The SIM provides a theoretical framework for understanding how social determinants interact with communication processes to influence health behaviors and outcomes [11]. This model posits that communication outcomes, including exposure, usage, attention, seeking, and processing of health information, are interlinked with social determinants to influence health outcomes (Figure 1). According to SIM, health information benefits are predominantly accessible to those with the capacity to access, process, and use it, highlighting potential for health technologies to widen but not narrow health disparities.

When applied to older adults’ engagement with telemonitoring technologies, the model complements the political economy of aging theory [12,13], which emphasizes how socioeconomic structures create and maintain inequalities in later life. Thus, as shown in Figure 1, we adapted Viswanath et al’s [11] 4-component framework into a 3-component framework centered on socioeconomic determinants (age, education, and income) and their relationship to health information exposure, seeking, and processing related to telemonitoring. Consistent with SIM, the digital health care revolution may disproportionately benefit those with greater socioeconomic resources, technological skills, and health literacy [14]. Friemel [15] showed that the digital divide persists among seniors, encompassing not just technology access but also differences in usage patterns, skills, and perceived benefits. Research has identified social determinants such as age, income, and education as significant factors influencing telemonitoring adoption [16,17].

Despite existing literature linking social determinants and communication outcomes with digital health adoption, significant gaps persist. Socioeconomic status (SES) complexity necessitates diverse social deprivation measures, tailored to research objectives and cultural contexts. While scholars commonly use income, occupation, and education levels to delineate social standing when exploring health disparities [18], Darin-Mattsson et al [19] suggest that in old age, educational attainment may have a weaker association with health compared to wealth, income, and tenure. These findings underscore the importance of focusing on low-income older adults when examining telemonitoring adoption motivators and barriers. Consequently, guided by SIM, we aimed to address the following research question (RQ):

RQ1: What are the motivators and barriers associated with the adoption of telemonitoring services among low-income older adults, as understood through the lens of information exposure, usage, attention, seeking, and processing?

Building upon Phase 1 findings, which highlighted the importance of affordability, convenience, and health benefits, as well as the challenges posed by preferences for in-person consultations, perceived limitations of remote monitoring, and diminished information processing abilities and health LOC beliefs. Phase 2 aimed to develop and test effective communication strategies to enhance telemonitoring adoption. While Phase 1 identified the key drivers of adoption, altering the technology itself or its accessibility through subsidies, redesign, or infrastructure is better addressed through policy or engineering, and falls out of the scope of health communication. Instead, Phase 2 was designed on the premise that user awareness of its benefits and the perceived relevance of those benefits are essential prerequisites for adoption, regardless of a service’s objective advantages [20]. That indicates that the critical question is not only whether these benefits exist, but whether they are being communicated in a way that resonates with older adults, as prior research has shown that older adults often fail to recognize the personal relevance of technologies even when aware of their potential benefits [21]. Accordingly, experimental stimuli of 2 modalities held core service information constant across framing conditions, varying only how this information was presented.

Framing holds particular promise as a major message format studied across diverse health contexts. It involves strategically presented information where attention is drawn toward specific details [8]. In the context of older adults’ technology adoption, framing effects can be particularly significant given the experience gap, the discrepancy between older adults’ life experiences and new technologies' demands [22]. This gap creates ambiguity about the value and relevance of technologies, making the framing of messages particularly influential in shaping perceptions and attitudes.

Framing theory posits that media messages shape audience perception and interpretation by highlighting particular aspects of an issue [8]. Iyengar [23] proposed an influential distinction between episodic framing and thematic framing. Episodic framing focuses on concrete exemplars, isolating single events or individuals while offering engaging narratives to depict broader issues via unknown and ordinary individuals [23]. Thematic framing focuses on a more societal context while offering statistical evidence. These framing approaches intersect with a sociocognitive perspective on aging, emphasizing how older adults’ perceptions, beliefs, and decision-making are shaped by both individual cognitive processes and social contexts [24]. In the health communication context, episodic messages might describe a specific individual’s experience with an illness, while thematic messages might present population-level statistics or attribute the issue to broader social factors [25,26]. In this study, the distinction was operationalized as follows: episodic messages featured a personal account from a named older adult regarding their use of health screening services, while thematic messages presented national survey statistics and referenced community health infrastructure (Experimental Stimuli in Methods section).

Findings from Phase 1 showed that only a small number of individuals believed they were fully responsible for their personal health, while the majority displayed a sense of health responsibility attributions to external parties such as the government or medical professionals. This divergence aligns with research in health communication indicating that responsibility for health outcomes is often attributed to a combination of individual and social or systemic factors [27]. Specifically, older adults’ sense of health agency often involves negotiating between personal responsibility and reliance on health care systems, particularly as physical vulnerabilities increase with age, consistent with evidence that older adults’ perceptions of aging shape their health and functioning [28]. How health messages frame this balance between individual and systemic responsibility may significantly impact their effectiveness. Iyengar’s [23] framework was originally designed to explain how episodic and thematic frames shape audience attributions of responsibility for social issues (ie, attributing responsibility to the individual vs society or institutions); thus, it offers a theoretically grounded perspective for examining whether older adults holding differing views on health responsibility resonate differently with these frames.

The distinction between episodic and thematic framing may be closely related to exemplification theory [29]. This theory distinguishes between exemplar-based information (ie, vivid and specific cases) and base-rate statistical information, elucidating the cognitive mechanisms underlying their differing effects. According to the theory, exemplar-based information exerts a disproportionate influence on judgment because it is cognitively more accessible and more readily perceived as representative of broader group characteristics compared to abstract statistical data [29]. Recent research strongly supports this effect: a meta-analysis found that across various communication contexts, exposure to exemplar-based information consistently produced stronger effects than exposure to base-rate information [30]; these effects were most pronounced regarding “first-order reality judgments,” such as assessments of prevalence or risk [31]. From this perspective, episodic framing can be viewed as a form of exemplar presentation, whereas thematic framing corresponds to the presentation of base-rate information; and this offers a cognitive-mechanism-level complement to explanations regarding the persuasive advantage of episodic framing [32]. These characteristics may make exemplar-based information particularly relevant for some older adults. Age-related declines in information processing capabilities [33], together with a preference for emotionally meaningful and concrete content [34], suggest that older audiences may be especially sensitive to the availability heuristic and representativeness heuristic underlying the exemplar effect. Therefore, we propose the following hypothesis:

H1: Compared to those exposed to thematic framing, older adults exposed to episodic framing report more positive (1) attitudes toward and (2) intention to use telemonitoring services.

Based on Ajzen’s [35] Theory of Planned Behavior (TPB), attitude is one of the predictors demonstrating strong and consistent relationships with intention to perform a behavior. This theoretical relationship has particular relevance for older adults given innovation resistance often observed in later life [36]. This resistance isn’t merely about technological difficulty but reflects complex attitudinal factors including concerns about privacy, perceived usefulness, and compatibility with existing routines and values [37]. Understanding how attitudes mediate the relationship between message framing and adoption intentions can therefore provide insights into the psychological mechanisms underlying older adults’ technology decisions. Similarly, studies have found evidence supporting significant relationships between attitudes and behavioral intentions in the context of telehealth acceptance and COVID-19 vaccination [38-40]. Thus, we posited:

H2: Attitude toward using telemonitoring services is positively associated with intention to use telemonitoring services.

H3: The relationship between framing and intention to use is mediated by attitude toward using telemonitoring services.

The moderating role of modality suggests the delivery method of messages through media platforms significantly influences message reception [41]. Defined modalities, categorized into video and photo-text, are informed by dual-coding theory [42], which posits that visual and verbal information is processed by distinct cognitive systems. For older adults, message modality may have heightened importance due to age-related changes in sensory processing and cognitive function. Charness and Boot [43] note that age-related changes in visual and auditory processing can affect how older adults engage with different media formats. This theoretical framework supports that videos, combining visual with conversational elements, can enhance comprehension and retention, making them particularly effective for presenting eHealth information to older adults [44].

Evidence suggests varying impacts based on the chosen modality. One study [45] found that both gain- and loss-framed photo-text stimuli significantly influenced health behavior intentions, whereas another [46] demonstrated that loss-framed videos increased perceived severity of health risks more effectively than gain-framed videos. de Graaf et al [47] and Occa and Suggs [48] highlighted the stronger impact of video messages on health behavior attitudes compared to photo-text modalities. Therefore, we proposed:

H4: Modality and framing would interact, such that participants assigned to the video condition would have a more positive attitude toward using telemonitoring services compared to participants assigned to the photo-text condition.

Wallston et al [49] defined health LOC as the degree to which individuals believe their health is controlled by their own behavior versus external forces. People with an internal LOC believe health status is determined by internal causes such as their own behavior. Meanwhile, those with an external LOC believe that one’s health depends on external causes or powerful others such as health professionals. Health LOC intersects with gerontological theories about control beliefs in later life. The selective optimization with compensation model [50] suggests successful aging involves adjusting control strategies as resources and capacities change. External LOC might represent an adaptive response to increasing health challenges rather than simply passive orientation [51]. Prior research supports a strong association between numerous sociodemographic variables and external LOC, with studies finding that those of older age, lower income, and lower education were more external in terms of health LOC [52,53].

Current literature found individuals with a stronger external health LOC are more likely to find thematic framing messages more appealing [54], as they emphasize societal or group responsibility. Similarly, Ceylan and Hayran [55] verified that LOC moderated the effects of prosocial and self-interested messages in following COVID guidelines, where individuals with higher external LOC were more influenced by prosocial messages. In the context of interactive health websites, those with higher external LOC are responding more strongly to vivid, exemplar-rich message formats [56]. However, other studies have shown contrasting or inconclusive results on the link between episodic or thematic framing and individual and societal attributions [57,58]. Hence, the following question was asked:

RQ2: To what extent does external LOC moderate the relationship between framing and attitude toward using telemonitoring services? The proposed research model and hypotheses are presented in Figure 2.

The research was conducted with FullHealth (provided by FullHealth Asia), a grassroots initiative that provides telemonitoring services through health screening and monitoring booths (Figure 3). This study received approval from the Institutional Review Board (IRB) of Nanyang Technological University (IRB-2022‐826). All participants provided written informed consent prior to participation in the study, with protocols designed to accommodate potential age-related needs in the consent process. Before each session, participants were briefed on the study’s goals and their participation rights, with consent for audio-recording obtained and anonymity assured through the use of pseudonyms. All data were deidentified by using numerical codes, and any identifying information mentioned in the interview transcripts was removed. The data were stored securely and were accessible only to research team members (DW, HB, APTQ, GHYH, PSH, ESMO, and EWJL). All participants received S$30 (US $23.40) compensation.

We implemented a 2×2 between-subjects experimental design with framing conditions (episodic and thematic) and media formats (video and photo-text), engaging 120 participants from January 16 to February 1, 2023. Participants were recruited using convenience and snowball sampling through FullHealth Asia, referrals, and social media. All participants were unfamiliar with the FullHealth booth.

Framing conditions: thematic framing highlighted societal issues using “Singapore” in titles, statistical data, national surveys, and references to older adult activity centers to reflect broader societal concerns. Episodic framing focused on the individual, using direct address (eg, “You”), personal stories, and specific experiences. Both framing approaches ensured clarity for older audiences.

Modalities: Stimuli maintained design consistency through uniform color schemes, fonts, and logos. Photo-text stimuli are provided in Multimedia Appendix 1; the 4 video stimuli are provided in Multimedia Appendices 2-5. Differences were marked by imagery, text, and narration changes for video content, accommodating potential age-related sensory and cognitive changes.

Crucially, the core information of service description, emphasizing that FullHealth’s booth was free, convenient, and offers comprehensive health checks (blood sugar, blood pressure, height, weight, heart, and eyes), remained consistent across the episodic and thematic conditions in both the English and Chinese versions. The only difference between the presentation styles lay in the narrative approach used to convey this information (personal narratives vs aggregate statistics combined with organizational background), isolating the framing manipulation from message content.

Participants were randomly allocated to 4 conditions: (a) episodic video, (b) episodic photo-text, (c) thematic video, or (d) thematic photo-text. After informed consent, participants assessed external LOC, viewed assigned stimulus, then completed attitude and intention measures plus manipulation checks. A debriefing session clarified any misinformation during the study. Participants received S$15 (US $11.70) as compensation.

External LOC: this moderating variable was measured using a 5-point Likert scale adapted from the Health Locus of Control Scale (HLCS) by Wallston et al [63], using items such as “Good health is largely a matter of good fortune,” with mean 3.13 (SD 0.60, Cronbach α=0.72).

Attitudes toward telemonitoring: participants’ attitudes were measured using a 5-point semantic differential scale, with items modified from Dabholkar and Bagozzi [64]. Participants rated their feelings toward using the FullHealth booth on dimensions such as “Good to bad” and “Harmful to beneficial,” with mean 4.03 (SD 0.47, Cronbach α=0.85).

Intention to use telemonitoring services: this was measured on a 5-point Likert scale, adapted from Boontarig et al [65], with items such as “I intend to use this FullHealth booth in the future,” mean 4.14 (SD 0.65, Cronbach α=0.90).

Manipulation check: for framing, participants assessed statements reflecting message focus. Items for episodic framing evaluated FullHealth booth experiences from an individual older adult’s perspective, while thematic framing assessment involved societal impacts and statistical evidence. For modality, 2 items checked participants’ recognition of the stimulus format (video or photo-text).

We analyzed the data using Hayes’ PROCESS macro (Model 9) [66] via SPSS Statistics (IBM Corp), incorporating gender as a covariate. Because PROCESS estimates a series of ordinary least squares regression equations rather than a simultaneously estimated structural equation model, the coefficients reported below reflect independent regression paths but not a jointly estimated path model; accordingly, global model fit indices (such as χ², Comparative Fit Index [CFI], and root-mean-square error of approximation [RMSEA]) are not applicable here. Message framing was the independent variable, with attitudes and intentions toward telemonitoring services as dependent variables. External LOC and modality were considered moderating variables. The complete set of model estimation coefficients (including all covariate, interaction paths, and conditional indirect effects at representative values of the moderators) is detailed in Multimedia Appendix 6, together with bootstrap SE and 95% CI.

A sensitivity power analysis indicated that, based on the sample size (N=120) and the 6-predictor moderated regression model used to test Hypotheses 1‐4 and RQ2, the study had 80% power (α=.05) to detect an effect of Cohen f²≥.067, an effect size falling between the conventional benchmarks for small (.02) and medium (.15) effects [67]. Post hoc power analyses regarding the key effects reported herein confirm that the study possessed sufficient power to detect the significant “framing× locus of control” interaction (f²=.095, power=.92) and the relationship between attitudes and intentions (f²=.173, power>.99). The post hoc statistical power for the main effects of framing, presentation modality, and gender, as well as the “framing× presentation modality” interaction, was relatively low (0.06‐0.10) (Table 1).

Most FullHealth users cited proximity as the primary reason for using the telemonitoring booth, allowing them to “save time” (U7, male, 57 years) and:

Nonusers also acknowledged “convenience” as a potential motivator. Given the observed mobility challenges, convenience appeared to be particularly significant, reflecting how aging-related changes shape technology preferences.

The “FOC [Free-of-Charge]” (U10, Female, 64 years) nature of the booth services encouraged usage among participants. While some noted that the cost advantage might be less significant compared to other subsidized older adult care services (eg, “This doctor I visit helps me save money. It’s not very expensive because I'm old” [N1, female, 75 years]), comparing the booth with a regular doctors’ visit, cost remained critical. When discussing more expensive health services, participants expressed concerns:

Overall, participants demonstrated high price sensitivity, highlighting how financial considerations become increasingly significant in postretirement life.

Many participants cited health benefits as a key motivator. These benefits included addressing specific health issues (eg, “it helps prevent me from getting a stroke again” [U3, male, 59 years]) and general health monitoring (eg, “if we have any illness, we will know” [U8, male, 63 years]). As one participant summarized:

Participants noted their increased susceptibility to illnesses and, making health monitoring particularly valuable, reflecting how health vigilance increases with advancing age.

A primary barrier was the preference for in-person medical consultations. Participants felt that face-to-face meetings provided better understanding of their health concerns (eg, “Seeing the doctor. Talk about our illnesses [...] clearer. We can understand” [U4, male, 76 years]; “When in front of a doctor, if we have any questions, we can ask, which is better” [U2, male, 57 years]), and offered a sense of familiarity (eg, “The doctor if we see often already, [we] will go find that doctor” [N2, female, 78 years]). These preferences highlight how technology adoption in later life is influenced by established relational patterns and trust.

The booth’s restricted monitoring capabilities were considered “very simple” (U9, male, 65 years) and “not comprehensive, it’s not full” (N3, female, 56 years), presenting a barrier to adoption. One participant noted:

This perception of limited functionality compared to traditional health care services influenced participants’ willingness to adopt telemonitoring.

Information processing posed a major barrier, especially for lower socioeconomic participants who struggled to understand health advertisements (eg,“[When shown a health advertisement on the National Steps Challenge] also can't really understand when looking” [N2, female, 78 years]); “[When asked what the National Steps Challenge Advertisement is talking about] Don’t know” (U8, male, 63 years). Information processing abilities varied notably based on education and income levels. Participant U4 (male, 76 years) with no formal education and earned less than S$1000 per month, demonstrated poor information processing skills (“[When asked about the TraceTogether Advertisement] never studied, don’t understand” [U4, male, 76 years]). In contrast, participant N3 (female, 56 years), with an O-level certificate and monthly income between S$3000 and S$3999, showed better comprehension (“[When asked about the TraceTogether Advertisement] To track people where they, in case of Covid, they know who is where, you know, Yeah, so they can trace the people and then inform them that, you know” [N3, female, 56 years]). These variations highlight how lifelong socioeconomic advantages and disadvantages accumulate to influence technology engagement.

Participants cited various media sources including “posters,” “pamphlets,” “flyers,” “radio,” “TV,” “newspapers,” “Facebook,” “WhatsApp,” “YouTube,” “Websites,” and “[digital display screens in] lift.” Although television and mobile phones were the most commonly owned, neither was frequently used by most participants. Multiple media sources were often mentioned (eg, “I think from the website, [...] flyers, [...] more news” [N3, female, 56 years]; “See the news. [...] I buy newspapers [...] Yes, all [Whatsapp and Facebook] got used, also use Youtube” [U2, male, 57 years]). The diverse media preferences suggested the importance of considering various modalities in health communication.

Participants displayed varying beliefs about control. Some demonstrated a sense of personal responsibility (eg, “This [health screenings] is [my] responsibility. If you don’t go, then you have a lot of illnesses, then blame who?” [U5, male, 86 years]); “For my blood pressure, if it’s high, I have to control [it]. If my blood sugar is high, my food intake has to [be] control[led]. I think it’s very useful for elderly, people like us to take care of our health” (U7, male, 57 years). However, most participants showed a tendency to be influenced by health care providers and government authorities (eg, “[When asked if the participant wanted to undergo a cataract surgery] I then refute back and say you’re a doctor, you then know what, I’m not a doctor how should I know” [N4, female, 73 years]; “over there (i.e Chinese medicine clinic) is the volunteer that comes up and helps me do (pump pressure)” [U3, male, 59 years]). These varying health LOC orientations suggested potential for further investigation (Figure 4).