The recruitment and data collection for this study took place between August and December 2023. Participants were recruited from clinics in a large National Cancer Institute–designated comprehensive cancer center located in Chicago, IL. Recruitment initially occurred through oncologist referral and digital advertisement in a survivorship clinic and a Spanish-speaking clinic to ensure inclusion of diverse patient populations. Subsequently, we expanded the recruitment channels to include the institution’s electronic health record patient portal, which became the primary recruitment channel. A message was disseminated to patients receiving cancer care through the patient portal inviting them to indicate interest in learning more about the study. Participants who expressed interest were then contacted via email with an introductory message and a link to complete the consent form and screening survey on REDCap (Vanderbilt University Medical Center). The survey served as a screening tool to identify participants eligible for follow-up interviews.

Eligibility for interview participation was determined based on predefined criteria. From the survey responses, participants were sampled for the semistructured interview if they (1) had a cancer history; (2) reported at least 2 health risk behaviors, defined as current smoking (every day or some days), BMI≥25 kg/m², or engaging in less than 150 minutes of physical activity per week; (3) being 18 years or older and proficient in English; and (4) indicated interest in health behavior change (responding yes or maybe to intention items), meaning they were considering at least 2 of the following health risk behaviors: smoking, insufficient physical activity, and poor quality diet.

The survey responders who met the inclusion criteria and consented to be contacted for an interview were contacted via email and invited to participate in the semistructured interviews. A total of 16 participants were enrolled and completed the semistructured interviews. Participant recruitment continued until no new insights were observed. All interviews were completed and transcribed prior to analysis. Preliminary analysis informed code development, and coding proceeded until data saturation was reached [78], and we no longer identified additional novel information.

Two senior researchers (Courtney Scherr and BS) led the design of a structured survey screener to collect demographic data, cancer history, current health behaviors, and participants’ attitudes toward behavior change. The survey included questions on motivation for change, perceived barriers, prior engagement with health promotion resources, and interest in participating in a digital health behavior change program.

The interview guide was collaboratively developed by the research team, incorporating input from oncologists, behavioral scientists, and patient advocates. The team had up to 4 meetings to collaboratively develop and iteratively refine the interview guide. The last meeting served as a pilot test where all the researchers who led the interviews tested the guide to establish familiarity, while other team members observed to assess clarity, relevance, and flow. Feedback from this informal pilot test informed minor refinements to question wording and sequencing prior to full data collection. The final interview guide consisted of 15 core questions covering key topics, including motivation for multiple health behavior change, perceived barriers to behavior change, cancer history and its influence on health behaviors, experiences with existing health promotion resources, and preference for ongoing support during intervention. See in Multimedia Appendix 1 for the interview guide, which contains the interview questions.

Interviews were conducted via the Zoom (Zoom Video Communications) videoconferencing platform, allowing for real-time verbal and nonverbal cue analysis. Each interview lasted between 30 minutes and 1 hour and was recorded. Two researchers were present during each interview: one acted as the primary interviewer, while the second took detailed notes (memos) to capture contextual nuances. Interviews followed a semistructured format, allowing flexibility for interviewers to probe and participants to elaborate on their experiences [79-81]. Participant recruitment continued until we reached data saturation, with no additional novel information emerging from subsequent interviews [78].

This section presents an in-depth exploration of the findings from our thematic analysis across 4 major themes and 12 subthemes to directly address our research questions. In discussing what has worked well in past experiences and their desires in future interventions, we categorized participants’ preferences mapped onto 3 psychological needs from SDT: competence support through patient empowerment, technology-enabled tracking, and expert accessibility; autonomy support through personalized, flexible program design that acknowledges existing knowledge and provides choice; and relatedness support through human-centered systems featuring health promotionists, peer support, and health care team integration.

Although not classified as a primary psychological need, a fourth theme emerged that highlights participants’ desire for integrated, comprehensive care systems that connect behavior change programs with broader health care ecosystems rather than functioning as isolated programs. Textbox 1 highlights the key themes and the subthemes of this study.

Participants consistently emphasized wanting programs that would build their sense of capability and effectiveness through ongoing recognition and reinforcement, comprehensive patient health education on lifestyle management, and technological enablement. These preferences reflected their desire to become more adept at managing their health rather than being passive recipients of care.

In addition to challenges related to competence, several participants described difficulties aligning recommended behavior change strategies with their personal routines and lived contexts. Participants often felt that the guidance they received was too generalized and did not account for the complexities of their daily lives. For example, one participant explained that although they were given behavioral change coaching, the recommendations did not reflect their individual circumstances and made implementation difficult.

This mismatch created a sense that behavior change strategies were externally imposed rather than self-directed, making them difficult to integrate and sustain. As a result, participants discussed that even well-intentioned recommendations were perceived as impractical or overwhelming, limiting participants’ ability to take ownership of their behavior change efforts. Accordingly, our study participants consistently emphasized wanting programs that could adapt to their individual circumstances, knowledge levels, and life constraints rather than requiring rigid adherence to standardized protocols. This preference for autonomy-supportive design reflected their desire to maintain control over their health journey while accommodating the complex realities of survivorship.

Guided autonomy and flexibility emerged as fundamental requirements, as participants rejected programs with inflexible scheduling or mandatory time commitments. Several participants emphasized that rigid time commitments and intervention plans represented significant barriers to program engagement. As the conversations unfolded, participants shared their preference for collaborative programs that could adapt to their complex lives rather than demanding rigid adherence. One participant outlining their preference for exercise programs highlighted that:

Like Participant 2, most participants preferred the weekly distribution of activities over daily requirements. They advocated for starting small with gradual build-up, recognizing behavior change as a journey requiring low-barrier entry points. The participant also framed time allocation as personal exploration rather than an external mandate imposed by the program. The concept of “protected time” suggested that participants wanted programs to help them identify sustainable time allocation strategies. Intervention programs could help participants identify these protected times. Once identified, health and wellness time might become “protected time” that survivors defend against competing life demands, which could facilitate adherence.

Similarly, the importance of guided autonomy and schedule flexibility became reinforced when participants highlighted their desire for flexible re-engagement and how they would prefer the opportunity to pause, resume, and re-enter their behavior change programs, noting that:

The preference for pause-and-resume features reflected the reality that survivors of cancer often navigate multiple health challenges and life changes simultaneously.

Put together, participants showed a preference for programs that offer low barrier entry, gradual progression, flexible scheduling across the week, varied intensity-duration levels for activity scheduling based on individual capacity, flexible reengagement options, and, most importantly, support with identifying what schedule works for each patient.

Autonomy through customization and fostering survivors’ sense of choice emerged as another important preference for some participants. Among the challenges participants reported was the assumption that all participants started from the same knowledge baseline. We found that survivors sought recognition as informed patients with valuable prior learning. They want programs that assess and respect existing knowledge rather than enforcing universal curricula, thereby supporting their need for autonomy. Survivors wanted programs that acknowledged their existing health knowledge and expertise and tailored content based on their experience level. As one participant explained about a workplace weight loss program:

The inability to skip known material transforms education from empowerment to obstacle, wasting time and engagement. This rigid structure and preset learning modules ignored the possible heterogeneous knowledge base among survivors of cancer and their varying educational needs and simply assumed uniform learning needs and pace. This feedback further underscored the need for adaptive interventions that could meet survivors at their current knowledge level.

In addition to acknowledging existing health knowledge, participants further highlighted their preference for choice and personal agency in intervention components. They wanted options for different activity types, duration commitments, dietary intake, and support levels rather than one-size-fits-all approaches. One participant, discussing their preference for choice on the intervention component they would like to receive, emphasized that:

This preference reflected participants’ understanding that survivors of cancer have diverse needs based on their diagnosis, treatment history, health status, and information knowledge. While participants articulated their preference for choice, they also communicated their desire for support from the intervention in making the best choices, denoting an overall interest in shared or negotiated autonomy.

Although not explicitly articulated, our analysis throughout participants’ narratives implicitly depicts that interventions should be designed to avoid controlling language, as participants consistently described preferences for collaborative rather than controlled approaches.

The convenience of virtual participation emerged as important for reducing barriers, particularly for those geographically distant from treatment centers. As one participant narrating their present ordeal with relocation and commute time noted that:

Like Participant 2, most participants expressed the value of virtual options, not as a preference for digital over human interaction, but as a practical solution to address barriers, allowing survivors to access support regardless of their proximity to major medical centers.

We assessed participants’ receptiveness to health promotion support [88] by introducing the concept of a health promotionist, defined as a trained coach who helps individuals make health behavior changes. The responses were positive, with participants expressing interest in receiving this type of support and willingness to share behavior change data with their health promotionist.

Participants highly valued having a dedicated health professional who could offer expertise and hold them accountable in a personalized manner. One participant captured this sentiment, stating:

This highlights a clear desire among participants for access to knowledgeable professionals who can provide tailored guidance instead of generic information or automated responses.

In addition to expert information, Participant 2 further highlighted the relevance of the supportive accountability aspect of health promotionist relationships. They shared that their past experience with human accountability significantly boosted their motivation compared to self-directed efforts. They recalled:

The participant mentioned that their physical fitness trainer would check on them whenever they missed an appointment. This relational accountability fostered a sense of mutual responsibility that the participant found highly beneficial. The human element introduced reciprocal dynamics that enhanced their commitment to health behaviors.

When asked about their preferred way of receiving support from a health promotionist, participants emphasized a preference for human-centered support that enhances relatedness through understanding, acceptance, and practical guidance rather than judgment or control. Describing this supportive approach and relationship they would value, participants emphasized their desire for nonjudgmental support and guidance through setbacks and challenges. One survivor, while describing an ideal interaction, emphasized their interest in receiving supportive responses to difficulties and mentioned that:

Participant 11 wanted the freedom to acknowledge their mistakes without the fear of harsh judgment. They appreciated the idea of a health promotionist recognizing that while their behavior might not be ideal, it can be addressed with understanding. This preference reflected participants’ understanding that sustainable behavior change involves navigating challenges and making mistakes, which requires supportive professionals who can help them and maintain motivation in the face of setbacks.

In addition to support from skilled health coaches, participants discussed their interest in receiving support from their peers. Participants reported finding profound value in connecting with other survivors who understood or could relate to their unique challenges and could provide emotional support through shared experience. One participant described the relief of shared experience:

This validation that Participant 6’s and their friend’s struggles were common rather than personal failures provided emotional support that professional guidance alone may not replicate.

However, while participants recognized value in connecting with other survivors, they emphasized the critical need for skilled facilitation when peer support is offered in group settings to maintain supportive rather than detrimental dynamics. One survivor cautioned about potential unmoderated group challenges, noting that:

Consequently, the participant proposed that interventions should incorporate skilled moderation that could manage group dynamics effectively, saying that: “So, it’s so important for the leader to know how to manage that and to bring it back to a center point for everybody in the group so that you can have fun together” (Participant 7). This preference highlighted participants’ understanding that peer support required structure and guidance to be beneficial rather than harmful.

Another suggestion that participants raised was empathy-centered communication that considers the struggles of cancer survivorship. Most participants preferred empathy and person-centered communication over clinical language. They emphasized the importance of using accessible language when health care professionals communicate with patients. For example, Participant 7, who reported experiencing information overload on receiving a 200-page book at diagnosis, while reflecting on the experience, highlighted the relevance and desire for human-to-human conversations. They mentioned that:

Participant 7 described feeling overwhelmed by the volume of information. The issue was not necessarily the information itself but how it was communicated; specifically, the absence of relational, human-centered communication to contextualize complex information. Participant 7 valued having the same medically accurate and helpful information but delivered through personal, empathetic interaction rather than clinical documentation.

Similarly, another participant recounted a negative and judgmental experience with an oncologist who, according to the participant, strongly emphasized weight management but did so in a manner perceived as dismissive and unhelpful, sharing that:

Furthermore, in highlighting their preferred communication style, another participant recalled an early encounter with a diabetic educator whose framing of behavior change felt condescending. They mentioned that:

Put together, these participants’ accounts emphasize how communication styles perceived as clinical, judgmental, or condescending can create emotional resistance and shape patients’ willingness to engage in future behavior change discussions.

An interesting tension emerged between participants’ autonomy and relatedness needs regarding family involvement, illustrating the complex interplay between psychological needs in intervention design. Participants expressed variable preferences for how they would want their family involvement in their participation, with some participants wanting family engagement, while others preferred keeping their health behavior change participation private. This complex interplay revealed how supporting one psychological need might potentially conflict with another, creating nuanced design challenges for behavior change programs. While some participants viewed family engagement as essential for creating supportive environments that enhanced their sense of connection, others prioritized personal autonomy and privacy in their health behavior change journey.

For example, Participant 7 highlighted the benefits of family involvement, noting the relevance of having their family members enrolled in exercise programs to bolster survivors’ motivation. They shared that:

This participant valued family engagement as a source of connection and ongoing support that would extend beyond formal program boundaries. Conversely, another participant’s experience highlighted the autonomy dimension. Expressing discomfort with family awareness of their participation in past behavior change intervention, they noted that:

For Participant 1, the embarrassment from disclosure and the desire to maintain privacy and personal agency over disclosure outweighed potential relational benefits.

This tension illustrates how supporting one psychological need, in this case, relatedness through family involvement, might potentially undermine another, autonomy through personal choice. Interventions must navigate this complexity by offering flexible approaches to family involvement that could accommodate different comfort levels and family dynamics while respecting participants’ autonomy in deciding their level of disclosure and family engagement.

While the previous subthemes focused on individual psychological needs, participants also expressed distinct preferences that operated at the program and health care system levels, emphasizing the importance of structural integration and comprehensive service coordination. This subtheme represents survivors’ narratives that effective behavior change interventions must function as coordinated components within broader health care ecosystems rather than isolated programs addressing individual psychological needs alone.

Survivors of cancer consistently expressed a desire for behavior change interventions that would seamlessly integrate with their existing health care infrastructure and provide access to comprehensive wellness services. Rather than viewing health behavior change as a stand-alone activity, participants envisioned programs that would function as coordinated components of their overall care ecosystem, connecting them to broader support networks and addressing their multifaceted health needs.

The desire for whole-person care approaches emerged as participants emphasized that their health needs extended far beyond cancer-related concerns. They wanted programs that could address the full spectrum of their health challenges rather than focusing narrowly on cancer survivorship issues. One participant articulated this comprehensive need:

This perspective reflected the reality that many survivors were managing multiple health conditions simultaneously and needed comprehensive care approaches that could address their complete health care needs.

To facilitate their need for whole-person care, participants discussed their interest in accessing complementary wellness services. In addition to physical therapy and nutrition counseling, they wanted connections to various healing modalities, including mental health support and mind-body practices like yoga that addressed the emotional challenges of survivorship. One survivor described their interest in this movement and stress relief practices, saying that:

This preference highlighted participants’ holistic understanding of health and wellness that extended beyond medical interventions to include therapies that could address pain, stress, and overall quality of life.

As part of measures to enhance holistic and comprehensive care delivery, participants shared their desire for continuity and coordination between their behavior change efforts and their ongoing medical care. Survivors wanted their clinical teams to be aware of and involved in their wellness activities to optimize care coordination. Participants also valued health care providers who took holistic approaches to their care, addressing not just medical issues but also supporting their overall well-being. One survivor described their appreciation for comprehensive provider support:

This desire for integrated, whole-person care suggested that participants wanted their behavior change efforts to be supported and reinforced by their trusted health care providers rather than existing in isolation.

Participants wanted their clinical teams to be informed about and involved in their wellness activities. One participant articulated their comfort with data sharing: “I think that them knowing what I am or I’m not working on between visits, I’d be fine to have them like integrated into that process” (Participant 2). This preference reflected participants’ views that behavior change efforts should complement rather than exist separately from their medical care, allowing for better care coordination and more informed clinical decision-making.

The preference for integrated and comprehensive care systems ultimately reflected participants’ understanding that sustainable health behavior change required coordinated support across multiple domains. They envisioned programs that could bridge the gaps between different aspects of their care, connect them to appropriate resources, and provide holistic support that addressed their physical, emotional, and practical needs as whole persons rather than just survivors of cancer.

These findings, ultimately, suggest that survivors of cancer seek more than improved health outcomes when participating in multiple behavior change interventions. They envision programs that respect their autonomy while providing robust support systems. Their preferred intervention considers time constraints by providing flexible, personalized, and convenient support while strengthening relationships with health promotionists and clinicians through coaching, accountability, and ongoing communication. They also emphasized that effective interventions should be integrated into routine clinical care and recognize that successful behavior change depends on addressing survivors’ physical, emotional, and social needs alongside their cancer care. Finally, survivors desire practical education delivered in an empathy-focused and accessible language structure that acknowledges both their accumulated health information and ongoing vulnerabilities in navigating life after cancer.

SDT [64,69,82] is a well-established framework for understanding motivation. Interventions grounded in this theory have proven effective in fostering adoption (initial uptake) and facilitating actual engagement (active participation) in health behaviors among diverse groups, settings, and behavioral activities [61,63,84-86,90]. It identifies the psychological elements and mechanisms that influence motivated behaviors [82], highlighting 2 primary forms of motivation: autonomous and controlled [83]. Changing autonomous motivation encourages individuals to engage in certain behaviors and sustain those behaviors over time, as reported in a meta-analysis review [91], which makes interventions grounded in SDT especially beneficial for health behavior change programs [92,93]. Based on this understanding and from the participants’ perspective, we identified opportunities to facilitate health behavior change and make survivors of cancer more autonomously motivated by designing interventions to align with their needs and preferences.

Our study revealed participants’ interests, support needs, and preferences that primarily mapped onto the 3 basic psychological needs: competence, autonomy, and relatedness, which, according to SDT, are vital psychological needs for promoting autonomous motivation. Autonomy involves feeling a sense of control and personal choice, which enhances motivation when individuals are empowered to make decisions about their health. Competence refers to a person’s capacity to handle challenges and successfully achieve health goals and objectives. It accounts for the necessary knowledge and skills to implement decisions. Relatedness pertains to establishing meaningful connections with others, which can enhance motivation by offering emotional support and fostering a sense of belonging. In the subsequent paragraphs, we discuss evidence-based recommendations for intervention design that specify the psychological mechanisms through which effective programs might operate.

Empowering patients to make informed decisions, advocate for their needs, and share decisions toward their health outcomes has been widely discussed in the literature [94-96], and its relevance has been highlighted as a motivation for action [97] toward better self-management of chronic disease [98-101]. Participants’ emphasis on evidence-based education, ongoing recognition and reinforcement, and technological enablement reflected their desire to become adept at managing their health rather than being passive recipients of care. When previously enrolled in health behavior change, participants saw the relevance of technological tools in enabling their participation by fostering accountability and transparency with the care team and improving patients’ health promotion literacy. They highlighted the value of technology in facilitating goal-setting, self-tracking, and monitoring, and receiving immediate feedback to enhance patient engagement.

Additionally, our study identified the desire of survivors of cancer for evidence-based health education and health literacy. Although operationally distinguished from patient empowerment, in that being health literate alone does not translate to an empowered patient [94,97], research posits that health literacy and empowerment work together to enhance self-management skills. Based on participants’ data, we argue that providing patients with the skills to make reasoned decisions and informed choices may facilitate their continued motivation to participate and engage with health promotion interventions. However, a critical implementation consideration for this finding is the need to empower patients with health education and literacy support, with careful attention to preserving survivors’ autonomy rather than reinforcing practitioner authority. While our study participants valued evidence-based information and expert support, empirical evidence demonstrates that seemingly supportive techniques can backfire if implemented in controlling ways. A meta-analysis reported that offering structure and information was associated with lower autonomous motivation, possibly because participants experienced these strategies as controlling or reinforcing the providers’ authority in guiding behavior change [86]. Our participants’ perspectives for implementing knowledge recognition systems that acknowledge survivors’ existing expertise, offering adaptive content, and ensuring expert accessibility through collaborative consultation could provide pathways for delivering essential information in autonomy-supportive ways.

The desire for recognition and positive reinforcement from health care providers further highlighted the social nature of competence building. Participants valued validation of their efforts from trusted professionals, suggesting that competence support requires patient empowerment through positive reinforcement and recognition from respected sources. Providing survivors of cancer with recognition and positive reinforcement can further improve their self-efficacy and empowerment toward adopting and maintaining healthier lifestyle choices.

Perhaps the strongest theme in our data was participants’ desire for autonomy expressed through a desire for shared decision-making and a preference for personalized, flexible program design that respected their individual circumstances and expertise. The survivors of cancer in our study expressed high interest in taking increased responsibilities and shared decision-making when they participate in health behavior change interventions as opposed to being passive recipients of care.

As part of their autonomy-supportive techniques, Teixeira et al [89] recommended identifying obstacles or sources of pressure for participants. Our study advances this contribution by specifying some concrete implementation requirements, including pause-and-resume features for life transitions and incremental physical activity intensity options to help participants self-manage challenges. Notably, corroborating Teixeira et al [89], we identified participants’ desire for personal agency and choice over intervention features vis-à-vis customization of intervention content. For example, participants wanted programs that would allow them to skip familiar content and build on existing expertise. We found that many survivors of cancer accumulate substantial health knowledge through their active treatment and are somewhat aware of behavior change requirements for survivorship. Programs should acknowledge this awareness to prevent inadvertently undermining autonomy by treating participants as universally lacking health literacy.

While autonomy has been identified as a major human need to facilitate internalization of health behaviors [67,68] and that choice is a positive motivator [102], it has also been argued that having many opportunities for choices might in fact be burdensome and demotivating [103] or generally energy-draining, thereby complicating decision-making. Our study echoes our participants’ vantage on their desire to exercise their capacity to reflectively endorse or reject certain intervention features, which is the SDT’s view of autonomy [89,104]. SDT endorses facilitating people’s experience of choicefulness or volition, which may happen with one, more, or fewer options, as the number of options is not, by itself, defining autonomy [69]. A person might have multiple choices but still lack a sense of autonomy, feeling instead burdened and frustrated by the decision-making process. On the other hand, an individual could find themselves with just a single choice, essentially having no options, yet still feel a sense of autonomy if they genuinely support that choice.

Moreover, our findings reveal that autonomy can be guided or negotiated, thereby giving intervention designers the flexibility to remain objective in intervention research design. We see incorporating frequent patient-centered interactions, in other words, check-ins, throughout the intervention lifecycle as a potential pathway for guiding or negotiating autonomy.

Although not explicitly articulated, our reflective analysis throughout participants’ narratives implicitly depicts that interventions should be designed to avoid controlling language, as participants consistently described preferences for collaborative rather than controlled approaches. This finding aligns with recent empirical evidence on SDT intervention mechanisms [86]. The meta-analysis demonstrated that theoretically predicted techniques like noncontrolling language and providing rationales significantly enhanced autonomy satisfaction and autonomous motivation, respectively.

In addition to needs and preferences that satisfy competence and autonomy, our study identified the support needs of survivors of cancer for relatedness through human-centered support. Relatedness accounts for people’s desire to feel accepted and close to others, as opposed to feeling isolated or rejected [64,65,69], and it reinforces motivation by linking health behaviors to one’s community. Our study participants consistently emphasized the value of human connection and support. The desire for health promotionists as central figures in care for survivors of cancer reflected a need for expert support and professional guidance, but, more importantly, meaningful relationships that could provide accountability and emotional support through the challenges of behavior change.

Additionally, while participants valued receiving support from dedicated health promotionists, they also noted their interest in receiving support from their peers. The survivors of cancer in our study discussed the opportunity to receive relational and emotional support from connecting with other survivors, for example, in group settings, who understood their unique situations and with whom they shared similar experiences. However, our participants highlighted the need for skilled moderation when peer support is offered in group settings to maintain supportive rather than detrimental dynamics. Moreover, a recent meta-analysis showed that group-based intervention activities demonstrated dual effects, in that participants reported increased feelings of connection and belonging during group interactions, yet experienced diminished individual competence perceptions when functioning independently [86]. This suggests that intervention designers should pay careful consideration to delivering group support to prevent inadvertently undermining personal self-efficacy.

An interesting tension emerged between participants’ autonomy and relatedness needs. Participants expressed variable preferences for how they would want their family involved in their participation, with some participants wanting family engagement, while others preferred keeping their health behavior change participation private. Interventions must navigate this complexity by offering flexible approaches to family involvement while respecting participants’ autonomy in deciding their level of disclosure and family engagement.

While Teixeira et al [89] identified adopting empathetic listening as a relatedness-supporting technique, our survivors articulated the specific qualities they want in health promotionist relationships, including incorporating nonjudgmental acceptance, practical guidance through setbacks, an empathy-centered communication style, and collaborative rather than controlling approaches. Paradoxically, an empirical study showed that personal engagement from practitioners, including showing individual interest and offering affirmations, was associated with reduced autonomy and relatedness satisfaction among participants [86]. The study suggested that this detrimental effect may occur because such approaches create participant dependence on the practitioner, or the results could reflect the heterogeneous effects of multiple techniques reviewed under the umbrella study. This study highlights survivors’ nuanced preferences for health promotionist relationships that balance support with autonomy, emphasizing collaboration rather than controlling involvement.

To facilitate relatedness, our participants consented to and expressed value in having their relevant behavioral participation and progress data shared with their primary and oncology care teams. This preference challenges the common practice of delivering behavior change interventions in isolation from survivors’ primary and oncology clinical care, suggesting instead that integrated approaches may more effectively support sustained engagement by leveraging existing trusted relationships and ensuring care continuity.

Beyond individual psychological needs, participants expressed distinct interest in support needs that operated at the program and health care system levels. The emphasis on integration with existing care infrastructure and access to comprehensive wellness services reflects survivors’ holistic understanding of health and their desire for integrated rather than fragmented support.

This systems-level perspective has important implications for how digital health interventions are conceptualized and implemented. Rather than functioning as standalone programs, effective interventions may need to serve as coordinated components within broader health care ecosystems, connecting participants to complementary resources and facilitating communication between different aspects of their care. Participants’ interest in complementary wellness services, including counseling, mental health support, and mind-body practices, suggests that survivors understand wellness as multidimensional and are seeking programs that can support their complete health and well-being rather than focusing solely on specific behaviors or cancer-related outcomes.

Our findings may also be understood in relation to other behavior change frameworks beyond SDT. First, the importance participants placed on accountability, encouragement, self-monitoring, and confidence-building aligns with social cognitive theory (SCT), which emphasizes self-efficacy, behavioral capability, and social support as key drivers of behavior change. From an SCT perspective, features such as coaching, feedback, and selective peer interaction may function not only as sources of encouragement but also as mechanisms for observational learning, modeling, and accountability that reinforce self-efficacy [105-108]. Prior work in cancer survivorship has shown that SCT-based physical activity and diet interventions can improve behavior, particularly in physical activity–focused programs, although results are more mixed for multiple-behavior interventions. Second, participants’ emphasis on recurrence risk, long-term health, and the need for clear, credible information about why behavior change matters is consistent with the health belief model, particularly its constructs of perceived susceptibility, perceived benefits, and perceived barriers. Survivors’ preferences for low-burden, trustworthy, and symptom-aware support may also reflect efforts to reduce cognitive and physical barriers while responding to salient cues to action, such as symptom changes or fear of recurrence. Empirical studies show that perceived benefits and barrier self-efficacy are associated with greater engagement in moderate-to-vigorous physical activity among survivors of cancer [109,110], and that perceived barriers, such as fatigue, comorbidities, or treatment side effects, remain central determinants of behavior change [111,112]. Third, participants’ descriptions of hesitation, trial-and-error, and varying preferences for when and how to engage in behavior change echo the transtheoretical model’s focus on stages of readiness. Transtheoretical model–based health behavior change studies suggest that preferences for specific forms of support, such as informational guidance, reminders, or other accountability features, which were also identified in this study, may shift over time, as individuals progress through interventions [113,114]. Our study highlights the importance of matching support to an individual’s current level of preparedness. Future work may study the relevance of stage-sensitive design in identifying how various forms of digital support may feel most useful at different stages of patients’ health behavior change. Taken together, these comparisons suggest that the preferences identified in our study, framed through motivation and pre-enrollment preferences, also support broader behavior change principles.

In addition to studies that used other theories, we identified some preference-focused studies in cancer survivorship that suggest that survivors value features such as tailored feedback, coaching, and optional social support [115-117], which we also identified in this study. While these prior studies identify which features are desirable, our findings extend this work by including perspectives from survivors of more diverse cancer types and examining multiple health behaviors. More importantly, our study elucidates the underlying motivational pathways, contextual constraints, and emotional experiences that shape these preferences.

Finally, as AI becomes increasingly ubiquitous, it offers a promising approach for making digital behavior change interventions more responsive to the lived realities of survivors of cancer [118,119]. Our findings highlight survivors’ preference for personalized, holistic, and context-sensitive support. Emerging evidence in the literature suggests that AI-enabled digital behavior change interventions can effectively deliver such support through personalization [120-122], adaptive coaching [123,124], predictive risk modeling [125,126], and multimodal data integration [127]. For example, a randomized trial among sedentary survivors of cancer demonstrated that an AI-enabled, bidirectional voice coach produced meaningful short-term increases in physical activity (steps per day), supporting the feasibility of AI-driven adaptive coaching [123]. More broadly, context-aware interventions using analytic and AI methods to integrate multimodal data such as patient-reported outcomes, wearable-derived activity and sleep signals, and clinical context have been used to automate individualized feedback or monitoring and were associated with improvements in behaviors, including physical activity and diet [128]

While we see the promise of AI, our study participants’ desire for encouragement and human connection that does not feel purely automated highlights careful considerations for fully automated approaches. Hybrid programs that combine AI scalability with human relational support may be best positioned to sustain engagement and acceptability [129]. We recommend that, rather than viewing technology and human support as alternatives, effective programs should leverage technology to augment and enhance human-centered survivorship care.

Our findings provide specific, actionable recommendations for developing digital health behavior change interventions that can address the complex motivational and support needs identified by survivors of cancer while maintaining focus on sustainable, autonomous motivation and engagement (Textbox 2).