This scoping review was systematically conducted following the methodological framework proposed by the Joanna Briggs Institute (JBI) [22]. The review was reported in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) checklist (Checklist 1) [23]. Ethical approval was not required for this scoping review, as no new data were collected.

The research questions of this scoping review were identified as follows:

A total of 1813 records were identified through searches of 5 electronic databases. After removing 48 duplicates, 1765 records remained for title and abstract screening, of which 1728 were excluded based on relevance. The full texts of the remaining 37 articles were assessed for eligibility, resulting in the exclusion of 28 studies. An additional 5 relevant studies were identified through reference list screening. Finally, 14 studies were included in this scoping review. The PRISMA flow diagram of study selection is shown in Figure 1.

The publication years of included studies ranged from 2015 to 2025. The majority (n=9) were conducted in the United States [28-36], while the remaining 5 were carried out in Ireland [37], Australia [38], the Netherlands [19], Singapore [39], and the Czech Republic [20], respectively. All studies were described as feasibility or pilot studies, including 4 qualitative studies [19,31,37,38], 3 randomized controlled trial (RCT) studies [28,29,32], 6 single-armed feasibility or pilot studies [20,33-36,39], and 1 prospective cohort study [30]. Sample sizes ranged from 8 to 108 participants. Among those, 4 studies recruited individuals with T2D [19,20,35,38], 3 targeted individuals undergoing or completing cardiac rehabilitation [28,31,33], and 2 included individuals with stroke [30,37]. The remaining studies enrolled participants with hypertension (n=1) [29], spinal cord injury (n=1) [34], acute coronary syndrome or those who had undergone coronary revascularization (n=1) [33], individuals who were overweight or obese (n=1) [36], and older adults with a chronic condition (n=1) [39]. More details of the characteristics of included studies are presented in Table 2.

Among the included studies, 13 reported on the design characteristics of JITAIs, with 2 studies using the same version of the JITAIs [31,32], resulting in a total of 12 unique JITAIs. Additionally, 8 of the JITAIs were grounded in theoretical frameworks [19,20,29,32-34,38,39], with social cognitive theory being the most commonly applied [32,34]. Overall, 6 JITAIs [28,30,32-34,37] were designed to primarily target PA behavior, 1 focused solely on sedentary behavior [36], and 5 addressed both sedentary behavior and PA [20,29,35,38,39]. Most studies used wearable devices, smartphone apps, or both to collect PA or sedentary behavior data. Wearable or sensor-based data were used in studies (n=11) [20,28,29,31-36,38,38,39] using Fitbit or smartwatch devices, activity trackers, smartphone accelerometers, or external accelerometers. Several studies (n=4) also incorporated EMA or self-report data. Additional contextual data, such as location, weather, time of day, or day of week, were used in Hietbrink et al [19], Golbus et al [28], Daryabeygi-Khotbehsara et al [38], Mair et al [39], and Novak et al [20]. Two studies [31,32] also included app-entered clinical or health data, such as symptoms, blood pressure, heart rate, medication use, or self-reported health measures.

Tailoring variables could be grouped into 5 categories: behavioral variables, contextual variables, self-reported momentary states, goal/progress variables, and clinical or safety-related variables. Behavioral variables included step count, sedentary bout duration, walking status, activity intensity, and recent PA patterns [20,28,29,31-36,38,39]. Contextual variables, including time of day, weather, location, weekday/weekend, or study phase [19,20,28,38,39]. Self-reported momentary states, such as mood, fatigue, restrictions, cravings, readiness, and energy [19,33,34]. Goal- or progress-related tailoring was used in 6 studies [20,29,33,34,38,39]; for example, Novak et al [20] tailored messages according to recent step counts, weekly step goals, and individual action plans. Clinical or safety-related tailoring, such as symptom-triggered safety responses, was mainly reported in Elnaggar et al [31] and Park et al [32].

Decision points also differed across studies. Some interventions (n=4) [20,34-36] used sensor-triggered decision points, where prompts were delivered when a behavioral threshold was detected. For example, Pellegrini et al [35] delivered prompts after 20 minutes of consecutive sedentary time; Thomas et al [36] tested prompts after 30, 60, or 120 minutes of sedentary behavior; Novak et al [20] used Fitbit-detected inactivity or walking patterns; and Hiremath et al [34] triggered feedback when moderate-intensity wheelchair-based activity was detected. Other studies (n=5) [28,31,32,38,39] used scheduled or semi-scheduled decision points. Golbus et al [28] and Daryabeygi-Khotbehsara et al [38] also used microrandomized decision points, where participants could be randomized to receive or not receive an intervention message at predefined times or contexts. EMA-based decision or assessment points were used in Hietbrink et al [19] and Sengupta et al [33].

Intervention options could be grouped into PA prompts (n=9) [19,20,28,31-34,38,38,39] sedentary break prompts (n=5) [20,35,36,38,39], goal-setting or planning support (n=7) [20,28,29,33,34,38,39], feedback and encouragement (n=8) [19,20,29,31-34,39], education or coping support (n=5) [19,20,31-33], and safety prompts, which were specifically reported in Elnaggar et al [31] and Park et al [32], where participants reporting symptoms such as chest pain or shortness of breath received safety-related guidance. Decision rules were mostly simple and rule-based, which could be classified as EMA-based if-then rules, sensor-triggered if-then rules, sedentary-threshold rules, scheduled rules, microrandomization rules, and coach- or staff-supported rules. EMA-based if-then rules were used in 2 studies. Hietbrink et al [19] matched text messages to EMA-reported activity, mood, location, restrictions, weather, or cravings, while Sengupta et al [33] used EMA and goal information to trigger reminders or reinforcement messages. Four studies [20,34-36] adopted sensor-triggered or sedentary-threshold rules, where prompts were delivered after predefined behavioral states such as inactivity, low-intensity walking, moderate-intensity wheelchair activity, or prolonged sedentary time. Scheduled rules were used in 4 studies [20,31,32,39], with messages delivered at fixed daily or weekly times. Probability-based or availability-based microrandomization was used in Golbus et al [28] and Daryabeygi-Khotbehsara et al [38], where participants were randomized to receive or not receive messages at predefined decision points, sometimes conditional on availability. Four studies [29,31-33] used coach- or staff-supported decision rules, where participant data were reviewed by coaches or study staff to support tailored feedback, motivational messages, or safety guidance.

The adaptation mechanisms could be grouped into sensor-based tailoring, contextual tailoring, self-report-based tailoring, goal/progress-based tailoring, and coach- or staff-supported tailoring in this study. Sensor-based tailoring was used in 6 studies [20,34-36,38,39], which adapted messages according to step count, sedentary time, activity intensity, or walking patterns. Contextual tailoring based on time, weather, location, or day of week was used in 5 studies [19,20,28,38,39]. Two studies [19,33] adopted self-report-based tailoring, which used EMA responses such as mood, fatigue, restrictions, cravings, or readiness for activity. Goal/progress-based tailoring (n=6) [20,29,33,34,38,39] adapted feedback according to whether participants were meeting activity or sitting goals. Four studies [29,31-33] used coach- or staff-supported tailoring, which involved study staff or health coaches reviewing participant data and providing individualized feedback. Furthermore, a total of 12 studies reported intervention duration of JITAIs, which ranged from 7 days to 6 months, and was typically 2 to 12 weeks. Further details regarding the design characteristics of the JITAIs are provided in Table 3.

Engagement and implementation-related measures were inconsistently reported across studies. Reported measures included study retention, wearable-device use, EMA completion, app use, goal setting, message exposure and responsiveness, intervention-delivery fidelity, usability, and satisfaction. Four studies [33,34,36,39] reported quantitative engagement and delivery-fidelity data. Hietbrink et al [19] and Daryabeygi-Khotbehsara et al [38] mainly reported qualitative acceptability or development-stage feedback, while Golbus et al [28], Bai et al [29], Elnaggar et al [31], and Park et al [32] provided limited objective usage or response data. Further details are presented in Multimedia Appendix 2.

Of the 14 included studies, 7 objectively measured PA. Three RCTs [28,29,32] reported preliminary evidence of a positive effect of JITAIs on promoting PA. In the study by Golbus et al [28], participants’ 60-minute step counts significantly increased within the first 30 days following the intervention. In the study by Park et al [32], the intervention group had a higher mean daily step count compared to the control group (P=.02). Similarly, Bai et al [29] found that participants in the Fitbit plus health coaching group increased their daily step count (P<.01) and moderate to vigorous physical activity (MVPA) (P<.05), with an improvement in daily MVPA compared to those in the Fitbit-only group (P<.05). Four nonrandomized studies reported PA outcomes, with three demonstrating positive effects [30,34,35]. Studies by Hiremath et al [34] and Pellegrini et al [35] reported increases in PA levels compared to baseline (both P<.05). The study by Lau et al [30] also showed that a higher proportion of light and moderate to vigorous PA was recorded among participants receiving JITAIs. However, the study conducted by Sengupta et al [33] found no significant change in PA levels compared to baseline.

Three studies [29,35,36] objectively measured sedentary behavior, including 1 RCT and 2 nonrandomized studies. In the RCT conducted by Bai et al [29], participants in the Fitbit plus health coaching group preliminarily demonstrated a reduction in sitting time (P<.01). Among the 2 nonrandomized studies, both reported a reduction in sedentary time or behavior compared with baseline (P<.05) [35,36]. Further details are provided in Multimedia Appendix 3.

To the best of our knowledge, this is the first scoping review to systematically map and synthesize the existing literature on JITAIs specifically aimed at reducing sedentary behavior and promoting PA in individuals with chronic disease. A total of 14 studies were finally included in this review, and we systematically summarized the key characteristics of delivery methods and components used in JITAIs designed to reduce sedentary behavior and promote PA, and explored the existing evidence regarding the evaluation of these interventions in improving PA levels and reducing sedentary time. Additionally, this review also identified reported barriers and facilitators associated with the use of JITAIs among individuals living with chronic diseases. The findings contribute to a clearer understanding of how JITAIs are currently operationalized and provide a foundation for guiding future research and intervention development in chronic disease management.

Among the included studies, most of them provided participants with instructions on how to operate these technologies. As the integration of wearable activity trackers becomes more common, user training and support are essential to maximize engagement and minimize technological barriers [20,28]. The widespread use of built-in accelerometers and GPS sensors indicates that sensor-based context detection is a central feature of current JITAIs [30,35]. By identifying timely opportunities for intervention, these technologies may support more personalized and responsive behavioral prompts. Additionally, the intervention options and decision rules identified in this review suggest that most JITAIs used relatively simple designs. The frequent use of PA prompts, goal-setting or planning support, and feedback/encouragement indicates that most interventions focused on motivating participants to initiate or maintain PA, whereas fewer studies specifically targeted sedentary breaks. Safety prompts were reported only in Elnaggar et al [31] and Park et al [32], suggesting that symptom- or risk-related tailoring remains limited despite its relevance for patients with chronic disease. Decision rules were also mostly predefined and rule-based, which, although practical and transparent, indicates that many current JITAIs are adaptive mainly to predefined behavioral or contextual triggers rather than continuously learning from individual responses over time. Future research could therefore explore more dynamic and data-driven adaptation strategies that can continuously learn from users’ evolving behaviors and intervention responses.

In this review, most included studies reported preliminary evidence of a positive effect of JITAIs in promoting PA among patients with chronic disease. JITAIs provide personalized, real-time support based on individual context, such as location, time of day, or recent activity levels, which may enhance the relevance and timeliness of prompts and promote behavioral engagement [15,16]. The use of wearable sensors and mobile apps may also support continuous monitoring and feedback, potentially enhancing motivation and self-awareness [15,16]. In addition, many interventions incorporated behavior change techniques such as goal-setting, reminders, and feedback, which have been shown to support PA improvement [40]. However, the findings from one study [33] found no significant change in PA levels compared to baseline. One possible explanation for the lack of significant change in PA levels in the study could be the exclusive inclusion of female participants. Factors such as caregiving responsibilities, differing motivational drivers, and social or cultural expectations may influence women’s engagement with PA interventions [41]. Three studies [29,35,36] reported preliminary evidence of a positive effect of JITAIs in reducing sitting time and promoting breaks in sedentary behavior. The effects may be related to the real-time and context-sensitive design of JITAIs, which allows prompts or support to be delivered when individuals are likely to be sedentary, such as after prolonged inactivity or during sitting-related contexts, including watching television or working at a desk [29,35,36]. The timely cues may help interrupt habitual sedentary patterns and encourage individuals to take breaks, stand, or engage in light activity. However, the overall evidence base for JITAIs in promoting PA and reducing sedentary behavior remains limited by small sample sizes, few RCTs, and short intervention durations. In this review, only one study [28] had an intervention duration of 6 months, whereas the remaining studies involved relatively shorter intervention periods. The limited durations may not be sufficient to determine whether JITAIs can produce sustained behavioral changes or maintain long-term user engagement. Future research could therefore use RCT designs adequately, include longer intervention and follow-up periods, and recruit more diverse chronic disease populations across different clinical and community settings to further establish the effectiveness of JITAIs.

The identification of multiple barriers to JITAI use among patients with chronic diseases highlights important challenges that need to be addressed to optimize their utility. Most studies reported that technological usability and literacy emerged as a significant obstacle. Issues such as small text sizes, rigid app settings, and unreliable detection of low-intensity or brief activities can compromise user experience and reduce engagement, aligning with previous findings, which emphasize the digital divide, particularly among older or less technologically experienced populations commonly affected by chronic conditions [42,43]. Future interventions can adopt inclusive design principles that accommodate users with varying levels of digital literacy. This involves simplifying user interfaces, enabling customizable text sizes and display settings, and ensuring device compatibility with a wide range of PA levels, particularly low-intensity movements common among individuals with chronic diseases. Incorporating user-centered design through participatory development processes may enhance both usability and acceptability by aligning intervention features with patient needs and preferences [31,37]. Additionally, the burden and intrusiveness of the intervention represent another critical barrier. Patients reported perceptions of excessive time demands, repetitive or not engaging content, and misalignment with their daily routines, which may undermine adherence [20,34,38,39]. The results underscore the necessity for interventions that are not only effective but also seamlessly integrate into users’ lifestyles, minimizing disruption and maximizing relevance. Moreover, perceived value and acceptability also influence intervention uptake. Limited perceived usefulness and psychological resistance, potentially stemming from skepticism toward new technologies or concerns about personal capability, suggest that patient education and motivational strategies are crucial components for successful implementation [19]. Privacy and data security concerns were reported in several studies, reflecting widespread apprehension regarding data sharing and protection, similar to the findings from Muhunzi et al [44]. Future interventions should ensure transparent data policies and robust security measures while granting users control over their personal information to build trust, enhance sustained engagement, and support the further development of JITAIs.

This scoping review has several limitations that should be acknowledged. First, although a comprehensive search strategy was used, some relevant studies may have been missed due to publication bias or limitations in database indexing. In addition, only peer-reviewed articles published in English were included, which may have excluded relevant studies published in other languages or gray literature. Second, the heterogeneity of the included studies in terms of population characteristics, intervention designs, and duration limited our ability to quantitatively synthesize findings or make direct comparisons across studies. Third, many included studies had small sample sizes and short follow-up periods, which may limit the generalizability of the findings and make it difficult to determine whether the reported effects can be sustained over time. Additionally, most studies were conducted in high-income countries, which may limit the applicability of findings to low- and middle-income settings. Therefore, the conclusions should be interpreted with caution. Future systematic reviews with meta-analyses and risk-of-bias assessments are warranted to more rigorously evaluate the effectiveness of JITAIs in promoting PA and reducing sedentary behavior.

This scoping review identified preliminary evidence suggesting that JITAIs may help reduce sedentary behavior and promote PA among patients with chronic diseases. However, the evidence supporting this conclusion remains limited by the small number of RCTs, small sample sizes, short intervention and follow-up periods, and heterogeneity across studies. Therefore, these findings should be interpreted as preliminary. Future research could use RCT designs with sufficient sample sizes, longer follow-up periods, and more diverse chronic disease populations to further establish the effectiveness and generalizability of JITAIs.