Health during childhood is a key predictor of health outcomes in adolescence and adulthood [1-5]. Poor physical health during childhood that continues into adulthood can result in higher health care costs and poorer quality of life [6]. As the prevalence of childhood ill health increases [7], it is important to consider novel approaches to help improve health outcomes in children. Behavioral interventions can be used to promote positive lifestyle behaviors (eg, engaging in physical activity [PA]) as well as support the management of chronic conditions (eg, promoting treatment adherence) [8].

For adults and adolescents, there is evidence that digital health interventions are effective in promoting health behavior change while also being cost-effective, appealing, and accessible [9-11]. However, in a systematic review of digital health interventions for children, of the 17 digital interventions identified, only 5 were found to be promising in producing a positive impact on clinical outcomes [12]. The lack of effective interventions indicates a need for progress in the field of digital health interventions for children, investigating technological innovations that meet the specific needs and contexts of children.

The development of wearables and sensors presents innovative solutions for delivering behavioral interventions. There are now a large number of commercial products that enable the real-time tracking and monitoring of personal health information and behaviors such as step count, heart rate, and sleep [13]. Devices such as smartwatches are well-suited for use as health intervention tools, as they are widely available and allow users to reflect on the health information collected via sensors in real time to gain a better understanding of their health and engage in self-management behaviors [14]. In this review, we define a smartwatch as a wrist-worn device that uses sensors to collect at least 2 types of personal health data (eg, steps and heart rate) and has an interactive display. With the popularity of adult trackers, there is also an established market for wearables and smartwatches for children [15]. These devices may be more appropriate for younger users, as they can be used independently from mobile devices, as device ownership is uncommon in young children [16].

There is a growing body of evidence for activity tracker and smartwatch health interventions for adults and adolescents, but despite the promise, there has been inadequate attention to their use with children [17,18]. Designing and deploying wearable health interventions for children needs specific attention due to the distinct developmental characteristics of this age group (social, emotional, and cognitive) and their greater reliance on parents and carers for support [8,19,20]. This includes considering how wearable hardware and software should be designed to meet the distinct needs of children. It also includes considering wider systems, such as family systems, for example, considering how multiple devices should be integrated into one intervention to meet the needs of different users (eg, adult and child) and how issues such as wearable data sharing and data privacy should be navigated. To understand these technical and sociotechnical issues, understanding the end users’ experience, such as the technology’s feasibility and acceptability, is essential [21].

Few reviews have focused on the design and user experience of wearables for children. Systematic reviews of the acceptability and feasibility of wearable devices to promote PA in children have identified that wearables can increase children’s motivation for PA using behavior change techniques but noted novelty effects and uncertainties over long-term use and impact [17,18]. However, these reviews had a narrow focus on wearables for PA and a broad age range; they were not designed to explore developmentally sensitive design implications for children across health contexts. Previous work has identified common design adaptations for children’s devices such as being more colorful, having fewer health metrics (eg, calories burned not included), and including features, such as games, to boost children’s engagement [15,22]. However, these did not use a systematic review methodology and are limited to presenting a description of design adaptations without analysis of the end users’ experiences or engagement with these design adaptations.

This systematic review aimed to investigate how smartwatches and activity trackers have been used across the full range of child health interventions (for children aged 5-11 y), spanning public health and prevention interventions to interventions for clinical populations, focusing on their acceptability and feasibility for children. This review aimed to build on the established literature on digital interventions for children by investigating design considerations to support young children’s engagement with the smartwatch and the data collected for their health. Therefore, we focused on devices with an interactive display that supports data feedback. In this review, we investigated the following research questions (RQs): (1) What are the population characteristics of those being targeted in the health-related smartwatch and activity tracker interventions for children aged 5 to 11 years? (2) What are the characteristics of the smartwatches and activity trackers being used for health-related applications with children aged 5 to 11 years (eg, device type and features included)? (3) What is the feasibility and acceptability of using smartwatches and activity trackers for health-related applications with children aged 5 to 11 years?

The literature search was conducted in March 2023 in the following databases: CINAHL, ACM Digital Library, PsycINFO, Embase, MEDLINE, Cochrane Library, IEEE Xplore, PubMed, Scopus, and Web of Science. The search included key terms for (1) children aged 5 to 11 years, such as “school child” or “minor” and (2) smartwatches, such as “smartwatch” or “wearable electronic device” (Multimedia Appendix 2 provides an example of the full list of search terms used in Embase).

Studies were included if they fulfilled the following criteria: (1) peer-reviewed or empirical studies with any study design; (2) published in English; and (3) the research involved a child (aged 5-11 y) using a smartwatch for health-related purpose, including behavior change intervention (physical, mental, or social). We use the term intervention to apply to any activity undertaken to improve health [24]. For the purpose of this review, a smartwatch is defined as a wrist-worn device that uses sensors to collect at least 2 types of personal health data (eg, steps and heart rate) and has an interactive display. We also extended this to include activity trackers with a visual display (excluding those that do not reflect on the data via the device), as they are largely similar to smartwatches but are typically cheaper with fewer sensors and can be used without a smartphone [25]. We selected devices that collect at least 2 types of personal health information and incorporate a visual, interactive display, thereby excluding simpler devices such as pedometers, which lack these functionalities and therefore do not meet our criteria of being a smartwatch or activity tracker. The devices could be used in conjunction with a partnering app if the child engages with the wearable device independently as part of its use.

Studies were excluded if they fulfilled the following criteria: (1) the study did not use a smartwatch or activity tracker with an interactive display; (2) the children in the study only wore the smartwatch or activity tracker as a passive data collection tool, and they did not otherwise interact with the device, by which we mean actively collect or view the data it collected as part of an intervention or feasibility assessment (eg, accuracy or validation studies); (3) the study included children aged <5 years or >11 years or do not report on the eligible age range specifically; or (4) the article was considered a review on previously reported studies, protocol, nonempirical or nonscientific paper (eg, book chapters), or when the full text was not available.

The search results were imported into Zotero (Corporation for Digital Scholarship) and duplicates were removed. Each title and abstract were independently screened by 2 authors (LT and SC) for relevance. If an inclusion decision could not be made based on the title and abstract, the articles were included for full-text review. Articles deemed relevant at the title and abstract screening stage were independently double-screened against the inclusion and exclusion criteria using the data management platform Rayaan (Rayyan Systems Inc), with reasons for exclusion recorded. Disagreements at both stages were discussed and resolved in meetings by the reviewers (LT and SC), with persistent disagreements being resolved through discussion with the wider research team (AB and JB). If the full text did not contain the information needed, 2 attempts were made to contact authors by email; if the information was not provided, the study was excluded.

Data related to our RQs were extracted by the primary author (LT), as described in Textbox 1. Due to the heterogeneity in the study designs and the data reported, we synthesized data using narrative and thematic synthesis [26,27]. The extracted data are summarized in Textbox 1. For qualitative data, open coding was conducted by the primary author on the available qualitative data referring to participants’ experiences of using the device, which included quotations from participants or authors’ interpretations of participants’ experiences. Themes were developed top-down using the technology acceptance model (TAM) [28], in which codes generated were grouped under each relevant theme of the TAM (perceived ease of use and perceived usefulness). The remaining analysis was data driven (bottom-up) to develop themes for any codes that fell outside the scope of the TAM. Wherever possible, quotes from the participants of the included studies (primary data) have been presented. In addition, to understand the characteristics of the device relevant to behavior change, we used the taxonomy of app features [29] and coded whether the features were present on the smartwatch device or partner app in each research study as shown in Multimedia Appendix 3.

Two authors conducted quality assessments independently, discussing disagreements to reach a consensus. As the studies varied in study design, we used different quality assessment tools. We used the Critical Appraisal Skills Programme systematic review checklist [31] for qualitative studies. This is a 10-item checklist with 3 responses: “yes,” “no,” and “cannot tell.” The Quality of Survey Studies in Psychology checklist [32] was used to assess articles that collected user feedback via surveys and questionnaires. This checklist comprises 20 items under 4 categories: introduction, participants, data, and ethics. Finally, the National Heart, Lung, and Blood Institute pre-post and controlled studies assessment tools were used. These are 12- and 14-item checklists with 5 responses: “yes,” “no,” “cannot determine,” “not applicable,” and “not reported” [33,34]. Full details of the quality assessment can be found in Multimedia Appendix 4 [35-49].

After deduplication, 3312 (69%) of the 4772 articles were identified from the database searches. After abstract and title screening, 150 (4.5%) articles were considered for a full-text review, and of these, 15 (10%) were identified as eligible for inclusion. In total, 3 articles reported on the same intervention; therefore, we identified 13 studies detailing children’s (aged 5-11 y) use of smartwatches for health-related purposes. A summary of the study details can be found in Table 1. Table 2 provides a summary of the population characteristics. We will refer to these 13 distinct studies unless otherwise indicated. Most (11/13, 85%) studies focused on using a single device, but 2 (15%) studies [35,36] used multiple devices during the study period. Additional devices not meeting the study criteria (n=5) are not reported here. The most common reasons for exclusion were participants being outside the target age range, wrong publication type, and wrong device type. The PRISMA flow diagram can be found in Figure 1.

This systematic review aimed to answer three RQs by identifying (1) the population characteristics of children targeted in health-related wearable interventions, (2) the characteristics of the devices used, and (3) the feasibility and acceptability of using these devices in wearable interventions for children aged 5 to 11 years. We found that most (10/13, 77%) of the studies focused on PA, with only 23% (3/13) focusing on clinical conditions. The devices were typically used to provide activity feedback to support behavior change, with additional features often available on an accompanying app. Many (10/13, 77%) studies included wider stakeholders (parents, teachers, and peers) and data sharing, either as part of the intervention or implicitly due to the device being linked to a partner app. The quantitative and qualitative user feedback shows that using smartwatches for health interventions is acceptable and feasible with children aged 5 to 11 years, though there were common usability issues (eg, discomfort, technical issues, and short battery life). Overall, several studies included in this review lacked sufficient detail for data extraction, were considered not to have a rigorous data analysis procedure, and thus are considered to be of poor quality (Multimedia Appendix 3 provides full quality assessment). Sample sizes varied from n=1 to n=137 and small sizes limit our ability to generalize their findings, which should be rectified in future research. There was a limited focus on child-centered design, with many studies using commercial adult devices that require parent involvement rather than developing bespoke child-friendly designs. Furthermore, when gathering user feedback, children’s views were often not the focus, with 3 of the 12 (25%) studies not gathering feedback from children but instead focusing on their parents.

All of the studies in this review used existing commercial smartwatches and activity-tracking devices. These provide limited data feedback, often only reporting step count on the device. More detailed information was generally only available through a partner app, which was usually under the control of a parent. This is consistent with previous research, which found that children were frustrated by the limited data feedback available in real time due to it being on the app rather than the device [50]. The devices and their partner apps included a number of behavior change techniques, such as feedback or monitoring and goal setting, both of which are consistent with previous child and adult research [51,52]. The studies also identified usability issues, including discomfort, technical challenges such as syncing issues, and logistical challenges such as poor battery life, which is consistent with previous research [50,53,54].

Despite this, the studies in this review reported high levels of engagement and enjoyment using these devices and their partner apps. In total, 2 (13%) of the 15 studies also reported a novelty effect, indicated by declining use across the study period. This effect has also been observed with adults’ use of wearable devices, where it is common to see initial high engagement levels, followed by abandonment after a few months [55]. One potential way to combat abandonment and support motivation is through the use of gamification, as research has shown that gamified apps have longer use than nongamified apps for PA [56]. Furthermore, while the children were mostly content with wearing the devices, in one of the studies participants described that children were not actively engaging with the data collected via the device. This highlights an important distinction between engagement with the data collection process and the use of or reflection on the data collected. According to the staged-based model of personal informatics by Li et al [57], reflection is an important distinct stage that takes place before action, in which individuals can enact behavior change with their new information. Previous research has shown that the way people reflect on their data can be impacted by both the phase of self-tracking and the feedback they receive such as data visualizations, which help individuals gain insights from their data [58]. This is particularly relevant as children can struggle to understand the data collected and visualized on smartwatches [59], which, in turn, may impact their engagement. For example, research with children aged 10 to 15 years has shown that children can misinterpret biofeedback metrics, including heart rate and calories [60,61]. Therefore, further work is needed to understand how to support children’s sensemaking of personal health data collected via wearables.

To ensure that children understand and can reflect on the data being collected via smartwatches, it is important to consider their needs and preferences by including them and other stakeholders in a user-centered design process [62]. Considering the end users’ experience and including these individuals in the design process has been shown to produce technology, which is more engaging and acceptable to users [63,64]. This is particularly relevant when using children’s wearables in health interventions, as research has shown that children conceptualize health differently from adults. It was found that children extended their understanding of health beyond traditional health measures to include other metrics, such as time spent playing outside [59], highlighting the importance of exploring children’s health-tracking interests. However, many studies in this review did not seem to take a user-centered approach to the design and choice of devices in their studies. For example, a few (3/13, 23%) studies in this review outlined the rationale for their choice of wearable and failed to detail the specific design features of the device or software that supported children to use it in relation to their health. In addition, most of the devices used across the studies were adult devices and not designed specifically for use by children. Moreover, the device partner apps included additional features and behavior change techniques, such as leaderboards, games, and rewards; however, these were often only accessible via a parent’s smartphone as young children did not have their own smartphone. This can lead to the issue identified in previous research of parents acting as “gatekeepers” to their child’s interaction with an intervention. This dynamic can result in children no longer being the focus of the design and generates additional “invisible” work for parents [50,51].

However, including parents in pediatric behavioral interventions has previously been shown to be a key component for achieving sustained behavioral change [8,12]. In this review, 77% (10/13) of the studies included additional stakeholders such as parents and teachers, but it was rare that they considered the child’s perspective on their inclusion. Ethical and privacy issues around collecting and sharing self-tracked data via wearables are rarely discussed [65]. Potapov and Marshall [66] found that teenagers had concerns about the consequences of sharing their data with others; however, this has not been thoroughly investigated in younger children. In this review, privacy concerns were not directly explored or raised by younger children. However, previous research that has investigated the impact of sharing children’s PA and sleep data with their parents or carer found that children reported being uncomfortable with the surveillance. Furthermore, shared access to the data appeared to influence trust between children and their parents or carers negatively, as parents used the data to assert further parental control and impose additional restrictions [67]. Similarly, when using co-design methods with families with children aged 7 to 15 years, it was found that parents and children raised concerns about the impact of data sharing on children’s privacy and autonomy, to which the authors suggested the need for child-controlled data sharing options to help set boundaries and minimize conflict [68]. As these privacy concerns were raised in studies with children older than the focus of this review, this may be indicative of a change that occurs during the transition to adolescence, whereby children are working toward increased independence and autonomy [69]. A robust investigation of end users’ views (children, parents, carers, and other stakeholders) on wearable data sharing and data privacy is an important research gap, with significant implications for design consideration for wearable interventions for children. This should be considered in future research, where it would be valuable to investigate both children’s current preferences for data sharing and privacy as well as their shifting views as they mature and the sociotechnical design considerations that could meet these changing needs.

In this review, most (9/13, 69%) studies focused on using smartwatches and activity trackers to promote PA. Future work should consider the potential of using these devices for more novel health applications. One growing use case for smartwatches within health care and health care research is just-in-time adaptive interventions (JITAIs) [70,71]. JITAIs use mobile and sensing technology to offer personalized and timely support [72] and have been used in a variety of adult and adolescent health domains including PA [70] and substance use behavior [71,73]. In this review, 23% (3/13) of the studies used the devices to deliver interventions in real time. This included using the watch display and SMS text messages to deliver emotional regulation strategies to children with ASD. However, most of the devices in this review only provided basic data summaries on the watch face. Only one study reported the use of a novel visualization using data analytics, in which an animated dragon was used to convey a risk score as part of a JITAI. This highlights the immense potential for novel uses of smartwatches in pediatric health and further work is needed to explore how smartwatch interventions can be adapted to be developmentally appropriate to convey health information to children in a way that is meaningful to support behavior change.

Digital health is an interdisciplinary field, and a key strength of this review is the interdisciplinary approach taken by searching both health science databases as well as human-computer interaction venues. The review aimed to investigate how wearables are designed and used for children (5-11 y) to understand developmentally sensitive design implications. To investigate this, we looked across the broad spectrum of child health behaviors and conditions, which means our review is not limited to PA interventions, making the findings relevant to a broader scope of researchers working within pediatric wearable interventions. However, we excluded studies that included our target age group (5-11 y) but also included older and younger children (eg, 5-18 y) that did not report the results separately, as it was not feasible to stratify the results by age. We also conducted a narrative synthesis, which has previously been criticized for lacking transparency [74]. To mitigate this, we have included primary data where possible in our qualitative summary. The limited qualitative data available in the articles also resulted in our synthesized themes having limited richness.

This systematic review addresses the gap in understanding children’s wearables by expanding the scope of the review to a broad range of health applications, with a focus on identifying developmentally sensitive design features and understanding the experience of younger children. We identified a total of 15 articles (13 distinct studies) that involved children aged 5 to 11 years using a smartwatch for health-related purposes. Most of these were using commercial smartwatches and activity trackers as tools to help increase PA, but other novel applications included uses to support children with ASD and asthma. Of those that gathered user feedback, experiences using the devices were positive, although there was an indication that the children may not have been engaging with the visualization of their data. Overall, the studies lacked a child-centered focus with (1) many (9/15, 60%) studies not including children’s opinions reflecting on the designs, (2) most of the functionality and data visualization was placed on partner apps usually under the control of parents, and (3) most smartwatches and activity trackers being used were adult devices. Further work is needed to better understand children’s perspectives on the design and evaluation of these devices while also investigating other potential areas of research beyond PA tracking for children, such as managing chronic health conditions. In addition, future research should provide richer details about the devices used, including the rationale for the choice of device, and who has access to the data to consider how this aligns with their research aim and user experience.