Compensatory network capital: negative and ambivalent ties in the conversion of health literacy into action

Abstract

Critical Health Literacy (CHL) emphasizes that health-related action is shaped by power, inequality and structural conditions, not by individual skills alone. This study examines a more specific and under-theorized link in that process: how negative and ambivalent social ties affect whether health-related competencies are enacted as concrete action in everyday healthcare. Drawing on 12 semi-structured interviews with adults living with multimorbidity in contexts of social vulnerability, the study explores how health-related knowledge, intentions, and professional advice are converted, delayed, redirected, or blocked in concrete health-related episodes. Analysis was abductive and used Compensatory Network Capital (CNC) as a sensitizing framework. Participants often described knowledge, motivation, and awareness of recommended care, but action appeared to be constrained when support was difficult to mobilize, poorly matched to the task, insufficiently recognized in institutional encounters, or when no helper was available to act on the person’s behalf. These patterns were analysed through four tie-level mechanisms: Activation, Function-specific Help, Recognition and Substitution. Negative and ambivalent ties appeared to operate not only as psychosocial stressors but also as practical barriers affecting help-seeking, follow-up, care coordination and digital navigation. The study contributes to CHL research by specifying a micro–meso relational layer through which health-related competencies may become actionable, partial or blocked in practice. It suggests that equity-oriented health promotion may need to address not only individual skills, but also the relational and organizational conditions that enable, constrain or block health-related action.

Graphical Abstract

Graphical Abstract.

Contribution to Health Promotion.

• Introduces CNC as a way of understanding when support becomes usable in real healthcare situations.

• Indicates how family, friends and other close ties can help or hinder people in acting on health advice.

• Indicates that negative and ambivalent relationships affect care not only through stress but also, for example, by delaying help-seeking and follow-up.

• Points to practical responses such as making it easier to ask for help, providing task-specific support, recognizing informal helpers and improving options for proxy action.

Introduction

Managing multimorbidity often requires more than understanding health information; it also involves treatment work and the continuous balancing of workload and capacity across everyday life and healthcare systems (Shippee et al. 2012, May et al. 2014). It involves a sequence of small but consequential tasks, such as interpreting written or digital messages, booking or changing appointments, arranging transport, attending consultations, understanding treatment instructions, making medication changes, and completing follow-up across fragmented services. These tasks are frequently embedded in everyday relationships and distributed across personal networks (Vassilev et al. 2013). Family members, friends, or peers may help make action possible, but social ties may also go unused, offer poorly matched advice, create conflict, or remain invisible to organizations (Rosland et al. 2010). In this article, social ties refer to family members, friends, peers, neighbours, and other informal helpers who may become involved in concrete health-related tasks.

A substantial body of epidemiological research has documented protective associations between social support and health outcomes (Berkman et al. 2000, Uchino 2006, Holt-Lunstad et al. 2010, Umberson and Montez 2010). However, much of this literature operationalizes support as availability, quantity, or perceived adequacy, which makes it less visible how specific ties are mobilized, matched to concrete healthcare demands, recognized by organizations, or misaligned with professional advice.

This article brings together social support, health literacy (HL), and Critical Health Literacy (CHL); therefore, a brief clarification of terminology is needed. HL refers broadly to individual and distributed competencies for accessing, understanding, appraising, and using health information (Sørensen et al. 2012). Functional HL concerns task-oriented capacities, such as understanding instructions, booking an appointment, or responding to a health-related message. CHL serves here as the article’s overarching action- and equity-oriented framework, concerned with reflection, judgement, power, and the social conditions of action. ‘Health-related competencies’ denotes the knowledge, understanding, motivation, intentions, and practical capacities that may, or may not, be enacted in concrete health-related action.

Compensatory Network Capital (CNC) does not reduce CHL-related action to individual skills. Rather, it specifies the relational and institutional conversion conditions under which health-related competencies, including CHL-related capacities, become actionable, partial, delayed, redirected, or blocked in practice (Bak et al. 2025). In this study, CNC is used to examine how support from social ties is converted, or fails to be converted, into concrete action by asking whether help is activated, task-specific, institutionally recognized and, where necessary, substitutive at the point of need. Here, CNC refers to the relational layer at micro–meso level linking individual competencies, concrete social ties and organizational or institutional conditions.

Existing research shows that relationships and competencies matter for health but says less about how negative and ambivalent ties interfere with conversion in concrete healthcare situations. Negative interactions are often measured as psychosocial stressors (Rook 1984, 2015, Brooks and Dunkel Schetter 2011), leaving under-specified how conflict, judgement, dependency concerns, or competing health logics may suppress help-seeking, misalign support, reduce recognition, or make safe substitution unavailable.

In this article, ‘concrete health-related action’ is used as an umbrella term. Empirically, however, the study focuses on bounded health-system navigation and care coordination tasks, including booking, attendance, digital or written correspondence, treatment instructions, medication changes, transport, accompaniment, and follow-up. Broader lifestyle behaviours are discussed only as implications for future research.

Multimorbidity is understood as the presence of two or more chronic or long-term conditions (Boyd and Fortin 2010, World Health Organization 2016). The study does not assume that multimorbidity itself implies low HL, limited decision-making capacity, or social vulnerability. Rather, the focus is on people with multimorbidity living in contexts of social vulnerability: social, material, and institutional conditions that may make health-related action more fragile, including limited resources, unstable or strained support, digital barriers, treatment burden, fragmented care pathways, and difficult access to recognized help (Mechanic and Tanner 2007, Braveman and Gottlieb 2014, May et al. 2023).

Digital competence is not treated here as a separate outcome, but as one practical condition shaping whether competencies can be enacted in digitalized healthcare episodes. It refers to navigating concrete digital tasks such as reading digital messages, using patient portals, booking appointments, responding to written communication, and managing digital follow-up. Because Distributed Health Literacy is abbreviated as DHL, digital health literacy is written in full throughout.

This conversion problem is particularly relevant to CHL. CHL moved health literacy beyond functional and individualized understandings by emphasizing reflection, judgement, collective agency, and action on social determinants of health (Nutbeam 2008, Chinn 2011, Abel and McQueen 2013, Abel and Benkert 2022, Sykes et al. 2025). However, even when people have knowledge, motivation, and critical awareness, concrete action may still depend on relational and institutional conditions. Capability-based approaches similarly underline that resources acquire practical significance through conversion processes that translate endowments into real opportunities for action (Sen 1999, Abel and Frohlich 2012, Pithara 2020, Rüegg and Abel 2022).

DHL has shown that health-related competencies are often shared and mobilized within networks (Edwards et al. 2015, Muscat et al. 2022). CNC builds on this insight but asks a more specific question: when does distributed competence become concrete action? It therefore treats social ties as analytically open: they may enable, partially enable, delay, redirect, or block health-related action. This is particularly salient in multimorbidity, where treatment burden, digital systems and fragmented pathways increase coordination demands (Hardman et al. 2020, Chauhan et al. 2024).

Research aim

Guided by CHL, the study asks: How do negative and ambivalent social ties shape the conversion of health-related competencies into concrete action among people with multimorbidity in contexts of social vulnerability, and how do these ties interact with institutional conditions in concrete health-related episodes? In addressing this question, the study also refines CNC as a sensitizing framework for analysing relational and institutional conversion conditions.

Conceptual framework

CNC is used here as an operational framework for analysing relational conversion processes in everyday healthcare. A fuller theoretical elaboration is provided elsewhere (Bak et al. 2025). CNC is defined as a health-specific subdimension of social capital that captures how immediate social networks influence the translation of health-related resources into action. Rather than focusing on network size, density, or generalized trust, CNC specifies tie-level mechanisms through which relations enable, constrain, or distort the enactment of competencies.

The four mechanisms are applied as sensitizing analytical constructs rather than fixed or sequential stages. Activation refers to whether relevant social ties are mobilized or withheld when individuals face health-related demands. Function-specific Help concerns whether support matches the practical, cognitive, digital or emotional demands of the situation. Recognition captures whether informal support is acknowledged within institutional encounters. Substitution denotes situations in which social ties act with or on behalf of the individual, such as booking appointments, attending consultations, or managing digital correspondence.

Full compensation requires alignment of all four mechanisms; empirically, cases more often show partial or disrupted configurations. Negative and ambivalent ties are therefore conceptualized not as the absence of support, but as relational processes that suppress activation, misalign help, undermine recognition, or hinder substitution.

Figure 1 presents the study’s conceptual pathway. It suggests how social support, health-related competencies, and contextual conditions may or may not be converted into action through CNC’s micro–meso relational conversion layer. The figure therefore illustrates the theory-of-change logic behind the study: action depends not only on competencies or support as resources, but on whether help is activated, task-specific, institutionally recognized, and substitutive at the point of need. Examples of multidimensional self-reported health literacy instruments include the Health Literacy Questionnaire (HLQ) (Osborne et al. 2013).

Figure 1.

Conceptual pathway for relational conversion into health-related action. The figure presents Critical Health Literacy as the article's overarching action- and equity-oriented frame. It shows how health-related endowments and perceived competencies may be converted into concrete health-related action through CNC's micro–meso relational conversion layer, consisting of Activation, Function-specific Help, Recognition and Substitution. The pathway retains a Sen-inspired distinction between endowments/resources, capabilities and functionings.

Table 1 summarizes the conceptual relationship between CHL, DHL, and CNC. It should not be read as a hierarchy. CHL provides the action- and equity-oriented framework; DHL highlights the social distribution of health-related competencies; and CNC specifies the micro–meso relational conversion mechanisms through which such competencies become actionable, partial, delayed, redirected, or blocked in concrete healthcare episodes.

Table 1.

Conceptual comparison of CHL, DHL and CNC.

Dimension	Critical Health Literacy (CHL)	Distributed Health Literacy (DHL)	Compensatory Network Capital (CNC)
Primary purpose	Cognitive and social skills development for critical appraisal, reflection, judgement and individual or collective action in relation to health, power, and social determinants (Nutbeam 2000, 2008, Chinn 2011, Osborne et al. 2013, Abel and Benkert 2022, Sykes et al. 2025).	Understanding how health-related knowledge, understanding, and practical know-how are shared, mediated, and mobilized through family members, peers, and wider social networks (Vassilev et al. 2013, Edwards et al. 2015, Muscat et al. 2022).	Explaining how health-related competencies and support are converted, delayed, redirected, or blocked through relational and institutional mechanisms in concrete healthcare episodes (Sen 1999, Abel and Frohlich 2012, Rüegg and Abel 2022, Bak et al. 2025).
Theoretical foundation	Functional, interactive, and critical health literacy; cognitive and skills development; critical appraisal; reflection, empowerment, social justice and action (Nutbeam 2000, 2008, Chinn 2011, Abel and McQueen 2013, Abel and Benkert 2022).	Distributed health literacy, social networks, social support, and chronic illness self-management (Vassilev et al. 2013, Edwards et al. 2015, Muscat et al. 2022).	Capability approach, Bourdieusian social capital, social conversion, and tie-level relational mechanisms (Bourdieu 1986; Sen 1999, Abel and Frohlich 2012, Rüegg and Abel 2022, Bak et al. 2025).
Analytical focus	Critical appraisal, reflection, judgement, power relations, structural conditions, and action (Nutbeam 2008, Chinn 2011, Abel and Benkert 2022, Sykes et al. 2025).	How close others mediate information, understanding, navigation, self-management, and action support (Vassilev et al. 2013, Edwards et al. 2015, Muscat et al. 2022).	Whether help is activated, function-specific, institutionally recognized, and substitutive at the point of need (Bourdieu 1986; Sen 1999, Bak et al. 2025).
Analytical level	Individual, collective, and macro–meso: competencies are developed and enacted within social, institutional, and structural conditions (Nutbeam 2000, 2008, Chinn 2011, Abel and Benkert 2022).	Meso/relational: health literacy is distributed across close ties, family members, peers, and personal communities (Vassilev et al. 2013, Edwards et al. 2015).	Micro–meso: tie-level mechanisms in specific health-related episodes, situated between individual competencies, social ties, and institutional settings (Abel and Frohlich 2012, Rüegg and Abel 2022, Bak et al. 2025).
View of social relations	Social and institutional conditions shape whether critical appraisal, judgement and action can be enacted (Chinn 2011, Abel and McQueen 2013, Abel and Benkert 2022, Sykes et al. 2025).	Primarily enabling and supportive: close others may interpret information, mediate encounters, and support action (Vassilev et al. 2013, Edwards et al. 2015, Muscat et al. 2022).	Analytically neutral: ties may enable, partially enable, delay, redirect, or block action; their value depends on situational convertibility (Rook 1984, 2015, Bourdieu 1986; Adebahr 2022, Bak et al. 2025).
Treatment of negative and ambivalent ties	Usually indirect: negative or ambivalent ties may be relevant through power, inequality, or social conditions, but are rarely specified as tie-level conversion mechanisms (Chinn 2011, Abel and Benkert 2022, Sykes et al. 2025).	Limited: negative or ambivalent ties are not central to most DHL formulations, which mainly emphasize supportive mediation and shared competence (Edwards et al. 2015, Muscat et al. 2022).	Explicitly theorized: negative and ambivalent ties may suppress activation, misalign support, undermine recognition or hinder substitution (Rook 1984, 2015, Adebahr 2022, Bak et al. 2025).
Conceptual clarity	Broad but increasingly specified: combines cognitive/skills development, critical appraisal, reflection, action, empowerment and equity (Nutbeam 2000, 2008, Chinn 2011, Abel and Benkert 2022, Sykes et al. 2025).	Conceptually relational, but overlaps with social support, social capital, and network perspectives; conversion into action remains less fully specified (Vassilev et al. 2013, Edwards et al. 2015, Muscat et al. 2022).	Distinct A–F–R–S mechanism vocabulary with clear analytical boundaries: Activation, Function-specific Help, Recognition and Substitution (Bourdieu 1986; Sen 1999, Bak et al. 2025).
Operationalization	Often qualitative or survey-based; CHL measurement varies across cognitive skills, critical appraisal, reflection, empowerment, and action (Nutbeam 2000, 2008, Chinn 2011, Abel and Benkert 2022).	Predominantly qualitative and relationally descriptive; limited standardized tools for measuring conversion of distributed competencies into action (Edwards et al. 2015, Muscat et al. 2022).	Designed for qualitative episode analysis and future quantitative operationalization through A–F–R–S, event-level outcomes, ego-network grids, and item development (Rüegg and Abel 2022, Bak et al. 2025).
Contribution to health equity	Highlights how health literacy is shaped by power, structural inequality, social determinants, and action opportunities (Nutbeam 2008, Chinn 2011, Abel and McQueen 2013, Abel and Benkert 2022, Sykes et al. 2025).	Shows that health literacy is not only an individual attribute but may also be shared across social ties and networks, with implications for relational resources and unequal support (Vassilev et al. 2013, Edwards et al. 2015, Muscat et al. 2022).	Explains relational and organizational barriers and facilitators in the conversion of competencies into action, including partial, delayed, redirected, and blocked conversion (Sen 1999, Rook 2015, Adebahr 2022, Bak et al. 2025).

CHL is represented here through both classic and recent work on cognitive/skills development, critical appraisal, reflection and action (Nutbeam 2000, 2008, Chinn 2011, Abel and Benkert 2022, Sykes et al. 2025). DHL is represented through work on socially distributed health literacy and networked self-management (Vassilev et al. 2013, Edwards et al. 2015, Muscat et al. 2022). CNC is used as a relational conversion framework that specifies how social ties become consequential through Activation, Function-specific Help, Recognition, and Substitution (Bak et al. 2025). The frameworks are adjacent but not equivalent.

Materials and methods

Study design and analytical approach

This qualitative, theory-informed study used an abductive approach to refine the emerging CNC framework (Timmermans and Tavory 2012). Abduction was appropriate because the aim was not to test predefined hypotheses, but to move between empirical material and conceptual development in order to clarify how social ties condition the conversion of health-related competencies into action. CNC served as a set of sensitizing concepts (Blumer 1954), while the analysis remained open to empirical variation and unexpected cases.

The reporting of the qualitative component was informed by COREQ, the Consolidated Criteria for Reporting Qualitative Research checklist for interviews and focus groups (Tong et al. 2007). Reflexive memo writing, documentation of analytical decisions, and peer debriefing supported methodological transparency.

Participants and recruitment

Participants were 12 adults living with multimorbidity, defined as having two or more chronic or long-term conditions. Recruitment followed two pathways. Approximately 15 eligible adults were approached by staff at three Danish municipal health promotion and rehabilitation services, where health centre staff informed them about the study and facilitated contact with the research team; eight agreed to participate and completed an interview. The exact number approached through the municipal services was not systematically recorded.

Four participants were recruited through follow-up contact with respondents from a small exploratory survey conducted in March 2025 in a closed Danish Facebook group for people living with type 2 diabetes and at least one additional chronic condition. Access was granted by the group administrator, participation was voluntary, and the survey included an optional question asking whether respondents could be contacted for a follow-up interview.

The Facebook survey generated 65 responses, many of which were partial. Five respondents consented to follow-up contact; all five were contacted by the first author, four completed a telephone interview, and one did not complete the interview. Because the invitation was posted openly in a closed group, the number of people who saw it could not be determined and no formal response rate could be calculated. Participant characteristics and empirical healthcare episodes are summarized in Table 2 at the beginning of the Results section.

Table 2.

Participant overview and illustrative health-system navigation episodes.

Pseudonym	Age group/gender	Condition/living situation	Main support context	Illustrative health-system navigation or care-coordination episode
Peter	60+/male	Diabetes, heart disease; living alone	Limited trusted support/socially costly help-seeking	Avoided sharing illness-related concerns; limited trusted help for follow-up.
Maria	50+/female	Diabetes, hypertension; living with spouse	Support available but functionally misaligned	Inconsistent diabetes/weight-management support; family and peer food practices could undermine dietary intentions.
Hans	50+/male	Depression, heart disease; living alone	Targeted informational support	Targeted informational support was relevant for understanding advice or navigating follow-up tasks.
Camilla	40+/female	Diabetes, chronic pain; living alone	Sporadic family involvement	Relied on family members to read institutional letters; follow-up depended on their availability.
Karen	60+/female	Chronic obstructive pulmonary disease (COPD), anxiety; living with spouse	Spouse support for medication routines	Handled many practical matters herself; municipal activities during work hours limited attendance.
Anna	50+/female	Diabetes, depression; living alone	Limited trusted support/socially costly help-seeking	Withheld concerns because involving relatives risked judgement, conflict, or being overruled.
Poul	50+/male	Diabetes, hypertension; living with spouse	Digital/proxy support constrained	Digital task advice did not become practical help; institutional/digital rules constrained proxy involvement.
Dennis	50+/male	Chronic pain, anxiety; living alone	Limited trusted support/socially costly help-seeking	Difficulty asking for help left potentially available support unused.
Mette	40+/female	Hypertension, autoimmune disease; living with spouse	Active family support	Sister/daughter helped with letters, calls or follow-up, but support remained administratively invisible and availability-dependent.
Julia	20+/female	Asthma, anxiety; living with family	Support available but functionally misaligned	Family support was available but not always matched to the concrete healthcare or navigation task.
Sisse	40+/female	Diabetes, obesity; living with spouse	Peer support	Peer support was relevant for diabetes-related information, motivation, and everyday health-management tasks.
Henrik	60+/male	Diabetes, cardiovascular disease; living alone	Limited trusted support/socially costly help-seeking	Family support was not activated in diabetes management; digital/administrative tasks could accumulate when capacity was low.

Episodes are illustrative rather than exhaustive. ‘Main support context’ is a qualitative, descriptive summary of the most salient support pattern or relational constraint discussed in the interview; it is not a formal scale of network involvement. Several participants described more than one episode.

Sampling followed principles of thematic sufficiency and maximum variation (Patton 2015), seeking heterogeneity across age, gender, education, living arrangements, employment status, and social/material vulnerability contexts. Social vulnerability was treated as a contextual sampling and analytical criterion rather than a fixed individual attribute. It was assessed through recruitment pathways and interview information on living situation, education and work history, economic strain, support structures, digital navigation, treatment burden, and experiences of fragmented or difficult access to recognized help.

Data collection

Data were collected through semi-structured in-depth interviews conducted face-to-face or by telephone according to participants’ preferences and health status (Kvale and Brinkmann 2009). Eight interviews took place in participants’ homes or another private setting chosen by the participant, and four follow-up interviews with Facebook survey respondents were conducted by telephone. Interviews lasted 30–60 minutes.

Interviews were conducted by the first author and one co-author, Gitte Thybo Pihl. The first author conducted the four telephone interviews and four in-home interviews; Gitte Thybo Pihl conducted four in-home interviews. Neither interviewer had a prior personal, clinical or care-related relationship with participants. Municipal staff assisted recruitment but did not participate in interviews. For Facebook-recruited participants, prior contact was limited to survey participation and voluntary consent to follow-up contact.

Interview guide

The interview guide explored experiences with health-related information, decision-making, everyday healthcare tasks, interactions with professionals and the involvement of social relations in managing such tasks. It was intentionally broad because negative and ambivalent ties could only be meaningfully identified in relation to participants’ wider relational context. Interviews therefore invited accounts of supportive, neutral, negative and ambivalent ties, while the analysis reported here focuses on conversion barriers: episodes in which ties delayed, redirected, constrained or blocked the translation of competencies into action.

Consistent with the definition introduced in the Introduction, the empirical analysis focused on bounded health-system navigation and care coordination tasks, including booking, attendance, digital or written correspondence, contact with professionals, transport or accompaniment, treatment instructions, medication changes, and follow-up. Broader lifestyle behaviours were not analysed as core empirical outcomes, except where participants linked them to a concrete healthcare instruction, follow-up task or service-navigation episode.

The interview guide was not formally pilot-tested as a separate procedure. It was tested and refined during the first three interviews, leading to minor adjustments in wording, sequencing and follow-up probes concerning concrete episodes, negative and ambivalent ties, institutional recognition and proxy/substitution practices. Because the core topics were unchanged, the first three interviews were retained.

Analytical procedure: data analysis and coding

Analysis followed a multi-stage abductive thematic process (Braun and Clarke 2006, Timmermans and Tavory 2012). First, all transcripts were read closely by the first author to identify episodes in which health-related knowledge, intentions or professional advice did not lead straightforwardly to observable action. Episodes were coded descriptively using in vivo phrasing where possible, with attention to situational context, relational dynamics and microprocesses shaping action or inaction.

Second, descriptive codes were synthesized into broader interpretive themes. The four CNC mechanisms—Activation, Function-specific Help, Recognition and Substitution—served as sensitizing concepts (Blumer 1954, Bak et al. 2025). Coding combined deductive sensitizing codes derived from CNC with inductive codes capturing negative, ambivalent and institutionally constrained relational processes that did not fit neatly into the initial framework.

Constant comparison within and across cases refined the boundaries of each mechanism. The purpose was not to allocate each episode to one exclusive category, but to identify which mechanisms were involved and how they interacted in the conversion, delay, redirection, or blocking of action. Selected transcripts, coded excerpts and emerging interpretations were discussed with co-authors during peer debriefing sessions, with particular attention to ambiguous cases.

Ambiguous cases were handled through memoing and return to the full interview context rather than by forcing excerpts into a single category. Where relevant, excerpts were coded as involving more than one mechanism. Negative and ambivalent ties were identified in accounts characterized by anticipated judgement, emotional strain, inconsistent or misaligned support, withdrawal, competing health logics, misinformation, dependency concerns or institutional disregard (Rook 1984, 2015, Adebahr 2022). Analytic memos documented decisions, uncertainties, alternative explanations and changes to the coding framework.

Reflexivity and analytic quality

Reflexivity was important because the study both applied and refined an emerging framework developed by the first author. The first author’s background in sociology, HL and social inequality informed sensitivity to relational and institutional dynamics, but also created a risk of confirming CNC too readily. To address this, CNC was treated as a sensitizing framework rather than a fixed coding template. Peer debriefing with co-authors challenged premature interpretations, tested whether excerpts supported proposed mechanisms and helped distinguish relational barriers, institutional conditions and their intersection.

Data management and equity, diversity and inclusion (EDI) considerations

Audio recordings were transcribed verbatim. Transcripts were pseudonymized, and identifying details such as names, places, workplaces, and other potentially recognizable information were removed or altered where necessary. Audio files, transcripts and analytic notes were stored securely in accordance with institutional data protection procedures and General Data Protection Regulation (GDPR). Access to identifiable material was restricted to the research team. Because the interview material contains personal and potentially identifiable information, transcripts are not publicly available. Non-identifiable methodological documentation, such as the interview guide and coding framework, may be made available upon reasonable request, subject to data protection requirements.

EDI considerations were addressed through flexible recruitment and interview procedures. Interviews were adapted to participants’ preferences and capacities, including in-home or telephone interviews, flexible timing and adaptation to fluctuating health and energy. Particular care was taken when discussing dependence on others, shame, conflictual family relations, digital barriers, illness burden and difficulty navigating services. The aim was to make participation accessible and ethically safe for participants who might otherwise be excluded from qualitative health promotion research.

Ethical considerations

The study adhered to national ethical standards for qualitative health promotion research. According to Danish rules on notifiable health research projects, this non-interventional interview study did not require review by a regional research ethics committee. All participants received written and oral information and provided informed consent prior to participation. Participation was voluntary, and participants were informed that they could pause, decline to answer questions, or withdraw at any time. Interviews involving sensitive descriptions of negative or strained relations were conducted with attention to participants’ emotional well-being and to minimizing discomfort.

Results

Sample overview and empirical focus

Before presenting the four CNC mechanisms, Table 2 provides an overview of the qualitative sample and the types of healthcare episodes discussed in the interviews. The interviews focused on bounded health-system navigation and care coordination tasks, including booking or changing appointments, attending consultations, responding to digital or written correspondence, contacting professionals, arranging transport or accompaniment, understanding treatment instructions, medication changes and completing follow-up tasks.

Through the interviews, participants described different forms of support, including emotional reassurance, informational interpretation, digital or administrative help, transport or practical assistance, accompaniment and proxy or co-action. The results below focus on episodes in which negative or ambivalent ties, sometimes in interaction with institutional conditions, delayed, redirected, constrained or blocked the conversion of health-related competencies into action.

The results are organized around the four CNC mechanisms: Activation, Function-specific Help, Recognition and Substitution. The mechanisms are presented separately for analytical clarity, but empirical episodes often involved several mechanisms simultaneously.

Analytical orientation

Across the interviews, participants described several forms of support: emotional reassurance, informational interpretation, practical help, digital or administrative assistance, transport or accompaniment, and proxy or co-action. These forms of support were not automatically equivalent to conversion capacity. Emotional support could be present without enabling a specific task; informational help could be useful but insufficient without practical follow-up; and digital or proxy assistance could enable action while remaining constrained by institutional rules, consent procedures, or availability of helpers.

The analysis therefore distinguishes between three types of conversion constraints. ‘Relational barriers’ refer to constraints arising within social ties, such as anticipated judgement, shame, conflict, dependency concerns, misinformation, or reluctance to burden others. ‘Institutional conditions’ refer to organizational, administrative, or digital arrangements that make support difficult to recognize or enact. ‘Relational–institutional intersections’ occur when a social tie is available or activated, but its conversion value depends on whether the healthcare or municipal system accommodates, authorizes, or recognizes the helper.

Activation: suppressed mobilization of relevant ties

Participants described situations in which they refrained from contacting family members or acquaintances despite recognizing a need for support. Non-mobilization was rarely due to the absence of a network; rather, it reflected concerns about relational risks, including anticipated conflict, judgement, stigma, or burdening others. See Table 2 for an overview of participants’ characteristics, including the pseudonyms used.

Henrik explained why he avoided discussing diabetes management with family:

I do not really talk to my family about it. They have strong opinions, and then it becomes a discussion instead of help.

Here, mobilization was withheld because involvement was expected to generate friction rather than facilitate action. Activation depended on relational safety rather than network availability.

Anna similarly described withholding concerns:

They always think they know better. So, I keep it to myself, even when I am unsure what to do.

Avoidance served as a strategy to prevent relational strain, yet it limited access to potential sources of support.

Institutional norms also shaped activation. Poul hesitated to seek help with digital booking:

I should probably have asked someone, but I did not want to seem stupid. You are supposed to be able to handle those things yourself.

Norms of self-reliance rendered help-seeking socially costly. When mobilization was suppressed, competencies remained unused despite being present.

Declining or not taking up available social support

In some interviews, participants described situations in which support from family members or others in their social networks was available or offered but not taken up. Help was declined or avoided for different reasons, including norms of self-reliance, reluctance to involve others, or previous negative experiences.

Camilla, a woman living alone with multiple chronic conditions, described how she had historically avoided accepting help even when it was available. She explained that she had long positioned herself as someone who managed on her own:

I’ve always been the kind of person who tries to please everyone and never takes the help that’s offered.

Only later in her illness trajectory did she begin to accept support, for example, by bringing a family member to medical appointments.

Similarly, Karen explained that although help from others was available, she typically handled practical matters herself: ‘I’m just good at managing on my own’.

She reflected that she might sometimes have benefited from asking for help but generally chose not to involve others.

For Dennis, living alone, the reluctance to seek support was linked to personal norms of independence. Although others around him were willing to assist, he explained that he found it difficult to ask for help: ‘I have difficulty asking for help and getting others involved’.

Peter described deliberately avoiding sharing illness-related concerns despite having acquaintances around him:

I don’t have anyone here I can confide in.

Analytically, Activation indicates that network availability and conversion capacity are not the same. Several participants had potential helpers, but help remained latent because asking involved social costs. Anticipated judgement, shame, fear of conflict, norms of self-reliance, or concern about burdening others suppressed mobilization before any practical help could occur, as illustrated by Henrik’s avoidance of family discussions, Anna’s withholding of concerns and Poul’s hesitation to ask for help with digital booking. This means that conversion could fail even when emotional, informational, or practical support existed somewhere in the network. The primary barrier in these episodes was relational: the social tie was present, but not safe or legitimate to activate when needed. Camilla’s and Karen’s accounts of not taking up available help further show that Activation may fail not only because support is absent, but because accepting support conflicts with self-reliance, prior relational experience, or identity. In CNC terms, competencies remained individualized because the first relational conversion step did not take place.

Function-specific Help: misaligned or contradictory support

When ties were activated, support did not always correspond to situational needs. Functional mismatch emerged when emotional reassurance or general advice replaced task-specific assistance.

Karen described motivational comments that did not translate into practical help:

They say things like ‘you just have to pull yourself together’ or ‘try harder.’ However, that does not help when you are the one who has to deal with it every day.

Relationally supportive intentions did not align with practical demands.

Poul recounted asking for help navigating digital systems:

He told me how easy it was and what I should do, but he did not actually help me do it. So, it just stayed undone.

Advice substituted for concrete assistance, leaving tasks incomplete.

Poul described emotionally affirming responses that conflicted with clinical guidance:

They said it was not that bad and that I should not make a big deal out of it. So, I did not go back for a while.

Here, support shaped actions that diverged from professional recommendations. Effective enactment depended on alignment between the provided and required support.

In many accounts, relational support took forms that were emotionally supportive or socially meaningful but did not facilitate concrete health actions.

These episodes illustrate why general support is not necessarily useful support. Participants could receive encouragement, advice, or reassurance while still lacking the assistance required by the task. Emotional support was sometimes valued but insufficient for digital navigation, transport, appointment booking, form completion, or follow-up. Informational advice could also become misaligned when it told participants what they should do without helping them complete the task. The analytical point is therefore that support must match the practical, cognitive, digital, or emotional demand of the episode. When this fit was absent, action could remain incomplete despite motivation and apparent support.

Maria, a woman with diabetes and hypertension, described how everyday social interactions sometimes conflicted with her intentions to follow dietary advice. Although she discussed healthy eating with a close friend who also had diabetes, their interactions often undermined these intentions:

We talk about eating healthy… but then he shows up with two cream cakes from the bakery.

Similarly, family support could conflict with dietary management. Maria explained how her children continued preparing traditional foods despite her diagnosis:

Even after I told them I had diabetes, they said: ‘Are we having creamy potatoes this weekend?

In these situations, relationships provided companionship and emotional support but simultaneously produced practical obstacles to behaviour change.

Functional mismatch also occurred where participants received advice without corresponding assistance. Maria described asking her doctor for help managing her weight and diabetes, but felt the response remained limited to a referral rather than practical guidance:

I asked the doctor if she would help me… and she could only offer me a meeting with the dietitian.

Participants also described cases where health-related programmes did not align with everyday constraints. Karen, who was still working full-time, explained that several municipal programmes were offered during working hours and therefore remained inaccessible:

They offered different activities, but they were during my working hours.

Here, support structures existed but were not functionally aligned with participants’ practical circumstances. As a result, available support did not translate into concrete assistance.

Across interviews, functional specificity therefore depended not only on the presence of support but also on whether the type, timing, and form of assistance corresponded to participants’ everyday needs.

Recognition: institutional accommodation of informal support

A third mechanism concerned whether informal support was acknowledged within institutional encounters. Even when ties were mobilized and functionally relevant, their contribution could be limited if organizations failed to accommodate them.

Mette attended consultations with her daughter:

She joins me because I forget things, but the staff still speak only to me and hurry. Later, we realise important details were missed.

Although support was present, institutional communication practices restricted its effectiveness.

Mette described assistance with correspondence:

My sister helps me with the letters and calls, but officially, it is still just me. So, when something goes wrong, it is my responsibility.

Support remained administratively invisible.

Digital systems further constrained relational involvement. Poul noted:

They say you can get help, but then everything must be done in your own name. So even if someone helps you, it does not really count.

Recognition operated at the interface between interpersonal relations and organizational procedures. Where accommodation was limited, the practical value of activated support diminished.

Recognition marks the point where relational support becomes dependent on organizational response. In these episodes, the primary issue was not always that the social tie was negative. Rather, support lost conversion value when it remained administratively invisible, was not accommodated in communication, or could not be incorporated into procedures. This represents a relational–institutional intersection: the helper may be activated and functionally relevant, but the organization still treats the individual patient as the only legitimate actor. Recognition, therefore, links CNC directly to organizational HL, because services shape whether informal helpers can become legitimate, consent-based, and practically usable sources of support in care processes (Brach et al. 2012, Farmanova et al. 2018). The conversion value of informal support depended in part on whether services enabled the involvement of helpers in legitimate, clear, and consent-based ways. This implication is developed further in the Discussion.

Substitution: fragile or blocked acting on behalf

Substitution occurred when others acted directly on behalf of individuals with limited capacity. In several accounts, substitution enabled follow-up but remained contingent.

Henrik described periods of cognitive and emotional exhaustion:

There are days when I just cannot deal with it. Then things pile up, and nothing happens.

In such moments, acting on behalf was necessary to sustain engagement.

Mette described reliance on her daughter:

She helps when she can, but she has her own life. If she is busy, then it just stops.

Substitution depended on availability rather than stable arrangements.

Institutional rules also restricted substitution. Poul explained:

They say someone can help you, but then they will not talk to them anyway. So, in the end, it still has to be me.

Digital authentication and consent requirements limited third-party involvement. Substitution was therefore situational and vulnerable to interruption.

Camilla described relying on family members when written communication from institutions became difficult to understand:

If my sister doesn’t have time, then I ask my niece: ‘Can you help me read this? I simply do not understand it’.

Such assistance allowed administrative tasks to be completed but depended on the availability of relatives and, therefore, remained informal and unstable.

Substitution was the most concrete form of relational conversion because another person temporarily carried out, co-produced or stabilized the required action. This included reading letters, managing digital correspondence, booking, accompanying, contacting services, or helping complete administrative tasks. However, substitution was also fragile. It depended on the helper’s availability, the participant’s willingness to delegate, and whether institutional or digital systems allowed proxy or co-action. The barrier was therefore often located at the intersection of relational dependence and institutional authorization. Substitution enabled action when individual capacity was insufficient, but it also exposed how health-system navigation may become contingent on informal, unstable, and sometimes unrecognized help.

Accumulated relational constraints

Across interviews, these mechanisms rarely appeared in isolation. Participants often described situations where suppressed activation, functionally misaligned support, and limited institutional recognition co-occurred. Consequently, participants frequently possessed relevant health knowledge and expressed intentions to follow professional advice, yet translating these competencies into concrete actions depended on relational and institutional conditions. Variation in observable practices, therefore, reflected differences in conversion conditions rather than differences in expressed competence.

Table 3 analytically separates the four CNC mechanisms for clarity. This separation should not be read as implying that the mechanisms occur in isolation. In the empirical material, conversion barriers often co-occurred within the same episode. For example, a participant might hesitate to activate help because of shame, receive support that did not match the task, and still face limited institutional recognition if a helper became involved. The table therefore summarizes the dominant analytical mechanism in each row while the surrounding text illustrates how mechanisms intersect in concrete episodes.

Table 3.

Negative and constraining relational mechanisms in the conversion of health-related competencies into action.

CNC mechanism	Relational condition observed in interviews	Typical empirical pattern	Illustrative interview evidence	Consequences for health action
Activation (A)	Suppressed mobilization of available ties	Participants avoided contacting family or acquaintances due to anticipated conflict, judgment, or norms of self-reliance.	‘I do not really talk to my family about it … it becomes a discussion instead of help’ (Henrik); ‘I am just good at managing on my own’ (Karen)	Competencies remain unused because support is not mobilized.
Function-specific Help (F)	Misalignment of available support and situational needs	Support took the form of encouragement or general advice rather than practical assistance.	‘He told me how easy it was … but did not actually help me do it’ (Poul)	Tasks remain incomplete despite expressed motivation.
Recognition (R)	Institutional practices fail to accommodate informal support	Relatives attended consultations or helped with administration, but were not formally acknowledged in communication or procedures.	‘My sister helps me with the letters, but officially it is still just me’ (Mette)	Informal support remains administrativey invisible and less effective.
Substitution (S)	Others act on behalf of participants when capacity is limited	Family members temporarily managed tasks such as understanding letters or attending appointments.	‘If my sister does not have time, I ask my niece to help me read it’ (Camilla)	Action becomes dependent on the contingent availability of others.

Mechanisms are separated analytically for clarity; in practice, several conversion barriers may co-occur within the same healthcare episode.

Discussion

Main findings: Compensatory Network Capital and the action dimension of Critical Health Literacy

This study contributes to CHL research by specifying a relational and institutional conversion layer between health-related competencies and concrete action. CHL has moved HL research beyond functional information processing towards critical reflection, judgement, empowerment, power, inequality, and action (Nutbeam 2000, 2008, Chinn 2011, Sykes et al. 2013, Abel and Benkert 2022). The present findings, however, suggest that CHL’s action dimension requires closer attention to the conditions under which competencies can be enacted in everyday healthcare episodes.

Participants often described knowledge, motivation, and awareness of recommended care, but action was delayed, redirected, or blocked when support was difficult to activate, poorly matched to the task, insufficiently recognized by services, or unavailable as a substitute when individual capacity was insufficient. Negative and ambivalent ties were therefore not only sources of psychosocial strain. They functioned as practical barriers to conversion, affecting help-seeking, digital navigation, follow-up, and care coordination.

Implications for Critical Health Literacy

The findings have three implications for CHL. First, they suggest that CHL’s action dimension should be understood as a conversion problem, not only as a competence or awareness problem. Nutbeam’s distinction between functional, interactive, and CHL moved the field beyond basic comprehension towards empowerment, participation, and critical engagement with social determinants of health (Nutbeam 2000, 2008). Chinn’s review further emphasized that CHL concerns the social and political conditions under which people interpret, assess, and act on health-related information (Chinn 2011). CNC extends this action-oriented tradition by specifying relational and institutional conditions that enable, constrain, delay, or block behavioural implementation.

Recent CHL scholarship also points to this unresolved implementation question. Reviews of CHL and health behaviour outcomes indicate that the field still has limited empirical knowledge about how reflection, judgement, and critical awareness translate into observable behaviours (Færevaag et al. 2026). Work on multidimensional HL and medication adherence further suggests that trust may mediate or moderate the extent to which competencies translate into action (Inanaga et al. 2024). The present study adds that trust is not only a patient–professional issue. It is also relational and institutional: people may trust advice, but still not act if help-seeking is socially costly or if helpers are not recognized in digital and organizational systems.

Second, CNC highlights that social ties should not be treated only as supportive resources for CHL. They are analytically ambivalent conversion conditions. A relation may enable action in one episode, remain unavailable in another, or constrain action through judgement, shame, conflict, misinformation, dependency concerns, or competing health logics. This extends CHL by showing how power and inequality also operate in everyday relational situations in which asking for help, accepting help, or delegating action may be socially risky.

Third, the findings suggest that CHL research should more clearly distinguish among competencies, conversion conditions, and realized action. A person may possess knowledge, critical awareness, or motivation, but still be unable to complete a concrete health-system navigation task. Conversely, action may become possible through relational and organizational arrangements that compensate for limited capacity, digital barriers, or illness burden. While CHL foregrounds reflection, power, and inequality, and DHL highlights the distribution of competencies across networks, CNC specifies the mechanisms through which such competencies become actionable, partial, delayed, redirected, or blocked in concrete care situations (Edwards et al. 2015, Muscat et al. 2022).

Organizational and digital health literacy: proxy action, consent, and privacy

The findings also raise a critical issue at the intersection of digital health literacy and organizational HL (Norman and Skinner 2006, Brach et al. 2012, Farmanova et al. 2018, Wamala-Andersson et al. 2026). In several episodes, Substitution involved informal or proxy assistance with digital messages, appointment booking, letters, patient portals, or administrative follow-up. In the Danish context, such tasks may also involve national digital identification systems such as MitID. Proxy assistance may enable action when individual capacity, digital confidence, or illness burden is insufficient, but it may also involve access to sensitive information and raise questions about consent, privacy, autonomy, and delegated digital authority (Wolff et al. 2022, Salmi et al. 2024).

CNC should therefore not be read as recommending informal workarounds, shared credentials or unregulated transfer of responsibility to relatives. Rather, the framework points to organizationally supported, consent-based, and revocable forms of assisted digital access. Services should clarify when and how informal helpers may be legitimately involved, how patients’ autonomy and privacy are protected, and what alternatives exist for people without trusted helpers. This is especially important in multimorbidity and social vulnerability, where digital systems may intensify dependence on informal support while making such support difficult to recognize or authorize.

Social vulnerability and conversion failure

Social vulnerability shaped the conversion environment in which CHL-related competencies were enacted. It did not simply mean fewer resources. It often meant that asking for help was more costly, support was more fragile, trusted helpers were overburdened or unavailable, and institutional encounters were more difficult to navigate. Social vulnerability may also weaken relational and institutional trust: previous experiences of judgement, conflict, fragmented care, or administrative difficulties may make it harder to ask for help, accept support, or believe that services will recognize informal helpers in legitimate ways.

A particularly important finding is the accumulation of relational and institutional conversion constraints. Activation barriers, functional mismatch, lack of recognition, and fragile substitution rarely appeared in isolation. Rather, they accumulated across interactions and care episodes, increasing the likelihood that knowledge, intentions, and professional advice would not be enacted. This suggests a relational form of inequality in the enactment of CHL: unequal access to the social and organizational conditions that enable action.

This interpretation resonates with studies of bridge-building and accompaniment. Ethnographic work from ‘Seeing the Invisible’ shows that people with formal entitlements may still struggle to access services when relational scaffolding is missing (Tonnesen et al. 2024). Evaluations of the Social Health Bridge-Building Programme similarly suggest that structured accompaniment and legitimized questioning can improve follow-up and coordination (Valentin et al. 2023). Such studies indicate that relational conversion capacity may be modifiable, but they also underline that support must be organized and recognized rather than assumed to exist in private networks. For this reason, social vulnerability should be treated as a conversion environment rather than as a simple participant characteristic.

As Table 1 summarizes, CNC is complementary to adjacent frameworks rather than a replacement for them. Candidacy theory conceptualizes access to healthcare as negotiated between individuals and services (Dixon-Woods et al. 2006). Treatment burden and illness-work perspectives illuminate the workload delegated to patients and their networks, as well as the capacity required to carry it out (Vassilev et al. 2013, May et al. 2014). DHL explains how health-related competencies are shared and mobilized across social networks (Edwards et al. 2015, Muscat et al. 2022). Social support research distinguishes between the structures, functions, and processes of support (House et al. 1988), while organizational HL highlights service-level accessibility and responsiveness (Brach et al. 2012, Farmanova et al. 2018). CNC adds a narrower mechanism vocabulary for analysing when these resources and conditions become, or fail to become, concrete action in specific healthcare episodes.

Implications for health promotion practice

For municipal health promotion and social prescribing, the findings suggest shifting attention from motivating individuals alone to identifying barriers to relational and institutional conversion. This may involve safe mobilization of support, task-aligned navigation assistance, formal involvement of relatives or peers, and accompaniment structures modelled on bridge-building initiatives (Valentin et al. 2023). Offers should be assessed not only for availability, but also for relational safety, task fit, and organizational recognition. For digital infrastructures, delegated digital authority, navigator roles, and simpler proxy pathways may strengthen Substitution, provided that privacy, autonomy, and consent are protected.

Future research

Future research should examine whether CNC mechanisms operate similarly across different forms of health-related action. The present study focused on bounded health-system navigation and care coordination tasks, such as booking appointments, attending appointments, digital correspondence, treatment instructions, medication changes, and follow-up. Broader forms of health behaviour, including medication adherence over time, physical activity, diet, smoking cessation, and everyday self-care routines, were not analysed as core empirical outcomes. Future studies should examine whether these different forms of action require different configurations of Activation, Function-specific Help, Recognition, and Substitution.

Future CHL research could use CNC to examine the behavioural implementation of reflection, judgement, and critical awareness, including whether people have access to relational and organizational conversion conditions that allow competencies to become action. Longitudinal and mixed-methods designs are relevant because repeated successes and failures in conversion may shape trust, help-seeking, perceived agency, and willingness to mobilize support. Survey-based CNC items, ego-network grids, and event-level indicators could also be used to test whether cumulative relational constraints are associated with observable outcomes such as booking, attendance, digital response, and follow-up.

Future studies should also compare recognized and unrecognized forms of proxy support. In highly digitalized systems, it would be important to examine whether authorized helpers, peers, or municipal navigators are more likely than informal, unrecognized helpers to complete tasks such as booking appointments, accessing portals, or following up. Such work could clarify when Recognition and Substitution become protective, when they create privacy or autonomy risks, and when people without trusted helpers require alternative forms of professional or community-based relational support.

Limitations

The study has several limitations. First, CNC offers a focused micro–meso lens and does not, by itself, capture broader macro-structural determinants such as welfare reforms, labour market exclusion, housing conditions, or healthcare financing. Relational conversion processes are embedded in these wider contexts and should be analysed alongside them (Link and Phelan 1995, Marmot 2010). CNC should therefore be understood as a delimited framework for analysing one mechanism-rich layer of health inequality, not as a total theory of health inequality.

Second, the qualitative design and recruitment through municipal services may have underrepresented people who were most disconnected from formal systems. Third, the findings are based on self-reported and retrospective accounts and cannot establish prevalence or causal direction. The findings should be interpreted as experienced mechanisms rather than observed interactional sequences. Fourth, CNC was used as a sensitizing framework in an abductive analysis; further qualitative, quantitative and mixed-methods studies are needed to assess its transferability, measurement validity, and practical utility.

Finally, the study did not directly observe clinical encounters, digital portal use or interactions between participants and helpers. The analysis therefore cannot determine how professionals interpreted informal support, nor how helpers themselves understood their role. Including helpers and professionals in future studies would allow a fuller analysis of the relational–institutional intersection, including disagreements about responsibility, consent, and appropriate support.

Conclusion

The findings suggest that CHL-related competencies are insufficient unless they can be enacted through relational and institutional conversion conditions. Adults living with multimorbidity in contexts of social vulnerability often possessed knowledge, motivation, and awareness of recommended care. However, action depended on whether support could be safely mobilized, matched to the task, recognized institutionally and, where necessary, enacted with or on behalf of the person.

CNC helps specify this conversion layer. It indicates how social ties and organizational arrangements may shape whether critical awareness, knowledge, motivation, and professional advice translate into concrete action, and how negative or ambivalent ties may delay, redirect, or block that process. The contribution is therefore not to replace CHL, but to make its action dimension more empirically visible and analytically precise under unequal relational and institutional conditions.

Author contributions

Carsten Kronborg Bak (Conceptualization, developer of CNC framework, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Visualization, Writing—original draft, Writing—review & editing), Tina Junge (Formal analysis, Project administration, Supervision, Writing—review & editing), Henrik Vardinghus-Nielsen (Conceptualization, Formal analysis, Validation, Writing—review & editing), Heidi Klakk (Formal analysis, Supervision, Validation, Writing—original draft, Writing—review & editing), Gitte Thybo Pihl (Formal analysis, Investigation, Writing—review & editing), and Henrik Bøggild (Formal analysis, Investigation, Supervision, Validation, Writing—review & editing). All authors reviewed and approved the final manuscript and agree to be accountable for the work.

Conflicts of interest

None declared.

Funding

This work was supported by Helsefonden [grant number 24-B-0180].

Data availability

The qualitative interview data (audio recordings and verbatim transcripts) contain personal and potentially identifiable information and therefore cannot be shared publicly. A de-identified codebook/metadata (e.g. coding framework, variable dictionary, and interview guide) and non-identifiable methodological documentation may be made available upon reasonable request to the corresponding author, subject to data protection requirements, institutional approvals, and a data use agreement (as appropriate).

Ethical approval

The study was conducted in accordance with the Declaration of Helsinki. Under Danish regulations governing research ethics review, non-interventional qualitative interview studies of this type do not require notification to or approval by a regional health research ethics committee. All participants received written and oral information and provided written informed consent prior to participation. The study adhered to national ethical standards for qualitative health promotion research. Pseudonyms were used to protect confidentiality, and data were handled in accordance with applicable institutional and data protection requirements.

Disclosure

During revision, the authors used ChatGPT (OpenAI) to support language editing and reference formatting. All scientific content, interpretations and references were critically reviewed and verified by the authors, who take full responsibility for the manuscript.