Abstract
Background
Patients undergoing maintenance hemodialysis require long-term complex treatment and strict self-management, and nonadherence to treatment is relatively common. Meaning in life, social support, and loneliness may be jointly associated with treatment adherence; however, the complex network of associations among these factors remains unclear.
Methods
In this cross-sectional study, 481 patients with end-stage renal disease undergoing maintenance hemodialysis were recruited using convenience sampling from the hemodialysis centers of three hospitals in a city in China. Data were collected between December 2024 and February 2025 using the Treatment Adherence Scale for Patients with End-Stage Renal Disease Undergoing Maintenance Hemodialysis, the Meaning in Life Questionnaire, the Perceived Social Support Scale, and the Loneliness Scale. Network analysis was used to construct a partial correlation network comprising eight nodes. Strength centrality and expected influence were used to identify relatively central nodes, bridge expected influence was applied to analyze the strength of cross-community associations, and network accuracy and stability were examined.
Results
In the network, treatment adherence showed positive partial correlations with presence of meaning (edge weight = 0.213) and family support (0.103), and negative partial correlations with emotional loneliness (−0.099) and social loneliness (−0.093). Social loneliness and significant other support were among the nodes with the highest strength centrality, while support from friends had the highest expected influence. Presence of meaning had the highest positive bridge expected influence, while social loneliness had the highest negative bridge expected influence. Strength centrality and expected influence showed good stability, with correlation stability coefficients of 0.595 for both indices.
Conclusion
The findings suggest that social loneliness, support from friends and significant others, and presence of meaning may be important nodes in the psychological network related to treatment adherence. These findings may inform the selection of candidate targets for future nursing intervention research.
Keywords: treatment adherence, meaning in life, social support, loneliness, network analysis
Introduction
End-stage kidney disease (ESKD) is a chronic condition that has become a serious public health challenge.1 The global number of patients receiving kidney replacement therapy has been increasing rapidly and is projected to reach 5.4 million by 2030.2 In China, maintenance hemodialysis (MHD) has become the predominant kidney replacement therapy for patients with ESKD.3 By the end of 2022, the number of dialysis patients in China had exceeded 850,000, making China the country with the largest dialysis population worldwide.4
However, maintenance hemodialysis is not a definitive curative treatment.5 Although it can prolong survival, patients must maintain good self-management and treatment adherence throughout the course of treatment to maintain clinical stability.6 Treatment adherence refers to the extent to which a patient’s behavior corresponds with the recommendations agreed upon with a healthcare provider.7 In the clinical management of maintenance hemodialysis, treatment adherence typically includes regular attendance at dialysis sessions, appropriate medication use, fluid intake restriction, and dietary restrictions.8 Notably, non-adherence is highly prevalent among patients undergoing maintenance hemodialysis.8,9 In China, non-adherence among maintenance hemodialysis patients is particularly common, with its prevalence reported to be as high as 70.8%.10 Long-term non-adherence behaviors may not only lead to inadequate dialysis but also increase the risk of poor blood pressure control, electrolyte disturbances, and cardiovascular complications.6,9,11 Given this clinical challenge, researchers have increasingly recognized that psychological distress and depletion of psychological resources associated with long-term dialysis may weaken patients’ willingness to engage in self-management and further reduce treatment adherence.12,13
Factors Associated with Treatment Adherence
Conservation of Resources (COR) theory provides a strong theoretical foundation for explaining factors associated with treatment adherence among patients undergoing maintenance hemodialysis.14 In chronically stressful contexts, such as long-term illness, ongoing treatment burden, and lifestyle restrictions, individuals’ psychological resources may be continuously depleted, thereby affecting their coping behaviors and ability to maintain health-related behaviors.13,14 Meaning in life may provide patients with an internal source of meaning and purpose when facing ongoing treatment,15 whereas social support reflects the practical assistance and emotional support that patients perceive as available from others.16 Meanwhile, loneliness reflects the subjective experience of unmet needs for social connection and emotional closeness,17 offering a complementary perspective on perceived inadequacies in patients’ interpersonal resources. Accordingly, this study included meaning in life, social support, and loneliness alongside treatment adherence in a single network to examine the interrelationships among personal meaning-related resources, interpersonal support, and unmet relational needs, as well as their associations with treatment adherence.
Resource-Protective Factors
Meaning in life refers to the extent to which individuals understand the significance of their lives and perceive their lives as having goals, values, and purpose.18 It serves as an important internal psychological resource that helps patients undergoing maintenance hemodialysis cope with the burden of long-term illness and the pressure of continuous treatment, while maintaining a sense of purpose, value, and self-continuity in the context of illness.19,20 Meaning in life comprises two dimensions: the presence of meaning and the search for meaning.21 Presence of meaning refers to the extent to which individuals perceive their lives as purposeful, valuable, and meaningful, whereas the search for meaning refers to the psychological tendency to actively explore and construct meaning in life.21 A strong sense of meaning may enable patients to find reasons to persevere throughout the painful treatment process, allowing them to face illness with a more positive attitude and thereby maintain greater treatment adherence.8,13,22 Search for meaning, in contrast, reflects a motivational state in which patients attempt to regain meaning-related resources and reconstruct life goals following illness-related disruption.18 A moderate level of searching for meaning may encourage patients to actively explore new lifestyles and strategies for adapting to illness.22 However, if patients remain in a prolonged state of searching without achieving a sense of meaning, the search for meaning may be associated with psychological resource depletion, thereby weakening their capacity to sustain treatment adherence and self-management.13,14,19 In addition, meaning in life is closely related to social support.23 Adequate support from family members, friends, and peers can enhance patients’ meaning in life.24,25
Social support refers to the instrumental assistance and emotional comfort that individuals obtain through their social networks.16 It is an important external resource that buffers stress and promotes treatment adherence. Social support is commonly conceptualized as comprising three dimensions: support from family, friends, and significant others.16 Family support represents the most fundamental resource for patients undergoing maintenance hemodialysis.26 Family members’ involvement in daily care, dietary supervision, and medication management can help reduce practical barriers to adherence.27 Friend support provides patients with emotional companionship, while support from fellow patients or peers can help them share illness-related experiences and alleviate the sense of social isolation caused by long-term dialysis.28,29 Significant other support refers to support perceived by respondents as being provided by individuals whom they personally regard as important.16 In the context of maintenance hemodialysis, healthcare professionals and fellow patients may constitute important sources of such support because patients engage in recurrent and structured interactions with them throughout ongoing treatment. Healthcare professionals may provide professional information, behavioral guidance, and emotional support, which are particularly relevant to the treatment demands of maintenance hemodialysis.5,30 In particular, healthcare professionals can provide patients with professional medical information and behavioral guidance.5 A good patient-provider relationship may enhance patients’ understanding, acceptance, and adherence to complex dialysis regimens.30 This narrowing of social support sources may reduce patients’ external resource reserves for coping with the stress of long-term dialysis, thereby increasing psychological distress, loneliness, and the risk of treatment non-adherence.13
Resource-Depleting Factors
Loneliness is a distressing subjective experience that arises when individuals perceive a discrepancy between their desired and actual social relationships, it comprises both social loneliness and emotional loneliness.31 Within the framework of COR theory, loneliness can be understood as a state of psychological resource depletion that may be associated with lower treatment adherence.14,32 Social loneliness stems from reduced social network resources. Patients undergoing maintenance hemodialysis may be forced to reduce work participation and routine social activities because of illness- and treatment-related constraints.33 As a result, they may lose health-related supervision and positive feedback from peers and the workplace, which in turn may weaken their treatment adherence and self-management behaviors.24,33 Emotional loneliness, in contrast, results from a lack of emotional support. Patients who lack close supportive relationships are more likely to experience helplessness and feelings of abandonment, which may reduce their willingness to continue treatment.13,34 In addition, loneliness may also be associated with resource-protective factors, such as social support and meaning in life.22,25 Adequate social support can provide these patients with emotional connection and a sense of belonging, helping to buffer the psychological stress caused by treatment burden and reduce loneliness in the context of illness.5,28 Conversely, patients undergoing maintenance hemodialysis who experience high levels of loneliness are more likely to develop negative perceptions of life goals and self-worth, which may lead to hopelessness about the future, perceived social isolation, and a belief that their existence lacks meaning.13,25,35
Application of Network Analysis in Treatment Adherence Research
Most existing studies have adopted traditional linear statistical approaches.25,26,32 Although such approaches are useful for identifying factors associated with treatment adherence, network analysis can provide a complementary perspective on the patterns of associations among the psychological variables themselves. In this context, network analysis has emerged as a valuable research method in nursing and psychology.36 Within the network analysis framework, variables are represented as nodes, and their associations are represented as edges, enabling the construction of network models that illustrate patterns of connections among variables. In a partial-correlation network, these connections represent pairwise associations conditional on the other included variables.37 Network analysis can identify central or highly connected variables within the network through centrality indices, helping identify candidate targets for future intervention research.38 In addition, although cross-sectional network analysis remains correlational in nature, its findings can still generate valuable theoretical hypotheses and inform model construction in subsequent causal research.36 From a clinical perspective, these findings may provide preliminary evidence to help prioritize candidate targets for future intervention studies aimed at improving treatment adherence.
Current Study
Further clarification is needed regarding how the distinct dimensions of meaning in life, social support, and loneliness are conditionally associated with one another and with treatment adherence in patients undergoing maintenance hemodialysis. In summary, guided by COR theory, the present study employed network analysis to examine resource-protective factors (presence of meaning, search for meaning, family support, friend support, and significant-other support), resource-depleting factors (social loneliness and emotional loneliness), and treatment adherence within a unified network framework. With the aim of generating hypotheses for future research, the study used an exploratory network analysis to: (1) construct a psychological network of factors related to treatment adherence among patients undergoing maintenance hemodialysis and clarify the patterns of interaction among these factors; (2) identify central protective and risk-related variables within the network, thereby determining key nodes in the system; and (3) identify the dimensions that serve as bridges connecting different theoretical domains, including social support, loneliness, meaning in life, and treatment adherence.
Methods
Participants and Procedure
Patients undergoing maintenance hemodialysis were recruited using convenience sampling from the hemodialysis centers of three hospitals in a city in China between December 2024 and February 2025. The inclusion criteria were as follows: participants were (1) aged 18 years or older; (2) had been clearly diagnosed with end-stage renal disease according to the Chinese Guidelines for the Diagnosis and Prevention of Chronic Kidney Disease (2017 Edition) and had received regular maintenance hemodialysis for at least 3 months; (3) conscious and able to comprehend and communicate adequately; and (4) willing to provide informed consent and participate voluntarily. The exclusion criteria were as follows: (1) severe or unstable diseases involving major organs, such as the heart, brain, liver, or lungs, that prevented completion of the questionnaire survey; (2) severe visual or hearing impairment that prevented questionnaire completion; and (3) a confirmed diagnosis of a mental disorder or cognitive impairment. Eligibility was assessed according to the inclusion and exclusion criteria before questionnaire administration. All 481 eligible patients invited to participate completed the questionnaire survey, and all were included in the final analysis. The participants had a mean age of 60.70 years (SD = 13.09); 348 were men (72.3%) and 133 were women (27.7%). Descriptive characteristics of the 481 participants are presented in Supplementary File S1.
This study was reviewed and approved by the Ethics Committee of the School of Psychology, Jiangxi Normal University (approval number: HSR-24-000712). Permission for patient recruitment and questionnaire administration was obtained from the heads of the hemodialysis centers at all three participating hospitals. Before data collection, participants were informed of the study’s purpose and procedures and provided informed consent. A trained nurse with relevant psychological knowledge conducted the survey during routine treatment hours, with no time limit for questionnaire completion. Participation was voluntary, and participants could withdraw at any time. Responses were kept confidential and were accessible only to the research team.
Measures
Treatment Adherence
Treatment adherence was assessed using the Treatment Adherence Scale for Patients with End-Stage Renal Disease Undergoing Maintenance Hemodialysis, which has demonstrated good reliability and validity in the Chinese context.39 The scale consisted of 23 items rated on a 5-point Likert scale. Participants were asked to respond to each item using a scale ranging from 1 (never) to 5 (always), with higher total scores indicating greater treatment adherence. In the present study, Cronbach’s α for this scale was 0.861.
Meaning in Life
Meaning in life was assessed using the Meaning in Life Questionnaire (MLQ), which was revised by Chinese scholars and has been validated as having good reliability and validity in the Chinese context.15,21 The questionnaire consists of 10 items and includes two subscales: presence of meaning and search for meaning, with each subscale comprising five items. Participants were asked to rate each item on a 7-point scale ranging from 1 (absolutely untrue) to 7 (absolutely true). Higher scores on each subscale indicate higher levels of the corresponding dimension. In the present study, Cronbach’s α was 0.883 for the total scale, 0.847 for the presence of meaning subscale, and 0.931 for the search for meaning subscale.
Social Support
Social support was assessed using the Chinese version of the Perceived Social Support Scale, which has been widely used and demonstrated good psychometric properties in the Chinese population.40 The scale consists of 12 items and includes three dimensions: family support, friend support, and significant other support, with each dimension comprising four items. Participants were asked to respond to each item on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). Higher scores indicate higher levels of perceived social support in the corresponding dimension. In the present study, Cronbach’s α was 0.874 for the total scale, 0.911 for the family support subscale, 0.914 for the friend support subscale, and 0.856 for the significant-other support subscale.
Loneliness
Loneliness was assessed using the Chinese version of the De Jong Gierveld Loneliness Scale, which has demonstrated good reliability and validity in Chinese populations.41 The scale consists of 11 items and includes two dimensions: social loneliness and emotional loneliness. The social loneliness dimension consists of seven items, and the emotional loneliness dimension consists of four items. Participants responded using a 3-point format with the options “no”, “more or less” and “yes”. Higher scores on each dimension indicate higher levels of the corresponding type of loneliness. In the present study, Cronbach’s α was 0.815 for the total scale, 0.854 for the social loneliness subscale, and 0.594 for the emotional loneliness subscale.
Statistical Analysis
First, SPSS 27.0 was used to assess common method bias and conduct correlation analyses. Second, the network analyses were conducted in R 4.4.2 using the packages bootnet, qgraph, mgm, and networktools. The specific analyses are described below.
Network Model Estimation
A partial correlation network was estimated using the R package bootnet. Specifically, the network was estimated using the EBIC graphical LASSO procedure, which applies the graphical least absolute shrinkage and selection operator to regularize the partial correlation matrix. Model selection was performed using the extended Bayesian information criterion (EBIC), with the tuning hyperparameter γ set to 0.5. This value is commonly recommended as it favors sparse network structures by reducing potentially spurious or weak edges, thereby improving the interpretability and robustness of the estimated network.37
In the resulting network, variables were represented as nodes, and the associations between nodes were represented as edges. Each edge indicates the partial correlation between two variables after controlling for all other variables in the network, with the magnitude of the partial correlation corresponding to the edge weight. The network was visualized using the R package qgraph.37
In addition, node predictability was estimated using the R package mgm. Predictability refers to the proportion of variance in a given node that can be explained by all of its neighboring nodes in the network. Thus, higher predictability values indicate that a larger proportion of the variance in a node is accounted for by its direct connections, providing an intuitive measure of how well each node can be predicted from the rest of the network.42
Centrality Estimation
To evaluate the relative importance of individual nodes within the network, four centrality indices were calculated using the qgraph package: strength, betweenness, closeness, and expected influence. Strength was defined as the sum of the absolute edge weights directly connected to a given node, reflecting the overall magnitude of its direct associations with other nodes. Betweenness centrality quantifies the extent to which a node lies on the shortest paths between pairs of other nodes. Closeness centrality was calculated as the reciprocal of the sum of the shortest path lengths from a given node to all other nodes.37 Expected influence was defined as the sum of the signed edge weights directly connected to a node. Unlike strength, expected influence retains the sign of associations and is therefore particularly informative in networks containing both positive and negative edges.43
Furthermore, all nodes were assigned to four theoretically defined communities: social support, loneliness, meaning in life, and treatment adherence. Bridge centrality indices were computed using the R package networktools to identify nodes that may serve as important connectors between different communities. Specifically, bridge strength, bridge closeness, bridge betweenness, and bridge expected influence were calculated.36 In the present study, particular attention was given to one-step and two-step bridge expected influence. One-step bridge expected influence refers to the sum of signed edge weights directly connecting a node to nodes belonging to other communities, thereby capturing its direct cross-community influence. Two-step bridge expected influence additionally incorporates indirect cross-community connections through one intermediary node, reflecting a node’s broader indirect influence across communities. Larger positive bridge expected influence indicates stronger positive cross-community associations, whereas larger negative values indicate stronger negative cross-community associations and suggest that the corresponding node may play a more prominent role in linking different theoretical domains.37
Accuracy and Stability Testing
The accuracy of edge-weight estimates and the stability of centrality indices were evaluated using the bootnet package. To assess edge-weight accuracy, nonparametric bootstrapping was performed to estimate 95% confidence intervals for edge weights. Narrower confidence intervals indicate greater precision of edge-weight estimates. In addition, bootstrap-based difference tests were conducted to examine whether nonzero edge weights and node centrality indices differed significantly from one another.37
To assess the stability of centrality indices, a case-dropping bootstrap procedure was used to calculate the correlation stability coefficient (CS) for strength and expected influence. The CS represents the maximum proportion of cases that can be removed while retaining, with 95% probability, a correlation of at least 0.70 between the centrality indices estimated from the subset data and those estimated from the full sample. In general, CS above 0.25 are considered acceptable, whereas values above 0.50 indicate good stability.37
Results
Preliminary Analyses
The results of Harman’s single-factor test showed that the first factor accounted for 17.81% of the total variance, which was below 40%. Descriptive statistics and correlation coefficients for the eight node scores are presented in Table 1, including means, standard deviations, observed minimum and maximum scores, skewness, and kurtosis. Across the eight node scores, skewness ranged from −1.002 to 0.982, and kurtosis ranged from −1.008 to 1.122. Most variables showed significant correlations with one another.
Table 1.
Descriptive Statistics and Correlation Analysis of the Variables
| M ± SD | min | max | Skewness | Kurtosis | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 FRI | 11.84 ± 7.70 | 4 | 28 | 0.451 | −0.999 | – | |||||||
| 2 FAM | 20.47 ± 6.34 | 4 | 28 | −1.002 | 0.196 | 0.188** | – | ||||||
| 3 SO | 16.67 ± 5.80 | 4 | 28 | −0.469 | −0.573 | 0.495** | 0.308** | – | |||||
| 4 SL | 14.18 ± 3.59 | 7 | 21 | −0.029 | −1.008 | −0.264** | −0.334** | −0.334** | – | ||||
| 5 EL | 8.34 ± 1.57 | 4 | 12 | −0.066 | 0.211 | −0.027 | −0.082 | −0.072 | 0.285** | – | |||
| 6 POM | 21.09 ± 6.37 | 5 | 35 | −0.052 | −0.359 | 0.169** | 0.292** | 0.218** | −0.330** | −0.151** | – | ||
| 7 SOM | 13.72 ± 7.64 | 5 | 35 | 0.982 | 0.070 | 0.331** | 0.086 | 0.158** | −0.263** | −0.023 | 0.384** | – | |
| 8 TA | 87.57 ± 12.10 | 23 | 115 | −0.608 | 1.122 | 0.151** | 0.239** | 0.199** | −0.276** | −0.211** | 0.376** | 0.110* | – |
Notes: N = 481. * p < 0.05, ** p < 0.01.
Abbreviations: FRI, friends support; FAM, family support; SO, significant other support; SL, social loneliness; EL, emotional loneliness; POM, presence of meaning; SOM, search for meaning; TA, treatment adherence.
Network Model Estimation
Figure 1 presents the estimated network structure of treatment adherence, meaning in life, loneliness, and social support in patients undergoing maintenance hemodialysis. In the overall network, node predictability ranged from 9.6% to 32.2%, with a mean of 24.75%, indicating that, on average, 24.75% of the variance in each node could be explained by its neighboring nodes. The nodes with the highest predictability were significant other support (32.2%), presence of meaning (31.3%), support from friends (31.0%), and social loneliness (29.4%), suggesting that a relatively large proportion of the variance in these nodes was explained by their neighboring nodes. In contrast, emotional loneliness showed the lowest predictability (9.6%), indicating that its variance was least accounted for by its neighboring nodes. The predictability values for all nodes are presented in Supplementary File S2.
Figure 1.
Psychological network of treatment adherence in patients undergoing maintenance hemodialysis.
Notes: Node colors indicate the predefined communities. Blue edges represent positive associations, whereas red edges represent negative associations. Thicker and darker edges indicate larger absolute edge weights.
Among the edges directly involving treatment adherence, the largest absolute estimated edge weight was observed for presence of meaning (0.213), indicating a positive partial correlation. Treatment adherence also showed a positive partial correlation with family support (0.103) and negative partial correlations with emotional loneliness (−0.099) and social loneliness (−0.093). These latter associations were relatively small in magnitude compared with the presence-of-meaning–adherence association. Across the network, 17 of the 28 possible edges were nonzero (60.71%). The largest positive estimated edge weight was observed between support from friends and significant other support (0.364), while the most negative estimated edge weight was observed between family support and social loneliness (−0.168). All edge weights are reported in Supplementary File S3.
Network Centrality
The results of the centrality analysis are shown in Figure 2. Social loneliness had the highest strength centrality (1.400) and closeness centrality (1.520), while significant other support had the second-highest strength centrality (0.839). Social loneliness and significant other support were tied for the highest betweenness centrality (1.180). In contrast, emotional loneliness had the lowest strength centrality (−1.960), closeness centrality (−2.010), and betweenness centrality (−1.510). These findings suggest that social loneliness and significant other support occupied relatively central positions in the network, whereas emotional loneliness was located at the periphery. With respect to expected influence, support from friends had the highest positive expected influence value (0.977), followed by significant other support (0.706), whereas social loneliness had the lowest expected influence value (−2.210). Overall, social loneliness showed the strongest negative expected influence and appeared to be one of the most central nodes in the network, whereas support from friends and significant other support showed relatively strong positive expected influence, suggesting their potential protective roles within the network. The results for all centrality indices are presented in Supplementary File S4.
Figure 2.
Centrality results of the network.
Notes: Centrality indices are presented as standardized z-scores.
Abbreviations: FRI, friends support; FAM, family support; SO, significant other support; SL, social loneliness; EL, emotional loneliness; POM, presence of meaning; SOM, search for meaning; TA, treatment adherence.
The results of the bridge centrality analysis are shown in Figure 3. Social loneliness had the highest bridge strength (1.758), while search for meaning had the highest bridge closeness centrality (0.820). Search for meaning and support from friends both had the highest bridge betweenness centrality (1.365). These findings suggest that social loneliness had the strongest direct connections with nodes in other communities, that search for meaning was positioned closest to nodes in other communities, and that search for meaning and support from friends may serve as intermediary nodes linking different communities. In terms of bridge expected influence, presence of meaning showed the highest positive values for both 1-step bridge expected influence (0.785) and 2-step bridge expected influence (0.693), whereas social loneliness showed the most negative values for both 1-step bridge expected influence (−2.285) and 2-step bridge expected influence (−2.309). These results suggest that, in cross-community associations, presence of meaning was associated with the strongest positive bridge expected influence, whereas social loneliness was associated with the most negative bridge expected influence and may represent a key risk node connecting different communities. The results for all bridge centrality indices are presented in Supplementary File S5.
Figure 3.
Bridge centrality results of the network.
Note: Bridge centrality indices are presented as standardized z-scores.
Abbreviations: FRI, friends support; FAM, family support; SO, significant other support; SL, social loneliness; EL, emotional loneliness; POM, presence of meaning; SOM, search for meaning; TA, treatment adherence.
Network Accuracy and Stability
As shown in Figure 4, the bootstrap 95% confidence intervals for the edge weights were relatively narrow, and the bootstrap means were generally close to the original sample estimates, suggesting good accuracy of edge-weight estimation. Bootstrap difference tests identified statistically significant differences between some edge weights, whereas other comparisons were not statistically significant.37
Figure 4.
Bootstrap confidence intervals and pairwise difference tests for edge weights.
Notes: Left panel: Red points indicate sample estimates, black points indicate bootstrap means, and gray areas show 95% bootstrap confidence intervals. Right panel: Black and gray cells indicate significant and nonsignificant differences between edge weights, respectively. Diagonal colors represent sample edge weights (blue, positive; red, negative).
Abbreviations: FRI, friends support; FAM, family support; SO, significant other support; SL, social loneliness; EL, emotional loneliness; POM, presence of meaning; SOM, search for meaning; TA, treatment adherence.
The results of the stability analysis are shown in Figure 5. Strength centrality and expected influence demonstrated good stability, with CS coefficients of 0.595 for both indices. This indicates that, even after dropping up to approximately 59.5% of the cases, the re-estimated strength centrality and expected influence values remained correlated with the original estimates at 0.70 or higher with 95% probability, suggesting good stability of these centrality measures.
Figure 5.
Stability results for strength and expected influence.
Abbreviations: FRI, friends support; FAM, family support; SO, significant other support; SL, social loneliness; EL, emotional loneliness; POM, presence of meaning; SOM, search for meaning; TA, treatment adherence.
Discussion
Based on the COR theory, this study explored the complex interrelationships among treatment adherence, social support, loneliness, and meaning in life in patients undergoing maintenance hemodialysis using a network analysis approach. The findings not only revealed complex associations among the variables but also identified central nodes and cross-community bridge nodes within the treatment-adherence-related network. This study provides a multidimensional perspective for understanding treatment adherence among patients undergoing maintenance hemodialysis and may inform the development of hypotheses for future intervention research.
Social Loneliness: A Key Network Node in Psychological Resource Depletion
Social loneliness was among the nodes with the highest strength centrality and also had the lowest expected influence and bridge expected influence values. These findings indicate that social loneliness occupied a relatively central position in the estimated network. Its association with treatment adherence was reflected in a negative partial correlation between the two variables. From the perspective of the COR theory, long-term dialysis treatment may limit patients’ participation in their previous occupational and social circles, leading to a substantial reduction in social network resources, which may in turn contribute to psychological defensiveness and withdrawal behaviors.14,44 The high centrality of social loneliness suggests that this variable occupies a central position among multiple negative psychological responses. Moreover, its strongly negative expected influence indicates that increased social loneliness may be associated with broader negative activation across the network. It may be associated with weaker meaning in life and poorer treatment adherence, and its network position may be relevant to the mobilization and utilization of social support resources.45,46 Notably, compared with social loneliness, emotional loneliness occupied a relatively peripheral position in the network. This may suggest that, for patients undergoing maintenance hemodialysis who are required to adhere strictly to dietary restrictions and regular dialysis schedules, a lack of external monitoring from social networks and peer feedback may be more strongly associated with poorer treatment adherence behaviors than subjective emotional deprivation alone.31
Social Support: An Important Network Resource in Psychological Resource Protection
With regard to protective resources, the findings highlighted the important roles of significant other support and support from friends in the network. First, significant other support showed relatively high strength centrality in the overall network, suggesting that significant others may serve as a protective hub in patients’ psychological network. For patients, significant other support may include healthcare professionals, who can provide authoritative medical guidance and emotional support, help patients gain a sense of disease control, and buffer negative emotional responses.5,30 But this interpretation is specific to the treatment context and does not redefine the significant-other dimension as provider-specific. Second, support from friends exhibited the highest positive expected influence in the entire network, indicating that friend support may be broadly associated with positive activation across the network and may serve as a core bridge connecting different network communities. In addition, edge weights showed that the strongest positive connection was observed between support from friends and significant other support. This finding suggests that, within the external resource pool of patients undergoing maintenance hemodialysis, support from friends and significant other support may be closely linked and may operate in a complementary manner. In clinical contexts, healthcare professionals may provide authoritative guidance, whereas peers or fellow patients may offer experiential sharing and emotional resonance.28 The combination of these two forms of support may be associated with stronger positive psychological resources and better treatment adherence.29 The network also showed the strongest negative association between family support and social loneliness, suggesting that among patients experiencing illness-related social withdrawal, adequate family support may help buffer social loneliness and serve as a basic safeguard against psychological resource depletion.26
Meaning in Life: A Core Bridge Between Psychological Resources and Treatment Adherence
Meaning in life plays an important bridging role between psychological resources and treatment adherence. First, the presence of meaning showed not only the strongest positive cross-community connection with treatment adherence but also the highest positive first-order and second-order bridge expected influence. This indicates that having clear life goals may be an important factor that supports patients in maintaining a high level of treatment adherence.8 When patients possess this meaning-related resource, they may reinterpret the distressing experience of dialysis treatment as a necessary part of pursuing their life goals, thereby demonstrating better adherence.47 Second, the search for meaning showed the highest bridge closeness centrality and bridge betweenness centrality in the network, suggesting that it may occupy a structurally important position in facilitating connections across network communities. When patients lose their original sense of meaning due to the impact of illness, a strong motivation to search for meaning may prompt them to actively seek new sources of meaning and external support.48 For patients undergoing maintenance hemodialysis, the search for meaning may not necessarily reflect psychological resource depletion.23 Instead, when a relatively well-established social support system is available, the search for meaning may be associated with connections between external support resources and the reconstruction of personal meaning, with possible relevance to loneliness and treatment adherence. These possibilities should be examined in future longitudinal or intervention studies.19,49,50
Clinical Implications
The findings highlight psychological domains that hospitals may consider when planning supportive nursing care for patients undergoing maintenance hemodialysis. First, hospitals may consider assessing patients’ social loneliness and perceived social connectedness during nursing consultations and organizing voluntary peer-support groups for sharing treatment experiences. Second, hospitals may explore combining professional guidance with support from trained peer volunteers through patient education and regular communication. Finally, hospitals may consider incorporating discussions of personal values and life goals into psychological nursing care, using approaches such as narrative nursing and life review where appropriate.
Limitations and Future Directions
Although this study provided valuable insights through network analysis, several limitations should be acknowledged. First, the cross-sectional design limited our ability to examine temporal dynamics and causal relationships within the psychological network related to treatment adherence among patients undergoing maintenance hemodialysis. Although this study is guided by COR theory, its theoretical interpretations should not be taken as evidence of causal effects. Future studies could employ longitudinal dynamic network analysis across multiple time points to explore changes in network structure and temporal associations, while intervention studies could test whether modifying these factors improves treatment adherence. Second, this study used self-report measures to assess patients’ psychological characteristics and treatment adherence during the same session, which may have introduced response bias, social desirability bias, recall bias, and common method bias. Future research could incorporate objective clinical indicators and medical records to conduct more comprehensive assessments and further validate the present findings. Furthermore, the emotional loneliness subscale showed low internal consistency in the present sample. This may be related to factors such as subscale length and scoring procedures, and the associated measurement error may have affected the estimated network associations, centrality, and bridge centrality. Findings concerning emotional loneliness, including comparisons with social loneliness, should therefore be interpreted cautiously. In addition, the network model in this study did not include clinical factors that may be associated with treatment adherence, such as dialysis vintage, comorbidities, economic burden, educational level, and marital status.51 Future studies could construct integrated network models incorporating both clinical and psychological variables. Finally, treatment adherence was not further categorized in this study, and the treatment adherence community consisted of only one node. This may have obscured the heterogeneity of different adherence behaviors; therefore, the results of bridge centrality indices should be interpreted with caution. Treatment adherence among patients undergoing maintenance hemodialysis can be divided into medication adherence, dietary adherence, fluid restriction adherence, and dialysis regimen adherence.39 These four dimensions may be associated with distinct psychological characteristics. Future research may consider this distinction and further investigate psychological networks related to different types of treatment adherence in this population.
Conclusion
Treatment adherence among patients undergoing maintenance hemodialysis was associated with a complex network of multiple psychological resources. Social loneliness may represent an important risk-related node associated with reduced network stability and lower treatment adherence, whereas significant other support and friends may represent candidate protective resources. Presence of meaning occupied a central bridging position between psychological resources and treatment adherence. Because this was a cross-sectional network analysis, these findings are correlational and hypothesis-generating rather than evidence of validated causal intervention targets. Future studies could evaluate interventions addressing social loneliness, strengthening support from healthcare professionals and peers, and maintaining or reconstructing meaning in life.
Acknowledgments
We would like to express our sincere gratitude to all the researchers and participants who contributed to this study.
Funding Statement
The authors received no funding for this study.
Data Sharing Statement
The datasets generated and/or analysed during the current study are not publicly available due to ethical restrictions. However, they are available from the corresponding author upon reasonable request. The analytical code for this study is available in Supplementary File S6.
Ethics Approval and Consent to Participate
This study was conducted in accordance with the Declaration of Helsinki. The study was approved by the Institutional Review Board of Jiangxi Normal University. The approval number is HSR-24-000712. Participation was voluntary, and written informed consent was obtained from all participants.
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.Bikbov B, Purcell CA, Levey AS, et al. Global, regional, and national burden of chronic kidney disease, 1990-2017. A systematic analysis for the Global burden of disease study 2017. Lancet. 2020;395(10225):709–14. doi: 10.1016/S0140-6736(20)30045-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Levin A, Tonelli M, Bonventre J, et al. Global kidney health 2017 and beyond: a roadmap for closing gaps in care, research, and policy. Lancet. 2017;390(10105):1888–1917. doi: 10.1016/S0140-6736(17)30788-2 [DOI] [PubMed] [Google Scholar]
- 3.Zhang H, Zhang C, Zhu S, Ye H, Zhang D. Direct medical costs of end-stage kidney disease and renal replacement therapy: a cohort study in Guangzhou city, southern China. BMC Health Serv Res. 2020;20(1):1–9. doi: 10.1186/s12913-020-4960-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wei T, Liang YX, Zhao XJ; Chinese Hospital Association Blood Purification Center Branch. Research report on the development of China’s hemodialysis. Chin J Blood Purif. 2024;23(j5):321–329. doi: 10.3969/j.issn.1671-4091.2024.05.001 [DOI] [Google Scholar]
- 5.Himmelfarb J, Vanholder R, Mehrotra R, Tonelli M. The current and future landscape of dialysis. Nat Rev Nephrol. 2020;16(10):573–585. doi: 10.1038/s41581-020-0315-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Daugirdas JT, Depner TA, Inrig J, et al. KDOQI clinical practice guideline for hemodialysis adequacy: 2015 update. Am J Kidney Dis. 2015;66(5):884–930. doi: 10.1053/j.ajkd.2015.07.015 [DOI] [PubMed] [Google Scholar]
- 7.Burkhart PV, Sabaté E. Adherence to long-term therapies: evidence for action. J Nurs Scholarsh. 2003;35(3):207. doi: 10.1111/j.1547-5069.2003.tb00001.x [DOI] [PubMed] [Google Scholar]
- 8.Beerappa H, Chandrababu R. Adherence to dietary and fluid restrictions among patients undergoing hemodialysis: an observational study. Clin Epidemiol Glob Health. 2019;7(1):127–130. doi: 10.1016/j.cegh.2018.05.003 [DOI] [Google Scholar]
- 9.Ghimire S, Castelino RL, Lioufas NM, et al. Nonadherence to medication therapy in haemodialysis patients: a systematic review. PLoS One. 2015;10(12):e0144119. doi: 10.1371/journal.pone.0144119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wu ZB, Yao XW, Tan QZ, et al. Current status of treatment compliance of young maintenance hemodialysis patients and its influence factors: a 378-case study. J Nurs China. 2020;27(1):47–51. doi: 10.16460/j.issn1008-9969.2020.01.047 [DOI] [Google Scholar]
- 11.Saran R, Bragg-Gresham JL, Rayner HC, et al. Nonadherence in hemodialysis: associations with mortality, hospitalization, and practice patterns in the DOPPS. Kidney Int. 2003;64(1):254–262. doi: 10.1046/j.1523-1755.2003.00064.x [DOI] [PubMed] [Google Scholar]
- 12.Cukor D, Coplan J, Brown C, et al. Depression and anxiety in urban hemodialysis patients. Clin J Am Soc Nephrol. 2007;2(3):484–490. doi: 10.2215/CJN.00040107 [DOI] [PubMed] [Google Scholar]
- 13.Shirazian S, Grant CD, Aina O, et al. Depression in chronic kidney disease and end-stage renal disease: similarities and differences in diagnosis, epidemiology, and management. Kidney Int Rep. 2017;2(1):94–107. doi: 10.1016/j.ekir.2016.09.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hobfoll SE, Halbesleben J, Neveu JP, Westman M. Conservation of resources in the organizational context: the reality of resources and their consequences. Annu Rev Organ Psychol Organ Behav. 2018;5:103–128. doi: 10.1146/annurev-orgpsych-032117-104640 [DOI] [Google Scholar]
- 15.Wang XQ. Psychometric evaluation of the meaning in life questionnaire in Chinese middle school students. Chin J Clin Psychol. 2013;21(5):764–767. doi: 10.16128/j.cnki.1005-3611.2013.05.008 [DOI] [Google Scholar]
- 16.Cohen S. Social relationships and health. Am Psychol. 2004;59(8):676–684. doi: 10.1037/0003-066X.59.8.676 [DOI] [PubMed] [Google Scholar]
- 17.Sidik SM. Why loneliness is bad for your health. Nature. 2024;628(8006):22–24. doi: 10.1038/d41586-024-00900-4 [DOI] [PubMed] [Google Scholar]
- 18.King LA, Hicks JA. The science of meaning in life. Annu Rev Psychol. 2021;72(1):561–584. doi: 10.1146/annurev-psych-072420-122921 [DOI] [PubMed] [Google Scholar]
- 19.Ouyang WH, Wang XQ, Cai JY, Pan SY, Li JY. A three-level meta-analysis of the relationship between meaning in life and depression. J Affect Disord. 2026;398:121045. doi: 10.1016/j.jad.2025.121045 [DOI] [PubMed] [Google Scholar]
- 20.Breitbart W, Pessin H, Rosenfeld B, et al. Individual meaning-centered psychotherapy for the treatment of psychological and existential distress: a randomized controlled trial in patients with advanced cancer. Cancer. 2018;124(15):3231–3239. doi: 10.1002/cncr.31539 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Steger MF, Frazier P, Oishi S, Kaler M. The meaning in life questionnaire: assessing the presence of and search for meaning in life. J Couns Psychol. 2006;53(1):80–93. doi: 10.1037/0022-0167.53.1.80 [DOI] [Google Scholar]
- 22.Hooker SA, Masters KS, Park CL. A meaningful life is a healthy life: a conceptual model linking meaning and meaning salience to health. Rev Gen Psychol. 2018;22(1):11–24. doi: 10.1037/gpr0000115 [DOI] [Google Scholar]
- 23.Cai JY, Wang XQ, Zhang T, Pan SY. Social support and meaning in life: a meta-analysis. J Happiness Stud. 2026;27(2):24. doi: 10.1007/s10902-026-01008-4 [DOI] [Google Scholar]
- 24.Picariello F, Moss-Morris R, Macdougall IC, Chilcot J. The role of psychological factors in fatigue among end-stage kidney disease patients: a critical review. Clin Kidney J. 2017;10(1):79–88. doi: 10.1093/ckj/sfw113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Stavrova O, Luhmann M. Social connectedness as a source and consequence of meaning in life. J Posit Psychol. 2016;11(5):470–479. doi: 10.1080/17439760.2015.1117127 [DOI] [Google Scholar]
- 26.Ahrari S, Moshki M, Bahrami M. The relationship between social support and adherence of dietary and fluids restrictions among hemodialysis patients in Iran. J Caring Sci. 2014;3(1):11–19. doi: 10.5681/jcs.2014.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kara B, Caglar K, Kilic S. Nonadherence with diet and fluid restrictions and perceived social support in patients receiving hemodialysis. J Nurs Scholarsh. 2007;39(3):243–248. doi: 10.1111/j.1547-5069.2007.00175.x [DOI] [PubMed] [Google Scholar]
- 28.Bennett PN, St Clair Russell J, Atwal J, Brown L, Schiller B. Patient-to-patient peer mentor support in dialysis: improving the patient experience. Semin Dial. 2018;31(5):455–461. doi: 10.1111/sdi.12703 [DOI] [PubMed] [Google Scholar]
- 29.Taylor F, Gutteridge R, Willis C. Peer support for CKD patients and carers: overcoming barriers and facilitating access. Health Expect. 2016;19(3):617–630. doi: 10.1111/hex.12348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kugler C, Vlaminck H, Haverich A, Maes B. Nonadherence with diet and fluid restrictions among adults having hemodialysis. J Nurs Scholarsh. 2005;37(1):25–29. doi: 10.1111/j.1547-5069.2005.00009.x [DOI] [PubMed] [Google Scholar]
- 31.National Academies of Sciences. Engineering, and Medicine. Social Isolation and Loneliness in Older Adults: Opportunities for the Health Care System. Washington, DC: National Academies Press; 2020. [PubMed] [Google Scholar]
- 32.Buecker S, Maes M, Denissen JJA, Luhmann M. Loneliness and the big five personality traits: a meta-analysis. Eur J Pers. 2020;34(1):8–28. doi: 10.1002/per.2229 [DOI] [Google Scholar]
- 33.Murali KM, Mullan J, Roodenrys S, et al. Strategies to improve dietary, fluid, dialysis or medication adherence in patients with end stage kidney disease on dialysis: a systematic review and meta-analysis of randomized intervention trials. PLoS One. 2019;14(1):e0211479. doi: 10.1371/journal.pone.0211479 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Jhamb M, Abdel-Kader K, Yabes J, et al. Comparison of fatigue, pain, and depression in patients with advanced kidney disease and cancer: symptom burden and clusters. J Pain Symptom Manage. 2019;57(3):566–575. doi: 10.1016/j.jpainsymman.2018.12.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Costin V, Vignoles VL. Meaning is about mattering: evaluating coherence, purpose, and existential mattering as precursors of meaning in life judgments. J Personality Soc Phychol. 2020;118(4):864–884. doi: 10.1037/pspp0000225 [DOI] [PubMed] [Google Scholar]
- 36.Jones PJ, Heeren A, McNally RJ. Commentary: a network theory of mental disorders. Front Psychol. 2017;8:1305. doi: 10.3389/fpsyg.2017.01305 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Epskamp S, Borsboom D, Fried EI. Estimating psychological networks and their accuracy: a tutorial paper. Behav Res Methods. 2018;50(1):195–212. doi: 10.3758/s13428-017-0862-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Bringmann LF, Elmer T, Epskamp S, et al. What do centrality measures measure in psychological networks? J Abnorm Psychol. 2019;128(8):892–903. doi: 10.1037/abn0000446 [DOI] [PubMed] [Google Scholar]
- 39.Zhang Y, Huang J. The development of medical treatment adherence scale for the end-stage renal disease patients with maintenance hemodialysis. Chin Gen Pract. 2013;16(34):3312–3316. doi: 10.3969/j.issn.1007-9572.2013.28.010 [DOI] [Google Scholar]
- 40.Yan BB, Zheng X. Researches into relations among social-support, self-esteem and subjective well-being of college students. Psychol Dev Educ. 2006;22(3):60–64. Available from: https://devpsy.bnu.edu.cn/CN/Y2006/V22/I3/60 [Google Scholar]
- 41.Yang B, Guo LL. Reliability and validity of the Chinese version of the De Jong Gierveld Loneliness Scale. Chin Gen Pract. 2019;22(33):4110–4115. doi: 10.12114/j.issn.1007-9572.2019.00.319 [DOI] [Google Scholar]
- 42.Haslbeck JM, Waldorp LJ. How well do network models predict observations? On the importance of predictability in network models. Behav Res Methods. 2018;50(2):853–861. doi: 10.3758/s13428-017-0910-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Robinaugh DJ, Millner AJ, McNally RJ. Identifying highly influential nodes in the complicated grief network. J Abnorm Psychol. 2016;125(6):747–757. doi: 10.1037/abn0000181 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Reid C, Seymour J, Jones C. A thematic synthesis of the experiences of adults living with hemodialysis. Clin J Am Soc Nephrol. 2016;11(7):1206–1218. doi: 10.2215/CJN.10561015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Cacioppo JT, Cacioppo S. The growing problem of loneliness. Lancet. 2018;391(10119):426. doi: 10.1016/S0140-6736(18)30142-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Holt-Lunstad J. Loneliness and social isolation as risk factors: the power of social connection in prevention. Am J Lifestyle Med. 2021;15(5):567–573. doi: 10.1177/15598276211009454 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Winger JG, Adams RN, Mosher CE. Relations of meaning in life and sense of coherence to distress in cancer patients: a meta-analysis. Psychooncology. 2016;25(1):2–10. doi: 10.1002/pon.3798 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Newman DB, Nezlek JB, Thrash TM. The dynamics of searching for meaning and presence of meaning in daily life. J Pers. 2018;86(3):368–379. doi: 10.1111/jopy.12321 [DOI] [PubMed] [Google Scholar]
- 49.Czekierda K, Banik A, Park CL, Luszczynska A. Meaning in life and physical health: systematic review and meta-analysis. Health Psychol Rev. 2017;11(4):387–418. doi: 10.1080/17437199.2017.1327325 [DOI] [PubMed] [Google Scholar]
- 50.Kim ES, Shiba K, Boehm JK, Kubzansky LD. Sense of purpose in life and five health behaviors in older adults. Prev Med. 2020;139:106172. doi: 10.1016/j.ypmed.2020.106172 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Denhaerynck K, Manhaeve D, Dobbels F, et al. Prevalence and consequences of nonadherence to hemodialysis regimens. Am J Crit Care. 2007;16(3):222–235. doi: 10.4037/ajcc2007.16.3.222 [DOI] [PubMed] [Google Scholar]
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/or analysed during the current study are not publicly available due to ethical restrictions. However, they are available from the corresponding author upon reasonable request. The analytical code for this study is available in Supplementary File S6.