The aging of the population is accompanied by an acceleration in the incidence of disability (1). Frailty, defined as a pre-disability state of initial impairment of intrinsic capacity, is a target for interventions aimed at improving function and delaying disability (2). Multicomponent lifestyle interventions aimed at promoting healthy aging have proven to be effective in the short term (3, 4).

During the COVID-19 pandemic, social distancing protocols and the subsequent demand for community spaces led to an increase in sedentary behavior in older adults (5), contributing to the progression of frailty and disability (6). An alternative approach to traditional face-to-face physical activity interventions that has gained special momentum is the use of mobile health (mhealth) (7). Technology-based interventions appear to positively influence physical activity levels in older adults (8) and offer the potential to reach individuals on a large scale while allowing for personalized programs. Despite the availability and potential of technology for enhancing physical activity (9), barriers to its adoption and use by older adults and in different care settings remain (10, 11).

It is widely recognized that there is a significant gap between the development of evidence-based interventions for public health and health promotion and their successful and sustainable implementation (12). Approaches for promoting physical activity in older adults using mhealth present unique challenges. Currently, most physical activity promotion interventions remain limited to the experimental or pilot phase, as their continuous implementation or scale-up poses large difficulties. These include a limited understanding of implementation strategies and a failure to match these to the needs of end users.

Engaging end users in the development of health promotion interventions and the design of digital solutions incorporating elements derived from participatory methodologies, conceived within the framework of patient and public involvement (PPI), is key to achieving strategies that are contextually adapted and conducive to their sustained adoption and implementation (13). Participatory design, now known as co-design, is hypothesized to have a strong and lasting impact on health outcomes and may represent a promising strategy for addressing complex health behaviors. Co-design in this context specifically refers to patients and caregivers working collaboratively with health and allied health professionals to improve service delivery by sharing knowledge and experience (14). Its goal is to optimize the implementation of evidence-based interventions according to the priorities and preferences of all stakeholders, enabling designed solutions to achieve maximum feasibility and sustainability.

The present study is part of the +AGIL Barcelona (AGILBcn) program (15), a complex community intervention co-designed by and for frail older adults, together with primary care teams and community stakeholders. The program encompasses different aspects of health including physical activity, nutrition, emotional wellbeing, sleep hygiene, cognitive screening and stimulation, loneliness, and medication review. The AGILBcn multicomponent exercise program consists of 10 face-to-face group sessions led by a physiotherapist in a primary care setting. The program is complemented by exercises performed at home and prescribed through the publicly available ViviFrail^® App (16). Results showed a positive impact on physical function at 3 months (17).

Due to the COVID-19 pandemic and the ensuing challenges facing health services, pressure to redesign the program in a virtual or semi-virtual format increased. However, despite the great potential of digital technology to enhance the promotion of healthy lifestyles in older adults, a lack of specific implementation strategies could even exacerbate health inequalities, increase costs, and jeopardize implementation in routine practice.

This work aims to identify implementation strategies for optimizing the accessibility, acceptability, and adaptability of mHealth interventions aimed at increasing physical activity, within the framework of AGILBcn or similar programs, and to assess their level of priority and feasibility through a co-design process aimed at ensuring equal and equitable participation of multiple stakeholders.

Through the +AGIL Café sessions, 27 possible strategies were suggested to adapt a multicomponent program aimed at enhancing physical activity in older adults, based on accessibility, acceptability and adaptability. These were grouped into 4 categories: general strategies for reducing digital barriers; specific strategies for facilitating the use of a digital application; specific strategies for facilitating the participation in virtual exercise groups; and strategies for facilitating external support, if needed. The priorities included improvement of digital literacy, assessment of training and technology support needs, technological support from family members, telephone feedback, personalized exercises, and exercise conducted in small groups.

Although mhealth interventions appear to be beneficial for increasing physical activity levels in OA (8, 30), there are still barriers to large-scale implementation, on personal, social, technological, and organizational levels (10). We present solutions for program adaptation that vary in complexity from single-component strategies to multifaceted and multilevel strategies (31). The variety of strategies proposed by our participants appears in line with the characteristics that m-health interventions for physical activity promotion should have (8, 30) and with the theoretical constructs for promoting and sustaining behavior change (BCT) (32). First, according to the existing literature, an essential strategy for increasing physical activity levels is to develop digital health-literacy training resources (8, 33). In our study, educational sessions, collaborative learning and paper or video guides represented priorities to improve self-efficacy and digital literacy at the individual, interpersonal and social/community levels. Second, for the participating health professionals, the assessment of access to digital infrastructure, social support and digital skills should be systematically and universally added to the comprehensive assessment of older people; this is a core element of AGILBcn (15) and is consistent with previous studies (34). Our findings are also consistent with the need for social and community support for the adoption and use of technology by OAs, for the resolution of technical problems, and for a decrease in the potential digital divide, as highlighted by other authors (8). This support might be provided by family members or by local networks (e.g., volunteers or peers).

Social interaction in face-to-face groups has been shown to benefit the adoption, increase and maintenance of physical activity (35, 36). In our study, older adults recognized virtual group delivery as an opportunity to remove some of the existing barriers to participating in face-to-face group programs and as a means to interact with peers, avoiding exposure to COVID-19. In contrast, CVS highlighted that the potential benefits of virtual delivery are undeniable, but that the pandemic has amplified the barriers to technology in OA, increasing their social isolation and loneliness (11). Controversy exists regarding the positive or negative impact of technology on loneliness, connectedness, and social support (33). Future interventions should seek to mitigate the social connectedness paradox of COVID-19 (37). Our groups emphasized the importance of combining non-digital alternatives to decrease the digital divide. Providing feedback is another important strategy to promote and maintain adherence to physical activity, and to trigger and sustain motivation for goal attainment (32). Among different options available [e.g., telephonic, via apps, wearable devices (38)], our participants still preferred to receive feedback by phone.

Attitudes of OA toward mhealth exercise vary (39). In our case, OA were willing to use technology-based exercise programs if they perceived them as useful or beneficial for achieving their goals. Interestingly, all participating groups paid little attention to safety and privacy in technology use, as compared to available evidence (10). OA focused on the adherence to and safety of home exercise performance, suggesting that technology is not an end in itself, but a mean.

While all the groups identified similar strategies, the results concerning priority and feasibility showed notable differences among groups. The OAs appreciated the limited size of participants in the virtual groups, the need for external support for participating in the intervention, personalization of the exercises, guarantee of access to technological support, incorporation of music in the virtual exercise sessions, and weekly telephone follow-up by HP. In contrast, significantly lower scores for these solutions were observed in CVS, and, in a smaller proportion, in HP.

CVSs scored lower on all items compared to the other two groups. In HPs, we observed a tendency to score higher for priority than for feasibility. This may be due to health professionals' experience regarding macro-, meso- and micro-level barriers to the implementation, scalability, integration and sustainability of mhealth interventions. Among the items for which HPs perceived feasibility to be higher than priority were: the use of messaging Apps (such as WhatsApp) to connect users, or as a vector for education; the creation of instruction booklets on the use of applications; and the incorporation of music in virtual exercise sessions. OAs perceived as a priority the design of an expert patient program and the sharing of information with family about the intervention, although ease of implementation was considered low, coinciding with HPs views. Conversely, the recruitment of local volunteers to provide support was deemed both a higher priority and a feasible step for all three groups.

We aimed to engage a wide range of stakeholders from an early stage to address the problem, identify strategies and prioritize them. This is in line with current policies, care practices and growing evidence on the importance of engagement and co-design for the development, implementation and adaptation of health promotion interventions and for the design of digital solutions (40, 41). However, there was some disparity in results regarding the potential benefits of this involvement concerning uptake and adoption (42). As in previous studies, the implementation of the co-design process was time- consuming, and it proved challenging to merge different stakeholder perspectives (43). In addition, involving OA in co-design was demanding, due to the extreme heterogeneity in physical and digital needs and capabilities (42).

As potential limitations of our study, the digital format of the AGIL Café sessions provided opportunities to participate in conversations during the COVID-19 pandemic, but was challenging due to technical limitations, such as signal loss, which resulted in certain segments in which the audio was missing (44). Although the platform allows the respondent to be seen, it is possible that we missed some non-verbal and body language cues, as participants often sit close to their cameras. Participants who were not technologically skilled required additional attention from the research team members, which led to a delay in the start of the sessions. The results should be interpreted with caution: the study was conducted in a particular area and with a particular group of participants, thus the results may not be completely generalizable; the integration of the different contributions made by the three different groups was limited to the final prioritization approach; and, finally, the phrasing of specific questions introduced the risk of being leading or suggestive (this, however, was necessary at the beginning of the sessions with OA, who had trouble understand more open questions).

As for strengths, the main advantage of the study was the combination of qualitative and quantitative research methods to provide complementary information. Another strength was the co-design approach involving all stakeholders, incorporating the diversity of perspectives of the AGILBcn program. Purposive sampling enabled us to recruit a wide range of participant types, although obviously selection bias cannot be completely excluded, as participation could have been skewed toward motivated individuals.

The present study provides practical solutions for implementing a technology-based, multicomponent program for older adults from a variety of perspectives, namely, those of older adults acting as end users, but also those of health professionals and professionals from the community. If confirmed by future studies in experimental and implementation research, these results might provide important considerations for policymakers, care providers, and practitioners, for designing, adapting, and implementing multicomponent, technology-based programs aimed to promoting physical activity in the older adult population. This can help to overcome barriers imposed by extreme conditions, such as the COVID-19 pandemic, and to improve adherence and enhance scalability to exercise programs.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Clinical Research Ethics Committee (CREC) of the Foundation University Institute for Primary Health Care Research Jordi Gol i Gurina (IDIAPJGol) (20/048-P) and by the Ethical Commission of Animal and Human Experimentation (Authorization Number CEEAH 5066) of the Autonomous University of Barcelona. The patients/participants provided their written informed consent to participate in this study.

LV-G, VD, LP, and MI contributed to the conceptualization of the study. LV-G, VD, LP, and LS-B performed the data collection. LV-G and VD were responsible for data curation and formal analysis. LV-G, VD, and MI were responsible for writing the initial draft. ER, PD, KK, MG-G, and CC-T reviewed and edited the draft. MI supervised the process of manuscript preparation. All authors agreed to be accountable for the content of the work. All authors contributed to the article and approved the submitted version.