Introduction
One of the distinguishing features of palliative care is its supportive structure, which extends not only to the patient but also to family members and caregivers. The International Association for Hospice and Palliative Care (IAHPC, Reference De Lima and Radbruch2018) emphasizes that the primary goal of palliative care is to enhance the quality of life for individuals and families facing the multidimensional burdens of serious illness and underscores the need for a holistic approach to care. In the literature, a caregiver is defined as an individual who assists the patient in maintaining daily activities, facilitates adherence to treatment regimens, shares emotional burdens, and often undertakes non-medical responsibilities (Hudson et al. Reference Hudson, Trauer and Graham2010, Reference Hudson, Remedios and Zordan2012). The role of caregivers becomes even more crucial in advanced stages of illness, with increased responsibilities for providing not only physical care but also emotional, social, and psychological support (Stajduhar et al. Reference Stajduhar, Giesbrecht, Mollison, Dosani and McNeil2020). Research has shown that caregivers may experience high levels of stress, burnout, anxiety, and depressive symptoms during prolonged care periods, reflecting the psychological dimension of caregiver burden (Oksuz et al.; Ünsar et al.). These findings indicate that caregiving is not merely supportive but also entails significant psychosocial risks.
The challenges faced by caregivers are described in the literature as “caregiver burden,” a multidimensional construct encompassing physical, psychological, social, and economic dimensions. Caregiver burden is not limited to physical fatigue and exhaustion; it also includes multifaceted outcomes such as sleep disturbances, musculoskeletal problems, depression, anxiety, stress, social isolation, role conflicts, and financial difficulties (Etters et al. Reference Etters, Goodall and Harrison2008; Adelman et al. Reference Adelman, Tmanova and Delgado2014). Therefore, psychological well-being should not be considered separate from caregiver burden; rather, anxiety and depressive symptoms represent important psychological components or manifestations of this burden. Numerous studies have demonstrated that higher caregiver burden is closely related to greater anxiety and depressive symptoms, indicating deterioration in the psychological component of caregiver burden (Pinquart and Sörensen Reference Pinquart and Sörensen2003). These risks become more pronounced during long-term care, and caregivers facing socioeconomic disadvantages, insufficient social support, and severe clinical conditions are reported to have a higher tendency toward suicidal ideation (O’Connor and Nock Reference O’Connor and Nock2014). Although anxiety and depressive symptoms are well-recognized psychological components of caregiver burden, suicidal ideation represents a more severe and clinically urgent outcome that may emerge when psychological distress becomes persistent, intense, or insufficiently supported. In palliative care settings, caregivers may be repeatedly exposed to anticipatory grief, end-of-life decision-making, functional decline of the patient, and prolonged emotional strain, all of which may increase vulnerability to suicidal ideation.
Beyond its psychological dimension, caregiver burden also affects multiple aspects of quality of life. High caregiver burden is associated with reduced physical functioning, weakened social relationships, financial strain, limitations in daily activities, and decreased overall life satisfaction (Covinsky et al.; Adelman et al.Reference Adelman, Tmanova and Delgado2014). Studies conducted in Turkiye have shown that as the level of caregiver burden increases, the caregivers’ quality of life meaningfully decreases and their psychological distress intensifies (Oksuz et al.; Ünsar et al.).
Although the relationships between caregiver burden, psychological distress, and quality of life have been widely examined among caregivers of patients receiving palliative care, suicidal ideation has received comparatively less attention as a clinically critical outcome in this population. Similarly, studies conducted in Türkiye have mostly focused on caregiver burden, depressive symptoms, or quality of life, whereas suicidal ideation among caregivers in palliative care settings remains insufficiently investigated. Therefore, this study aimed to investigate suicidal ideation among caregivers of patients receiving palliative care and to examine its relationship with caregiver burden, depressive symptoms, and quality of life. By focusing on suicidal ideation as the primary clinically relevant outcome, this study seeks to provide evidence that may help identify caregivers at increased psychological risk in palliative care settings.
Materials and methods
Study design and population
This study included caregivers aged 16 years and above for patients in the Adult Palliative Care Unit of Erzurum City Hospital. Participation was voluntary for all participants. Throughout the patients’ hospitalization, written and verbal informed consent was obtained from all individuals. The study was designed as descriptive cross-sectional research, and no medical interventions were performed on the participants.
Caregivers were defined as informal/family caregivers who were relatives of the patients and who provided regular physical, emotional, or practical support during the palliative care process. In this study, first-degree family caregivers were defined as spouses, parents, or children of the patient. Other relatives were categorized separately. Professional or paid caregivers were excluded from the study. Caregivers aged 16 years or older were eligible for inclusion because individuals in this age group may actively undertake caregiving responsibilities in the local sociocultural context. In Türkiye, adolescents aged 16 and above are frequently involved in family caregiving responsibilities, particularly within extended family structures where caregiving duties are shared among household members regardless of age. This threshold was therefore selected to reflect real-world caregiving practices in the study setting rather than an arbitrary cutoff. For caregivers under the age of 18, informed consent was obtained in accordance with institutional ethical requirements, including additional consent procedures involving a legal guardian. In the case of the presence of more than 1 caregiver, only 1 caregiver per patient was included in the study; the primary caregiver, defined as the person who assumed the main responsibility for the patient’s daily care and decision-making support, was invited to participate. Inclusion criteria were accompanying the patient as a caregiver for patients hospitalized in the palliative care unit, agreeing to participate voluntarily in the study, and having sufficient cognitive capacity to complete the questionnaires. Individuals were excluded if they had language barriers preventing comprehension of the Turkish questionnaires, severe hearing or speech impairments that hindered communication, marked cognitive impairment, or a history of severe psychiatric disorders. Caregivers who declined participation, did not complete the questionnaires, or withdrew from the study at any point were not included in the final analysis. Eligibility was determined during routine clinical contact by the responsible physician and palliative care team. No additional structured study assessment was performed for non-participants. Caregivers deemed unsuitable for participation due to heavy psychosocial burden in the terminal phase, as assessed by the physician/team, and were not invited to participate in the research.
Demographic information collected from participants included age, educational level, and degree of relationship to the patient. Educational level was categorized into 4 groups: less than primary school education, primary school, high school, and university or higher. Relationship to the patient was classified as first-degree relatives (spouse, parents, and children) and other relatives.
Clinical features of study population
The diagnoses of the palliative patients included in the study were classified into 2 main groups. The neurological disease group includes dementia (Alzheimer’s disease and other dementias), cerebrovascular events, Parkinson’s disease, multiple sclerosis, amyotrophic lateral sclerosis, paraplegia, meningioma, glioblastoma multiforme, and other brain disorders. Although glioblastoma multiforme is a malignant central nervous system tumor, it was categorized under the neurological disease group because the primary palliative care needs of these patients were predominantly related to neurological impairment, cognitive decline, seizures, and functional dependency. The systemic/other diseases group includes heart failure, chronic kidney failure, chronic obstructive pulmonary disease, pneumonia, various malignant neoplasms (colon, stomach, pancreas, esophagus, rectum, and bladder), myelodysplastic syndrome, urinary system infections, decubitus ulcers, humerus and femur fractures, catheter infections, and atherosclerotic cardiovascular diseases.
Tests and scales
Assessments were conducted in the Gerontology Unit of the Adult Palliative Care Service. In this study, several standardized and validated instruments were used to evaluate caregivers’ burden, depressive symptoms, quality of life, and suicidal ideation. Caregiver burden was assessed using the Caregiver Strain Index (CSI), originally developed by Robinson (Reference Robinson1983). The CSI consists of 13 dichotomous items, and a total score of 7 or higher indicates a high level of caregiver strain. The Turkish validity and reliability of the CSI have been previously established, demonstrating acceptable psychometric properties for use in Turkish populations (Ugur and Fadiloğlu Reference Ugur and Fadiloğlu2010).
Depressive symptom severity was evaluated using the Hamilton Depression Rating Scale (HAM-D), developed by Hamilton et al. (1960). The scale categorizes depression severity as follows: scores of 0–7 indicate no depression, 8–16 mild depression, 17–23 moderate depression, and ≥24 severe depression. The Turkish version of the HAM-D has been shown to be valid and reliable in clinical and non-clinical samples (Aydemir et al. Reference Aydemir, Köroğlu and Akdemir1997). Quality of life was assessed using 2 complementary instruments. The OPQOL-Brief (Older People’s Quality of Life Questionnaire – Brief), originally developed by Bowling et al. (Reference Bowling, Hankins and Windle2013), was used to evaluate quality of life specifically in older populations. Higher scores indicate better perceived quality of life. The Turkish adaptation of the OPQOL-Brief has demonstrated satisfactory validity and reliability (Caliskan et al. Reference Caliskan, Sengul Aycicek and Ozsurekci2019). In addition, general health status was assessed using the SF-36 Health Survey (Short Form-36 Health Survey), developed by Ware and Sherbourne (Reference Ware and Sherbourne1992), which evaluates health-related quality of life across multiple physical and mental domains. The Turkish validity and reliability study of the SF-36 was conducted by Kocyigit et al. (Reference Kocyigit, Aydemir and Olmez1999), confirming its suitability for use in Turkish populations.
Suicidal ideation among caregivers was assessed using the Beck Scale for Suicide Ideation (BSS), originally developed by Beck et al. (Reference Beck, Kovacs and Weissman1979). The scale consists of items scored between 0 and 38, with higher scores indicating more severe suicidal ideation. The Turkish version of the BSS has been validated and shown to have acceptable reliability (Ozcelik et al. Reference Ozcelik, Ozdel and Bulut2015).
For the evaluation of clinical history and medical information related to the palliative care process, patient files and the hospital’s electronic health record system were meticulously reviewed.
Statistical analysis
Data analysis was performed using Statistical Package for the Social Sciences (IBM SPSS) Statistics version 26.0. Categorical variables were summarized as frequency (n) and percentage (%), while the normality of numerical variables was assessed based on skewness and kurtosis values, and analytic tests via the Kolmogorov–Smirnov test. Accordingly, parametric tests were used in the analyses, including independent-samples t-test, 1-way ANOVA, and Pearson’s correlation analysis. When interpreting the correlation coefficient, values between 0.00 and 0.30 were considered low, 0.30–0.70 moderate, and 0.70–1.00 high correlation (Büyüköztürk Reference Büyüköztürk2020). For the comparison of variables converted to ordinal data, such as the CSI and the HAM-D, according to demographic characteristics, the Chi-square test was used. In cases where the assumptions of the Chi-square test were not met, Fisher’s Exact Test was applied (Büyüköztürk Reference Büyüköztürk2020). The effects of physical and mental health (MH) (SF-36), CSI, and the OPQOL-Brief for older adults on participants’ levels of depression were examined using multiple linear regression analysis. In the application of multiple linear regression in this model, the linearity of the relationship between independent and dependent variables was assessed using scatter plots, while the normality of the scores was evaluated based on skewness and kurtosis values, as well as by examining histograms and Normal P–P plot graphs. The review indicated that the model met both the linearity and normality assumptions. In this study, statistical significance was set at p < 0.05 and p < 0.01.
Results
A total of 75 caregivers participated in the study. Regarding the gender distribution of patients, 58.7% were male (n = 44), and 41.3% were female (n = 31). Analysis of age distribution showed that 65.3% of patients were aged 75 years and above (n = 49), while 34.7% were aged 74 years and below (n = 26). The mean age of patients was calculated as 76.28 ± 12.72 years. In terms of disease diagnosis, 66.7% of participants had neurological diseases (n = 50), and 33.3% had systemic or other diseases (n = 25). Among caregivers, 61.3% were female (n = 46) and 38.7% were male (n = 29). The mean age of caregivers was 49.32 ± 14.43 years, with age groups distributed equally: 16–46 years (33.3%, n = 25), 47–55 years (33.3%, n = 25), and 56 years and above (33.3%, n = 25). Regarding relationship to the patient, 66.7% of caregivers were first-degree relatives (spouse, parent, or child) (n = 50), while 33.3% were other relatives (n = 25) (Table 1).
Demographic characteristics of participants
Table 1 Long description
Demographic characteristics are reported by variable and category, with columns for N and percent. Patient age includes 26 participants age 74 years or younger at 34.7 percent and 49 participants age 75 years or older at 65.3 percent. Disease diagnosis includes neurological diseases with 50 participants at 66.6 percent and systemic and other diseases with 25 participants at 33.3 percent. Relative age is evenly split across three groups: 16 to 46 years, 47 to 55 years, and 56 years or older, each with 25 relatives at 33.3 percent. Education level includes less than primary school with 9 at 12 percent, primary school with 33 at 44 percent, high school with 20 at 26.7 percent, and university or higher with 13 at 17.3 percent. Relationship to patient includes first degree relative with 50 at 66.7 percent and other with 25 at 33.3 percent.
According to the CSI, 52% (n = 39) of caregivers had high caregiver strain. Examination of the HAM-D revealed that 76% of participants showed no signs of depression (n = 57), whereas 24% had mild depression (n = 18). Based on the BSS, all participants (100%, n = 75) were classified in the low-risk category. The mean score for the physical health (PH) component of the SF-36 Quality of Life Scale was 67.8 ± 21.4 points, with the highest subdomain being physical functioning (75.3 ± 21.7 points) and the lowest being pain (62.6 ± 31.9 points). Moderate scores were obtained in the physical role difficulty and general health subdomains. The mean score for the MH component was 66.4 ± 16.0 points; within this component, social functioning had the highest score (86.5 ± 19.2 points), while energy had the lowest (54.1 ± 18.6 points). Moderate scores were reported for the MH and emotional role difficulty subdomains. The mean score for the OPQOL was 56.3 ± 6.0 points.
Comparisons of SF-36 subdomains and the Quality of Life Scale according to demographic characteristics were presented in Table 2. In analyses comparing SF-36 and OPQOL scores by demographic variables, no statistically significant differences were found in PH (p = 0.82), MH (p = 0.33), or OPQOL (p = 0.99) scores according to patient gender. Similarly, no significant differences were observed in PH (p = 0.69), MH (p = 0.11), or OPQOL (p = 0.40) scores across age groups. When caregiver outcomes were compared according to education level, only the MH subdomain differed significantly across education groups (p = 0.024). The highest MH score was observed among caregivers with high school education, whereas the lowest score was observed among those with university or higher education. Post-hoc analyses revealed that caregivers with a high school education had higher MH scores than those with no primary education or university-level education. In contrast, PH (p = 0.471) and OPQOL scores (p = 0.812) did not differ significantly across education levels. Comparisons by disease diagnosis revealed no significant differences in PH (p = 0.29), MH (p = 0.37), or OPQOL (p = 0.83) scores. No significant differences were found in PH (p = 0.50), MH (p = 0.75), or OPQOL (p = 0.88) scores according to caregiver gender.
Comparison of SF-36 subscales and Quality of Life Scale by demographic characteristics
Table 2 Long description
Mean and standard deviation are reported for Physical Health, Mental Health, and OPQOL across patient and caregiver demographic categories, with a p row for each variable. For patient gender, males had Physical Health 67.3 plus or minus 21.4, Mental Health 64.9 plus or minus 17.0, and OPQOL 56.3 plus or minus 6.4; females had 68.5 plus or minus 21.7, 68.5 plus or minus 14.6, and 56.4 plus or minus 5.4, with p values 0.82, 0.33, and 0.99. For patient age, those 74 years or younger had 69.1 plus or minus 23.1, 62.3 plus or minus 15.9, and 55.5 plus or minus 7.0; those 75 years or older had 67.1 plus or minus 20.6, 68.5 plus or minus 15.9, and 56.8 plus or minus 5.4, with p values 0.69, 0.11, and 0.40. Education level includes Less than primary school A, Primary School B, Highschool C, and University and higher D; Mental Health means are 61.2 plus or minus 16.2, 67.1 plus or minus 16.6, 73.5 plus or minus 11.9, and 57.3 plus or minus 15.9, with p 0.024 marked with an asterisk, while Physical Health p is 0.471 and OPQOL p is 0.812. Disease diagnosis compares Neurological Diseases with Systemic and other diseases, with Physical Health 69.6 plus or minus 21.3 versus 64.1 plus or minus 21.5, Mental Health 67.6 plus or minus 18.1 versus 64.0 plus or minus 10.7, and OPQOL 56.2 plus or minus 6.3 versus 56.6 plus or minus 5.3; p values are 0.292, 0.365, and 0.829. Caregiver gender shows males at 65.6 plus or minus 25.1, 65.6 plus or minus 17.3, and 56.5 plus or minus 6.7, and females at 69.1 plus or minus 18.9, 66.8 plus or minus 15.4, and 56.3 plus or minus 5.5, with p values 0.497, 0.747, and 0.877. Caregiver age groups are 16 to 46 years, 47 to 55 years, and 56 years and above, with Physical Health 77.3 plus or minus 17.7, 64.1 plus or minus 22.3, and 62.0 plus or minus 21.4, and p 0.024 marked with an asterisk; Mental Health is 67.4 plus or minus 16.2, 68.0 plus or minus 13.5, and 63.8 plus or minus 18.3 with p 0.882, and OPQOL is 58.0 plus or minus 4.2, 56.8 plus or minus 6.0, and 54.3 plus or minus 7.1 with p 0.415. Relationship to patient compares First Degree Relative with Other, with Physical Health 68.4 plus or minus 22.0 versus 66.5 plus or minus 20.5, Mental Health 67.0 plus or minus 15.9 versus 65.1 plus or minus 16.6, and OPQOL 56.0 plus or minus 6.3 versus 57.0 plus or minus 5.3; p values are 0.715, 0.623, and 0.482.
* Post-hoc analyses revealed that caregivers with a high school education had higher mental health scores than those who had no primary education or a university education.
However, a significant difference in PH scores was observed according to caregiver age (p = 0.024); PH scores were higher among caregivers in the 16–46 age group compared to other age groups. MH (p = 0.88) and OPQOL (p = 0.42) scores did not vary by caregiver age. Analyses by relationship to patient showed no significant differences in PH (p = 0.72), MH (p = 0.62), or OPQOL (p = 0.48) scores.
In analyses comparing HAM-D scores by demographic characteristics (Table 3), no statistically significant differences in depression levels were found according to patient gender (χ 2 = 0.625, p = 0.429), age groups (χ 2 = 0.186, p = 0.666), disease diagnosis (χ 2 = 0.00, p = 1.000), caregiver gender (χ 2 = 1.283, p = 0.257), caregiver age (χ 2 = 1.316, p = 0.518), or relationship to patient (χ 2 = 0.00, p = 1.000). Although different distributions were observed between the rates of no depression and mild depression across education levels, this difference was not statistically significant (χ 2 = 5.134, p = 0.147). It was noteworthy that the rates of mild depression were relatively higher among caregivers without primary education and university-educated caregivers, although this did not reach statistical significance.
Comparison of Hamilton Depression Rating Scale categories by demographic characteristics
Table 3 Long description
Counts and percentages for two Hamilton Depression Rating Scale categories, No Depression and Mild Depression, are listed by patient and caregiver characteristics, with a chi square statistic and p value for each variable. Patient gender: male 32 no depression 72.7 percent and 12 mild 27.3 percent; female 25 no depression 80.6 percent and 6 mild 19.4 percent; chi square 0.625, p 0.429. Patient age: 74 years or less 19 no depression 73.1 percent and 7 mild 26.9 percent; 75 years or more 38 no depression 77.6 percent and 11 mild 22.4 percent; chi square 0.186, p 0.666. Education level: less than primary school 5 no depression 55.6 percent and 4 mild 44.4 percent; primary school 28 no depression 84.8 percent and 5 mild 15.2 percent; high school 16 no depression 80.0 percent and 4 mild 20.0 percent; university and above 8 no depression 61.5 percent and 5 mild 38.5 percent; chi square 5.134, p 0.147. Disease diagnosis: neurological diseases 38 no depression 76.0 percent and 12 mild 24.0 percent; systemic and other diseases 19 no depression 76.0 percent and 6 mild 24.0 percent; chi square 0.00, p 1.000. Caregiver gender: male 20 no depression 69.0 percent and 9 mild 31.0 percent; female 37 no depression 80.4 percent and 9 mild 19.6 percent; chi square 1.283, p 0.257. Caregiver age: 16 to 46 years 18 no depression 72.0 percent and 7 mild 28.0 percent; 47 to 55 years 18 no depression 72.0 percent and 7 mild 28.0 percent; 56 years or more 21 no depression 84.0 percent and 4 mild 16.0 percent; chi square 1.316, p 0.518. Relationship to patient: first degree relative 38 no depression 76.0 percent and 12 mild 24.0 percent; other 19 no depression 76.0 percent and 6 mild 24.0 percent; chi square 0.00, p 1.000.
χ 2: Chi-square statistic; p: p-value (Fisher’s Exact Test applied where Chi-square assumptions were not met).
According to the results of Pearson’s correlation analysis, there was a moderate, positive, and significant relationship between PH and MH (r = 0.556, p < 0.01). PH was negatively correlated with caregiver stress level (r = –0.235, p = 0.042), suicidal ideation (r = –0.232, p = 0.046), and depression (r = –0.355, p = 0.002), and positively correlated with OPQOL (r = 0.384, p = 0.001). MH showed significant negative correlations with caregiver stress level (r = –0.447, p < 0.01), suicidal ideation (r = –0.295, p < 0.01), and depression (r = –0.607, p < 0.01), and a significant positive correlation with OPQOL (r = 0.515, p < 0.01). Caregiver stress level was positively and significantly correlated only with depression (r = 0.324, p = 0.005). Suicidal ideation was positively correlated with depression (r = 0.454, p < 0.01) and negatively correlated with OPQOL (r = –0.278, p = 0.016). There was a moderate, negative, and significant relationship between depression and OPQOL (r = –0.304, p = 0.008) (Table 4).
Comparison of relationships between scales
Table 4 Long description
Pearson correlation coefficients are listed between the columns SF 36 Phy, SF 36 Mental, CSI, BSS, and HAM-D and the rows Mental Health, CSI, BSS, HAM-D, and OPQOL. In the Mental Health row, SF 36 Phy is 0.556 with double asterisks. In the CSI row, SF 36 Phy is minus 0.235 with one asterisk and SF 36 Mental is minus 0.447 with double asterisks. In the BSS row, SF 36 Phy is minus 0.232 with one asterisk, SF 36 Mental is minus 0.295 with double asterisks, and CSI is 0.188. In the HAM-D row, SF 36 Phy is 0.355 with double asterisks, SF 36 Mental is minus 0.607 with double asterisks, CSI is 0.324 with double asterisks, and BSS is 0.454 with double asterisks. In the OPQOL row, SF 36 Phy is 0.384 with double asterisks, SF 36 Mental is 0.515 with double asterisks, CSI is minus 0.129, BSS is minus 0.278 with one asterisk, and HAM-D is minus 0.304 with double asterisks. Asterisks appear next to some coefficients to mark statistical significance.
r = Pearson’s correlation coefficient.
* p < 0.05.
** p < 0.01.
The multiple linear regression model (Table 5) examining the variables affecting HAM-D scores was found to be significant (F(8,66) = 5.609; p < 0.001). The model explained 40.5% of the total variance (R 2 = 0.405). Examination of the regression coefficients for the independent variables revealed that only the MH subscale was a significant predictor of HAM-D scores (β = −0.511; p < 0.001). Other variables (PH, CSI, OPQOL, patient age, caregiver age, caregiver gender, and relationship to patient) did not have a significant effect on depression levels (p > 0.05). According to the model coefficients, each 1-unit increase in MH score was associated with a 0.113-point decrease in HAM-D score.
Multiple regression analysis of factors affecting participants’ depression levels
Table 5 Long description
Multiple regression results list each predictor with columns for B, Std. Error B, beta, t, p, and a 95 CI column that is blank for all rows. Physical Health has B minus 0.018, Std. Error B 0.021, beta minus 0.112, t minus 0.890, and p 0.377. Mental Health has B minus 0.113, Std. Error B 0.032, beta minus 0.511, t minus 3.558, and p 0.001 with two asterisks. Caregiver Strain Index has B 0.132, Std. Error B 0.174, beta 0.085, t 0.759, and p 0.450. Older Peoples Quality of Life Questionnaire, OPQOL, has B minus 0.006, Std. Error B 0.068, beta minus 0.010, t minus 0.090, and p 0.929. Patient Age has B minus 0.021, Std. Error B 0.028, beta minus 0.075, t minus 0.746, and p 0.458; Caregiver Age has B minus 0.035, Std. Error B 0.026, beta minus 0.144, t minus 1.381, and p 0.172. Relationship to patient of the Caregiver has B 0.182, Std. Error B 0.733, beta 0.025, t 0.249, and p 0.804; Caregiver Gender has B minus 0.637, Std. Error B 0.711, beta minus 0.089, t minus 0.895, and p 0.374. Model summary reports R 0.636, R2 0.405, and an F statistic with p reported as 0.000 and also noted as less than 0.01.
B = unstandardized regression coefficient; Std. error = standard error of B; β (Beta) = standardized regression coefficient; t = t-statistic testing the null hypothesis that B = 0; p = p-value; 95% CI = 95% confidence interval for B.
Discussion
This study comprehensively examined the relationships among quality of life, PH, MH, depression, suicidal ideation, and caregiver stress levels in caregivers of older adults receiving inpatient palliative care followed in an adult palliative care unit. The present study primarily highlights suicidal ideation as a clinically important outcome among caregivers of older adults receiving inpatient palliative care. While caregiver burden, depressive symptoms, and quality of life have been widely examined in palliative care populations, suicidal ideation remains a less frequently addressed but critical manifestation of severe psychological vulnerability. In this study, suicidal ideation was evaluated in relation to caregiver burden, depressive symptoms, and quality of life, providing a broader perspective on psychological risk among family caregivers in inpatient palliative care settings. The findings suggest that high caregiver stress may adversely affect caregivers’ psychological well-being, caregiving capacity, communication with the healthcare team, and participation in decision-making during the inpatient palliative care process. In addition, physical and MH scores were negatively associated with depressive symptoms, suicidal ideation, and caregiver stress, indicating that poorer quality-of-life domains are closely linked to greater psychological vulnerability. These findings are consistent with previous literature emphasizing that psychological distress among caregivers is closely intertwined with physical functioning, quality of life, and the overall caregiving experience.
In the present study, suicidal ideation was considered a clinically important indicator of severe psychological vulnerability among caregivers. It should be noted that although all participants fell within the low-risk category on the BSS according to standard clinical cut-offs, the continuous BSS scores nonetheless demonstrated significant correlations with depression and quality of life, indicating that subclinical variation in suicidal ideation may still hold clinical relevance in this population. This is particularly relevant in inpatient palliative care, where caregivers may face prolonged emotional distress, anticipatory grief, uncertainty regarding prognosis, and involvement in difficult care-related decisions. Previous evidence suggests that family caregivers may represent a vulnerable group for suicidal ideation, particularly when caregiving is accompanied by high psychological distress, depressive symptoms, and reduced perceived support (Park et al. Reference Park, Kim and Shin2013; O’Dwyer et al. Reference O’Dwyer, Janssens and Sansom2021). In end-of-life care contexts, caregivers’ perceptions of poor patient quality of life have also been associated with increased risk of suicidal ideation during bereavement, further supporting the need to address suicide-related risk in palliative caregiving (Abbott et al. Reference Abbott, Prigerson and Maciejewski2014). Therefore, suicidal ideation may reflect an advanced stage of accumulated psychological strain and may identify caregivers who require more immediate psychosocial support. Incorporating suicide-related screening questions into caregiver assessments, particularly for those with high caregiver burden, depressive symptoms, or impaired quality of life, may help clinicians identify individuals at increased risk and facilitate timely referral to MH services.
López et al. (Reference López, Perez-Rojo and Noriega2024) demonstrated that quality of life is particularly closely linked to MH, mood, and social functioning. Similarly, Georgi et al. (Reference Georgi, Dragomirecká and Tichá2025) reported that depression is one of the most important psychological factors determining individuals’ quality of life. The finding in this study that MH is the strongest predictor of depression is fully aligned with these results.
The positive association between caregiver stress and depression underscores that caregiver burden in the palliative care setting appears to be associated with the patient’s psychological state, a finding that may reflect the interconnected nature of the patient–caregiver relationship. However, given the cross-sectional design of this study, causal or directional interpretations cannot be made. Shefeek et al. (Reference Shefeek, Joy, Olickal, Nezrin, George and Thankappan2024) reported that caregiver burden in palliative care can negatively influence the emotional state of both the caregiver and the patient. Pop et al. (Reference Pop, Cojan Minzat, Ursu, Ursu and Puia2024) further suggested that caregiver stress increases over time and that this burden is indirectly reflected in the patient’s care processes. Similarly, this study found a significant association between caregiver stress and patient depression levels.
The negative association of OPQOL scores with depression and suicidal ideation, and their positive association with physical and MH, are consistent with findings from recent quality of life studies. Zhang et al. (Reference Zhang, Zhang, Xiang, Li and Xu2025) reported that depression is a key determinant of reduced quality of life in older adults, particularly noting that social isolation and low mental well-being strengthen this relationship. Argyriou et al. (Reference Argyriou, Dimitriadou, Saridi, Toska, Lavdaniti and Fradelos2024) also demonstrated the direct impact of depression on quality of life. The findings of this study are in line with these results, emphasizing the close relationship between OPQOL and psychological well-being.
The findings of this study should also be interpreted within the sociocultural context of Türkiye, where family-based caregiving remains the predominant model of support for patients with serious illness. Cultural and religious norms in Turkish society place strong emphasis on filial responsibility, and caregiving duties are often assumed by close family members – particularly women – rather than being delegated to institutional or professional care services (Ünsar et al.). This may partly explain the high proportion of female caregivers observed in the present sample. Additionally, suicidal ideation remains a highly stigmatized topic in Turkish society, and concerns about social judgment or family reputation may lead caregivers to underreport such thoughts even in confidential research settings. This cultural tendency toward underreporting may partly account for the finding that all participants were classified as low risk on the BSS, despite significant correlations between suicidal ideation scores, depression, and quality of life. Furthermore, palliative care is a relatively recently established service model within the Turkish healthcare system, and caregivers may have limited prior exposure to structured psychosocial support mechanisms, which may heighten their vulnerability during the caregiving process.
Overall, the analyses demonstrate that caregivers’ depression levels are particularly strongly and inversely related to their MH status, while PH, caregiver stress, quality of life, and demographic characteristics do not independently explain depression. MH emerged as the strongest psychosocial determinant at the correlation level and as the sole significant predictor of depression in the multiple regression model.
Finally, the findings of this study carry several implications for clinical practice in palliative care settings. First, routine psychosocial screening, incorporating validated instruments for caregiver strain, depressive symptoms, and suicidal ideation, should be integrated into standard care protocols at the time of patient admission to palliative care units, with periodic reassessment during prolonged hospitalizations. Second, given that MH emerged as the strongest predictor of depressive symptoms, screening protocols should prioritize brief, validated MH assessments as a first-line triage tool, given their strong association with depressive symptoms in this population. Third, caregivers identified with high caregiver strain, low MH scores, or elevated suicidal ideation should be referred to psychosocial support services, including counseling or psychiatric consultation, in a structured and timely manner rather than on an ad hoc basis. Fourth, given the cultural stigma surrounding suicidal ideation discussed above, screening should be conducted in a private, non-judgmental manner, potentially through trained social workers or psychologists rather than general clinical staff, to encourage honest disclosure. Finally, institutional policies should formally designate a responsible team member, such as a palliative care social worker or liaison psychiatrist, to coordinate psychosocial follow-up for caregivers, ensuring that identified risk does not go unaddressed after initial screening. In addition, structured caregiver support programs incorporating psychoeducation, psychosocial support, and stress-management strategies may provide a more systematic approach to addressing caregiver needs, as supported by recent evidence from intervention studies. Caregiver support should also extend beyond active caregiving when needed, including appropriate follow-up and bereavement support after the end of the caregiving period.
Limitations and strengths
This study has several limitations that should be considered when interpreting the findings. The cross-sectional design precludes causal inference regarding the relationships among caregiver burden, depressive symptoms, suicidal ideation, and quality of life. In addition, the study was conducted in a single palliative care center with a relatively small sample size, which may limit the generalizability of the findings to other care settings and caregiver populations. The use of self-report instruments may also have introduced reporting bias and social desirability bias, particularly for sensitive outcomes such as depressive symptoms and suicidal ideation. Moreover, some caregiving-related variables, including total duration of caregiving, daily caregiving hours, and length of hospitalization, were not systematically recorded. The absence of these variables may have limited our ability to fully account for contextual factors that could influence caregiver burden, depressive symptoms, quality of life, and suicidal ideation. Additionally, no a priori power analysis was performed, which may have reduced the ability to detect small but clinically meaningful associations. Finally, the lack of longitudinal follow-up prevented evaluation of changes in caregivers’ psychological outcomes over time. Despite its limitations, this study provides valuable insights into the psychological well-being of caregivers in a palliative care setting by simultaneously evaluating caregiver burden, depression, suicidal ideation, and health-related quality of life using validated instruments. The inclusion of both physical and MH dimensions and the focus on caregivers in a real-world clinical setting strengthen the clinical relevance of the findings.
Conclusion
Overall, the study demonstrates that MH, depression, and caregiver stress are critical factors in the context of palliative care and exert strong effects on quality of life. Routine screening for depressive symptoms and psychological distress in palliative care units is recommended, and given the strong association between OPQOL scores and psychological variables, individualized psychosocial interventions may be implemented for caregivers. To further analyze patient–caregiver interactions and causal relationships, larger, multicenter longitudinal studies are recommended. In this context, suicidal ideation should be considered an important warning sign that may indicate increased psychological risk among caregivers and may require further psychosocial assessment and referral when appropriate.
Author contributions
Conceptualization: E.U.K, M.G.; data collection: K.G.S., E.S.; statistical analysis: K.G.S, M.G.; writing – original draft: K.G.S; and writing – review and editing: all authors.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Competing interests
The authors declare no conflict of interest.
Ethical approval
The study was approved by the Ethics Committee of Erzurum Faculty of Medicine, University of Health Sciences, on December 11, 2024, with approval number 2024/12-229. The research was conducted in accordance with the ethical principles of the Declaration of Helsinki.
Ethics and consents
Written informed consent was obtained from all participants prior to inclusion in the study.
Open access
