Aging and Health Research 2 (2022) 100052

Contents lists available at ScienceDirect

Aging and Health Research

journal homepage: www.elsevier.com/locate/ahr

Physical and mental health of family caregivers of older parents and

grandchildren in China
Yujun Liu a,∗, M. Courtney Hughes b, Karen A. Roberto c, Jyoti Savla d
a
School of Family and Consumer Sciences, Northern Illinois University College of Health and Human Sciences, 1425 W. Lincoln Hwy, DeKalb, IL, USA
b
School of Health Studies, Northern Illinois University, DeKalb, IL, USA
c
Center for Gerontology and the Institute for Society, Culture and Environment, Virginia Tech, Blacksburg, VA, USA
d
Department of Human Development and Family Science, Center for Gerontology, Virginia Tech, Blacksburg, VA, USA

a r t i c l e i n f o a b s t r a c t

Keywords: Background: Families have been recognized as the primary source of care for younger and older family members
Family caregiving in Chinese tradition. The aim of this study was to examine the association between informal caregiving for aging
Mental health parents and/or grandchildren and health outcomes of family caregivers in mainland China.
Physical health
Methods: Using data from the 2013 and 2018 waves of China Health and Retirement Longitudinal Study
China
(CHARLS), we conducted univariate comparisons of demographics characteristics between informal caregivers
Grandparenting
Caregiver health and non-caregivers using t-test and chi square test. We used linear regression and negative binomial regression to
examine the physical and mental health outcomes of family caregivers aged 45 and over living in mainland China.
This cross-sectional study was based on the self-reported measures of caregiving status, demographic factors, and
health outcomes.
Results: Compared to non-caregivers, informal caregivers were more likely to be women, younger in age, married,
living in rural areas, and with less years of education. Older age, being male, being married, living in an urban
area, having a higher level of education, and having better self-reported health were associated with better
mental health outcomes among family caregivers. Being male, younger age, and working, and having better self-
reported health, were associated with better physical function among family caregivers. Caregivers who only
cared for their grandchildren reported greater life satisfaction, fewer depressive symptoms, and a higher level of
physical function than non-caregivers. In addition, caregivers who cared for only their parents reported fewer
depressive symptoms than non-caregivers.
Conclusions: Health practitioners and policymakers should consider the needs for additional resources for family
caregivers of older adults and children. Targeted policies and programs may improve family caregivers’ health,
especially for female family caregivers living in rural areas.

1. Introduction by 2050 [7]. As the number of children per family decreases and as life
expectancy increases in China, fewer adult children or children-in-law
When faced with illness or disability, families have long been rec- will be available to support their older parents [8,9]. Researchers and
ognized as the primary source of care for older people [1,2]. In China, practitioners are concerned about the strain on adult children without
which has over one-fifth of the world’s older population [3], providing siblings to share caring responsibilities [10–12].
care for the aging population is a strong cultural value. Adult children Another aspect of Chinese cultural norms, rooted in the Confucian-
are traditionally expected to provide care for their aging parents [4]. ism philosophy which promotes family harmony [3,13], is multigenera-
However, changes in the demographic characteristics of China’s fami- tional co-residence, or a “networked” family living arrangement. In this
lies have resulted in fewer adult children caregivers available to support latter situation, relatives live apart but close [6], allowing grandpar-
their older parents. The “4–2–1″ paradigm, consisting of four grandpar- ents to be involved in their grandchildren’s care daily without claiming
ents, two adult children, and one grandchild, represents the anticipated a custodial responsibility. With its cultural reliance on filial piety and
family structure in China in upcoming decades [5,6]. China’s older adult practices focusing on collective family interests [6], China represents an
support ratio (number of adults aged 25–64 years per number of persons ideal setting for examining health outcomes when considering multiple
older than 64 years) is projected to fall from nearly 13 in 2000 to 2.1 types of family caregiving relationships.

∗
Corresponding author.
E-mail address: lyujun@niu.edu (Y. Liu).

https://doi.org/10.1016/j.ahr.2021.100052
Received 30 August 2021; Received in revised form 15 December 2021; Accepted 26 December 2021
Available online 28 December 2021
2667-0321/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Y. Liu, M.C. Hughes, K.A. Roberto et al. Aging and Health Research 2 (2022) 100052

Findings from previous studies conducted in western countries sug- adults living in nursing homes or care facilities were not sampled, but
gested that sociodemographic characteristics such as race/ethnicity, in- the survey followed up with respondents of Wave 1 who later entered
come, education, and marital status differed significantly between fam- care facilities. The same study procedures were used in 2013, and the
ily caregivers and non-caregivers [4,8,9,14]. For example, using data survey followed up with 15,770 respondents (84.8% of the total sample)
from the U.S. Centers for Disease Control and Prevention’s Behavioral from the 2013 wave.
Risk Factor Surveillance System, Anderson and colleagues (2013) re-
ported that caregivers in the U.S. were significantly more likely to be 50 2.1.1. Sampling procedures and data collection
to 64 years of age, female, non-Hispanic black, having some college edu- At the provincal level, individuals from 28 provinces were sampled.
cation, and married. In studies that went beyond describing caregiver’s On the neighborhood level, the sample used administrative villages
characteristics, findings suggested that providing care could threaten (cun) in rural areas and community (shequ) in urban areas, as primary
family caregivers’ physical and mental health, which may compromise sampling units (PSUs). Three PSUs were selected within each unit at
their ability to continue in their caregiver role [17]. Family caregivers the provincial level, using probabilities proportional to size sampling
reported fatigue and sleep disturbances, lower immune functioning, in- procedures. After final sampling work in the PSU completed, a letter
creased insulin levels and blood pressure and greater risk for cardio- to potential respondents was sent to each selected household to inform
vascular disease [11,17]. Caregiver’s burden and strain also have been them about the study. In each sampled household, a short screening
found to be directly associated with caregivers’ engagement in adverse form was used to identify whether the household had a member meeting
health behaviors such as smoking and greater alcohol consumption [18]. the age criteria. Individuals aged 45 years and above were selected as
Furthermore, several studies have demonstrated a relationship between main respondents. If this person was married, his or her spouse was also
caregiving and anxiety, stress, and depression [19]. interviewed. If an age-eligible person was too frail to answer questions,
Earlier research in the U.S. reported negative health effects of car- a proxy respondent was identified to help him/her to answer questions,
ing for grandchildren. In particular, extensive and custodial grandpar- usually a spouse or knowledgeable adult child in the household. Indi-
enting (providing primary care for the grandchild) has been associated viduals were interviewed by CHARLS research team members who were
with poor health outcomes for grandparents, including elevated depres- professionally trained for conducting face-to-face interviews.
sive symptoms, declined life satisfaction and more functional limita-
tions [20]. In contrast, recent studies conducted in China, Hong Kong,
2.2. Current study sample
and Taiwan found health advantages for caregivers who provided oc-
casional, extensive, or even custodial care to grandchildren [21]. Com-
We selected the study sample from the combined CHARLS datasets
pared to older adult non-caregivers, grandparent caregivers experienced
(N = 11,203), including CHARLS 2013 and CHARLS 2018. We included
reduced depressive symptoms, better self-rated health, greater life sat-
all the participants in the combined dataset for the analysis of descrip-
isfaction and fewer functional health limitations [22,23].
tive statistics. Among all respondents, 3787 participants (about 35% of
Although China has the largest and fast growing older population,
the total sample) were identified as caregivers for grandchildren and
and family members are expected to provide care for the older adults
1083 were identified as caregivers for older parents (10% of the total
and grandchildren, little is known about the sociodemographic char-
sample). About 3.4% (N = 373) participants were caregivers for both
acteristics of family caregivers in mainland China [13,15,16]. Further-
grandchildren and older parents. When respondents provided care to
more, to our knowledge, there are no studies in Eastern or Western coun-
parents or grandchildren both at CHARLS 2013 and CHARLS 2018, we
tries investigating caregiving for both older parents and grandchildren
used their data from CHARLS 2018 to examine their recent caregiving
and its potential relationship with long-term caregiver’s health.
experience.
Our study aimed to describe the characteristics of family caregivers
in mainland China. We aimed to examine the associations between care-
giver demographic characteristics and caregiving roles and long-term 2.3. Measures
health outcomes of caregivers. Our research based on Pearlin’s stress
process model [5,24], which many researchers in Western countries The sample for the regression and binominal analysis included par-
have used to examine how providing care could influence caregivers’ ticipants who had given personal care to their old parents and/or grand-
physical, emotional, and social health [25,26]. We chose Pearlin’s four children for one month or more during the past 12 months. Participants
domains (the background and context of stress, the primary and sec- who responded “Yes” to the question “Did you or your spouse take care
ondary stressors, the mediators of stress, and the outcomes of stress) to of your parents or parents-in-law during the last year in assisting them
study caregiver health as a potential consequence of several interrelated in their daily activities or other activities (e.g., household chores, meal
factors [24]. Our goal is to improve knowledge about the long-term preparation, laundry, going out, grocery shopping, financial manage-
health outcomes of caregivers for older parents and/or grandchildren ment, etc.)” were categorized as caregivers for parents. Participants who
while interpeting possible primary and secondary stressors upon which answered “Yes” to the question “During last year, did you/ your spouse
policymakers and practitioners could consider effective interventions. spend time in taking care of your grandchildren” were categorized as
caregivers for grandchildren. Those who answered “Yes” to both ques-
2. Methods tions were categorized as caregivers for both. We characterized the sam-
ple by two caregiving factors (caregiving status and time spent on care-
2.1. Data source and sample giving) and six demographic variables (including age, gender, marital
status, urban/rural residency, education, and working status) (Table 1).
Data came from the China Health and Retirement Longitudinal Study Our dependent variables were four self-reported measures of health sta-
(CHARLS), a nationally representative and publicly available longitudi- tus, two related to physical health and two related to mental health.
nal survey which include a wide range of individual and household in-
formation about people aged 45 years and over and their partners living 2.4. Statistical analysis
in private households in mainland China. CHARLS was designed to be
comparable with the Health and Retirement Study (HRS) of U.S. and In preparation for conducting the statistical analysis for the current
related aging surveys around the world while being sensitive to the spe- study, we ran tests and removed outliers and data from respondents
cific conditions of China. The baseline wave of CHARLS was collected other than the selected participants (data from responses from other
in 2011 and included 17,500 individuals from 10,000 households in 28 people in the household). We used the full information maximum likeli-
provinces, 150 counties/districts, and 450 villages/communities. Older hood (FIML) estimation, a pragmatic missing data estimation approach,

2
Y. Liu, M.C. Hughes, K.A. Roberto et al. Aging and Health Research 2 (2022) 100052

Table 1 between the caregivers for parents only and caregivers for grandchildren
Definitions of study variables used in CHARLS. only, namely the caregivers for parents tended to be younger, married,
Variable Definition and with lower education levels than the caregivers for grandchildren.
Compared to non-caregivers, the family caregivers of both parents and
Dependent
grandparents were more likely to be women, younger in age, married,
Physical health
Physical function Indicated 4-point scale (1 = no difficulty to living in rural areas, and having less years of education.
4 = absolutely cannot do it) responses to ability to
complete activities of daily living (dressing, bathing or 3.1. Findings of regression and negative binomial analysis
showering, eating, transferring (getting into or out of
bed), toileting, and continence) and instrumental
activities of daily living (doing household chores, As shown in Tables 3 and 4, family caregivers caring only for grand-
preparing hot meals, grocery shopping, making phone children reported greater life satisfaction, fewer depressive symptoms,
calls, and taking medications). and a higher level of physical function than non-caregivers. Family care-
No. of chronic diseases Indicated number of listed chronic diseases ranged from givers caring only for parents also reported fewer depressive symptoms.
1 to 14.
Overall, the greater the amount of time caregivers spent care provid-
Chronic diseases included: hypertension, dyslipidemia,
diabetes, cancer, chronic lung diseases (bronchitis, ing, the lower their life satisfaction and higher their level of physical
emphysema), liver disease, heart diseases, stroke, function. Better mental health outcomes were associated with the fol-
kidney diseases, digestive diseases, psychiatric problem, lowing caregiver characteristics: older age, being male, higher level of
memory-related diseases, arthritis or rheumatism,
education, being married, having better self-reported health, and living
asthma.
Mental health in urban areas. There was no relationship between working and men-
Life satisfaction Indicated Likert scale response (1 = completely satisfied tal health outcomes. Better physical function was negatively associated
to 5 = not at all satisfied) to “Please think about your with older age but positively associated with being male, working, and
life-as-a-whole. How satisfied are you with it? Are you having better self-reported health (Table 4). Additionally, no relation-
completely satisfied, very satisfied, somewhat satisfied,
ship existed between marital status or urban/rural residency and phys-
not very satisfied, or not at all satisfied?”
Depression Assessed using the 10-item Center for Epidemiological ical function and disease status.
Studies Depression Scale [27].
Independent 4. Discussion
Caregiving status Caregiving status was categorized into caregiver for
parents only, caregiver for grandchildren only and
caregiver for both parents and grandchildren. This study expands the growing literature on the health implications
Time spent on caregiving Response to “Approximately how many weeks per year of family caregiving on adults in developing countries. First, it situates
and how many hours per week did you spend last year grandparenthood and caregiving in a new context in which family care-
taking care of your parents or parents-in-law?” givers are involved in caring for multiple generations. Such dual burden
Covariates
Age Age of respondent
of caregiving is likely to increase at a population level as older adults’
Gender Gender of respondent life expectancy and health status continue to improve [2]. Second, it
Marital status Indicated yes to married or cohabiting highlights the need for effective interventions to improve the health out-
Urban/rural residency Urban or rural residency status comes of family caregivers. A scoping review found that a wide range
Education Level of education completed
of such interventions exists in high-income countries [31]; however, in-
Working status Indicated yes to earn a wage, run own business, or
unpaid family business work terventions suitable for lower-income countries are lacking [32]. One
reason may be that lower- and middle-income countries are focused on
chronic disease prevention for the growing number of residents facing
using Stata 14 (StataCorp LP, College Station, TX, USA) [28,29] to ad- more illnesses than for caregiver interventions [33,34]. Third, we char-
dress missing data. FIML has been shown to produce unbiased parame- acterized our sample using various independent variables which allowed
ter estimates and standard errors under missing at random (MAR) and us to more precisely identify which factors were associated with caregiv-
missing completely at random (MCAR) conditions. FIML requires that ing health outcomes. Our findings are consistent with previous research
missing values be at least MAR [29]. For all continuous variables, we in western countries, showing that disparities in mental health outcomes
ran tests for skewness and kurtosis, checked for multivariate normality, for family caregivers are based on demographic factors including gen-
and conducted transformations for data not normally distributed. der, education, and metropolitan status [19,20].
We described all study variables using means and frequency. We Unlike previous studies that reported poor physical health among
also conducted univariate comparisons of informal caregivers and non- grandparent caregivers [35], we found that family caregivers who
caregivers using t-test and chi square test. This cross-sectional study was cared only for their grandchildren had better physical health than non-
based on the self-reported measures of caregiving status, demographic caregivers. One explanation for this finding may be that taking care of
factors, and health outcomes. We used a negative binomial regression to children may involve being physically active to “keep up” with them.
determine the association of the independent variable (caregiving sta- To test this hypothesis, The Healthy Grandparenting Project in Belgium, a
tus) on physical function (i.e., ADL/IADL) and the number of chronic clinical trial currently ongoing, is examining physical activity levels in
conditions of the caregivers. This type of analysis is used with skewed caregiving grandparents and investigating its health impact compared
dependent count variables when the variance is larger than the mean to the physical activity levels of non-caregiving grandparents and older
[30]. All analyses were controlled for age, marital status, gender, edu- adults without grandchildren [36].
cation, residence, working status, and self-reported health. In addition, We also found that compared with non-caregivers, family caregivers
we performed linear regression models to examine the associations be- who provided care for parents or grandchildren reported lower depres-
tween caregiving, depression symptoms, and life satisfaction using the sive symptoms and higher life satisfaction. We postulated that the cul-
2018 wave data. tural value of filial and family obligation may have served as a buffering
effect, weakening the association between caregiver stress and depres-
3. Results sive symptoms. It is possible that by providing care to their parents, the
adult children caregivers in mainland China have achieved the sense of
Descriptive characteristics of the informal caregivers are provided fulfilling their family obligations and their role as filial children [37],
in Table 2. There were many consistencies among the characteristics of and consequently not feeling caregiving stress and depressive symp-
the various caregiver status groups. However, some differences existed toms often reported in studies conducted in western nations [38]. In

3
 Y. Liu, M.C. Hughes, K.A. Roberto et al.
Table 2
Univariate comparisons of informal caregivers and non-caregivers, CHARLS data, 2013, 2018.

t-test

Group 1 Caregivers for Group 2 Caregivers for Group 3 Caregivers Group 4 Non-Caregivers Group 1 Group 2 Group 3
Continuous Variable Parents (n = 1083) Grandchildren (n = 3787) for Both (n = 373) (n = 6333) vs 4 p value vs 4 p value vs 4 p value

Depressive 20.55 (4.14) 18.89 (3.45) 20.29 (3.90) 19.83 (3.61) 5.93 <0.001 12.88 <0.001 2.38 0.03
Symptoms,
Score, Mean (SD)
Age, Year, Mean (SD) 51.54 (6.04) 58.44 (8.76) 52.43 (6.54) 60.07 (11.63) 23.59 <0.001 7.45 <0.001 12.57 <0.001
Total Caregiving 2.52 (0.80) 2.57 (0.78) 2.36 (0.85) – – – –
Time
(Log-transformed hours),
Mean (SD)
Chi2
Categorical Variable Group 1 p value Group 2 p value Group 3 p value
vs 4 vs 4 vs 4
4

Gender, n(%) 161.28 <0.001 275.77 <0.001 86.22 <0.001

Male 404 (37.30) 1552 (40.98) 125 (33.51) 3675 (58.02)
Female 679 (62.70) 2235 (59.02) 248 (66.49) 2658 (41.98)
Marital Status, n(%) 445.61 <0.001 698.93 <0.001 180.33 <0.001
Married or Cohabiting 1006 (92.89) 3210 (84.76) 352 (94.36) 3369 (59.50)
Single (widowed, divorced, never 77 (7.10) 567 (15.24) 21 (5.64) 2297 (40.50)
married)
Urban/rural 81.57 <0.001 39.25 <0.001 12.51 <0.001
Residency, n(%)
Urban 378 (27.68) 1052 (27.78) 109 (29.76) 1250 (22.01)
Rural 703 (71.33) 2735 (72.22) 264 (70.24) 4399 (77.99)
Education, n(%) 124.72 <0.001 134.34 <0.001 53.46 <0.001
Less than high school 830 (75.31) 3332 (88.00) 300 (80.60) 5509 (87.00)
High school 166 (16.70) 234 (6.20) 41 (11.10) 677 (10.70)
College and above 87 (7.99) 221 (5.80) 32 (8.30) 147 (2.30)

Aging and Health Research 2 (2022) 100052

Note. CHARLS: China Health and Retirement Longitudinal Study; Bonferroni adjustments were used for multiple comparisons.
Y. Liu, M.C. Hughes, K.A. Roberto et al. Aging and Health Research 2 (2022) 100052

Table 3
Results of the regression analysis for the associations between caregiving activities at CHARLS 2013 and life satisfaction and depressive symptoms among
caregivers at CHARLS 2018.

Life Satisfaction Depressive Symptoms

b SE p value b SE p value

Covariates (2013
wave)
Age 0.086 0.002 <0.001 0.049 0.009 0.08
Marital Status (Ref. 0.078 0.050 <0.001 −0.015 0.270 0.06
Not married/in a
relationship)
Gender (Ref. Men) 0.011 0.031 0.07 0.147 0.161 <0.001
Education 0.023 0.009 0.12 −0.112 0.045 <0.001
Urban/rural 0.040 0.038 0.17 0.073 0.198 <0.001
residency (Ref.
Urban)
Working Status (Ref. 0.014 0.037 0.57 −0.020 0.196 0.11
Not working)
Self-reported Health 0.041 0.023 0.008 −0.023 0.011 0.002
Caregiving Status
(Ref. Not caregiving)
Care Only for 0.021 0.008 0.10 −0.033 0.017 0.003
Parents
Care Only for 0.032 0.012 0.001 −0.021 0.003 <0.001
Grandchildren
Care Both 0.019 0.004 0.27 0.028 0.002 0.38
Caregiving Time −0.013 0.090 0.03 0.033 0.252 0.24
(Log transformed)
R2 0.27 0.46

Note. SE: standard error.

Table 4
Results of the negative binomial analysis for the associations between caregiving activities at CHARLS 2013 and physical function and disease status of caregivers
at CHARLS 2018.

Physical Function No. Chronic Conditions

b SE p value b SE p value

Covariates (2013
wave)
Age 0.004 0.002 0.08 0.018 0.003 <0.001
Marital Status (Ref. −0.006 0.041 0.15 −0.128 0.097 0.13
Not married/in a
relationship)
Gender (Ref. Male) −0.069 0.077 <0.001 −0.033 0.055 0.57
Education −0.028 0.012 0.72 −0.022 0.015 0.55
Urban/rural 0.045 0.051 0.96 0.033 0.067 0.77
residency (Ref.
Urban)
Working Status (Ref. −0.129 0.049 0.08 −0.354 0.063 <0.001
Not working)
Self-reported Health 0.031 0.067 0.03 −0.119 0.090 0.06
Caregiving Status
(Ref. Not caregiving)
Care Only for −0.012 0.003 0.10 0.031 0.012 0.12
Parents
Care Only for 0.063 0.013 0.008 −0.017 0.003
Grandchildren
Care Both 0.026 0.006 0.99 0.023 0.006 0.71
Caregiving Time 0.013 0.090 0.02 0.033 0.252 0.08
(Log-transformed)
Model Fit
AIC 10,301.51 8178.40
BIC 10,345.24 8226.51

Note. SE: standard error; Ref., Reference category; AIC, Akaike information criterion; BIC, Bayesian information criterion.

this way, caregivers providing care for their older parents may be rec- were more likely to report better self-rated health, higher life satisfac-
ognized and praised by their parents and others in their social network tion, and fewer depressive symptoms. We postulated the positive finding
as filial and commendable, providing positive encouragement for the could be attributed to the influence of Chinese culture of multigenera-
adult children caregivers. Thus, receiving positive psychological affir- tional co-residence, or a “networked” family living arrangement [6].
mation and support may moderate or reduce depressive symptoms in Grandparents could have the support and involvement of the parents
adult children caregivers. Consistent with a previous study from Tai- of the grandchildren. Emotional and social support might mediate the
wan [21], compared with noncaregivers, multigenerational caregivers relationship between caregiving and health among caregivers for grand-

5
Y. Liu, M.C. Hughes, K.A. Roberto et al. Aging and Health Research 2 (2022) 100052

children. Our findings underscore the need for future research to clarify Acknowledgements
the moderating role of the cultural values of filial obligation for adult
children caregivers for aging parents and the mediating effect of social The authors declared no potential conflicts of interest with respect to
and emotional support on the association between caregiving and health the research, authorship, and/or publication of this article. The authors
outcomes among grandparent caregivers in China. had no receipt of funding to declare.

References
4.1. Limitations
[1] Fuller-Iglesias H, Webster NJ, Antonucci TC. The complex nature of family sup-
Unlike most previous studies about Chinese family caregivers which port across the life span: implications for psychological well-being. Dev Psychol
rely on regional, non-representative samples, this study draws on pub- 2015;51(3):277–88. doi:10.1037/a0038665.
[2] Pezzin LE, Pollak RA, Schone BS. Complex families and late-life outcomes among
licly available data from a nationally representative longitudinal sur- elderly persons: disability, institutionalization, and longevity. J Marriage Fam
vey, making the findings more generalizable, transparent, and replica- 2013;75(5):1084–97. doi:10.1111/jomf.12062.
ble. However, the study is not without its limitations. First, the find- [3] Lu J, Liu Q. Four decades of studies on population aging in China. China Popul Dev
Stud 2019;3(1):24–36.
ings are based on self-reported data, which may not be as accurate [4] Sun F. Caregiving stress and coping: a thematic analysis of Chinese fam-
as biomedical measurements. However, Ning et al. showed moderate ily caregivers of persons with dementia. Dementia 2014;13(6):803–18.
agreement between self-reported data and biomedical measurements doi:10.1177/1471301213485593.
[5] Yang X, Hao Y, George SM, Wang L. Factors associated with health-related qual-
using the CHARLS dataset [39]. Second, we were limited to the basic ity of life among Chinese caregivers of the older adults living in the commu-
questions about for whom the respondent cared. Thus, we could not nity: a cross-sectional study. Health Qual Life Outcomes 2012;10(1):143–243.
analyze secondary stressors (e.g., work stressors, relationship stressors, doi:10.1186/1477-7525-10-143.
[6] Yang J, He Z. Continuity or change? Chinese family in transitional era. Popul Res
balancing childcare) that earlier studies have shown may combine with
2014;38(2):36–51. Retrieve from http://rkyj.ruc.edu.cn/EN/Y2014/V38/I2/36.
caregiving to impact caregivers’ physical and mental health. Third, work [7] Zhao Y, Hu Y, Smith J, Strauss J, Yang G. Health outcomes and socio-economic
status was categorized as working or non-working as provided in the status among the mid-aged and elderly in China: evidence from the CHARLS national
CHARLS dataset. Future studies should include information about vari- baseline data. J Econ Ageing 2014;3:29–43. doi:10.1016/j.jeoa.2014.05.001.
[8] Gui T, Koropeckyj-Cox T. I am the only child of my parents:” Perspectives on future
ous types of employment to assess the influence of workload and work- elder care for parents among Chinese only-children living overseas. J Cross Cult
related stress on caregiving outcomes for persons in mainland China. Gerontol 2016;31(3):255–75. doi:10.1007/s10823-016-9295-z.
Fourth, although the effect of caregiving status on the two mental health [9] Shen Y, Yeatts DE. Social support and life satisfaction among older adults in
China: family-based support versus community-based support. Int J Aging Hum Dev
indicators is statistically significant, the effect size is small. We recom- 2013;77(3):189–209. doi:10.2190/AG.77.3.b.
mend further research examining this relationship by cross-validating [10] Liang Y, Wu W. Exploratory analysis of health-related quality of life among the
the findings using other datasets and indicators of mental health. empty-nest elderly in rural China: an empirical study in three economically de-
veloped cities in eastern China. Health Qual Life Outcomes 2014;12(1):59 -59.
doi:10.1186/1477-7525-12-59.
[11] Lin K, Yin P, Loubere N. Social support and the “left behind” elderly in rural
4.2. Implications China: a case study from Jiangxi province. J Community Health 2014;39(4):674–
81. doi:10.1007/s10900-014-9864-4.
[12] Pei X, Luo H, Lin Z, Keating N, Fast J. The impact of eldercare on adult chil-
Considering the cultural context of caregiving, specifically filial and
dren’s health and employment in transitional China. J Cross Cultural Gerontol
family obligation, this study sheds new light on the relevant predispos- 2017;32(3):357–72.
ing and enabling factors that practitioners and policymakers should con- [13] Muyskens K. Will Confucian values help or hinder the crisis of elder care in modern
sider when developing and delivering programs and services for fam- Singapore? Asian Bioeth Rev 2020;12:117–34. doi:10.1007/s41649-020-00123-5.
[14] Wang H, Chen K, Pan Y, Jing F, Liu H. Associations and impact factors between
ily caregivers in mainland China. For example, while there are health- living arrangements and functional disability among older Chinese adults: E53879.
related benefits related to caregiving for grandchildren or parents, our PLoS ONE 2013;8(1). doi:10.1371/journal.pone.0053879.
study findings suggested that there is also room for improvement in the [15] Liu Y, Fu R, Roberto KA, Savla J. Depressive symptoms among adult children care-
givers in China: moderating effects of working status and gender. Aging Ment Health
health outcomes of family caregivers from certain demographic groups. 2019;23(10):1391–9. doi:10.1080/13607863.2018.1488943.
For example, females often carry the burden of caregiving roles. Stud- [16] Zeng L, Zhu X, Meng X, Mao Y, Wu Q, Shi Y, et al. Responsibility and bur-
ies showed that compared with male caregivers, female caregivers may den from the perspective of seniors’ family caregivers: a qualitative study
in Shanghai, China. Int J Clin Exp Med 2014;7(7):1818–28. Retrieve from
face higher levels of caregiving stress, have fewer social resources, and http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4132150/pdf/ijcem0007-1818.pdf.
report lower levels of psychological and physical health [33]. Health sys- [17] Zehner Ourada VE, Walker AJ. A comparison of physical health outcomes for
tems should consider providing resources, including mental health and caregiving parents and caregiving adult children. Fam Relat 2014;63(1):163–77.
https://doi-org.auth.lib.niu.edu/10.1111/fare.12046.
wellness education, to help female caregivers cope with their respon-
[18] Tough H, Brinkhof MW, Fekete C. Is informal caregiving at odds with opti-
sibilities. Another group in need of mental health resources are family mal health behaviour? A cross-sectional analysis in the caregiving partners of
caregivers living in rural areas. Xu et al. described the disparities in persons with spinal cord injury. Health Psychol Behav Med 2020;8(1):526–42.
doi:10.1080/21642850.2020.1838282.
mental health service utilization in China based on metropolitan status
[19] Gouin J, Estrela C, Desmarais K, Barker ET. The impact of formal and informal sup-
and suggested increasing insurance reimbursement in rural areas would port on health in the context of caregiving stress. Fam Relat 2016;65(1):191–206.
help increase service uptake [39]. https://doi-org.auth.lib.niu.edu/10.1111/fare.12183.
In conclusion, our study contributes to the international literature on [20] Pulgaron ER, Marchante AN, Agosto Y, Lebron CN, Delamater AM. Grandparent
involvement and children’s health outcomes: the current state of the literature. Fam
family caregiving for older parents and grandchildren using evidence Syst Health 2016;34(3):260.
from mainland China. We identified both beneficial and determential [21] Ku LJE, Stearns SC, Van Houtven CH, Lee SYD, Dilworth-Anderson P, Konrad TR. Im-
effects of caregiving for family caregivers’ physical and mental health. pact of caring for grandchildren on the health of grandparents in Taiwan. J Gerontol
Ser B Psychol Sci Soc Sci 2013;68(6):1009–21.
Future research should consider using longitudinal data to explore the [22] Zhou J, Mao W, Lee Y, Chi I. The impact of caring for grandchildren on grand-
potential moderating and mediating factors that can account for positive parents’ physical health outcomes: the role of intergenerational support. Res Aging
health consequences of family caregiving. 2017;39(5):612–34. doi:10.1177/0164027515623332.
[23] Lou VW, Lu N, Xu L, Chi I. Grandparent–grandchild family capital and self-
rated health of older rural Chinese adults: the role of the grandparent–
parent relationship. J Gerontol Ser B Psychol Sci Soc Sci 2013;68(4):599–608.
Declaration of Competing Interest doi:10.1093/geronb/gbt040.
[24] Pearlin LI, Mullan JT, Semple SJ, Skaff MM. Caregiving and the stress process: an
The authors declare that they have no known competing financial overview of concepts and their measures. Gerontologist 1990;30(5):583–94.
[25] Haley WE, Roth DL, Howard G, Safford MM. Caregiving strain and estimated risk
interests or personal relationships that could have appeared to influence for stroke and coronary heart disease among spouse caregivers: differential effects
the work reported in this paper. by race and sex. Stroke 2010;41(2):1–13. doi:10.1161/STROKEAHA.109.568279.

6
Y. Liu, M.C. Hughes, K.A. Roberto et al. Aging and Health Research 2 (2022) 100052

[26] Mausbach BT, et al. A longitudinal analysis of the relations among stress, depres- [33] Hu M, Ma C, Sadarangani T, Wu B. Social-behavioral interventions for Asian and
sive symptoms, leisure satisfaction, and endothelial function in caregivers. Health Hispanic American dementia caregivers: an integrative review. Aging Health Res
Psychol 2012;31(4):433–40. doi:10.1037/a0027783. 2021:100027. doi:10.1016/j.ahr.2021.100027.
[27] Chin W, Choi E, Chan K, Wong C. The psychometric properties of the Center for [34] Xu Junfang, et al. Rural-urban disparities in the utilization of mental health in-
Epidemiologic Studies Depression Scale in Chinese primary care patients: factor patient services in China: the role of health insurance. Int J Health Econ Manag
structure, construct validity, reliability, sensitivity and responsiveness. PLoS ONE 2018;18(4):377–93. doi:10.1007/s10754-018-9238-z.
2015;10(8):1–16. doi:10.1371/journal.pone.0135131. [35] Chang YP, Lorenz RA, Phillips M, Peng HL, Szigeti K. Fatigue in fam-
[28] Collins L, Schafer J, Kam C. A comparison of inclusive and restrictive strate- ily caregivers of individuals with dementia: associations of sleep, depres-
gies in modern missing data procedures. Psychol Methods 2001;6(4):330–51. sion, and care recipients’ functionality. J Gerontol Nurs 2020;46(9):14–18.
doi:10.1037//1082-989X.6.4.330. doi:10.3928/00989134-20200527-01.
[29] Enders C, Bandalos D. Relative performance of full information maximum likelihood [36] Vermote M, Deliens T, Deforche B, et al. The impact of non-residential grand-
estimation for missing data in structural equation models. Educ Psychol Pap Publ child care on physical activity and sedentary behavior in people aged 50 years and
2001;64 hp://digitalcommons.unl.edu/edpsychpapers/64. over: study protocol of the Healthy Grandparenting Project. BMC Public Health
[30] Long JS, Freese J. Regression models for categorical dependent variables using stata. 2021;21:38. doi:10.1186/s12889-020-10024-9.
Stata press; 2006. [37] Ahmed HU, Jielin ZA. Behavioural Study on the influences of Confucianism in Chi-
[31] Gaziano TA, Pagidipati N. Scaling up chronic disease prevention interventions in nese society. Philos Prog 2017:109–32. doi:10.3329/pp.v61i1-2.44205.
lower- and middle-income countries. Annu Rev Public Health 2013;34:317–35. [38] Lai DW. Filial piety, caregiving appraisal, and caregiving burden. Res Aging
doi:10.1146/annurev-publhealth-031912-114402. 2010;32(2):200–23. doi:10.1177/0164027509351475.
[32] Hinton L, Tran D, Nguyen T, et al. Interventions to support family caregivers of [39] Ning M, Zhang Q, Yang M. Comparison of self-reported and biomedi-
people living with dementia in high, middle and low-income countries in Asia: a cal data on hypertension and diabetes: findings from the China Health
scoping review. BMJ Glob Health 2019;4:e001830. and Retirement Longitudinal Study (CHARLS). BMJ Open 2016;6:e009836.
doi:10.1136/bmjopen-2015-009836.