This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Aging, is properly cited. The complete bibliographic information, a link to the original publication on http://aging.jmir.org, as well as this copyright and license information must be included.
Through the increasingly aging population, the health care system is confronted with various challenges such as expanding health care costs. To manage these challenges, mobile apps may represent a cost-effective and low-threshold approach to support older adults.
This systematic review aimed to evaluate the quality, characteristics, as well as privacy and security measures of mobile apps for older adults in the European commercial app stores.
In the European Google Play and App Store, a web crawler systematically searched for mobile apps for older adults. The identified mobile apps were evaluated by two independent reviewers using the German version of the Mobile Application Rating Scale. A correlation between the user star rating and overall rating was calculated. An exploratory regression analysis was conducted to determine whether the obligation to pay fees predicted overall quality.
In total, 83 of 1217 identified mobile apps were included in the analysis. Generally, the mobile apps for older adults were of moderate quality (mean 3.22 [SD 0.68]). Four mobile apps (5%) were evidence-based; 49% (41/83) had no security measures. The user star rating correlated significantly positively with the overall rating (
There is an extensive quality range within mobile apps for older adults, indicating deficits in terms of information quality, data protection, and security precautions, as well as a lack of evidence-based approaches. Central databases are needed to identify high-quality mobile apps.
Demographic change continues worldwide [
Mobile and internet technologies such as mobile apps offer possible approaches to increase the empowerment of older adults, support social activities, prevent cognitive and physical decline, decrease loneliness, and provide assistance in everyday activities [
Mobile apps may offer many advantages for older adults to complement traditional health care behavior, as they can be cost-effective if implemented on a large scale and used independently of time and location [
Nevertheless, uptake and acceptance of mobile apps by older adults are rather low [
Smartphones have become an integral part of everyday life, even for older adults [
There are many mobile apps available in the app stores [
Users can have problems identifying mobile apps that will effectively and safely support them in their health care [
To close this research gap, our study has systematically searched for mobile apps in the European app stores with a focus on older adults. Hence, their general characteristics, aims, methods, content, and quality were assessed using a multidimensional instrument, the German version of the Mobile Application Rating Scale (MARS-G) [
Privacy and security features
Quality criteria based on the MARS-G (engagement, functionality, aesthetics, information)
Correlation between the user star rating and the MARS-G overall rating
Prediction of overall quality due to the obligation to pay fees
The systematic review was based on the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA statement) according to Moher and colleagues [
A web crawler was used to systematically screen the European Google Play and App Store for eligible mobile apps with the search terms “old,” “dementia,” “memory,” “mnemonic,” “elderly,” “senior,” “maturity,” “retiree,” “seniority,” and “aided recall.” The search string to identify mobile apps for older adults resulted from findings of self-conducted focus groups with older adults, caretakers, and physicians followed by an expert discussion (EMM, LS, HB, MD, DD, and NW). The web crawler is a search engine that systematically searches the internet and country-specific app stores such as Google Play and the App Store for eligible mobile apps [
All identified mobile apps were listed in a central database, and the first results were screened by the reviewers (AP, DS, MD, MS, LS, DD, and NW). The screening was conducted via an Access (Microsoft Corp) file. Every mobile app was screened by two reviewers. Disputes were discussed with a supervisor (EMM). To be included in this review, mobile apps had to meet the following inclusion criteria: (1) designed for older adults or older adults, their caregivers, and relatives; (2) available and downloadable in the official Google Play or the App Store; (3) in German or English (in accordance with the reviewers’ language skills); (4) functional to enable an assessment (eg, no device problems); and (5) usable independently of other software (eg, software on smartwatches). Duplicates were automatically and manually excluded. Nonworking links were tried several times. The reviewers excluded mobile apps that did not meet the inclusion criteria according to the title, mobile app description, given images, or comments of mobile app users in the app stores in the first step.
On May 8 and 9, 2019, an additional manual search of mobile app recommendations in the app stores took place by a reviewer (AP) to identify further relevant mobile apps. This should ensure an up-to-date and comprehensive search for mobile apps. Additionally to the previous search terms, the following German and English search terms were used: “seniors,” “older adults,” “Alzheimers,” “memory games,” “retirement,” “pills,” “dementia,” “memory,” “senior health,” and “emergency call.” The search terms to identify mobile apps for older adults resulted from findings of self-conducted focus groups and were developed in an expert discussion (EMM, LS, HB, MD, DD, and NW). In addition to technical terms, relevant synonyms and alternatives used by end users were added to the extracted search terms [
For the MARS-G analysis, the mobile apps were downloaded and checked regarding the inclusion criteria and their functionality for the review (eg, no device problems). Technical problems were validated on at least two devices. The mobile apps were downloaded and installed either on an iPad mini (Apple Corp; model MK9N2FD/A; operating system 12.1), a MediaPad X2 (Huawei Device Co; model GEM-701L; operating system 5.0.1), or an iPhone 6 (Apple Corp; model A1586; operating system 12.2).
The quality assessment of the mobile apps was conducted by two independent reviewers (AP, DS, MD, MS, LS, DD, or NW) using the MARS-G [
The MARS-G evaluation tool is a reliable and valid scale for the quality assessment of mobile apps [
The classification page of the MARS-G was used to examine mobile app characteristics. It contains descriptive and technical information about the mobile app: (1) name, (2) platform, (3) content-related subcategory, (4) store link, (5) price, (6) user star rating, (7) aims, and (8) methods [
The assessment of privacy and security features based on MARS-G is on a descriptive level (eg, availability of privacy policy, imprint). All features were assessed based on downloaded mobile apps, and only information that was disclosed within the mobile app or its description in the app stores was investigated.
The categorization of mobile apps for older adults according to Cunha and colleagues [
Mobile app categories for older adults with exemplary topics according to Cunha et al [
Categories | Exemplary topics |
Diagnostic | Cognitive impairments, physical and mental illnesses |
History | Monitoring of vital parameters such as blood pressure, and organization of daily activities |
Improve | Relaxation, speech-to-text, text-to-speech, risk assessment, magnifying glass, medication recognition, pictogram-to-speech, communication portals, and social networks |
Informative | Healthy living, education, and psychoeducation about mental and physical illnesses |
Interface | Mobile apps for conversion to a user-friendly interface |
Measurement | Physical activity, pedometer, and GPS tracking |
Protection | Drug reminder, help requests, and localization |
Simulation | Simulation of diseases, impairments, or appearance |
Trainer | Memory, relaxation, logical thinking, fitness, and cognitive speed |
Tutorial | Accident rehabilitation, sign language, improvement of self-esteem, and improvement of communication |
The multidimensional quality rating of the MARS-G includes 19 items on 4 different subdimensions, which are evaluated on a 5-point Likert scale (1=inadequate, 2=poor, 3=acceptable, 4=good, and 5=excellent): (1) engagement—5 items (entertainment, interest, individual adaptability, interactivity, target group); (2) functionality—4 items (performance, usability, navigation, motor and gestural design); (3) aesthetics—3 items (layout, graphics, visual appeal); and (4) information—7 items (accuracy of app description, goals, quality of information, quantity of information, quality of visual information, credibility, evidence base) [
For the evaluation of the overall rating and quality, the total score was calculated from the 4 subdimensions [
Item 19 on the information subdimension was used to assess whether empirical studies were available for a mobile app. This item was investigated by searching the mobile app name in Google Scholar, PubMed, Google, and the developers’ or providers’ websites for existing efficacy and effectiveness studies [
Bivariate correlations between the user star rating and the MARS-G ratings were calculated. Also, bivariate correlations between the user star rating and the number of security and privacy measures were determined. The user star ratings were extracted from the app stores. The user star rating from Google Play and the App Store can be assigned on a scale of 1 to 5 stars and is displayed to mobile app seekers in the app stores as a cumulative average of individual ratings [
To examine whether the obligation to pay fees is a predictor of overall quality, an exploratory regression analysis was conducted in which the predictor was dummy coded (1=obligation to pay fees, 0=no obligation to pay fees). Mobile apps that required an initial payment for use were defined as “obligation to pay fees.” Mobile apps that were not priced at the time of purchase or had a free basic version were defined as “no obligation to pay fees” [
A
The web crawler identified 1154 mobile apps, of which 11.01% (127/1154) were found to be eligible by initial screening (
Flowchart of the mobile app selection process.
Of the mobile apps, 64% (53/83) were from Google Play and 36% (30/83) were from the App Store. There were no significant mean differences in the MARS-G overall rating between mobile apps from different stores (t81=1.399,
On average, the mobile apps for older adults had 3.36 (SD 1.79) aims, with a maximum of one mobile app having 8 aims. Most common aims were improvement of well-being (54/83, 65%), entertainment (39/83, 47%), reduction of stress (37/83, 45%), and reduction of anxiety (29/83, 35%). Aims classified under other aims (23/83, 28%) included, for example, disease education (2/83, 2%) and screening for Alzheimer disease (3/83, 4%).
Frequency of objectives of mobile apps for older adults. Multiple naming of objectives for one mobile app was possible. Data are given for n=83 mobile apps.
On average, the mobile apps used 2.88 (SD 1.81) methods. The number varied from 1 to 9 methods. The most common methods were monitoring and tracking (26/83, 31%), data collection and measurement, feedback, and gamification (each 25/83, 30%) as well as information and education and tips and advice (each 23/83, 28%). Some mobile apps included memory, reminder, amplifier (16/83, 19%), strategies, skills, training (12/83, 14%) and resource orientation (11/83, 13%). Only a few mobile apps included physical exercises (7/83, 8%), mindfulness and gratefulness, and tailored interventions (each 5/83, 6%), acceptance, pursuing own goals and relaxation exercises (each 3/83, 4%), and traditional medicine (2/83, 2%) or alternative medical intervention elements and exposition (each 1/83, 1%). Methods classified under other methods (23/83, 28%) included, for example, personalization (7/83, 8%), social networking features (4/83, 5%), and emergency button and contacts (1/83, 1%). None of the mobile apps included serious games, breathing exercises, hypnotherapy or EMDR.
Frequency of methods used in mobile apps for older adults. Multiple naming of different methods in one mobile app was possible. Data are given for n=83 mobile apps.
The average number of security and privacy measures was 2.07 (SD 2.76). Of the included mobile apps, 49% (41/83) had no data protection precautions. Most frequently (30/83, 36%), a contact, contact person, or imprint was given. Only in 7% (6/83) emergency functions were available; 5% (4/83) provided data transmission security.
Privacy and security measures found in mobile apps.
Data protection precaution | Valuea, n (%) |
Allows password use | 22 (27) |
Requires a log-in | 20 (24) |
Has a privacy statement | 28 (34) |
Requires active confirmation of a consent form | 14 (17) |
Information on dealing with the data | 14 (17) |
Notes on financing/conflict of interest | 14 (17) |
Contact/contact person/imprint | 30 (36) |
Data transmission security | 4 (5) |
Emergency functions available | 6 (7) |
Security strategies for mobile phone loss | 20 (24) |
Other security strategies | 0 (0) |
aMultiple naming of different data protection precautions for one mobile app are possible.
According to the categorization of Cunha and colleagues [
The overall rating showed an excellent level of interrater reliability (2-way mixed ICC .97, 95% CI .97-.98). According to Portney and Watkins [
Significant positive bivariate correlations were found between overall rating and subdimensions (
Graphical representation of the distribution of the Mobile Application Rating Scale (German version) overall rating, and the four subdimensions. The median, the interquartile distance as well as the range and outliners were given (n=83 mobile apps).
Correlations between the mean values of the four MARS-G subdimensions, overall rating and user star rating.
Characteristics | MARS-Ga | ||||||||||
Engagement | Functionality | Aesthetics | Information | Overall rating | |||||||
|
|||||||||||
Engagement | —b | — | — | — | — | — | — | — | — | — | |
Functionality | .52 | <.001 | — | — | — | — | — | — | — | — | |
Aesthetics | .62 | <.001 | .54 | <.001 | — | — | — | — | — | — | |
Information | .55 | <.001 | .33 | .002 | .58 | <.001 | — | — | — | — | |
Overall rating | .83 | <.001 | .68 | <.001 | .85 | <.001 | .83 | <.001 | — | — | |
User star ratingc | .27 | .03 | .11 | .38 | .19 | .13 | .32 | .01 | .30 | .01 |
aMARS-G: German version of the Mobile Application Rating Scale.
bNot applicable.
cCorrelations were calculated with 69 mobile apps since the user star rating was missing for 14 apps.
Four (5%) mobile apps were evidence-based. For Lumosity [
The user star rating and overall rating correlated significantly positively with
There were no bivariate correlations between the overall rating or the four subdimensions and the obligation to pay fees (
In this study, we systematically examined the quality of 83 mobile apps for older adults in the European commercial app stores using a reliable and valid rating instrument. Furthermore, we assessed general characteristics, aims, methods, content, and privacy and security measures of the mobile apps for older adults. In general, the mobile apps were of moderate quality with a wide range of quality ratings. This result is in line with findings from other systematic mobile app reviews using the MARS [
The generally low information quality with a wide range is also in line with the results of other systematic reviews [
Moreover, users are confronted with data and security issues, as 49% of the mobile apps contained no security or data protection measures, and those that do exist lack clarity. The literature implies that concerns about the lack of data protection measures represent an essential usage barrier for older adults [
Furthermore, the efficacy and effectiveness of mobile apps for older adults are poorly examined [
Top-ranked mobile apps often have a high user star rating, which is discussed as an indicator of mobile app quality [
According to our results, the obligation to pay fees did not predict mobile app quality. In previous studies, it was partly implied that paid mobile apps are more credible, trustworthy, and recommendable and are more likely to promote users’ health and well-being [
Most mobile apps could be assigned to the trainer category. Training mobile apps such as fitness and cognitive exercises for the prevention of neurodegenerative diseases as well as social media mobile apps are mostly used by older adults [
One strength of this study is the use of traditional systematic review methodology, such as systematic search, independent screening, and quality evaluation of the included mobile apps on a reliable scale. The multidimensional MARS-G enabled an objective, reliable, and valid rating [
However, due to the high frequency of new and further developments as well as the continuous technological progress of the mobile app market [
Another limitation is the country-specific search for mobile apps in the German and British app stores. Different mobile apps are offered in various countries since the selection of countries in which a mobile app is available is determined by the developers [
Furthermore, mobile apps were not tested for a longer time, as in days or weeks. Therefore, some aspects of the mobile apps may not have been detected, and some errors may have remained hidden.
Additionally, we assessed privacy and security measures on a descriptive level, and the included data is based on information within the mobile apps and description in the app stores. Future studies should conduct an in-depth analysis of privacy and security measures in mobile apps for older adults (eg, analyzing if they transmit data using services provided by Facebook or Google) [
Since the user star rating is invalid to assess mobile app quality, publicly available expert mobile app ratings could help older adults as well as their relatives, caregivers, and health care professionals (eg, physicians) to select a high-quality mobile app. Publicly available MARS ratings by experts on a wide range of health topics on databases like Psyberguide and mHAD [
In the future, efficacy and effectiveness studies should be implemented for mobile apps. At present, there is a lack of high-quality studies that prove the long-term benefit, effectiveness, and safety of mobile app use for older adults [
Promotion measures as reimbursement of costs of mobile apps with proven effectiveness through health care providers and targeted information campaigns on existing high-quality mobile apps for older adults and their relatives could help them to integrate high-quality mobile apps into their daily lives [
The potential inherent in mobile apps to support a healthy, active, and safe life for older adults has not yet been sufficiently explored. The study was able to indicate that currently available mobile apps for older adults are on average of moderate overall quality. In particular, deficiencies could be found in information quality, evidence-based approach, data protection, and security measures. However, some mobile apps were of high quality, were based on evidence, and had sufficient data protection, and therefore, could provide suitable support. The user star rating and the obligation to pay fees did not provide valid orientation aids. Annually conducted reviews and publicly available expert mobile app ratings could help older adults and their relatives as well as caregivers to select a high-quality mobile app.
Preferred Reporting Items for Systematic Reviews and Meta-analyses 2009 checklist.
Included mobile apps with name, store, developer, version, price, user star rating, Mobile Application Rating Scale, German version (MARS-G), subdimensions and overall rating sorted by MARS-G overall rating.
intraclass correlation
Mobile Application Rating Scale, German version
Preferred Reporting Items for Systematic Reviews and Meta-analyses
The authors would like to the thank Jiaxi Lin, Rüdiger Pryss, Robin Kraft, Pascal Damasch, and Philipp Dörzenbach for their support in the development of the search engine and their support in the mHAD project. We also thank Milena Engelke for her assisting in the screening of the mobile apps.
EMM, YT, LS, and HB developed the study design. AP, DS, MD, LS, MS, NW, and DD collected the data. AP, EMM, and YT ran the statistical evaluations. AP and EMM wrote the first draft of the article. All authors contributed to the current version of the article and have approved the final paper.
EMM, YT, LS, and HB developed, and run the German Mobile Health App Database project (MHAD). The MHAD is a self-funded project at Ulm University with no commercial interests. LS, HB and EMM received payments for talks and workshops in the context of e-mental-health. LS reported receiving personal fees from Psychotherapy Training Institutes and Clinics outside the submitted work. This does not alter our adherence to JMIR policies on sharing data and materials. All other authors declare no conflicts of interest.