The review article by Förster, Happe and colleagues on the AS-Tra project, published in THERAPY 2024-I (Förster et al., 2024), outlines the development and iterative optimisation of a self-administered assessment and training system designed for elderly people aged 70 and above.

This article situates the initiative within the now burgeoning field of digital health interventions for older adults at risk of sarcopenia and frailty. Between 2024 and 2026, several systematic reviews and meta-analyses have been published, reporting predominantly positive – albeit heterogeneous – effects of digital interventions on muscle function, balance and mobility (Dai et al., 2025; Chen et al., 2025; Makizako et al., 2025).

Central components of the AS-Tra system – the instrumented Timed Up and Go Test, digital dynamometry and exergaming via force measurement platforms – have each been independently validated in parallel studies (Böttinger et al., 2024; Polo-Ferrero et al., 2026; Leung et al., 2026).

The working group’s own follow-up publications confirm the developmental rationale outlined in THERAPY 2024-I: By the third iteration, usability reached a “good” level (Förster et al., 2026a), and a four-week pilot study demonstrated feasibility under real-world conditions (Förster et al., 2026b).

At the same time, it is becoming evident that two questions left unresolved in the original article – behaviour change based on the transtheoretical model, and effectiveness under real-world conditions – continue to be only partially supported by current international research findings (Fawole et al., 2025).

Förster et al. (2024) describe the AS-Tra project (Assistance System for Sustainable Improvement of Nutritional and Mobility Status of Older People Aged 70 Years and Older Under Consideration of the Transtheoretical Model of Health Behavior Change) as an assistive technology system comprising two central components: a tablet-based app and a measurement and training station (MuTS).

The MuTS comprised pulse measurement, handgrip strength measurement, an automated TUG chair (aTUG-Chair) and the THERA-Trainer senso as an interactive training device. The target group comprises independently living elderly people aged 70 and above with early signs of malnutrition or mobility deficits (Happe et al., 2024).

Requirements and preferences of the target group were systematically collected across five focus groups involving n = 21 participants (mean age 78.5 years) (Happe et al., 2024); THERAPY 2024-I reports an improvement in System Usability Scale (SUS) scores across the first two cycles of structured usability testing, increasing from 70.63 to 84.55 – progressing from “acceptable” to “good”.

The planned next steps comprised simulated independent use, a field pilot study and, finally, a two-arm RCT.

This roadmap, outlined in THERAPY 2024-I, can now be cross-referenced with the working group’s follow-up publications from 2025/26, together with concurrent reviews from the international field.

Meta-analytic evidence for digital health interventions in sarcopenia and frailty

The article in THERAPY 2024-I argues primarily at a conceptual level that assistive technology systems could enable early detection and prevention (Förster et al., 2024). This assumption can now be tested quantitatively.

To date, the most comprehensive meta-analysis synthesises 16 RCTs involving older adults with frailty or sarcopenia, identifying significant effects of digital interventions on muscle mass (WMD 1.01; 95% CI 0.08–1.94), gait speed (WMD 0.09), Timed Up and Go Test (WMD –0.52 s), 30-second Chair Stand Test (WMD 2.19), balance (SMD 0.61) and quality of life (Dai et al., 2025).

Interestingly, hand grip strength – the key outcome measure in AS-Tra – shows no significant improvement following digital interventions in this analysis; by contrast, a parallel meta-analysis of 11 RCTs involving 757 patients with a formal diagnosis of sarcopenia found a significant, albeit small, effect on hand grip strength (SMD 0.28) and a moderate effect on skeletal muscle mass (SMD 0.35) (Chen et al., 2025).

Both reviews rate the certainty of evidence as low to moderate according to GRADE and note substantial heterogeneity (Dai et al., 2025; Chen et al., 2025).

For healthier, community-dwelling older adults without a formal sarcopenia diagnosis, a third meta-analysis (13 RCTs, n = 742) shows that digital interventions improve hand grip strength, usual gait speed and sit-to-stand performance, but not appendicular muscle mass (Makizako et al., 2025).

A narrative review, together with a recent systematic review, consistently concludes that digital tools raise awareness of malnutrition and reduce screening burden for healthcare professionals, with co-designed, user-centred interfaces consistently enhancing acceptance (Amatrudo et al., 2026).

A review focusing specifically on app-based interventions among community-dwelling older adults identified, across six studies involving 818 participants, a significant effect on step count (+799 steps/day), whereas the evidence relating to nutritional outcomes was described as “relatively unresearched”, yielding mixed findings (French et al., 2024).

This evidence base supports the conceptual premise underlying the AS-Tra project, while simultaneously highlighting where expectations require recalibration: Effects on muscle mass and hand grip strength are, at best, small, and the nutritional impact of digital tools remains poorly substantiated (French et al., 2024). For a system explicitly designed to address both domains, this is a key consideration.

Exergaming and the THERA-Trainer senso in context

The original article selects the THERA-Trainer senso as the training component on the basis of focus-group preferences (Förster et al., 2024; Happe et al., 2024). The evidence base for exergame-based training in older adults has since grown stronger and supports this decision, albeit with certain limitations.

A meta-analysis of 14 RCTs (n = 724) involving commercial exergames in residential care settings demonstrates consistent improvements in balance: a reduction in TUG time (SMD –0.87 s) and a mean increase of 2.88 points on the Berg Balance Scale, with higher training doses (≥ 9 weeks) associated with greater treatment effects (Leung et al., 2026).

In order to classify AS-Tra, it is essential to recognise exergaming as an effective form of balance and mobility training. Compared with conventional balance training, additional effects are not evident in all analyses; however, findings vary according to the technology used, training design and dosage.

A meta-analysis by Cieślik et al. (2024), comprising 28 studies and 1,457 participants, demonstrates moderate to larger effects specifically for motion-capture systems (SMD 0.48–0.86).

For force-plate-based systems – among which the THERA-Trainer senso, with its five measuring plates, is functionally classed – the evidence for improvements in functional mobility remains, to date, less pronounced, though fundamentally positive. Clinically relevant effects appear to be achievable primarily at a training duration exceeding 134 minutes per week.

Sturnieks et al. (2024) provide a particularly robust evidence base for efficacy in a randomised controlled trial involving 769 participants. A home-based step-exergaming programme significantly reduced fall rates over a twelve-month period (IRR 0.74; 95% CI 0.56–0.98), whereas passive cognitive training demonstrated no comparable effect.

For an assistive system such as AS-Tra, this is an important external argument: Exergame-based training can make a relevant contribution to fall prevention, provided it is delivered at an adequate dosage, undertaken regularly, and designed to be cognitively and motorically challenging.
GET IN TOUCH

The transtheoretical model in a digital context

The conceptual grounding of AS-Tra within the transtheoretical model (TTM) points to a field that remains, as yet, poorly substantiated by evidence (Förster et al., 2024).

A systematic review published in 2025, examining TTM-based interventions to increase physical activity in older adults, identified only three studies (two RCTs, one non-RCT, n = 1,474). The review found low certainty of evidence, with no significant effects on physical activity or self-efficacy reported in the RCTs; where effects did occur, these were most likely to emerge during the maintenance phase (Fawole et al., 2025).

A 2024 RCT employing nurse-led motivational interviewing based on the TTM (n = 117, 24 weeks), by contrast, reports marked transitions from the contemplation to the action phase (70.4% vs. 9.3% in the control group) and higher step counts (Sönmez Sarı & Kitiş, 2024).

The discrepancy shows that TTM-based interventions are effective when combined with personal interaction; in purely digital, self-directed usage scenarios without such interaction, efficacy remains, as yet, unclear.

The follow-up publications of the AS-Tra working group

The follow-up investigations of the AS-Tra system, announced in THERAPY 2024-I, are now available as two publications.

Iterative development for a use that is as independent as possible

Förster et al. (2026a) describe the complete three-stage usability process involving a total of 34 participants with a mean age of 78.60 ± 5.59 years; 70.93% were female. The System Usability Scale (SUS) score rose from 70.63 in the first cycle to 84.55 points in the second cycle, before reaching 78.18 points in the third cycle.

In the third cycle, more independent use was simulated: Participants remained alone in the laboratory without study personnel present, received audio instructions, and were monitored via webcams and eye-tracking glasses.

Of a total of 607 tasks, 63.11% were completed without assistance. A further 14.92% were achieved following repetition or rephrasing of the instruction, and 20.33% following a direct solution cue. Only 1.15% remained unsolved.

The SUS score, although lower than in the second cycle, remains within the “good” rating score, demonstrating that outcomes obtained under close supervision cannot be fully extrapolated to unsupervised, independent use. The authors therefore stress the importance of suitable support functions and testing conditions that closely reflect real-world conditions.

Four-week feasibility pilot study

Förster et al. (2026b) subsequently investigated the AS-Tra system over a four-week period. Nine older adults took part, with one participant withdrawing from the study prematurely. The mean age was 80 ± 5 years, with women accounting for 50%.

Grip strength, Timed Up and Go, and 5-Time Chair Rise Test were assessed at baseline, and after one, two and four weeks. In addition, participants used the nutrition app, wore an activity sensor for seven days, and completed regular training sessions at the measurement and training station.

Usability was rated as good, with an average SUS score of 79.3 ± 13.4 points. The overall system also received a predominantly positive assessment, with a mean Evaluation Overall System (EOS) score of 2.01 ± 0.99 on a scale of 1 to 5.

Technical issues affecting individual components of the measurement and training station occurred in no more than 17% of sessions in each case.

The tablet-based nutrition diary was associated with the most pronounced difficulties. Software crashes occurred in 17 of the 30 planned data entries; overall, problems or errors were recorded in 22 of the 28 sessions in which the software was actually used.

By contrast, questionnaires, activity sensors and automated functional assessments generally proved to be practicable.

The pilot study therefore provides initial confirmation of the feasibility of data collection and the good acceptance of the overall system, but does not confirm its health-related efficacy. Before the planned randomised controlled trial can be conducted, the nutrition diary in particular must be further developed, both technically and in terms of user experience.

This observation is consistent with the findings of the narrative review by Amatrudo et al. (2026). Digital tools can facilitate the identification of malnutrition risk and support nutritional management. However, consistent involvement of the target group, co-design and age-appropriate user interfaces are essential to ensure acceptance and reliable use among older adults.

Conceptual strengths. Few of the digital health interventions for older adults currently published in the literature combine inpatient-based, standardised assessments (grip strength, iTUG, chair rise test) with app components designed for home use. The “measurement station as an anchor point – app for everyday use” approach is rarely described in the literature, distinguishing AS-Tra from purely wearable and app-based interventions. The focus group work also confirmed requirements specific to the target group, revealing strong interest among older adults in self-monitoring, notably heart rate and grip strength (Happe et al., 2024).

Empirically robust core components. Between 2024 and 2026, the assessment methods selected in the original article underwent independent validation: digital dynamometers demonstrated excellent agreement with the Jamar reference standard (Polo-Ferrero et al., 2026), smartphone-based iTUG applications proved to be reliable self-assessment tools (Böttinger et al., 2024), and exergame-based training using force plates showed consistent balance effects across meta-analytic evidence (Leung et al., 2026).

Unresolved questions on efficacy. At the meta-analytic level, the effects of digital interventions on functional outcome measures (TUG, chair-stand test, gait speed, balance) are now consistently documented, whereas effects on muscle mass and grip strength remain small or inconsistent (Dai et al., 2025; Chen et al., 2025). For nutritional outcomes – the second core pillar of AS-Tra – the evidence base remains heterogeneous (French et al., 2024). The planned two-arm RCT by the working group would, for the first time, address the evidence of efficacy in both domains under conditions of independent use by individuals aged 70 and above with emerging deficits.

Methodological challenge. The decline in the SUS score between cycle 2 and the sole-use scenario in cycle 3 illustrates a challenge that other reviews also raise: Usability, when measured in structured test settings, systematically overestimates everyday usability (Förster et al., 2026a). Ambient assisted living studies emphasise that the effects of digital systems often only stabilise over a period of several months and are heavily dependent on the home environment (Chaudhary et al., 2026).

Behaviour change theory. While the TTM is theoretically well-founded, empirical evidence supporting it remains limited. Fawole et al. (2025) found low certainty of evidence and no significant effects in RCTs; TTM approaches were effective primarily when combined with personal interaction. For a system explicitly designed to promote independent use, this gives rise to a specific research contribution: AS-Tra can demonstrate whether digital representations of TTM phases can offset the efficacy disadvantage of purely digital TTM applications.

The AS-Tra system described in THERAPY 2024-I remains conceptually robust even after three iterative development cycles. The roadmap published by Förster et al. (2024) has since been substantiated by the working group’s 2025/26 follow-up reports: usability reaches a good level even under conditions of sole use (Förster et al., 2026a), and the feasibility of combined data collection has been demonstrated (Foerster et al., 2026b).

Over the same period, the international body of research has grown substantially more robust: meta-analyses consistently show moderate effects of digital interventions on functional mobility (Dai et al., 2025), smaller and heterogeneous effects on muscle parameters (Chen et al., 2025; Dai et al., 2025), and continued limited efficacy in the nutritional field (French et al., 2024).

The specific scientific contribution of AS-Tra – the combination of stationary, ambient-instrumented assessments with everyday app use in individuals aged 70 and above with emerging deficits – remains underrepresented in the international field and confirms continuation of the programme.

AS-Tra is pursuing a conceptually convincing and equally engaging approach: standardised functional assessments, interactive training, activity tracking and digital support for nutrition and exercise are combined within a single system. The combination of a central measurement and training station with everyday app use sets the project apart from purely wearable- or app-based solutions.

Another positive aspect is the early involvement of older people in the development process. The publications to date demonstrate good usability and fundamental technical feasibility. The principles employed – digital force measurement, automated mobility testing and exergame-based training – are also fully consistent with current scientific thinking.

Its greatest strength is also its main challenge: AS-Tra combines mobility, nutrition and behaviour change within a complex intervention. The health benefits will therefore depend substantially on whether these components function reliably, are utilised to a sufficient degree, and can be sustainably integrated into the daily routines of older adults.

The problems encountered with the nutrition diary and the greater need for support during independent use demonstrate that technical stability and low-threshold guidance remain crucial in this regard.

Further clinically focused research will need to determine whether the good levels of acceptance and technical feasibility observed also translate into measurable improvements in mobility, nutritional status and activities of daily living. In contrast to current findings from aggregated evidence, the critical evaluation of suitable primary endpoints could be decisive for the quality and validity of future outcomes.