Abstract
Objectives
Family health resources are critical elements for rural households to withstand health shocks. However, the mechanisms through which these resources are linked to the economic risk of diseases remain unclear. Grounded in the Conservation of Resources (COR) theory, this study investigates the association of family health resources with the economic risk of diseases among rural households.
Methods
Utilizing data from the 2020 wave of the China Family Panel Studies (CFPS), the economic risk of diseases was quantified using the Three-Stage Feasible Generalized Least Squares (FGLS) method. Path analysis was employed to examine the direct effects and the mediating roles.
Results
The average score of family health resources among rural households was 3.34 ± 0.73. A total of 400 households (9.42%) were identified as high-risk. Family health resources significantly predicted a significantly lower economic risk of diseases (P < 0.001). Mechanism analysis revealed that resource quantity is associated with economic risk through three mediating functions: positive effects from accumulation and investment function, and a negative effect from compensation function.The mediation effects of these three functions accounted for 12.492%, 2.580%, 7.196% of the total effect, respectively. Compared to overall resource-disadvantaged households, the resource-balanced advantaged households demonstrated stronger mediation proportions through the accumulation function and investment function—29.774% and 9.548%, respectively.
Conclusion
Family health resources are negatively associated with the economic risk of diseases by promoting the accumulation and investment functions while a lower reliance on the compensation function. Targeted optimization of family health resource allocation is needed to strengthen resilience against disease-related economic risk.
Keywords: Family health resources, The economic risk of diseases, Resource management functions, Path analysis
Introduction
Health shocks not only impair individual physical and mental health but also exert adverse effects on household economic status. This economic risk of diseases poses a universal challenge facing households globally [1]. The UHC Global Monitoring Report 2025 reveals that despite improvements in global healthcare service coverage, approximately 2.1 billion people still face financial hardship when accessing health services, with 1.6 billion falling into or deepening poverty due to health expenditures. Illness imposes an immediate strain on household finances through high out-of-pocket medical expenses and may also reduce or even eliminate the labor capacity of those affected. This leads to substantial productivity losses [2], and crowds out essential consumption and productive investments, thereby weakening the ability of patients and their families to generate future income and jeopardizing their long-term economic welfare [3, 4]. The economic consequences are amplified for rural households due to their reliance on manual labor, which makes income contingent upon physical health [5]. This challenge is notably pronounced in China, where a rapid epidemiological transition has fundamentally altered the nature of household economic risks. While acute infectious diseases were once the dominant concern, the disease burden is now overwhelmingly driven by chronic conditions. According to data released by the National Health Commission of China, chronic non-communicable diseases (NCDs) now account for 88.5% of total deaths and over 70% of the disease burden, posing potential risks to families. Given the magnitude of this problem, identifying factors associated with a lower economic risk of diseases is a critical academic and policy imperative.
Existing literature has extensively explored the determinants of the economic risk of diseases from a multi-dimensional perspective. First, regarding the disease itself, characteristics such as severity, type, and duration directly dictate the magnitude of the financial burden. Severe or chronic conditions not only incur high direct medical costs but also lead to long-term labor loss and productivity decline, thereby acting as the primary stressor [6, 7]. Second, at the micro-level, individual and household endowments influence the internal capacity to resist these shocks. Individual factors like education, health literacy and lifestyle influence the proactive adoption of preventive behaviors [8, 9], while household-level factors (e.g., financial status, dependency ratio) may influence the immediate economic buffering capacity [9–11]. Third, at the macro-level, societal determinants such as social support networks, regional socioeconomic development, and health policy design constitute the external safety net that moderates the impact of shocks by providing risk-sharing mechanisms [12–15].
Furthermore, there is a growing consensus that relying solely on external protection is insufficient. Beyond external protection and support mechanisms, enhancing the endogenous coping capacity of individuals and households is critical for mitigating the economic risk of diseases, preventing cycles of economic hardship, and achieving sustainable economic security [13, 16]. This has shifted focus toward the resources and capabilities inherent within the household itself. However, family health resources—defined as all internal and external resources accessible to households that promote member health and well-being [17–19]—have received limited attention as a potential key determinant. Besides, previous studies have generally examined the role of family resources from a single dimension—such as financial or social capital [20, 21]—or have focused exclusively on observable coping behaviors, such as the utilization of medical services [22], to evaluate family strategies for managing risks. However, empirical evidence on how the overall stock and structural composition of these resources influence economic risk, as well as the specific pathways involved, remains limited. To sum up, a critical gap exists in understanding this process: what are the precise mechanisms through which diversified family health resources translate into the capability to predict a lower economic risk of disease.
According to the Conservation of Resources (COR) theory, this study further proposes a mechanism hypothesis regarding the functional pathways through which family health resources influence the economic risk of diseases. The COR theory posits that individuals and families are motivated to obtain, retain, and protect valuable resources to mitigate stress from potential threats [23, 24]. Furthermore, the Resource Caravan principle posits that resources travel together in aggregated sets suggesting that the aggregate level and configuration of the resources are critical for their capacity to cope with threats [25]. This principle provides the theoretical foundation for our use of Latent Profile Analysis (LPA) to identify distinct structural configurations of family health resources. Secondly, the theory posits that individuals and families proactively invest current resources to acquire additional resources and reduce the probability of future resource loss. Third, the Loss Spiral principle states that resource-depleted individuals and families are disproportionately vulnerable to further resource loss, making them dependent on external resource supplementation to halt the downward cycle.
Guided by the COR theory, we propose that households engage in a series of behaviors to manage their resources to maintain a reasonable amount and structure, collectively termed “Resource Management Functions”: (1) The Accumulation Function: building a resource surplus (e.g., assets) to buffer against future uncertainty, reflecting the COR emphasis on resource preservation and surplus maintenance; (2) The Investment Function: allocating resources to enhance capital, thereby reducing the probability of shocks, derived from the Resource Investment principle; and (3) The Compensation Function: mobilizing reserve resources to offset losses after a shock occurs, preventing a “loss spiral”. We hypothesize that the effective operation of these integrated resource management functions serves as a key potential mediating pathway linking static family health resources to dynamic resilience against economic risk.
Therefore, this study aims to empirically investigate the association and mechanisms linking family health resources to the economic risk of diseases among rural households in China. Specifically, the objectives of this study are to: (1) evaluate the status of family health resources in rural China in terms of quantity and structure; (2) verify the direct associations of the quantity and structure of family health resources with a lower economic risk of diseases; (3) examine the mediating role of resource management functions (accumulation, investment, and compensation) in this relationship; and (4) identify the specific pathways through which different dimensions of resources operate. By elucidating these mechanisms, this study expects to provide evidence-based insights for enhancing household-level risk management capabilities in developing contexts.
Methods
Data source and sample selection
The study sample was drawn from the 2020 wave of the China Family Panel Studies (CFPS), which was conducted by the Institute of Social Science Survey (ISSS) at Peking University. The CFPS employed a multi-stage probability sampling design and collected information at three levels: individual, household, and community. The individual-level dataset contains a wide range of variables, including basic demographic information (age, gender, marital status, household size, etc.), education, employment and income, social relationships, financial assistance, and health-related indicators (such as self-rated physical and mental health, evaluated health status, health insurance coverage, health expenditure, and utilization of health services).
In this study, the sample was initially drawn from 11,620 households included in the CFPS family economic dataset. The financial respondent (resplpid) was used as the matching key to merge relevant variables from the individual-level dataset. After conducting preliminary data cleaning and variable alignment across the two datasets, a merged sample comprising 10,322 households was obtained. Subsequently, 5,732 non-rural households were excluded based on household registration information, resulting in a subsample of 4,590 rural households. Further exclusion was applied to 342 households with missing values in key variables such as age, gender, and marital status. The final analytic sample consisted of 4,248 rural households.The following figure (Fig. 1) presents the flowchart for sample screening.
Fig. 1.
Flowchart of the sample selection process
Variables and definitions
Dependent variable
The economic risk of diseases in this study is defined as the probability that a household will fall into economic distress in the future due to health shocks. Unlike static ex-post measures, this concept represents a forward-looking perspective, serving as a comprehensive prediction of future risks based on multiple household indicators, which aligns with the concept of poverty vulnerability. Therefore, the economic risk of diseases in this study was measured using the three-stage Feasible Generalized Least Squares (FGLS) method proposed by Chaudhuri. The specific formula is as follows:
 |
2-1 |
In this equation, 
represents the expected health-related poverty vulnerability, which reflects the probability that a household will fall into or return to poverty due to illness. Ф(•) denotes the cumulative distribution function of the standard normal distribution. The terms
and 
correspond to the expected value and the expected variance of the logarithm of per capita household income, respectively. The poverty line (Z) is defined based on the widely recognized World Bank threshold of $1.90 per person per day. After adjusting for the exchange rate and inflation, this corresponds to an annual per capita income of ¥4,903.74 in 2020. The calculated risk index ranges from 0 to 1, indicating the probability that a household will experience illness-induced poverty in the future. To categorize households for descriptive purposes, we adopted a risk threshold of 30%. A value equal to or above 0.3 suggests that the household faces a 30% or greater risk of falling into poverty and is therefore classified as vulnerable households, reflecting limited current capacity for health-related poverty prevention. Conversely, a value below 0.3 designates non-vulnerable households. To preserve the full variance of the data and maximize statistical power, the continuous probability score of the economic risk of diseases (ranging from 0 to 1) was employed as the dependent variable in all subsequent regression models and path analyses.
Independent and mediating variables
The independent variable in this study was family health resources, which—guided by the Conservation of Resources (COR) Theory and existing literature—were categorized into seven dimensions: human health resources, financial health resources, physical health resources, informational health resources, healthcare resources, social network, and policy-related health resources (Table 1). Human health resources focused on the stock of health capital within the household, encompassing both quantity and quality. Following prior studies, the quality dimension was reflected by the number of household members with chronic diseases and those with poor self-rated health, while the quantity dimension was captured by household size as a proxy for potential caregiving capacity [26–28]. Financial health resources emphasized liquidity available for maintaining and improving health and were measured using total household income and per capita income [29]. Physical health resources referred to tangible assets related to health promotion and disease prevention, proxied by household expenditure on health-related products. Informational health resources captured access to health knowledge, measured by whether any household members used mobile devices to obtain online health information—an indicator increasingly relevant under the worldwide shift from traditional to digital health information channels [28, 30, 31]. Healthcare resources were assessed using perceived quality of medical services, which reflects the extent to which regional health systems meet residents’ needs [32]. Social network resources referred to external social ties and were measured by neighborhood support, indicating the availability of community-level social capital during times of need [33]. Lastly, policy-related health resources were measured by whether household members were covered under the basic health insurance scheme, given its essential role in China’s social protection system [34, 35].
Table 1.
Measurement Indicators of Family Health Resources and Resource Management Functions among Rural Households
| Variable Dimension | Measurement Indicator | Weight | Indicator Description |
|---|---|---|---|
| Human Health Resources | Household Size | 0.912 | Total number of family members (persons) |
| Number of Chronic Disease Patients in the Household | 0.033 | The number of household members with chronic diseases, derived from the item “Has anyone in your household had a chronic disease in the past six months?”; this was reverse-coded for analysis. | |
| Number of Family Members with Poor Self-Rated Health | 0.055 | The number of household members with poor self-rated health, based on the question “How would you rate your own health status?”; this was reverse-coded in the analysis. | |
| Financial Health Resources | Total Household Income | 0.492 | Total annual net household income (CNY), log-transformed for analysis. |
| Per Capita Household Income | 0.508 | Per capita household income (CNY), calculated as total annual household income divided by household size, also log-transformed. | |
| Physical Health Resources | Health Product Expenditure | 1 | Expenditure on health-related products (CNY), based on the item “How much did your household spend on health-related products in the past 12 months?”, including items such as fitness equipment and health supplements. |
| Informational Health Resources | Access to Online Information | 1 | Access to potential health information, proxied by whether household members used mobile devices to obtain health information. |
| Healthcare Resources | Perceived Quality of Medical Services | 1 | Perceived quality of medical services, derived from the item “How would you rate the level of medical care at this facility?”; rated on a 1–5 scale, with higher scores indicating better quality. |
| Social Network | Neighborhood Support | 1 | Based on the question “If you need help, would your neighbors assist you?” (1–5 scale; higher = more support) |
| Policy-Related Health Resources | Basic Health Insurance Coverage | 1 | Based on “Which types of medical insurance do your family members have?”; measured by number of members covered |
| Resource Accumulation Function | Asset Accumulation Capacity | 1 | Total value of accumulated household assets (CNY), log-transformed for analysis. |
| Resource Investment Function | Frequency of Physical Activity | 1 | Frequency of participation in physical and recreational exercise activities, measured on a scale from 0 to 7, with higher scores indicating higher frequency. |
| Resource Compensation Function | Share of Formal Transfer Income | 1 | Share of formal transfer income, calculated as the ratio of total government subsidies, pensions, social security payments, and minimum living allowance to total annual household income. |
The mediating variable in this study was the resource management functions. The functions were conceptualized to include three core components: the resource accumulation function, the investment function, and the compensation function (Table 1). The resource accumulation function primarily captured the household’s ability to accumulate wealth over time, directly reflecting the COR emphasis on resource preservation and surplus maintenance as a buffer against future threats. When coping with health shocks, families often prioritize the use of liquid assets and only resort to selling fixed assets or restructuring property when liquid resources are insufficient [36, 37]. Therefore, the total value of accumulated household assets was used to measure this function, as it reflects both the level of accumulated health resources and the household’s actual capacity to withstand health-related financial risks. The investment function focused on proactive health-promoting behaviors, derived from the COR resource investment principle—the idea that individuals and families invest current resources to enhance future capital stocks. Health-related behaviors constitute an investment in health capital, with returns realized as extended healthy life expectancy and sustained labor productivity. Prior studies have shown that the frequency of physical exercise can indicate the level of health investment [38, 39]. Accordingly, this study used physical activity frequency as the indicator for the investment function. The compensation function referred to external support obtained by the household to offset the economic burden of illness. This component was grounded in the COR theoretical corollary of resource loss and substitution: when primary resources are depleted, individuals will attempt to substitute alternative resources to fill the void. While some studies have emphasized the importance of informal income transfers and caregiving assistance, due to data availability, this study adopted formal transfer income as a proxy for compensation. Formal transfer payments (e.g., social insurance, government subsidies) were conceptualized as institutionalized compensation mechanisms designed to smooth consumption and prevent health shocks from triggering a poverty trap. This indicator reflects the extent of financial support households received from government and social security systems [40].
Covariates
In this study, demographic characteristics were included as covariates to control for potential confounding effects on the main analytical models. Specifically, age was grouped into three categories (1 = ≤ 60 years, 2 = 61–75 years, 3 = > 75 years), and gender was coded as 1 = male and 2 = female. Marital status was categorized as 1 = married and 2 = divorced/widowed/never married. Educational attainment was divided into three levels: 1 = primary school or below, 2 = junior high school, and 3 = senior high school or above. The household dependency ratio was classified into three levels based on the ratio of non-working to working household members: 1 = 0, 2 = 0–0.7, and 3 = ≥ 0.7. These covariates were included in the regression models to account for basic socio-demographic differences among households.
Statistical analysis
The entropy weight method was used to assign weights to indicators of family health resources and resource management functions, while latent profile analysis (LPA) was employed to classify family health resources into different structural types. The entropy weight method calculated indicator weights across dimensions, enabling the aggregation of standardized indicators into a composite quantity of family health resources, where higher scores indicated better overall resource levels. LPA was then conducted to identify group-level patterns in health resource distribution. Model fit was assessed using the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and entropy. In general, lower AIC and BIC values indicate better model fit, while entropy values closer to 1 suggest greater classification accuracy. An entropy value above 0.8 is typically considered indicative of reliable classification.
Based on the above, this study employed descriptive statistics and regression analysis to examine the distribution of family health resources among rural households and their role in managing economic risk, and to preliminarily explore the association between health resources and the economic risk of diseases. For categorical variables, frequency and percentage were used to describe household characteristics, family health resources, resource management functions, and the economic risk of diseases. For continuous variables, mean ± standard deviation was reported. Group differences in continuous variables were assessed using independent samples t-tests, while chi-square tests were used for categorical variables. Further, multivariate linear regression models were constructed with the economic risk of diseases as the dependent variable, family health resources and resource management functions as the key independent variables, and household characteristics as control variables, to assess the associations of the explanatory variables with the economic risk of diseases.
This study further applied path analysis to systematically examine the mechanism through which family health resources are associated with a lower economic risk of diseases. Specifically, the analysis focused on the potential mediating effect of resource management functions in the relationship between family health resources and the economic risk of diseases, aiming to reveal how health resources are linked to risk outcomes through functional pathways. The mediation path analysis was conducted using the SPSS PROCESS macro.
 |
2-2 |
 |
2-3 |
 |
2-4 |
 |
2-5 |
 |
2-6 |
Equations (2–2) presented the baseline regression model used to estimate the total effect of the quantity of family health resources (
) on the economic risk of diseases (
), where 
represents the health resource score of households 𝑖. Building on this, Eqs. (2–3), (2–4), and (2–5) introduced three types of resource management functions as mediating variables—namely, the resource accumulation function (
), the investment function (
), and the compensation function (
)—to examine how HRi is associated with each functional pathway. Equations (2–6) incorporated all three mediators simultaneously to assess their potential mediating roles in the relationship between 
and 
. The significance of the indirect effects (
, 
, 
) was tested using the bootstrap method.Bootstrap sampling was set as 5000 resamples. If the 95% confidence interval of an indirect effect excluded zero, the mediating effect was considered statistically significant. When the indirect effect and total effect were in the same direction, it was identified as a mediation effect, and the proportion of the indirect effect in the total effect was reported. If they were in opposite directions, the result was classified as a suppression effect, and the absolute ratio of the indirect effect to the direct effect was reported.
To further explore the mechanisms through which different structural types of family health resources are associated with the economic risk of diseases, dummy variables were created for the structural types and incorporated into the path analysis. The bootstrap method was used to test the significance of relative mediation effects. A 95% confidence interval that did not include zero indicated that the relative mediation effect was statistically significant [41]. When the direction of the relative mediation effect was consistent with the total effect, its proportion in the total effect was reported. If the directions were opposite, a suppression effect was indicated, and the absolute ratio of the indirect effect to the direct effect was reported. In this study, latent profile analysis was conducted using the tidyLPA package in R 4.4.2, while all other statistical analyses were performed using Stata 16.0 and SPSS 26.0.
Ethical announcement
All methods used in this study were implemented in accordance with relevant guidelines and regulations of the China Family Panel Studies (CFPS). All participants joined CFPS voluntarily and signed a consent form before participation. The CFPS project was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-14010). Since this study involved the analysis of de-identified secondary data, no additional ethical review was required.
Results
Based on entropy-weighted scores for each dimension of family health resources, this study employed latent profile analysis (LPA) to identify the structural types of family health resources. Model fit results are presented in Table 2. The findings showed that when the number of latent classes was set to three, the entropy value was close to 1 and markedly higher than those of the four-class and five-class models. Additionally, the three-class model exhibited relatively lower AIC and BIC values and significantly higher minimum posterior classification probabilities, indicating better model fit, stronger class separation, and greater reliability. Therefore, the structure of family health resources among rural households was categorized into three types.
Table 2.
Model Fit Indices for Latent Profiles of Family Health Resources in Rural Households
| Type | AIC | BIC | Entropy | Minimum Classification Probability | Maximum Classification Probability |
|---|---|---|---|---|---|
| 1 | -21668.565 | -21445.145 | 1.0000 | 1.0000 | 1.0000 |
| 2 | -32962.995 | -32688.507 | 0.9999 | 1.0000 | 1.0000 |
| 3 | -37444.506 | -37118.951 | 0.9997 | 0.9997 | 1.0000 |
| 4 | -38125.688 | -37749.066 | 0.9839 | 0.6686 | 0.9998 |
| 5 | -38159.473 | -37731.783 | 0.9025 | 0.4066 | 0.9997 |
Based on the characteristics illustrated in Fig. 2, the three structural types were named accordingly. As scores for human health resources and policy-related health resources were below 0.5 across all types and showed no significant intergroup differences, they were not used as the basis for naming. Type 1 (black dashed line) was characterized by particularly low scores in informational health resources and physical health resources, with an overall lower resource level compared to the other two groups, and was thus labeled the Overall Resource-Disadvantaged Type. Type 2 (black solid line) had scores above 0.5 across most dimensions with relatively balanced resource distribution, and was labeled the Resource-Balanced Advantaged Type. Type 3 (gray solid line) showed the highest score in informational health resources, and was thus labeled the Information Resource-Advantaged Type.
Fig. 2.
Latent Profile Analysis–Based Typology of Family Health Resource Structures
Table 3 presents the distribution of household demographic characteristics, health resources, and resource management functions between high-risk and low-risk households. Based on the international poverty line standard (USD 1.90 per person per day) and the high-risk threshold of 0.3 for the economic risk of diseases, the results showed that a total of 400 households (9.42% of the sample) were identified as high-risk households (risk ≥ 0.3). This indicates that although these households were not currently in poverty, they had a 30% probability of falling into economic distress due to health shocks within the next year, placing them in a category of potentially high economic risk of diseases requiring targeted policy attention and intervention. Significant differences in demographic characteristics were observed between high-risk and low-risk households (P < 0.001). Specifically, households with members aged 61 or older, female household heads, Divorced/Widowed/Never Married status, Not Employed, Primary School or Below education level, and a household dependency ratio of ≥ 0.7 were more likely to exhibit high economic risk of diseases.
Table 3.
Distribution of demographic characteristics, family health resources, and resource management functions among rural households
| Variables | Category | Total Sample, N (%) | Low-Risk Households, N (%) | High-Risk Household, N (%) | χ2/t | P |
|---|---|---|---|---|---|---|
| Age | ≤ 60 years | 3152 (74.20) | 3082 (80.09) | 70 (17.50) | 888.199 | < 0.001 |
| 61–75 years | 966 (22.74) | 713 (18.53) | 253 (63.25) | |||
| > 75 years | 130 (3.06) | 53 (1.38) | 77 (19.25) | |||
| Gender | Male | 2569 (60.48) | 2360 (61.33) | 209 (52.25) | 12.499 | < 0.001 |
| Female | 1679 (39.52) | 1488 (38.67) | 191 (47.75) | |||
| Marital Status | Divorced/ Widowed/ Never Married | 644 (15.16) | 518 (13.46) | 126 (31.50) | 91.666 | < 0.001 |
| Married | 3604 (84.84) | 3330 (86.54) | 274 (68.50) | |||
| Employment Status | Not Employed | 670 (15.77) | 516 (13.41) | 154 (38.50) | 171.704 | < 0.001 |
| Employed | 3578 (84.23) | 3332 (86.59) | 246 (61.50) | |||
| Educational Attainment | Primary School or Below | 2074 (48.82) | 1731 (44.98) | 343 (85.75) | 243.454 | < 0.001 |
| Junior High School | 1446 (34.04) | 1398 (36.33) | 48 (12.00) | |||
| Senior High School or Above | 728 (17.14) | 719 (18.69) | 9 (2.25) | |||
| Household Dependency Ratio | 0 | 2115 (49.79) | 2075 (53.92) | 40 (10.00) | 1290.625 | < 0.001 |
| 0-0.7 | 1581 (37.22) | 1502 (39.03) | 79 (19.75) | |||
| ≥ 0.7 | 552 (12.99) | 271 (7.05) | 281 (70.25) | |||
| Quantity of Family Health Resources | 3.34 ± 0.73 | 3.41 ± 0.71 | 2.67 ± 0.57 | 24.252 | < 0.001 | |
| Structural Types of Family Health Resources | Overall Resource-Disadvantaged Type | 1868 (43.97) | 1529 (39.73) | 339 (84.75) | 302.277 | < 0.001 |
| Information Resource-Advantaged Type | 1874 (44.11) | 1838 (47.77) | 36 (9.00) | |||
| Resource-Balanced Advantaged Type | 506 (11.92) | 481 (12.50) | 25 (6.25) | |||
| Resource Accumulation Function | 0.45 ± 0.32 | 0.47 ± 0.32 | 0.27 ± 0.29 | 12.366 | < 0.001 | |
| Resource Investment Function | 0.15 ± 0.30 | 0.15 ± 0.30 | 0.13 ± 0.31 | 1.073 | 0.283 | |
| Resource Compensation Function | 0.13 ± 0.24 | 0.10 ± 0.21 | 0.39 ± 0.34 | -16.513 | < 0.001 |
Significant differences were also found in the distribution of family health resources and resource management functions between the two groups. Households with lower overall levels of family health resources, classified as the Overall Resource-Disadvantaged Type, with lower resource accumulation and investment function levels, but higher resource compensation function levels, were more likely to exhibit high economic risk of diseases.
By controlling for covariates, this study further examined the association between family health resources and the economic risk of diseases (Table 4). The results showed that an increase in the quantity of family health resources was significantly associated with a lower economic risk of diseases. Regarding the structural types of health resources, compared to households classified as the Overall Resource-Disadvantaged Type, those in the Information-Advantaged Type and Balanced-Advantaged Type had significantly lower economic risk of diseases. In terms of resource management functions, both the resource accumulation function and the resource investment function were negatively associated with the economic risk of diseases—higher levels of these functions were linked to lower risks. Conversely, the resource compensation function was positively associated with the economic risk of diseases, indicating that stronger reliance on compensatory support was linked to higher economic risk of diseases.
Table 4.
Regression analysis of the association between family health resources and the economic risk of diseases in rural households
| Variables | model 1 | model 2 | model 3 | model 4 |
|---|---|---|---|---|
| Age | 0.071*** | 0.024*** | 0.058*** | 0.011** |
| (13.35) | (5.21) | (11.02) | (2.41) | |
| Gender | 0.014*** | 0.006 | 0.015*** | 0.007* |
| (3.16) | (1.59) | (3.46) | (1.89) | |
| Marital Status | -0.045*** | -0.046*** | -0.040*** | -0.039*** |
| (-7.63) | (-8.99) | (-6.91) | (-7.90) | |
| Employment Status | -0.033*** | -0.028*** | -0.023*** | -0.018*** |
| (-5.41) | (-5.23) | (-3.90) | (-3.51) | |
| Educational Attainment | -0.023*** | -0.026*** | -0.018*** | -0.019*** |
| (-7.91) | (-9.94) | (-6.09) | (-7.53) | |
| Household Dependency Ratio | 0.072*** | 0.018*** | 0.066*** | 0.013*** |
| (19.81) | (5.65) | (18.56) | (4.21) | |
| Quantity of Family Health Resources | -0.034*** | -0.026*** | ||
| (-10.54) | (-8.28) | |||
| Structural Types of Family Health Resources | ||||
| (ref- Overall Resource-Disadvantaged Type) | ||||
| Information Resource-Advantaged Type | -0.049*** | -0.043*** | ||
| (-11.18) | (-10.19) | |||
| Resource-Balanced Advantaged Type | -0.048*** | -0.040*** | ||
| (-7.72) | (-6.57) | |||
| Resource Accumulation Function | -0.062*** | -0.062*** | ||
| (-9.58) | (-11.05) | |||
| Resource Investment Function | -0.017** | -0.023*** | ||
| (-2.48) | (-3.80) | |||
| Resource Compensation Function | 0.112*** | 0.107*** | ||
| (11.93) | (13.12) | |||
| _cons | 0.061*** | 0.118*** | 0.054*** | 0.131*** |
| (3.52) | (9.22) | (3.17) | (10.47) | |
| R2 | 0.37 | 0.187 | 0.404 | 0.243 |
| Adj. R2 | 0.37 | 0.19 | 0.4 | 0.24 |
*** indicates P < 0.001; ** indicates P < 0.01; * indicates P < 0.05. Standard errors were shown in parentheses
As shown in Fig. 3(A), the quantity of family health resources was significantly and positively associated with both the resource accumulation function and the resource investment function (P < 0.001), and also showed a direct negative association with the economic risk of diseases(P < 0.001). In addition, health resource quantity was negatively associated with the resource compensation function (P < 0.001), indicating that households with greater health resources relied less on compensatory mechanisms. In Fig. 3(B), compared to the Overall Resource-Disadvantaged Type, both the Balanced-Advantaged Type and the Information-Advantaged Type households not only showed a direct negative association with the economic risk of diseases (P < 0.01), but also demonstrated an indirect negative association with the economic risk of diseases through higher levels of their accumulation and investment functions. The findings further revealed that, relative to the Overall Resource-Disadvantaged Type, households of the Information-Advantaged Type exhibited lower dependence on the compensation function, which was further linked to a lower economic risk of diseases.
Fig. 3.
Path coefficients diagram of the mechanism linking family health resources to the economic risk of diseases in rural households
Table 5 presents the results of the mediation analysis examining the indirect pathways linking family health resources to the economic risk of diseases among rural households. The findings indicate that in the association between a greater quantity of family health resources and a lower economic risk of diseases, the resource accumulation function and resource investment function played positive potential mediating roles, while the resource compensation function served as a negative mediator. Specifically, the indirect effect of the accumulation pathway was β = − 0.018 (95% CI excluding zero), accounting for 12.492% of the total effect; the mediation effect through the investment function was also significant, contributing 2.580% of the total effect. These results suggest that health resource quantity is associated with a lower economic risk of diseases through higher levels of accumulation and investment functions. In contrast, the indirect effect through the compensation pathway was β = − 0.011 (95% CI excluding zero), accounting for 7.196% of the total effect, indicating that greater health resource quantity is linked to a lower economic risk of diseases through a reduced reliance on compensatory mechanisms—thus, the compensation function exerts a negative mediating effect.
Table 5.
Mediating Effects in the Pathways Linking Family Health Resources to The Economic Risk of Diseases in Rural Households
| Path | Standardized Coefficient / Effect Value | Standard Error | Lower Limit of Confidence Interval | Upper Limit of Confidence Interval | Proportion of Indirect Effect (%) |
|---|---|---|---|---|---|
| Quantity of Family Health Resources | |||||
| Family Health Resource Quantity →Accumulation →The Economic Risk of Diseases | -0.018 | 0.003 | -0.024 | -0.013 | 12.492 |
| Family Health Resource Quantity → Investment →The Economic Risk of Diseases | -0.004 | 0.002 | -0.007 | -0.000 | 2.580 |
| Family Health Resource Quantity →Compensation →The Economic Risk of Diseases | -0.011 | 0.003 | -0.017 | -0.005 | 7.196 |
| Total Indirect Effect | -0.033 | 0.005 | -0.042 | -0.024 | 22.267 |
| Structural Types of Family Health Resources | |||||
| Resource-Balanced Advantaged Type → Accumulation →The Economic Risk of Diseases | -0.047 | 0.008 | -0.063 | -0.033 | 29.774 |
| Resource-Balanced Advantaged Type → Investment →The Economic Risk of Diseases | -0.015 | 0.006 | -0.027 | -0.005 | 9.548 |
| Resource-Balanced Advantaged Type → Compensation →The Economic Risk of Diseases | 0.013 | 0.009 | -0.003 | 0.031 | n.sa |
| Relative Direct Effect (Resource-Balanced Advantaged Type) | -0.110 | 0.042 | -0.192 | -0.028 | – |
| Relative Total Effect (Resource-Balanced Advantaged Type) | -0.159 | 0.043 | -0.243 | -0.076 | – |
| Information Resource-Advantaged Type → Accumulation →The Economic Risk of Diseases | -0.022 | 0.005 | -0.032 | -0.012 | 12.169 |
| Information Resource-Advantaged Type → Investment →The Economic Risk of Diseases | -0.004 | 0.002 | -0.008 | -0.001 | 2.028 |
| Information Resource-Advantaged Type → Compensation →The Economic Risk of Diseases | -0.011 | 0.006 | -0.023 | -0.001 | 6.197 |
| Relative Direct Effect (Information Resource-Advantaged Type) | -0.141 | 0.029 | -0.199 | -0.084 | – |
| Relative Total Effect (Information Resource-Advantaged Type) | -0.178 | 0.030 | -0.237 | -0.119 | – |
“a” indicates the presence of a suppression effect in the path, but the indirect effect was not statistically significant (95% CI includes zero)
Furthermore, the accumulation function, investment function, and compensation function also served as mediators in the relationship between health resource structural types and the economic risk of diseases. Compared with households in the Overall Resource-Disadvantaged Type, those in the Resource-Balanced Advantage Type showed statistically significant relative mediation effects through the accumulation function (β = − 0.047) and investment function (β = − 0.015), accounting for 29.774% and 9.548% of the total effect, respectively. For households classified as the Information-Dominant Type, the relative mediation effects through the accumulation (β = − 0.022), investment (β = − 0.004), and compensation functions (βe = − 0.011) accounted for 12.169%, 2.028%, and 6.197% of the total effect, all of which were statistically significant. These results indicate that structural advantages in family health resources are associated with a lower economic risk of diseases through higher levels of accumulation and investment functions and a reduced dependence on the compensation function.
Discussion
This study confirmed that both the aggregate quantity and structural configuration of family health resources were significantly associated with lower economic risk of diseases among rural households. Households with greater resource quantity and those with a more balanced structure exhibited significantly lower economic risk. Path analysis further indicated that these resources were linked to risk outcomes through optimizing resource management functions, specifically through higher levels of accumulation and investment and a reduced reliance on compensation. The possible reasons for the above results are as follows. The accumulation and investment functions primarily operate as ex-ante active defense mechanisms. Meanwhile, the compensation function typically serves as an ex-post passive remedy. Structural advantages (such as information resources) act as catalysts that amplify the efficiency of active defenses. Households with balanced resource structures can more effectively activate accumulation and investment strategies.
The accumulation function primarily operates before a health shock occurs, helping to build resilience. The path analysis indicated that family health resources were negatively associated with economic risk through higher levels of the accumulation function. Based on existing health resources, rural households can leverage the accumulation function to build financial assets, thereby creating a “cushioning effect” that is associated with lower economic risk of diseases. Economic resources accumulated over time enhance a household’s ability to access necessary medical services during sudden health shocks, serving as a buffer against financial impact and helping to avert a vicious cycle of economic distress and disease. Before a health shock occurs, rural households with low or temporarily reduced incomes—especially those lacking access to formal credit—tend to accumulate resources and increase savings as a proactive strategy to guard against potential economic risk of diseases [42, 43].Following a health shock, households often experience labor loss, leading to declines in wage and business income. This typically forces them to adjust their consumption patterns and increase precautionary savings to manage future expenditure uncertainties. Although such adjustments often come at the expense of spending on education, recreation, and preventive care, which may compromise long-term development, they remain an effective defensive strategy for preserving liquidity and reducing immediate economic risk of diseases. This suggests that the core of risk defense lies in robust asset accumulation. Policy interventions should therefore focus on enhancing the household’s capacity for wealth accumulation to ensure households have a sufficient “material buffer” to withstand shocks.
The investment function also comes into play before the occurrence of health shocks. Our findings suggest that this function contributes to health risk reduction potentially through sustained health-promoting behaviors, which are expected to maintain health capital, support a stable labor supply, and thereby alleviate the economic burden associated with illness. Existing studies have shown that health investment behaviors—such as maintaining a nutritious diet, avoiding tobacco and excessive alcohol consumption, engaging in regular physical activity, and ensuring adequate sleep—can significantly improve blood pressure and glucose levels, prevent the onset of chronic diseases, or improve health outcomes for individuals already living with such conditions [44]. By safeguarding the physical capacity to work, sustained health capital helps households avoid the income disruptions that often follow health shocks, thereby reinforcing the stability of household finances when crises occur [45]. These findings suggest that initiatives aimed at fostering health-promoting behaviors could serve as a valuable complement to asset-building strategies, potentially strengthening rural households’ proactive defense against health-related economic shocks. In the Chinese context, recent policy efforts—such as the Healthy China 2030 initiative and the expansion of community-based health education programs—have begun to prioritize health literacy and preventive behaviors, providing an existing institutional platform through which such health-promoting interventions could be delivered.
The compensation function is typically mobilized after a shock as a safety net. The path analysis indicated that greater family health resources were associated with a reduced reliance on the compensation function, which in turn was linked to a lower economic risk of diseases. Households with richer and more balanced health resources tended to rely less on compensation, because their internal buffers reduced the need for emergency external assistance. For resource-depleted households, compensation functions such as relying on health insurance reimbursements to offset hospitalization costs, seeking medical assistance to cover out-of-pocket expenses [46, 47], or depending on support from relatives and friends for caregiving needs [48] are typically activated when households experience substantial depletion of health capital and lack sufficient internal resources to manage the crisis. This interpretation aligns with extensive evidence from China’s rural health insurance, which plays an indispensable role in preventing households from falling into poverty traps following health shocks. Therefore, the policy objective should not be to uniformly reduce reliance on compensation, but rather to optimize the functional mix: by strengthening accumulation and investment as ex-ante defenses, compensation can be preserved as a robust safety net for those who genuinely need it.
Structural configurations of family health resources matter for economic risk resilience. Compared to “Overall Resource-Disadvantaged” households, both “Resource-Balanced Advantage” and “Information Resource-Advantaged” households tended to exhibit lower levels of economic risk of diseases. For this balanced group, the mechanism likely operates through the absence of critical shortages: a balanced and rational allocation of family health resources to avoid the “bucket effect” caused by critical resource shortages that may weaken the overall capacity for preventing risks. Existing literature has established that structural imbalances in external health resources, such as insufficient human resources and pharmaceutical shortages, directly impede continuity of care and exacerbate financial burdens [49]. Our study reveals a parallel mechanism at the micro-household level, indicating that internal structural imbalances are equally detrimental. Even if a household possesses certain assets, a “weakest link” in other dimensions may prevent the effective utilization of these assets.
Furthermore, informational health resources, in particular, function as a lever within the resource system. Adequate information enhances household members’ understanding of disease prevention, treatment options, and reimbursement policies, thereby improving the allocative efficiency of limited health resources [50]. Conversely, a deficit in information may hinder engagement with external social network resources and limit the acquisition of critical support. This informational asymmetry undermines rational health-related decision-making [51] and financial planning, ultimately constraining the household’s potential for maintaining long-term economic resilience [31]. In policy terms, this suggests a dual-track approach—strengthening foundational health assets across multiple dimensions for all households, while simultaneously targeting health information deficits, which may disproportionately constrain the resource-poor. China’s current emphasis on cultivating grassroots health literacy is precisely oriented toward strengthening informational resources for rural households.
Despite its contributions, this study has several limitations that warrant acknowledgment.
First, regarding the research design, the reliance on cross-sectional data restricts our ability to establish strict causal sequencing. Although we controlled for key demographic covariates, the reported path coefficients may be subject to omitted variable bias (e.g., unobserved wealth endowments or health history) and must be interpreted as conditional associations rather than strict causal effects. Furthermore, the positive association between the compensation function and economic risk is susceptible to reverse causality, as China’s means-tested transfer programs typically target households already facing elevated risk. Future panel-data studies employing instrumental variable or fixed-effects approaches would be valuable to further address this concern and disentangle these temporal dynamics.
Second, utilizing a general-purpose national survey necessitated the use of proxy indicators with inherent constraints. Specifically, the investment function was proxied solely by physical activity, capturing only a fraction of multidimensional health behaviors. The compensation function, measured by formal transfer income, inevitably excludes non-monetary informal support (e.g., caregiving) and simultaneously captures both institutional compensation and underlying socioeconomic disadvantage. Furthermore, measuring the accumulation function via total assets captures a static stock rather than a dynamic process. Finally, the inability to fully control for unobserved confounders—such as regional fixed effects and specific occupational types—may introduce omitted-variable bias. Future research utilizing specialized surveys should incorporate comprehensive multi-indicator measures to further validate the pathways identified in this study.
Conclusion
In conclusion, when family health resources are sufficient and structurally balanced, they are directly and indirectly associated with a lower economic risk of diseases through resource management functions. Family health resources are linked to a lower economic risk of diseases through higher levels of the resource accumulation and investment functions, alongside a reduced reliance on the compensation function. Establishing long-term risk resilience requires shifting governance focus from resource infusion to capacity enhancement. Policy efforts should focus on enhancing households’ capacity to optimize resource configuration through informational empowerment, strengthening their ability to accumulate assets as a material buffer, and fostering their capability for proactive health investment. Ultimately, this facilitates a transition for rural households from passive risk coping to active risk management.
Acknowledgements
JW is funded by National Science Foundation of China (grant numbers 72074086，72574074). We thank the China Family Panel Studies team for providing data and training in using the datasets.
Authors’ contributions
Conceptualization: Y.Z., S.L., J.W.; Methodology: Y.Z., S.L., J.W., D.H., Y.A.; Formal analysis: Y.Z., S.L., S.G.; Writing – original draft: Y.Z., S.L.; Writing – review & editing: Y.Z., S.L., J.W., D.H., S.G., X.D., Y.A., X.L., Y.Y.; Data curation: S.L., S.G., X.L.; Software: D.H., X.D.; Validation: D.H., X.D.; Supervision: J.W.; Project administration: J.W., Y.Y.; Funding acquisition: J.W.; Investigation: S.G.; Visualization: X.D.; Resources: X.L. Y.Z. and S.L. contributed equally to this work and share first authorship. All authors read and approved the final manuscript.
Funding
The research was supported by:
(1) Research on Optimization Strategies for Medical Service Supply Response in Remote Rural Areas Based on Digital Ecological Dilemma Governance (grant numbers 72574074);
(2) Research on dynamic optimization of coping strategies on health poverty risk for rural older households (grant numbers 72074086)
Data availability
The datasets generated and analysed during the current study are available in the China Family Panel Studies (CFPS) repository, https://cfpsdata.pku.edu.cn/.
Declarations
Ethics approval and consent to participate
The China Family Panel Studies (CFPS) project was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-14010).As this study involved secondary analysis of publicly available de-identified data, no additional ethical review was required. (Clinical trial number: not applicable.)
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yixuan Zhong and Sangsang Li contributed equally to this work.
References
- 1.Chen S, Kuhn M, Prettner K, Yu F, Yang T, Bärnighausen T, et al. The global economic burden of chronic obstructive pulmonary disease for 204 countries and territories in 2020-50: a health-augmented macroeconomic modelling study. Lancet Glob Health. 2023;11:e1183–93. 10.1016/S2214-109X(23)00217-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Luengo-Fernandez R, Walli-Attaei M, Gray A, Torbica A, Maggioni AP, Huculeci R, et al. Economic burden of cardiovascular diseases in the European Union: a population-based cost study. Eur Heart J. 2023;44:4752–67. 10.1093/eurheartj/ehad583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Arredondo A. Universal coverage and economic burden from epidemiological changes of diabetes in Latin America. J Glob Health. 2016;6:020309. 10.7189/jogh.06.020309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hex N, Bartlett C, Wright D, Taylor M, Varley D. Estimating the current and future costs of Type 1 and Type 2 diabetes in the UK, including direct health costs and indirect societal and productivity costs. Diabet Med. 2012;29:855–62. 10.1111/j.1464-5491.2012.03698.x. [DOI] [PubMed] [Google Scholar]
- 5.Zhao X, Wang X, Ren Z, Zhang C. The impact of health shocks on poverty vulnerability: evidence from rural households in China. Front Public Health. 2025;13:1–12. 10.3389/fpubh.2025.1657269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bernatsky S, Panopalis P, Pineau CA, Hudson M, St Pierre Y, Clarke AE. Healthcare Costs of Inflammatory Myopathies. J Rheumatol. 2011;38:885–8. 10.3899/jrheum.101083. [DOI] [PubMed] [Google Scholar]
- 7.Bassi A, Dodd S, Williamson P, Bodger K. Cost of illness of inflammatory bowel disease in the UK: a single centre retrospective study. Gut. 2004;53:1471–8. 10.1136/gut.2004.041616. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ervin JL, Milner A, Kavanagh AM, King TL. The double burden of poverty and marital loss on the mental health of older Australian women; a longitudinal regression analysis using 17 annual waves of the HILDA cohort. Soc Psychiatry Psychiatr Epidemiol. 2021;56:1059–68. 10.1007/s00127-020-02019-z. [DOI] [PubMed] [Google Scholar]
- 9.Liao P, Zhang X, Zhang W. Endogenous health risks, poverty traps, and the roles of health insurance in poverty alleviation. Health Econ Rev. 2022;12:25. 10.1186/s13561-022-00370-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang W, Chen K, Xiao W, Du J, Qiao H. Determinants of health poverty vulnerability in rural areas of Western China in the post-poverty relief era: an analysis based on the Anderson behavioral model. BMC Public Health. 2024;24:459. 10.1186/s12889-024-18035-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Nguyen TT, Do TL, Halkos G, Wilson C. Health shocks and natural resource extraction: A Cambodian case study. Ecol Econ. 2020;169:106517. 10.1016/j.ecolecon.2019.106517. [DOI] [Google Scholar]
- 12.Li Y, Marquez R. Can government subsidies and public mechanisms alleviate the physical and mental health vulnerability of China’s urban and rural residents? Int J Equity Health. 2023;22:59. 10.1186/s12939-022-01805-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cui Z, Li E, Li Y, Deng Q, Shahtahmassebi A. The impact of poverty alleviation policies on rural economic resilience in impoverished areas: A case study of Lankao County, China. J Rural Stud. 2023;99:92–106. 10.1016/j.jrurstud.2023.03.007. [DOI] [Google Scholar]
- 14.Zhang Y, Zhao W. Social capital’s role in mitigating economic vulnerability: Understanding the impact of income disparities on farmers’ livelihoods. World Dev. 2024;177. 10.1016/j.worlddev.2023.106515. [DOI]
- 15.Zhuang X, Lau YY, Chan WMH, Lee BSC, Wong DFK. Risk and resilience of vulnerable families in Hong Kong under the impact of COVID-19: an ecological resilience perspective. Soc Psychiatry Psychiatr Epidemiol. 2021;56:2311–22. 10.1007/s00127-021-02117-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Liu W, Xu J, Li J, Li S. Rural Households’ Poverty and Relocation and Settlement: Evidence from Western China. Int J Environ Res Public Health. 2019;16. 10.3390/ijerph16142609. [DOI] [PMC free article] [PubMed]
- 17.Rotegård AK, Moore SM, Fagermoen MS, Ruland CM. Health assets: a concept analysis. Int J Nurs Stud. 2010;47:513–25. 10.1016/j.ijnurstu.2009.09.005. [DOI] [PubMed] [Google Scholar]
- 18.Sacks E, Morrow M, Story WT, Shelley KD, Shanklin D, Rahimtoola M, et al. Beyond the building blocks: integrating community roles into health systems frameworks to achieve health for all. BMJ Glob Health. 2018;3. 10.1136/bmjgh-2018-001384. [DOI] [PMC free article] [PubMed]
- 19.Olpin E, Hanson CL, Crandall A. Influence of Social Media Uses and Gratifications on Family Health among U.S. Parents: A Cross-Sectional Study. Int J Environ Res Public Health. 2023;20:1910. 10.3390/ijerph20031910. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ünver Ş, Alkan Ö, Çelik AK, Abar H. Determining factors affecting regional poverty levels in Türkiye: a comparison of ordered discrete choice models. J Health Popul Nutr. 2025;44:37. 10.1186/s41043-024-00722-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ma X, Hu Q, He J, Li C, Chen K, Wang W, et al. Association between sanitary toilets and health poverty vulnerability among rural western Chinese adults aged 45 years and older: A cross-sectional study. PLoS ONE. 2024;19:e0308688. 10.1371/journal.pone.0308688. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Yu Q, Zhu C, Feng S, Xu L, Hu S, Chen H, et al. Economic Burden and Health Care Access for Patients With Inflammatory Bowel Diseases in China: Web-Based Survey Study. J Med Internet Res. 2021;23. 10.2196/20629. [DOI] [PMC free article] [PubMed]
- 23.Hobfoll SE, Halbesleben J, Neveu J-P, Westman M. Conservation of Resources in the Organizational Context: The Reality of Resources and Their Consequences. In: Morgeson F, editor. Palo Alto: Annual Reviews. 2018;5:103–28. 10.1146/annurev-orgpsych-032117-104640 [DOI]
- 24.Tibbetts RH, Lutter SB. Associations of Health and Financial Resources With Stress: Applying the Theory of Conservation of Resources. J Financial Couns Plann. 2018;29:60–7. 10.1891/1052-3073.29.1.60. [DOI] [Google Scholar]
- 25.Wang A, Tang C, Zhou L, Lv H, Song J, Chen Z, et al. How surface acting affects turnover intention among family doctors in rural China: the mediating role of emotional exhaustion and the moderating role of occupational commitment. Hum Resour Health. 2023;21. 10.1186/s12960-023-00791-y. [DOI] [PMC free article] [PubMed]
- 26.Jiao X, Wang X, Zhang J. Household expenditures on education and the relative poverty in China. Finance Res Lett. 2025;72:106476. 10.1016/j.frl.2024.106476. [DOI] [Google Scholar]
- 27.Zhou Z, Zhao Y, Shen C, Lai S, Nawaz R, Gao J. Evaluating the effect of hierarchical medical system on health seeking behavior: A difference-in-differences analysis in China. Soc Sci Med. 2021;268:113372. 10.1016/j.socscimed.2020.113372. [DOI] [PubMed] [Google Scholar]
- 28.Zhou Y, Bai Y, Wang J. The impact of internet use on health among older adults in China: a nationally representative study. BMC Public Health. 2024;24:1065. 10.1186/s12889-024-18269-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wu Y, Guo Z, Qi D. Analysis of the effect of social capital on rural household poverty. Soc Policy Adm. 2023;57:333–48. 10.1111/spol.12864. [DOI] [Google Scholar]
- 30.Ding H, Zhang C, Xiong W. Associations between Mobile Internet Use and Self-Rated and Mental Health of the Chinese Population: Evidence from China Family Panel Studies 2020. Behav Sci (Basel). 2022;12:221. 10.3390/bs12070221. [DOI] [PMC free article] [PubMed]
- 31.Wei X, Liu Y, Liu Y. Study on the Impact of Internet Usage, Aging on Farm Household Income. Sustainability. 2023;15. 10.3390/su151914324. [DOI]
- 32.Hu H, Jian W, Fu H, Zhang H, Pan J, Yip W. Health service underutilization and its associated factors for chronic diseases patients in poverty-stricken areas in China: a multilevel analysis. BMC Health Serv Res. 2021;21:707. 10.1186/s12913-021-06725-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cao D, Zhou Z, Liu G, Shen C, Ren Y, Zhao D, et al. Does social capital buffer or exacerbate mental health inequality? Evidence from the China Family Panel Study (CFPS). Int J Equity Health. 2022;21:75. 10.1186/s12939-022-01642-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Wang N, Xu J, Ma M, Shan L, Jiao M, Xia Q, et al. Targeting vulnerable groups of health poverty alleviation in rural China- what is the role of the New Rural Cooperative Medical Scheme for the middle age and elderly population? Int J Equity Health. 2020;19:161. 10.1186/s12939-020-01236-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ma M, Li Y, Wang N, Wu Q, Shan L, Jiao M, et al. Does the medical insurance system really achieved the effect of poverty alleviation for the middle-aged and elderly people in China? Characteristics of vulnerable groups and failure links. BMC Public Health. 2020;20:435. 10.1186/s12889-020-08554-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Kolesar RJ, Erreygers G, Van Damme W, Chea V, Choeurng T, Leng S. Hardship financing, productivity loss, and the economic cost of illness and injury in Cambodia. Int J Equity Health. 2023;22. 10.1186/s12939-023-02016-z. [DOI] [PMC free article] [PubMed]
- 37.Thomas AR, Dash U, Sahu SK. Illnesses and hardship financing in India: an evaluation of inpatient and outpatient cases, 2014-18. BMC Public Health. 2023;23:204. 10.1186/s12889-023-15062-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Moss JL, Pinto CN, Shen C. Prevalence of cancer risk behaviors by county-level persistent poverty. Cancer Epidemiol. 2025;94. 10.1016/j.canep.2024.102735. [DOI] [PubMed]
- 39.Aue K, Roosen J, Jensen HH. Poverty dynamics in Germany: Evidence on the relationship between persistent poverty and health behavior. Soc Sci Med. 2016;153:62–70. 10.1016/j.socscimed.2016.01.040. [DOI] [PubMed] [Google Scholar]
- 40.Liu J, Wang C, Wang J, Yang C. Decomposition of the impact of government and private transfers on income inequality in China. Rev Dev Econ. 2025;29:203–25. 10.1111/rode.13086. [DOI] [Google Scholar]
- 41.Hayes AF, Preacher KJ. Statistical mediation analysis with a multicategorical independent variable. Br J Math Stat Psychol. 2014;67:451–70. 10.1111/bmsp.12028. [DOI] [PubMed] [Google Scholar]
- 42.Habib SS, Perveen S, Khuwaja HMA. The role of micro health insurance in providing financial risk protection in developing countries- a systematic review. BMC Public Health. 2016;16. 10.1186/s12889-016-2937-9. [DOI] [PMC free article] [PubMed]
- 43.Do MH. The Role of Savings and Income Diversification in Households’ Resilience Strategies: Evidence from Rural Vietnam. Soc Indic Res. 2023;168:353–88. 10.1007/s11205-023-03141-6. [DOI] [Google Scholar]
- 44.Huang Y, Wang S, Tian L, Zhang X, Liu S, Zhu Z, et al. Healthy lifestyle habits, educational attainment, and the risk of 45 age-related health and mortality outcomes in the UK: A prospective cohort study. J Nutr Health Aging. 2025;29. 10.1016/j.jnha.2025.100525. [DOI] [PMC free article] [PubMed]
- 45.Zheng L, Hou W, Huang M. Impact of health investment on household income distribution: insights from China’s longitudinal survey data. Front Public Health. 2024;12. 10.3389/fpubh.2024.1346133. [DOI] [PMC free article] [PubMed]
- 46.Fu Y, Chen M, Si L. Multimorbidity and catastrophic health expenditure among patients with diabetes in China: a nationwide population-based study. BMJ Glob Health. 2022;7:e007714. 10.1136/bmjgh-2021-007714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Fu X-Z. The comparison of catastrophic health expenditure and its inequality between urban and rural households in China. Health Econ Rev. 2022;12:19. 10.1186/s13561-022-00365-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Jacobs JC, Lo J, Van Houtven CH, Wagner TH. The impact of informal caregiving on U.S. Veterans Health Administration utilization and expenditures. Soc Sci Med. 2024;344:116625. 10.1016/j.socscimed.2024.116625. [DOI] [PubMed] [Google Scholar]
- 49.Tang S, Yao L, Ye C, Li Z, Yuan J, Tang K, et al. Can health service equity alleviate the health expenditure poverty of Chinese patients? Evidence from the CFPS and China health statistics yearbook. BMC Health Serv Res. 2021;21. 10.1186/s12913-021-06675-y. [DOI] [PMC free article] [PubMed]
- 50.Chen K, Qiu J, Wang W, Hu Q, Xu N, Qiao H. Identification and analysis of influencing factors of multidimensional health poverty in rural areas of Northwest China. Sci Rep. 2024;14. 10.1038/s41598-024-80628-3. [DOI] [PMC free article] [PubMed]
- 51.Huang H-T, Kuo Y-M, Wang S-R, Wang C-F, Tsai C-H. Structural Factors Affecting Health Examination Behavioral Intention. Int J Environ Res Public Health. 2016;13. 10.3390/ijerph13040395. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and analysed during the current study are available in the China Family Panel Studies (CFPS) repository, https://cfpsdata.pku.edu.cn/.