Quality of life and related factors among cancer caregivers in China


Lie Wang, PhD, Department of Social Medicine, School of Public Health, China Medical University, No. 92 Beier Road, Heping District, Shenyang, 110001, China. Email: liewang@mail.cmu.edu.cn


Aims:  The present study was designed to investigate quality of life (QOL) among caregivers of cancer inpatients, and to explore the related factors with the long-term goal to improve QOL of cancer caregivers in China.

Methods:  A cross-sectional study was performed during the period January–March 2008. Subjects comprised 358 caregivers of cancer inpatients. The questionnaires pertaining to QOL, sociodemographic characteristics, caregiving demands, available social resources, caregiving situation, and caregiver's health status were used. Data were collected by nurses through interviews.

Results:  The average QOL score of these subjects was 5.26. General linear model analysis showed that caregiver's QOL was significantly associated with, in standardized estimate sequence, spousal relationship, patient's activities of daily living, and chronic disease (no). Overall spousal relationship to patient had the strongest association with caregiver's QOL.

Conclusions:  Psychological well-being and social concerns of cancer caregivers were disrupted more seriously than spiritual well-being and physical well-being. In addition, spousal relationship, patient's disability, and caregiver's health status could affect the caregiver's QOL.

CANCER REMAINS ONE of the leading causes of morbidity and mortality worldwide. According to the World Cancer Report estimated by the World Health Organization, cancer rates have increased at an alarming rate, from 10 million new cases globally in 2000 to 15 million in 2002,1 and it has been estimated that, in 2020, the world's cancer population will increase to 7.5 billion, which will include 15 million new cancer cases and 12 million cancer deaths.2 However, cancer has an effect not only on patients, but also on family caregivers, because cancer patients need long-term care due to the long duration of the disease and the treatment, such as surgery, chemotherapy and radiotherapy. In a study of family caregivers of cancer patients, 66.8% of caregivers quit work or made major life changes in order to provide care, 68.0% of families lost most of their savings, and 55.0% of families lost their main source of income.3 Therefore, the health and well-being of cancer caregivers have attracted accelerating interest in recent years, and have been studied well in Western countries.4–7

Quality of life (QOL) is a term used widely to indicate an individual's general well-being, and has been applied in the field of caregiving for years. Studies conducted in Western countries have reported that factors associated with caregiver's QOL during the cancer caregiving period could be divided into four categories: (i) sociodemographic characteristics; (ii) caregiving demands; (iii) available psychological and social resources; and (iv) caregiving situation.8–10

In China, the Center for Health Information and Statistics, Ministry of Health reported a total of 2.1 million cancer cases in 2000 and this number has kept increasing. In 2007, according to the report from the Ministry of Health of the People's Republic of China, cancer has become the fifth leading cause of hospitalization.11,12 As older people are at high risk for cancer, the absolute number of cancer patients is expected to be prodigious in the near future because of the large population base and aging problem. In Chinese culture, due to the Confucian thought of filial piety and the Chinese law, it is regarded as an obligation for family members, both spouses and children, to provide care to patients. Even during the period of hospitalization, family caregivers, not nurses, have the responsibility to take care of inpatients directly. This fact, together with the increasing trend of cancer and the only-child policy, increase the likelihood that QOL of Chinese cancer caregivers will be affected as much as those in Western countries or even more seriously. However, up to now, there are few studies that evaluate the QOL of cancer caregivers in mainland China.

The present study was designed to assess the QOL of cancer caregivers and to explore the related factors. As it is the first attempt in mainland China, we initially focused on cancer inpatients because inpatients usually place a greater burden on family caregivers than outpatients and it is more feasible to collect cancer cases in hospital due to the concentration of patients.13 Thus, QOL, sociodemographic characteristics, caregiving demands, available social resources, caregiving situation, and caregiver's health status were considered to assess QOL of caregivers of cancer inpatients and to explore the related factors with the long-term goal to improve QOL of cancer caregivers.



A cross-sectional study was conducted to fulfill the aim outlined above. Due to the infeasibility of obtaining a complete list of cancer patients from China, patients and their caregivers were recruited from the oncology inpatient units of the Affiliated Hospitals of China Medical University by a convenience sampling strategy. We made contact with the caregivers by referral from the doctors in charge of patients. The caregiver was defined as the family member who: (i) was identified by the patient as the person most often involved in his or her care, regardless of any family tie; (ii) was over 18 years of age; and (iii) had no diagnosed cancers or cognitive impairments. Inpatients who were: (i) undergoing treatment in the Affiliated Hospitals of China Medical University; (ii) at least 16 years of age; and (iii) diagnosed with a malignancy, such as lung cancer, stomach cancer, and colorectal cancer, were approached in order to access cancer caregivers. Exclusion criteria for inpatients and their caregivers were persons who could not communicate with interviewers in Mandarin. During the period January–March 2008, of nearly 1300 inpatients identified by the doctors-in-charge as appropriate, a total of 490 caregivers were recruited, and these people became our study population. Among this population, we received complete responses from 358 primary caregivers who were willing to participate (effective response rate 73.1%), and these caregivers became our subjects. These subjects were the caregivers of patients with lung cancer (24.6%), stomach cancer (18.2%), colorectal cancer (15.6%), breast cancer (11.5%), leukemia (9.1%), liver cancer (7.4%), esophageal cancer (4.3%), and others (9.3%).


Dependent variables

Caregiver's QOL was measured using the QOL-Family Version that is offered by the American National Medical Center and the Beckman Research Institute with Cronbach's α = 0.69 and test–retest reliability = 0.89. The original questionnaire was translated into Chinese and adjusted by two experts of social medicine with consideration of Chinese-specific status. After permission was obtained from the original developer, two items relating to religious activities were excluded because most Chinese people do not have religious beliefs. Back translation was done by two bilingual translators and the back-translated version was found to be comparable with the original version. The Chinese version of QOL comprised 35 items, grading from 0 (worst) to 10 (best) and included the assessment of physical well-being, psychological well-being, social concerns and spiritual well-being of the caregivers. Among all the items, several have reverse anchors and need to be reversed by scores, and they are 1–4, 6, 13–20, 22, 24–29, and 31.14 The total score can be obtained by adding all of the items and then creating a mean score. The internal consistency of the questionnaire was Cronbach's α = 0.71 in this study, and a confirmatory factor analysis confirmed the construct validity of the questionnaire (comparative fit index = 0.887, incremental fit index = 0.889, Tucker–Lewis index = 0.874, goodness-of-fit index = 0.825).

Independent variables

  • 1Sociodemographic characteristics included gender, age, occupation (employed/unemployed), education (below high school/high school and above), monthly household income per person (below 1500 yuan/person or 1500 yuan and above/person), and relationship to patient (spouse/parent/child/daughter-in-law/son-in-law/sister or brother).
  • 2Caregiving demands comprised five items: (i) patient's activities of daily living (ADL); (ii) duration of cancer (months); (iii) treatment modality (surgery, radiotherapy, chemotherapy); (iv) end-stage caregiving (yes/no); and (v) monthly medical cost (yuan). ‘Patient's ADL’ was evaluated with the Activity of Daily Living Scale designed by Lawton and Brody in 1969, a 14-item self-rating scale with responses from 1 (able to perform independently) to 4 (unable to perform at all).15 It includes six items termed as basic ADL (i.e. bathing, dressing, walking, toileting, tooth brushing and eating), and eight items termed as instrumental ADL (i.e. meal preparation, shopping, clothes washing, telephone using, finances managing, transferring, doing housework, and medicine taking). Higher total scores of the 14 items indicate a higher level of dependence of patients. The overall Cronbach's α of the questionnaire was 0.93 in this study population. ‘Duration of cancer’ was determined as the time (months) after initial diagnosis of cancer obtained by medical records. ‘Treatment modality’ was indicated through three items, surgery, radiotherapy and chemotherapy, obtained from medical records, which represented active treatment. ‘End-stage caregiving’ was determined with the question ‘Is the care recipient (patient) in the end-stage of illness?’ obtained from medical records, which were supplied by the doctor (MD) in charge of him/her. ‘Monthly medial cost’ was indicated as the monthly cost caused by treatment after diagnosis (yuan) reported by caregivers.
  • 3Available social resources comprised of two items according to the current Chinese situation: (i) patient's health insurance status (have/haven't); and (ii) total number of available caregivers. Patient's health insurance status was determined with the question ‘Does the patient have any kind of health insurance?’ For assessment of ‘total number of available caregivers’, subjects were asked to provide the number of persons who have been sharing caregiving tasks with him/her since diagnosis.15
  • 4The caregiving situation was assessed based on two items: (i) total time of caregiving (hours); and (ii) co-residence (yes/no). ‘Total time of caregiving’ was measured with two questions: ‘Since the diagnosis of cancer, how long have you been providing direct care to patients (days)?’ and ‘When it was your turn to take care of the patients, how many hours did you usually do per day (hours/days)?’ The combination of two responses was used as an indicator of ‘total time of caregiving’ (hours). ‘Co-residence’ was measured by asking a question, ‘Do you usually live with the patient, both before and after diagnosis of cancer?’
  • 5Health status of caregivers referred to their chronic disease history. It was assessed dichotomously as ‘presence’ if any disease (e.g. hypertension, cardiovascular disease, diabetes and arthritis) had been diagnosed.

Data collection

The protocol of this study was approved by the Research Ethics Committee, China Medical University. Before performing the investigation, nurses were trained for two weeks in order to exclude implications and inducements during the interview. After obtaining the written informed consent about the conduct of the survey, face-to-face interviews were conducted through these trained nurses with patients, and with caregivers in the absence of patients, respectively.

Statistical analysis

The distributions of QOL in categorical items were tested with the Student's t-test and one-way anova, and the significance level was set at P < 0.05 (two-tailed test). When the one-way anova was found to be significant, least-significant-difference-t-test was done to perform multiple comparisons. Pearson's correlation coefficients were used for all continuous variables and Spearman's correlation coefficients were used for categorical or nonparametric variables. Colinear variables among continuous data and agreement between categorical variables were identified. All items that were related to QOL (P < 0.05) in univariate analysis were entered into the general linear model (GLM) to explore the factors associated with QOL. After the caregiver's age was fixed in the model, items with P > 0.15 were eliminated one at a time in the sequence of P-value. When a variable was eliminated, if the change of any remaining parameter estimate was greater than 20%, this variable would remain in the model as a confounder. In this study, no confounders were found during item elimination. Colinear data and agreement data were adjusted while performing GLM analysis and the model with the highest model R2 was selected. spss 11.5 for Windows was used for all data analyses.


The subject characteristics are shown in Table 1, and caregiver QOL scores are depicted in Table 2. In the subscale of physical well-being, fatigue was the greatest problem, which damaged the physical health of caregivers. In terms of psychological well-being, the family member's initial diagnosis was the most oppressive shock for caregivers, rated as a mean of 0.59. Caregivers reported that the greatest disruption in social concerns was the distress that family felt about the patient's illness. The item reflected that the most serious problem of spiritual well-being was the uncertainty about the family member's future. The QOL scores were also summarized according to subscales with the greatest disruption in psychological well-being followed by social concerns, spiritual well-being, and physical well-being. The overall QOL score for caregivers was 5.26.

Table 1.  Characteristics of the caregivers (subjects, n = 358) and patients
  Mean ± SD
  1. ADL, activities of daily living; QL, lower quartile; QOL, quality of life; QU, upper quartile.

QOL score 5.26 ± 1.48
Age 46.9 ± 12.3
  Median (QLQU)
Patient's ADL score 22.0 (14.0–32.0)
Duration of cancer (months) 4.0 (2.0–11.4)
Monthly medical cost (yuan) 10000 (5000–20000)
Total number of available caregivers 2.0 (1.0–2.0)
Total time of caregiving (h) 540 (180–2880)
  n (%)
SexMale150 (41.9)
Female208 (58.1)
Caregiver's occupationEmployed160 (44.7)
Unemployed198 (55.3)
Education<High school142 (39.7)
High school and above216 (60.3)
Monthly household income<1500/ 219.8/ 19650190 (53.1)
(yuan/ US dollars/ JPN yen per person)1500/ 219.8/ 19650∼168 (46.9)
Relationship to patientSpouse179 (50.0)
Child99 (27.7)
Parent11 (3.0)
Others69 (19.3)
SurgeryYes185 (51.7)
No173 (48.3)
RadiotherapyYes68 (19.0)
No290 (81.0)
ChemotherapyYes235 (65.6)
No123 (34.4)
End-stage caregivingYes202 (56.4)
No156 (43.6)
Patient's health insurance statusHave289 (80.7)
Haven't69 (19.3)
Co-residenceYes235 (65.6)
No123 (34.4)
Chronic diseasePresence82 (22.9)
Absence276 (77.1)
Table 2.  Caregiver QOL scores (n = 358)
ItemsMean ± SD
  1. QOL, quality of life.

Physical well-being6.43 ± 2.54
 1. Fatigue5.20 ± 3.62
 2. Appetite changes7.15 ± 3.43
 3. Pain or aches7.76 ± 3.42
 4. Sleep changes5.81 ± 3.67
 5. Rate your overall physical health6.25 ± 2.68
Psychological well-being4.69 ± 1.68
 6. How difficult is it for you to cope as a result of your family member's disease and treatment?3.96 ± 3.39
 7. How good is your overall quality of life?5.49 ± 2.80
 8. How much happiness do you feel?5.63 ± 3.44
 9. Do you feel like you are in control of things in your life?8.53 ± 2.44
10. How satisfying is your life?6.73 ± 3.28
11. How is your present ability to concentrate or to remember things?5.11 ± 3.42
12. How useful do you feel?8.44 ± 2.58
13. How distressing was your family member's initial diagnosis for you?0.59 ± 1.78
14. How distressing were your family member's cancer treatments for you?1.53 ± 2.49
15. How distressing has the time been since your family member's treatment ended?3.06 ± 3.01
16. How much anxiety do you have?4.06 ± 3.60
17. How much depression do you have?6.09 ± 3.85
18. Are you fearful of a second cancer for your family member?5.51 ± 4.35
19. Are you fearful of recurrence of your family member's cancer?1.56 ± 2.98
20. Are you fearful of the spreading (metastasis) of your family member's cancer?1.68 ± 3.11
21. Rate your overall psychological well-being5.99 ± 2.91
Social concerns5.19 ± 1.71
22. How distressing has your family member's illness been for your family?1.54 ± 2.59
23. Is the amount of support you receive from others sufficient to meet your needs?7.26 ± 3.40
24. To what degree has your family member's illness or treatment interfered with your personal relationships?5.66 ± 3.93
25. To what degree has your family member's illness or treatment interfered with your sexuality?6.72 ± 3.96
26. To what degree has your family member's illness or treatment interfered with your employment?5.89 ± 4.03
27. To what degree has your family member's illness or treatment interfered with your activities at home?5.16 ± 3.96
28. How much isolation is caused by your family member's illness or treatment?6.43 ± 3.86
29. How much financial burden resulted from your family member's illness or treatment?3.70 ± 3.70
30. Rate your overall social well-being5.99 ± 2.54
Spiritual well-being6.05 ± 1.91
31. How much uncertainty do you feel about your family member's future?2.85 ± 3.12
32. Has your family member's illness made positive changes in your life?4.15 ± 3.60
33. Do you have a purpose/mission for your life or a reason for being alive?8.36 ± 3.09
34. How hopeful do you feel?7.97 ± 3.22
35. Rate your overall spiritual well-being6.91 ± 2.66
Total score QOL5.26 ± 1.48

The results of univariate analyses of items related to caregiver's QOL are shown in Table 3. Among continuous variables, patient's ADL (r = −0.21), duration of cancer (r = −0.11), total number of available caregivers (r = 0.21) and total time of caregiving (r = −0.27) had significant correlations with QOL of caregivers. Of categorical variables, significant differences of caregiver QOL score were observed for four variables: monthly household income, relationship to patient, co-residence, and chronic disease.

Table 3.  Univariate analysis of items related to caregiver's QOL (n = 358)
  1. * P < 0.05; ** P < 0.01. ADL, activities of daily living; QOL, quality of life.

Age −0.08
Patient's ADL −0.21**
Duration of cancer (month) −0.11*
Monthly medical cost (yuan) −0.10
Total number of available caregivers 0.21**
Total time of caregiving (h) −0.27**
  QOL score Mean ± SD
SexMale5.36 ± 1.41
Female5.19 ± 1.52
Caregiver's occupationEmployed5.23 ± 1.41
Unemployed5.28 ± 1.53
Education<High school5.14 ± 1.40
High school and above5.34 ± 1.52
Monthly household income (yuan/ person)<15005.10 ± 1.46
1500∼5.44 ± 1.48*
Relationship to patientSpouse4.88 ± 1.44
Parent & Child5.50 ± 1.41
Others5.87 ± 1.41**
SurgeryYes5.32 ± 1.46
No5.22 ± 1.51
RadiotherapyYes5.24 ± 1.48
No5.29 ± 1.49
ChemotherapyYes5.23 ± 1.50
No5.31 ± 1.43
End-stage caregivingYes5.21 ± 1.49
No5.27 ± 1.45
Patient's health insurance statusHave5.23 ± 1.46
Haven't5.37 ± 1.54
Co-residenceYes4.79 ± 1.42
No5.80 ± 1.43**
Chronic diseasePresence4.77 ± 1.39
Absence5.41 ± 1.47**

In this study, ‘duration of cancer’ and ‘total time of caregiving’ were found to be colinear data and substantial agreement (Kappa value = 0.68) was found between ‘relationship to patient’ and ‘co-residence’. The results of general linear model analysis for exploring the factors related to caregiver's QOL is shown in Table 4, and it yielded a higher R2 value of 0.141 for the model containing the relationship to the patient rather than co-residence. By fixing the caregiver's age in the model and with adjustment for colinear data and agreement data, the items that were significantly associated with caregiver's QOL were, in the sequence of standardized parameter (β), spousal relationship, patient's ADL, and chronic disease. Overall spousal relationship with patients had the strongest association with caregiver's QOL.

Table 4.  The general linear model analysis for exploring the factors related to caregiver's QOL
Variables Parameter estimate (B)Standardized estimate (β)95% confidence interval for BModel R2
Lower boundUpper bound
  1. * P < 0.05; ** P < 0.01. ADL, activity of daily living; QOL, quality of life.

Intercept 5.78 4.826.75 
Age (fixed) 0.010.09−0.010.03 
Relationship to patientSpouse/other−1.03−0.35**−1.44−0.63 
Parent & child/other−0.36−0.11−0.800.08 
Patient's ADL −0.03−0.20**−0.04−0.01 
Chronic diseaseNo/yes0.470.13*0.100.84 
Monthly household income per person −0.27−0.09−0.560.030.141**


To our knowledge, the present study is the first attempt in mainland China to investigate QOL among cancer caregivers. In this study population, the average QOL score of cancer caregivers was 5.26, and the most serious problems that affected the caregiver's QOL in different dimensions were fatigue, family member's initial diagnosis, distress that family felt for patient's illness, and uncertainty about family member's future. According to the results reported by Juarez, sleep changes, initial diagnosis, distress from the illness, and uncertainty were the most serious problems in the four dimensions, respectively, in the primary family caregivers of cancer patients,16 all of which were also found in our study, except for sleep changes. Furthermore, psychological well-being and social concerns of caregivers got more disrupted than spiritual well-being and physical well-being, which was consistent with other studies.16,17 This suggests it is necessary to pay more attention to the psychological health and social well-being of cancer caregivers.

Spousal relationship has been well documented as one of the crucial factors affecting QOL of caregivers.9,10,18 In our study, variables of spousal relationship to patient and co-residence were in high agreement (Kappa value = 0.68) and GLM analysis yielded a higher R2 value of 0.141 for the model containing the relationship to patient than that for the model containing co-residence. In China, Confucian thought results in traditional expectations that the needs of family members should be placed before one's own. Thus, family caregivers generally provide direct care to patients, regardless of daily co-residence. Even during the period of hospitalization, it is the family caregiver, not a nurse, who directly takes care of an inpatient's daily life. According to the results of our study, spousal relationship has the strongest impact on caregiver's QOL. This is probably related with the fact that spouses are less likely than other caregivers to receive assistance, and they may be identified as the most vulnerable group.19 Among family caregivers, spouses tend to provide the most extensive and comprehensive care, for example help with the care recipient's ADL and instrumental ADL, whereas they also experience more distress and strain than any other family caregivers due to their intimacy with their care recipients, which inevitably results in the deterioration of their physical and psychological well-being.20 Spouses may also have worse outcomes because of age-related declines in psychological and physical resources.10 Therefore, it seems necessary to pay more attention to spousal caregivers.

ADL is an indicator of functional disability. Independent performance of ADL is the base of independent daily life.21 As for caregivers, the amount of caregiving tasks mainly depends on the ADL of care recipients. The lower level the care recipient's ADL becomes, the more caregiving demands the caregiver faces. As a result, QOL of caregivers is affected.3,22 This adverse impact of patient's ADL on QOL of caregivers was also found in the present study. However, in general, cancer patients' ADL worsen inevitably with the course of cancer. Correspondingly, for cancer caregivers, the risks of being caught by physical fatigue caused by managing physical assistance and psychological burden caused by assisting multiple tasks or tasks that are difficult or unappealing are expected to increase.22–24 Therefore, QOL among cancer caregivers seems to be affected more seriously by patient's ADL.

In addition to patient's ADL, health status (chronic disease) of caregivers was also found to be another important risk factor affecting caregiver's QOL in this study, coinciding with the conclusion reported by Morimoto, Schreiner and Asano.25 In comparison to the caregivers with good physical function, the same caregiving demand is expected to be heavier for those with worse physical function. Therefore, caregivers who have health problems of their own tend to be more distressed, as reported by Pitceathly and Maguire,18 and report a lower level of QOL, which also seems to increase the risk of being unable to continue caregiving, as reported by Kasper, Steinbach and Andrews.26

The R2 value was 0.141 in this study, which was very low. The probable reason may be related with the fact that psychological distress was not included in our study. At the beginning, we worried so much about the feasibility of conducting this investigation and we did not have the confidence to include the education level of the study population, therefore few and comparatively easy measurements were considered in order to obtain convincing data. As psychological well-being of cancer caregivers suffered from the greatest disruption among all four dimensions in our study, psychological distress may be an important factor associated with caregiver's QOL. Thus, the absence of it may be the reason for the lower R2 value.

Limitations of this study included the following. Psychological distress was not included in exploring the factors related to caregiver's QOL. However, this study represents the first attempt in mainland China to assess QOL of cancer caregivers. In light of the feasibility of conducting this survey and China-specific culture, few and comparatively easy measurements were considered in the present study. Further study with full considerations should be conducted in the future. The study population only comprised of inpatients in the Affiliated Hospitals of China Medical University with consideration to the practicability of sampling. Data about outpatients should also be collected. The study was limited by its cross-sectional design, and the conclusions drawn from this study need to be confirmed in future population-based and prospective studies. Also, the study sample was not recruited randomly.

In summary, according to the results of this study, there are increasing disruptions to a caregiver's psychological well-being and social concerns, which suggest it is necessary to pay more attention to the psychological health and social well-being of cancer caregivers. Spousal relationship and physical health status of patients and caregivers seemed to have considerable effect on a caregiver's QOL.


Our special thanks to the participants and nurses in the study.