Flow diagrams of identification and retention of apps and websites are presented in Figures 1 and 2. Overall, 28 apps were considered for inclusion. Reasons for exclusion following the app screening process are listed in Additional File 1. We identified 25 apps from the searches in Google Play and the Apple App Store. Of these, 12 fulfilled inclusion criteria (LifeCurve; StopFalls; Otago Exercise Programme; Spiro100; Nymbl Balance Training; Moves4Me; Stannah Balance; Wysefit; Keep On Keep Up; Exercise Plan for Seniors; Hearty Seniors; and Senior Beginner Workout).. We identified three further apps from our database search for apps in development, one intervention study (ActiveLifestyle) (24) and two study protocols (StandingTall and eLiFE) (25,26) involving app-based interventions. Out of these 15 apps, 11 were available for download in UK app stores. We contacted developers to gain access to the remaining four apps, and were granted access to two of them (StandingTall; Nymbl Balance Training).. We were unable to gain access to the ActiveLifestyle app and the eLiFE app is not planned to be made publicly available following feasibility RCT testing (27). Thus, these two apps were excluded from the review. A total of 13 apps were evaluated.
We identified 24 relevant websites for inclusion in the review. As there were no available published studies evaluating the effectiveness of the included apps or websites, we were unable to perform process evaluations, risk of bias assessments, or meta-analyses. Full narrative summaries of all apps and websites detailing the aim, target population, description, platform, and evidence evaluation are provided in Additional File 1.
Characteristics of apps and websites
Ten of the 13 apps were commercially developed and three were developed by universities. Six were available on both Android and iOS platforms, four were iOS only, and three were Android only. Of the 12 apps currently available for download, eight were free to use, and four involved a subscription service. Most of the 24 websites originated from the USA (n = 11) or the UK (n = 6), with the remainder from Canada (n = 4), Australia (n = 1), Singapore (n = 1), and Europe (n = 1). Thirteen were provided by commercial organisations, six were provided by government, three were not-for profit, and two were academic. Detailed characteristics are provided in Additional File 1.
Table 1 provides an evaluation summary of the 13 apps. On average, total MARS scores differed between raters by 0.57 points (SD = 0.38; difference range = 0.02–1.30). Inter-rater agreement between raters on BCT scores was strong (Kalpha = 0.89; 95% CI: 0.78, 0.97).
Evidence for effectiveness
In seven apps, most of the promoted exercises featured in evidence-based exercise programmes, such as Otago and FaME (5). Five apps included at least some evidence-based exercises. One app (Senior Beginner Workout) did not contain any evidence-based exercises. Only one of the apps was explicitly based on an existing evidence-based exercise programme: the Otago Exercise Programme app (Otago). None of the apps have direct evidence supporting their effectiveness, and only one has recently been assessed using an RCT design, with results yet to be published (StandingTall)..
Use of BCTs
The mean number of individual BCTs across all apps was five (SD = 2.30; range: 1.5–8.5). Figure 3 (panel A) illustrates the prevalence of BCTs across apps. Frequently included BCTs belonged to the following categories of the 93-item BCT taxonomy: shaping knowledge (12 out of 13 apps); and comparison of behaviour (11 out of 13 apps). Video-based apps (Wysefit and Spiro100) contained the fewest BCTs. Apps containing more than five individual BCTs (the top 50% in terms of the number of BCTs applied) included StandingTall, Move4Me, LifeCurve, Otago Exercise Programme, Nymbl Balance, and Keep On Keep Up.
MARS Quality Ratings
The mean total MARS score was 3.56 out of 5 (SD = 0.32). Scores ranged from 2.78 (StopFalls) to 4.09 (Nymbl Balance).. Most of the apps (8 out of 13) received a score of ‘good’, but two of these (LifeCurve and Stannah Balance) could be interpreted as ‘acceptable-good’ with scores falling almost exactly in between the two categories. Five of the apps were rated as ‘acceptable’ (Table 1)..
Table 1 provides an evaluation summary for the 24 websites. Kalpha agreement was acceptable for credibility (0.77; 95% CI: 0.60, 0.89), and strong for senior friendliness (0.88; 95% CI: 0.83, 0.92) and BCT ratings (0.85; 95% CI: 0.73–0.95).
Evidence for Effectiveness
In 21 of the websites, most of the featured exercises also appear in evidence-based programmes. Two of the websites included at least some evidence-based exercises. Only one website (closingthegap.ca) did not appear to contain any evidence-based exercises. As expected, none of the websites have been subject to randomised trial design, or other evaluations of effectiveness. Three of the websites (profound.eu.com, betterhealthwhileaging.net, and caringseniorservice.com) contained video demonstrations of the Otago Exercise Programme, for which there is a strong evidence-base for effectiveness.
Use of BCTs
The mean number of individual BCTs across all websites was 3.88 (SD = 1.44; range: 2–7). Figure 3 (panel B) illustrates the prevalence of BCTs across websites. Frequently included BCTs belonged to the following categories of the 93-item BCT taxonomy: shaping knowledge (23 out of 24 websites); comparison of behaviour (19 out of 24); and natural consequences (19 out of 24). Websites containing more than 4.5 individual BCTs (the top 50% in terms of number of BCTs applied) included buffalorehab.com, csp.org.uk, eldergym.com, fallsassistant.org.uk, go4life.nia.nih.gov, healthhub.sg, nhs.uk/live-well, and profound.eu.com.
Quality of Websites
Overall, the websites scored an average of 71.74%, indicating good quality. Eight were considered excellent, 12 were considered good, four were considered fair, and none of the websites were considered poor. The mean credibility score across all websites was 4.60 out of a possible 7 (65.77%) (SD = 1.15; range = 2–7), indicating that compliance with the HONCode standards for health-related websites was good. Only two websites stated that they were HONCode compliant, with mayoclinic.org meeting all seven standards, and healthline.com meeting six. Three of the websites (caringseniorservice.com, eldergym.com, and unitypoint.org),, scored poorly on credibility (score <50%) The mean senior friendliness score across all websites was 31.08 out of a possible 40 (77.71%) (SD = 4.13; range = 23–38), indicating that overall website senior friendliness was excellent. None of the websites scored lower than 20 (<50%) on senior friendliness.