Age and Ageing 2022; 51: 1–10 © The Author(s) 2022. Published by Oxford University Press on behalf of the British Geriatrics https://doi.org/10.1093/ageing/afac243 Society. All rights reserved. For permissions, please email: journals.permissions@oup.com This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com SYSTEMATIC REVIEW Effectiveness of home-based exercise delivered by digital health in older adults: a systematic review and meta-analysis Lilian Solis-Navarro1,2,3 , Aina Gismero2 , Carles Fernández-Jané2 , Rodrigo Torres-Castro3,4,5 , Mireia Solá-Madurell6 , Clara Bergé7 , Laura Mónica Pérez8 , Joan Ars8,9 , Carme Martín-Borràs1,10 , Jordi Vilaró2 , Mercè Sitjà-Rabert2 1 Departmentof Physical Therapy, Facultat Ciències de la Salut Blanquerna, Universitat Ramon Llull, Barcelona, Spain 2 Global Research on Wellbeing (GRoW), Facultat Ciències de la Salut Blanquerna, Universitat Ramon Llull, Barcelona, Spain 3 Department of Physical Therapy, University of Chile, Santiago, Chile 4 Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain 5 Department of Pulmonary Medicine, Hospital Clínic, University of Barcelona, Barcelona, Spain 6 Associació de Dones no Estàndards, Barcelona, Spain 7 Department of Physiotherapy, School of Health Sciences, TecnoCampus-Pompeu Fabra University, Barcelona, Spain 8 RE-FiT Barcelona Research Group, Vall d’Hebron Institute of Research & Parc Sanitari Pere Virgili, Barcelona, Spain 9 Medicine Department, Universitat Autonoma de Barcelona, Barcelona, Spain 10 Facultat de Psicologia, Ciències de l’Educació i l’Esport (FPCEE), Universitat Ramon Llull, Barcelona, Spain Address correspondence to: Mercé Sitjà-Rabert, Global Research on Wellbeing (GRoW), Facultat Ciències de la Salut Blanquerna, Universitat Ramon Llull, Barcelona, Spain. Tel: +(34) 932533256. Email: mercesr@blanquerna.url.edu Abstract Background: regular physical exercise is essential to maintain or improve functional capacity in older adults. Multimorbidity, functional limitation, social barriers and currently, coronavirus disease of 2019, among others, have increased the need for home-based exercise (HBE) programmes and digital health interventions (DHI). Our objective was to evaluate the effectiveness of HBE programs delivered by DHI on physical function, health-related quality of life (HRQoL) improvement and falls reduction in older adults. Design: systematic review and meta-analysis. Participants: community-dwelling older adults over 65 years. Intervention: exercises at home through DHI. Outcomes measures: physical function, HRQoL and falls. Results: twenty-six studies have met the inclusion criteria, including 5,133 participants (range age 69.5 ± 4.0–83.0 ± 6.7). The HBE programmes delivered with DHI improve muscular strength (five times sit-to-stand test, −0.56 s, 95% confidence interval, CI −1.00 to −0.11; P = 0.01), functional capacity (Barthel index, 5.01 points, 95% CI 0.24–9.79; P = 0.04) and HRQoL (SMD 0.18; 95% CI 0.05–0.30; P = 0.004); and reduce events of falls (odds ratio, OR 0.77, 95% CI 0.64–0.93; P = 0.008). In addition, in the subgroup analysis, older adults with diseases improve mobility (SMD −0.23; 95% CI −0.45 to −0.01; P = 0.04), and balance (SMD 0.28; 95% CI 0.09–0.48; P = 0.004). Conclusion: the HBE programmes carried out by DHI improve physical function in terms of lower extremity strength and functional capacity. It also significantly reduces the number of falls and improves the HRQoL. In addition, in analysis of only older adults with diseases, it also improves the balance and mobility. Keywords: Home-based exercise, older adults, digital health intervention, physical function, systematic review, older people 1 L. Solis-Navarro et al. Key points • Regular physical exercise is essential to maintain or improve functional capacity in older adults. • No systematic review focuses on the effects of home-based exercise (HBE) delivered with digital health interventions (DHI) on physical function. • Home-based exercise (HBE) programmes carried out by using digital health interventions (DHI) improved physical function and reduced the falls in older adults. • In older adults with diseases, home-based exercise (HBE) using digital health interventions (DHI) improves the balance and mobility Introduction of 2019 pandemic has shown how important it is to make The older population is growing worldwide. Between 2015 good use of it and have clear information about its effective- and 2050, the proportion of the world’s population over ness, especially in a population where the digital divide is still 60 years will nearly double from 12 to 22% [1]. The chal- a reality [14]. lenge for this population is to achieve healthy ageing [1]. There are currently different studies on the effects of The clinical guidelines widely report recommendations for DHI and exercise in older people, although no systematic physical activity as a central axis of healthy ageing [2, 3]. Reg- review focuses on the effects on physical function. The main ular exercise leads to ageing actively and satisfactorily because objective of this systematic review is to evaluate the effects it is associated with physical, functional, psychological and of home-based exercise (HBE) programmes delivered with cognitive improvement [4]. Moreover, physical exercise is DHI on physical function in community-dwelling older the basis for treating many diseases, such as hypertension, adults. stroke, osteoporosis, metabolic syndrome, obesity, cancer and depression, among others [4–6]. Different aspects of physical functioning deteriorate with Methods increased age and physical inactivity [7]. Therefore, regular Protocol and registration physical exercise is essential for healthy ageing. It can help prevent or manage many costly chronic diseases that affect We performed a systematic review using Preferred Reporting older adults. It can also reduce the risk of developing func- Items for Systematic Reviews and Meta-analyses (PRISMA) tional limitations and premature death [4]. In addition the guidelines [15]. The review was registered in the Interna- World Health Organisation (WHO) states that older people tional Prospective Register of Systematic Reviews (PROS- over 65 years old should dedicate 150 min a week to do PERO) CRD42021192499. moderate physical activity or 75 min of vigorous physical activity [2]. Criteria for considering the studies in this review Physical exercise is a planned, structured and repetitive We included randomised clinical trials (RCTs) in older adults physical activity done to improve and/or maintain one or over 65 years of age who participate in HBE programmes more of the basic capacities of physical fitness [8]. The delivered with DHI. We included studies if they fulfilled the physical function should be assessed in older adults with following criteria: (i) original research; (ii) people older than objectively quantitative measures, which include: muscu- 65 years; (iii) reported intervention of HBE programme; (iv) lar strength, gait speed, balance, mobility, cardiorespiratory used a DHI and (v) reported a physical function measure. endurance, physical performance and functional capacity [5, The main outcome was the physical function within 9, 10]. which was considered: muscle strength, gait speed, balance, This physical exercise can be performed in different sce- mobility, cardiorespiratory endurance or physical perfor- narios such as rehabilitation centres, gym’s, public parks or at mance tests [16]. The secondary outcomes were functional home. The exercise programme is considered home-based if ability, health-related quality of life (HRQoL), falls and the physical exercise is performed in an informal and flexible adherence to treatment. DHI includes those related to the place such as the individual’s house. The programme, should internet, for example the use of telephone calls, and video have clear goals and include monitoring, follow-up visits, games, when these play a significant role in the delivery and calls from health professionals or self-monitoring diaries dosage of the exercise programme. [11]. Digital health intervention (DHI) is the use of digital, mobile and wireless technologies to support the achievement Search strategies and data resources of health objectives [12]. It describes the general use of infor- We reviewed six databases: Embase, PubMed/MEDLINE, mation and communications technologies (ICT) for health CINAHL, Web of Science, Cochrane Central Register of and includes mHealth and eHealth [12]. DHI combined Controlled Trials (CENTRAL) and ScienceDirect from their with physical exercise may be an opportunity to promote inception to 1 December 2021. We imposed no language or more active and healthy ageing [13]. The coronavirus disease publication restrictions. 2 Effectiveness of home-based exercise delivered by digital health in older adults The terms selected were combined using Boolean logical Results operators (OR, AND and NOT). Moreover, we did a man- ual search of the references that were included in the selected Study selection articles. From the 4,323 identified references, 26 articles were finally included. We removed 1,176 duplicated studies and screened Reviewing procedure and study selection 3,147 studies. Finally, we had 167 studies assessed based on the full text. We excluded 84 studies for wrong pop- Two investigators (LSN–AG) performed the review inde- ulations, 26 for wrong intervention, 15 for wrong study pendently. The 1st step consisted of reviewing the titles designs, 7 studies for being conference abstracts, 5 for wrong and abstracts of all the references retrieved by the database outcome and 4 for duplicate data (Supplementary Figure 1, searches (LSN–AG) and identifying the studies that met the Supplementary data are available in Age and Ageing online). inclusion criteria. Next, we selected all articles that were deemed potentially eligible by at least one of the reviewers. In the 2nd step, the reviewed the full texts, and a decision on Characteristics of the included studies inclusion or exclusion was made according to the predefined Eleven studies were performed in Europe [19–29], six in selection criteria (LSN–AG). A 3rd reviewer (MSR) solved North America [30–35], six in Asia [36–41] and four in any disagreement in any step. All studies that did not fulfil Oceania [24, 42–44]. Out of the 26 studies, 20 were unicen- the predefined criteria were excluded, and their bibliographic tric RCTs [20–22, 26–30, 32, 33, 35–44], and 6 were multi- details were listed with the specific reason for exclusion. centric RCTs [19, 23–25, 31, 34] (Supplementary Table 1). Data extraction Participants Two authors (LSN–AG) extracted the data independently In total, 5,133 participants were enrolled in the included and in duplicate, using a standardised protocol and reporting studies, 2,542 (66.4% females) in the intervention group forms. The following information was extracted from each (IG) and 2,591 (65.0% females) in the control group (CG). included study: author(s), year of publication, publication Thirteen studies were performed in healthy and/or sedentary details, population characteristics (sample size, age and gen- community-dwelling older adults [22, 24–26, 28–31, 37, der), type of physical exercise performed and typology of 38, 41, 42, 44], and 13 included pathologies like cardiorespi- DHI, description of the test, test/scale for each variable and ratory problems [23, 33, 40, 43], balance disorders [20, 27, a description of the results. This compilation was done using 34, 36], musculoskeletal disorders [32, 39], cancer [19, 35] the COVIDENCE© platform [17]. If some relevant data and others [21]. were not included in the article, the author was contacted to request the information. Interventions Methodological quality assessment The DHI used in the articles selected were nine mobile applications [23, 26–30, 39, 40, 44], six phone calls [19, An assessment of the methodological quality of the primary 22, 25, 32, 34, 37], three websites [20, 21, 33], three DVDs articles was carried out using the Cochrane Collaboration [35, 36, 41], two exergames [24, 42], two videoconferences tool for assessing the risk of bias (the Cochrane Handbook [38, 43] and one embodied conversational agent [31]. The for Systematic Reviews of Interventions; [18]). Two review- exercise interventions were multicomponent (two or more ers (CF–MS) independently assessed the risk of bias of the components as strength, balance or endurance) in 17 articles studies. A 3rd author (RTC) was consulted for discrepancies [19, 20, 24–29, 32–37, 39, 41, 43], physical activity incen- that could not be resolved. tive in three articles [21, 22, 31], balance in three articles [30, 42, 44], endurance in two articles [23, 40] and strength in Data synthesis and analysis one article [38]. We reported summaries of the association between the The interaction was via communication with the therapist interventions and the outcomes for each study in terms in 16 articles, automatic in seven articles [24, 26, 29–31, 42, of the mean differences of absolute values. We obtained 44], one with the support of home helper [27] and two did combined measurements of effect for each primary outcome not report. The communication frequency was, at least, every through meta-analysis under a random-effect model due to week in 17 articles [20, 23, 24, 26–32, 36–39, 42–44] with the expected heterogeneity between the studies. Statistical eight studies with feedback every session [24, 27, 30, 38, 39, heterogeneity was measured through the I2 statistic and 42–44]. classified as low (I2 < 25%), moderate (I2 25–50%) or The comparator groups were heterogeneous including high (I2 > 50%) [18]. Subgroup analysis was performed usual care [20, 23, 25, 27, 33, 42–44], lifestyle intervention (whenever possible) according to type of population [28, 35, 38], no intervention [34, 37, 41], wait list [21, 32], classifying the intervention in healthy older adults or with education [24, 36], exercise without DHI [29, 30, 39] and diseases depending on the population described by the exercise/physical activity recommendations [19, 22, 26, 31, author. 40] (Supplementary Table 2). 3 L. Solis-Navarro et al. Figure 1. Forest plot for mobility (TUG) Risk of bias assessment 35]. The forest plot showed that both groups had similar Due to insufficient information, most trials were scored values (MD 0.22 points; 95% CI −0.05 to 0.49; P = 0.12, unclear in the selection bias category (random sequence I 2 = 16%; Supplementary Figure 4, Supplementary data are generation and allocation concealment). Most trials were available in Age and Ageing online). If we analysed by type of scored as unclear or high risk for blinding of participants and population, the studies with healthy older adults showed personnel. Close to half of the authors did not report it, and that both groups had similar values (MD 0.13 points; 95% those who did report it raised the difficulty of blinding due CI −0.10 to 0.36; P = 0.26, I 2 = 0%). A similar result was to the intervention’s nature. About two-thirds of the trials observed in the studies that included older adults with were unclear or high risk of detection, attrition and report- diseases (MD 0.56; 95% CI −0.38 to 1.50; P = 0.24, ing bias. However, most of the trials provided insufficient I 2 = 65%) in comparison with the CG. information to assess whether a critical risk of bias existed Seven studies examined the strength of lower limbs with for other sources of bias (Supplementary Figures 2 and 3, five times sit-to-stand test (5STS) [24, 30, 34, 36, 41, 42, Supplementary data are available in Age and Ageing online). 44] analysing 543 participants in the IG and 577 in the CG [24, 30, 34, 36, 41, 42, 44]. The forest plot showed that IG significantly improved in −0.56 s (95% CI −1.00 Main findings to −0.11; P = 0.01, I 2 = 2%). If we analysed by type of population, the studies with healthy older adults the IG Physical function showed a significant improvement (MD −0.51; 95% CI Thirteen studies examined mobility with the timed up and −0.95 to −0.08; P = 0.02, I 2 = 0%) in comparison with the go test (TUG) [20, 24, 25, 27, 28, 30, 32, 34, 36, 39, CG (Figure 2). 41, 42, 44] analysing 717 participants in the IG and 764 Ten studies reported the balance with Berg balance scale participants in the CG [20, 24, 25, 27, 28, 30, 34, 39, (BBS), functional reach test (FRT), SPPB balance score [20, 41, 42, 44]. The forest plot showed that both groups had 25, 30, 34–37, 39, 42, 44] analysing 457 participants in similar values (SMD −0.05; 95% confidence interval, CI the IG and 517 participants in the CG [20, 25, 30, 34, −0.16 to 0.05; P = 0.33, I 2 = 0%). If we analysed by type 35, 37, 39]. The forest plot showed that both groups had of population, the studies with healthy older adults showed similar values (SMD 0.17; 95% CI −0.12 to 0.45; P = 0.25, that both groups had similar values (SMD 0.00; 95% CI I 2 = 64%). If we analysed by type of population, the studies −0.12 to 0.12; P = 0.99, I 2 = 0%) and in the studies that with healthy older adults showed that both groups had included older adults with diseases the IG showed a signifi- similar values (SMD 0.20; 95% CI −0.76 to 1.17; P = 0.68, cant improvement (SMD −0.23; 95% CI −0.45 to −0.01; I 2 = 76%) and in the studies that included older adults with P = 0.04, I 2 = 0%) in comparison with the CG (Figure 1). diseases the IG showed a significant improvement (SMD Eight studies examined physical performance with the 0.28; 95% CI 0.09–0.48; P = 0.004, I 2 = 1%) in comparison short physical performance battery (SPPB; [19, 24, 26, with the CG (Figure 3). 28, 29, 34, 35, 44]) analysing 307 participants in the Six studies assessed the gait speed [19, 24, 30, 34, 41, 44] IG and 332 participants in the CG [19, 24, 28, 29, 34, analysing 293 participants in the IG and 286 participants 4 Effectiveness of home-based exercise delivered by digital health in older adults Figure 2. Forest plot for muscular strength of lower limbs (5STS) Figure 3. Forest plot for balance (BBS, FRT, SPPB balance score) in the CG [19, 24, 30, 34, 41, 44]. The forest plot showed I 2 = 2%). If we analysed by type of population, the studies that both groups had similar values (SMD −0.07; 95% CI with healthy older adults showed that both groups had sim- −0.23 to 0.10; P = 0.42, I 2 = 0%) (Supplementary Figure 5, ilar values (OR 0.85; 95% CI 0.68–1.07; P = 0.17, I 2 = 0%) Supplementary data are available in Age and Ageing online). and in the studies that included older adults with diseases the We did not analyse by type of population due to insufficient IG showed a significant reduction in risk of falls (OR 0.60; number of studies. 95% CI 0.42–0.87; P = 0.01, I 2 = 2%) in comparison with Three studies examined the functional ability with the the CG (Figure 4). Barthel index (BI) [20, 27, 39] analysing 155 participants Eight studies examined the falls efficacy scale-international in the IG and 156 participants in the CG. The forest (FES-I; [20, 24, 29, 36, 39, 42, 44]). The studies that plot shows that the IG had a significant improvement in reported the score between groups post-intervention 5.01 points (95% CI 0.24–9.79 points; P = 0.04, I 2 = 39%; analysed 371 participants in the IG and 377 participants in Supplementary Figure 6, Supplementary data are available the CG [20, 24, 29, 39, 42, 44]. The forest plot showed that in Age and Ageing online). both groups had similar values (MD −0.05 points; 95% CI −1.27 to 1.17; P = 0.94, I 2 = 44%). If we analysed by Falls type of population, the studies with healthy older adults Seven studies examined the falls events [24–27, 34, 36, 44] showed that both groups had similar values (MD 0.16 analysing 1,048 participants in the IG and 1,083 participants points; 95% CI −1.32 to 1.64; P = 0.83, I 2 = 54%). A in the CG. The forest plot showed that the odds ratio (OR) similar result was observed in the studies that included was reduced 0.77 in the IG (95% CI 0.63–0.94; P = 0.007, older adults with diseases (MD −1.05; 95% CI −2.98 5 L. Solis-Navarro et al. Figure 4. Forest plot for falls events to 0.88; P = 0.29, I 2 = 0%) in comparison with the CG only older adults with diseases, it improved the balance and (Supplementary Figure 7, Supplementary data are available mobility. in Age and Ageing online). Our review included older adults with different base- line characteristics. Although the majority of studies were Quality of life conducted in apparently healthy older adults, a significant Twelve studies examined the HRQoL [20, 23–25, 27, 28, proportion was performed in older adults with diseases. 31, 33, 36, 40, 43, 44] analysing 536 participants in the Therefore, the potential for improvement in these popula- IG and 563 participants in the CG [20, 24, 25, 40, 43, tions is different, since patients with diseases have a lower 44]. The forest plot shows that the IG had a significant functional status [45]. improvement in respect to the CG (SMD 0.18; 95% CI This could explain why, when analysed at the entire pop- 0.05–0.30; P = 0.004, I 2 = 95%). If we analysed by type of ulation, only the 5STS showed a significant difference post- population, the studies with healthy older adults showed intervention. On the other hand, when analysing the disease that both groups had similar values (SMD 0.07; 95% CI group, TUG and balance were added with significant results. −0.26 to 0.40; P = 0.67, I 2 = 80%) and in the studies This is even more evident if we analysed some outcomes that that included older adults with diseases the IG showed a turned out to be favourable to the intervention in the entire significant improvement in HRQoL (SMD 2.77; 95% CI population, but that the effect is really due to the studies that 0.41 to 5.12; P = 0.02, I 2 = 97%) in comparison with the had patients with diseases, such as HRQoL, or the number CG (Figure 5). of falls. Of all the evaluations, only the 5STS showed a significant Adherence post-intervention change in all patients. Our results are similar to those observed by Mañas et al., who also analysed Adherence was reported in more than half of the articles [19, HBE but focused on resistance exercise [46]. And, as in our 20, 24–26, 28, 29, 31–33, 36, 39, 40, 42, 43]. There were review, no significant changes were observed in performance different forms to describe the adherence, like sessions com- or mobility tests such as SPPB or TUG. The 5STS is a widely pleted [19, 28], and patients that completed the programme used test in older adults, by itself or as part of the SPPB; it is [20, 32], with a threshold of expected exercise [25, 26, 36, used as a functional measure of strength of lower limbs, and 39]. has clinical relevance because the ability to go from sitting to standing position reflects an important functional skill in Discussion older people [9]. In addition, the 5STS is related to strength and balance, elements incorporated in the different training The HBE programmes carried out by DHI improved programmes, so it would be a possible explanation for this physical function in terms of lower extremity strength and improvement [47]. functional capacity. It also significantly reduced the number Current recommendations indicate the importance of of falls and improved the HRQoL. In addition, if we analysed training lower limb strength in older adults, making it a 6 Effectiveness of home-based exercise delivered by digital health in older adults Figure 5. Forest plot for health-related quality of life (HRQoL) very frequently training task, safe to perform without direct Although more than half of the studies reported adher- supervision and easy to progress (overload; [48]). Different ence to exercise [19, 20, 24–26, 28, 29, 31–33, 36, 39, 40, exercise guides, for example Otago, incorporate sitting and 42, 43], the concept of adherence is not well-established and standing exercises in their base routine, so it is very possible varies between the studies [56], confusing adherence with that this task is trained and, therefore, susceptible to change attendance [19, 28], that is, the number or percentage of in a test that measures it in the same motor gesture [49]. sessions to which patients are assisted [56]. Another way HBE programmes did not show an improvement in to describe exercise adherence is to count the number of balance. However, when considering only the studies with dropouts during your exercise intervention [56]. Following older adults with diseases, a positive effect in favour of the this last conception, a good adherence to the programme intervention was observed. A possible explanation for this could be interpreted when the percentage of its participants improvement could be because some studies were specifically who finished the intervention is high [20, 32]. In this sense, aimed at patients at risk of falling [20, 27, 34, 36], and adherence has been described as the degree to which a the studies that performed a multicomponent intervention person’s behaviour correlates with the intervention planned, included the balance training. thus it would be related to the degree to which the target In this systematic review, most of the studies included intensity and volume are achieved [25, 26, 36, 39]. In the multicomponent exercises complying with the recommen- selected studies, adherence rates are highly variable, so it was dations of the clinical guidelines [48]. The effectiveness of not possible to compare adherence results, and this factor multicomponent exercise has been demonstrated in the liter- could influence that we do not observe effects in the majority ature in other populations with chronic non-communicable of outcomes assessed. diseases or sedentary [50, 51], also associated with DHI [51] Another critical point to consider is whether HBE with or in home-based setting [52]. DHI is effective compared with face-to-face programs. This Interestingly, the studies that evaluated the total number is particularly relevant since home physical training pro- of falls showed that the IG reduced falls by 23%, being up to grams, which aim to improve health functions at a distance 40% in the group of older adults with diseases. However, this using telecommunication strategies, could provide equitable fact was not observed when the FES-I scale was evaluated. access to training/rehabilitation services [57], thus allowing One possible explanation is that the FES-I is a subjective many isolated people to slow down their loss of autonomy measure of ‘fear of falling’ or, more appropriately, ‘concerns [58]. In addition, although we could not compare it in this about falling’ [53]; instead, the events of falls are objective. study due to the heterogeneity of the comparators, a recent Both evaluations are complemented by what we consider RCT comparing HBE through videoconference versus face- relevant so that both are reported. to-face in older people between 70 and 80 years old found The functional ability evaluated through the BI showed similar effects on body composition and cardiorespiratory an improvement. Although the literature suggests the assess- fitness, suggesting that when face-to-face training is not ment of both basic and instrumental activities of daily living possible, HBE through videoconference should be advised (ADL) in older people [54, 55], the small number of studies [59]. that report it is striking. A possible explanation could be that many studies are done in community-dwelling and the ADL Strengths and limitations scales tend to reach a ceiling very quickly, so they are not Our study has several strengths. This review shows that HBE usually used in exercise programmes. with different types of technological support are effective in 7 L. Solis-Navarro et al. improving physical function in older people and are more so 2. Bull FC, Al-Ansari SS, Biddle S et al. World Health Orga- in people with pathologies. This is important because people nization 2020 guidelines on physical activity and sedentary with a higher burden of disease, with multiple pathologies or behaviour. Br J Sports Med 2020; 54: 1451–62. frail are the ones who have more problems attending com- 3. Piercy KL, Troiano RP, Ballard RM et al. the physical activity munity programmes, so home interventions can be effective guidelines for Americans. JAMA 2018; 320: 2020–8. 4. Mora JC, Valencia WM. Exercise and older adults. Clin in different aspects of physical function, such as strength Geriatr Med Med 2018; 34: 145–62. of lower limbs, mobility and balance, all important aspects 5. American College of Sports Medicine. ACSMs Guidelines for to help maintain the functional independence of the older Exercise Testing and Prescription. 2021 https://www.acsm.o person. rg/read-research/books/acsms-guidelines-for-exercise-testi Our study has some limitations. Most of the studies ng-and-prescription (2 November 2021, date last accessed). included populations with different types of diseases or 6. Sun F, Norman IJ, While AE. Physical activity in older people: comorbidities, which may limit the extrapolation of the a systematic review. BMC Public Health 2013131 2013; 13: results and recommendations to the entire spectrum of older 1–17. https://doi.org/10.1186/1471-2458-13-449. people; even though our results were statistically significant. 7. Sherrington C, Fairhall NJ, Wallbank GK et al. Exercise for The heterogeneous nature of populations implies that many preventing falls in older people living in the community. subjects have different pathophysiological behaviours and Cochrane Database Syst Rev 2019; 2019: 204–5. 8. Caspersen CJ, Powell KE, Christenson G. Physical activity, conditions, a wide spectrum of severity, as well as the impli- exercise, and physical fitness: definitions and distinctions cation or impact that suffering from associated comorbidities for health-related research. Public Health Rep 1985; 100: may influence the magnitude of reported results. In addition, 126–31. there was heterogeneity in the CG since some of them 9. Soubra R, Chkeir A, Novella JL. A systematic review performed no interventions, lifestyle interventions or only of thirty-one assessment tests to evaluate mobility exercised without DHI, hindering the comparison between in older adults. Biomed Res Int 2019; 2019: 1–17. them. For example, some authors delivered written recom- https://doi.org/10.1155/2019/1354362. mendations for exercise [24, 26], others through videos [35], 10. Mijnarends DM, Meijers JMM, Halfens RJG et al. Validity and some through verbal recommendations [28]; however, and reliability of tools to measure muscle mass, strength, and it was not measured that the proposed exercises complied physical performance in community-dwelling older people: a with the principles of training, so it cannot be concluded systematic review. J Am Med Dir Assoc 2013; 14: 170–8. 11. Ashworth NL, Chad KE, Harrison EL, Reeder BA, that only the DHI was really the difference. Finally, we do Marshall SC, Cochrane Musculoskeletal Group. Home not compare the applied DHIs. Mainly because the studies versus center based physical activity programs in older that have an app are recent and those that have phone calls are adults. Cochrane Database Syst Rev 2005; CD004017. older. Future reviews should consider the differences between https://doi.org/10.1002/14651858.CD004017.pub2. DHI interventions in order to determine those which are 12. World Health Organization. Monitoring and Evaluating Dig- more efficient in these populations. ital Health Interventions: a Practical Guide to Conducting Research and Assessment. Geneva PP - Geneva: World Health Conclusion Organization, 2016. 13. Ienca M, Schneble C, Kressig RW, Wangmo T. Digital health The HBE programmes carried out by using DHI improved interventions for healthy ageing: a qualitative user evalua- physical function in terms of lower extremity strength and tion and ethical assessment. BMC Geriatr 2021; 21: 1–10. functional capacity and also reduced the number of falls https://doi.org/10.1186/s12877-021-02338-z. 14. von Humboldt S, Mendoza-Ruvalcaba NM, Arias-Merino and improved the HRQoL in older adults. Additionally, for ED et al. Smart technology and the meaning in life of older those older adults with diseases, it improves the balance and adults during the Covid-19 public health emergency period: a mobility. cross-cultural qualitative study. Int Rev Psychiatry 2020; 32: 713–22. Supplementary Data: Supplementary data mentioned in 15. Moher D, Liberati A, Tetzlaff J et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA the text are available to subscribers in Age and Ageing online. statement. Ann Intern Med 2009; 151: 264–9. Declaration of Conflicts of Interest: None. 16. Jadczak AD, Makwana N, Luscombe-Marsh N, Visvanathan R, Schultz TJ. Effectiveness of exercise interventions on phys- Declaration of Sources of Funding: This study was ical function in community-dwelling frail older people: an supported by a grant from Col·legi de Fisioterapeutes de umbrella review of systematic reviews. JBI database Syst Rev Catalunya (R02/20). Implement reports 2018; 16: 752–75. 17. Covidence. https://www.covidence.org/ (5 April 2022, date last accessed). References 18. Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic 1. World Health Organization. Ageing and Health. 2021 Reviews of Interventions version 6.1 (updated September https://www.who.int/news-room/fact-sheets/detail/ageing-a 2020). Cochrane, 2020. www.training.cochrane.org/handboo nd-health (2 May 2022, date last accessed). k. Accessed, https://doi.org/10.1002/9781119536604. 8 Effectiveness of home-based exercise delivered by digital health in older adults 19. Arrieta H, Astrugue C, Regueme S et al. Effects of a physical 32. Goode AP, Taylor SS, Hastings SN, Stanwyck C, Coffman activity programme to prevent physical performance decline CJ, Allen KD. Effects of a home-based telephone-supported in onco-geriatric patients: a randomized multicentre trial. J physical activity program for older adult veterans with chronic Cachexia Sarcopenia Muscle 2019; 10: 287–97. low back pain. Phys Ther 2018; 98: 369–80. 20. Bernocchi P, Giordano A, Pintavalle G et al. Feasibility and 33. Tomita MR, Tsai BM, Fisher NM et al. Effects of multi- clinical efficacy of a multidisciplinary home-telehealth pro- disciplinary internet-based program on management of heart gram to prevent falls in older adults: a randomized controlled failure. J Multidiscip Healthc 2009; 2: 13–21. trial. J Am Med Dir Assoc 2019; 20: 340–6. 34. Arena SK, Wilson CM, Boright L, Peterson E. Impact of 21. Volders E, Bolman CAW, De Groot RHM, Verboon P, the HOP-UP-PT program on older adults at risk to fall: a Lechner L. The effect of active plus, a computer-tailored randomized controlled trial. BMC Geriatr 2021; 21: 520. physical activity intervention, on the physical activity of older https://doi.org/10.1186/s12877-021-02450-0. adults with chronic illness(es)-a cluster randomized controlled 35. Salerno EA, Gothe NP, Fanning J, Peterson LL, Colditz trial. Int J Environ Res Public Health 2020; 17: 2590. GA, McAuley E. Effects of a DVD-delivered random- https://doi.org/10.3390/ijerph17072590. ized controlled physical activity intervention on functional 22. Von Bonsdorff MB, Leinonen R, Kujala UM et al. Effect of health in cancer survivors. BMC Cancer 2021; 21: 1–9. physical activity counseling on disability in older people: a 2- https://doi.org/10.1186/s12885-021-08608-8. year randomized controlled trial. J Am Geriatr Soc 2008; 56: 36. Boongird C, Keesukphan P, Phiphadthakusolkul S, Rat- 2188–94. tanasiri S, Thakkinstian A. Effects of a simple home-based 23. Scherrenberg M, Zeymer U, Schneider S et al. EU-CaRE exercise program on fall prevention in older adults: a 12- study: could exercise-based cardiac telerehabilitation also be month primary care setting, randomized controlled trial. cost-effective in elderly? Int J Cardiol 2021; 340: 1–6. Geriatr Gerontol Int 2017; 17: 2157–63. https://doi.org/10.1016/j.ijcard.2021.08.024. 37. Chang M, Huang YH, Jung H. The effectiveness of the 24. Gschwind YJ, Eichberg S, Ejupi A et al. ICT-based system exercise education programme on fall prevention of the to predict and prevent falls (iStoppFalls): results from an community-dwelling elderly: a preliminary study. Hong Kong international multicenter randomized controlled trial. Eur J Occup Ther 2011; 21: 56–63. Rev Aging Phys Act 2015; 12: 10. https://doi.org/10.1186/ 38. Hong J, Kim J, Kim SW, Kong HJ. Effects of home- s11556-015-0155-6. based tele-exercise on sarcopenia among community-dwelling 25. Iliffe S, Kendrick D, Morris R et al. Multicentre cluster elderly adults: body composition and functional fitness. Exp randomised trial comparing a community group exercise pro- Gerontol 2017; 87: 33–9. gramme and home-based exercise with usual care for people 39. Li CT, Hung GK, Fong KN, Gonzalez PC, Wah SH, Tsang aged 65 years and over in primary care. Health Technol Assess HW. Effects of a home-based occupational therapy telere- (Rockv) 2014; 18: 1–106. https://doi.org/10.3310/hta18490. habilitation via smartphone for outpatients after hip fracture 26. Mansson L, Lundin-Olsson L, Skelton DA et al. Older surgery: a feasibility randomised controlled study. J Telemed adults’ preferences for, adherence to and experiences of Telecare 2020; 28: 239–47. two self-management falls prevention home exercise pro- 40. Liu WT, Wang CH, Lin HC et al. Efficacy of a cell phone- grammes: a comparison between a digital programme based exercise programme for COPD. Eur Respir J 2008; 32: and a paper booklet. BMC Geriatr 2020; 20: 209. 651–9. https://doi.org/10.1186/s12877-020-01592-x. 41. Yamada M, Aoyama T, Hikita Y et al. Effects of a DVD-based 27. Mézière A, Oubaya N, Michel-Pellegrino V et al. Exercise seated dual-task stepping exercise on the fall risk factors among interventions with trained home helpers for preventing loss community-dwelling elderly adults. Telemed e-Health 2011; of autonomy and falls in community-dwelling older adults 17: 768–72. receiving home heath physical therapy T4H: a randomized 42. Schoene D, Lord SR, Delbaere K, Severino C, Davies controlled pilot study. J Geriatr Phys Ther 2021; 44: E138– TA, Smith ST. A randomized controlled pilot study 49. of home-based step training in older people using 28. van den Helder J, Mehra S, van Dronkelaar C et al. Blended videogame technology. PLoS One 2013; 8: e57734. home-based exercise and dietary protein in community- https://doi.org/10.1371/journal.pone.0057734. dwelling older adults: a cluster randomized controlled trial. 43. Tsai LLY, McNamara RJ, Moddel C, Alison JA, McKenzie J Cachexia Sarcopenia Muscle 2020; 11: 1590–602. DK, McKeough ZJ. Home-based telerehabilitation 29. Van Het Reve E, Silveira P, Daniel F, Casati F, De Bruin via real-time videoconferencing improves endurance ED. Tablet-based strength-balance training to motivate and exercise capacity in patients with COPD: the randomized improve adherence to exercise in independently living older controlled TeleR Study. Respirology 2017; 22: 699–707. people: part 2 of a phase ii preclinical exploratory trial. J 44. Delbaere K, Valenzuela T, Lord SR et al. E-health Standing- Med Internet Res 2014; 16: e159. https://doi.org/10.2196/ Tall balance exercise for fall prevention in older people: results jmir.3055. of a two year randomised controlled trial. BMJ 2021; 373: 30. Bao T, Carender WJ, Kinnaird C et al. Effects of long- n740. https://doi.org/10.1136/bmj.n740. term balance training with vibrotactile sensory augmentation 45. Fong JH. Disability incidence and functional decline among among community-dwelling healthy older adults: a random- older adults with major chronic diseases. BMC Geriatr 2019; ized preliminary study. J Neuroeng Rehabil 2018; 15: 5. 19: 323. https://doi.org/10.1186/s12877-019-1348-z. https://doi.org/10.1186/s12984-017-0339-6. 46. Mañas A, Gómez-Redondo P, Valenzuela PL, Morales 31. Bickmore TW, Silliman RA, Nelson K et al. A randomized JS, Lucía A, Ara I. Unsupervised home-based resistance controlled trial of an automated exercise coach for older adults. training for community-dwelling older adults: a J Am Geriatr Soc 2013; 61: 1676–83. systematic review and meta-analysis of randomized 9 L. Solis-Navarro et al. controlled trials. Ageing Res Rev 2021; 69: 101368. efficacy scale-international (FES-I). Age Ageing 2005; 34: https://doi.org/10.1016/j.arr.2021.101368. 614–9. 47. Muñoz-Bermejo L, Adsuar JC, Mendoza-Muñoz M et al. Test- 54. Veronese N, Custodero C, Demurtas J et al. Compre- retest reliability of five times sit to stand test (FTSST) in hensive geriatric assessment in older people: an umbrella adults: a systematic review and meta-analysis. Biology (Basel) review of health outcomes. Age Ageing 2022; 51: afac104. 2021; 10: 510. https://doi.org/10.3390/biology10060510. https://doi.org/10.1093/ageing/afac104. 48. Izquierdo M, Merchant RA, Morley JE et al. International 55. Briggs R, McDonough A, Ellis G, Bennett K, O’Neill exercise recommendations in older adults (ICFSR): expert D, Robinson D. Comprehensive geriatric assessment consensus guidelines. J Nutr Health Aging 2021257 2021; 25: for community-dwelling, high-risk, frail, older people. 824–53. Cochrane Database Syst Rev 2022; 2022. https://doi.o 49. Thomas S, Mackintosh S, Halbert J. Does the ‘Otago exercise rg/10.1002/14651858.cd012705.pub2. programme’ reduce mortality and falls in older adults?: a 56. Collado-Mateo D, Lavín-Pérez AM, Peñacoba C et al. Key systematic review and meta-analysis. Age Ageing 2010; 39: factors associated with adherence to physical exercise in 681–7. patients with chronic diseases and older adults: an umbrella 50. Zhang M, Wang W, Li M, Sheng H, Zhai Y. Efficacy review. Int J Environ Res Public Health 2021; 18: 1–24. of mobile health applications to improve physical https://doi.org/10.3390/ijerph18042023. activity and sedentary behavior: a systematic review 57. Russell TG. Physical rehabilitation using telemedicine. J and meta-analysis for physically inactive individuals. Telemed Telecare 2007; 13: 217–20. Int J Environ Res Public Health 2022; 19: 4905. 58. Bigot L, Langeard A, Moussay S, Gauthier A, Quarck G. https://doi.org/10.3390/ijerph19084905. Activité physique à domicile pour les seniors: revue de la 51. Batrakoulis A, Jamurtas AZ, Metsios GS et al. Comparative question et proposition d’une pratique optimisée. Mov Sport efficacy of 5 exercise types on cardiometabolic health in over- Sci Mot 2019; 103: 27–37. weight and obese adults: a systematic review and network 59. Langeard A, Bigot L, Maffiuletti NA et al. Non-inferiority meta-analysis of 81 randomized controlled trials. Circ Car- of a home-based videoconference physical training program diovasc Qual Outcomes 2022; 15(6): e008243. https://doi.o in comparison with the same program administered face- rg/10.1161/CIRCOUTCOMES.121.008243. to-face in healthy older adults: the MOTION randomised 52. Lee H, Lee SH. Effectiveness of multicomponent home-based controlled trial. Age Ageing 2022; 51: afac059. https://doi.o rehabilitation in older patients after hip fracture surgery: rg/10.1093/ageing/afac059. a systematic review and meta-analysis. J Clin Nurs 2022. https://doi.org/10.1111/jocn.16256. 53. Yardley L, Beyer N, Hauer K, Kempen G, Piot-Ziegler C, Todd C. Development and initial validation of the falls Received 12 June 2022; editorial decision 15 July 2022 10