Abstract
Objective
As digitalization accelerates, understanding the relationship between digital health literacy (DHL) and health-related quality of life (HRQoL) among older adults is critical. This study investigated the complex, double-edged impact of DHL on HRQoL, specifically examining the serial mediating roles of social support and loneliness among Chinese older adults.
Methods
A cross-sectional study was conducted using questionnaire data from 469 older adults in China. A serial mediation model was constructed and tested using structural equation modeling (SEN) to analyze the pathways from DHL to HRQoL.
Results
The model revealed a surprising and significant negative total effect of DHL on HRQoL. However, DHL was positively associated with both social support (β = 0.45, p < .001) and, unexpectedly, also with loneliness (β = 0.57, p < .001). The analysis confirmed a significant positive serial mediating pathway, where higher DHL increased social support, which in turn reduced loneliness, ultimately leading to better HRQoL. Despite this positive indirect effect, the strong positive association between DHL and loneliness contributed to the overall negative effect on HRQoL.
Conclusion
These findings provide theoretical insights for developing health promotion policies for the elderly that emphasize integrated interventions that combine digital empowerment with psychosocial support.
Introduction
As China rapidly advances towards its “Healthy China 2030” goals amidst profound digitalization, the well-being of its over 260 million older adults has become a central public health priority. 1 Digital health technologies are widely promoted as a panacea to enhance Health-Related Quality of Life (HRQoL), with the assumption that improving older adults’ digital health literacy (DHL)—their ability to access and use online health information—will directly lead to better health outcomes.2,3 However, a perplexing paradox emerges from the real world: many digitally adept older adults report not a sense of empowerment, but heightened health anxiety, social comparison, and feelings of isolation.4,5 This stark contradiction between policy expectations and lived realities challenges the simplistic “digital-empowerment-equals-well-being” narrative. It underscores a critical need to unpack the complex psychosocial mechanisms that mediate the true impact of DHL on HRQoL, which forms the central impetus for this study.
Conceptually, higher DHL should empower older adults to make informed health decisions, engage in preventive behaviors, and better manage chronic conditions, thereby leading to improved HRQoL.6,7 However, the empirical evidence presents a more complex and inconsistent picture. While certain studies confirm this positive association, others suggest that high engagement with digital health information can lead to information overload, heightened health anxiety, and negative social comparisons, potentially harming HRQoL. 8 Therefore, it is evident that the relationship between DHL and HRQoL is not a straightforward linear one, but is likely mediated by complex psychosocial mechanisms.
This inconsistency in the findings of previous studies points to a significant research gap. While some studies confirm the benefits of DHL in facilitating informed health decisions,9–11 others highlight its adverse effects, such as inducing ‘cyberchondria’ or creating a ‘social displacement effect,’ where online engagement diminishes meaningful offline interactions.12–14 Although social support and loneliness are widely recognized as powerful determinants of well-being in later life, they have rarely been examined together as serial mediators within a unified framework to elucidate this paradox. Previous studies have investigated the relationship between DHL and social support, or between social support and loneliness15–17; however, the sequential pathway—whereby DHL influences social support, which subsequently affects loneliness, and ultimately HRQoL—remains a theoretical and empirical ‘black box.’ This issue is particularly relevant in the Chinese context, where collectivist values and strong family ties render social connectedness a cornerstone of well-being, yet this very fabric is being reshaped by digitalization and social change.
According to the Buffer-Gain Theory, social support as a crucial resource that not only buffer against stress but also directly promoting well-being. 18 DHL can facilitate the acquisition of social support by enabling older adults to connect with peers, family, and healthcare providers through digital platforms.19–21 Loneliness a persistent negative emotional arising from a perceived deficit in social relationships, is a major risk factor for poor HRQoL among the elderly.22–24 The impact of DHL on loneliness is ambiguous; it could alleviate it by fostering new social connections or exacerbate it by substituting shallow online interactions for deep, face-to-face contact.25–27 The Behavior-Cognitive Theory (BCT) provides a lens for understanding these dynamics, suggesting that health literacy (a cognitive asset) influences health-related behaviors, such as seeking support or social engagement, which in turn shape psychological and health outcomes.28,29
Although existing studies have preliminarily confirmed the direct association between DHL and HRQoL, there is a notable lack of in-depth exploration of the complex transmission pathways through which DHL influences HRQoL in older adults via psychosocial mechanisms. Specifically, it is unclear whether DHL enhances social support, which in turn alleviates loneliness, ultimately leading to better HRQoL. This sequential mechanism is particularly relevant in the Chinese context, where collectivist values place a strong emphasis on social connectedness. 21
Therefore, this study aims to construct and test a serial mediation model to fill this gap. As illustrated in the conceptual framework (see Figure 1), we propose that DHL’s effect on HRQoL is channeled through the mediating roles of social support and loneliness. Based on this model, we hypothesized that DHL would not only have a direct effect on HRQoL (hypothesis 1, H1) but would also exert indirect effects through three pathways: (1) a positive path via increased social support (H2); (2) a positive path via reduced loneliness; and most importantly (H3), (3) a serial mediation path whereby higher DHL enhances social support, which in turn reduces feelings of loneliness, ultimately improving HRQoL (H4). By validating these hypotheses, this study seeks to provide a nuanced understanding of the psychosocial mechanisms linking digital skills to well-being among Chinese older adults, and offering robust evidence to resolve the current theoretical debate and, more importantly, to guide the development of more nuanced and human-centered digital health policies that avoid the potential pitfalls of technological solutionism. The conceptual model of the study.
Methods
Participants
This cross-sectional study analyzes data from a survey conducted as part of the “Healthy Living Condition of Community Residents and Healthy Community Building Project,” a school-local cooperation initiative. The survey was carried out in Wuxi City, Jiangsu Province, from January to May 2024. A total of 21 communities (13 urban and 8 rural) were included. The target population for the original survey comprised adults aged 18 years and older who were capable of independently completing the questionnaire and had prior internet experience. Individuals lacking internet experience or those unable to complete the questionnaire independently (e.g., due to severe cognitive or physical limitations) were excluded.
General characteristics of participants (N = 469).
+ In China, the hukou system serves as a crucial demographic variable. The possession or lack of a local hukou significantly influences an individual’s and family’s access to quality health services and social support, as well as the formation of social networks.
Measurement instruments
Health-Related Quality of Life (HRQoL)
HRQoL was assessed using the HRQOL Comprehensive Assessment Scale developed by Ware et al.(1996). 30 This scale comprises 12 questionnaire items categorized into two dimensions: Physical Health (8 items) and Mental Health. A Kaiser-Meyer-Olkin (KMO) measure of .904 and a significant Bartlett’s test of sphericity (p < 0.001) indicated excellent sampling adequacy for factor analysis. Sample questions included: “During the past month, I have limited strenuous exercise due to health problems,” “During the past month, I have reduced my work hours or other daily activities because of emotional problems,” “Overall, my quality of life has been good in the past month,” and “During the past month, my quality of life has been very good.” Participants rated items on a 6-point Likert scale (1 = completely inconsistent to 6 = completely consistent). The score for each dimension was calculated by averaging its respective items. A higher average scores indicate a better HRQoL. The overall reliability of the scale, indicated by Cronbach’s α = 0.725, was considered satisfactory.
Digital Health Literacy (DHL)
The variables for assessing DHL were based on the Digital Health Literacy Assessment Scale developed by Liu (2022). 31 This measurement was adapted to include 12 questionnaire items divided into two dimensions (KMO = .931, p< 0.001): access and assessment (6 items) and interaction and application (6 items). The items included statements such as, “I am interested in learning about health knowledge or skills on the Internet,” “I will check the health information available on the Internet,” and “I will take care of my own and others’ privacy and security when searching for and sharing information online.” Items were rated on a 6-point scale (1 = never to 6 = always). The overall social support score was calculated by averaging all eight items, where higher scores denote a greater level of perceived social support. The scale’s reliability was adequate (Cronbach’s α = 0.717).
Social support
Social support was assessed using the modified Medical Outcomes Study Social Support Survey (mMOS-SS) scale adapted by Gómez-Campelo et al. (2014). 32 This scale comprises eight items categorized into two dimensions: instrumental support (including financial, informational, and daily living assistance), and emotional support (KMO = .798, p < 0.001). Sample items include “someone who can provide you with the information you need,” “someone who can give you financial or monetary assistance,” and “someone who can share your joys and concerns with you.” Items were rated on a 6-point scale (1 = never to 6 = always). The overall social support score was calculated by averaging all eight items, where higher scores denote a greater level of perceived social support. The scale’s reliability was adequate (Cronbach’s α = 0.717).
Loneliness
This study assessed participants’ levels of loneliness using the UCLA-8 (Version 3) brief Chinese version scale, which was adapted by Zhou et al.(2012) 33 from Hays and and DiMatteo’s (1987) UCLA Loneliness Scale. 34 The scale comprises eight items(KMO = .923, p < 0.001), including statements such as “I often feel a lack of companionship,” “I often feel left out,” and “I often feel that no one can be trusted.” Responses were given on a 6-point scale (1 = never to 6 = always). The total score was calculated by averaging the eight items, with higher scores indicating greater feelings of loneliness. The scale demonstrated good reliability in our study (Cronbach’s α = 0.824).
Statistical analysis
This study employed statistical analysis using SPSS 25.0, and AMOS 25.0. Initially, descriptive and correlation analyses were conducted with SPSS to assess the sociodemographic characteristics of the respondents. Following this, structural equation modeling (SEM) was performed using AMOS to investigate the serial mediating effects of DHL and social support on loneliness and its subsequent impact on HRQoL. Our analytical approach consisted of five key steps: (1) Descriptive analyses were conducted to ascertain the sociodemographic characteristics of the respondents. (2) Exploratory factor analysis and confirmatory factor analysis were utilized to evaluate the structural validity of the data. (3) Reliability tests were performed to assess the consistency of the measurements. (4) Correlation analysis was executed to examine the relationships among the primary variables. (5) Finally, SEM was employed to analyze the relationship between DHL (independent variable) and HRQoL (dependent variable) while testing the serial mediation effects of social support and loneliness.
Result
Descriptive statistics and correlation analysis of main variables
Descriptive analysis of the main variables.
As shown in Appendix Table 1, correlation analysis revealed significant correlations among all main variables. Notably, DHL was positively correlated with both social support (r = .448, p < .001) and loneliness (r = .568, p < .001), and negatively correlated with HRQoL (r = -.311, p < .001). Furthermore, a multicollinearity assessment confirmed no issues, with all variables exceeded 0.1 (range: 0.476-0.987), and the Variance Inflation Factor (VIF) remained below 10 (range: 1.013-2.101).
Mediation analysis with Structural Equation Modeling (SEM)
Results of the multiple mediation model analysis.
*p<.05, **p<.01, ***p<.001.
x2=65.777, df=9, CFI=.922, TLI=.924, RMSEA=.082.
+ DHL = digital health literacy; SS = social support; LON = Loneliness; HRQoL= health-related quality of life.
The path coefficients of the multiple mediation model.
Results of the multiple mediation effect test.
+ DHL = digital health literacy; SS = social support; LON = Loneliness; HRQoL= health-related quality of life.
Discussion
This study investigated the complex relationship between DHL, social support, loneliness, and HRQoL among Chinese older adults. Our findings present a paradox: while DHL offer pathways to improved well-being, its overall effect on HRQoL appears to be negative, acting as a ‘double-edged sword’. This is primarily because the strong, detrimental impact of increased loneliness outweighs the modest benefits gained from enhanced social support. These results challenge the simplistic views of digital empowerment and provide a nuanced understanding of the psychosocial costs and benefits of digital engagement in later life.
This ‘double-edged sword’ finding must be interpreted through the unique lens of contemporary Chinese society. The strong positive association between DHL and loneliness, which was particularly unexpected, can be explained by the specific characteristics of older adults in China. As relative newcomers to the digital world, or ‘digital immigrants,’ their motivation to learn digital skills is often instrumental. 35 For example, it serves to maintain contact with children who have migrated to distant cities due to rapid urbanization. While this increases their perceived social support (our DHL → social support path), these digital interactions are often fragmented and asynchronous, failing to replace the deep emotional nourishment of face-to-face companionship that is highly valued in Chinese culture. Furthermore, the legacy of the one-child policy has created a substantial population of ‘empty-nest’ and ‘solo-living’ elders who face significant structural risks of loneliness.36,37 For this group, extensive engagement with online health information may inadvertently become a solitary activity that displaces real-world social participation, turning the digital world into a ‘sophisticated cage’ that amplifies, rather than alleviates, their underlying sense of isolation. This tension, arising from the interplay between socio-structural changes and cultural values, is crucial for understanding our results.
The first significant finding of this study was the negative total effect of DHL on HRQoL, driven by a powerful indirect pathway through loneliness. Against H1, higher DHL did not directly translate to better HRQoL. Why would being more digitally savvy about health lead to worse outcomes? We propose several potential explanations. (1) Higher DHL may lead to increased health anxiety and information overload. Older adults with greater digital skills may be exposed to a vast and often contradictory sea of health information, including frightening symptom checkers and misinformation, which can trigger anxiety and reduce subjective well-being. 38 (2) It could be a case of substitution effect, where online health-related activities replace higher-quality, face-to-face interactions with family, friends, or healthcare providers, thus weakening substantive social bonds. (3) Negative social comparisons on health-focused online platforms can be detrimental. Seeing peers who appear healthier or are managing their conditions more successfully can lead to feelings of inadequacy and depression. This complex reality refutes the simple assumption that digital literacy is an unalloyed good.
Another primary finding of our study is that our model elucidates two opposing mechanisms. In alignment with the Buffer-Gain Theory, the pathway DHL → social support → HRQoL represents a positive mechanism. Higher DHL enables older adults to utilize digital tools to maintain social ties and access informational support, which subsequently enhances HRQoL. This finding is consistent with existing research that underscores the significance of social connections for healthy aging,39,40 particularly within the collectivist context of Chinese culture. 41 However, this positive pathway is overshadowed by a more potent negative mechanism: DHL → loneliness → HRQoL. The strong positive correlation between DHL and loneliness was particularly unexpected, suggesting that for many older adults, digital engagement may not alleviate but rather coexist with, or even exacerbate, feelings of isolation. This, coupled with the substantial negative impact of loneliness on HRQoL, creates a prevailing negative force that ultimately detracts from well-being. These findings highlight that the quantity of digital interaction cannot substitute for the quality of human connection.
Finally, the study confirmed the serial mediation pathway: DHL → social support → loneliness → HRQoL. This finding represents a significant theoretical contribution, illustrating the sequential interconnections among these factors. It demonstrates that social support serves as a protective buffer, with increased social support significantly reducing feelings of loneliness. This underscores that the positive effects of DHL are contingent upon its capacity to foster authentic social support, which subsequently alleviates the pervasive risk of loneliness. By integrating these variables into a cohesive model, our study advances beyond previous research that examined these factors in isolation, providing robust empirical evidence for a cognitive-behavioral framework in which literacy influences social resources, which in turn shape psychological states and health outcomes.
Based on the findings of this study, we propose the following recommendations to inform policy and practice, particularly in the context of China’s ‘National Strategy for Actively Responding to Population Aging.’ (1) Transition from ‘Digital-Only’ to ‘High-Tech, High-Touch’ Integrated Interventions. Our results serve as a crucial warning for ongoing ‘Smart Eldercare’ projects: merely equipping older adults with digital skills is insufficient and may even be counterproductive. Future interventions should integrate digital literacy training with structured, offline psychosocial support. For instance, community centers could utilize health applications not only for information dissemination but also to organize offline group activities such as health seminars, walking clubs, or intergenerational tech-buddy programs, thereby transforming online connections into tangible social bonds. (2) Elevate Loneliness to a Core Public Health Priority in the Digital Age. Given the overwhelming negative impact of loneliness observed in our model, it should be regarded as a key clinical risk factor. We strongly advocate for the integration of routine loneliness screening (e.g., utilizing the UCLA-8 scale) into the ‘National Basic Public Health Service Standards’ for older adults. This would enable community health workers to identify at-risk individuals early and provide targeted support, thereby mitigating the unintended social consequences of digitalization. (3) Design for Meaningful Connection, Not Just Information Access. Technology developers and policymakers must shift their focus from creating platforms for passive information consumption to designing for high-quality social interaction. Digital health platforms should incorporate features that facilitate meaningful engagement, such as moderated peer-support groups, online-to-offline event organization, and tools that simplify intergenerational communication, ensuring that technology serves as a bridge to, rather than a substitute for, genuine human connection.
Limitations
This study presents several limitations. First, the specific mechanisms driving the antagonistic effects of the mediating variable were not fully explored. Specifically, the “double-edged” nature of digital health literacy (DHL) may stem from a reliance on online health information that weakens real-world social connections. This counterintuitive pathway contradicts our initial hypothesis regarding the link between DHL and HRQoL. Future research should further investigate the psychosocial mechanisms underlying this phenomenon. Second, the generalizability of the findings is limited by the geographic scope of the sample. The sampling distribution was not sufficiently randomized, resulting in a concentration of participants in specific regions. Future studies should employ rigorous random sampling across a broader geographic range to reduce selection bias and enhance representativeness. Third, although this study examined a serial mediation model, other potential mediators or moderators—such as health behaviors, self-efficacy, and digital technology usage patterns—warrant consideration. Future work should expand the theoretical model to include these variables and employ a longitudinal design to better establish causal relationships. Fourth, this study did not examine the differential impacts of specific DHL components (e.g., information acquisition, assessment, and application skills) on health outcomes. Future studies should use multidimensional measurement tools to parse how distinct DHL dimensions influence social support, loneliness, and HRQoL.
Future directions
Beyond addressing these limitations, our findings open several avenues for future research. First, longitudinal design is essential for tracking the dynamic interplay between these variables over time and establishing clearer causal pathways. Second, mixed-methods studies that combine quantitative surveys with in-depth qualitative interviews would be invaluable in uncovering the subjective experiences and coping strategies of older adults as they navigate the digital health landscape. How do they personally weigh the convenience of information against the potential for social loss? Third, and most importantly, the field should advance towards intervention research. Designing and rigorously evaluating the effectiveness of the proposed ‘high-tech, high-touch’ integrated interventions would provide the most direct and actionable evidence for policymakers. Finally, future models could explore other critical variables, such as the moderating role of intergenerational relationship quality or the differential impacts of various types of digital technology use (e.g., social media versus health information portals). Cross-cultural comparative studies would also be highly valuable to test whether the ‘double-edged sword’ phenomenon is unique to collectivistic societies like China or a more universal feature of digital aging.
Conclusion
This study clarifies the complex mechanisms linking DHL to HRQoL among Chinese older adults, specifically examining the mediating roles of social support and loneliness. Specifically, results indicate that while DHL promotes health by enhancing social support, it may simultaneously exacerbate loneliness, leading to negative health outcomes. This paradox challenges the assumption that technological empowerment invariably leads to improved health, highlighting the critical influence of psychosocial factors in digital health interventions. Consequently, the implementation of digital health technologies for older adults must go beyond technical skills training and functional utility to address broader impacts on social connectedness and mental well-being. Future strategies should adopt a holistic approach that balances improved technology accessibility with efforts to strengthen social networks and mitigate loneliness, ultimately empowering older adults to achieve better health outcomes. Collectively, these findings offer novel insights into the digital divide and provide actionable implications for optimizing digital health services in an aging society.
Footnotes
Ethical considerations
This paper was reviewed and considered human subjects exempt by Jiangnan University Medical Ethics Committee.
Consent to participate
The participants were informed about the purpose of this research and provided written informed consent before administering the questionnaire. To maintain confidentiality, personal identifiers were omitted from the questionnaires and the collected data were stored in a secure, password-protected database without identifiers.
Author contributors
This work is the result of a collaboration between D. W, C. J, S.S and F. G. Authors have equally contributed, reviewed, and improved the manuscript. Conceptualization, S. S, D. W and C. J; Methodology, S. S., C. J and F. G.; Writing—original draft preparation, C.J and D. W; Writing—review and editing, S. S. and D. W; Funding acquisition, D. W, S. S and F. G. All authors have revised the final manuscript. All authors have read and agreed to the published version of the manuscript.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work is supported by General Project of Philosophy and Social Science Research in Colleges and Universities of Jiangsu Province (2022SJYB0950).
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to their containing information that could compromise the privacy of research participants.
Appendix
Appendix Table 1
Correlation analysis of the main variables. *p<.05, **p<.01
X1
X2
X3
X4
X5
X6
X7
Physical health (X1)
1
Mental health (X2)
.267**
1
Access & evaluate (X3)
.104*
.071
1
Interact & apply (X4)
-.295**
-.112*
.622**
1
Instrumental support (X5)
.160**
.269**
.380**
.225**
1
Emotional support(X6)
.192**
.285**
.338**
.199**
.572**
1
Loneliness(X7)
-.737**
-.243**
.159**
.375**
-.121*
-.192**
1