioural impairments affect patient mobility, communication, personality, executive function, and judgement. Hence, caregivers are required to coordinate complex clinical regimens and also simultaneously adapt to rising practical and supervisory demands.4 Caregiving in glioma differs from many other cancer types because disease often produces early neuro-oncological symptoms in patients that can disrupt personality, behaviour, communication, and functional independence.5 Functional dependence may increase after surgery, during chemotherapy and radiotherapy, or with tumour progression. This requires caregivers to take greater responsibility for medication management, monitoring of symptoms, transportation, assistance with activities of daily living, and decision-making support.6 Due to these patient-related impairments and care demands, caregivers may develop “caregiver burden,” which is a multidimensional construct consisting of time dependence, physical strain, emotional distress, social restriction, and financial pressure.7 Consistent caregiver burden might contribute to reduction in caregiver health-related quality of life (QoL), increased anxiety and depressive symptoms, sleep disruption, and reduced participation in employment and social activities. This has potential downstream effects on patient care through reduced caregiving capacity and increased dyadic stress.8 Conceptually, caregiver outcomes in glioma may be understood as downstream consequences of patient-related neurological and functional impairment interacting with intensity, duration, and context of informal provision of care. Empirical studies in patients with glioma suggest that caregiver burden is common and clinically meaningful.6, 7, 8, 9, 10 Previous studies have shown that a significant number of family caregivers of patients with glioma report moderate to severe levels of burden, with respect to reduction in sleep duration and increased anxiety and depressive symptoms.9 Similar work in glioblastoma caregivers has also highlighted that burden is not uniform across clinical settings and might vary according to treatment phase and patient clinical characteristics.10 This reinforces the importance of disease and treatment-related factors in shaping caregiving demands. These findings align with broader neuro-oncology literature in which caregiver distress arises not only from practical care tasks, but also from prognostic uncertainty, the need for vigilance regarding neurological deterioration, and the psychological impact of witnessing progressive cognitive and functional loss in the patient.11 Caregiver QoL and support needs may be particularly noticeable during transitions in care progression, treatment failure, and end-of-life periods when symptom burden escalates or care coordination intensifies.12 In a multicentre retrospective study focusing on the last months of life in glioblastoma, caregivers reported low QoL alongside symptoms suggestive of burnout and a strong need for targeted psychological support and caregiver training. Qualitative data similarly depict caregiving as an isolating role characterised by feeling unprepared, under-informed, and insufficiently supported by health care systems.13 Recent caregiver narratives emphasise the importance of proactive education, clearer anticipatory guidance, and recognition of caregivers as core members of the care team. This invisible patient phenomenon where caregiver needs remain secondary to tumour-directed care has particular salience in glioma because rapid clinical changes can outpace caregivers' ability to adapt.14 Across brain cancer more broadly, qualitative synthesis indicates that supportive care experiences are shaped by persistent informational, psychosocial, and practical unmet needs for both patients and informal caregivers, with gaps spanning communication, access to services, and navigation of palliative and rehabilitation resources.15 Recent qualitative work in glioblastoma caregivers reinforces these themes, describing overwhelming burden, challenges coping with the caregiving role, and deficiencies in caregiver support, including limited education and confusion regarding hospice and palliative pathways. Collectively, these data suggest that caregiver burden and QoL are core outcomes in glioma care, reflecting both the intensity of caregiving tasks and the broader psychosocial ecology in which caregiving occurs.15 Despite growing attention, the magnitude of caregiver burden and QoL impairment in glioma caregiving, as well as the frequency of related psychosocial outcomes such as anxiety, depression, and distress, remain variably reported across studies using different instruments and sampling frames. Therefore, this review was undertaken to synthesise evidence on caregiver burden, QoL, and associated psychosocial outcomes among informal caregivers of patients with glioma. In this review, caregiver burden and caregiver QoL were considered the primary outcomes, while anxiety, depression, and psychological distress were treated as related but conceptually distinct secondary psychosocial outcomes rather than interchangeable components of burden. Methods This systematic review and meta-analysis was conducted and reported in accordance with the PRISMA 2020 statement. A PRISMA flow diagram was used to document the study selection process.16 Eligibility criteria Studies were selected if they satisfied the following criteria. Participants Informal (unpaid) caregivers, including family members, partners, and friends, who provided practical, emotional, or decision-making support to adult or paediatric patients with glioma (any grade, including glioblastoma). Studies in mixed brain tumour samples were eligible if glioma-specific caregiver data were extractable. No age-based exclusions were made. Comparators For comparative meta-analysis, eligible studies included a comparison group of caregivers supporting patients with other cancer types (non-glioma). For single-group synthesis, no comparator was required. Outcomes Studies reported at least one of the following caregiver outcomes, measured using validated instruments or clearly described tools: i) Caregiver burden; ii) QoL; iii) Anxiety; iv) Depression; v) Psychological distress. For the single-group meta-analysis (glioma caregivers only), burden, anxiety, depression, and distress were synthesized as proportions meeting a study-defined or instrument-validated threshold (e.g., “clinically significant” symptoms). For the comparative meta-analysis, all outcomes (including QoL) were synthesized as standardized mean differences (SMDs) between glioma caregivers and caregivers of other cancers. Study designs Observational studies (cross-sectional, cohort, case–control) and baseline data from interventional studies were eligible if relevant caregiver outcomes were reported prior to intervention delivery. Qualitative-only studies, case reports/series without extractable quantitative outcomes, conference abstracts without sufficient data, editorials, narrative reviews, and animal studies were excluded. Other information Studies were required to report sufficient quantitative information to compute proportions (numerator/denominator) and/or effect sizes (means/SDs or equivalent convertible statistics). When multiple reports described the same cohort, the most complete dataset was used. Search strategy Electronic database searches were conducted from inception to January 2026 in MEDLINE (PubMed), Embase, Scopus, PsycINFO, and Cochrane Library. In addition, reference lists of included studies and relevant reviews were screened to identify additional eligible studies. Final search strategy is available in Supplementary File 1. Study selection process Two reviewers independently screened titles/abstracts for eligibility, followed by full-text screening of potentially eligible reports. Discrepancies were resolved through discussion; if unresolved, a third reviewer adjudicated. Reasons for full-text exclusions were recorded and reported. Data extraction and management Two reviewers independently extracted data using a piloted standardized extraction form. When required information was missing or unclear, attempts were made to contact study authors. The primary outcomes were caregiver burden and caregiver QoL. Secondary outcomes were caregiver anxiety, depression, and psychological distress. These secondary outcomes were conceptualized as distinct indicators of caregiver psychological morbidity that may accompany or be associated with caregiving burden, rather than as direct components of the burden construct itself. For each outcome, we extracted the following: instrument name, whether the outcome was reported as continuous (mean/SD) and/or categorical (above a threshold), threshold definition (instrument-validated cut-off, study-defined cut-off, or clinical categorization), any available information on scoring approach or study-specific classification used to define clinically significant burden or symptoms. These details were extracted because differences in outcome measurement and threshold definitions were anticipated to be an important source of between-study heterogeneity, particularly for meta-analyses of prevalence. For comparative analyses, we extracted outcome data separately for glioma caregivers and other-cancer caregivers within the same study. Risk of bias assessment Risk of bias for observational studies was assessed independently by two reviewers using the Newcastle–Ottawa Scale (NOS) for cohort/case–control studies and an adapted NOS for cross-sectional studies.17 Domains assessed included selection/representativeness, comparability/confounding control, and outcome ascertainment/statistical reporting. For transparency, domain-level judgments for each included study are presented in Table 2. Studies receiving stronger ratings across selection, comparability, and outcome domains were considered at lower risk of bias, whereas studies with important limitations in sample representativeness, limited control for confounding, or less robust outcome assessment/reporting were judged to be at moderate or high risk of bias.18 Disagreements were resolved by consensus or third-party adjudication. Table 2Risk of bias assessment of the included studies (N = 10).Table 2Study (Year)SelectionComparabilityOutcomeRisk of biasAu (2022)Two starsTwo starsTwo starsModerateChoi (2012)Three starsTwo starsThree starsLowFlores (2014)Three starsTwo starsTwo starsLowGulino (2024)One starOne starOne starHighHeckel (2018)Three starsTwo starsTwo starsLowJacobs (2014)Three starsTwo starsThree starsLowRenovanz (2018)Two starsTwo starsTwo starsModerateSacher (2018)Three starsTwo starsTwo starsLowTrad (2015)One starZero starOne starHighWasner (2013)One starOne starOne starHighRisk of bias was assessed using the Newcastle–Ottawa Scale (NOS) for cohort/case–control studies and an adapted NOS for cross-sectional studies. Overall judgments were based on study performance across the selection, comparability, and outcome domains, with lower ratings across multiple domains contributing to moderate- or high-risk classifications. Risk of bias assessment of the included studies (N = 10). Risk of bias was assessed using the Newcastle–Ottawa Scale (NOS) for cohort/case–control studies and an adapted NOS for cross-sectional studies. Overall judgments were based on study performance across the selection, comparability, and outcome domains, with lower ratings across multiple domains contributing to moderate- or high-risk classifications. Statistical analysis For caregiver burden, anxiety, depression, and distress, the effect measure was proportion of caregivers meeting clinically significant symptom/burden thresholds (as defined by validated instrument cut-offs or study-defined criteria). Where multiple thresholds were reported, the most widely accepted clinical cut-off for the instrument was prioritized; otherwise, the authors’ primary definition was used. For QoL, the effect measure was the standardized mean difference (SMD), calculated with 95% confidence intervals. Directionality was harmonized so that positive SMD values consistently represented better status among glioma caregivers. All quantitative syntheses were performed in Stata software (StataCorp, College Station, TX, USA) using the “metaprop” command for pooled proportions and the “metan” command for pooled SMDs. Given expected clinical and methodological heterogeneity (differences in glioma grade, treatment phase, caregiver relationship, and outcome instruments), a random-effects model was used for all meta-analyses. For burden, anxiety, depression, and distress among glioma caregivers, pooled prevalence estimates were generated using “metaprop” with appropriate variance stabilization (Freeman–Tukey double arcsine transformation), and results were back-transformed for interpretability. Studies that did not provide a clear numerator/denominator or could not be converted to a proportion using a defensible threshold were excluded from proportion pooling but were retained for qualitative synthesis. For studies with a comparator group of caregivers of other cancers, outcome means and SDs were pooled using “metan” to estimate pooled SMDs. If a study reported multiple eligible instruments for the same construct (e.g., two depression scales), a prespecified hierarchy was applied prioritizing the most commonly used and validated scale in oncology caregiver research; otherwise, the instrument most frequently reported across included studies was selected to avoid double-counting. If multiple time points were reported, baseline or the time point most comparable across studies (typically closest to diagnosis/treatment initiation) was preferentially extracted; alternative time points were explored in sensitivity analyses when data permitted. Statistical heterogeneity was assessed using Cochran's Q (with p-value) and quantified using I2 with corresponding interpretation based on conventional thresholds. Between-study variance was estimated as τ2 under the random-effects model. Sensitivity analysis was performed to assess the robustness of the results by excluding studies at high risk of bias. Publication bias was assessed using a funnel plot with Egger's test for depression outcome as it had a reasonable number of studies. Results Search results A total of 5559 records were identified across databases (PubMed 2,134, Embase 1,192, Cochrane 221, Scopus 1,230, PsycINFO 782), and 1334 duplicates were removed, leaving 4225 records for title/abstract screening. After excluding 4109 records, 116 full-text reports were assessed for eligibility. Of these, 106 full-text reports were excluded for the following reasons: different population (n = 81), including non-glioma or non-brain tumour populations (n = 34), mixed tumour/cancer samples without extractable glioma-specific caregiver data (n = 21), patient-only samples without caregiver data (n = 12), and professional rather than informal caregiver populations (n = 14); no relevant outcomes of interest reported (n = 19); and insufficient quantitative data for extraction/synthesis (n = 6). The remaining 10 studies were included in the review (Fig. 1).10,13,14,19, 20, 21, 22, 23, 24, 25 PRISMA 2020 flow diagram of study selection Fig. 1. PRISMA 2020 flow diagram showing the identification, screening, eligibility assessment, and final inclusion of studies evaluating caregiver burden, quality of life, and psychosocial outcomes among informal caregivers of patients with glioma. Characteristics of the included studies Across the 10 included studies, designs were heterogeneous, comprising cross-sectional comparisons/mixed-methods studies, longitudinal descriptive/trajectory analyses, matched case–control studies, multicentre cross-sectional surveys, prospective cohort studies, and one multicentre retrospective end-of-life study. The evidence base was geographically diverse, with studies conducted in the USA (n = 3), Germany (n = 4), Australia (n = 1), France (n = 1) and Italy (n = 1). Most samples focused on caregivers of high-grade glioma/glioblastoma patients (often WHO grade III/IV), including caregiving during chemoradiation, post-surgery survivorship, and the end-of-life phase (Table 1). Table 1Characteristics of the included studies (N = 10).Table 1Study (Year)Study DesignOrigin of StudyCharacteristics of ParticipantsIntervention group detailsControl group detailsOutcome measurement instrumentsNumber of Samples in Intervention Group#Number of Samples in Control Group#Mean Age (Int / Con)Male/female (intervention/control)Au (2022)Cross-sectionalUSA (7 academic cancer centers)Primary caregivers of adults living with glioblastomaCaregivers of patients with Frontal Lobe lesionsCaregivers of patients with Non-frontal Lobe lesionsGBM-Specific Instrument for caregiver burdenCQOL-C for QoL5611158 (SD 12) (Overall Caregiver Mean)Overall Caregivers: 62M / 105FChoi (2012)Cross-sectionalUSA (Western Pennsylvania)103 family caregivers of high-grade gliomaCaregivers assigned to High or Moderate distress/burden trajectories (e.g., "High-decreasing" or "Consistently High" groups).Caregivers assigned to Low or Low-decreasing distress/burden trajectories.CRA (Schedule) for caregiver burdenCESD-10 for depressionPOMS for anxiety4846Total Sample: 51.4 (SD 11.4)Total Sample: 27M : 76FFlores (2014)Matched Case-ControlFrance (Marseilles)Caregivers of glioma patients (WHO grade II-IV)Caregivers of patients with primary gliomasCaregivers of patients with other cancers (breast, lung, etc.)CarGOQoL (Index) for QoL4718851.6 / 52.9314M:36F / 56M:132FGulino (2024)Retrospective cohortItaly (Palermo)Caregivers of deceased glioblastoma patientsCaregiving during the end-of-life phase (last 3 months)None (Single arm)End-of-life research questionnaire developed and validated by Sizoo et al, which assesses caregiver burden, depression and anxiety.24NA48.78 (SD 10.9) / N/A8M:16F / N/AHeckel (2018)Cross-sectionalGermany (Erlangen)Informal caregivers (family/friends) supporting home care.Caregivers of patients with Glioblastoma WHO-Grade IV.Caregivers of patients with non-brain tumors (e.g., abdomen, head/neck).BSFC for caregiver burdenPHQ for depression and anxiety171157.5 / 59.55M:12F / 2M:9FJacobs (2014)Cross-sectionalUSA (Chicago, IL)Primary caregivers (73.3% spouses) of adults with Grade III/IV glioma.Caregivers of patients with malignant glioma.Caregivers of patients with lung, breast, or prostate cancer.CQOL-C for QoLDT scale for distress4526351.4 / NR (Caregivers).16M:29F / NR.Renovanz (2018)Cross-sectionalGermany (Mainz, Ludwigsburg, Ulm)Caregivers and Glioma Patients (84% HGG).Caregivers (mostly life partners) of glioma patients.Glioma patients (matched subgroup for caregiver comparison).CPQ for depression and anxiety969656 (14.8) / 52 (14)129M:103F (Total sample)Sacher (2018)CohortGermany (Leipzig)Matched patient-caregiver dyads; HGG patients (WHO III/IV) at least 3 months post-surgery.Primary informal caregivers of HGG patients.NAMultidimensional Mood State Questionnaire for psychological distress and the Center for Epidemiology Studies Depression Scale for depression outcome, 12-Item Short Form Medical Outcome Questionnaire or SF-12 for caregiver burden45NA54.16 (Caregivers)12M:33F (Total sample)Trad (2015)Cross-sectionalAustralia (Southwestern Sydney)Informal caregivers of adult primary brain tumor (PBT) patients (High-grade, low-grade, and benign).Caregivers supporting patients with High-Grade Glioma (HGG) at initial diagnosis.Caregivers supporting patients with Benign Brain Tumors (BBT) at initial diagnosis.DT scale for distress226NR / NRNR / NRWasner (2013)Cross-sectionalGermany (Munich)Caregivers of PMBT patients (Astrocytoma III / Glioblastoma)Informal family caregivers (spouses, parents, children)NABSFC for caregiver burdenSEIQoL-DW for QoLHADS-D for depressionHADS-A for anxiety27NA51.0 (Range 34-80) / NA5M : 22 F / NACG - Caregiver; HGG - High-grade glioma; GBM - Glioblastoma multiforme; WHO - World Health Organization; RCT - Randomized controlled trial; SD - Standard deviation; NA - Not applicable; NR - Not reported; PBT - Primary brain tumor; PMBT - Primary malignant brain tumor; Pt - Patient.Abbreviations: CRA, Caregiver Reaction Assessment; CESD-10, 10-item Center for Epidemiologic Studies Depression Scale; POMS, Profile of Mood States; CarGOQoL, CareGiver Oncology Quality of Life questionnaire; BSFC, Burden Scale for Family Caregivers; PHQ, Patient Health Questionnaire; CQOL-C/CQOLC, Caregiver Quality of Life Index-Cancer; DT, Distress Thermometer; CPQ, Caregivers’ Perspective Questionnaire; SEIQoL-DW, Schedule for the Evaluation of Individual Quality of Life–Direct Weighting; HADS-D, Hospital Anxiety and Depression Scale–Depression; HADS-A, Hospital Anxiety and Depression Scale–Anxiety# The number of participants in each group differs from the total sample size because only those included in the analysis are reported under the group-specific counts. Characteristics of the included studies (N = 10). CG - Caregiver; HGG - High-grade glioma; GBM - Glioblastoma multiforme; WHO - World Health Organization; RCT - Randomized controlled trial; SD - Standard deviation; NA - Not applicable; NR - Not reported; PBT - Primary brain tumor; PMBT - Primary malignant brain tumor; Pt - Patient. Abbreviations: CRA, Caregiver Reaction Assessment; CESD-10, 10-item Center for Epidemiologic Studies Depression Scale; POMS, Profile of Mood States; CarGOQoL, CareGiver Oncology Quality of Life questionnaire; BSFC, Burden Scale for Family Caregivers; PHQ, Patient Health Questionnaire; CQOL-C/CQOLC, Caregiver Quality of Life Index-Cancer; DT, Distress Thermometer; CPQ, Caregivers’ Perspective Questionnaire; SEIQoL-DW, Schedule for the Evaluation of Individual Quality of Life–Direct Weighting; HADS-D, Hospital Anxiety and Depression Scale–Depression; HADS-A, Hospital Anxiety and Depression Scale–Anxiety # The number of participants in each group differs from the total sample size because only those included in the analysis are reported under the group-specific counts. Sample sizes varied widely, ranging from 24 caregivers in an end-of-life glioblastoma cohort to 308 caregivers in a cross-sectional comparison study, with several studies enrolling approximately 100–200 participants. The mean age of caregivers was approximately 48.8–58.0 years, with one study reporting an age range of 34–80 years, and most studies had predominantly female caregivers. Comparative groups (where available) have included caregivers of other cancer types like breast, lung, prostate, or other non-brain tumours. Outcome measurement approaches have varied across studies and all the tools were summarized in Table 1. Domain level risk of bias assessments were summarized in Table 2. This provides study-specific judgments across NOS domains and overall classification. Overall risk of bias rating showed that five studies rated to have low risk, two moderate, and three high risk of bias. Studies rated as low risk had better ratings across all three domains (selection, comparability, and outcome domains), by having representative sampling, clear outcome measurement, and appropriate statistical reporting. Studies rated as moderate risk had fewer limitations in representativeness or outcome reporting despite having appropriate comparability. Studies classified as high risk had weaknesses across almost all the domains, particularly in selection, confounding control, and outcome ascertainment, which reduced confidence in their estimates. Caregiver burden In the single-group meta-analysis of caregiver burden (3 studies), the pooled prevalence of clinically significant burden among glioma caregivers was 60% (random-effects; 95% CI 35% to 83%) (Fig. 2). Individual study estimates ranged from 33% to 75%. There was substantial heterogeneity across studies (I2 = 83.46%, χ2 = 12.09, df = 2, P = 0.002), with between-study variance τ2 = 0.16. Sensitivity analysis excluding high-risk-of-bias studies was not feasible for caregiver burden, because two of the three studies contributing to this pooled estimate were judged to be at high risk of bias, leaving insufficient studies for a meaningful re-analysis. Forest plot of pooled prevalence of caregiver burden among informal caregivers of patients with glioma Fig. 2. Forest plot showing the pooled prevalence of clinically significant caregiver burden among informal caregivers of patients with glioma. Pooled estimates were generated using a random-effects meta-analysis of proportions. Squares represent study-specific prevalence estimates, horizontal lines indicate 95% confidence intervals, and the diamond represents the pooled prevalence estimate. Caregiver QoL In the comparative meta-analysis of caregiver QoL, 3 studies comprising 710 participants contributed data. The pooled SMD (random-effects, DerSimonian–Laird) showed no statistically significant difference in QoL between caregivers of patients with glioma and caregivers of other tumour patients (SMD = −0.21, 95% CI −1.15 to 0.72; z = −0.45, P = 0.655) (Fig. 3). Individual study effects were inconsistent in direction (SMD range: −1.06 to 0.61). Between-study heterogeneity was considerable (Q = 49.45, df = 2, P < 0.001; I2 = 96.0%), indicating substantial variability across included studies. None of the three studies had high risk of bias and hence, sensitivity analysis was not applicable. Forest plot of comparative quality of life between caregivers of patients with glioma and caregivers of patients with other cancers Fig. 3. Forest plot showing the pooled standardized mean difference (SMD) for caregiver quality of life comparing informal caregivers of patients with glioma versus caregivers of patients with other cancers. A random-effects inverse-variance model was used. Negative SMD values indicate poorer quality of life among glioma caregivers relative to comparison groups, whereas positive values indicate better quality of life among glioma caregivers. Caregiver depression, anxiety and distress Across 6 studies, the pooled prevalence of depressive symptoms among caregivers of patients with glioma was 35% (random-effects; 95% CI 14% to 60%) (Fig. 4). There was very high heterogeneity across studies (I2 = 94.02%, χ2 = 83.57, df = 5, P = 0.003), with between-study variance τ2 = 0.34. After excluding studies judged to be at high risk of bias, the pooled prevalence decreased from 35% to 21% (95% CI 4% to 46%) (Supplementary Fig. S1). This reduction suggests that inclusion of high-risk-of-bias studies may have contributed to a higher primary pooled estimate for depressive symptoms. Forest plot of pooled prevalence of depressive symptoms among informal caregivers of patients with glioma Fig. 4. Forest plot showing the pooled prevalence of clinically significant depressive symptoms among informal caregivers of patients with glioma. Pooled estimates were generated using a random-effects meta-analysis of proportions. Squares represent study-specific prevalence estimates, horizontal lines indicate 95% confidence intervals, and the diamond represents the pooled prevalence estimate. In the pooled single-group analysis of anxiety among caregivers of patients with glioma (4 studies), the prevalence of clinically significant anxiety was 36% (random-effects; 95% CI 19% to 55%) (Fig. 5). Heterogeneity was substantial (I2 = 78.16%, χ2 = 13.74, df = 3, P < 0.001), with between-study variance τ2 = 0.11. After excluding studies judged to be at high risk of bias, the pooled prevalence increased slightly from 36% to 40% (95% CI 31% to 50%) (Supplementary Fig. S2), indicating that the anxiety estimate remained broadly comparable and was not materially reduced after sensitivity analysis. Forest plot of pooled prevalence of anxiety among informal caregivers of patients with glioma Fig. 5. Forest plot showing the pooled prevalence of clinically significant anxiety among informal caregivers of patients with glioma. Pooled estimates were generated using a random-effects meta-analysis of proportions. Squares represent study-specific prevalence estimates, horizontal lines indicate 95% confidence intervals, and the diamond represents the pooled prevalence estimate. In the meta-analysis of psychological distress among caregivers of patients with glioma (3 studies), the pooled prevalence of clinically significant distress was 68% (random-effects; 95% CI 53% to 82%) (Fig. 6). Heterogeneity was substantial (I2 = 73.66%, χ2 = 7.59, df = 2, P = 0.02), with between-study variance τ2 = 0.05. After excluding studies judged to be at high risk of bias, the pooled prevalence decreased from 68% to 61% (95% CI 53% to 69%) (Supplementary Fig. S3). Although the estimate was slightly lower, clinically significant distress remained highly prevalent among caregivers of patients with glioma. Forest plot of pooled prevalence of psychological distress among informal caregivers of patients with glioma Fig. 6. Forest plot showing the pooled prevalence of clinically significant psychological distress among informal caregivers of patients with glioma. Pooled estimates were generated using a random-effects meta-analysis of proportions. Squares represent study-specific prevalence estimates, horizontal lines indicate 95% confidence intervals, and the diamond represents the pooled prevalence estimate. Supplementary Fig. S4 shows the symmetrical funnel plot for depression indicating no clear evidence of publication bias and it was confirmed by a non-significant Egger's test (P = 0.699). Discussion This review synthesizes quantitative evidence on psychological morbidity and QoL among caregivers supporting people with glioma. Across included studies, caregiver outcomes were consistently impaired, but magnitude varied substantially by study context, measurement approach, and clinical phase. Overall, pooled estimates indicate that adverse caregiver experiences are common, with a sizeable proportion reporting clinically significant burden and emotional symptoms. The most prominent signal from pooled data was the high prevalence of distress, suggesting that glioma caregiving frequently involves an intense psychological strain profile that extends beyond routine caregiving demands and is likely shaped by neuro-oncological disease characteristics. However, these pooled findings should be interpreted with caution because the number of included studies was limited overall, and several outcome-specific analyses were based on only a small number of studies. With respect to caregiver burden, pooled prevalence indicated that a substantial proportion of caregivers experience clinically significant burden, implying that many caregivers face meaningful constraints on their daily functioning and role capacity. When considering caregiver burden, earlier reports in neuro-oncology often describe moderate-to-severe burden levels and identify patient functional dependence and neuropsychiatric symptoms as key correlates.26,27 The pooled burden prevalence in this review is broadly consistent with that literature, while the heterogeneity mirrors the diversity of caregiving contexts previously described.10,27 In addition, heterogeneity may have been amplified by differences in the instruments used to assess caregiver burden and by variation in how clinically significant burden was operationalized across studies. The comparative finding for QoL showing no clear pooled difference versus caregivers of other cancers both agrees and conflicts with prior evidence, depending on the comparator literature considered. Many studies outside neuro-oncology report substantial QoL impairment among caregivers of advanced cancers, particularly where symptom burden, treatment intensity, and end-of-life caregiving predominate, which could narrow observed differences between tumour types. In that sense, the absence of a pooled difference may reflect the fact that caregiving for other advanced cancers can also be profoundly taxing when care needs are intense.20,28 Conversely, neuro-oncology literature suggests that glioma caregivers may experience a particularly distinctive form of burden because patient cognitive, behavioural, and communication changes can disrupt relationships, shared decision-making, and daily family functioning in ways that are less prominent in many other cancers. The inconsistent direction of study-level effects in this review is compatible with that interpretation and also indicates that “other cancer” is not a uniform comparator category.29 For depression and anxiety, pooled prevalence estimates indicate that clinically significant symptoms are frequent among glioma caregivers, reflecting a substantial mental health burden in this population. Prior caregiver research in cancer populations demonstrates prevalence estimates that vary widely depending on tools and thresholds, but consistently indicate clinically meaningful symptom levels.30,31 This problem is more relevant in the present review, as depression, anxiety, and distress were assessed using different screening instruments across included studies. Clinically significant outcomes were not defined using identical thresholds across the studies. This measurement variability could have contributed to the high heterogeneity observed in the pooled analysis. When compared to previous literature on caregivers of other oncology patients, the prevalence of caregiver burden, anxiety, depression, and distress found in this review appears clinically significant and, in some ways, might be comparable to or higher than estimates reported in mixed cancer caregivers.32, 33 Caregivers of patients with advanced forms of cancer are known to have impaired QoL and higher psychological morbidities. This is because symptom burden, treatment intensity, and end-of-life care needs for the patients are high.32,33 However, glioma caregiving might be challenging as caregivers must manage not only physical dependency and treatment-related demands, but also progressive cognitive impairment, behavioural disturbance, communication difficulty, personality changes, and loss of decision-making capacity of the patients. Comparative QoL meta-analysis in this review did not show significant difference between glioma caregivers and caregivers of other cancers. However, this finding should not be taken as evidence that caregiving experience is equivalent across tumour types. High heterogeneity and inconsistent direction of study level effects suggest that differences may depend on comparator population and clinical contexts. In advanced non-central nervous system cancers, caregiver burden can be severe, particularly in palliative or terminal phases. Taken together, these findings suggest that glioma caregiving is not uniformly “worse” than caregiving in all other cancers, but it is likely to be uniquely complex and, in many settings, especially psychologically demanding. The overall pattern of findings aligns with the broader psycho-oncology literature demonstrating elevated caregiver burden and psychological morbidity among caregivers of cancer patients, while also reflecting features that are particularly salient in neuro-oncology.11,30 Prior studies in cancer caregiving consistently show that informal caregivers experience higher rates of anxiety, depressive symptoms, and reduced QoL compared with population norms, with burden influenced by caregiving intensity, financial toxicity, and perceived lack of support.33 The present synthesis is congruent with this general evidence base, but the notably high pooled distress and substantial burden estimates support the view that glioma caregiving can be uniquely challenging because of the additional neurocognitive, behavioural, and relational consequences of the disease. Thus, although the absolute severity of burden may overlap with that seen in other advanced cancer caregiving settings, the pattern and drivers of distress in glioma appear to be particularly complex. This review provides a focused synthesis of caregiver burden and psychosocial outcomes in glioma caregiving, but several limitations should be considered when interpreting the findings. This review was not prospectively registered in PROSPERO because registration could not be completed during the study period due to service limitations on the PROSPERO platform. Nevertheless, the review was conducted in accordance with the PRISMA 2020 guidelines, and the review protocol was established prior to study screening and data extraction. The overall evidence base was smaller, as only 10 studies met eligibility criteria, and pooled analysis of some outcomes was based on only three studies. Hence, pooled estimates for most outcomes should be interpreted cautiously, as a limited number of contributing studies reduces statistical power, precision, restricts the ability to explore heterogeneity, and limits the robustness and generalizability of pooled estimates. Study quality was mixed, and domain-level assessment showed that some of the included studies had important methodological limitations with respect to sample selection, comparability, and outcome ascertainment. These limitations are relevant because they may have influenced pooled prevalence estimates. Sensitivity analyses excluding high-risk-of-bias studies showed that depressive symptoms decreased from 35% in the primary analysis to 21%, while psychological distress decreased from 68% to 61%. In contrast, anxiety increased slightly from 36% to 40% after exclusion of high-risk studies. These changes indicate that study quality may have contributed to variability in some pooled estimates, particularly depressive symptoms, although the overall conclusion that glioma caregivers experience substantial psychosocial morbidity remained unchanged. Heterogeneity was significant across most of the outcomes. This might be due to differences in study design, caregiver population, tumour grade, disease phase, care setting, and variability in outcome measurement. Studies used different instruments to assess burden, QoL, depression, anxiety, and distress, and thresholds for outcomes were not uniform across studies. Such measurement variability is important in pooled analysis of prevalence, as it can influence both absolute estimates obtained within individual studies and comparability of results across studies. Finally, a small number of studies available for individual outcomes limited our ability to conduct subgroup analysis for factors like glioma grade, active treatment or end of life care, caregiver relationship, and instrument type. Despite these limitations, the findings of this review have clear practical relevance. High pooled prevalence of psychological distress suggests that caregiver screening in glioma should be incorporated as a structured component of neuro-oncology care. As distress had the highest burden amongst the outcomes assessed in this review, a brief validated distress screening approach may be a pragmatic first step in routine practice. This will be useful at clinically vulnerable time points such as initial diagnosis, chemotherapy and radiotherapy initiation, radiological progression, functional decline, and end of life care planning. Implications for nursing practice Findings of this review have important implications for oncology nursing and neuro-oncology supportive care. Higher prevalence of distress, along with higher caregiver burden, anxiety, and depressive symptoms, indicates that caregiver wellbeing should be treated as a clinically important aspect of comprehensive glioma care rather than addressing it informally. Oncology nurses are often the health care professionals with the highest level of contact with patients and families across diagnosis, treatment, follow-up, progression, and end-of-life care. This places them in a key position to identify caregiver burden early and facilitate appropriate support. From a practical perspective, the present findings support incorporating structured caregiver assessment into routine neuro-oncology pathways. Distress may serve as a pragmatic first-line screening target because it was a prominent psychosocial outcome identified in this review and may function as an early signal of caregiver difficulty. The review also suggests that nursing interventions should extend beyond recognition of caregiver burden. Caregivers may benefit from guidance regarding cognitive, behavioural, and functional changes in patients with glioma, practical training in symptom monitoring and daily care tasks, education about available supportive services, and timely referral to psycho-oncology, social work, rehabilitation, or palliative care teams. Hence, oncology nursing practice can play a central coordinating role in linking caregiver assessment to actionable supportive care pathways, thereby improving both caregiver wellbeing and the sustainability of home-based care. Recommendations for future research Future research should aim to strengthen the evidence base through larger, methodologically robust studies using standardized outcome measures and clearly reported thresholds for clinically significant burden and psychological symptoms. Prospective longitudinal studies are also needed to better understand how caregiver outcomes evolve across the glioma trajectory, including diagnosis, active treatment, progression, survivorship, and end-of-life care. Comparative studies involving caregivers of patients with other cancer types would also be valuable to determine more precisely which aspects of glioma caregiving are shared across oncology settings and which are distinctive to neuro-oncology. Conclusions This review suggests that caregivers of patients with glioma experience substantial psychosocial burden, characterized by high levels of distress, followed by caregiver burden, anxiety, and depressive symptoms. Although caregivers of patients with other advanced cancers also experience such burden, glioma caregiving may be unique and complex because of the combination of impact of neurological decline, cognitive and behavioural change, and rapid shifts in patient dependency. The limited number of available studies, variable study quality, and high heterogeneity across outcomes reinforce the need for more standardized research, while the current evidence already supports greater attention to structured caregiver assessment and supportive care integration within neuro-oncology practice. CRediT authorship contribution statement Haimeng Wei: Conceptualization, Methodology, Writing – Original Draft. Shuo Zhang: Data Curation, Formal Analysis, Visualization. Shuai Tao: Data Curation, Formal Analysis, Resources, Software, Supervision, Writing – Review & Editing. Jan Jansen: Supervision. Jing Li: Formal Analysis, Resources, Software, Supervision, Validation, Writing – Review & Editing. Xiaonan Liu: Project Administration, Supervision, Validation, Writing - Review & Editing. All authors have read and approved the final manuscript. Ethics statement Not required. Data availability statement Data availability is not applicable to this article as no new data were created or analyzed in this study. Declaration of generative AI and AI-assisted technologies in the writing process No AI tools/services were used during the preparation of this work. Funding This work is supported by Tiantan Hospital Institutional Fund for Nursing Research (Grant No. 2025-TYHL-15; No.2023-YQN-06) and Tiantan Hospital Institutional Fund for Management Research (Grant No. TYGL202521). The funders had no role in considering the study design or in the collection, analysis, interpretation of data, writing of the report, or decision to submit the article for publication. Declaration of competing interest The authors declare no conflict of interest. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.apjon.2026.100978. Supplementary material The following are the Supplementary data to this article. Multimedia component 1 Sensitivity analysis of pooled prevalence of depressive symptoms after excluding high-risk-of-bias studies Supplementary Fig. 1. Sensitivity analysis forest plot showing the pooled prevalence of depressive symptoms among informal caregivers of patients with glioma after exclusion of studies judged to be at high risk of bias. This analysis was performed to assess the robustness of the primary pooled estimate. Sensitivity analysis of pooled prevalence of anxiety after excluding high-risk-of-bias studies Supplementary Fig. 2. Sensitivity analysis forest plot showing the pooled prevalence of anxiety among informal caregivers of patients with glioma after exclusion of studies judged to be at high risk of bias. This analysis was performed to assess the robustness of the primary pooled estimate. Sensitivity analysis of pooled prevalence of psychological distress after excluding high-risk-of-bias studies Supplementary Fig. 3. Sensitivity analysis forest plot showing the pooled prevalence of psychological distress among informal caregivers of patients with glioma after exclusion of studies judged to be at high risk of bias. This analysis was performed to assess the robustness of the primary pooled estimate. Funnel plot for the meta-analysis of depressive symptoms among informal caregivers of patients with glioma. Each point represents an individual study plotted according to its effect size (ES) and standard error [se(ES)]. The solid vertical line indicates the pooled effect estimate, and the dashed diagonal lines represent the pseudo 95% confidence limits. The plot should be interpreted cautiously because the number of included studies was small, which limits the reliability of visual assessment of small-study effects or publication bias. References Associated Data Supplementary Materials Multimedia component 1 Data Availability Statement Data availability is not applicable to this article as no new data were created or analyzed in this study.