Principal Findings
Overview
This study provides valuable insights into the attitudes and experiences of Spanish retirement-age adults regarding mHealth interventions grounded in the PSD framework and identifying BCTs. The findings suggest that while older adults recognize the potential benefits of mHealth for promoting healthier lifestyles and are open to using them, several barriers hinder their adoption. The use of different PSD principles and BCTs (eg, social support) could enhance the persuasiveness and usage of these tools.
Primary Task Support
Participants discussed what content they would like to receive and how to receive it to facilitate promoting a healthy lifestyle. Their preferred content in order was social/cultural activities, nutrition, physical activity, mental health, and relaxation. This prioritization aligns with the biopsychosocial model of health, which underscores the need to address physical, mental, and social well-being holistically [49].
In addition, previous literature highlights users’ preference for a holistic understanding of health and well-being, incorporating stress and emotions rather than focusing solely on behaviors [44]. Consistently, they would like to be motivated not only by outlining the health consequences (BCT: 5.1. Information about health consequences) but also through emotional benefits (BCT: 5.6. Information about emotional consequences). While lifestyle interventions frequently use health consequences, emotional ones are usually ignored [39,50].
This preference aligns with self-determination theory [51], which can offer valuable insights on what type of consequences could be outlined to foster autonomous motivation, which is related to higher adherence [52,53]. In this regard, highlighting personally relevant consequences (eg, being independent for longer) or enjoyable ones (eg, fun, stress relief), rather than focusing only on health or disease prevention, may foster integrated/identified or intrinsic motivation, respectively [24]. Both forms of autonomous motivation could lead to higher engagement and maintenance of the target behavior [51,54]. Thus, outlining these consequences in an app could enhance its effectiveness in promoting health behaviors.
Regarding how they would like to receive the desired content, there was consensus on the need for tailoring and personalization based on users’ profiles and age groups, which was consistent with previous research [44,55]. Participants valued tailored diets (eg, what ingredients and liquids you should consume considering your usual activity), advice on behavior substitution (BCT: 8.2. Behavior substitution, such as ideas for substituting unhealthy snacks with healthier ones), and meaningful and progressive goals (BCTs: 1.1. Goal setting and 8.7. graded tasks: eg, start this week doing 1 hour of strength exercises, this improves your muscle function and will help you be strong and independent).
The importance of setting goals that are personally relevant to foster motivation also appeared in a previous qualitative study [44], and as already discussed, is essential to promote autonomous motivation and therefore long-term adherence [52,53]. Setting personally meaningful goals might be especially relevant in this population, as the socioemotional selectivity theory [56] proposes that, as individuals age, only activities that are meaningful and offer well-being are prioritized. In this regard, when an app suggests and sets a goal, identifying and outlining relevant consequences for the individual may improve the likelihood of its achievement.
One of the most appreciated features of mHealth was self-monitoring (BCTs: 2.3-4. Self-monitoring of behavior and of outcomes of behavior), as it allowed users to track their progress effectively. Participants noted that these tools provided valuable feedback and increased motivation. Although previous qualitative studies have not frequently highlighted this, this population widely uses self-monitoring apps [57]. Furthermore, self-monitoring has been proven to be an effective BCT in diverse meta-analyses [39,58], becoming an essential tool in any lifestyle intervention.
However, feedback valence played a critical role according to our participants. When negative feedback (BCTs: 2.2.7. Feedback on behavior or outcomes of behavior) was expected, some participants avoided introducing their data or even provided inaccurate information. This preference was found in previous qualitative studies: participants outlined the importance of feeling supported and receiving encouraging messages even on bad days, when they do not accomplish their goals [44].
This preference was further corroborated by a systematic review including qualitative and mixed studies [59], finding that positive and encouraging content was preferred, whereas negative feedback elicited negative feelings, discouraging app use. These reactions may reflect emotion regulation strategies. When faced with negative feedback, emotions such as anger or shame may arise, which some individuals attempt to avoid by disengaging from the source (eg, the app) or using distraction techniques [60].
Nevertheless, when looking at quantitative studies, there is controversy in the study of feedback: some studies find that feedback is useful in the promotion of health behavior, regardless of whether it is negative or positive [61], while others find that negative feedback could be less effective [62]. Furthermore, negative feedback can decrease self-efficacy [63] and make future efforts less likely.
Importantly, it is possible that negative feedback is more acceptable when it provides informative content or instructions that users were not previously aware of; for example, pointing out a lack of protein in a meal that was assumed to be balanced. In contrast, when feedback highlights known unhealthy behaviors, it may trigger feelings of guilt or shame, prompting users to avoid the app altogether as a means of emotional regulation.
In fact, other studies, mostly in the field of learning experiments, have found that how feedback is delivered is more important than whether it is positive or negative. For example, a meta-analysis showed that negative feedback decreased intrinsic motivation compared with positive feedback, but this effect was less pronounced when negative feedback included instructions for improvement [64]. Although this has been less explored in health behaviors, evidence suggests that gain-framed messages are overall more effective than loss-framed messages in promoting healthy behaviors [65,66], which could be considered when designing negative feedback messages in this context.
These findings offer a novel and valuable perspective on feedback in mHealth, highlighting the importance of avoiding strictly negative feedback and suggesting approaches to make it more acceptable and constructive. For example, combining negative feedback with actionable suggestions and using gain-framed messages. However, future studies should investigate these strategies in greater depth and determine how to deliver feedback effectively in mHealth interventions tailored to older adults.
Dialogue Support
One of the main ways in which an intervention is delivered through an app is through messages and dialogue. In fact, a meta-analysis found that in web-based interventions, a higher use of dialogue support elements was related to higher user adherence [42]. However, along with the Social support principle, these were the least implemented principles. Considering its importance, it is essential to understand how user-app communication should be established.
In this regard, our participants preferred a more human-like interaction, appreciating simple gestures such as asking “How are you doing today?” or receiving messages designed to make them smile or laugh. This reflects a preference for a social-oriented communication style, in contrast to a more task-oriented one, and is closely linked to the concept of social presence (the feeling of being with another [67,68]). Prior research has shown that social-oriented communication and higher perceived social presence foster trust, enjoyment, and emotional connection, which in turn increase engagement with digital tools [69,70].
However, communication preferences vary across individuals and cultural contexts. According to context theory [71], some people prefer high-context communication, which relies on implicit meanings, nonverbal cues, and social nuance, while others prefer low-context communication, characterized by direct, explicit messages and a focus on task efficiency. These preferences have been shown to moderate user engagement: for individuals who value relational and contextual cues (ie, high-context, socially oriented users), social presence enhances engagement; conversely, for task-oriented, low-context users, utilitarian features are more impactful [70].
Our findings suggest that participants in this Spanish sample tended to favor a socially oriented, high-context style of communication, which aligns with Spain’s classification as a high-context culture [72]. Recognizing and tailoring to these cultural and individual preferences is essential for effective mHealth design.
To support this type of communication, conversational artificial intelligence, whether through chatbots or IVR systems, could serve as a facilitator. Participants proposed the integration of simple interactive systems, such as IVRs. This aligns with prior findings showing that older adults often appreciate audio feedback features [55]. Furthermore, using a voice can increase social presence, and for those with preferences for social-oriented and high-context communication, it could be more appealing [73,74].
Participants also expressed a strong preference for avoiding technical terms and anglicisms, which was previously identified as a barrier [55].
These findings highlight the importance of culturally adapted communication, as preferences in terminology (eg, acceptance of anglicisms) and communication style (eg, high-context communication) can vary across cultures and generations. For instance, anglicisms might be more accepted by younger populations [75,76], and their usage differs depending on the language and cultural context [77]. Adopting a communication style that avoids unnecessary technical jargon and anglicisms, while fostering high-context communication and a strong sense of social presence, may better address the cultural and linguistic preferences of older Spanish adults.
Participants also mentioned accessibility as a critical design element. They noted that sometimes apps do not have an adequate letter-screen size ratio for this population, which appears as a usual barrier leading to difficulty in reading and interacting with the app effectively [55]. This aligns with existing guidelines recommending the use of large, legible fonts and touch-friendly interfaces to accommodate older users [78].
Regarding the messages and notifications, participants appreciated positive reinforcement, particularly receiving messages from the app that praised their achievements (BCT: 10.4. Social reward). This is consistent with previously found results outlining the importance of praise and reward messages to sustain user motivation [44,79].
Last, participants viewed reminders or prompts (BCT: 7.1. Prompts/cues) as useful tools to encourage behavior. This is consistent with prior literature, where prompts have been identified as facilitators in a systematic review of qualitative studies [55] and as an effective BCT for promoting physical activity in older adults [50]. However, participants emphasized that receiving too many notifications would be tiring, which could lead to alert fatigue [80], a commonly identified barrier [55]. To address this issue, participants suggested either a moderate use of reminders or allowing users to personalize the frequency and timing of notifications. This approach could prevent alert fatigue while maintaining the effectiveness of prompts to encourage health behaviors.
System Credibility
Participants mentioned the importance of an app free of publicity. Advertising in mHealth apps is often perceived as intrusive, which can erode user confidence and decrease engagement [81]. Ensuring a design free of advertisements could reinforce the app’s credibility and promote a sense of reliability among users.
Previous literature found concerns about security, data privacy, and trust in its management as a barrier to the adoption of eHealth tools [21,55]. However, in this study, such concerns were primarily associated with issues of digital literacy, derived from the fear and uncertainty of how to use apps safely. This highlights the critical need for clear communication about data security. A clear message about security and showing private handling of user data without advertisements will improve the perceived reliability of the app, particularly with cybersecurity concerns among aging people.
For instance, apps could include straightforward explanations of how personal data is protected, avoiding overly technical language to ensure comprehension among older users. Additionally, showcasing features such as encrypted data storage and the absence of third-party sharing could enhance the perceived reliability of the app, particularly for aging populations that are increasingly concerned about cybersecurity risks.
In addition to offering a secure, advertisement-free experience, system credibility can be further enhanced by making the app’s origin and affiliations transparent, such as clearly stating its development by reputable health institutions. Research has shown that perceived provenance (who developed the app and whether it is supported by trustworthy sources) has a strong influence on credibility among older adults [82]. Users may feel more confident when they know the app is endorsed by public health agencies, universities, or well-known hospitals. Displaying this information visually (eg, institutional logos, disclaimers, or mission statements) could reinforce legitimacy. Similarly, stating a credible source for the information provided could improve trust in the app [82,83]. For example, a common issue found in previous research was the low quality of information provided, without specifying a credible source or guaranteeing its correctness [82,83].
mHealth developers can reduce trust-related barriers and foster greater acceptance of these tools among older adults by addressing these issues directly: communicating clearly, eliminating intrusive advertising, highlighting institutional credibility, including high-quality information from credible sources, and visibly signaling data protection.
Social Support
Social support and involvement emerged as one of the most mentioned and valued topics, with participants discussing its influence in different ways and through multiple approaches. This is consistent with previous literature, as social support has been systematically found as an essential component when studying the acceptance, use, and adherence to eHealth tools in older adults [23,55,84,85]. Social support contributes not only to increased engagement but also to better health outcomes, as it fosters a sense of community and shared motivation [39]. Especially, external social support from family members, caregivers, or health care providers has been found to predict greater use and adherence to eHealth technologies [23,55,84,85]. Despite its recognized importance, social support is usually the least implemented principle in mHealth, along with dialogue support [41].
In this study, participants expressed a strong interest in features that facilitated social interaction within the app. First, they suggested the inclusion of tools for organizing or recommending group activities (BCT: 3.3. Social support practical; eg, meeting to get to know a part of the city and count the steps). Second, they appreciated the idea of group chats where users could share their experiences, difficulties, and barriers, supporting each other (BCT: 3.1. Social support unspecified). These preferences indicate a clear demand for mechanisms that foster peer-to-peer interaction and mutual encouragement. Interestingly, these findings align with prior studies that have identified online support groups as a desirable feature for older adults [44,84]. Including these features could facilitate engagement, enjoyment, and also make health behaviors (eg, walking) an opportunity for social affiliation (an intrinsic goal from the self-determination theory), which could lead to autonomous motivation and better adherence to the behavior.
Participants also mentioned that instructional videos could be useful for demonstrating behaviors (BCT: 6.1. Demonstration of the behavior, 4.1. Instructions on how to perform a behavior) when other options are unavailable. However, some participants were reluctant to follow such videos, suggesting that their preferences may vary based on individual comfort levels with this format. This variability aligns with previous research indicating that while many older adults exhibit positive attitudes toward video-based interventions and adhere to this format [86], others may prefer alternative methods of instruction, such as written guides or interactive tutorials. For instance, the Institute on Aging [87] highlighted that older adults often favor written materials for their clarity and accessibility, as these formats are perceived as straightforward and less overwhelming. Additionally, interactive tutorials were shown to enhance engagement by allowing users to learn at their own pace, accommodating varying levels of digital literacy and familiarity with technology. These findings emphasize the importance of offering diverse instructional formats to meet the heterogeneous preferences and needs of older adults effectively.
Overall, these results point out the importance of integrating social support features into mHealth tools for older adults. By enabling social interaction and offering multiple instructional formats, developers can address individual differences and enhance both engagement and adherence.
Attitudes, Barriers, and Design Preferences
Similar to previous research, our findings indicate an overall positive attitude toward the use of eHealth and mHealth, as well as openness to use them [88]. However, certain concerns and barriers influencing the use of these tools arose, such as (1) concern for technology dependence and (2) low digital literacy.
Regarding technology dependence, participants criticized that tasks previously performed manually (such as route planning) are now often outsourced to technology, reflecting that excessive reliance on digital tools can lead to a decline in memory retention and navigational abilities. Studies partially support this concern, showing that overly relying on technology can lead to negative cognitive consequences, such as reduced memory retention and attention span. However, recent studies have shown that technology can also improve cognitive functioning, potentially enhancing memory, attention, and executive function when used appropriately [89-91]. Understanding this hesitance and concerns is critical for eHealth developers, who could explicitly address them by incorporating features designed to counteract cognitive decline. For instance, integrating cognitive training tasks or memory-enhancing games could encourage users to perceive technology as a tool for improving, rather than replacing, their cognitive skills. Furthermore, the use of these tools could also be framed as a learning challenge that could be cognitively stimulating.
Low digital literacy appeared in terms of participants’ perceived lack of competence and fear of severe consequences if used incorrectly. This is consistent with prior reviews, indicating that older adults often experience low self-efficacy, lack of confidence, and a general aversion or fear toward technology [55]. These challenges lead to a cycle of avoidance and missed opportunities to benefit from digital tools.
Participants identified several design elements that could help overcome this barrier. They emphasized the importance of a user-friendly and intuitive app design that minimizes cognitive and technical demands; an app that is not excessively demanding or time-consuming. In fact, it has been found that information or system features overload in mHealth relates to higher resistance and reluctance to use these tools in older adults [80]. Thus, reducing this cognitive load is essential to foster engagement and minimize dropout rates among older users [78].
Participants suggested additional actionable ways to improve usability, such as incorporating clear, universally understandable symbols and short tutorials for onboarding. How to design an adequate tutorial would be another key point, for which different guidelines for older adults can be followed (eg, avoid cognitive low and burden on working memory) [92]. Further training on technology might also be useful to address this barrier. These features can ease the learning process and enhance users’ confidence in navigating the app. This aligns with previously identified facilitators in the literature, which recommend prioritizing simplicity and providing step-by-step guidance to improve user experience [55,78].
Limitations
This study provides valuable insights into the perspectives of Spanish retirement-age adults regarding mHealth interventions; however, several limitations should be considered when interpreting the findings.
First, although the number of focus groups aligns with prior recommendations for qualitative research [27], data saturation was not formally assessed and therefore cannot be confirmed. Thus, it is possible that some influencing factors were not identified. Second, the sample size was relatively small (n=19) and predominantly female. Although the initial recruitment strategy did not target any specific gender, once a gender imbalance became evident, additional efforts were made to recruit male participants. Despite these efforts, participation remained predominantly female (15/19, 79%). This limitation, common in research [93], and particularly with this population [94,95], may limit the generalizability of the results and highlights the need for more targeted strategies to engage men in future research. Therefore, while the findings shed light on important themes, a larger and more diverse sample, including more male participants, could provide a broader understanding of mHealth preferences among this population.
Additionally, all participants were recruited from Madrid, an urban area, which may not fully capture the experiences and needs of individuals from rural regions with different technological adoption rates and cultural contexts. Moreover, while information on participants’ employment (current or prior to retirement) was collected as an indirect indicator of socioeconomic status, other demographic variables such as education level and income were not systematically assessed. Expanding the geographic scope and using a more comprehensive demographic assessment in future studies could enhance the contextualization of the findings.
Additionally, while the PSD and BCT frameworks provided a valuable structure for organizing and interpreting the data, they present certain limitations. Specifically, some salient themes that emerged (such as attitudes, barriers, and easy-to-use design) did not clearly align with any of the pre-established categories within these models. This suggests that although these frameworks are useful for identifying actionable techniques and persuasive strategies, they may not fully capture all relevant aspects for older adult users. Future research could consider complementing these models with additional theoretical perspectives to ensure a more holistic understanding of user needs.
Finally, even though several preferences are highlighted, these might not always be feasible for app design and can actually differ across individual characteristics such as digital literacy or prior technology use. While we report general information on these variables when available to contextualize participants’ quotes, the small sample size and qualitative design do not allow for meaningful subgroup comparisons. Future quantitative studies could build upon these findings to explore, for example, whether preferences for features such as human-like interaction or group-based activities differ according to digital literacy level or other relevant variables. Such studies could further inform the personalization of mHealth tools and contribute to reducing digital health disparities in this population.