Corresponding Author: Hock Hai Teo, Department of Information Systems and Analytics, School of Computing, National University of Singapore, 13 Computing Drive, 117417 Singapore, Singapore (gs.ude.sun.pmoc@hhoet)
Received 2021 Mar 25; Revised 2021 Jun 1; Accepted 2021 Jul 5.Copyright © The Author(s) 2021. Published by Oxford University Press on behalf of the American Medical Informatics Association. All rights reserved. For permissions, please email: journals.permissions@oup.com
This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
The data underlying this article are available in the article and in its online supplementary material.
Mobile-based interventions have the potential to promote healthy aging among older adults. However, the adoption and use of mobile health applications are often low due to inappropriate designs. The aim of this systematic review is to identify, synthesize, and report interface and persuasive feature design recommendations of mobile health applications for elderly users to facilitate adoption and improve health-related outcomes.
We searched PubMed, Embase, PsycINFO, CINAHL, and Scopus databases to identify studies that discussed and evaluated elderly-friendly interface and persuasive feature designs of mobile health applications using an elderly cohort.
We included 74 studies in our analysis. Our analysis revealed a total of 9 elderly-friendly interface design recommendations: 3 recommendations were targeted at perceptual capabilities of elderly users, 2 at motor coordination problems, and 4 at cognitive and memory deterioration. We also compiled and reported 5 categories of persuasive features: reminders, social features, game elements, personalized interventions, and health education.
Only 5 studies included design elements that were based on theories. Moreover, the majority of the included studies evaluated the application as a whole without examining end-user perceptions and the effectiveness of each single design feature. Finally, most studies had methodological limitations, and better research designs are needed to quantify the effectiveness of the application designs rigorously.
This review synthesizes elderly-friendly interface and persuasive feature design recommendations for mobile health applications from the existing literature and provides recommendations for future research in this area and guidelines for designers.
Keywords: mobile application, user-centered design, interface design, persuasive features, healthy aging
The aging population is increasing rapidly all over the world. The proportion of people aged 65 or above is expected to reach 12% and 23% worldwide by 2030 and 2100, respectively, 1 which would put enormous pressure on health and social service systems. 2 To alleviate this pressure, healthcare providers are considering designing mobile-based interventions to promote healthy lifestyles, support disease prevention and management, and improve access to health services. 3 The global mobile health market was estimated at $35.1 billion in 2020 and is expected to increase to $145.7 billion by 2027. 4 However, acceptance and continued use of mobile health (mHealth) applications are often low among the elderly, significantly plaguing their utility. 5 One important reason is that most mHealth applications available in the market do not carefully consider the unique needs, preferences, and capabilities of elderly users, resulting in low usability. 5 , 6
Compared with younger populations, older adults often face additional challenges in using mHealth applications due to limited perceptual, motor, and cognitive capabilities. 7 , 8 In particular, the aging process will negatively affect visual and hearing abilities, hand-motor functions, and information processing capacity. 9–12 Moreover, elderlies tend to experience reduced motivational orientation, 6 further hampering the adoption and sustainable use of mHealth applications. Indeed, older adults will not adopt a technology if they do not perceive the benefits of using it. 13
It is therefore essential to provide sufficient support for older adults when designing mHealth applications for them. Several review articles have focused on evaluating the benefits of mHealth applications for chronic disease management and healthy lifestyle promotion in older adults, 14–17 and some have discussed practical challenges in application design. 15 , 16 , 18 However, none of these reviews have compiled and reported mHealth application design recommendations for older adults. On the other hand, some articles have analyzed aging barriers for mHealth application usability and proposed several interface guidelines. 6 , 19–21 However, these studies neither summarized the current evidence of the effectiveness of interface design nor investigated persuasive feature design. This review aims to address these gaps by identifying, synthesizing, and reporting recommended mHealth application designs that facilitate application adoption and promote healthy aging. Specifically, we seek to identify elderly-friendly interface designs that increase mHealth application acceptability and usability and persuasive features that increase adherence to the delivered mobile interventions.
We followed the guideline of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement. 22 With the help of a librarian, 2 reviewers (JY and NL) searched PubMed, Embase, PsycINFO, CINAHL, and Scopus databases (Jan 01, 2010, to Apr 30, 2021) to identify articles with a focus on the design of mHealth applications for older adults. We chose 2010 as the start period because mHealth applications for older adults have only grown exponentially in the last decade. 23
We searched the title and abstract of the articles using 3 groups of keywords that were combined using an AND operator: (a) mHealth application-related terms, (b) application design-related terms, and (c) older adult-related terms. The detailed search strategy can be seen in Supplementary Appendix A.
We structured our inclusion and exclusion criteria based on the PICOS (Population, Interventions, Comparisons, Outcomes, Study) framework (see Table 1 ). Two members of the research team (JY, NL) independently screened the titles and abstracts of the retrieved articles and assessed the full texts of all potentially eligible studies against the inclusion criteria. Interrater reliability of the screening based on the title and abstract resulted in high agreement (Cohen’s kappa = 0.74), while the interrater reliability of the screening based on the full text resulted in moderate agreement (Cohen’s kappa = 0.67). Disagreements were resolved through group discussion with a third investigator (SST) until 100% agreement was achieved.
Inclusion and exclusion criteria based on the PICOS model
An older cohort.
An older person can be defined as one whose chronological age is 60 years or above, or 65 years or above. 24 In this review, we used 60 years as a more conservative threshold.
Any mHealth applications with any mobile devices, such as mobile phones and tablets.
We excluded health-related interventions that only used mobile devices (eg, short messages), wearables, and smart home devices without mobile apps.
Two team members (JY, NL) extracted and synthesized information using a piloted data extraction form. Specifically, they individually extracted the following items: study characteristics (ie, participant characteristics, country, study design, mHealth application descriptions, and health domain), application design features (ie, user interfaces, persuasive features, theory), mobile interventions, comparison or control groups, evaluation outcomes (ie, end-user perceptions, health-related outcomes), and key findings. Disagreements were resolved through group discussion with another 2 investigators (HHT and KYN) until 100% agreement was achieved.
We identified a total of 6337 articles using the initial search strategy ( Figure 1 ). We removed 1569 duplicate records using Endnote’s automatic duplication finder, leaving 4768 unique articles. After screening titles and abstracts, 175 articles were eligible for full-text review. We excluded 104 articles because the studies did not focus on mHealth applications or did not evaluate the applications with an elderly cohort, leaving 71 included articles. We included 3 studies identified through searching the reference lists of included studies. Thus, 74 articles were kept in the final analysis.
Flowchart of the literature search and selection processes.
Table 2 presents details of study characteristics. All of the included studies enrolled participants aged 60 years or more. Four studies did not report participants’ gender. 38 studies did not reveal participants’ educational background, and 39 studies did not reveal their experience with mobile devices and digital technologies. Sample sizes ranged from 1 to 503, in which 46 studies had a sample size smaller than 30.
Mean (M): not reported.
(Pilot study: usability study, feasibility study)
Usability: 65+, M = 73.2;
Feasibility: 65+, M = 76
Usability: 9 years of education;
Feasibility: 10.53 years of education
Usability: M = 64.9;
(Pre-post intervention study)
(Pre-post intervention study)
38 had Android smartphones; 2 had feature phones on which mobile apps were not available.
(Pilot study, within-subject, cross-over experiment)
(Usability study, interviews)
Mixed methods research
Mixed methods research
Mixed methods research
Mixed methods research
(Pre-post between-subjects experiment)
(Pre-post intervention study)
(Pre-post intervention study)
(Usability test, mixed design experiment)
Mixed methods research
Mixed methods research
Mixed methods research
(Pilot study, focus group)
Mixed methods research
Of the 74 included studies, a variety of quantitative (N = 47), qualitative (N = 9), and mixed-methods designs (N = 18) were used to evaluate mHealth applications (see Table 1 for details). We further assessed the quality of the 12 RCT studies based on the Cochrane Risk of Bias Assessment Tool (see Supplementary Appendix B for details). 99 Nine studies reported random sequence generation. 46 , 48 , 52 , 54 , 59 , 69 , 71 , 74 , 90 Only 3 studies explicitly stated that the allocation was concealed. 54 , 59 , 69 Only 2 studies were double-blinded, 54 , 74 and 1 study was blinded to neither researchers nor participants. 71 Using 80% completed participants as a threshold, 7 studies had a low risk of incomplete outcome data. 34 , 46 , 54 , 58 , 69 , 74 , 90 One study had a high risk of selective reporting bias. 38 Nine studies suffered from other sources of bias, such as a small sample size, 34 , 46 , 48 , 52 , 54 , 58 , 69 , 71 a biased sample, 54 , 59 , 71 , 90 and a lack of baseline data. 54
The mHealth applications described in the 74 included studies were mainly designed for disease prevention and self-management (N = 28), physical and cognitive function improvement (N = 27), social inclusion and well-being (N = 6), healthy dieting (N = 5), and health monitoring and reporting (N = 5). The remaining 3 articles covered more than 1 health domain, such as using a conversational agent to offer training modules for healthy dieting, physical health, and cognitive health. 97
The mHealth application designs were mainly assessed along 2 dimensions: end-user perceptions and health-related outcomes (see Supplementary Appendix C for details). Examples of end-user perceptions included usability, feasibility, acceptability, adherence, compliance, engagement, satisfaction, adoption and use, and other self-reported usage experience such as facilitators, barriers, and suggestions for future improvement. On the other hand, health-related outcomes included health-related behavior changes, physical and cognitive functions, psychosocial and emotional well-being, quality of life, and health knowledge.
Our analysis identified a total of 9 elderly-friendly interface design recommendations (see Box 1). We grouped these designs into 3 categories based on the addressed needs of elderly users: perceptual limitations, motor coordination problems, and cognitive and memory deterioration.
Vision impairment
Large font size Use Sans Serif family font style (eg, Arial, Verdana) Use bold font for key points Avoid special font styles (eg, italics, underline, all caps) Enable users to customize the font and text propertiesEnsure high contrast (eg, use dark texts on a light interface background, use differentiated button color from the interface
Limit number of colors Use basic and distinctive colors Use consistent background colors Use culturally appropriate colors (eg, use Arabic basic colors for Arabian users) Color coding of buttonsProvision of audio alternatives
Provide audio options Ensure loud audio volume Add vibrations during each auditory toneMotor coordination problems
Use simple touchscreen gestures (eg, swiping, tapping, dragging, dropping) Avoid complex touchscreen gestures (eg, scrolling, zooming) Use large and structured buttonsMinimize text entry
Support voice control Button-only interface (eg, use buttons and sliders to answer questions)Cognitive and memory deterioration
Simple and consistent layout
Provide a simple and consistent layout Ensure adequate white space between lines and buttons Adaptive layout based on the screen size Avoid large blocks of texts Avoid texts over images Use culturally appropriate layout (eg, use right to left reading for Arabian users)Simple and clear navigation
Flat navigation structure Place main functions in the home screen Reduce menu options Provide a clear return button and a static menu on every page Present information in the form of texts, pictures, and videos Use labeled buttons Organize related topics into groups Use age-appropriate and common languagesFifteen studies emphasized the importance of font design to improve application readability. 25 , 37–39 , 41–43 , 46 , 50 , 53 , 63 , 67 , 81 , 86 , 87 One often-mentioned design aspect is using large font size. 25 , 38 , 39 , 41–43 , 46 , 50 For example, Goumopoulos et al 81 recommended using a font size of 36 to 48 points on mobile phones. Other specific font design recommendations included using Sans Serif family font style, 67 , 81 using bold font for key points, 53 avoiding special font styles, 53 and enabling users to customize font and text properties. 53
Designers emphasized the importance of choosing an appropriate color to cater to older adults. One suggestion is to ensure high contrast to enhance the content readability for the aging population, such as using dark texts on a light interface background. 26 , 40 , 43 , 46 , 50 , 51 , 67 , 81 , 86 Another suggestion is to keep it simple, such as limiting the number of colors, 29 , 43 using basic and distinctive colors, 38 using consistent background colors, 81 and using culturally appropriate colors. 25 Finally, it is suggested to employ color-coding, such as assigning appropriate and unique colors for each button. 27 , 33 , 43 , 50 , 53 , 63 , 66 , 88 , 91 For example, Harte et al 63 used red for the “I have fallen” button and green for the “I am OK” button.
Applications should provide audio alternatives to ease the burden on vision-impaired elderly users. One suggestion is to read out texts to avoid misinterpretation and ensure accessibility to elderly users. 28 , 62 , 81 Designers can also use loud audio volume and add vibrations during each auditory tone for older adults with hearing difficulty. 55 , 67
Gestural interactions represent how elderly users interact with applications via a set of touchscreen gestures. One design guideline is to use simple gestures (eg, tapping) and avoid complex ones (eg, scrolling, zooming). 26 , 37 , 39 , 47 , 63 , 80 , 84 In one study, elderly users perceived tapping to be easier than swiping and hence preferred to use a tap-only interface than an interface that supported both tapping and swiping. 84 Another guideline is to use large and structured buttons to facilitate user interactions with applications. 38–40 , 43 , 61 , 63 , 67 , 83 , 86 , 87 , 90 However, no study has specified appropriate button sizes. Prior human-computer interaction (HCI) studies have suggested using rectangle buttons larger than 15.9 × 9.0 mm and square buttons between 16.51 mm and 19.05 mm square to facilitate user interactions. 100 , 101
Text input can be challenging for motor-impaired elderly users. Specific strategies to reduce text input include supporting voice control, 80 , 81 , 96 and using a button-only interface. 27 , 47 For example, Goumopoulos et al 81 supported voice commands to facilitate elderly users with motor skill problems to interact with the mHealth application.
Layout refers to the location of data elements on the application interface. It is crucial to provide a simple and consistent layout so that elderly users could easily process and comprehend the application content. 36 , 38 , 52 , 65 , 81 , 90 Specific layout design suggestions include ensuring adequate white space, 26 , 40 , 81 adapting layout based on the screen size , 89 avoiding large blocks of texts, 81 avoiding texts over images, 81 and using a culturally appropriate layout. 25
Navigation design describes how to guide users through an application via a set of predefined steps. Overall, designers suggested keeping the navigation structure simple and straightforward, such as using a flat navigation structure, 29 , 40 , 50 , 55 , 89 placing main functions in the home screen, 86 reducing menu options, 61 and providing a clear return button and a static menu on every page. 38
Multimedia presentation describes an application interface that presents information in the form of texts, images, and videos. This practice is strongly recommended by several studies because multiple sensory cues could effectively ease the information processing for older adults by providing sensory awareness. 35 , 46 , 50 , 55 , 58 , 66 , 70 , 85 , 87
Finally, it is important to ensure that the application content is easy to understand for elderly users. For example, designers suggested using labeled buttons so that users could easily comprehend the functionality of each button. 25 , 40 , 63 Other suggestions include organizing related topics into groups, 81 and using age-appropriate and common languages. 25 , 51 , 70 , 81
We have also identified 5 categories of persuasive features that provided motivational affordance for older adults to adhere to mobile-based interventions: reminders, social features, game elements, personalized interventions, and health education (see Figure 2 and Supplementary Appendix D).
Summary of persuasive features.
Twenty articles included reminders as part of their mobile interventions to remind elderly users to complete important daily tasks such as appointment and event scheduling, physical activities, nutrition and water intake, and medication intake. 27 , 29 , 34–36 , 38 , 41 , 45 , 46 , 48 , 52 , 57 , 69 , 78 , 80 , 81 , 88 , 91 , 98 , 102 Among these studies, 2 studies found that older adults appreciated reminders and reported that reminders provided a sense of modernity, independence, and control in task completion. 29 , 46 Another study examining a HIV prevention and self-care application found that medication reminder was one of the most frequently used features among elderly users. 48 In contrast, Farsjø and Moen 88 found that meal notifications of a nutrition application did not work as expected, and only one user paid attention to this feature. Also, Loh et al 41 revealed that reminders of a mobile-based geriatric assessment-driven intervention were not useful for 3 elderly patients who already had an appointment and medication tracking system. On the other hand, 4 studies examined the impacts of reminders, and the results were mixed. 36 , 41 , 45 , 98 Specifically, Mertens et al 45 found that reminders of the Medication Plan app significantly improved medication recording adherence. Hackett et al 36 showed that a reminder application improved the the completion of the Remember to Drink tasks for older adults with dementia and mild cognitive impairment. 36 In contrast, the other 2 studies suggested that reminders alone may be inadequate to promote behavioral changes. Loh et al 41 found that reminders were effective at improving medication adherence but not physical activities. Likewise, reminders increased medication adherence and fish and water intake, but not physical activities and social communication with friends and relatives. 98
The fulfillment of psychological and social needs is essential when designing mHealth applications. The most commonly used social feature is coaching or professional care services (N = 16). 34 , 38 , 41 , 42 , 54 , 57 , 58 , 61 , 62 , 69 , 72 , 78 , 81–84 Six out of the 16 studies reported positive end-user perceptions of this feature. 38 , 42 , 54 , 61 , 78 , 86 For example, Jongstra et al 38 found that elderly participants enjoyed discussing their lifestyle goals with coaches and appreciated their support. Moreover, Van Het Reve et al 78 examined the messages sent and received on a strength–balance training app and found that most interactions occurred between caregivers and participants rather than between participants.
Second, 8 studies incorporated peer support by enabling users to communicate, interact and network with peer users. 35 , 38 , 75 , 78 , 81–84 In particular, 1 study examining Senior App Suite showed that elderly users were satisfied with the social networking service and used it 3 times per week on average. 81 Further, 2 studies evaluated the effectiveness of InTouch, a mobile application that supports multi-modality messages (ie, audio, wave, picture, and video messages) to increase social connectedness. They found that messages brought positive communication and relationship changes among elderly users. 83 , 84 They further noted that audio messages were the easiest to use and were used most frequently, whereas wave messages were perceived as useless and were used the least.
Third, 6 studies introduced collaborative activities such as group-based online exercises, 54 , 75 , 78 collaborative digital games, 62 , 75 , 78 offline group discussions , 34 and offline social activities. 82 Only 1 study examined the effect of collaborative activities, and the results were not encouraging. 34 It designed mobile-based interventions for cognitive training that involved offline group discussions and social interactions. It showed that although elderly participants created social ties throughout the training intervention, they did not experience an improvement in cognitive and psychosocial outcomes.
Finally, 3 studies evaluated the impact of social features as a whole. 54 , 75 , 78 Specifically, Baez et al 54 examined the effects of tablet-based online group exercises and found that there were no significant differences in social well-being outcomes between the group training group and the individual training group. The other 2 studies compared two exercise training interventions delivered by ActiveLifestyle application (ie, an individual version and an social version) against a control group using training plans on paper sheets. 75 , 78 Van Het Reve et al 78 found that both individual and social groups had improved gait and physical performance compared with the control group. Silveira et al 75 found that social features were more effective at stimulating training compliance and healthy behavior changes. They further noted that 83% of participants felt motivated by external monitoring, followed by emotional support (75%) and collaborative games (58%).
Seven studies employed digital games to promote healthy aging. 26 , 31 , 43 , 51 , 70 , 74 , 76 The evaluation results were mixed. 43 , 51 , 70 , 74 For example, Shake et al 74 evaluated Bingocize, a serious game for exercising and health education, and found that it improved quality of life in older adults by improving physical and cognitive performance. In contrast, Li et al 70 designed and assessed 5 exergames among community-dwelling older adults. They found that although the exergames brought exercise enjoyment, they failed to improve self-efficacy, reduce loneliness, and improve life satisfaction.
We also analyzed specific game elements from the included articles, such as avatars (N = 9), 28 , 38 , 44 , 51 , 54 , 74 , 80 , 96 , 97 goal-setting (N = 11), 38 , 59 , 64 , 69 , 72 , 75–78 , 86 , 97 self-monitoring and tracking (N = 25), 27 , 30 , 32 , 38 , 40 , 41 , 46 , 48–50 , 52 , 54 , 56 , 57 , 69 , 70 , 75 , 76 , 79 , 86–88 , 91 , 95 , 97 performance feedback (N = 16), 40 , 43 , 53 , 54 , 61 , 62 , 65 , 66 , 69–71 , 86 , 87 , 92–94 reward and challenges (N = 6), 27 , 36 , 65 , 66 , 74 , 75 and peer competition (N = 2). 69 , 75 Regarding avatars, Chi et al 80 examined a human-operated conversational agent embodied in a dog or cat avatar. Overall, elderly participants felt comfortable with the agent, even though they felt worried that their social interactions might be adversely affected if their emotional attachment to the avatar became too strong. Bott et al 96 investigated the impact of a relational conversational agent in the form of an animated avatar and found that hospitalized patients who received the avatar had lower incidences of delirium, loneliness, depression, and falls than control patients. Finally, Stal et al 97 compared user preferences of two embodied conversational agent appearances for health assessments (ie, an older male agent and a young female agent) and found that elderly users did not perceive an added value of the agent. Regarding other game elements, Aure et al 86 found that the most regularly used feature of the Appetitus application was the self-monitoring dietary function. Silveira et al 75 reported that 67% of elderly participants of ActiveLifestyle application felt motivated by the goal-setting and self-monitoring features, followed by positive and negative reinforcement via rewards and praises (50%), and the peer performance comparison feature (42%). Two studies examining an attention training application revealed that elderly users appreciated working through challenges and perceived negative feedback as distracting and frustrating. 65 , 66
Thirteen studies provided personalized interventions to motivate elderly users to participate in health-promoting programs, such as physical and cognitive training, nutrition advice, therapy, and rehabilitation. 34 , 35 , 41 , 54 , 57–59 , 64 , 68 , 71 , 73 , 77 , 89 Of these, 10 studies investigated the effectiveness of the personalized interventions as a whole with promising results. 34 , 35 , 41 , 54 , 57–59 , 68 , 71 , 77 For example, Loh et al 41 provided a tailored mobile intervention to support elderly cancer patients and found that the intervention was effective at decreasing symptom severity and health care utilization. The remaining 3 studies did not examine the intervention's effectiveness. 64 , 73 , 89
27 studies provided education on various health topics to older adults. 25–27 , 32 , 35 , 37 , 38 , 41 , 42 , 48 , 52 , 59 , 67–69 , 72–75 , 82 , 87 , 91–94 , 96 , 97 Shake et al 74 examined the efficacy of a health education program that covered fall risks and osteoarthritis and found that elderly users’ health knowledge of the two topics increased over time. Hsieh et al 67 reported that elderly participants perceived a fall risk mHealth application to be useful in learning their risk of falling. Sun et al 52 assessed a diabetes self-management app and found that elderly users had improved self-management skills and knowledge of diabetes. In contrast, Algilani et al 91 investigated the feasibility of a mHealth application for early assessment and management of patient-reported concerns. The results revealed that some users provided positive opinions regarding the self-care advice, whereas others failed to appreciate its usefulness and availability. The remaining studies did not report the benefits of their education feature.
Our review provided specific and actionable mHealth application design recommendations to develop elderly-oriented interface and persuasive features. We classified the included studies into 3 categories based on the design aspects. Category I includes 43 studies that proposed interface design recommendations to cater to the perceptual, motor, and cognitive deterioration of older adults. Category II contains 66 studies that incorporated persuasive features to provide motivational affordance toward healthy behavior changes. In contrast, Category III includes 37 studies that discussed both elderly-friendly interface and persuasive feature designs. Category III studies are particularly important because appropriate interface designs could reduce users’ confusion and frustration and thus improve application acceptability and usability, while persuasive features could motivate elderly users to adherence to mobile interventions and improve health outcomes. Future work in this field should consider both application design aspects based on our derived recommendations to assist older adults in achieving healthy aging.
We found that 6 studies focused on psychosocial well-being, 80–85 and 10 studies focused on mental disease prevention and self-prevention. 26 , 29 , 31 , 33–36 , 43 , 44 , 51 However, it is important to note that older adults are vulnerable to social isolation and psychological distress. They may suffer from declined functional abilities, age-related diseases, and decreased socioeconomic status after retirement. 103–106 Worse still, a recent WHO report indicates that approximately 15% of older adults were suffering from mental health problems, 107 and most of them were reluctant to seek professional healthcare services. 108 Considering the potential of mobile-based interventions to improve the accessibility and efficacy of mental healthcare services, 109 we suggest future studies should devote more effort to the design and evaluation of mHealth applications for the mental well-being of the elderly.
Our analysis also revealed that only 5 studies (6.8%) based their design elements on theories, which included social cognitive theory, 38 behavior change wheel, 28 technology acceptance model, 25 theory of planned behavior, 64 and unified theory of acceptance and use of technology. 60 We recommend future research to propose theory-based application designs to provide a clearer understanding of the mechanism underlying each single design feature and, consequently, inform the theories of aging. 13 For example, using the complexity literature as a theoretical basis, researchers have found that although older adults perceived a high comprehensiveness recommendation agent (that elicited detailed product attribute preferences and provided more product recommendations) to be more complex, they also perceived it as more beneficial than its counterpart. 110 This contradicts the prevalent view that older adults should use simpler digital technologies.
Another important observation is that many studies proposed more than one design element and assessed the application as a whole. Consequently, our knowledge of end-user perceptions and the effectiveness of each single persuasive feature is relatively limited, particularly for peer support, collaborative activities, digital games, goal-setting, self-monitoring, performance feedback, rewards and challenges, peer comparison, and health education features ( Figure 2 ). Moreover, elderly users who provided positive ratings on application usability may perceive specific features as difficult to use or useless. 28 , 30 , 39 , 41 , 47 , 55 , 64 , 65 , 79 , 80 , 85 , 87 , 89 , 91 , 92 We thus recommend future research to investigate each design element in detail.
Finally, although 41 studies assessed the effects of mHealth applications on health-related outcomes, the strength of the evidence was limited due to poor research design. Specifically, approximately half of these studies (N = 22, 53.7%) did not have a comparison or control group. Moreover, 9 out of the 12 RCTs suffered from significant biases in study design. 34 , 38 , 46 , 48 , 52 , 58 , 59 , 71 , 90 Therefore, stronger research designs are needed to rigorously quantify the effectiveness of application designs.
We acknowledge that this review has several limitations. First, we restricted our scope to English-written, peer-reviewed journal articles. Our review could benefit from the inclusion of relevant conference publications and articles published in other languages. Second, we only included mHealth applications from the existing literature. Future research may consider conducting a more comprehensive evaluation of mHealth applications available for elderly users in the market.
This review identifies, synthesizes, and reports elderly-friendly interface and persuasive feature designs for mHealth applications from the existing literature. We derived 9 interface design recommendations that addresses the perceptual, motor, and cognitive deterioration of older adults. We also compiled 5 categories of persuasive features that may motivate older adults to achieve healthy aging. Overall, we provide specific and actionable suggestions for elderly-friendly mHealth application design and provide directions for future research in this field.
This research was supported by funds from the Singapore Ministry of Education Academic Research Fund Tier 1 (grant number: R-253-000-150-114) and The University of Sydney—National University of Singapore Partnership Collaboration Award 2019 Round.
NL and HHT designed the systematic review. JY, NL, and SST screened the title and abstracts of the included articles. JY, NL, NKY, and HHT extracted and synthesized the article information. All authors were involved in the data analysis, drafting, editing, and proofreading of the manuscript.
Supplementary material is available at Journal of the American Medical Informatics Association online.
