Each year, approximately one-quarter to one-third of adults aged >65 years living in the community experience a fall, a finding consistently reported across international populations [1-3]. A fall is defined as “an unexpected event in which the participant comes to rest on the ground, floor, or lower level” [4]. The consequences of falls range in severity from minor bruising to debilitating fractures or even death [5]. Beyond physical injuries, falls can also have considerable psychosocial effects, as concerns about falling and loss of independence can lead to reduced social engagement and diminished quality of life (QoL) [6].

Falls also place a substantial financial burden on health care systems worldwide due to hospitalizations and the need for long-term care following injurious falls [7]. Many risk factors for falls, including balance and strength deficits, cognitive decline, and reduced muscle mass, are associated with aging and low levels of physical activity [8]. Regular physical activity can mitigate these issues by improving muscle strength, balance, and overall physical function, as well as enhancing cognitive performance, mood, and QoL [9-11]. Strong evidence suggests that physical activity programs, particularly those incorporating balance and functional exercises, can reduce both the rate of falls and the number of older people who experience falls in the community [12]. Nevertheless, adherence to these programs is often modest and tends to decline over time, with systematic reviews reporting full adherence rates as low as 21% for home-based programs [13] and only approximately half of participants remaining adherent at 12 months [14,15].

Exergaming offers a promising alternative by providing an immersive and engaging experience, incorporating gamified elements such as feedback, achievements, and progress tracking to enhance motivation and adherence [16]. Exergames, broadly defined as active video games that integrate gameplay with physical exercise, use real-time motion detection to track players’ movements [17]. These games may be delivered via animated gaming platforms or within 2D or 3D virtual reality (VR) environments and are increasingly being explored as a tool for fall prevention [18,19].

As exergaming can be tailored to an individual’s functional ability, it may make physical activity more accessible, enjoyable, and acceptable for older adults, addressing barriers such as low motivation and physical limitations [16,20]. As a fall prevention tool, exergaming interventions often gamify physical activities targeting balance, strength, 3D movement (eg, dance), endurance [20-22], and cognitive function [23,24]. These interventions can be provided to individuals or groups in a wide range of settings, including home-based settings, community settings, and hospitals, with or without facilitation by a health professional.

Several systematic reviews and meta-analyses have examined exergaming in older adults, reporting improvements in intermediate outcomes, such as balance, physical function, cognitive outcomes, and concerns about falling [25-32]. These findings suggest that exergaming may improve important fall-related risk factors and may enhance exercise engagement through interactive feedback and gamification. However, comparatively less attention has been given to whether these interventions translate into reductions in clinically measured falls and injurious falls, which remain the most meaningful outcomes for patients, clinicians, and health services. In addition, previous reviews have given limited consideration to implementation-relevant factors, such as adherence, acceptability, adverse events, and cost-effectiveness, all of which are important when evaluating the real-world feasibility of integrating exergaming into fall prevention pathways. Furthermore, prior syntheses have rarely distinguished between usual care, evidence-based active exercise controls, and non–evidence-based active comparator interventions, limiting the interpretation of whether exergaming offers benefit beyond established fall prevention practice. A more clinically focused synthesis of the available evidence is therefore needed.

To address these gaps, the aim of this systematic review and meta-analysis was to evaluate the effectiveness of exergaming interventions for preventing falls and injurious falls in adults aged ≥60 years. The secondary objectives were to synthesize evidence relating to QoL, concerns about falling, balance confidence, adherence, acceptability, adverse events, and cost-effectiveness and to examine findings according to comparator type.

We followed Cochrane systematic review methods [33]. The review was conducted and reported in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 statement, using the PRISMA 2020 expanded checklist [34], the PRISMA 2020 for Abstracts checklist [35], and the PRISMA-S extension for reporting literature searches [36]. The study protocol was prospectively registered with PROSPERO (CRD42020214721).

Searches were completed first in March 2023 and updated using the same search strategy in February 2025. The following databases were searched from inception to this date: MEDLINE (Ovid), Embase (Ovid), CINAHL Plus (EBSCO), PsycINFO (Ovid), and Cochrane Central Register of Controlled Trials (CENTRAL) (Ovid). The PICOS framework was used to develop eligibility criteria around the themes of exergames, falls, and older people (Table 1). The search strategy combined database-specific controlled vocabulary terms (eg, MeSH, Emtree, CINAHL Headings, and APA Thesaurus) with free-text terms. Full search strategies for all databases, including database-specific adaptations, are provided in Supplementary Material S1 in Multimedia Appendix 1. No date restrictions were applied, but searches were limited to studies published in English in accordance with the eligibility criteria. Where published reports lacked sufficient methodological or outcome detail, study authors were contacted for clarification or to request additional data. In addition to database searches, reference lists of included studies and relevant systematic reviews were screened, and forward citation searching was conducted manually by 2 independent reviewers (CE-T and YY) using Google Scholar to identify any additional eligible studies.

We included randomized controlled trials from any setting that evaluated the effects of exergaming interventions on falls or injurious falls in older adults aged ≥60 years (Table 1). This threshold aligns with international definitions of older populations (eg, World Health Organization [2]) and allows inclusion of a broad range of relevant studies. We included trials with generally healthy participants and participants with conditions that may increase the risk of falls (eg, dementia, stroke, Parkinson disease, and frailty).

We included trials in which the main purpose of the study was to investigate the effectiveness of exergaming. We excluded studies that focused solely on the design of exergaming interventions.

The review considered all types of exergaming interventions and all delivery methods. This included all interactive active games and immersive or nonimmersive VR platforms, and interventions of any length and duration were eligible. Control interventions included usual care, standard exercise programs with an evidence base in fall prevention (eg, strength and balance training), as well as those with no evidence base in fall prevention (eg, an educational leaflet on exercise and low-intensity or low-frequency exercise) [12].

The types of exercise primarily targeted by the intervention were categorized according to the Prevention of Falls Network Europe (ProFaNE) taxonomy [37], as follows: (1) gait, balance, coordination, and functional training; (2) strength or resistance training; (3) flexibility; (4) 3D exercise (eg, Tai Chi or dance); (5) general physical activity; (6) endurance (eg, treadmill walking); and (7) other kinds of exercise.

The primary outcomes of interest were (1) rate of falls (number of falls over a period and falls per person-year), (2) number of fallers (number of people experiencing one or more falls), and (3) number of people who experienced one or more injurious falls [38].

The secondary outcomes were (1) health-related QoL (measured using a validated scale such as the EQ-5D [39]); (2) concern about falling (measured using a validated scale such as the Falls Efficacy Scale-International [40]); (3) balance confidence (measured using a validated scale such as the Activities-Specific Balance Confidence Scale [41]); (4) adherence to the intervention (defined as the extent to which participants completed the prescribed exergaming program, including attendance, exercise frequency, duration, or retention); (5) acceptability of the intervention (defined as participants’ perceptions of usability, satisfaction, enjoyment, and willingness to engage with or continue the program, as reported by study authors); (6) adverse events (defined as any negative outcome resulting either directly or indirectly from the assigned treatment [42], measured as the number of people who experienced one or more adverse events); and (7) cost-effectiveness of the intervention. These outcome definitions were based on measures reported in the included studies.

This review was prospectively registered in PROSPERO (CRD42020214721) and conducted in accordance with the registered protocol. One methodological amendment was made during the review process. We initially planned to combine all active intervention comparators into a single analysis; however, important conceptual and clinical differences were identified between evidence-based fall prevention interventions and non–evidence-based active controls. These comparator groups were therefore analyzed separately to improve the interpretability and clinical relevance of the findings. This decision was made prior to final data synthesis. No other substantive deviations from the registered protocol were made.

Titles and abstracts from the database searches were imported into the Rayyan web app for systematic reviews [43], and duplicate records were removed. Records were then screened for relevance by 2 independent reviewers (selected from a pool of CE-T, LM, EDB, ES, KD, YY, AM, KG, RC, and AA), and full texts of studies considered potentially relevant were assessed for eligibility. The reasons for exclusion at the full-text stage were recorded and are reported in Supplementary Material S2 in Multimedia Appendix 1. The same prespecified eligibility criteria (Table 1) were applied at title or abstract and full-text screening, and the table served as the screening framework for reviewers. No automated decision-making tools were used, and all decisions were made by reviewers. Any disagreements were resolved through discussion with a third independent reviewer from the pool.

Data were extracted from included studies by 2 independent reviewers (KG and YY) using a standardized data extraction form in an Excel (Microsoft) spreadsheet, and an independent reviewer (CE-T or AA) checked for accuracy. Extracted data included author details, country, year of publication, participant characteristics (eg, age, gender, and health status), study characteristics (eg, design, location, setting, sample size, and length of follow-up), intervention details (eg, exergaming device and dose), outcomes measured, and key results.

We used PROGRESS-Plus criteria (place of residence, race/ethnicity/culture/language, occupation, gender/sex, religion, education, socioeconomic status, and social capital) [44] when extracting participant characteristics.

The TIDieR (Template for Intervention Description and Replication) checklist was used to describe the included interventions and controls [45].

The Excel version of the Cochrane risk of bias (RoB 2.0) tool was used to assess the risk of bias [46,47]. Two independent reviewers assessed the risk of bias for each study, and any disagreements were resolved through discussion with an additional reviewer. Risk of bias was classified as high, low, or some concerns, across the following domains: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result.

Reporting bias, including publication bias and selective reporting, was considered during synthesis. Formal statistical assessment (eg, funnel plots or Egger test) was not undertaken because of the small number of studies included in each meta-analysis, making such methods unreliable. Potential reporting bias was therefore considered qualitatively when interpreting the findings.

We conducted GRADE assessments on the overall evidence for each primary outcome, split by comparator group [48]. Two independent reviewers assessed each outcome (CE-T and AM or LM and SA), and any disagreements were resolved through discussion. Overall bias per outcome was based on the majority rating across domains.

Findings were synthesized narratively and summarized in overview tables to facilitate comparison across interventions. Where 2 or more studies contributed sufficiently clinically comparable data, pooled overall or comparator-specific meta-analyses were conducted in Review Manager (RevMan; version 5.4) [49]. Outcomes informed by single studies are presented for consistency of effect estimation but interpreted narratively. Incidence rate ratios (IRRs) with 95% CIs were calculated for fall rate outcomes, risk ratios (RRs) with 95% CIs for binary outcomes (number of fallers and number of participants experiencing one or more injurious falls), mean differences (MDs) for continuous outcomes reported on the same scale, and standardized MDs (SMDs) where different measurement scales were used. Fixed effects or random effects models were selected according to the clinical comparability between studies. Where pooling was not appropriate because of limited study numbers, heterogeneous outcome measures, or inconsistent reporting, results were synthesized narratively in accordance with Synthesis Without Meta-analysis guidance [50].

Planned subgroup analysis for the primary outcomes included participant age (older vs younger populations), health condition (eg, healthy, stroke, Parkinson disease, and frailty), and intervention setting (community, residential care home, hospital, assisted care facility, sheltered housing, and retirement community). However, these analyses were not feasible because too few studies contributed data within each subgroup. Comparator-specific analyses were undertaken instead, as these provided the most clinically interpretable basis for exploring variation in effect estimates.

We identified 3830 records, and after removal of duplicates, 3320 (86.7%) were screened on title and abstract, and 225 (5.9%) full-text reports were assessed for eligibility. Of these, 216 (96%) full-text reports were excluded, most commonly because of ineligible study design, inappropriate intervention, ineligible population, or absence of relevant fall outcomes. A full list of studies excluded at the full-text screening, together with reasons for exclusion, is provided in Supplementary Material S2 in Multimedia Appendix 1. Nine randomized controlled trials involving 1385 participants met the inclusion criteria, with 51% (701/1385) in intervention groups and 49% (684/1385) in control groups. Studies varied considerably in participant populations, intervention formats, comparator types, and follow-up duration (Figure 1).

Participant characteristics and PROGRESS-Plus aspects of equity are presented in Table 2. In total, 59% (817/1385) were women (intervention groups: 416/817, 51% and control groups: 401/817, 49%). The mean age of participants was 72.8 (SD 7.9; intervention groups: 72.9, SD 7.8, and control groups: 72.6, SD 8.0) years. The mean age of participants in the no contact or usual care control studies was 72.8 (SD 6.9) years; in the active control studies (evidence based), the mean age was 73.1 (SD 9.0) years; and in the active control studies (non–evidence based), the mean age was 72.6 (SD 7.9) years. Four studies recruited participants based on a diagnosis of Parkinson using clinical criteria [51-54]. Participants were recruited in 3 studies on the basis that they had a history of falls [51,53,55], and 1 study specifically recruited participants with mild cognitive impairment [53].

No study included data on the research locale (eg, urban or rural), 2 studies noted race, ethnicity, or culture [56,58], and 1 study each presented data on the participants’ language [58], occupation [56], number of years in education [53], and socioeconomic status [56]. Three studies included some information on aspects, which could contribute to social capital (ie, living arrangements) [56-58].

The effectiveness of exergaming interventions is presented according to comparator type: usual care, evidence-based active controls, and non–evidence-based active controls. This approach was used to distinguish the effects of exergaming against usual care or no intervention, against established exercise programs with evidence for fall prevention, and against active control activities not specifically designed to reduce falls.

Three studies, including 652 participants, compared the use of exergames against usual care or no intervention [54,56,57]. Forest plots for this comparison are presented in Figure 3.

Three studies (n=210) reported on the use of exergame interventions and an evidence-based active control [52,55,58]. Forest plots for this comparison are presented in Figure 4.

Three studies (n=523) reported on the use of exergame interventions and a non–evidence-based active control [51,53,59]. Forest plots for this comparison are presented in Figure 5.

This review synthesized evidence on the effectiveness of exergaming interventions for fall prevention in older adults by examining falls and injurious falls as primary outcomes alongside implementation-relevant secondary outcomes. Comparator-specific analyses suggest that exergaming interventions may reduce fall rates when compared with usual care and some active comparators, although the magnitude and certainty of effect varied across comparisons. When data were pooled across all comparator groups, exergaming was also associated with an overall reduction in fall rate; however, this broad summary estimate should be interpreted cautiously because substantial heterogeneity indicates important variation between intervention contexts, participant groups, and comparator conditions. Moderate-certainty evidence also indicated that exergames may reduce the number of older adults experiencing one or more falls at 12-month follow-up compared with usual care. However, evidence for injurious falls was limited to a single study, and confidence in several pooled estimates was reduced by substantial heterogeneity, small numbers of contributing studies, and imprecision.

These findings suggest that exergaming may have value as an adjunct to established exercise-based fall prevention approaches, particularly where motivation, adherence, and sustained participation in conventional exercise programs are challenging. At the same time, the review highlights that exergaming interventions are not homogeneous, and effectiveness is probably influenced by intervention design, level of supervision, participant health status, comparator type, and follow-up duration.

Evidence for secondary outcomes, including QoL, concerns about falling, balance confidence, adherence, acceptability, and cost-effectiveness, was limited, of low certainty, or inconsistently reported. Adverse events were uncommon and were not clearly attributed to the interventions, but the small number of studies reporting implementation outcomes limits broader conclusions regarding long-term feasibility, sustainability, and economic value. Overall, these findings suggest that exergaming may offer a promising additional approach within fall prevention; however, the current evidence base is not yet robust enough to support widespread routine adoption.

Previous systematic reviews have generally reported beneficial but mixed effects of exergaming on intermediate outcomes, such as balance, mobility, physical function, concerns about falling, and cognitive performance, but have offered limited synthesis of clinically meaningful fall outcomes [64-74]. This review extends this literature by focusing specifically on falls and injurious falls, by examining implementation-related outcomes relevant to real-world delivery, and by considering effects according to comparator type. This provides a more clinically relevant understanding of where exergaming may offer a benefit within fall prevention pathways.

Our findings are broadly consistent with previous reviews in suggesting that exergames can be effective, but they also confirm that these effects are not uniform and may depend on intervention design, comparator type, and duration of follow-up. Previous reviews, including Chen et al [75], have reported short-term improvements were more commonly seen than long-term effects, and this was evident in our included studies. This highlights the importance of longer follow-up periods, in line with ProFaNE recommendations, to establish whether short-term gains translate into sustained reductions in falls over time.

Variation in game design, exercise challenge, alignment with the Systems Framework for Postural Control [76], progression, feedback, and professional supervision may also explain differences in effectiveness across studies. Standardizing key intervention components and clearer reporting of intervention dose and support needs would improve comparability across future studies and improve confidence in findings [45]. This is particularly relevant given that adherence was stronger in studies with structured supervision, consistent with previous work emphasizing declining motivation in unsupported programs [77].

Economic and implementation evidence is also limited. Only 2 included studies reported intervention costs and only one considered cost-effectiveness [56], reflecting the limited economic evaluation seen across the wider exergaming literature [78,79]. Similarly, acceptability and usability were also underreported, despite their importance for long-term use of digital interventions among older adults [16,77,80,81]. Together, these gaps indicate that future exergaming studies should move beyond efficacy measures alone and address whether exergames represent a viable and scalable fall prevention approach for real-world adoption [82,83].

This review has several strengths, including a comprehensive multi-database search, prospective protocol registration, comparator-stratified analyses, and use of established tools including RoB 2.0 and GRADE. However, several limitations should be considered.

Our search was limited to studies published in English, which may have introduced language bias. There was also considerable clinical and methodological heterogeneity across studies in participant populations, intervention design, comparators, outcome reporting, and follow-up duration, which limited the precision and generalizability of pooled estimates. Exergame interventions varied widely in the types of equipment used, the extent to which games were purpose designed and the degree to which they were based on known effective strength and balance exercises (eg, Otago [84]), the amount of supervision by physiotherapists or health care professionals, and the duration and frequency of delivery. Furthermore, in several studies, exergames were delivered as part of a wider package of support, making it difficult to isolate the individual contribution of the exergame.

Notably, the studies involving support did not report on the amount or types of support needed by participants. Future research should therefore collect and report this information to better understand its impact on outcomes and to help assess the cost of delivery of the intervention in the real world.

Participant health status also varied considerably: 4 studies included people living with Parkinson disease or mild cognitive impairment, and 3 studies specifically recruited people with a history of falls. Such comorbidities, known to affect strength, balance, and falls, may have impacted the study outcomes [85-88]. Although we planned to conduct subgroup analyses, including those based on health status and PROGRESS-Plus, as outlined in the PROSPERO protocol, the relatively small number of studies included in meta-analyses limited the feasibility of more extensive subgroup analyses and formal assessment of publication bias. Many of the studies also had small sample sizes, limiting their ability to detect intervention effects and increasing the risk of type II error. While pooled “overall” estimates are presented, these should be interpreted with caution, and subgroup analyses by comparator type are likely to provide more clinically meaningful insights.

In addition, pooled estimates were calculated using the standard DerSimonian-Laird random effects approach in RevMan, which, although widely used, may provide less conservative interval estimates than alternative methods such as Hartung-Knapp-Sidik-Jonkman when only a small number of studies are available. Follow-up periods ranged from 1 to 12 months but were often relatively short, with 22% of studies including a follow-up of less than 6 months, and 33% at 6 months. Given that the ProFaNE recommends a minimum follow-up of 12 months in fall prevention trials, because these interventions may produce delayed effects, requiring sustained adherence over time to demonstrate efficacy [12], the longer-term sustainability of exergaming effects remains uncertain.

The findings of this review suggest that exergames may play a role as a complementary component of fall prevention programs for older adults, particularly given the evidence that they may reduce fall rates and the number of people experiencing a fall when compared with usual care. However, the current evidence does not support exergames as a stand-alone replacement for established exercise-based fall prevention interventions.

While the included studies varied considerably in intervention design and dosage, several interventions associated with favorable outcomes incorporated structured, progressive exercise delivered multiple times per week over several weeks or months. This is broadly consistent with established fall prevention exercise principles, which emphasize sufficient frequency, intensity, and progression [12,89]. Given the heterogeneity of interventions and limited evidence on optimal dosage, specific recommendations for exergaming interventions cannot yet be made, and further research is needed to determine the most effective training parameters. However, given the variability in intervention design and delivery, standardized guidelines for exergame design and delivery could be developed to ensure greater consistency across interventions. Purpose-designed serious exergames with appropriate theoretical underpinnings should be developed to unlock the full potential of exergame-enhanced fall prevention interventions [90]. This would facilitate the integration of exergaming with established exercise-based fall prevention programs such as Otago [91] or FaME [92], potentially optimizing effectiveness.

Adherence remains a recognized challenge. Most studies reported good adherence rates when structured support, such as supervision by physiotherapists, was provided, suggesting policies promoting supported or remotely monitored delivery may improve engagement and effectiveness [77]. Ensuring equitable access will therefore require investment in user-friendly technology, digital literacy programs, and technical support to reduce barriers to sustained participation among older adults [93-95].

Finally, our review highlights the need for longer follow-up periods and more robust economic evaluations before exergames can be confidently integrated into routine service delivery. Given the preliminary evidence suggesting cost-effectiveness within National Institute for Health and Care Excellence thresholds, further evaluations are needed to confirm these findings. Policymakers should consider further evaluations to guide future resource allocation and determine whether exergames represent a financially sustainable addition to health care and social care fall prevention pathways.

This review suggests that exergaming interventions may help reduce falls among older adults, particularly in comparisons against active intervention control groups, while pooled analyses also indicate potential overall benefit across diverse intervention settings. However, these findings should be interpreted cautiously because effect estimates varied considerably between studies and were limited by substantial heterogeneity, modest sample sizes, and relatively few long-term follow-up assessments.

Evidence for injurious falls and implementation-relevant outcomes, such as adherence, acceptability, QoL, and cost-effectiveness, remains limited or inconsistently reported. As such, although exergaming appears to offer a potentially engaging and clinically relevant adjunct to existing fall prevention programs, the evidence base is not yet sufficiently robust to support widespread routine implementation. Larger, methodologically rigorous trials with standardized fall reporting and longer follow-up are needed to determine which exergaming formats, doses, and delivery models are most effective for sustained fall prevention in older adults.