We conducted a scoping review following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [29]. The final protocol was registered prospectively with the Open Science Framework on March 23, 2023. This study was conducted in the context of the ABOARD (A Personalized Medicine Approach for Alzheimer’s Disease) project, which aims for a future with personalized, patient-orchestrated diagnosis, prediction, and prevention of AD [30]. It is a large-scale research project carried out by a multidisciplinary consortium in the Netherlands.

We defined digital health tools as “smart devices and connected equipment that improve health by having patients actively interact with them,” such as mobile apps, digital platforms, or wearables [14]. We searched online search engines and databases in January 2023, including PubMed, IEEE Xplore, Ovid, and Web of Science. The search strategy comprised terms related to AD and dementia, (pre-)disease stages, digital tools, and different purposes of digital tools (eg, “diagnosis,” “dementia prevention,” or “prediction”), and their synonyms). The query was co-developed with the research team. The final search queries for the various engines and databases can be found in Multimedia Appendix 1. Definitive search results (ie, records) were exported to Rayyan software [31]. Duplicate records were automatically identified by Rayyan, and potential missing duplicates were manually identified by the researchers and subsequently removed. Two researchers (CHN, TJdR) independently reviewed all identified records based on our inclusion and exclusion criteria. When in doubt, an extra researcher (TE) reviewed these records as well. Articles were included if they described research on a digital tool developed in the field of AD or dementia, with people without an AD or dementia diagnosis as the intended primary end users, that is, those interacting with the tool. Additional inclusion criteria comprised full text availability, peer-reviewed articles written in English or Dutch, the articles addressing humans (not animals), and articles containing original empirical research data. As the aim of this study is to provide a state-of-the-art overview of recent digital tools reported on in academic literature, we applied a 10-year filter (2013‐2023) to include relevant tools while excluding outdated technological developments (Textbox 1).

Titles and abstracts were reviewed on these inclusion criteria. For the remaining set, full-text articles were screened. The researchers compared their screening decisions and discussed these until they reached consensus.

A data extraction sheet was developed based on our aims. For the overview of available tools, we (CHN, KKMK, TJdR) extracted information on the name of the tool, purpose of the tool, place of use (ie, home or clinical setting), and self-guidance (ie, fully self-guided or researcher/clinician-assisted). Data on study samples (ie, the total number of participants, cognitive status, age, sex, educational attainment, ethnicity, and digital literacy) were extracted by KKMK and checked by TJdR. We extracted data that were explicitly mentioned in the demographic section of an article, as well as data that were implicitly mentioned in the inclusion and exclusion criteria of the article (eg, if digital literacy was an inclusion criterion, we assumed that the sample was probably digitally literate). Data regarding the stage of developmental maturity and stage of evaluation were extracted and categorized (TJdR) in accordance with the World Health Organization guide [32]; in case of doubt, these were discussed with others (TE, LNCV, EMAS). Early maturity stages comprise a preprototype, prototype, or pilot product [32]. Mid-maturity comprises the demonstration phase that assesses the costs and implementation requirements, in which the effect of the tool is tested in an uncontrolled situation limited to a certain population or geography [32]. The advanced stage of maturity comprises scale-ups (ie, tools that are ready to be optimized and scaled up across multiple [sub]national and population levels) or integrated and sustained programs (ie, integrating tools in a broader health system and determining components of the enabling environment to maximize impact at a large scale) [32]. Stage of evaluation is related to the aforementioned stages of maturity and informs on the level of evidence for a digital tool, thus informing on implementation fidelity. The early stage of evaluation comprises feasibility (ie, assessing if the tool works in a given context), usability (ie, assessing if the tool can be used by users), and efficacy testing (ie, assessing if the tool can achieve the intended results in a controlled setting) [32]. Mid-stage evaluation encompasses effectiveness testing, which informs if the tool can achieve the results in a noncontrolled setting [32]. The late stage of evaluation addresses implementation issues regarding the tool’s uptake, integration, and sustainability in a given context, including policies and practices [32]. If multiple stages of maturity and evaluation were present, we chose the most advanced stage (eg, if early and mid-maturity were described, we categorized the study as mid-mature). In this study, we assessed the stage of maturity and stage of evaluation of a tool in relation to a specific study context and population as described in a scientific article. For instance, a Wii is a mature digital tool by itself, but it was evaluated as an early-stage tool when the scientific article was a pilot study among people with cognitive complaints that assessed the usability of the tool. We (KKMK, TJdR) also extracted data on future directions bottom-up from Future Research or Future Directions sections in the Discussion section of the included articles.

Next, a narrative synthesis on the main findings in relation to our main objectives was written by TJdR. We grouped the included articles by tool category (ie, primary prevention, secondary prevention, daily life support, self-administered screening, and decision support). We defined dementia prevention tools as tools that aim for primary (healthy people and people with SCD) and secondary dementia prevention (people with MCI), including both single- and multidomain interventions. Daily life support tools comprised tools that aim to support daily life for people living with cognitive complaints, enhance overall well-being in relation to the cognitive complaints, and improve quality of life. The category self-administered screening tools comprised self-administered AD/dementia-risk assessment tools and tools that assess cognitive functioning, including memory, language, and perception. These types of tools may involve questions and/or tasks to complete that could indicate whether there is any ground for concern and do not replace a detailed clinical assessment required for a formal diagnosis of dementia or MCI [33]. Decision support tools consist of tools containing a health care–related decision aid for people without an AD or dementia diagnosis. Subthemes resulting from the data extraction of reported future directions were grouped into overarching themes.

The search yielded a total of 3056 records, of which 2811 records remained after removing duplicates. After screening the titles and abstracts, 408 articles were subjected to full-text screening. Finally, 39 articles met the inclusion criteria and were found suitable for data extraction (Figure 1).

The following study populations could be identified: people with MCI (12/39, 31%), cognitively healthy people (11/39, 28%), (subjective) cognitive impairment (9/39, 23%), or “no dementia” (1/39, 3%). Six articles (15%) included people with a variety of cognitive abilities (Table 1). Most articles included people aged 60 years and older (>50% >60 years: 23/39, 59%), with more female representation (>50% female: 27/39, 69%) and highly educated study participants (>50% high education: 13/39, 33%; Table 1). Twelve articles (12/39, 31%) reported data on the ethnic background of participants. These articles predominantly included White people or those of European ancestry (Table 1). Of the 39 articles, 11 (28%) explicitly discussed the digital literacy levels of their study participants. These articles revealed that the majority of their samples consisted of individuals proficient in digital skills. In 2 articles, around half of the participants lacked experience in using digital technologies (Table 1). Regarding the inclusion and exclusion criteria of the included articles, we found some articles (9/39, 23%) mentioning digital literacy and/or having internet access and/or the possession of a smartphone, tablet, or computer as an inclusion criterion [34-42]. None of the articles mentioned educational attainment or ethnicity as an inclusion criterion. For one study, the recruitment was done fully online, potentially resulting in a sample of people with sufficient digital literacy skills [43].

Table 2 provides an overview of the included articles, together with a description of the author, year of publication, country, name of tool, purpose of tool, type of tool, fully self-guided (yes/no), home setting (yes/no), stage of maturity, and stage of evaluation. Almost 70% of these tools have been developed since 2018 (27/39, 69%). A variety of types of tools was found, including web-, tablet-, or smartphone-based apps (n=25); websites (n=12); and game consoles or video games (n=2) (Table 2). Tools were mainly used in the home setting (n=30) and fully self-guided (n=30) (Table 2). If tools were used in a clinical setting (n=9; Table 2), this was usually under the supervision of a health care professional or researcher.

We identified 4 main categories of tools (Table 2). The first category was tools aimed at dementia prevention, including primary prevention (14/39, 36%) and secondary prevention (11/39, 28%). Primary prevention–related tools generally comprised lifestyle-related risk reduction, whereas secondary prevention–related tools generally focused on enhancing cognitive functioning. Both primary and secondary prevention tools did so via single-domain interventions, such as cognitive training [36,37,43,45,46,48-51,53,57,59] (primary prevention: 8/15, 53%; secondary prevention: 4/10, 40%), or multidomain interventions, such as improving physical activity, social activity, mental activity, lifestyle, and attitude [34,35,38,39,41,44,47,52,54-56,58,73] (primary prevention: 7/15, 47%; secondary prevention: 6/10, 60%). Second, daily life support tools (8/39, 21%) frequently aimed to help people manage symptoms and/or decrease the impact of cognitive complaints on daily life, for instance, via memory support or enhancing quality of life. The main aim of the self-administered screening-related tools (4/39, 10%) was to assess current cognitive functioning and/or the risk of developing AD/dementia. Finally, decision aids (2/39, 5%) had an overarching goal of supporting shared decision-making regarding certain health care decisions (eg, via preference elicitation).

Most tools were in an early maturity stage of evaluation/testing (31/39, 80%) [32], comprising preprototyping (1/35, 3%), prototyping (15/35, 43%), or pilot testing (19/35, 54%), and not yet in the stage of demonstration in uncontrolled conditions (Table 2). Overall, the stage of evaluation [32] mainly comprised efficacy testing (18/40, 45%), followed by usability testing (12/40, 30%) and feasibility testing (10/40, 25%; Table 2). Usability and feasibility testing were often combined with each other, whereas efficacy or effectiveness testing were usually performed independently. Some studies combined multiple forms of maturity and evaluation testing (eg, prototyping and pilot testing or feasibility and usability testing), explaining why there are more total testing methods than papers. Eight tools were in a mid-maturity stage and evaluation stage (8/39, 21%).

Five main themes related to future research directions were identified (Table 3). First, many suggested directions were related to early tool development, whereby authors suggested an array of directions surrounding future early development steps or recommended early development steps in general, such as the need for (further) prototyping or developing a digital tool via co-design (Table 3).

Second, some studies comprised a digital tool in the early development stage and looked forward to preparing for the mid-maturity stage. These studies, for instance, contained future directions surrounding implementation research or mentioned the need for efficacy testing of their digital tool (Table 3). Third, some studies even mentioned future directions regarding advanced development stages, such as digital tool testing in multiple settings. Fourth, themes regarding the diversity and inclusivity of both the study population and the content of the tool were mentioned as future directions (Table 3). Finally, multiple subthemes regarding study advancements were mentioned. These were usually mentioned after reflecting on the study and its limitations, such as the need for methodological changes. Some of these study advancement directions were mentioned in light of next steps needed, such as longitudinal randomized controlled trial testing (Table 3).

This scoping review sought to create a state-of-the-art overview of digital tools for people without an Alzheimer’s disease (AD) or dementia diagnosis as currently reported in the scientific literature. Our scoping review shows a considerable number of tools that people without an AD or dementia diagnosis can use by themselves, with almost 70% of articles published in the last 5 years (27/39, 69.2%). Digital tools focused on training cognitive functioning, improving lifestyle, improving daily life functioning, assessing current cognitive functioning and/or the risk of developing AD or dementia, and decision support regarding brain-related health care. Yet, the majority of digital tools reported on were directed at primary or secondary prevention, including both single-domain and multidomain interventions. These findings underscore the recent trend toward dementia prevention efforts in dementia-related health care [11-13,74].

We found no digital tools that calculate an individual’s dementia risk estimate and few self-administered screening-related tools that people can use by themselves. The shift toward early prediction and diagnosis may be too recent to already have resulted in scientific reports on these types of digital tools [8]. In addition, this shift toward earlier disease phases and greater responsibility for individuals to take an active role in their own brain health and dementia prevention comes with new ethical, legal, and societal implications such as the risk of misinformation, potentially enlarging existing health inequities, the balance between responsibility and autonomy, data privacy issues, ownership of sensitive health data, conflict with other non–health-related values people hold in life, the risk of increasing anxiety, stigmatization, medicalization, and increased social pressure on the individual [75-78]. These issues may need to be addressed first in order to develop prediction and self-administered screening tools that are safe and acceptable to use by people themselves, without professional involvement or assistance. Nevertheless, some digital prediction and diagnostic tools are already available for memory clinic physicians [79-82]. Given the trend toward greater individual responsibility and timely diagnosis, we expect that more tools will emerge in the coming years that cater toward individuals with cognitive complaints or those at risk as the primary end users.

We identified fewer digital tools to support daily life for people with SCD or MCI than expected [83,84]. For instance, only 2 decision-support tools could be identified in academic literature in this review. In other populations, research shows that decision-support tools may enhance shared decision-making processes, increase patient knowledge and empowerment, reduce decisional conflict, and improve patient-provider communication [85,86]. Therefore, developing decision-support tools may also become increasingly relevant for the field of AD and dementia as recent biomedical developments allow for early diagnosis, dementia risk prediction, and dementia prevention, as this comes with increasingly complex decisions to be made by individuals and their care partners (for instance, what treatment burden and risk is someone willing to accept) [87]. We did not find any wayfinding digital tools that people with MCI could use as end users. We found some articles on wayfinding systems; however, we had to exclude these as the person with MCI was not the main end user. These finding thus highlight a potential research gap and opportunity for future support.

Most digital tools for cognitively unimpaired people or people with SCD or MCI in the context of AD or dementia as described in the literature are still in (early) development, in which usefulness, feasibility, and initial efficacy are investigated, rather than in a phase of widespread implementation and effectiveness testing. Primary dementia prevention tools seem to be slightly further in terms of maturity and evaluation, indicated by 5 articles that report on tool effectiveness [37,39,43,46,49]. Our overview of digital tools, including their stages of maturity and evaluation, shows that the evidence gathered for digital tools for cognitively unimpaired people or people with SCD or MCI in the context of AD or dementia in the literature is limited. Health care policies should therefore not yet be based on the availability of evidence-based digital tools.

With regard to study samples, our findings show an overrepresentation of women aged 60 years and older and underrepresentation of diverse study populations regarding digital literacy and ethnicity. The World Health Organization states that digital technology plays a crucial role in achieving universal health coverage and adds that digital tools are essential for promoting health and serving the disadvantaged [88,89]. Development and evaluation processes of digital tools should thus take an inclusive approach, with representative study samples when investigating the usability and feasibility of digital tools [90]. Study samples in research on digital tools in the context of dementia prevention and self-administered screening should be diverse in terms of sex, gender, race and/or ethnicity, religion, disability, educational level, socioeconomic position, and people with marginalized status (eg, Indigenous Peoples), as these characteristics can affect a person’s risk of dementia [74,91]. In addition, it is important to include people with varying levels of digital health literacy (DHL). DHL encompasses the ability of individuals to seek, comprehend, and apply health information from digital sources, using various technologies and digital platforms [92]. It involves the skills and knowledge required to navigate, evaluate, and use digital tools, websites, and apps to access health-related information and services [92]. Although commonly used methods to assess DHL skills are available [93,94], we found that current research on digital tools for dementia prevention and self-administered screening lacks the reporting of such DHL skills. Approximately one-quarter of papers discussed the computer-, Wii-, and smartphone-related expertise of participants; however, these participants were often digitally proficient. Additionally, around 70% of papers did not report anything on the DHL skills of their participants. Moreover, people can experience barriers, such as difficulty navigating the system, that hamper the successful use and adoption of digital tools due to a chronic disease or general aging processes [95-97]. These barriers are well-known, but people with low DHL nevertheless remain underrepresented in research. In our ongoing research, we emphasize the importance of including people with varying educational attainment, digital proficiency, and digital acceptability as these factors seem to influence people’s intention to use and satisfaction with digital tools for dementia prevention (see [98]).

The results from this scoping review also highlight the importance of oversampling culturally diverse and underrepresented populations during recruitment to result in more diverse study samples. Other recommendations regarding inclusivity include engaging in dedicated usability testing sessions and using inclusive design guidelines (eg, Web Content Accessibility Guidelines and the upcoming European Accessibility Act requirements) [99-102], as well as engaging in co-design, in which a representative group of (potential) end users directly cooperate in a creative way with developers and researchers throughout the entire (tool) design process [103]. Specific co-design methods exist for working together with people with cognitive complaints and/or dementia, such as carefully introducing new topics, using probes to involve people in abstract thinking, guiding caregivers toward a supporting role, and personalizing the length of an interview or session [104]. Tools are developed that address a need of end users and are acceptable, usable, and inclusive for end users. Finally, from an accessibility perspective, it is recommended to offer digital tools as part of a larger intervention containing (nondigital) alternatives.

Among the strengths of this scoping review is the thorough process of article inclusion, conducted by 3 reviewers, ensuring researcher triangulation. As for limitations, we only included Dutch and English articles. We also may have missed some articles, since we did not explicitly search for digital tools on cognitive training or on specific dementia-related risk factors, such as hearing loss prevention or high blood pressure. In our search query, we also searched for “diagnostic tools” and “screening” instead of “self-administered screening tools,” which may have resulted in the exclusion of relevant articles since they were not identified through the search query. Moreover, we did not perform snowballing. We included tools on dementia prevention, daily life support, AD/dementia diagnosis and risk assessment, and decision support. We may have missed some tools that improve brain health but are not explicitly linked to brain health in the context of AD or dementia. For instance, digital tools targeting lifestyle in general, word games, or exergames may also help to reduce dementia risk, but these did not appear in our search results if they were not explicitly linked to AD or dementia. Likewise, certain assistive technologies and tools targeting people with dementia and their informal caregivers might also be useful in earlier disease stages, such as lifestyle-monitoring systems [105]. We did not include these articles since we focused on cognitively unimpaired people or people with SCD or MCI in the context of AD or dementia.

A variety of future (research) directions were reported in the included papers, mainly focusing on improvements in the early development stages, such as the need for (further) efficacy or feasibility testing and the need for (further) prototyping. Many included papers also mentioned that they would focus on effectiveness testing in future research in a real-world setting and/or in large-scale clinical randomized controlled trials. In addition, implementation research is considered crucial to translate digital tools into practice [106,107]. Previous research shows that successful implementation of digital tools in the AD context proves to be difficult [19,22,23]. Some factors hindering implementation might relate to common characteristics of the target group of people with or at high risk for AD/dementia, such as older age, cognitive impairment, impaired hearing, problems with vision, and comorbidities [3,25,26]. Personal factors of the person at risk, such as attitudes and beliefs, might also influence people’s intention to use digital tools for health (care) purposes [108-110]. Moreover, systematic implementation issues on an interpersonal level (eg, social influences of others on digital tool use) and systemic level (eg, facilitating health care policies) should be addressed [111]. If these factors are not taken into account, digital tools may only benefit those who are already better off and structural inequities may reinforce themselves, enlarging health disparities [112]. Thus, an overarching examination of factors associated with implementation success of digital tools in the context of cognitively unimpaired people or people with SCD or MCI in the context of AD or dementia is needed to allow for successful implementation in practice for all. This scoping review captured tools described in the academic literature. It must be noted that some digital tools included in our study as early stage of maturity and evaluation are currently already in mid-mature phases among other populations. However, it may be that their follow-up development has not been reported in academic literature or that follow-up papers assessing the mid-maturity stage of maturity and evaluation did not show up in our search results (eg, CogniFit [113]). It is also likely that there is a broad variety of digital tools on the market that are not reported on in academic literature. This implies that even though we found that the majority of digital tools as reported on in academic literature are considered to be in early maturity stages, the actual stage of maturity and evaluation may be different than reported on in academic literature. Future research may combine these findings from academic literature with digital tools as reported on in gray literature for a more comprehensive overview, focusing on performance accuracy, stage of evaluation, and stage of maturity. Future research could create an overview of digital tools developed in the context of AD or dementia for cognitively unimpaired people or people with SCD or MCI reported on outside of academic literature.

This scoping review shows that the majority of digital tools for cognitively unimpaired people or people with SCD or MCI reported on in academic AD and dementia literature comprise primary and secondary dementia prevention–related tools (64%). Tools that support people who do not have AD or dementia in activities of daily living are also common (20%). Digital tools are tested among people who vary in cognitive abilities, but reported studies were skewed toward highly educated women aged 60 years and older. Few papers reported on the ethnicity and digital literacy of their study population. Almost 80% of digital tools as reported on in academic literature are in (early) development, comprising early stages of maturity (eg, preprototyping, prototyping, or pilot testing) and early stages of evaluation (eg, efficacy testing, usability testing, and feasibility testing). The scientific evidence gathered for digital tools for people without an AD or dementia diagnosis is limited. Future (research) directions comprise the need for ensuring diversity, equity, and inclusion, as well as the need for implementation research. We therefore encourage further research and policy development to support the transition of promising tools from the early development stage to later stages, thereby allowing for increased use among people without an AD or dementia diagnosis.